WILTSHIRE

(1) Amesbury, 19 Boscombe Road (SU 16724 40996): a ditch and refuse pit containing second- to fourth-century Roman pottery were recorded during an evaluation on land close to the substantial Romano-British settlement at Butterfield Down.Footnote ²⁴¹

(2) Swindon, land at Lydiard Tregoze (SU 10355 85010): investigations revealed a series of early Roman intercutting pits and ditches. The pits, cut into an alluvial deposit, are provisionally interpreted as clay extraction pits associated with nearby pottery kilns. The alluvium sealed a few early Roman features.Footnote ²⁴²

(3) Westbury, 30 Wellhead Lane (ST 87241 50270): an evaluation was undertaken on the site where previous excavations had revealed the remains of a Romano-British settlement. The remnants of a wall foundation, a possible yard surface and deposits related to burning and collapse were found. The proximity of the wall to a Roman lime kiln suggests that the wall is also of this date.Footnote ²⁴³

(4) Wilsford, River Avon restoration works (SU 13348 39148): a watching-brief during restoration works on the river Avon revealed a small assemblage of unabraded Roman pottery, suggesting settlement activity in the vicinity.Footnote ²⁴⁴

DORSET

(1) Wimborne Minster, land south of Leigh Road (SZ 0276 9964): evaluation revealed a farmstead which included ditched enclosures, a possible crop-drying oven and an associated field-system. Finds date to the third century. A Roman road conjectured to have passed through the site was not present but may have been entirely truncated by ploughing.Footnote ²⁴⁵

BERKSHIRE

(1) Horton, Kingsmead Quarry (Area 15 East) (TQ 01947 75327): an evaluation revealed the remains of a probable bridge of later prehistoric or Romano-British date, which would have crossed the narrowest part of one of the several water channels on the site.Footnote ²⁴⁶

HAMPSHIRE

(1) Basingstoke, Manydown, land at Battledown Farm Scheduled Monument and Kite Hill (SU 59590 50295 and 59950 51280): a detailed gradiometer survey was undertaken at the two sites. Both were shown to have multi-phase archaeological remains, probably Iron Age to Roman, including sequences of enclosures and probable trackways. The survey did not identify any clear evidence for the presence of a purported Roman road.Footnote ²⁴⁷

(2) Bramley, land at Minchens Lane (SU 653 597): an evaluation recorded a Romano-British pit.Footnote ²⁴⁸

(3) Chineham, Razor's Farm (SU 465538 56175): an evaluation on land proposed for residential development revealed the northern edge of a late Iron Age–early Roman settlement, part of which had been recorded during excavations on an adjacent plot of land.Footnote ²⁴⁹

(4) Corhampton and Meonstoke, Shavards Farm (SU 617 210): earth resistance survey of an area of 0.9 ha further defined the remains of a villa discovered during road widening in the 1930s. At least four separate Roman buildings are present including the previously partially excavated aisled hallFootnote ²⁵⁰ east of the A32 road that bisects the site, with remains of three further buildings, including a structure of octagonal plan, located west of the road in the meadows bordering the river Meon.Footnote ²⁵¹

(5) Godshill, Cockley Bushes, Leadenhall (SU 198 155): a circular platform feature terraced into the valley side, located in field survey, was excavated and shown to be linked to charcoal burning. Finds consisted only of a small number of struck flakes and a quantity of fire-affected flint, but charcoal yielded radiocarbon dates of 2046 ± 35 BP and 1939 ± 35 BP (SUERC-65053 and 65054), placing the charcoal-burning activity in the late Iron Age to early Roman period. A second platform site was surveyed, c. 100 m to the east, and a third was also identified in the Cockley Bushes area, but not surveyed. These platform sites are a new monument type for the New Forest, and suggest that woodland management, charcoal preparation and processing was the primary activity to be associated with them.Footnote ²⁵²

(6) Havant: (a) Warblington (SU 734 059): a Ground Penetrating Radar (GPR) survey was conducted on the villa siteFootnote ²⁵³ over an area of 2.4 ha and identified building remains that correspond well with the results of earlier earth resistance survey of the site.Footnote ²⁵⁴ The GPR survey complements the existing surveys, slightly extending both the area of coverage and providing some useful information regarding the depth of the surviving structural remains, particularly where building rubble appears to obscure the earth resistance results. This suggests the main villa building developed through a series of phases that may, in part, be identified through the realignment of the walls. Some evidence for differing construction methods was also found through the variable depth of the wall footings.Footnote ²⁵⁵

(b) 59–61 West Street (SU 714 062): three areas were excavated, of which two produced evidence of Roman activity. In the south-eastern area a number of post-holes and pits were dated to the first and second centuries a.d. Undated post-holes appeared to form part of a rectangular structure along with the Roman features. In the north-eastern area an early Roman ditch extended south-east to north-west across the site.Footnote ²⁵⁶

(7) Silchester (Calleva Atrebatum) (SU 64 42): excavation in Insula III continued for a third season. An area of 20 m by 15 m was opened in the north-east corner of the insula in order to re-excavate the trenches of the Society of Antiquaries’ excavation of 1891 and evaluate their findings. No trace was found of the postulated walls (‘Traces of House’) or of any structural connection with the foundations found in the south-east corner and re-excavated in 2013–14.Footnote ²⁵⁷ Instead it was possible to trace from the profiles of the 1891 trenches an occupation sequence from the late Iron Age, with evidence of successive Roman timber buildings and associated occupation at the intersection of the north–south and east–west streets, through to the fifth century. A first-century burnt horizon and a substantial, late first/early second-century deposit of iron-smithing waste were identified, while the ‘hypocaust’ on the plan of 1891 proved to be a keyhole-shaped, brick-built oven also of late first/early second-century date. At least two phases of a late third- and fourth-century timber-framed building with dwarf flint walls fronting onto the east–west street with a gravelled yard to the rear were also identified.Footnote ²⁵⁸

(8) South Wonston, Project Wellesley, Worthy Down, service families accommodation (SU 466 350): evaluation targeted features previously identified as cropmarks and by geophysical survey. A single boundary ditch contained pottery dated c. a.d. 240–400. Other features were undated.Footnote ²⁵⁹

(9) Southwick, land at Pigeon House Lane (SU 64670 08682): a watching-brief during groundworks associated with installation of a solar panel array revealed a layer of gravel, considered potentially to be the remnants of the Wickham to Chichester Roman road.Footnote ²⁶⁰

(10) Winchester: (a) Barton Farm (SU 4757 3162): an excavation recorded a series of multi-period funerary enclosures to the north of the site, dating from the Bronze Age to the Roman period. These included a first-century a.d. ditched enclosure, with three inhumations cutting into the base of the ditch, and a third- to fourth-century cemetery enclosure with multiple phases of burials and cremations in the enclosure ditch. To the west was a separate, unconnected second- to third-century enclosure, containing a number of both unurned and urned cremations. Excavation of an area in the south-east of the site recorded a section of a first- to second-century aqueduct ( fig. 29), approximately on the course previously proposed,Footnote ²⁶¹ potentially supplying Roman Winchester from a source near Itchenstoke.Footnote ²⁶²

FIG. 29. Barton Farm, Winchester. Aqueduct channel looking north. (© PCA)

(b) Colebrook Street (SU 484 292): a watching-brief recorded the southern side of a third- to fourth-century east–west-aligned road metalling. Seen in section at the junction of Colebrook Street and the Broadway, it is interpreted as part of the main Roman street through Winchester.Footnote ²⁶³

(c) North Walls Fire Station (SU 4844 2967): an excavation recorded the remains of a third- to fourth-century timber-piled structure, overlain by alluvial deposits.Footnote ²⁶⁴

(d) Provost Cells (SU 4782 2926): an excavation recorded two phases of Roman rampart, part of the city defences, dating to the first and third century. A second- to third-century pit was recorded cutting into the earlier rampart.Footnote ²⁶⁵

(e) Victoria House, Victoria Road (SU 4795 2997): an excavation within Winchester's northern Roman cemetery recorded eleven inhumation burials dated to the fourth century. These included one infant, one neonatal and one prone burial. Other parts of the site had been previously investigated in the 1970s.Footnote ²⁶⁶

WEST SUSSEX

(1) Alfoldean (TG 117 328): investigations by a variety of individuals and organisations over the last decade have considerably improved an earlier interpretation of the development of this roadside settlement.Footnote ²⁶⁷ The enclosure, defined by a bank and two ditches adjacent to the river Arun, is typically cited as an example of a mansio. While rooms which could have been associated with a mansio were identified in Winbolt's excavations in the 1920s, the full layout of the main building was only clarified by the Time Team 2005 investigations, thus proving its existence. Their geophysical survey to the south also mapped the roadside enclosures and trackways on the east side of Stane Street. More recent geophysical surveys have provided further information on the ditched boundaries to the west of Stane Street (although not those adjacent to it which were recorded in a trench dug in 1983). A field-system that does not respect the alignment of Stane Street has now been identified. It is probably pre-Roman in origin but is likely to have survived into the Roman period. Perpendicular to the field-system and on its northern edge is a straight double-ditched feature. Although this may be a Roman road, especially as it is aligned on Stane Street's crossing-point over the river Arun, it appears to stop, or at least change alignment, where it meets a rectangular enclosure to the west. A second double-ditch arrangement may also represent a road that appears to bypass the mansio enclosure ( fig. 30).

FIG. 30. Alfoldean, overall plan of roadside settlement.

None of these major new boundaries has been excavated so their dating is unproven. However, it seems likely that Stane Street was constructed through a pre-existing field-system (possibly of middle Bronze Age origins). The enclosures associated with the roadside settlement were arranged perpendicular to Stane Street while the pre-existing fields were retained. The previous suggestion that the mansio buildings pre-dated the construction of the main enclosure is still likely to hold true. Once the enclosure had been constructed the adjacent fields and trackways were amended to respect it, resulting in the creation of the bypass route to the west.Footnote ²⁶⁸

(2) Hurstpierpoint, land off Chalkers Lane (TQ 28286 17599): an evaluation targeted features identified during an earlier geophysical survey. These included part of an outer enclosure ditch and two sections of a ring-gully, perhaps the drip-gully of a roundhouse, with a shallow pit located between them. Three more ditches roughly running north–south were exposed within the enclosure. The date range is late prehistoric to early Roman.Footnote ²⁶⁹

(3) Strettington (SU 89751 07076): evaluation identified a number of ditches representing two enclosures. Pottery from these features spans the Roman period.Footnote ²⁷⁰

EAST SUSSEX

(1) Bexhill, north-east Bexhill Gateway Road and East Sussex Business Centre (TQ 7840 0920): evaluation showed that activity during the Roman period was concentrated on the summit of a ridge in the central area of the site. Small enclosure ditches and the presence of a single cremation are suggestive of short-term settlement. This appears to be typical of iron-working sites in this part of the Weald, which tend to favour high locations, where the basal Wadhurst Clay, rich in sideritic mudstone, overlies the Ashdown Beds.Footnote ²⁷¹

(2) Plumpton, Plumpton College (TQ 360 147): in 2013/14 a programme of geophysical surveys (both resistivity and magnetometry) and limited trial-trenching was designed to more fully assess the extent, nature and condition of the villa complex first discovered during the 1970s.Footnote ²⁷² In 2015 excavation further revealed the eastern half of the villa. The exposed remains comprised a south-facing corridor which fronted part of a line of (probably) three main rooms separated by sub-divided passages, and terminated, at the eastern end at least, in projecting wing-rooms. The front (southern) wall of the eastern wing-room is internally apsidal, whilst its outer face is straight. Much of the large room to the north of the investigated wing-room contains the remains of what is thought to be an inserted rectangular corn-drying oven. Just to the east of the villa a concentration of flint rubble may have been the floor of a timber-framed building. Dating evidence from the site spans the third and early fourth centuries. Evidence for the nature of some of the flooring, now all damaged by ploughing, includes finds of both tesserae and small mosaic cubes.Footnote ²⁷³

(3) Ringmer, Bridge Farm, Wellingham (centred at TQ 4320 1440): a third season of excavation was undertaken at this Roman enclosed roadside settlement first discovered in 2011, examining a 40 m square area over the north-east corner of the double-ditched enclosure at the point of its intersection with the road heading to the north (Margary Road 14, London–Lewes). Three main phases of activity were indicated, dated by coins. The enclosure ditches were dug in the late second century, a dark layer formed above the road in the late third century and the mid-fourth century was represented by a layer affected by pedogenesis (including the effects of recurrent flooding by the river Ouse). Two Hadrianic denarii and a single siliqua of Honorius, found by metal-detecting, indicated some earlier and later activity. Other finds included c. 10,000 sherds of pottery, a red jasper intaglio, a ring-key and a stylus. Over 200 kg of iron-working residue was collected, some perhaps used in a surface of the London road.Footnote ²⁷⁴

SURREY

(1) Epsom, former Nescot College Animal Husbandry Centre, Reigate Road (TQ 2227 6207): excavation was carried out in three areas of the site following an evaluation undertaken in 2014. Early Roman features in the north-east of the site included three enormous quarry pits along with smaller pits and isolated post-holes. The quarries related to the extraction of chalk and flint which appears to have been undertaken on an industrial scale. This material may have been used locally, yet the quantities of what must have been removed suggest that this resource was being exported out of the immediate area, possibly using the Hogsmill River as a conduit. The quarries were closed by the second century, although one in particular, which could be determined as a ‘shaft’, may have been backfilled by the end of the first century. All the quarries were closed with material associated with the practice of selective deposition, and the shaft appears to have been the focus of such activity. Human remains in disarticulated form were recovered from the bottom of this feature and have at present been estimated as representing a minimum of 53 individuals. These remains were accompanied by animal bone groups comprising 67 partial dogs, seven partial pigs and four partial foetal horses along with pottery vessels, coins, a gaming-piece, a brooch and a spindle-whorl. A near-complete human skeleton lying in the prone position was also revealed in the shaft and, along with the disarticulated remains, suggests that excarnation was being practised in this area at the time. A bone-handled knife was recovered above the articulated body. The remaining features of this phase included two north-west–south-east-aligned ditches representing an enclosure or possible field boundary in the north-eastern corner of the site, along with the final infilling of a sinkhole which cut through a Bronze Age droveway in the south-west of the site. Activity appears to have resumed during the third century. Two of the large quarries had final fills dating to this period which again suggested that selective deposition was practised, including small finds, animal carcasses and in one instance a neonate human. In the north-east corner of the site quarrying started again and three small quarries or ‘mini shafts’ were backfilled with material suggesting that they were excavated during the latter half of the third century. Selective deposition appears to have continued and finds within the quarries included dog, badger, a human infant and a pick made from red deer antler tine. The lower legs of a human inhumation within a disturbed grave were also associated with this phase of activity.Footnote ²⁷⁵

KENT

(1) Canterbury (Durovernum Cantiacorum): (a) former Peugeot Garage, Rhodaus Town (TR 1492 5736): excavationFootnote ²⁷⁶ in advance of development for student accommodation identified evidence for early Roman quarrying, land boundaries and a late Roman cemetery. The site, situated adjacent to previous investigations at 24a Old Dover Road,Footnote ²⁷⁷ Canterbury Police Station,Footnote ²⁷⁸ Augustine HouseFootnote ²⁷⁹ and Rhodaus Town,Footnote ²⁸⁰ forms part of an emerging extramural landscape fronting the Roman road from Dover and immediately outside the late third-century town wall ( fig. 31). Quarrying for clay and underlying river gravels dating from the late first century was evidenced on a large scale, with comparable activity identified on adjacent excavations encapsulating a combined area of at least 0.9 ha. The activity was contemporary with the initial formation of Roman Canterbury and presumably provided raw materials for laying of streets and major public buildings. A series of parallel boundary ditches extended south from the area of quarrying towards an east–west-aligned hollow way. The hollow way, established in the late Iron Age, continued in use through the Roman period and beyond. There was little evidence for land use. A single cremated burial interred within a vessel was dated to the late second century. From the fourth century a formal cemetery extended across the land between the quarry and the hollow way, with a new east–west ditch separating the cemetery from the still extant quarry. Previously two inhumation burials had been identified within the former garage area.Footnote ²⁸¹ During the present excavation, a further 219 inhumation burials were identified, aligned east–west and in ordered rows. The cemetery included the remains of infants as well as adults, the majority placed in timber coffins. With the exception of a small number of personal objects, notably hairpins and a comb, there were in general no grave goods. One distinctive burial, orientated west–east, contained two silver belt fittings, preliminary assessment of which indicates a likely early fifth-century date and a central or eastern European origin.

FIG. 31. Canterbury, extramural features south-east of the town. (© Canterbury Archaeological Trust)

(b) 49 St Peter's Street (TR 1470 5798): excavationFootnote ²⁸² in a cellar prior to installation of pumps as part of a flood-relief scheme uncovered a deliberate dump of pottery, brick, tile and other debris, deposited c. a.d. 60–80 to create an artificial platform. On this platform were the remnants of a building, comprising an internal clay floor and occupation deposits. No wall remains survived, so it was not possible to identify whether the structure was of timber or masonry construction. During the first century a.d. the ground surface was raised again and flint metalling laid over the newly raised ground. This metalling perhaps formed a yard within or adjacent to an associated building. Occupation material built up across the metalling before a change of use was suggested by an overlying sequence of clay floor and occupation deposits, almost certainly lying within the room of a building. The latest activity perhaps spanned the early to mid-second century to at least the mid-third century. A cesspit of third- to fourth-century date contained a waterlogged basal fill from which bioarchaeological assemblages were recovered.

(c) Westgate Gardens (TR 1436 5785, centred): further excavationFootnote ²⁸³ outside and to the north of the site of London Gate uncovered another early roadside ditch and earlier surfaces of the Roman road. Areas of metalled hard standing next to the road appear to indicate early Romano-British ribbon development pre-dating the construction of the wall that redefined the town c. a.d. 270–90. (d) 38b St Dunstan's Terrace (TR 1424 5825): a watching-briefFootnote ²⁸⁴ maintained during ground reduction prior to house construction revealed metallings probably associated with those recorded in 2012Footnote ²⁸⁵ and also the junction of two minor streets.

(2) Dover, South Foreland Lighthouse (TR 3593 4330): a community excavationFootnote ²⁸⁶ on ground immediately south-east of the lighthouse uncovered a post-built structure set within a shallow terrace in the chalk. Late second- to third-century pottery and coins minted between a.d. 293 and 297Footnote ²⁸⁷ (including one of Allectus) were recovered. The structure might perhaps have formed a watch-post on this strategic coastal vantage point.

(3) East Farleigh, Oaklands, Lower Road (TQ 722 622): excavation concentrated on large Roman drainage ditches at the east end of the buildings.Footnote ²⁸⁸ A trench at the west end of Building 5 concentrated on lifting what is hoped to be decorated fallen external wall-plaster. An Iron Age silver coin dated was found in hill-wash overlying the drainage ditch.Footnote ²⁸⁹

(4) Eccles, Burham to Hythe and Ditton Mains Water Scheme (TQ 71997 60607): a watching-brief immediately west of the site of the Eccles Roman villa recorded a number of Romano-British features, including ditches, gullies and possible remnants of foundations. The latter may relate to outlying structures of the villa complex.Footnote ²⁹⁰

(5) Folkestone, East Wear Bay (TR 2407 3700): ongoing excavationFootnote ²⁹¹ 50 m north of the villaFootnote ²⁹² uncovered a quern manufacturing site active during the late Iron Age and early Roman period. Over 60 quern fragments in various stages of manufacture were recovered, associated with a metalled surface covered by greensand dust and debitage. A late Iron Age or early Roman roundhouse was also identified, truncated by ditches which were backfilled during the first to second century.Footnote ²⁹³

(6) Herne Bay, Herne Bay Golf Course (TR 1716 6664): evaluationFootnote ²⁹⁴ prior to redevelopment revealed late Iron Age/early Roman activity. Sample excavated features contained pottery dating up to the second century a.d. Together the evidence suggests that the area forms part of a Romano-British landscape of enclosures and settlement.

(7) Manston, World's Wonder, Manston Road (TR 3435 6607): evaluation trenchingFootnote ²⁹⁵ prior to construction of a solar farm located two early Roman cremation burials. One comprised a vessel placed within an amphora; the second comprised a group of three vessels, one of which dated to a.d. 70–130. A third potential cremation burial was indicated. Ditches/linear features, a possible hollow way and a pit were also located.

(8) Peter's Village and Medway Crossing (TQ 7220 6220): a watching-brief revealed the continuation of a shallow north-east–south-west-oriented ditch of late Iron Age or Roman date previously recorded in 2014.Footnote ²⁹⁶

(9) Sellinge, land at Ashford Road (TR 1037 3801): a 15-trench evaluation recorded a number of large ditches and two pits. Most ditches had multiple recuts, suggesting continued use and boundary re-establishment throughout the Roman period.Footnote ²⁹⁷

(10) Sittingbourne, Bredgar C of E Primary School (TQ 8806 6027): evaluationFootnote ²⁹⁸ in advance of the construction of new school buildings uncovered a wall foundation, ditch and deposits of demolition material indicating that the structure(s) identified on the western side of the school during excavations in 2004Footnote ²⁹⁹ almost certainly extended eastwards.

(11) Sturry, Woodlands Farm, Calcott Hill (TR 18113 63258): a watching-brief recorded a north-west–south-east-aligned early Roman ditch in the north of the site.Footnote ³⁰⁰

(12) Swanscombe, Ebbsfleet Green (formerly Northfleet West Sub-station), Southfleet Road, (TQ 0920 3030): of 55 evaluation trenches excavated, one located on high ground in the north of the site contained elements of three ditches. Two of these contained small quantities of first- to second- and first- to third-century a.d. pottery respectively, while a shallow gully contained pottery of first- to second-century date.Footnote ³⁰¹

(13) Upper Harbledown, Homestall Wood (TR 1175 5892 centred): following initial survey,Footnote ³⁰² excavation sampled the north-west side of the main polygonal enclosure and the eastern side of the inner rectangular enclosure. At the former point (Site A), a section of the bank showed it to be 8 m wide, surviving to a height of 1 m and composed of layers of brown clay and gravel laid in conformity with the ground slope. A primary bank appeared to be derived from the underlying grey clay subsoil, the surface of which bore traces of shallow excavation, sealed by the main bank. Evidence for a front revetment was lacking but the ditch edge was marked by placed flints and possible stake-holes. No datable finds were recovered. The high water table prevented investigation of the single external ditch or the rear of the bank. On the site of the inner enclosure (Site B), excavation identified a low bank external to a waterlogged ditch, immediately north of an apparent causewayed entrance. A well-laid gravel surface, sterile of finds, underlay the bank as did a series of post-holes close to the outer edge of the ditch. The external edge of the bank was cut by a series of possible wheel-ruts from traffic following the outer perimeter. The bank produced a Romanised cream flagon base, native grog-tempered pottery and local imitations of Gallic wares of the late first century b.c. or early first century a.d. A simple leaf-shaped iron arrow or bolt-head from a high level may be post-Roman. At three points within the interior of the larger enclosure recent ground disturbances have yielded more pottery of the late Iron Age or early Roman period including grog-tempered ware and terra rubra. Amphora sherds recovered from a tree hole close to Site A appeared to be of Campanian black-sand fabric, the body form consistent with Dressel 1 vessels of the first century b.c.

On the adjacent hill to the west, within Willows Wood, a brief walk-over survey identified a single bank and external ditch defining a roughly oval area of 8 ha, centred at TR 1091 5889. A linear bank and ditch midway between this and the main Homestall enclosure's western side was traced for 600 m northwards from the stream crossing the line of Watling Street west of Harbledown Lodge.Footnote ³⁰³

(14) Whitfield, Menzies Road (TR 3018 4429, centred): new building work on part of the former Old Park estate on a chalkland spur, some 3.3 km north-west of Dover (Portus Dubris), revealed over 30 clay-cut featuresFootnote ³⁰⁴ including 23 pits of varying shapes and sizes, six post-holes and three ditches (one recut). Many of the features had sterile fills but about one third produced Roman pottery provisionally dated to the late first and second century a.d. Other evidence for Roman habitation in the area has previously been recorded.Footnote ³⁰⁵