CAERPHILLY

(1) Gelligaer, Gelligaer cemetery (ST 1348 9733): a field evaluation noted the presence of a Roman road incorporating a fragment of Roman glass and two sherds of Roman pottery in its road surface. The road, which was aligned north-east–south-west, is likely to be that which left the north-east gate of the late first-century fort. Part of a wall or roadside structure was also recorded, which is also likely to date to the Roman period. Roman pottery and the base of a Roman pillar were recovered from two other trenches, indicating potential Roman activity at some distance away from the road and its associated structures.Footnote ¹

FIG. 1.

Key to numbered sections of Part I (Sites Explored).

CARMARTHENSHIRE

(1) Llandovery, Llanfair Ar Y Bryn (SN 77026 35221): significant archaeological remains were recorded in an exposed section during groundworks.Footnote ² Below post-medieval and modern contexts a series of features was exposed which was most likely of Roman date and associated with the forts. Based on the typical layout of auxiliary forts, the site area would have been located close to the centre of the early, late first-century fort. In the later, reduced sized fort, the site would have been located on top of the via praetoria and it is possible that a layer of pitched cobbles in the section may have been part of this roadway. Below the cobbles a series of layers and cut features suggested post-holes and beam-slots, indicating the remains of possible timber buildings associated with the first fort. The work confirms the presence of further remains associated with the forts in the immediate surrounding area.Footnote ³

GWYNEDD

(1) Caernarfon, Ysgol Yr Hendre, Llanbeblig (SH 4897 6236): scattered across the excavation area were 18 features, roughly figure-of-eight-shaped in plan. These consisted of two adjoining pits — one with evidence of in situ burning; the other, while containing charcoal, generally lacked traces of burning. The evidence suggested that the fire chamber was an enclosed oven, roofed with turves, possibly resting on branches, and the other pit was for raking out. Radiocarbon dates showed the ovens to be late first-century a.d. and suggested that they were concentrated around the probable date of the construction of the fort of Segontium, which lies 300 m to the east, in about a.d. 77. The similarity of the ovens to those found in the Roman marching camp at Kintore, Aberdeenshire, suggests that they can be interpreted as temporary Roman military ovens.Footnote ⁴ Although a large area was investigated in the project, no ditches identifiable as belonging to a Roman camp were identified; however, it has been argued that not all temporary camps had ditches.Footnote ⁵ It is therefore proposed that the ovens may indicate the presence of a Roman camp, not otherwise detectable, which might have been related to the construction of the fort.Footnote ⁶

(2) Maentwrog (SH 66907 41219): on the flood plain of the Afon Dwyryd a gas pipeline trench revealed a deposit of dark grey silt containing a dense concentration of pieces of wood, branches and hazelnuts, at a depth of between 0.9 m and 1.1 m below the present surface. The view of this feature in the base of the pipe trench did not allow the identification of any structure and it was initially assumed to be material deposited by the river; however the assessment of the timber strongly suggests that it was a structure. The exact route of the Roman road from Segontium to Tomen y Mur is not known, but if a line is drawn between the ends of the most likely routes on either side of the flood plain it crosses almost exactly where the wood deposit was found.Footnote ⁷ It therefore seems possible that the structure was the timber foundation for a causeway leading possibly to a bridge across the river and that this was Roman in date. Dendrochronological dating will be carried out in an attempt to prove this.Footnote ⁸

(3) Tremadog, Y Bryn, Bryn-y-Fynwent (SH 55701 40083): a lime kiln of Roman date was found dug into the upper east slope of the low rounded hill. The initial construction cut was roughly circular in plan with steep sides and a flat base. It measured 6 m in length, 4.5 m wide and was 2 m deep. A dry stone wall up to 1.3 m thick lined the cut. Within the main chamber a mortared stone shelf had been constructed, and the base of the main chamber was lined with a very hard concrete-like ‘bowl’, which appeared to be part of the main construction. The interior contained a fused friable mass of white calcined blocks with evidence in places of vitrification. Within what appeared to be a rake-out deposit were found three sherds of black burnished ware and a piece of mortarium rim. Adjacent to the kiln was a smaller kiln or oven, which was oval in shape, measuring 1.5 m in diameter and 0.5 m deep. The flue was stone-lined and the bedrock in the base of the feature was heat-affected. Radiocarbon dates obtained from both features support the Roman date suggested by the pottery. Not far away a deposit of roofing slates in a slipped stacked formation in organised rows was found. All of the unbroken slates were roughly hexagonal in shape with a single offset hole for suspension, some of which still contained the remnants of iron nails. Nearby was found a fragment of a lathe-turned, highly polished, plain shale bracelet. Also in this area were two small pits containing hammerscale, which would seem to indicate smithing activity on the site. It is possible that all these features were related to the construction of the bath-house excavated nearby in the early twentieth century,Footnote ⁹ but the bath-house is not well dated, so the exact relationship is difficult to establish.Footnote ¹⁰

MOMMOUTHSHIRE

(1) Caerwent, Museum Cottage (ST 4686 9071): features revealed included a section of wall running along the southern foundation trench and a wall between Rooms 2 and 3 of House IIN.Footnote ¹¹ The finds included Roman pottery, brick, tegulae, tesserae and painted wall-plaster; all of these had been removed from their original contexts within or as part of the Roman buildings.Footnote ¹²

(2) Gwehelog Fawr, Lancayo Farm (SO 3630 0300): the substantial stone foundations of a structure, probably built of cob, were revealed to the south-east of the Roman temple and within the marching camp.Footnote ¹³ The structure is thought to be of Roman or earlier date. The wall foundation ended abruptly with a substantial post-hole on each side. Attempts to find any corresponding walling on the site were unsuccessful and no satisfactory interpretation for the structure has so far been forthcoming.Footnote ¹⁴

(3) Monmouth: (a) Agincourt Square (SO 50743 12826): a watching-brief recorded a light brown sandy mixture, probably natural, at 1.4 m. This excavation was interesting for its negative results as a few metres to the north previous excavations exposed a ground surface with pre-Flavian pottery, at 1.07 m below the road surface,Footnote ¹⁵ which was believed to be inside the known Roman fort.Footnote ¹⁶

(b) The Barton (SO 50803 12853): an evaluation trench on the frontage of the site revealed a 0.7 m-deep ditch which produced early Roman pottery. The ditch was very similar to others found in the upper parts of Monmouth where the perfectly straight-edged shallow ‘V’-shaped cuts extend some 0.6 m to 1 m into natural. Only one steep side of this ditch was exposed as the other was below the section; the ditch extended to 2.7 m below the modern ground surface. The ditch produced a small assemblage of Roman pottery which included two very small fragments of South Gaulish samian. These fragments could support the suggestion that the ditch is an early military feature, possibly a drainage ditch, inside the Roman fort. There was a thin deposit of charcoal in the upper fill of the ditch, together with lumps of burnt clay, presumably daub, and a large stone, but the main fill was dark clean silt. A single small glass gaming counter was recovered from the silt.Footnote ¹⁷

(c) Glendower Street (SO 50824 12746): a watching-brief recorded a pit running in line with the street on the north side of the trench, cut into the natural. The length, east–west, was 1.5 m at the top of the pit. The upper fill of the pit was of hard red clay containing charcoal, burnt bone and Roman pottery. This layer peeled off cleanly from the lower surface at a depth of 1.4 m. This was heavy with charcoal and burnt bone, probably bovine, with Roman pottery. On the southern side of the pit was a sticky brown loam with charcoal and Roman pottery. Much of the rest of the excavation was cut into deposits which had been disturbed by earlier service trenches. Natural clay was reached at 1.45 m.Footnote ¹⁸

(d) Monmouth School (SO 50967 12716): excavations revealed a pair of parallel ditches, aligned on the presumed orientation of the fort (north-west–south-east) and filled with large quantities of iron slag, bone and Roman pottery. These were possibly drainage ditches located within the fort, or perhaps the remains of the defensive ditches truncated by later developments.Footnote ¹⁹

(e) 20 Monnow Street (SO 50725 12762): an evaluation excavation confirmed that the site lies over the frontage of an early Roman fort. The military ‘V’-shaped ditch was excavated and found to be running at a slight angle across the trench. It appeared to be steeper on the outside (south) edge, although the true angle was obscured as there had been a slump on the south-western half and a post-hole had cut through the south-eastern edge. The lower ditch fill on the south-western side was devoid of pottery but the higher fill on that side produced an assemblage of Roman pottery. The main fill on the south-eastern side of the ditch contained the most datable Roman pottery: two sherds of South Gaulish samian of the middle to late first century. There was an area of iron-working over the natural to the east of the ditch. The hollow was similar to the shape of the base of a furnace but the fired clay was not vitrified, as is often the case in bloomery furnaces, and the slag was rather drossy, more like the result of a smithing process, rather than a small bloomery furnace. The slag appeared to be lightly spread over the back-fill of the Roman ditch, suggesting that it was of a later date. The whole of the excavated area was covered by a gritty brownish deposit containing some Roman pottery but not easily distinguished in colour from the succeeding layer which was similar in nature to soils separating Roman from medieval levels on other Monmouth sites, including the site next door at No. 24 Monnow Street.Footnote ²⁰

(f) Wyebridge Street (SO 51062 12843–SO 50911 12737): a watching-brief recorded a layer of iron slag waste from bloomery furnaces directly on top of prehistoric alluvial sands. The layer incorporated a rim sherd of a Central Gaulish samian form 33 cup.Footnote ²¹

NEWPORT

(1) Caerleon (Isca): (a) Caerleon Secondary School (ST 33527 90549): a field evaluation revealed a probable occupation layer containing Roman pottery with a date-range from the mid-first century to the fourth century, but no structures of Roman date were noted.Footnote ²²

(b) Hillside Lodge, Lodge Road (ST 32759 191004): laying of a new driveway at the front of the property disturbed six Roman cremation burials, five in pottery vessels and one in a glass container with a lead lid.Footnote ²³

POWYS

(1) Beulah, Caerau (SN 9239 5016): test pitting was undertaken to refine the picture of the extent of the vicus, partially known from geophysics.Footnote ²⁴ No evidence for extramural activity was identified to the south-west of the fort and the well-attested civilian occupation on the north-west did not continue beyond a Roman road that bypassed the fort. On the south-east side, limited activity close to the fort gateway is unlikely to have reflected settlement. North-east of the fort, earthworks suggested one or more platforms that could be the stances for buildings, and the test pits confirmed this; one also revealed a small ditch.Footnote ²⁵

(2) Caersws, Craig y Nos (SJ 0312 9180): a watching-brief in the area of the vicus identified a ditch and a parallel gully, together with a metalled surface; the pottery evidence confirmed activity during the second and third centuries.Footnote ²⁶

(3) Hindwell, vicus (SO 2605 6055): excavations concentrated on the eastern extent of the vicus, investigating part of the settlement, the Roman road and part of a triple-ditch system. Buildings and ancillary activity appear to spread out for about 30 m on either side of the road leading from the east gate of the fort and defined, at least on the south side, by a narrow, V-shaped ditch. Settlement evidence continues for about 160 m beyond the fort defences, occupying an area of around 0.7 ha. The excavations uncovered several pits within an area that is presumed to lie behind structures fronting on to the road, evidence from which indicated the presence of smithing activity within the immediate vicinity. A preliminary examination of the pottery suggests the presence of some earlier pottery, possibly of Neronian date, but an absence of later Roman material. The eastward extent of the vicus appears to be bounded by a system of triple ditches which respect the Roman road. An investigation of the innermost ditch showed it to be about 2.65 m wide and 1.2 m deep and although there was no evidence to suggest the position of an accompanying bank, it may be presumed to have been on the interior, given the generally ‘punic’ profile of the ditch. At present the only dating evidence is a probable Neronian Gallo-Belgic butt-beaker, which might suggest that the ditch system was associated with the fort itself, possibly as an outwork, rather than built as a defence for the vicus. In the light of the excavated evidence a re-examination of the geophysical survey results shows a number of possible post-holes on the edge of the Roman road, suggesting a possible four- or six-posted gateway set inside the innermost ditch.Footnote ²⁷

(4) Llanfihangel Cwmdu, Pen-y-gaer (SO 168 219): further work in the vicus revealed that to the west of the road, running from the south gate of the fort, there was a complex of buildings associated with one or more side roads.Footnote ²⁸ The nature and extent of the settlement in this area is not completely clear, but there was significant industrial activity in the form of iron smithing, and it seems likely that the buildings housed various small-scale industries.Footnote ²⁹

HIGHLAND

(1) Belladrum (NH 517 421): excavationFootnote ³⁰ of a metal-detecting find recovered a scattered hoard of 36 denarii (to Antoninus Pius) and Iron Age settlement remains.

MORAY

(1) Clarkly Hill (NJ 131 675): continuing excavationFootnote ³¹ produced further Roman finds from this Iron Age settlement, notably a glass sherd from one roundhouse and two modified sherds of samian ware (Dr 37) in a foundation cache at the foot of a standing stone. Further denarii from the scattered hoard were also recovered.Footnote ³²

ANGUS

(1) Stracathro (NO 617 657): further resistivity and magnetometry surveyFootnote ³³ revealed most of the fort's double-ditch system with ‘parrot beak’ gates, the probable annexe attached to the south-west corner, and the temporary camp to the west. Magnetic anomalies within the fort may represent barrack blocks, while circular features to the west may be roundhouses.

PERTH AND KINROSS

(1) Dalginross: (a) (NN 772 210): geophysical surveyFootnote ³⁴ over the fort's north-east corner confirmed the line of the double-ditch system and outer enclosure where it was not clear on aerial photographs. Internal features were noted in the fort but not the enclosure. (b) Penfillan (NN 7732 2104): three sections across the single-phase V-shaped outer enclosure ditch,Footnote ³⁵ c. 4.0 m wide and 1.0–1.4 m deep, revealed an ‘ankle-breaker’ slot in the base and a narrow layer of slumped turf on the inner edge; vestiges of the turf rampart were also found. A shallow external V-sectioned linear drainage ditch terminating at an angle just short of the enclosure could be of Roman or later date.

(2) Near Callander (NN 63 08): radiocarbon dating of a recently-recognised old find of a leather shoe confirmed its Roman period date; the style is an unusual hybrid one.Footnote ³⁶

(3) Castle Craig (NN 97604 12714): further excavationFootnote ³⁷ of this lowland broch produced more Roman finds, including samian, glass and a brooch.

STIRLING

(1) Stirling: (a) Kildean (NS 7801 9529): an evaluationFootnote ³⁸ found no trace of the Roman road at this point. (b) Beechwood Park (NS 79305 92025): geophysical surveyFootnote ³⁹ located possible traces of the Roman road line.

EAST DUNBARTONSHIRE

(1) Bearsden, 16 Roman Road (NS 5463 7203): a watching-briefFootnote ⁴⁰ over the south-east corner of the fort annexe indicated that Victorian housing had destroyed Roman remains in the area.

FALKIRK

(1) Castlecary: (a) (NS 79 77): metal-detecting c. 200 m south of the fort produced a fine copper-alloy lion-headed mount and a damaged enamelled lunate cavalry harness pendant.Footnote ⁴¹ (b) Allandale Football Pitch (NS 788 786): evaluation workFootnote ⁴² located the badly-truncated remains of the Wall base, and the better-preserved Ditch and upcast mound, on the Ordnance Survey line.

(2) Falkirk: (a) Elgin Place (NS 8903 7961): a ground-penetrating radar surveyFootnote ⁴³ confirmed the revised position for the Wall Ditch suggested by trial-trenching in the area.Footnote ⁴⁴ (b) Comely Park House (NS 891 795): further trial-trenchingFootnote ⁴⁵ confirmed the dog-leg in the alignment of the Ditch in this area, and found possible traces of the Military Way c. 35 m to the south. A fortlet has been hypothesised to explain this realignment;Footnote ⁴⁶ limited trenching uncovered no trace. Radiocarbon dating of a charcoal layer showed the ditch was half-filled by the early medieval period.

(3) Polmont, Old Polmont Meeting Hall (NS 938 792): an evaluationFootnote ⁴⁷ found traces of the Military Way.

WEST LOTHIAN

(1) Castle Greg (NT 0502 5925): resistivity surveyFootnote ⁴⁸ ( fig. 2) over the upstanding fortlet revealed traces of the internal road and drain network, possible buildings and rampart-back ovens.

FIG. 2.

Resistivity survey of Castle Greg fortlet. (Courtesy of Edinburgh Archaeological Field Society per Ian Hawkins)

EAST LOTHIAN

(1) Inveresk, Eskmills (NT 3394 7230): two linear ditches recorded in a watching-briefFootnote ⁴⁹ to the north-west of the fort may be part of the Roman period field-system, although no dating evidence was recovered.

SOUTH LANARKSHIRE

(1) Crawford, Clyde Wind Farm (NS 9634 1864): a sectionFootnote ⁵⁰ across the Clyde valley road some 3 km SSE of Crawford fort revealed a foundation, c. 8 m wide and up to 0.25 m thick, of large cobbles and sub-rounded stones set directly onto the old ground surface, with a 0.25 m-thick levelling deposit of angular coarse gravel which also filled the voids. It was surfaced with a 0.05 m-thick layer of coarse gravel; possible kerb stones were noted on the east. A quarry pit and dump of stones, perhaps unused road metalling, were also found.

SCOTTISH BORDERS

(1) Kippilaw (NT 556 292): excavationFootnote ⁵¹ over a coin scatter revealed a hoard of 79 denarii and a single dupondius to Severus (a.d. 207), buried on the edge of a small stream valley. The number of die-linked coins suggests a source from a military paychest.

(2) Synton (NT 48 21): a scattered hoard of 225 denarii to Commodus was found in the vicinity of an Iron Age hillfort. Excavation of a 15 by 15 m trench found the pit which held the hoard, but no other archaeological remains.Footnote ⁵²

DUMFRIES AND GALLOWAY

(1) Birrens (NY 2190 7518): magnetometer surveyFootnote ⁵³ cast fresh light on enclosures known from aerial photography in the field west of the upstanding fort ( fig. 3). The high visibility of the triple ditches of an earlier (Flavian?) fort, aligned north-west–south-east, suggests rubbish was used to infill them; their irregular spacing might point to more than one phase. This fort's interior is obscured by roads and buildings associated with a triple-ditched annexe for the later fort. One road runs to the fort's west gate, the other at a north-west–south-east angle across the annexe, this latter road being densely lined with buildings on both sides. Activity spread west to the defences, but is notably absent closer to the fort. Ditches of a field-system of uncertain date to the west lie beyond and respect the annexe ditches.

FIG. 3.

Magnetometry survey of Birrens. (Courtesy of Dr R. Jones)

(2) Carzield (NX 968 818): a watching-briefFootnote ⁵⁴ on cable trenches inside the fort examined part of the north-west rampart and ditch and remains of internal buildings.

SOUTH TYNESIDE

(1) South Shields fort (Arbeia) (NZ 365 679): excavation continued in the same 10 by 17 m area, 20 m south-west of the south angle of the fort.Footnote ⁵⁵ The road running along the south-east edge of the excavation, described last year, proved to have a long history, with many earlier surfaces. The earliest of these may be contemporary with a deep dump of clay, at least 8 m wide, which extends across the entire trench to the north-west of the road, with the appearance of upcast from an early ditch or even the rampart of an earlier fort (facing north with the road running behind?).

The earliest road surface seen was followed by a period of disuse, silting and rubbish dumping, which sealed the heel or ‘back’ of the clay bank just described. Then the road was remade at a higher level. Contemporary with this the two metal-working pits described last year were cut into this clay dump. The north-east pit had been used for gold- and silver-smithing: an unfinished hexagonal silver ring and a small flake of gold were recovered from the lower fill. Coupled with several fragments of crucibles, this indicates that high value items were being crafted here. The elongated south-west pit resolved into a metalled sunken walkway leading from the reinstated road and into the working pit. Further south-west the corner of a rectilinear cut into the clay dump may have represented a building associated with the craft phase.

The backfills of the cut features were stratigraphically earlier than the many later surfaces of the road and a stone building previously described,Footnote ⁵⁶ and contained pottery which on Hadrian's Wall is no earlier than the mid-third century in date (notably Campanian ‘black sand’ amphora), showing that the craft activity was probably of third-century date and was superseded by a further phase of extramural building in the late third/early fourth century. At this time the road may have led to an annexe on the south-east side of the fort.Footnote ⁵⁷

NORTHUMBERLAND

(1) Vindolanda (NY 769 663): (a) vicus: the 2012 excavations were primarily focused on the remains of the third-century extramural settlement in an area between the third-century bath-house and a Romano-Celtic temple to the west of the fort.Footnote ⁵⁸ The excavations linked the site of this temple to the west wall of the third-century baths, some 65 m to the east. In this area, during excavations in 1930, Eric Birley had excavated a finely constructed aqueduct channel (see fig. 4), which had been set above earlier water-filled ditches. The work in 2012 demonstrated that the stone aqueduct channel had merely been the last and most permanent version of a long-established network of ditches and aqueducts in the area, all of which were cut through clay banks to manage the distribution of water from a series of powerful natural springs. The pottery from many of the channels proved to be pre-Hadrianic in date. Below the pre-Hadrianic water channels were the remains of at least one extramural pre-Hadrianic timber building. This structure had been built in the usual method of wattle and daub fences with foundations set into the natural boulder clay, a style of construction used widely within the timber forts at Vindolanda. A small trench, approximately 4 by 5 m, explored up to 20 per cent of this structure and produced many perfectly preserved items from within the laminated layers on the floor of the building. These included over sixty items of footwear, tent panels, and fragments of three different stylus tablets.

FIG. 4.

Vindolanda: the 2012 excavation area in the vicus.

To the west of the aqueduct channel, and no longer connected to the stone channel blocks located by Eric Birley, was a stone-built water-tank. The builders of the water-tank had tapped into a natural fissure in the bedrock below and this has proved to be one of the main sources of water for the Roman period site. It is likely that such springs were the primary reason why the Roman settlement was placed where it was in the immediate landscape.

To the east of the stone aqueduct channel a large defensive wall and ditch system of Antonine date was uncovered. This wall and associated ditch system may have been designed to entirely encircle the extramural settlement in the Antonine period c. a.d. 140–160. An Antonine variant of the later aqueduct had been built to cross the ditches and carry the fresh spring water into the settlement through the foundations of the defensive wall.

A final area of exploration to the south of the Romano-Celtic temple and water-tank uncovered third-century roads and cobbled areas as well as the northern defensive ditch from the Period IV Vindolanda fort c. a.d. 105–120. Artefacts from this area included pottery lamps, wine amphorae, face pots, and an inscription on stone from one of the Vindolanda pre-Hadrianic garrisons, I Tungrorum. The inscription was dedicated to a Tungrian female deity named as ‘AHVARDVA’ (see below, Inscriptions, No. 5). A stone statue of the goddess Fortuna (minus her head) was also recovered from this vicinity during the excavations.

(b) The field to the north of the site: after three years of exploratory trenching, it can now be confirmed that there were multiple periods of military occupation in the field, from the late first century to the first half of the second century, with a substantial gap from the middle to the end of the second century before a re-occupation in the third century. There is as yet no material that can solidly date occupation to the a.d. 70s, but it is almost certainly pre-Hadrianic in date (the material recovered in 2012 is currently being processed).

In 2009 and 2010 two large rectangular buildings were found; one incorporated an internal colonnade.Footnote ⁵⁹ These structures were built using large stones similar to those used in most third-century extramural settlement buildings at Vindolanda and almost certainly show that the third-century vicus extended well into this field. In 2012 a large ditch was found located to the north and west of the aisled building. Excavation and geophysical evidence suggests that this ditch continues between the two third-century stone buildings known in the field. Furthermore, a series of 25 coins of the 260s and 270s (Gallienus, Cornelia Salonina, Victorinus, Claudius II, Tetricus, and Tetricus II), found throughout the fill and right to the bottom of the ditch cut, securely dated the ditch fill to the 270s or soon thereafter. This feature and its contents suggest that this part of the vicus was partially defined if not defended by a ditch before it was abandoned, c. a.d. 280.

CUMBRIA

(1) Rockcliffe, Guard Mill (NY 33690 67392): the north-east corner of a temporary camp was recorded from the air in 2010.Footnote ⁶⁰

CUMBRIA

(1) Maryport (Aluana), Camp Farm (NY 042 375): a second season of excavations to determine the archaeological context of the famous ‘Maryport altars’ found during the 1870s.Footnote ⁶¹ Following on the 2011 excavations, which had demonstrated the altars were reused as packing in post pits, two undisturbed pits were identified, one of which included a substantially complete Roman altar dedicated to I.O.M. by T. Attius Tutor, prefect of cohors I Baetasiorum (see below, Inscriptions, No. 3). Excavation of the ancient pits also allowed the team to characterise the working methods of antiquarian diggers and to demonstrate that there had been at least one phase of antiquarian investigation prior to the 1870s diggings, sometime before 1725 ( fig. 5).

FIG. 5.

The Maryport pits 2012, view looking south. (Photo: Maryport Project)

Several ditches were also investigated within the excavation area. Of these the most significant was a rough-cut, sub-circular feature or ‘ring ditch’ which cut through traces of the early occupation of the site. The filling of this ditch, which included part of a Roman altar, was dated to the late fourth century or later, on account of its inclusion of fragments of Crambeck Parchment Ware. This ditch was found to truncate an earlier surface in one area, but no coherent traces of earlier activity remained elsewhere.

The sub-circular ditch was in turn cut by post-pits in two places, indicating that it preceded the construction of the timber structures associated with these pits. These structures were clearly substantial and the pits indicate that they were constructed with 300 mm-square timber uprights. The interpretation of the structures remains problematic. No floor surfaces survive. Their importance is, however, suggested by both their size and location, standing as they do at the most visible point on site. The arrangement of posts on the western side of the complex appears to indicate the presence of an apse, while the use of smaller post-pits to the east may indicate an elaborated entry-way there.

Immediately north-east of these timber buildings lay a cemetery, which was separated from the structures by a fence. This cemetery included stone-lined long-cists and consisted of three north–south and four (later) east–west-oriented graves. One of the north–south-oriented graves yielded a textile fragment with a calibrated radiocarbon date of a.d. 240–340 at 95.4% confidence (SUERC-44284 (GU29199)); another yielded a nummus of the house of Valentinian in its fill. Two of the east–west graves cut their north–south predecessors. Soil conditions mitigated against the survival of human bone, but it was possible to demonstrate that at least one of the deceased was a child.Footnote ⁶²

(2) Maryport area: 118 silver denarii and 1 sestertius. Denarii: Mark Antony (32–1 b.c.), 3; Nero (a.d. 54–68), 2; Galba (a.d. 69), 2; Vespasian (a.d. 69–79): Vespasian, 34; Titus Caesar, 3; Domitian Caesar, 2; Titus (a.d. 79–81), 1; Domitian (a.d. 81–96), 4; Nerva (a.d. 96–8), 1; Trajan (a.d. 98–117), 19; Hadrian (a.d. 117–38): Hadrian, 20; Sabina, 1; Antoninus Pius (a.d. 138–61): Antoninus Pius, 13; Diva Faustina I, 4; Marcus Aurelius Caesar, 5; Faustina II, 1; Marcus Aurelius (a.d. 161–80): Divus Antoninus, 1; Faustina II, 2. A worn and damaged sestertius is possibly an issue of Hadrian. Probably deposited c. a.d. 161–76.Footnote ⁶³

COUNTY DURHAM

(1) Binchester (Vinovia) (NZ 2102 3134): a fourth season of the new campaign of excavation took place in 2012 and saw continued work in the dextral half of the praetentura, close to the east corner of the fort, as well as in that part of the vicus lying beyond the south-east defences.Footnote ⁶⁴ The latest surface of the intervallum was exposed along with a drain on its outer edge into which smaller drains serving the adjacent building debouched. A much larger drainage culvert was encountered immediately adjacent to the end wall of the building. In the area outside the fort to the south-east, investigation continued of buildings on the north-east side of Dere Street. Set into the surface of the road in front of these buildings are a number of stone blocks about 0.25 m square. Each one of these has a shallow square recess cut into its surface clearly designed to receive a timber post indicating that in a late stage these buildings had a veranda constructed along their frontage. Some of these recessed blocks were found to have been re-used and incorporated into the footings of the latest of a series of extensions to the frontage of the southernmost building within the excavated area thought to be part of the regimental bath-building.

The removal this year of the heavy stone surfacing in the smaller of the two rooms exposed so far, followed by excavation of the underlying deposits, exposed some fascinating structural details of this part of the baths. Close to the east face of the trench a pair of opposed semi-circular niches were revealed in the side walls of this chamber. These presumably once contained statues of deities appropriate to such a building, while the overall dimensions of the room suggest it was the entrance corridor into the complex. Immediately before the niche in the northern wall was a blocked-up doorway that had originally given access into the neighbouring room. When the blocking occurred a new doorway had been inserted just beyond the niche. Contemporary with these developments was the insertion of a wall across the width of the corridor incorporating a centrally-placed doorway which still has its monolithic jambs largely intact. In terms of both form and size this doorway is very similar to that in the western side of the circular hypocausted room uncovered in the 1870s a few metres to the east.Footnote ⁶⁵

(2) Greta Bridge, The Morritt Arms (NZ 0850 1328): excavation in association with building works at the latter property, at a spot immediately beyond the north-east defences of the Roman fort, exposed elements of a substantial masonry building.Footnote ⁶⁶

(3) Langley Moor (land north of Brandon Lane and west of Angus Street) (NZ 2495 4064): a 90 m length of Roman roadFootnote ⁶⁷ was excavated in advance of development. The average width of the road was 6.5 m, with well-defined kerbs and a damaged but surviving metalled surface and no roadside gullies. The road overlay ditches of late pre-Roman Iron Age date, but it was not possible to determine whether these were still in use when the road was built. The road continues exactly the alignment previously seen to the south-west between Brancepeth and Brandon and confirms Wright's surmiseFootnote ⁶⁸ that the road was heading for a crossing of the River Browney at or close to Stone Bridge (NZ 258 414).Footnote ⁶⁹

NORTH YORKSHIRE

(1) Aldborough (Isurium Brigantum) (SE 4060 6632): fieldwork in 2012 comprised two principal elements. First, the magnetometer surveyFootnote ⁷⁰ of the site continued, with substantial areas covered across the northern part of the walled area, and on the extramural zones to the east and west of the townFootnote ⁷¹ . The northern intramural area provided exceptionally clear results, and showed details of the planned street grid and large areas of domestic buildings as well as public monuments, including the probable public baths. Surprisingly, the area immediately inside the wall was largely unoccupied but revealed evidence that may indicate the presence of an earlier defensive circuit. Second, Ground Penetrating Radar surveyFootnote ⁷² in the churchyard was completed providing further information about the forum. The buried remains in the immediate vicinity of the church were fragmentary, but the work confirms that the medieval church sat in the centre of the forum square, although further research is needed to clarify the relationship with the walls excavated to the north in 1770.Footnote ⁷³

(2) Askrigg, Askrigg village to Horrabank Farm (NY 9500 9132–NY 9525 9170): a 400 m length of Roman road has been identified on LIDAR imagery running north-east from Askrigg village ( fig. 6). It survives as an agger 5 m wide running up Stony Bank to Horrabank Farm underlying later strip fields and field boundaries. This is part of a road that runs north-north-east from Bainbridge towards Greta Bridge via Feetham in Swaledale, Stang Bridge in Langthwaite, Arkengarthdale, and Stang Top towards Scargill. It was recorded by Warburton who walked part of it in 1717.Footnote ⁷⁴ He mapped in on his map of Yorkshire in 1720 where he showed it running to Barnard Castle. However current research indicates that it runs to Greta Bridge.Footnote ⁷⁵

FIG. 6.

Bainbridge to Greta Bridge Roman road. The line of the agger between Askrigg village and Horrabank Farm, underlying strip fields and field boundaries. (© The Environment Agency)

(3) Barlby: 4 silver denarii and 5 base silver/base metal radiates: denarii: Elagabalus (a.d. 218–22), 1; Severus Alexander (a.d. 222–35), 2; Maximinus (a.d. 235–8), 1. Radiates: Gordian III (a.d. 238–44), 1; Trebonianus Gallus (a.d. 251–3), 1; joint reign of Valerian and Gallienus (a.d. 253–60): Valerian, 1; Gallienus, 1; Postumus (a.d. 260–9), 1. Deposited a.d. 260–9 or later.Footnote ⁷⁶

(4) Brearton: 2 silver denarii of Vespasian (a.d. 75) and Titus (a.d. 80).Footnote ⁷⁷

(5) Catterick, Pallet Hill Quarry (SE 2318 9910): a magnetometer survey has been carried out on the site of the Roman temporary camp to the east of Catterick racecourse originally detected as a cropmark.Footnote ⁷⁸ Problems in geo-referencing had resulted in a degree of uncertainty as to whether there was more than one camp represented by the cropmark data.Footnote ⁷⁹ The results would seem to confirm that the cropmark record represents a single camp measuring 160 m by 240 m (3.84 ha) ( fig. 7). On the north side of the camp the principal feature is a central gateway protected by an external traverse. The breach here is occupied by three parallel anomalies (trenches), which are presumably defensive features. Between the central gateway and the north-east corner is another, apparently undefended, breach which is mirrored by a similar one on the south side of the camp. Two probable external traverses or tituli were detected along the eastern side of the camp, although these do not correspond with breaches, and are perhaps indicative of the camp having more than one phase of use. The lack of data for the south-western part of the camp precludes any consideration of the camp's overall symmetry.

FIG. 7.

Pallet Hill Quarry, Catterick: magnetometer survey of Roman temporary camp. (© Archaeological Services WYAS)

Two smaller ditched enclosures have been detected at the north-eastern edge of the survey area, whilst other potential features were detected to the east of the camp. There are virtually no internal magnetic anomalies, although there is one cluster that corresponds with the southernmost external traverse along the eastern side. Whilst the geophysical data indicate that the cropmarks probably represent a single camp plan, there is a suggestion of another enclosure to the south-east. Although of similar orientation and form, the geophysical response suggests that the construction of this potential enclosure was different, perhaps defined principally by an earthwork bank, although there is no certainty that it is of Roman date.Footnote ⁸⁰

(6) Near Stokesley: hoard consisting of 39 silver siliquae:Footnote ⁸¹ Julian (a.d. 360–3) Arles, 2; Valens: (a.d. 364–7) Rome, 2; (a.d. 367–75) Trier, 2; uncertain mint, 1; (a.d. 375–8) Trier, 1; Gratian: (a.d. 367–75) Trier, 1; (a.d. 375–8) Trier, 1; (a.d. 379–383) uncertain mint, 1; Valentinian II: (a.d. 375–83) Trier, 1; (a.d. 378–83) Aquileia, 1; House of Valentinian (a.d. 379–88) Trier, 1; Magnus Maximus (a.d. 383–8) Trier, 1; Theodosius (a.d. 389–94) Milan or Trier, 1; Arcadius (a.d. 392–95) Trier, 6; uncertain mint, 2; Eugenius: (a.d. 392–4) Trier, 1; Lyon, 1; Honorius (a.d. 395–402) Milan, 5; Arcadius or Honorius: (a.d. 395–402) uncertain mint, 1; Milan, 3; House of Theodosius (a.d. 388–408) uncertain mint, 2; Constantine III (a.d. 407–8) Lyon, 1; Uncertain emperor/issue, 1. Deposited a.d. 407–8 or later.Footnote ⁸²

(7) Near Wetherby: hoard consisting of 38 silver siliquae and 2 bronze nummi: a.d. 355–64: Lyon, 6; Arles, 8; uncertain mint, 2; a.d. 364–7: Lyon, 6; Arles, 1; Thessalonica, 1; uncertain mint, 2; a.d. 367–75: Trier, 6; a.d. 375–8: Trier, 1; a.d. 395–402: uncertain mint, 1; Irregular, 4. Deposited c. a.d. 378–402. One highly clipped coin of a slightly later type (a.d. 395–402) could have been deposited well into the fifth century after a long period of circulation. The difference in the degree of clipping of this coin and the others in the group suggests that they were deposited some time before this and that this later coin may be a stray find. The fact that the hoard was found near a spring would point to a votive or ritual context for its deposition.Footnote ⁸³

(8) Ugthorpe: silver denarius of Faustina II under Antoninus Pius. Deposited a.d. 192 or later.Footnote ⁸⁴

YORK

(1) York (Eboracum) Haymarket hostel and car park, Peasholme Green, York (SE 6075 5192): located to the south-east of the fortress Eboracum within part of the canabae. Evidence was found for the Holocene landscape and environment at the time of the establishment of the fortress. The two earliest indicators of Roman activity were a V-shaped ditch and an inhumation burial. The ditch was interpreted as a continuation of a similar feature identified in the adjacent Hungate excavation.Footnote ⁸⁵ The size, extent and stratigraphic position of the ditch suggest it was an early boundary feature of some significance. The following phase of land-use was dominated by pit digging, possibly resulting from the exploitation of clay deposits in the area. Following this the area was put to agricultural land-use. Subsequently the ground level was raised when large-scale landscaping was undertaken. More than a metre in depth of redeposited natural sandy silt was used to level off the natural ground slope towards the south and south-east in the direction of the River Foss, presumably to improve what would have been a low-lying and probably seasonally flooded area of land. The final phase of Roman activity comprised a group of sparsely distributed inhumation burials cut into this deep terracing layer, all of which were prone and laid with the head oriented approximately north-east. A small greyware pot had been placed at the head end of one of these burials; otherwise they were devoid of grave goods.Footnote ⁸⁶

EAST RIDING OF YORKSHIRE

(1) Howden area: 1 base silver denarius of Severus Alexander (a.d. 222–35) and 1 base silver radiate of Philip I (a.d. 244–9).Footnote ⁸⁷

(2) Fangfoss: 5 base silver/base-metal radiates: Trebonianus Gallus (a.d. 251–3), 1; Valerian and Gallienus (joint reign) (a.d. 253–60): Valerian, 1, Salonina, 1; Claudius II (a.d. 268–70), 2. Deposited a.d. 268–70 or later.Footnote ⁸⁸

WAKEFIELD

(1) Nostell Priory, new visitor car park (SE 409 172): excavations in 2009 revealed the remains of a multi-period Romano-British settlement. Prior to the excavation nothing was known of a Roman presence at Nostell apart from the occasional sherd of possibly Roman pottery encountered in earlier watching-briefs on the site and a suggestion that the Roman road from Elslack to Bradford (Margary 721) may have continued past the site on the line of the modern A638.

Five phases were found. An Iron Age phase consisted of a small number of shallow ditches indicative of land division and an enigmatic, massive ditch that was backfilled and later re-cut in the early Roman period.

The earliest Roman phase ( fig. 8) was dated by Flavian-Trajanic pottery of types associated solely with Roman military sites. In addition, the presence of wasters indicated the operation of a previously unknown nearby kiln producing both fineware and mortaria. Additional support for the military nature of this phase comes from other material culture found in unphased contexts including bessales and a box-flue tile that probably originated in a hypocaust and a quern of twin-hopper design that is usually associated with military establishments. The most likely explanation for this is that the remains represent a vicus adjacent to an as yet unidentified Roman fort.

FIG. 8.

Nostell Priory: features from the early Roman phase (black = certain, grey = probable).

Three further phases at the site characterised by land divisions, pits and an unusual double-flued crop-drier were also identified, dating from the mid-second to the late third or fourth century. The pottery was of types familiar from rural settlement elsewhere in the region.

Several undated features were present, the most significant of which was a group of burials, including a stone-lined grave, located towards the eastern limits of the main excavated area. A complete absence of bone (animal or human) surviving in any of the features on the site hindered interpretation, so it is not clear whether these were originally full burials or ‘empty’ graves as occasionally found elsewhere. Near the graves an undated 3 m-diameter unbroken ring ditch has been identified as a possible shrine.Footnote ⁸⁹

(2) Wrenthorpe, Ruskin Avenue (SE 3205 2304): excavation produced evidence of two Romano-British enclosures: Feature 1, although only partially excavated, appeared to be circular, based on the geophysical survey and excavation evidence, while Feature 2 was irregular in shape and defined by a curvilinear ditch bisected by a shallow linear ditch. The Feature 2 enclosure appeared to have been segmented, though plough truncation of the site had occurred. Apart from the internal division within Feature 2 no other features were identified. The excavated sections of Feature 2 indicated that a bank had existed at least in part around the external circuit. The preferred interpretation for the function of the enclosures is that they were linked to the control and management of livestock.

Although there were only a small number of finds, the assemblage was dominated by Roman ceramics; mainly second-century in date, and the pottery is typically domestic in nature. The exception was a single sherd of third-century mortaria and a single lithic flake, although these may be intrusive and residual respectively.Footnote ⁹⁰

DONCASTER

(1) Doncaster: 86 silver denarii and pottery sherd fragments: Mark Antony (32–1 b.c.), 2; Galba (a.d. 69), 1; Vespasian (a.d. 69–79): Vespasian, 6; Titus Caesar, 1; Domitian (a.d. 81–96), 5; Nerva (a.d. 96–8), 1; Trajan (a.d. 98–117), 19; Hadrian (a.d. 117–38), 20; Sabina, 1; Antoninus Pius (a.d. 138–61): Antoninus Pius, 8; Diva Faustina I, 4; Marcus Caesar, 4; Faustina II, 4; Marcus Aurelius (a.d. 161–80): Marcus Aurelius, 5; Lucius Verus, 1; Lucilla, 2; Divus Antoninus, 2. The hoard total now stands at 396 denarii and a closer terminus post quem is supplied by a late coin of Marcus Aurelius, suggesting deposition in a.d. 175 or later.Footnote ⁹¹

(2) Fingley and Rossington Regeneration Route Scheme (FARRS) (SK 5894 9931–SK 6295 9942): following a scheme of geophysical survey and fieldwalking, trial trenching of a 33 ha site confirmed the existence of extensive enclosures or field-systems located close to Rossington Roman fort. The archaeological features revealed during the evaluation were largely undated, but the presence of a redeposited sherd of Romano-British pottery in a probable post-medieval ditch hints at a similar date for some of the ditches, although a prehistoric date cannot be ruled out. The ditches towards the western end of the route are likely to be Romano-British in origin, and represent boundary ditches or short-lived animal enclosures. The lack of any finds strongly suggests that any settlement lay away from the site. The area closest to Rossington fort was devoid of Roman features and it is likely that the area was pasture or scrubland. Peat deposits were also revealed in the central part of the site, which was radiocarbon dated to the Early Bronze Age.Footnote ⁹²

(3) Rossington Inland Port (SK 59185 98775): the excavation of 50 evaluation trenches revealed probable Romano-British features in the centre of the site, located on a small rise in the landscape. Part of a square (60 m by 60 m) double-ditched enclosure identified by prior geophysical survey was revealed and several unabraded fragments of Romano-British greyware and samian pottery were recovered from both ditches; the pottery is undiagnostic and broadly dates from the second to fourth centuries and is indicative of a nearby settlement. Ditches revealed in other trenches formed part of a possible brickwork field-system to the north of the enclosure. A large water-hole or well was revealed to the east of the enclosure, the fills of which contained two sherds of Romano-British pottery. Ditches revealed in the north-west of the area corresponded to a system of brickwork enclosures identified as geophysical anomalies in previous phases of work.Footnote ⁹³

(4) Sprotbrough: 2 silver denarii of Julia Domna (under Septimius Severus and Caracalla). Deposited a.d. 196–211 or later.Footnote ⁹⁴

(5) Stainforth, Ashfields land north of Kirton Road (SE 6650 1275): evaluation excavations in advance of a proposed colliery waste storage site followed on from geophysical survey which had identified an extensive area of features suggested as Iron Age and Romano-British field-systems and settlement sites extending over 12 ha. Roman period ditches formed small enclosures or were part of a much wider network of field-systems dating to the second to third or fourth centuries. The evidence for Romano-British activity appears to cluster into two main areas: towards the western fringe and in the central western part of the site which lies on a slightly higher ridge of sand. Post- and stake-holes uncovered throughout the excavations hint at wider structural alignments outside the limit of the trenches. The artefactual evidence is indicative of settlement activity with primarily locally-produced pottery recovered, but also a small number of tile/brick fragments, a flagstone and a coin. Evidence of metalworking activity was also found in the form of smithing slags and possible furnace lining. Charred plant remains recovered from the Romano-British features are typical of those of settlement waste and activities.Footnote ⁹⁵

SHEFFIELD

(1) Plumbley area (2012 T481): 5 silver denarii; 1 unidentified c. mid-second century b.c. Roma/biga type; Q. Fufius Calenus and P. Mucius Scaevola (70 b.c.), 1; Mark Antony and M. Barbatius Pollio (41 b.c.), 1; Augustus (Rome, struck 19–4 b.c.), 1; Vespasian (Rome, a.d. 74), 1. Deposited a.d. 74 or later.Footnote ⁹⁶

CHESHIRE

(1) Knutsford area: 56 silver denarii and 2 sestertii, 2 finger-rings, 2 brooches and pottery sherds.Footnote ⁹⁷ Denarii: Mark Antony (32–1 b.c.), 1; Galba (a.d. 69), 1; Vitellius (a.d. 69), 2; Vespasian (a.d. 69–79), 3; Domitian (a.d. 81–96), 3; Nerva (a.d. 96–8), 1; Trajan (a.d. 98–117), 8; Hadrian (a.d. 117–38), 9 + 1 sestertius; Antoninus Pius (a.d. 138–61): Antoninus Pius, 9 + 1 sestertius; Diva Faustina I, 2; Marcus Aurelius Caesar, 2; Faustina II, 1; Marcus Aurelius (a.d. 161–80): Marcus Aurelius, 4; Lucius Verus, 1; Faustina II, 2; Diva Faustina II, 1; Commodus Caesar, 1; Commodus, 1; Commodus (a.d. 180–92), 4. Deposited a.d. 190–1 or later.

The artefacts consisted of two late second-century silver gilt trumpet brooches, one smaller brooch and two silver finger-rings set with a red semi-precious stone (possibly carnelian), one engraved with a winged figure.Footnote ⁹⁸

(2) Middlewich: (a) Sea Bank (SJ 7054 6625): an evaluation of a small plot of land c. 140 m to the east of the parish church and south of the main area of Roman settlement in the town, revealed alluvial deposits between c. 1.5 m and 2 m deep, which overlay Boulder Clay. The alluvium derives from the adjacent valley of the River Croco which has been much altered in the recent past and is now occupied by the Trent and Mersey Canal. These deposits sealed a number of post-holes, pits, and gullies, some of which produced Roman pottery and fragments of briquettage. This evidence suggests that, like Nantwich a little to the south, Middlewich may have areas where well-preserved and waterlogged Roman deposits survive beneath significant accumulations of more recent alluvium.Footnote ⁹⁹

(b) Land off Jersey Way (SJ 7065 6660): a c. 2 ha site within the area of the Roman settlement at Middlewich was subject to evaluation in 2008.Footnote ¹⁰⁰ Those parts of the site where Roman period remains had been identified were subject to excavation in 2012 prior to residential development. Ditches excavated in the northern part of the site seemingly represent two phases of Roman land division and field management on different alignments. The earliest contained pottery dated to the early/mid-second century, whilst palaeoenvironmental assessment of samples taken from the later phase has indicated that the Roman landscape had been largely cleared and was dominated by grasses, with some evidence for arable cultivation. The alignment of a track that crossed the central part of the site seemed to respect the later ditches, and also appeared to separate different zones of activity.

The shallow foundation trenches for a timber building, 6 m wide and over 9 m long, were excavated in the southern part of the site. This may represent first-century activity, and appears to have been dismantled by the early second century. Fragments of a late first-century facepot, a wine-strainer and a second-century triple ring vase, all suggest elements of ritual practice.

A series of large intercutting pits lying to the north of the track had a more industrial character and produced pottery dating to the late first to early/mid-second century. Whilst the function of these pits is presently uncertain, clear evidence for salt-working was provided elsewhere on the site. The physical remains included three timber-lined wells, a large timber-lined tank, and the base of a brine-boiling hearth. A substantial assemblage of briquetage recovered from these features includes hearth or oven fragments, and a range of pedestals, other supports, and carefully made tile-like objects, or plates. Only a single rim fragment from a potential evaporation vessel was noted, suggesting that the brine was boiled in lead pans rather than ceramic containers. Several of the deeper features were waterlogged, and the preservation of organic material proved to be excellent. A worn dupondius of Trajan (a.d. 103–17) and an iron blade were discovered at the base of one of the wells, whilst a folded sheet of lead had been placed carefully into an overlying deposit, together with a large sherd of a samian bowl dating to a.d. 150–75.

Preliminary dating has indicated considerable activity during the Flavian period, although none of the samian ware is earlier that c. a.d. 65/70. Following a reduction in the pottery supply during the Trajanic period, there was a huge increase in Hadrianic material. The quantities of samian and amphorae point to a supply comparable to military and major urban sites. Activity during the third and early fourth centuries is attested by fragments of pottery, although this was only recovered in small amounts.Footnote ¹⁰¹

(3) Warrington, The Horizon Centre, Loushers Lane (SJ 6159 8681): an evaluation of the former school was carried out in connection with proposals for the redevelopment of the site for residential purposes. It demonstrated that across much of the site, sand extraction and recent landfill had destroyed any archaeological remains that might have been present. In the north-east corner of the site, however, Roman remains were identified. These lay in close proximity to walls discovered during work carried out in connection with extensions to the former centre in the 1990s. These formed the western element of a Roman building investigated during the 1930s in the carpark to the east.Footnote ¹⁰² This building produced evidence of a hypocaust and painted wall-plaster and lay to the east of the Roman industrial centre at Wilderspool, from which it was separated by the Lumb Brook, a minor tributary of the Mersey which has now been obliterated by recent industry.Footnote ¹⁰³

LINCOLNSHIRE

(1) Bittesby: five silver denarii: Roman Republic (P. Plautius Hypsaeus, 70 b.c.), 1; Vespasian (a.d. 69–79), 1; uncertain Flavian (a.d. 69–96), 1; Antoninus Pius (a.d. 138–61): Antoninus Pius, 1; Diva Faustina, 1.Footnote ¹⁰⁴

(2) Gainsborough: a pierced denarius of Diadumenian (as Caesar). Deposited a.d. 217–18 or later.Footnote ¹⁰⁵

(3) Waddingham (SK 9672 9518): a large Roman temporary camp identified from the air ( fig. 9). The camp measures approximately 240 m by 320 m (ditch to ditch) and lies approximately 270 m east of Ermine Street (the A19). The southern half of the camp and most of the west side were seen and recorded as cropmarks during the flight, though only the south-east corner was easily distinguished. Follow-up examination of Google Earth imagery taken in 2007 revealed a much clearer record of the south-west corner and western side of the camp, displaying a clearly marked entrance. The northern side of the camp is not apparent but must have lain along the top of the slope defining the shallow valley of a small watercourse.Footnote ¹⁰⁶

FIG. 9.

Waddingham Roman camp. Areas of the image have been selectively enhanced to improve visibility of the cropmark evidence. (NRHE 28304/45 25-05-2012 Dave MacLeod © English Heritage)

(4) Wragby (TF 131 783; 131 793; 143 786): the final year of an intensive fieldwalking programme in the parish of Wragby has identified three dense scatters of Roman pottery, building material and artefacts that are located to the north-west, north and north-east of the modern village and are suggestive of the presence of Roman period buildings. Also the fieldwalking programme has found that the vast majority of fields within the parish produce moderate quantities of abraded Roman ceramic tile, pottery and occasional metal objects, suggestive of intensive cultivation and manuring in the Roman period.

A limestone coffin was discovered utilised as a garden feature at the former offices of Hugh Bourne Developments on Louth Road, Wragby; this conforms to both the size and tooling marks of similar Ancaster-made coffins known in the Lincolnshire area. Its provenance is currently being researched.

The artefactual concentrations encountered within the fieldwalking survey indicated that each of the three potential Roman structures was mainly occupied from the second to the fourth century a.d., with two of the sites producing small quantities of late Iron Age and first-century pottery. All three of the sites have produced material suggesting the presence of hypocausts, with one of the sites possibly containing tessellated or mosaic floors, based on the finds of individual tesserae.

The first site (TF 131 783) is situated to the south of an adjacent field that was reported on last year.Footnote ¹⁰⁷ The new building is located on a level area on an east-facing slope within a shallow valley to the west of the B1202 Market Rasen Road. The scatter of building debris suggests a small building, possibly associated with the potential courtyard complex identified to the north. The vast majority of the artefactual material is clay roof-tile with some hypocaust-tile fragments. The abundant pottery is mainly utilitarian local coarse wares with a small quantity of imported and regional fine wares.

The second site (TF 131 793) is situated on a slight southern slope within the northern limits of the parish. It is bisected by a farm track that facilitates access to Barn Farm from the B1202 Market Rasen Road. The scatter of building debris and pottery indicates a building covering an area of some 45 m by 30 m, with the presence of hypocaust material and a single painted wall-plaster fragment. The pottery recovered from this area is dominated by local grey wares with the occasional fragment of regional and imported fine wares and a Southern Spanish amphora sherd.

The third site (TF 143 786) lies on a gentle south-eastern slope to the north-east of the village, north of Wire Hill Lane. The scatter of building material indicates a single building covering an area of 70 m by 40 m with several dense scatters of pottery and tile to the south and north-east that may derive from further structures. The pottery consists of local grey wares with occasional regional fine ware fragments.Footnote ¹⁰⁸

NOTTINGHAMSHIRE

(1) Clifton, A453 Widening Scheme M1 Junction 24 to A52 Site 28 (SK 5441 3312): excavation following geophysical survey and evaluation examined a late Iron Age enclosure and several subsequent phases of Romano-British occupation.Footnote ¹⁰⁹ The primary enclosure was over 70 m wide. A shallow and truncated crouched inhumation within the enclosure probably dated from the earliest phases of the site. In the second century a.d. the northern arm of the enclosure ditch was infilled, and a large stone building constructed within a smaller enclosure. The building was 12 m wide and over 16 m long and was defined by sections of stone foundations with post-sockets, beamslots and large post-holes. A post-hole or small pit within the building contained a deliberately placed grey ware vessel. A possible cellared workshop lay outside the structure; a quernstone was found nearby and metalworking slag was recovered from ditches in this area. Additional enclosures and paddocks were added in the second century but the main building had gone out of use by the third century a.d. Large amounts of building material came from later pits and robbed out post-holes; pottery within these pits suggests continued occupation elsewhere on the site. A second burial, of a young adult female, was located within the entrance to a later enclosure.

DERBYSHIRE

(1) Smalley, Lodge House (SK 4173 4497): an evaluation comprised eight trial trenches spread over two fields. Previous geophysical survey over the western half of the development area identified possible enclosures and a possible ditched trackway in the north of the site; features in the south included a group of ditch-type anomalies which did not correspond with known post-enclosure field boundaries. In Field 1 a large 4.5 m-wide enclosure ditch and an adjacent smaller enclosure corresponded with features identified by the geophysical survey. Both enclosure ditches contained large unabraded fragments of Roman pottery. The enclosures were broadly contemporary and rural domestic settlement was probably located within the smaller enclosure. There was no evidence of a ditched trackway in the north of the site and no Roman features were identified in Field 2 to the south.Footnote ¹¹⁰

STAFFORDSHIRE

(1) Stoke on Trent area: (a): thirteen AR denarii deposited c. a.d. 160–1 or later.Footnote ¹¹¹ The earliest coin is of Vespasian, the latest closely dated coin is of Antoninus Pius TR P XXIII (a.d. 159–60) but also included are two coins minted for Faustina II which could have been in production at the very end of the reign. (b) About 261 AR denarii and fragments, deposited c. a.d. 161–76.Footnote ¹¹² It was difficult to estimate precisely the total number of coins in the hoard due to the degree of fragmentation. Possible structural remains in the area were subsequently investigated.Footnote ¹¹³ These are thought to consist of part of a collapsed boundary wall.

SHROPSHIRE

(1) Wroxeter and Uppington, Shrewsbury Resilience Scheme (SJ 568 094–539 089): a detailed gradiometer survey was conducted on a 30 m-wide corridor centred upon the route of the new water main, some 4.2 km long, between Uckington pumping station and Atcham, passing c. 1 km north of the Roman town of Wroxeter (Viroconium Cornoviorum). The route crosses a number of possible archaeological sites identified from aerial photography, including possible marching camps and a Roman cemetery. Towards the eastern end of the survey area, a number of linear and pit-like anomalies were identified within a quiet magnetic background. West of Norton, several ditch-like anomalies appear in close proximity to a possible Roman marching camp, although the association between the geophysical anomalies and cropmarks has not been proven. Good agreement is seen between the geophysical survey and cropmark evidence east of Tern Lodge, where rectilinear ditches and a circular region of increased magnetic response were identified. South-west of Tern Bridge, anomalies including a network of rectilinear ditches were identified close to a possible Roman cemetery and pottery production site.Footnote ¹¹⁴

HEREFORDSHIRE

(1) Leintwardine: 18 Republican and early Imperial AR denarii deposited a.d. 58–9 or later.Footnote ¹¹⁵

WARWICKSHIRE

(1) Barford, Westham Lane (SP 2679 6057): evaluation subsequent to geophysical survey identified several ditched enclosures, mostly aligned north-west–south-east, with some aligned east–west, and mostly concentrated in the eastern part of the site. The few finds suggest that these are likely to represent agricultural enclosures and a possible trackway, with settlement located nearby but not within the area examined.Footnote ¹¹⁶

(2) Bedworth: two uninscribed North Eastern Iron Age gold staters and one North Eastern stater inscribed ‘VEP’, deposited a.d. 20–50 or later, are additional to two earlier (1994 and 2011) finds, bringing the total number of individual Iron Age coins found from this hoard to 16.Footnote ¹¹⁷

(3) Kineton, land east of Southam Road (SP 33971 51394): an evaluation identified a concentration of ditches, pits and post-holes, many containing late Iron Age and early Roman pottery and probably forming part of a settlement previously recorded as a cropmark.Footnote ¹¹⁸

(4) Shilton: six AR denarii fused together (including one of Trajan and two of Hadrian), deposited a.d. 117–38 or later.Footnote ¹¹⁹

(5) Shipston-on-Stour, Campden Road (SP 2495 4050): an evaluation identified a possible occupation layer of early Roman date. This was cut by Roman ditches and pits, potentially the remains of a rural settlement. The finds included ceramic building material, fired clay, animal bone and ironworking slag.Footnote ¹²⁰

(6) South Warwickshire: four Roman Republican and one Roman Imperial denarius of Claudius I are additional finds from a hoard dated c. a.d. 63–4.Footnote ¹²¹

(7) Willoughby, Barby Pools Marina (SP 5184 6976): an evaluation recorded the edge of a Roman farming settlement; the core of the settlement, previously identified by geophysical survey, lies beyond the site. The remains comprised ditches, a row of six pits, and a probable roundhouse. Pottery from these features dates to the first to third centuries a.d.Footnote ¹²²

LEICESTERSHIRE

(1) Burrough on the Hill, Burrough Hill hillfort (SK 7605 1195): the third season of excavation by the University of Leicester examined three areas of the hillfort interior (Trenches 5, 6 and 7, of which only the first contained Roman features), targeting anomalies revealed during the earlier geophysical survey. In Trench 5 a spread of pits around 50 m inside the hillfort entrance dated from the Iron Age to early Roman period. The Iron Age pits were deep rock-cut features, perhaps for storage. The presence of Roman pits was unexpected, as hitherto activity dating to the Roman period was restricted to the northern part of the site. The pits produced a wide array of finds including large pottery and bone assemblages and numerous small finds, including a complete loomweight, querns, metalwork, spindle whorls and a decorated blue glass bead. One of the early Roman pits also revealed evidence for a human burial, containing a skull, part of an arm and a rib, as well as a quarter of a rotary quern (lying above the individual's head), an inverted horse skull, a copper-alloy penannular brooch and remains of a copper-alloy bowl.Footnote ¹²³

(2) Desford, Hunts Lane (SK 471 032): evaluation failed to locate evidence for possible enclosures suggested by aerial photographs and geophysical survey. A single abraded sherd of Roman pottery was recovered from medieval ridge and furrow.Footnote ¹²⁴

(3) Glenfield, Glenfield Park (SK 5329 0535): an evaluation within the Eastern Employment Area identified four gullies and two pits dating to the Late Iron Age to Roman transitional period. Within the Western Employment Area, evaluation identified Roman ditches and pits; fragments of kiln bars and wasters from the site indicate that pottery production was occurring nearby.Footnote ¹²⁵

(4) Hungarton, Village Farm, Baggrave Road (SK 684 076): a strip, plan and sample excavation was undertaken in a field to the west of the farm. Features revealed within the area of a wind turbine base included a substantial Roman ditch broadly attributable to the second to fourth centuries a.d. Post-dating this ditch was a structural slot or gully possibly representing a late Roman timber building. A possible post-pad may have formed part of the same structure. A small collection of Roman pottery and tile was recovered from these features. An undated archaeological feature was observed within the electricity cable trench linking the wind turbine base to Village Farm. Scatters of Roman pottery have previously been found in the vicinity of Village Farm (though not in this field) and a large collection of Roman material, including pottery, animal bone, lead casting waste, a third-century coin and a bronze strip, was found during the excavation of a trench within the farmyard in 2010.Footnote ¹²⁶

(5) Ibstock, land at Ravenstone Road (SK 405 112): evaluation identified a pit alignment of possible Iron Age date and two undated ditches, perhaps part of a ditch system associated with a known Roman settlement to the north of the site. Only a single sherd of Iron Age pottery was recovered.Footnote ¹²⁷

(6) Leicester (Ratae): (a) junction of Highcross Street and Vaughan Way (SK 58309 04771 centre): evaluation revealed Roman features in all but one of seven trenches examined, with features ranging in date from the late first century through to the fourth century a.d. Along the northern side of the site, compacted gravel street metalling for the east–west Roman street running between Insulae IV and X was recorded in the western half of Trench 1 and the northern end of Trench 7. Early Roman features either pre-dating this street or contemporary with its early use were found in Trenches 1 and 4. These appear to date to the late first century or possibly the early second century a.d. In Trench 1 a beam-slot/gully appeared to be at right-angles to the street; whilst in Trench 4 early occupational trample and yard surfaces were of broadly contemporary date. Later Roman activity next to the street was recorded in Trench 7. Here a narrow gully running parallel with the street, post-holes and the robber trench for a road-side wall were all dug into soil containing mid- to late second-century pottery. Activity continued into the third century and a substantial concrete floor was also present lying on soil containing mid- to late third-century pottery. The noticeable absence of Roman roof-tiles in this trench may suggest that the floor was some sort of heavy-duty external surface and the wall a boundary wall surrounding a property rather than part of a building.

Further south in the centre of Insula X, was evidence for timber and masonry buildings of uncertain plan. The timber buildings, found in Trenches 2 and 5, survived as beam-slots and floor surfaces; these were a mixture of earth and concrete, each floor often separated by a layer of made ground containing pottery dating to the latter half of the second century, redeposited roof-tiles and painted wall-plaster. Robbed wall footings for masonry walls were found in Trenches 2, 3, 4 and possibly 5. In Trench 3, the corner of a building was present; whilst in Trench 4 in-situ floors and robbed walls suggest a sizeable Roman building. Further west, closer to the line of the north–south street leading to the town's north gate, the remains of the early timber building in Trench 5 were sealed beneath a substantial mortar and stone floor, perhaps part of a later building. Several Roman pits in Trenches 3 and 5 contained late second-century to late third-century material. The largest fragment of mosaic pavement found in Leicester in recent years was located in Trench 2. The panel, of which approximately a quarter survives, measures c. 3.2 m by c. 2.6 m. Its pattern, picked out in red and grey tesserae, includes a hexafoil central motif surrounded by an octagonal band, heart-shaped leaves and swastika-meander. The pavement may belong to the same building as a robber trench and floor make-up in this trench. The robber trench, on the same orientation as the mosaic and the town's street grid, appeared to be an external wall with yard surfaces present beyond it to the east. The floor make-up contained early fourth-century pottery.Footnote ¹²⁸

(b) 40–46 Western Road (SK 578 038): further work on the inhumation cemetery found on the former Equity Shoe site in 2010 completed the removal of the remaining skeletons, bringing the total assemblage to 58. This is probably still only a minute percentage of the cemetery population in the vicinity which, based on nineteenth-century records, must extend over a large area between the river and Narborough Road and the old line of the Fosse Way. Radiocarbon dating of six of the burials from the 2010 evaluation has produced calibrated dates (at 95% confidence) of a.d. 50–140, a.d. 80–240, a.d. 170–340, a.d. 230–340, a.d. 230–390 and a.d. 250–410, which cover a broader period than originally suspected. Evidence for the edge of a paleochannel found in 2010 confirms that the inhumations are restricted to rising ground above the river edge in the site's south-western corner. The only other archaeological feature of note in the cemetery was a substantial pit of uncertain function, possibly Roman in origin, which had been dug through some of the burials. Burial traditions in the assemblage are diverse, with alignments varying from north–south through to east–west; the latter are typically later than the former. Graves appear to be scattered haphazardly across the cemetery, clustering in small concentrations which might represent family groups. Many were buried in coffins, some with grave goods, including pottery, hobnailed shoes, jewellery (beads, brooches, bangles), and one with a small glass vessel. One grave contained three pottery vessels, four bangles and a glass hairpin, all placed beside the person's right leg whilst they were also wearing hobnailed shoes and a chain-link necklace. Seven of the burials were prone and one had been decapitated.Footnote ¹²⁹

(7) Lutterworth, land at Leaders Farm (SP 5302 8423): evaluation revealed significant evidence of Iron Age and Roman date, consisting of a probable roundhouse, enclosures, and ditches.Footnote ¹³⁰

(8) Market Harborough, Glebe Road (SP 744 867): an evaluation and subsequent excavation revealed Roman enclosures and a droveway with a date range from the late first century a.d. and continuing until the fourth century, lying at the western margin of a probable substantial field-system. Two inhumation burials were recorded, including a decapitation burial, probably interred in the fourth century. During the sixth or seventh centuries three isolated pits indicated Saxon settlement within the vicinity.Footnote ¹³¹

(9) Thorpe-by-Water, Tudor House, Main Street (SP 893 964): a watching-brief revealed a ditch containing pottery ranging in date from the mid-first to early fourth centuries, and two post-holes, within the area of a new driveway. Within the area of the new extension on the west side of the house was a shallow Roman gully.Footnote ¹³²

NORTHAMPTONSHIRE

(1) Crick, land off the A5/A428 Junction 18 of the M1 (SP 5706 7304): a pit containing charcoal and fired-clay hearth or oven furniture was recorded. On typological grounds, the fired clay is of probable Iron Age or Romano-British date.Footnote ¹³³

(2) Daventry: (a) DIRFT II, Zone 3 (SP 5639 7370): an evaluation recorded field boundary ditches of probable Iron Age/Roman date. One ditch defined an area of possible livestock enclosures/paddocks and a small number of possible shallow pits and post-holes. The paucity of finds from these features, combined with their low-lying, stream-side setting, suggests that they were peripheral to the main focus of occupation and were probably agricultural.Footnote ¹³⁴ (b) Land at Site 10 Middlemore Farm (SP 5643 6511): evaluation adjacent to a known Roman settlement produced no further evidence related to it.Footnote ¹³⁵

(3) Deanshanger, Stratford Road (SP 7693 3977): excavation on land north of Stratford Road revealed a trackway and ditches dating from the late first century to early second century a.d., contemporary with enclosures recorded to the south. They went out of use in the early second century, possibly as part of land reorganisation with the establishment of the villa to the south.Footnote ¹³⁶

(4) Earls Barton, 9 High Street (SP 8523 6395): evaluation identified a Roman settlement, possibly a small farmstead. The remains comprised enclosures, pits, a keyhole-shaped oven, an occupation layer, and a hearth. Pottery mostly dated to the second and third centuries a.d., with some later Roman fabrics also present.Footnote ¹³⁷

(5) Irchester: (a) Victoria Park (SP 910 668): a detailed watching-brief and targeted excavation were carried out across an extensive area to the west of the Roman walled town of Irchester ahead of the commercial development of Victoria Business Park, during 2004–2005. Iron Age activity included two ditched enclosures, one of which was probably a small farmstead. The Roman remains were more extensive, with the investigation of the south-western corner of the Roman town revealing substantial remains of the town's defensive stone wall and three defensive ditches. The partial exposure of the Roman town's western suburban occupation revealed the remains of buildings sited adjacent to a metalled road. Other features such as ditches, pits and wells were identified, dating to the first to third centuries.Footnote ¹³⁸

(b) Roman Town Research Excavation (SP 919 668): two trenches were excavated within the area of the Roman walled town, where magnetometer survey in 2010 had provided a detailed plan of the internal arrangement of the town ( fig. 10). The trenches were targeted on specific areas to provide information to inform future management strategies. Along the line of the northern defences the intention was to establish the impact of recent ploughing and the potential damage caused by recent illegal grazing by horses. The second area was the Romano-Celtic square shrine first uncovered by Baker in the late nineteenth century, to characterise its current state of preservation and the nature of the antiquarian excavations.

FIG. 10.

Irchester Roman town, magnetometer survey of the walled town, showing roads and buildings.

On the northern defences, the Roman and earlier levels were found to be deeply buried and had not been damaged. A sandy subsoil was cut by a ditch that may have defined part of a land parcel pre-dating the erection of the town defences, and there were also remains of two structures. The remains of one wall were cut through indicating that the defences disregarded previous alignments. This was reminiscent of the situation within the southern defences, where a building lying at an angle to the line of the wall was apparently also truncated. The pottery dates these remains to the first and second centuries a.d., with hardly any material of third-century date. The apparent hiatus of activity after the second century is of potential significance as this is when the defences would have occupied this area.

Part of the ground plan of the Romano-Celtic square shrine was revealed ( fig. 11). This demonstrated that neither Baker's excavations nor subsequent arable cultivation have done substantial damage to the remains of the shrine. It was also evident that Baker had adopted a style of wall chasing whereby the upper surface and faces of masonry walls were exposed. In most instances his excavations had not penetrated more than a single course below the present surface, although in places the entire surviving face appears to have been exposed. It was also shown that, in keeping with many shrines, the whole interior had been raised to create a podium.Footnote ¹³⁹

FIG. 11.

Irchester Romano-Celtic temple, showing the inner cella and the outer ambulatory walls, left, and the precinct wall, with a gate pier, and the metalled road, right.

(6) Kettering, land east of Kettering, Phase A (SP 904 775): evaluation comprised the excavation of 253 trenches over c. 205 ha. The results of an earlier geophysical survey were used to inform the trench locations. The evaluation identified eight distinct areas of activity dated to the later Iron Age and earlier Roman periods. Six areas of predominantly later Iron Age date appeared to represent small farmsteads. Finds from these sites comprised largely pottery and animal bone. The two remaining areas differed in character and dated to the earlier Roman period. One comprised a large number of ditches, post-holes and pits from which a substantial assemblage of pottery, comprising largely storage jars and kitchen wares, was recovered. This site seems to be a farmstead, on a larger scale than the Iron Age examples already noted. In the second area of Roman activity no structural features were identified, but a substantial pottery assemblage, comprising fine wares and kitchen wares, was found within ditches. In addition, a sunken-featured building containing pottery of fifth- to ninth-century a.d. date was excavated; an early Saxon iron bell was recovered from this structure.Footnote ¹⁴⁰

(7) Northampton, Cherry Orchard School (SP 7849 6232): an excavation was carried out at the former Cherry Orchard School in the area where a well-preserved Roman pottery kiln, dated to the first century a.d., was excavated prior to the construction of the school in the 1950s.Footnote ¹⁴¹ Although much of rest of the site was quarried in the nineteenth and twentieth centuries, further Roman remains were identified. Two parallel boundary or enclosure ditches lay close to the kiln. The southern ditch contained large quantities of kiln-bar fragments. To the west were two further parallel ditches with an entrance through the western ditch. It is possible that the two ditch groups formed a sub-rectangular enclosure, or a field-system. The ditches broadly date from the mid-first century a.d. into the mid/late second century. A grave dug into one of the earlier boundary ditches contained the remains of a middle-aged male.Footnote ¹⁴²

(8) Towcester: (a) Moat Lane Regeneration (SP 6931 4884): an evaluation identified the stone foundations of a possible building, a robbed-out wall, a possible yard surface, an inhumation burial and a ditch. All of these lay to the rear of existing properties fronting Watling Street. Tentative evidence was also found for a large ditch close to the conjectured line of the town's Roman defences. The finds assemblage included second- to fourth-century a.d. pottery, animal bone and part of a copper-alloy bracelet.Footnote ¹⁴³

(b) Belle Baulk (SP 6844 4884): two main concentrations of features were excavated in advance of residential development on the west side of Towcester, outside the Roman town defences. The smaller area comprised pits of middle Bronze Age date. In the larger excavation area there were three sides of a double-ditched rectangular enclosure of Roman date. The ditches drained down slope towards the River Tove. Within the enclosure was a rectangular building, 20 m long by 8 m wide, defined by 27 large post-holes or post-pits, with a lean-to portico structure along the north side ( fig. 12). A large sub-rectangular timber and clay-lined holding-tank was associated with the building, and drained into the surrounding enclosure ditch. The pottery assemblage was of mid-first- to mid-second-century a.d. date, containing a greater proportion of flagons and jars than other pottery forms. The building and its tank were perhaps part of an industrial process such as brewing or fulling cloth, although it may have previously been a simple barn.Footnote ¹⁴⁴

FIG. 12.

Towcester, Belle Baulk, showing the double-ditched enclosure, the post-hole building and the square feature.

(c) Old Greens Norton Road, Central Woolgrowers (SP 6872 4935): investigation of a small enclosure immediately behind the former Roman Watling Street frontage, to the north of the Roman town, indicated that the frontage had been lost to modern roadside landscaping. The enclosure was roughly rectangular, formed by a subdivision of longer continuous ditches that lay perpendicular to Watling Street. Within the enclosure lay a possible ancillary yard building. Pits in the immediate vicinity contained fairly large unabraded pottery sherds in utilitarian forms, dating predominantly from the late first to second centuries a.d. Soil samples highlighted the presence of coal and the probable mixing of industrial or craft waste with other deposits that included disarticulated human bone from a ditch fill. A possible rear boundary plot ditch was identified, and immediately west of it were two cremation burials in urns, roughly contemporary with the disuse of the enclosure.Footnote ¹⁴⁵

(9) Weston by Welland: six North Eastern Iron Age silver half units inscribed VOLISIOS DVMNOVE[LLAVNOS], deposited a.d. 20/30–50 or later. All of the coins appear to have been struck with the same obverse and reverse dies. This suggests that they probably remained together as part of the same group from the time of their production until they were lost or deliberately buried.Footnote ¹⁴⁶

CAMBRIDGESHIRE

(1) Alconbury, Enterprise Zone Enabling Works, Alconbury Airfield (TL 196 770): evaluation located a very dense area of features (pits and/or ditches) dating up to at least the end of the second century a.d. Further Roman remains were found between two other latest Iron Age domestic areas revealed by the evaluation.Footnote ¹⁴⁷

(2) Duxford, land East of Moorfield Road (TL 4820 4703): an evaluation revealed a series of late Iron Age/early Roman ditches and pits indicative of boundary enclosures within a small, rural settlement. The range and variety of the ceramic assemblage suggested the presence of a dwelling, probably a roundhouse, in the immediate area.Footnote ¹⁴⁸

(3) Godmanchester, Godmanchester Primary School (TL 24449 70821): subsequent to a geophysical survey, a community excavation opened up three trenches on the supposed route of Ermine Street. Roman rubbish pits and post-holes were found, as well as traces of a Roman gravel surface which may be associated either with Ermine Street or the Via Devana.Footnote ¹⁴⁹

(4) Huntingdon district: a hoard of 139 base metal radiates deposited a.d. 274 or later.Footnote ¹⁵⁰

(5) Willingham, Spong Drove (TL 41765 71310): a watching-brief on excavation of trenches and topsoil stripping identified three discrete spreads of archaeological material lying within the subsoil horizon along the line of an access road within an area identified previously as containing remains of Iron Age and Roman settlement. The spreads appeared to be areas of plough disturbance relating to buried archaeological features and contained late Iron Age to Roman pottery.Footnote ¹⁵¹

(6) Yaxley, Library, No. 266 Broadway (TL 1758 9233): a small excavation found a substantial early Roman ditch. Its fills contained a large quantity of pottery kiln-bars and other kiln furniture from a pottery production site, along with a relatively unabraded pottery assemblage consisting of some 400 sherds dating to c. a.d. 70–85. These constitute a potentially regionally important Roman pottery assemblage. The group includes most of a Roman cheese-press.Footnote ¹⁵²

BEDFORDSHIRE

(1) Great Denham, Biddenham Loop (TL 0210 4815): approximately 1.8 ha of open area excavation was undertaken in advance of the construction of a sports complex.Footnote ¹⁵³ Evidence for two trackways, defined by parallel ditches 11 m apart, and a ditched enclosure was found. These are located to the north of a known and partly excavated farmstead.Footnote ¹⁵⁴

(2) Sandy, Saddlers, 7 Swaden (TL 18204 49059): excavations covering c. 280 m², undertaken in advance of construction of a house extension,Footnote ¹⁵⁵ revealed a substantial late Bronze Age/early Iron Age ditch and second- to third-century Roman settlement remains, comprising pits, ditches, structural slots and an urned cremation burial. The site lies c. 500 m north of the Roman town of Sandy and probably represents part of a farmstead situated on the Baldock to Godmanchester road. No evidence for this route way, which it had been suggested ran through the site, was revealed.

(3) Shefford, Shefford Lower School (TL 0903 4624): a programme of observation, investigation and recording was undertaken during construction works.Footnote ¹⁵⁶ The presence of Roman quarry pits and a boundary ditch, along with the small quantity of finds and the absence of features to their west, suggests that the edge of the settlement containing a substantial aisled building has now been identified.Footnote ¹⁵⁷

HERTFORDSHIRE

(1) Bishops Stortford North (TL 480 226): evaluation of a site of approximately 148 ha was divided into 20 areas delineated by existing field boundaries. Sixty trenches were targeted on the results of a previous geophysical survey. Potentially significant remains were identified in four of the areas examined. These included a large late Bronze Age to early Roman occupation site in Area 13. Elsewhere were features representing a series of prehistoric or later field-systems. The majority of dated features were assigned to the late Bronze Age or late Iron Age periods.Footnote ¹⁵⁸

(2) Braughing, land at Pentlows Farm (TL 3965 2530): a few ditches, gullies and pits were identified in five of 13 evaluation trenches, with Roman dating material in two of these trenches.Footnote ¹⁵⁹

(3) Bushey, 23 Glencoe Road (TQ 13086 95350): a watching-brief and evaluation recovered Roman building material in medieval and later features. Although the material is residual, it appears to indicate a phase of occupation during the Roman period either on the site or in its immediate vicinity, adding to what is currently a very small body of data for Roman occupation in the area.Footnote ¹⁶⁰

(4) Sandridge: a hoard of 159 AV solidi, deposited c. a.d. 408 or later, comprising issues of Gratian (a.d. 367–83), 1; Theodosius (a.d. 379–95), 7; Valentinian II (a.d. 375–92), 11; Arcadius (a.d. 383–408), 4; Honorius (a.d. 393–423), 98; and Arcadius (a.d. 383–408), 38. No Roman pottery was recovered, suggesting an organic container or no container at all.Footnote ¹⁶¹

BUCKINGHAMSHIRE

(1) Weston Turville (SP 85487 12328): an earth resistance survey covering some 2.7 ha was targeted upon a rectangular enclosure identified in a recent gradiometer survey of an area c. 500 m south of the line of Akeman Street. The plan suggested that the enclosure was of Roman date. In addition to the main enclosure the second survey identified several other rectangular sub-enclosures on a similar orientation to the outer enclosing ditch.Footnote ¹⁶²

MILTON KEYNES

(1) Castlethorpe, South Street, Maltings Farm (SP 8015 4435): a watching-brief revealed part of a Romano-British farmstead and associated enclosures. Features comprised pits, an oven, the footings of a postulated barn, identified from stone pads, small enclosures and a large north–south-aligned ditch on the east side of the settlement.Footnote ¹⁶³

OXFORDSHIRE

(1) Abingdon: (a) Abingdon Abbey Shopping Centre, Queens Street (SU 497 971): an evaluation and subsequent excavation of the site revealed Roman pits. The site was located within the Iron Age oppidum, but the absence of prehistoric evidence may be due to the presence of a bank in this area, which may have been levelled during the Roman period.Footnote ¹⁶⁴

(b) Thornhill Walk (SU 4927 9803): a watching-brief revealed features and finds indicating occupation during the second century a.d. Lower layers, perhaps related to a shallow Roman quarry pit, contained a mix of prehistoric and Roman pottery. A post-hole and a pit were identified as Roman. Other gullies and post-holes across the site are either Roman or undated. Three parallel Roman walls on a north-west/south-east alignment were uncovered; two probably formed the sides of a small building, which may have had a wooden front and back walls, but modern disturbance makes this difficult to prove. Between the two walls a rubble spread may have formed a floor surface, on which were iron nails and slag, indicating possible use of the building as a workshop. A larger wall to the north was not straight and was therefore perhaps a property boundary wall. A possible make-up layer for a yard surface was seen on its north side. The site was seemingly abandoned during the third century a.d.Footnote ¹⁶⁵

(2) Ambrosden, Springfield Farm (SP 6070 1940): an evaluation consisting of 12 trenches revealed a small number of ditches in the north-east corner of the site, one of which contained Romano-British pottery. These ditches all followed a broadly south-west to north-east alignment. There were very few finds, suggesting that this was part of a field-system rather than being close to a settlement focus.Footnote ¹⁶⁶

(3) Banbury, land at Southam Road (SP 455 431): two areas were evaluated. East of Southam Road (SP 4585 4311) a range of features was identified but most were undated; finds included some Roman pottery. West of Southam Road (SP 4526 4305) none of the identified features was dated.Footnote ¹⁶⁷

(4) Burcot, Lynden Close (SU 5644 9593): a late Iron Age/early Roman ditch on an approximate north–south alignment was recorded traversing the length of a new basement.Footnote ¹⁶⁸

(5) Cholsey, 34 Honey Lane (SU 58822 86280): a watching-brief was conducted during ground reduction in advance of development. The principal north–south Roman road through the county from Dorchester to Silchester was uncovered beneath the former house during the removal of concrete footings. A chalky bedding deposit supported the road agger, and sections showing the agger and road surface metalling along with a probable roadside ditch were recorded. Other possible ditches either side of the road may indicate the shifting alignment of the road over a long period of time.Footnote ¹⁶⁹

(6) Didcot, Great Western Park (SU 5079 8995): excavations, which began in 2010, uncovered a large Iron Age settlement comprising up to 60 roundhouses. Hundreds of grain storage pits, several quarries, human burials, domestic rubbish and dumps of pottery and animal bones were recorded. Evidence of a later settlement, dating to the late Iron Age and Roman period, was also uncovered. Features included enclosure ditches, several wells and six ‘corn-drying ovens’. In the late Roman period, a modest stone-built villa with a hypocaust was constructed. This appeared to deliberately avoid the site of the Iron Age settlement.Footnote ¹⁷⁰

(7) Dorchester-on-Thames, Discovering Dorchester Project (SU 5774 9410): the Dorchester Project training excavation in the allotments, which overlie the south-west quarter of the Roman town, continued in the same 30 by 20 m area as in previous years, with the main north–south Roman road through the town at its eastern margin. Work concentrated on a number of Roman ditches and related deposits. The Roman road was also partly sectioned, revealing a number of superimposed surfaces. It is possible that there was a roadside ditch on the west side of the road from early in the period, although if so this was sealed by later layers. A fragment of mortar floor suggests that there may also have been at least one roadside building in the same general area. Further west, the east frontage of a building at the west end of the site, some 25 m from the road line, was at least c. 10 m long. This building probably lay within an enclosure defined on its eastern side by a palisade set in trenches which extended north and south beyond the edges of the site. A gap in the palisade trenches aligns with the mid-point of the east wall of the building, suggesting that there was an entrance at that point. The palisade trench, possibly of second-century date, was later recut as a boundary ditch, retaining the entrance opening.

Later ditches transformed the use of the area. A ditch lying alongside the Roman road in the south-east corner of the site curved away from it to the north-west, eventually running beyond the edge of the excavated area in a westerly direction through the opening in the north–south palisaded/ditched boundary and cutting the associated building. This curvilinear ditch, perhaps of later fourth-century date, had a small entrance in its northern side, between the road and the now disused enclosure/building complex to the west. What it itself enclosed is uncertain, although it must have been more than just the extensive midden deposits located in the southern part of the site, but any potential high status settlement components would have lain south of the excavated area. Two more east–west ditches were found running close to the north edge of the site. These were deeper and more steep-sided than the curvilinear ditch, but were of roughly similar late Roman date. Their function is uncertain, but if they too were enclosure ditches they would have related to an area north of the site.

The curvilinear ditch and the adjacent palisade ditch terminal to the south were cut by a pair of pits. One of these was perhaps very late Roman in date, and contained part of a decorated bone comb along with other finds, while the second pit produced early Anglo-Saxon pottery. Most notable amongst the large quantities of finds was a fragment from a late Roman pottery vessel decorated with figures of deities in niches.Footnote ¹⁷¹

(8) Eynsham, Station Road, Abbey Barns (SP 4318 0913): a watching-brief encountered possible Roman features, including at least one pit. Residual Roman pottery was recovered.Footnote ¹⁷²

(9) Faringdon, Park Road (SU 2910 9450): evaluation indicated the presence of Roman settlement spanning the late first to early fourth centuries a.d. focused on the sandy capping of the Corallian ridge deposits, with no evidence of any activity on the lower lying clay. Roman features comprised ditches, including a possible droveway or path, on two postulated alignments. WNW–ESE ditches may represent the earlier system, but no perpendicular features were dated. The later system comprises ditches dating from the mid-third century onwards aligned north-east/south-west and north-west/south-east.Footnote ¹⁷³

(10) Marcham, Anson Field (SU 4579 9675): two phases of evaluation were carried out subsequent to geophysical survey. Bronze Age features and an extensive Iron Age settlement were identified and a little late Roman pottery was recovered.Footnote ¹⁷⁴

(11) South Leigh, Gill Mill Quarry (centred SP 375 077): towards the south-east end of the quarry two separate parts of quarry area Tar Farm 7, together covering almost 2 ha, were examined. In the south-east area (SP 3870 0655) the fills of an undated palaeochannel were cut by ditches defining a trackway aligned NNW–SSE, linking at its northern end to WSW–ENE-aligned boundaries recorded previously. These features were probably of second-century date. The only other features in this area were three pits, probably contemporary with the use of the trackway. To the west (SP 3855 0660) an open area contained a penannular gully and an adjacent L-shaped gully, close to which were some of a scattered group of 15 cremation burials. Two of the latter were contained in vessels of early Roman date, and several of the others produced groups of small nails, also suggesting a Roman date, but were otherwise unaccompanied. Such early Roman cremation burials are unusual in a rural context in this period. To the north-west the adjacent quarry area (Tar Farm 8, centred SP 3840 0675, also roughly 2 ha in extent) was stripped under archaeological supervision and planned. Excavation had only just commenced when curtailed by flooding. Features revealed include parts of the complex of ditches defining a road running WNW–ESE from the centre of the large Roman settlement at Gill Mill to the west; parts of this road, probably established in the first half of the second century a.d., had been previously recorded in areas both east and west of Tar Farm 8. Plots both north and south of the road were defined by ditches running perpendicular to its alignment. One of the former appears to contain gullies defining a round building of more than one phase, while on the south side of the road a patch of stone may be a platform for another building. South-east of the latter the roadside plots extending towards the River Windrush contain very few Roman features.Footnote ¹⁷⁵

(12) South Oxfordshire: five base silver denarii and 11 radiates, deposited after c. a.d. 271.Footnote ¹⁷⁶

(13) Wantage, Stockham House (SU 392 888): twenty-nine geotechnical test pits were monitored. One feature of archaeological interest was observed: an east–west-aligned ditch dated by pottery to the late Iron Age or early Roman period. A subsequent evaluation uncovered a further ditch, which contained a single sherd of first- to second-century pottery.Footnote ¹⁷⁷

NORFOLK

(1) Deopham: two AR denarii of Caracalla as Caesar under Septimius Severus deposited a.d. 197–8 or later.Footnote ¹⁷⁸

(2) Gressenhall: two AR siliquae of Theodosius and Honorius deposited a.d. 402 or later. Both coins are heavily clipped.Footnote ¹⁷⁹

(3) Hingham (near): an AV solidus of Arcadius, mint of Milan, additional to an original find of 26 solidi and 4 siliquae, deposited a.d. 402 or later.Footnote ¹⁸⁰

(4) Mattishall (near): an Iron Age silver unit (East Anglian Pattern/Horse type inscribed ECEN) and 3 Roman Republican denarii, additional to an earlier find, deposited a.d. 20–50 or later. The original group of 15 Iron Age and 23 Roman coins was certainly from either one scattered hoard or, less likely, from two such scattered hoards. The present group of coins comes from the same deposit(s).Footnote ¹⁸¹

(5) Norwich (near): an AV solidus of Theodosius and 6 AR siliquae, deposited after c. a.d. 402. Most of the silver coins are quite heavily clipped, suggesting a deposition date probably c. a.d. 410.Footnote ¹⁸²

(6) Old Catton, Spixworth Road (TG 238 130): an excavation and watching-brief recorded a series of drainage ditches representing two phases of field-system dating from the late Iron Age to early Roman period. A few isolated Roman pits with charcoal-rich fills were also recorded and interpreted as possible cooking pits. Artefactual evidence suggests that the site had been abandoned by c. a.d. 100.Footnote ¹⁸³

(7) Quidenham: a hoard of 7 AR siliquae and 143 base metal coins to a.d. 402.Footnote ¹⁸⁴

(8) Scottow: thirty-two AE coins from Trajan to Gordian III. The latest coin in the group that can be accurately dated belongs to the years a.d. 238–44. This is very corroded and so its condition when deposited cannot be ascertained. The hoard is certainly very late in date for an aes hoard and could be connected to the recycling of sestertii and sub-denominations in a later period. Whether it was a currency hoard or a scrap hoard, this group of coins was most likely lost or deposited at some time within the third quarter of the third century.Footnote ¹⁸⁵

(9) South Norfolk: fifty-nine AR denarii (Republic, 36; Mark Antony, 2; Augustus, 5; Tiberius, 16) deposited after a.d. 14–37. The latest coins, of Tiberius, show some wear and had plainly circulated for a number of years.Footnote ¹⁸⁶

(10) Swaffham (near): four Iron Age East Anglian silver units: 3 Pattern/Horse types of Anted and 1 plated uninscribed unit of regular Face/Horse type. Deposited a.d. 20–50 or later.Footnote ¹⁸⁷

(11) Tacolneston: two Iron Age East Anglian silver units of Pattern-Horse types, found accreted together. Deposited a.d. 20–50 or later.Footnote ¹⁸⁸

SUFFOLK

(1) Bury St Edmunds (near): (a) three Iron Age East Anglian silver units (1 uninscribed ‘Face/Horse regular’ type, 1 Pattern/Horse type inscribed ECEN, and 1 Pattern/Horse type of Anted), deposited a.d. 20–50 or later.Footnote ¹⁸⁹ (b) Two AR siliquae of Gratian and Arcadius, deposited a.d. 402 or later.Footnote ¹⁹⁰

(2) Eriswell, Sewage Works, RAF Lakenheath (TL 7288 8082): an excavation area of c. 270 square metres adjacent to Caudle Head mere (where spring water surfaces from a buried watercourse which is now controlled) was sampled revealing 1.5 m of accumulated deposits, including peats below occupation soils dating from the first millennium b.c. up to the mid-first millennium a.d. At least seven phases of Roman activity consisted of enclosure or droveway ditches and fence lines, interspersed with dumped soil. Finds included the probable ritual deposit of a horse's head, and a first-century cremation burial.Footnote ¹⁹¹

(3) Exning, Windmill Hill (TL 6287 6589): an evaluation recorded two mid-Roman boundary ditches, possibly part of a field-system or enclosure.Footnote ¹⁹²

(4) Haughley, Chilton Leys (TM 0396 5997): in addition to prehistoric remains, Roman features included a pottery kiln. This was intact from its perforated floor downwards ( fig. 13), with the permanent kiln floor resting on what was probably a tongue support. The kiln has been tentatively dated to the mid-first to early second century. Adjacent to it was a group of clearly associated post-holes that presumably formed an associated structure. Roman material at the northern end of the site probably represents the edge of a second area of occupation with pits, post-holes and a watering hole or well.Footnote ¹⁹³

FIG. 13.

Haughley, Chilton Leys, early Roman pottery kiln.

(5) Long Melford, Long Melford Primary School (TL 8642 4530): excavation on the east side of an extensive settlement or small town was carried out prior to the construction of a new area of playground and two new classrooms. A small number of pits, aligned ditches and post-holes dated from the mid-first to second centuries and included metal-working residues and a crucible fragment. A late first-century cremation in a pot was associated with three further pottery vessels. Three inhumation burials included one with two samian ware dishes and a flagon, deposited in the late second or early third century. Two of the graves were large with possible funerary structures containing the bodies. These add to the evidence for dispersed inhumations of mid-second-century and later date in several areas of the settlement.Footnote ¹⁹⁴

(6) Orford (near): an East Anglian Iron Age silver unit inscribed Anted, deposited c. a.d. 20–50 or later, found within the area of a previously known coin hoard of four Iron Age silver units. An aureus of Nero (RIC 63, struck c. a.d. 66–7) found at the same time is probably related to a separate episode of deposition or loss.Footnote ¹⁹⁵

(7) Withersfield, Hanchet End (TL 647 464): following trial trenching an area of 4.5 ha was excavated, revealing a landscape used from the Iron Age to the late Roman period with a possible Anglo-Saxon building and evidence for a fifth- to sixth-century inhumation burial. The main phases of activity included enclosures, field-systems, a single roundhouse, pits, structural remains and three burials, indicating a largely agricultural landscape.Footnote ¹⁹⁶

(8) Woodbridge (near): eight AR denarii from Nero to Marcus Aurelius are addenda to an earlier find, deposited c. a.d. 172–6 or later. The original find contains the latest coin but the denarius of Nero is now the earliest in the hoard.Footnote ¹⁹⁷

ESSEX

(1) Braintree, 37–45 Clare Road (TL 7515 2276): the site, on the north slope of the valley of the River Brain, is at the open end of a triangular area formed by the convergence of Stane Street (Rayne Road) and the Roman Sudbury to Chelmsford road (London Road). Evaluation and excavation led to the significant discovery of a hitherto unknown metalled track or roadway, at least 5.8 m wide, leading north-east, through cemetery areas, towards the west side of the built-up area of Roman Braintree. However, there were no finds from the gravel, and, given the presence of a roadside ditch in other local Roman roads, the lack of a roadside ditch here is a problem to be addressed by future work. The lack of obvious repair to the track surface may mean it was only in use for a short while.Footnote ¹⁹⁸

(2) Colchester (Camulodunum): (a) 1 Rawstorn Road (TL 99074 25000): a watching-brief at a site in the western extramural suburb of the Roman town, south of the London road which entered the town through the Balkerne Gate, revealed 5 m of a 150 mm-thick gravel surface. The only find from the gravel was a samian ware sherd dated a.d. 120–200. Roman gravel lanes have been discovered during previous excavations close to this site, at both the former St Mary's Hospital site (now called Balkerne Heights) and at Crouch Street. The orientation of the gravel is similar and possibly parallel to that of the main Roman road to London (to the north), and similar to the Roman lane discovered at Crouch Street.Footnote ¹⁹⁹ If the gravel is of Roman date, then at least its upper part can be dated to the second century or later.Footnote ²⁰⁰

(b) 17 West Lodge Road (TL 98546 24806): a watching-brief at a site 700 m west of the walled town, within an extensive Roman cemetery area which flanks the main Roman Colchester to London road, revealed pottery and a disturbed cremation burial, probably of mid-first- to second-century date. A second possible burial was represented by sherds of ‘egg-shell’ ware in a charcoal-rich cut 600 mm below modern ground level. Roman pottery was also recovered from the topsoil.Footnote ²⁰¹

(c) Abbey House, Flagstaff Road (TL 9971 2476): two evaluation trenches dug outside the medieval precinct of St John's Abbey revealed two Roman ditches, three Roman pits and building debris indicative of a nearby substantial Roman building with mortared foundations and tessellated floors. No Roman burials were found in this extramural location.Footnote ²⁰²

(d) Colchester Royal Grammar School, 6 Lexden Road (TL 9868 2484): two evaluation trenches within the extensive cemetery west of the Roman town revealed a north-west/south-east orientated Roman road at a depth of 1.3–1.6 m below modern ground.Footnote ²⁰³ This is probably the Roman road first identified by schoolmaster A.F. Hall in the 1930s. Excavations known to have taken place on the site of the road in the 1950s may explain the presence of two backfilled trenches, one containing a large group of Roman pottery (presumably discarded into the trench with the backfill). Among the large quantity of finds was a piece of worked limestone (probably Purbeck), which might be from the outside of a monument such as the temple-tomb discovered at CRGS in 2006, 90 m south-east of this site.Footnote ²⁰⁴

(e) Garrison Athletics Facility, Abbey Field (TL 9946 2433): two evaluation trenches between the all-weather pitch and the Garrison athletics track, within the large Roman cemetery south-west of Colchester, revealed a ditch containing Roman tile.Footnote ²⁰⁵

(f) Garrison Alienated Land Area S2 (north) (TL 9972 2207): the Berechurch Dyke is the only known defensive earthwork on the eastern side of the oppidum of Camulodunum. Following the recent discovery of a new sector of this dyke at Hyderabad Barracks,Footnote ²⁰⁶ it is now thought that the dyke pre-dates the Roman conquest. As part of the on-going archaeological project accompanying the Colchester garrison redevelopment, an excavation (Site K) revealed the dyke ditch and part of the dyke bank. A line of post-holes was probably part of a timber revetment fronting the dyke bank, although a later date cannot be entirely excluded.Footnote ²⁰⁷

(g) North Primary School, John Harper Street (TL 9919 2584): watching-briefs during foundation work for a new classroom within the area of a Roman suburb containing high-status buildings identified a Roman ditch, and a deposit of demolition debris from a Roman building, probably of second- or third-century date. The debris consisted of crushed painted wall-plaster and tile. No in-situ foundations or other remains were observed, but the debris confirms the nearby presence of Roman buildings.Footnote ²⁰⁸

(h) Sixth Form College, North Hill (TL 9926 2545): a watching-brief on the machine-digging of foundation trenches for new buildings in the north-west corner of the college grounds revealed late Roman demolition debris, including fragments of painted wall-plaster and brick/tile, a late third-century coin, and Roman pottery.Footnote ²⁰⁹ The top of an east–west Roman foundation was also uncovered in two places. These remains perhaps formed part of the same Roman building (Building 212) excavated nearby in 2005–6.Footnote ²¹⁰

(i) ‘The Stockwell’, 18 West Stockwell Street (TL 9945 2542): monitoring during renovation at the former Stockwell Arms revealed the north–south-aligned gravelled street dividing Insulae 3 and 4 of the Roman town.Footnote ²¹¹

(j) Westhouse Farm, Bakers Lane (TL 973 261): an evaluation located a Roman field boundary ditch containing a large fragment of Roman tegula outside and south of an Iron Age enclosure to the west of the Moat Farm Dyke (part of the extensive and complex earthwork defensive system of Camulodunum).Footnote ²¹²

(k) Elmstead Road/Boundary Road, University of Essex (TM 0242 2424): a watching-brief during infrastructure works for the University of Essex's new research park revealed a pit containing Roman brick fragments which are probably derived from a nearby building and pot sherds. Other sherds may be derived from disturbed Roman burials on the lower valley slope, and late Roman sherds indicate some activity further upslope.Footnote ²¹³

(3) Great Dunmow, 46 High Street (TL 6284 2174): monitoring of this site within the area of the Roman small town revealed two areas of compacted gravel (which coincide precisely with the projected course of Roman Stane Street), and a ditch which may be a roadside ditch. Associated pottery dates the gravel to the second century a.d., possibly later.Footnote ²¹⁴

(4) Harlow, land at West Sumners (TL 4286 0706): following on from a fieldwalking and metal-detector survey undertaken in 2010, an evaluation recorded a number of regularly laid out linear features and a large curvilinear enclosure probably related to land division and exploitation in the late Iron Age and early Roman periods.Footnote ²¹⁵

(5) Stanford-Le-Hope, Butts Lane (TQ 6783 8179): a 61 trench evaluation revealed a total of 52 features including ditches, gullies, possible ring gullies and post-holes of prehistoric and Roman date, although pottery from the ditches was mostly abraded and lacked diagnostic types. The ditches and gullies suggest field-systems, with possible settlement activity in the vicinity.Footnote ²¹⁶

(6) Stebbing, Throws Farm (TL 6584 2248): three trial-trenches produced one sherd of Roman grey ware. This was from the trench closest to the line of the Roman Stane Street, immediately south of the site, but there were no archaeological features to indicate roadside settlement.Footnote ²¹⁷

CITY OF LONDON

(1) Aldgate Traffic Lights, Aldgate High Street (junction of St Botolph Street, Jewry Street and Minories), EC3 (TQ 33562 81165): small sections of the Roman City Wall, consisting of ragstone pieces bonded with chalky mortar, were observed in two trenches running north–south across Aldgate Road. Part of the Roman gate of Aldgate was recorded in a trench on the Aldgate roundabout. This consisted of an east–west oriented outer wall (which probably joined the City Wall, although this area has been truncated by modern concrete). The eastern end of the wall returns toward the north, which likely formed the outer wall of the gate. It is formed of an outer face of roughly hewn ragstone blocks encasing a chalk rubble and ragstone core bonded with the same chalky mortar seen earlier. The thickness of the walls (c. 2 m) indicated that Aldgate was part of a substantial defensive structure.Footnote ²¹⁸

(2) 100 Bishopsgate, EC1 (TQ 3243 1372): the main phases of watching-brief uncovered no remains of Roman date due to extensive truncation by the basements of the previous building. Subsequent excavation in an area on the south side of Camomile Street revealed natural brickearth truncated by dense clusters of Roman pits which would have stood in the back-plots of Roman buildings fronting Ermine Street (now Bishopsgate). In addition, evidence was found for industrial processes, including tanning and metal working, on plots which would have backed onto the City Wall or the fringes of Roman London.Footnote ²¹⁹

(3) Bucklersbury House, Queen Victoria Street, EC4 (TQ 32549 81002): natural London Clay was observed in the base of the Walbrook stream and along the stream sides, at levels between 0 and 1.5 m OD. The basal fills of the Walbrook and fluvial deposits were recorded, containing many well preserved metal items including military artefacts. The valley sides on both west and east had been subject to extensive land reclamation throughout the Roman period, and this material was very high in organic content and contained preserved leatherwork ( fig. 14). At the very base of the sequence were wattle fencelines running parallel to the stream, and basic timber buildings with plank floors. Upstanding timber piles suggest the presence of jetties and riverside structures. A human skull was associated with one of these structures. Along the eastern side of the site was a sequence of early timber drainage features and clay and timber buildings above these reclamation deposits ( fig. 15). There had been a fire event sometime in the early second century a.d., thought to be the Hadrianic fire of a.d. 120–5.

FIG. 14.

Bloomberg Place, London, upholstered leather object, probably from a high quality piece of furniture or the interior of a vehicle such as a chariot, with stitched decoration showing a warrior fighting mythical creatures. (© MOLA)

FIG. 15.

Bloomberg Place, London, timber foundation beams from Roman building. (© MOLA)

At the south-west corner of the site on the gravel terrace west of the Walbrook the ground was much higher and drier; here early Roman buildings, land boundaries and other activity were excavated. In the north-east corner of the site in the late first/early second century a.d. a large building was constructed on massive clusters of piles extending through the soft deposits below. This building had a timber plank floor in one room and mortar floors to the south. Corridors were clearly defined, delineated by surviving clay and timber walls. Two areas of polychrome mosaic were excavated. Later Roman evidence in the same area consisted of two large buildings with masonry foundations, separated by an external alleyway. Within the rear of one of these buildings were two ovens. Towards the south of this area was a fourth-century square timber well, which contained a large assemblage of coins and a hoard of pewter vessels ( fig. 16). These features are contemporary with the temple of Mithras which stood 10 m to the south of the site, and artefacts that may relate to the function and use of the temple itself were found.Footnote ²²⁰

FIG. 16.

Bloomberg Place, London, hoard of pewter vessels from fourth-century timber-lined well. (© MOLA).

(4) 100 Cheapside, EC2 (TQ 32445 81195): test pits within the City Tavern public house exposed dark silty clay deposits containing first- to second-century a.d. pottery and pieces of tegulae. These were interpreted as Roman or post-Roman ‘dark earth’ deposits.Footnote ²²¹

(5) Cannon Street, EC4 (TQ 32591 80916): a dropshaft excavated on the south side of Cannon Street adjacent to the junction with Dowgate Hill revealed remains suggesting the presence of an early Roman building with a number of floor surfaces and a possible hearth. A robber trench was also observed. A solid mass of baked brickearth was interpreted as the debris of the building's wall and a substantial phase of burning which may have destroyed the structure could relate to the Hadrianic fire in the a.d. 120s. Horizons of fire debris from this event have previously been observed in Walbrook, NNE of the site.Footnote ²²²

(6) 24–26 Minories, EC3N (TQ 33660 81065): previous work in the Minories had established the presence of an extramural cemetery together with an east–west-aligned road which ran through it. A compact layer of yellow gravel probably associated with this road was observed in two trenches. In the first, on the western side of the site, a sherd of Rhineland white ware, dated a.d. 50–300, was apparently rammed into the surface. Due to truncation the surface only covered an area of 700 mm by 300 mm, although it did continue into the east- and south-facing sections within the trench. A similar but heavily truncated gravel surface was also observed c. 5 m to the east of this trench.Footnote ²²³

(7) Centurion House, 24 Monument Street, EC3 (TQ 32930 80720): work continued from 2011Footnote ²²⁴ in three separate but adjacent areas: Zone 5 to the north-east, Zone 4 immediately south of Zone 5, and a crane base located towards the south-east. Much of the evidence for timber revetments and building construction was an extension of features originally identified in 1981. Late first-century timber waterfront structures were recorded in Zone 4 and the crane-base. A piled revetment supporting a foreshore gravel embankment was the northernmost and earliest timber feature. It was superseded by a partially collapsed east–west revetment situated along the south limit of Zone 4. In the crane base, a more substantial rectangular structure consisting of large stacked timber beams has been interpreted as part of a ‘landing stage’. This in turn was superseded by a stacked-beam quay. Both the landing stage and the quay are known to extend east–west across the breadth of the site. The landing stage was truncated by construction of open-fronted warehousing separated by a clay-lined timber drain. Masonry wall foundations relating to the warehouses and remains of the timber drain were recorded in Zone 4. In Zone 5, later first- and second-century development included building construction on a terrace to the rear of the warehouses. Remains included mortar, timber and opus signinum floors, associated beam-slots and substantial remains of ragstone wall foundations.Footnote ²²⁵

(8) 8–10 Moorgate, 3–4 Kings Arms Yard, 8–10 Telegraph Street and 16–17 Tokenhouse Yard, EC3 (TQ 32708 81346): excavation followed on from work in 2010,Footnote ²²⁶ which consisted of three trenches in Zone 4 and revealed three phases of clay and timber Roman buildings thought to date from the late first to the mid-second centuries. An associated gravel alleyway and steps probably led down to the Walbrook stream, a tributary of which crossed the eastern area of the site. In 2012 excavation in Trench 2A in the north-eastern area of the site revealed a timber revetment and two timber box drains, all aligned east–west. Trench 3, towards the western side of the site, picked up the western edge of one of the properties found in 2010. Immediately north of this were features associated with the open air: a possible kiln and a water hole. The latter may have been for animals, with green staining of the natural sandy gravel suggesting prolonged use of this area. In Zone 4 a number of Roman buildings, with associated gravel pathways and a total of six timber wells (one of which had a skeleton at its base), were recorded. Two circular wattle structures were also seen. The presence of stone walls, as part of a later property, suggests activity on the site from the first to the third centuries. A number of early medieval buildings had been built on top of dumping over the Roman phase of construction. They comprised turf walls, with gravel alleyways between them. A watching-brief in Zone 1, at the very western edge of the site (fronting Moorgate) revealed a further three Roman timber wells, pits and a kiln.Footnote ²²⁷

(9) International House, Dukes Place, EC3 (TQ 33455 81215): work continued from 2010. In one test pit, a silty deposit containing a single amphora fragment was recorded above a layer of redeposited brickearth. Below this a silty deposit containing pottery of a.d. 120–250 and an imbrex tile of a.d. 50–160 was observed. However, as the deposit was only seen within a small slot, it is not possible to determine whether this was the fill of a pit or part of a layer.Footnote ²²⁸

(10) Moorgate Telephone Exchange, Fore Street, EC2Y (TQ 32585 81640): a number of Roman features were cut into the natural brickearth. These consisted of three pits, the base of an east–west running ‘ankle-breaker’ military-type ditch, and a second east–west ditch which had been partially truncated by a well, which was also of Roman date.Footnote ²²⁹

(11) 5–7 St Helen's Place, EC3 (TQ 33211 81303): following an evaluation in 2011, excavation extended from the north elevation of St Helen's church and a watching-brief was carried out on the sub-basement. The excavation revealed natural brickearth overlain by extensive sub-soils of probable Roman date. The earliest activity identified consisted of pitting and dumping, with quarry pits being filled and overlaid by deposits containing pottery and building material, including painted wall-plaster. There was some evidence for Roman structures and buildings including the plaster rendered brickearth sills of a clay and timber building in the north part of the site and, in the south part of the site, the base of a stone wall foundation with an associated trampled floor surface. The latter was located near a number of large cess-pits and a square-cut Roman well which may have been timber-lined. Burnt floor- and roof-tiles, daub with wattle impressions, and pottery were recovered from these deposits. The early Roman city ditch identified at the nearby Baltic Exchange and projected to run through the site has not been located. Archaeological survival in the sub-basement was limited to the bases of deep-cut features such as large pits and wells which cut the natural clay and gravel deposits beneath.Footnote ²³⁰

(12) Three Quays House, Lower Thames Street, EC3 (TQ 33350 80550): a series of timber piles was observed during excavation of a new sewer heading along the northern perimeter of the site. These were on roughly the same alignment as the later Roman waterfronts seen in the previous phase of work. They were located to the north of the late third-century riverside defensive wall and may represent the earliest of these structures on the site. Due to project depth restrictions, they were not excavated. The timber piles were sealed by a large assemblage of samian ware.Footnote ²³¹

BARKING AND DAGENHAM

(1) 62–96 Abbey Road, Barking, IG11 7BT (TQ 4407 8346): an evaluation recorded a north–south-aligned, V-shaped Roman ditch below later fluvial deposits. The location of the feature in an area close to the River Roding suggests that the ditch had a drainage function.Footnote ²³²

HACKNEY

(1) Principal Place, Worship Street, EC2 (TQ 3333 8208): excavation in the south-east corner of the site in an area formerly occupied by nineteenth/twentieth-century railway viaducts revealed substantial remains of an extramural cemetery located north of Roman London. Twenty-five inhumations were recorded as well as evidence of cremation and a probable bustum burial (a burial pit over which was laid a funeral pyre which on burning was allowed to collapse into it). In addition, four fairly substantial north–south-aligned ditches were recorded which probably pre-date the use of the site as a cemetery. The area to the north of the cemetery showed evidence for Roman brickearth and gravel quarrying. Towards the southern end of the cemetery was a cut feature containing 132 coins (19 gold, 113 silver). The gold coins dated from a.d. 364–92, while the silver coins appeared to be not earlier than the late fourth century and may even have been in use during the early fifth century.Footnote ²³³

(2) Land at Holywell Lane, Shoreditch, E1 6HY (TQ 33430 82320): an evaluation revealed a possible Roman field boundary ditch and an agricultural layer below medieval deposits.Footnote ²³⁴

HAMMERSMITH

(1) Fulham Palace, Bishop's Avenue, SW6 6EA. Stage 2 Restoration and Revival Project (TQ 2401 7613): a watching-brief to the north of the east lawn recorded a north-east–south-west-aligned Roman ditch, with evidence of re-cutting, and one small pit, which yielded pottery dating to the third–fourth centuries. These features were sealed by soil layers containing additional sherds of pottery datable to the same period. A third- to fourth-century Roman coin was recovered from a later soil horizon in the same area.Footnote ²³⁵

HAVERING

(1) Harold Wood Hospital, Gubbins Lane, RM3 (TQ 5420 9035): a further phase of excavation uncovered additional evidence of occupation during the first century a.d., including a curvilinear ditch, representing a possible enclosure, a series of pits and post-holes, and a cremation containing the remains of a 5–10-year-old child. Of further note is the presence of remains of a small pony or donkey from the fills of the enclosure as, whilst the identification of these remains is by no means certain, very few examples of donkeys are known from the Roman period. A fragment of beehive quern in Worms Heath Puddingstone may be the first example found north of the River Thames.Footnote ²³⁶

HOUNSLOW

(1) Duchess Gate, Syon Park, Brentford, TW8 (TQ 16920 77020): excavations revealed a Roman pit and other features associated with a rural settlement next to the London–Silchester road previously recorded to the north-east at other sites in Syon Park and Brentford. Some features were covered by dark earth containing pottery dated to a.d. 350–400. The evidence extends the known limits of the settlement site c. 30 m further south-west.Footnote ²³⁷

LAMBETH

(1) Lambeth Palace, Lambeth Palace Road, SE1 (TQ 30795 79241): work adjacent to the perimeter wall of the north gardens revealed natural sands and gravels cut by several pits, gullies and ditches of Roman date. At the north end of the site was a large flat-bottomed ditch probably used for water management. It contained pottery of third- to fourth-century date as well as brick, daub and flue-tiles suggesting the presence of buildings nearby. It was also found to cut an earlier north-east–south-west-aligned ditch observed during excavations in 1986.Footnote ²³⁸

LEWISHAM

(1) Honor Oak Park Sports Ground, Brockley Rise, SE23 (TQ 36505 74248): following work in 2009, excavation in the footprint of the proposed sports pavilion building revealed a previously identified north-west–south-east running Roman road, evidenced by a shallow, concave linear ditch. This was filled with gravelly sand overlain by sandy gravel which appears to have formed the road surface. Running parallel on the east side of this feature was a V-shaped linear feature interpreted as a roadside ditch. This contained fragments of a Verulamium region white ware flagon dated a.d. 70–120. No corresponding ditch was evident on the west side due to truncation from levelling to form the playing fields.Footnote ²³⁹

NEWHAM

(1) Stratford, Old Rokeby School, Pitchford Street (TQ 3889 8405): work at School 21 (formerly Rokeby School) comprised the modification of the existing school buildings and the construction of a sports hall on the west side of the site.Footnote ²⁴⁰ In the area of new build, mitigation consisted of ten trenches dug on the site of new foundations. These uncovered ditches and small pits of first- and second-century a.d. date at the west edge of the site. The features were presumably peripheral to the focus of occupation found at the nearby Stratford Market Depot Jubilee Line Extension.Footnote ²⁴¹

SOUTHWARK

(1) 177 Grange Road (TQ 3349 7918): a thin layer of silty sand subsoil containing pottery of early second-century a.d. date was cut by a number of late second-century features including a pit, two post-holes and a series of north–south-aligned linear ditches. These were sealed by ploughsoil dating to the fourth century.Footnote ²⁴²

(2) 430 Old Kent Road, SE1 (TQ 33946 78081): an evaluation trench in the centre of the site revealed that the natural gravels had been cut by a north–south-aligned Roman ditch. This contained tile dated a.d. 120–150 and pottery of mainly first- to second-century date, but including two sherds of Oxfordshire white ware (a.d. 240–400). A subsequent excavation located a further five Roman ditches cut into the natural gravels, including part of one previously identified in 1995. These were probably field boundaries and/or part of a drainage system. Although a Roman road is thought to follow the line of the Old Kent Road adjacent to the site, no evidence for this was found.Footnote ²⁴³

(3) New London Bridge House, 25 London Bridge Street, SE1 (TQ 32820 80230): following work in 2010, excavation of the basement area was completed. A series of Roman sand and gravel extraction pits was backfilled with large quantities of domestic pottery and animal bone. Ground consolidation was followed by the construction of masonry buildings during the second century a.d. At the western end of the site, the continuation of a building recorded in 2010 was observed, with the addition of a small building to the south. Towards the centre of the site, a series of east–west-aligned chalk wall foundations bordered a mortar floor with three hypocaust pilae. A north–south-aligned wall had been robbed out. In places, the wall foundations were supported by wooden piles driven into the base of the foundation trench. Between these two areas, a large area of chalk cobbles was excavated, overlain by mortar and fragments of a tessellated floor, beneath demolition deposits. One probable late Roman burial was cut into these. Large parts of the site were sealed by ‘dark earth’ deposits which probably represent a mixture of deliberate dumping and waterlain material, dated by finds to a.d. 250–400.Footnote ²⁴⁴

(4) Thameslink Programme: Western Approach Viaduct, London Bridge Station, SE1 9RL (TQ 3291 8022): a series of investigations revealed 16 NNW–SSE orientated, closely-set timber piles, dendrochronologically dated to between a.d. 59 and a.d. 83 and interpreted as structural foundation piles.Footnote ²⁴⁵

(5) Thameslink Programme: London Bridge Improvement Works, Weston Street, SE1 2TF (TQ 3329 8010): an excavation recorded a series of Roman ground-raising deposits, containing large fragments of worked stone and timbers, sealing alluvial deposits indicative of repeated flooding episodes, presumably relating to the former ‘Guys Channel’.Footnote ²⁴⁶

(6) Ewer Street Car Park, SE1 0NX (TQ 3206 8012): a series of investigations recorded three Roman ditches truncated by a series of eighteenth-century features and a layer of alluvium which appeared to have accumulated during the Roman period. A human skull was recovered from the backfill one of ditches.Footnote ²⁴⁷

(7) 161–171 Abbey Street, SE1 3NS (TQ 3380 7945): an evaluation recorded a possible Roman consolidation/reclamation layer sealed by medieval alluvial deposits.Footnote ²⁴⁸

(8) 434–452 Old Kent Road (TQ 3400 7807): the truncated remains of a Roman roadside ditch encountered on an adjacent site in 1995 were recorded within the north-west corner of the development. Evidence for possible associated Roman period boundary ditches aligned on Watling Street was also uncovered in the south of the site. An undated layer of probable ploughsoil was observed within the southern part of the site sealing the ditches.Footnote ²⁴⁹

SUTTON

(1) Carshalton, Carshalton War Memorial Hospital, The Park, SM5 3DB (TQ 2793 6396): a series of excavations between 2008 and 2012 recorded evidence of late Iron Age–early Roman agricultural activity confirming that a farmstead or small settlement lay very close to the War Memorial Hospital, possibly to the south and east of the main excavated areas. The features uncovered consisted of a series of pits and a number of deep shafts dug through the natural chalk, interpreted as grain storage silos, which appear to be enclosed by a series of contemporary small gullies.Footnote ²⁵⁰

GLOUCESTERSHIRE

(1) Hardwicke, Cursey Lane (SO 8994 2819): several abraded sherds of Roman pottery and a fragment of tegula recovered from medieval furrow fills suggest a Roman presence in the vicinity of the site. Alternatively, the artefacts may have arrived at the site within manuring spreads or by other means.Footnote ²⁵¹

(2) Tewkesbury, New Hospital, Howells Road (SO 8959 3264): excavation identified Roman features including a ditched enclosure, possibly a cattle corral. A trackway or droveway probably linked a putative Roman settlement in the Oldbury area with the River Swilgate. Elements of a Roman field-system were also recorded, along with a small number of pits and post-holes. Other Roman features included an infant burial next to an upturned Roman pot containing burnt bone, and a possible well or water hole. Finds included large quantities of animal bone and first- to fourth-century a.d. Roman pottery as well as coins, a brooch and other metal artefacts, all of which suggest proximity to a settlement.Footnote ²⁵²

(3) Winchcombe, 70 Greet Road (SP 02500 29258): a watching-brief during construction revealed a number of Roman ditches. These were poorly dated but related to features noted in the area in earlier evaluations to the west and north. The features comprised early Roman ditches probably infilled by the second century a.d. and then a second phase of Roman activity represented by two roadside ditches and the beamslots and post-hole of a structure lying adjacent to the ditched road.Footnote ²⁵³

SOMERSET

(1) Backwell: two radiates (of Gallienus (sole reign) and Carausius) and 23 nummi, deposited a.d. 340 or later.Footnote ²⁵⁴

(2) Frome II: an AR miliarensis (a.d. 375–395, Trier) and 60 AR siliquae deposited c. a.d. 388–95 or later and forming part of a previously discovered hoard.Footnote ²⁵⁵ Some of the siliquae were fragmentary and two were found to fit with fragments from the original find, making 58 additional siliquae only.Footnote ²⁵⁶

(3) Kingston St Mary, Lower Marsh Farm (ST 225 282): evaluation subsequent to geophysical survey of a potential solar farm site identified two parallel ditches in Area 2, corresponding to cropmark evidence for a double-ditched enclosure of Roman date. Features in Areas 4 and 5 related to a Roman field-system and an earlier, possibly middle to late Iron Age D-shaped enclosure.Footnote ²⁵⁷

(4) Woolavington: forty-four base metal radiates, deposited a.d. 274 or later, found on a black organic patch interpreted as possible decayed leather.Footnote ²⁵⁸

DEVON

(1) North Tawton, Den Brook (SS 6870 0025): the remains of the east–west Roman road running through the southern part of the site were identified in evaluation trenching and an area excavation. The road survived in relatively good condition with much of the stone metalling still in situ. Two wheel-ruts were apparent in its surface and a ditch was located on the southern edge. The northern edge was truncated and disturbed by the double hedgerow and ditch of the present parish boundary, but the road was over 7 m wide. Another ditch lay just to the south of the road on the same alignment. A group of features in the western part of the site was undated.Footnote ²⁵⁹

CORNWALL

(1) Newquay, land at Tregunnel Hill (SW 8050 6125): an excavation identified the flexed burial of a teenager buried with two copper-alloy brooches dating to the mid- to late first/early second centuries a.d.Footnote ²⁶⁰

(2) St Blazey: a hoard of 45 base-metal radiates deposited c. a.d. 271 or later.Footnote ²⁶¹ The hoard closes with a final issue of Victorinus (a.d. 269–71) and contains a large proportion of the pre-debasement issues of Postumus (over 46 per cent). Radiate hoards closing with Victorinus form a comparatively unusual category noted for their diverse compositions. For discussion of such hoards see E. Besly on the Bassaleg hoard from Gwent.Footnote ²⁶²

WILTSHIRE

(1) Charlton, Charlton St Peter (SU 108 565): a caesium magnetometer survey provided a detailed plan of the character and layout of this known villa complex; quantities of sandstone, box-flue tiles and third- to fourth-century pottery have previously been found at the site. The overall settlement area defined by the magnetometer survey is trapezoidal in form, bounded by ditch anomalies to the north and south with the area in between further divided into a series of sub-rectangular enclosures containing a distribution of intense, positive magnetic responses indicative of pits, quarries and, perhaps, semi-industrial features such as hearths, ovens and furnaces. Three distinct areas of rectilinear negative anomalies located within the western part of the enclosure complex provide confirmation of the substantial masonry buildings identified by the previous earth-resistance survey.Footnote ²⁶³ Particular internal detail was revealed within the eastern and most elaborate of the three buildings identified, where individual rooms and corridors are apparent. An intense magnetic response to the south of the latter suggests the presence of fired features within the building. Compared to the previous earth-resistance data, the magnetic anomalies appear to suggest the presence of a much smaller second building, of simple rectangular plan, to the west. Beyond this lies a larger less well defined rectangular building that may represent an aisled hall structure typically found in such settlements. In the absence of intrusive evaluation, the relationship between the three building ranges is presently unclear. The enclosure containing the east building appears to be sub-divided into separate quadrants with possible evidence for semi-industrial activity segregated away from the domestic occupation in the north-east. There is less evidence for internal activity within the remainder of the wider ditched enclosure system east of the concentration of buildings, suggesting areas of paddocks or stock enclosures, although the presence of semi-industrial anomalies increases to the east of the settlement and additional anomalies here may represent the foundation gullies of timber structures. A distribution of surface noise in the same area may be related to ceramic building material or possibly an east–west field boundary of more recent origin shown on historic OS mapping.

Concentric ditch-type anomalies to the west of the Roman settlement may represent a trackway or, perhaps, be part of a large curvilinear enclosure complex lying mostly outside the area of survey. This is partly replicated as a broad, low resistance response in the previous geophysical survey, when it was suggested to be a former meander of the River Avon, possibly providing a water supply to a bath-house suite in the adjacent villa complex.Footnote ²⁶⁴ However, from consideration of the more recent wider area magnetic data, the anomaly now seems less likely to be of fluvial origin.Footnote ²⁶⁵

(2) Chippenham, Hunters Moon Farm, Easton Lane (ST 9028 7162): evaluation subsequent to geophysical survey confirmed the presence of small scattered clusters of features. One pit at the south-eastern edge of the site contained later prehistoric and Roman pottery. An Iron Age and Roman site lies immediately east of this location.Footnote ²⁶⁶

(3) Pewsey: two AR siliquae are addenda to a hoard of 30 siliquae dated to a.d. 402 or later.Footnote ²⁶⁷

(4) Salisbury, Fugglestone Red (SU 1123 3347): evaluation of a proposed new cemetery site revealed a shallow cut feature containing a fill of possible industrial waste including a small quantity of Roman pottery.Footnote ²⁶⁸

(5) South Marston, Sevor Farm (SU 198 885): evaluation was targeted on four areas of archaeological features identified by geophysical survey. Area 1 contained two large parallel boundary ditches on an east–west alignment, and a possible sub-rectangular enclosure of Roman date towards the northern end. A number of pits were potentially of similar date to the surrounding features. In Area 2 a large occupation spread included a possible stone wall and Roman pottery; ceramic building material including box-flue tile suggested the presence of a high-status Roman building in the near vicinity. Further features were undated. In Area 3 a ring ditch contained a significant quantity of Roman finds within its uppermost fill. A series of ditches may have formed part of a field-system or occupation features associated with the ring ditch. Pottery was dated predominantly to the second to fourth centuries a.d. Only one pottery-dated feature was earlier than the Roman period. Rural settlement may have developed in tandem with the small town of Durocornovium (Wanborough) to the south.Footnote ²⁶⁹

(6) Tidworth, Perham Down Defence Training Estate (SU 2625 4930): examination of a new access track identified two small ditches at the southern end of the route. These may form part of a Roman field-system recorded from aerial photographs.Footnote ²⁷⁰

(7) Tisbury: two AR light miliarenses (fragmentary), deposited c. a.d. 378 or later.Footnote ²⁷¹

DORSET

(1) Iwerne Minster area: an AR finger-ring set with a denarius of Elagabalus (illegible reverse), dated after a.d. 218–22.Footnote ²⁷²

HAMPSHIRE

(1) Chawton area: a hoard of 105 Iron Age AV staters (98 of Verica, 1 of Cunobelin and 6 of Epaticcus), deposited c. a.d. 30–45 or later. A plated Roman silver denarius of Vitellius (a.d. 69) and a late Bronze Age to early Iron Age arrowhead found with the coins were not considered to be associated with the hoard.Footnote ²⁷³

(2) Chineham, Razors Farm (SU 6517 56291): evaluation trenches positioned in the south-east of the site, corresponding with cropmark evidence and geophysical survey results examined a farmstead in use from the late Iron Age/early Roman period to the early part of the second century a.d. This settlement was defined by a system of enclosures, field boundary ditches and occupation debris. Dispersed small pits and undated post-holes were also identified.Footnote ²⁷⁴

(3) Itchen Abbas, Sunnybank (SU 5364 3292): a third phase of evaluation adjacent to a possible pond barrow associated with a number of satellite burials and pits of late Roman and early Saxon date, forming part of an extensive cemetery which extends north of the site, confirmed the presence of archaeological features to the west of the barrow. A pit, which may mark the western limit of the site, was probably linked to an adjacent pit found in the Phase I evaluation that produced Romano-British pottery. These features were related spatially with other pits and graves, also detected in the Phase I evaluation.Footnote ²⁷⁵

(4) Silchester (Calleva Atrebatum) (SU 64 42): the sixteenth season of excavation saw good progress in revealing the earliest occupation on the site of Insula IX. It is now possible to discern three phases of late Iron Age occupation (Period 0), two of which share orientations in common with the previously reported early ditch dating from c. 40 b.c.,Footnote ²⁷⁶ while the third appears to represent a new arrangement of rectangular buildings and compounds oriented with the north-west/south-east-aligned street or lane ( fig. 17). The footprint of the large timber hall previously reported now exceeds 27 m in length.Footnote ²⁷⁷ It belongs to the middle Iron Age phase, c. 20–1 b.c. There is little evidence so far of Tiberio-Claudian occupation, c. a.d. 20–40. The evidence for a conquest-period military occupation, c. a.d. 44–9, remains largely confined to further finds of military artefacts alongside the early (pre-Flavian) laying out of the north–south street. There are now some 30 military items of first-century date from Insula IX and some 150 from Calleva as a whole. Thereafter dense occupation (Period 1), comprising rectangular and circular buildings, builds up in the Insula in the Claudio-Neronian period. It can be subdivided into two phases ( fig. 17). Two further wells of this period were excavated, one containing a rich assemblage of mineralised plant and seed remains.Footnote ²⁷⁸

FIG. 17.

Silchester Insula IX, late Iron Age and early Roman phase plans.

SURREY

(1) Abinger, Cocks Farm villa (TQ 106 474): three small trenches were opened in the area of the villa to check the location of and expand on excavations in the 1990s.Footnote ²⁸¹ Only the base of foundations survived but these were sufficient to show at least two phases of building. In the centre of the adjacent field a large trench examined part of the area of a circular anomaly found by magnetometer survey set at the crest of a slope up from south and west, on the edge of a wider plateau. Feature survival was poor at the very base of the plough level but it was possible to confirm the existence of a sub-circular enclosure with traces of interior pits and gullies with related Roman (and possibly late Iron Age) pottery. Finally a trench was placed downslope from the lime kilnFootnote ²⁸² on the edge of a modern lane in a deep holloway. It showed that the Roman period ground surface dropped to a level that may represent a trackway cut into the side of a natural slope that became a precursor of the holloway. On its surface several small lumps of chalk (out of place in the Greensand) suggested that this was the route by which the charge for the kiln was delivered from the chalk Downs to the north.

(2) Ashtead, Ashtead Common villa and tileworks (centred TQ 1775 6017): a seventh season of work comprised further excavation on the newly discovered building, on the buildings previously examined by A.W.G. Lowther and on the tile kilns.Footnote ²⁸³ It was possible to confirm that the main part of the building west of Lowther's ‘house’ was a rectangle about 6 by 14.5 m. Its later, slighter, northern addition could not be traced further but it is likely that it turned back in some form to wrap around the main building. A line of post-holes and some other features further east may be associated with the structure but this is not yet clear. It was very noticeable that the northern foundations of the main building and its addition had shifted and heaved in parallel, probably the effect of being built on clay. Finds included more first- and second-century pottery, fine glassware, window-glass, and a few pieces of stucco. There were also many fragments of combed and relief-patterned flue tile.

A trench was placed over the junction of Rooms 4, 6 and 7 in Lowther's ‘house’. It was possible to take advantage of a roughly circular disturbance over the junction, perhaps caused originally by a tree, to gain a great deal of information about the structure and relationship of the various walls. Room 6 appeared to be primary, with very deep foundations against which the shallower foundations of Room 4 had been added. No evidence could be found for the wall between Rooms 4 and 6 having originally continued to the south, or that between Rooms 6 and 7 to the east, either of which might have explained the odd L-shape of Room 7. The well-known continuous box-tile jacketing of Room 6 was found to be still in situ as were some of the herringbone bricks of the floor of Room 7, with evidence for an earlier floor level below. In Room 4 there was no evidence for an earlier floor, just the make-up for the parallel below-floor channels found by Lowther at the eastern end of the room and now proven to run right across the room as he postulated. They were set at a considerably higher level than the base of the hypocaust in Room 6. A small trench at the separate bath-house confirmed the location of the walls of Lowther's Rooms A and B and also confirmed that the wall of A was curving as recorded in 1925. Some fragments of combed and relief-patterned flue tile were also located in this area.

Work on the kilns concentrated on the stokehole area and the kiln frontage ( fig. 18). As previously explained,Footnote ²⁸⁴ there were two kilns, one built on top of the remnants of the other. Constraints imposed by the presence of veteran trees made it impossible to examine the eastern end of the stokehole, which may have opened straight into the main clay pit (perhaps explaining the lack of evidence for a drain). There were two main levels: a deep charcoal layer associated with the first kiln and then a levelling layer of burnt clay and tile fragments to raise the area to the operating level of the second kiln. This had much less surviving associated charcoal. The tile walls of the first-period central flue had been badly damaged by burning nearer the stokehole end and had been rebuilt with fresh tiles at some stage. This flue was bottomed in a section and shown to have several thin charcoal levels separated by thin layers of burnt clay. These had built up the floor level considerably to what was presumably the final operating level, which had large tiles serving as a floor in front of the stokehole. The whole of the tegula structure at the end of the first flue seen in part the previous year was examined. On either side of the central flue the stokehole had tile-built retaining walls; that to the south was at a marked angle whereas the one to the north continued more or less at right angles to the flue and only for a short distance. Some phasing for elements of these walls was suggested by their relationship to the later and earlier charcoal levels in the stokehole. Beyond the stokehole retaining walls, the tile-built front walls of the kilns also showed two periods, the later one clearly bonded into the later central flue side walls. The rebuilding of the lower kiln flue and the considerable build up of charcoal levels associated with it must suggest that the first kiln was in use for a considerable period of time. Work will continue in 2013, which will be the last year of fieldwork for the current project.

FIG. 18.

Ashtead, tile kiln.

(3) Charlwood: an AR quinarius (Roman Republic, uncertain issue) and 5 AR denarii, deposited c. a.d. 141–61 or later.Footnote ²⁸⁵

(4) Ewell, Roman roadside settlement (centred TQ 2219 6293): excavation took place on part of a 10 by 80 m strip of meadow (soon to be used as an extension of a cemetery), thought on the basis of previous work to fall within the area of a roadside settlement on the London–Chichester road.Footnote ²⁸⁶ A number of Roman pits, gullies and a ditch were identified but an episode of deep ploughing has destroyed most of the later levels on the site. Two gullies appear at right angles to the projected line of Stane Street that is thought to run through the meadow, and a ditch, which was either recut or was partly on the line of a later gully, ran parallel to it. One very large pit was cut by a later gully and another pit, containing a substantial part of a Dressel 20 amphora, was cut into the ditch. The trench produced a range of first- to fourth-century pottery, including imported fine wares, together with coins and brooches. 180 coins were retrieved mainly by metal detecting and of the 71 that could be identified, 69 are of Roman date, mostly late third- to fourth-century.Footnote ²⁸⁷

(5) Flexford (SU 938 509): a fourth season of excavation on the 10 ha Romano-British rural settlement produced further evidence for buildings and flint surfaces.Footnote ²⁸⁸ At the north end of the site, near the flint floors reported earlier, 21 post-holes have been found in four parallel rows which, if they represent the remains of one structure, could indicate an aisled building with a nave 4 m wide, 10 m wide eaves to eaves, and at least 14 m long. The post-holes are typically 30 cm in diameter and 30–50 cm deep with flint packing. There was no trace of a floor and few finds apart from coarse pottery in four post-holes. Nearby ditches, however, contained large quantities of pottery dating from the early second to third century. An area of burning close to one post-hole produced a radiocarbon date from the late third or fourth century, suggesting that the building no longer existed by that time.

Two more substantial flint surfaces have been found towards the south of the site. One lines a hollow about 10 m in diameter and up to 60 cm deep. This may be a pond lined to protect the clay base for use perhaps for retting flax or washing animals or wool without creating a muddy area. The fill contained building remains and third- and fourth-century pottery and coins. A second flint surface was at least 20 m by 5 m with a ditch on one side, traces of iron smithing and in places considerable quantities of third- and fourth-century pottery. A second feature from a group of magnetic anomalies at the south of the site has revealed another rectangular pit approximately 2.5 m by 1.2 m by 0.5 m deep and lined with 2–10 cm of burnt clay. This pit had ash and sand and over 10 cm of charcoal and burnt logs at the base. Mixed with the charcoal were small patches of calcined bone; about 200 g of bone was recovered which may be a mixture of human and animal. This seems to have been an area of cremation pyres but it is not clear whether these were bustum burials. A sample from the charcoal gave a radiocarbon date of a.d. 260–390.

The site has produced 13 Roman steelyard weights, 12 brooches and 252 Roman coins. Two thirds of the coins have been identified to a Reece period of which 27 per cent are pre-a.d. 138 and 50 per cent are dated a.d. 138–92. The remaining 23 per cent are in the range a.d. 222–c. 405. Both the coin distribution and the ratio of weights to brooches are unusual, although Westhawk Farm, Kent has similarities.Footnote ²⁸⁹ Details of the coins have been uploaded to the Portable Antiquities website.

(6) Thorpe, Church of England School, Rosemary Lane (TQ 0193 6871): evaluation revealed mainly Bronze Age features. The base of a Roman vessel was recovered from a small ‘V-shaped’ ditch that cut a Bronze Age boundary feature. Some second-century pottery was recovered, mostly from topsoil.Footnote ²⁹⁰

EAST SUSSEX

(1) Hailsham, Woodholm Farm (TQ 5775 1070): excavation following geophysical survey and evaluation revealed a series of ditches and gullies aligned north-east/south-west and north-west/south-east representing boundaries and field-systems. Pottery dating evidence indicated an emphasis on the late Iron Age–early Roman period, though the sherds were in poor condition. A small number of discrete features including pits, post-holes and a small burnt pit were also revealed.Footnote ²⁹¹

KENT

(1) Canterbury (Durovernum Cantiacorum): (a) 45–47 Wincheap (TR 1436 5717): excavationFootnote ²⁹² prior to residential development identified two separate sequences of clay, plaster and metalled floor surfaces representing the interiors of two probable buildings dated to c. a.d. 50–300. An area of intense burning above the final clay floor of one building potentially represented the remnants of a hearth. Both structures were overlain by occupation/soil deposits dated to c. a.d. 175–400. (b) 6–8 Station Road West (TR 1450 5820): evaluationFootnote ²⁹³ revealed one inhumation burial and indications of eight others, all potentially part of the adjacent late Roman cemetery excavated in 2011.Footnote ²⁹⁴ A number of post- and stake-holes were also recorded. (c) 38b St Dunstan's Street/Ryde Street (TR 1420 5825): evaluationFootnote ²⁹⁵ revealed the metallings and roadside ditch of a south-west/north-east-aligned street, probably the late first- or second-century route identified some 25 m to the south-west at Ryde Street.Footnote ²⁹⁶ Charcoal, daub and ceramic building material beneath the metallings suggested occupation pre-dating the road similar to that recorded at other sites in the vicinity.Footnote ²⁹⁷ (d) London Road (TR 1397 5814): evaluationFootnote ²⁹⁸ prior to housing development revealed tentative evidence for a late Iron Age field-system and roundhouse and the southern edge of Watling Street. (e) Giles Lane (TR 1340 5940): excavationFootnote ²⁹⁹ revealed a prehistoric hilltop settlement, where a roundhouse within a massive enclosure appeared to be occupied into the late Iron Age.

(2) East Farleigh, Oaklands, Lower Road (TQ 7280 5359): trial pits in the woods south of Buildings 3 and 5 of the villa siteFootnote ³⁰⁰ failed to locate any further buildings. A trial trench 6 m long across the east end of Building 3 located a north–south wall demolished to below the floor level of Building 3. This wall, demolished by a.d. 240, is parallel to the demolished east wall of Building 2.Footnote ³⁰¹

(3) Lenham area: a hoard of 1,062 base-metal radiates (and 19 fragments), deposited c. a.d. 283 or later. Sherds of pottery, copper-alloy vessel fragments and slag were also found.Footnote ³⁰²

(4) Lynsted, Claxfield Farm (TQ 9467 6115): excavationFootnote ³⁰³ ahead of brickearth quarrying recovered a set of decorated Roman tweezers from a ditch which formed part of a series of enclosures spanning the late Bronze Age to late Iron Age.

(5) Maidstone, Church Street (TQ 7622 5586): excavationFootnote ³⁰⁴ revealed a series of north-east–south-west-aligned ditches, most of late Iron Age date, but one, containing pottery of late first- to early second-century date, appeared to be a Romano-British continuation of the earlier features. A cluster of pits all contained evidence of cess, but also metalworking and domestic waste and pottery dating through to the second and third centuries a.d. A hearth and flue were filled with carbon and pottery. The pottery from the site was predominantly of first- or second-century date.

(6) Margate, Capital House, Northdown Road (TR 3575 7115): excavation and watching-brief work prior to residential developmentFootnote ³⁰⁵ revealed a multi-phase ditch provisionally dated to the late Iron Age and possibly associated with hilltop settlement identified to the north-west.Footnote ³⁰⁶ Contemporary features included pits, drainage ditches, gullies and an extensive series of post-holes suggesting potential structures. Two inhumation burials were located on the edge of one of the ditches associated with mid- to late Iron Age pottery.