Monument record 3519/0/1 - Romano-British Circular Building

Excavation uncovered a large circular stone built foundation, dated to the 3rd century AD; since no other building debris was found the rest of the building may have been built in wood or it may have been a shrine. Excavated evidence provides a date from the mid 1st to the mid 3rd century. The building has also been interpreted as being possibly part of a Roman villa.

{1} Building remains were found on the east side of a small valley during deep ploughing in the early 1960s and first explored by Hall and Nickerson over a week in 1964. Their excavation revealed part of the remains of a substantial circular stone structure, which was later totally stripped and recorded by E. Greenfield, over 5 weeks in 1965, in advance of their destruction by the land-owner.
In the area of the structure were two ditches found to date to the C1st and C2nd AD. Ditch 2 lay beneath the circular building, and although no stratigraphic relationship was established between the building and Ditch 1 it was assumed to also belong to an earlier system of land allotment.
A deposit of pottery in the upper fills of Ditch 1 was dated 150-200 AD, possibly contemporary with the occupation of the overlying structure.
The building itself comprised the footings of a circular rotunda, containing four post pads and four internal walls, the main circuit was about 50ft in diameter and lay c.1ft below the ground surface. The masonry had been disturbed by stone robbing as well as cultivation and no occupation levels survived. The small quantity of pottery from the robber trench infill does not form a distinct group separate from the rest of the finds suggesting that it is either residual, or that robbing took place soon after the building was abandoned. Medieval cultivation furrows extended through the building remains, clearly removing stones from the footings they crossed.
The walls of the building were of local limestone in a drystone construction, employing pitched limestone arranged in multiple courses within a foundation trench cut down to the natural sand. None of the stones had been dressed. The footings were between 2.5 and 3ft in width around most of the circuit, but slightly thicker where they overlay the earlier ditch. Around most of its circuit the wall survived only in one or two courses, but over the earlier ditch three or four courses had survived. It is not known how high the original walls must have been but the size of the footings suggests that the stonework extended some way above the Roman ground surface.
Four large post pads were revealed by Greenfield, of which three had been noted in the earlier excavation. These formed a nearly square arrangement in the centre of the structure, each one roughly on a cardinal point of the compass [these are described in detail]. The arrangement suggests their use for upright roof supports.
The remains of four internal walls survived [described in detail].
In addition to the main stone features three small post holes were found, one with post-packing. A gully is also recorded, running across the south-east corner of the excavated area in a south-south-west to north-north-east direction, both ends extending beyond the limits of excavation. Part had been removed by a medieval furrow, but a feature c.18ins wide was observed. A depression was also recorded beside the eastern limit of excavation, but no further information was present.
Several irregular areas of limestone were recognised around the structure and Greenfield interpreted the linear spread of stone to the north-east of it as a metalled entrance flanked with drainage sumps. The latter were actually furrow bases and the limestone probably a scatter resulting from cultivation so this interpretation is no longer tenable.
The majority of pottery recovered dated from the C1st and C2nd AD, storage jars and bowls being the dominant forms.
Two brooches were also recovered, both having a C1st or early C2nd date; only one brooch was stratified, from the adjacent gully, where it accompanied 8 pot sherds dating to about the second quarter of the C2nd, along with a small rim fragment of a later colourcoat vessel.
The structure may have been associated with the Romano-British settlement 1.5km to the south-east but no contemporary occupation is recorded closer than this. The site occupied part of an area already utilised for agriculture during the C1st and C2nd. No material survived to date the building’s construction, but a predominance of C2nd material around suggests this was the period of occupation.
Such circular structures are not uncommon. Some have been identified as rural shrines, but such cases normally yield votive objects and are set within a wider ritual context. It is likely that such structures may have had a domestic function, thus Romanising a native British building type. No British circular Roman shrines contain internal divisions of the nature found here, which further suggests a domestic function.
The majority of pottery recovered dated from the C1st and C2nd AD, storage jars and bowls being the dominant forms. Part of a single mortaria was recovered from the tilth. Material from the robber trenches was of C1st to early C3rd AD and may suggest that occupation ceased in the C3rd as later forms are only represented in the ploughsoil by a single sherd. Deposits of pottery within the upper level of Ditch 1 (Hall and Nickerson) were dated 150-200 AD and may be contemporary with occupation of the structure.
The pottery recorded is a typical domestic assemblage. Part of 8 Samian vessels were identified from Greenfield’s excavation, all but one of Antonine date thus corresponding with the date suggested by the coarse pottery.
A single unstratified coin of Hadrian was also found. In addition 15 sherds of early to middle Saxon pottery were recovered from the ploughsoil, most plan body sherds of vessels of indeterminate form.

{2} Further work on the round-house at Bozeat revealed the remainder of the interior, as well as a surrounding area of metalling (Fig.12). A cambered surface flanked by drains indicated an entrance on the east side. Pottery from the metalled area suggests use in the second half of the C2nd. [See reference {1} for rejection of interpretation of evidence for the entrance.] The building overlay the ditches of an earlier Romano-British field system but none of its remains can be precisely related to the air photographic evidence.

{3} Occupation of this site dates from the mid first century to the mid- third century, the circular stone footings being constructed in the third century. No contemporary occupation layers were found so the purpose of the footings is speculative. As no destruction layer of stone in the ploughsoil was found, as is common with stone buildings, it is probably that either the building above ground was wood or that the footings never supported a building.

{4} The remains of the structure comprised a circular stone foundation, with a central arrangement of four circular stone-filled post pads. Although a religious use can sometimes be identified for buildings of this plan-form a domestic use is more likely in this instance since no evidence of ritual activity was recovered by the excavation. Domestic structures of this type occur at other sites in Northamptonshire. The building overlay the ditches of an earlier Romano-British field system.

{6} A circular building c.15m in diameter, possibly a shrine, was exavated in 1964. It had internal partitions and four central posts, and was surrounded by an area of metalling. A cambered surface flanked by drains on the east indicated an entrance. It was said to date from the second half of the C2nd.

{8} A circular footing with a central square arrangement of post pads, in addition four internal walls were present. The outer wall had an overall diameter of between 50-53ft and was between 2ft6in and 3ft wide. It was built of local limestone, the individual stones were not large and were arranged in a pitched fashion.
The individual post pad settings were arrangements of stone 3ft and 4ft in diameter, with variant in scale. The size of the pads would allow the construction of a square of between 4.26m and 4.87m across. The internal walls defined an area of approximately 3.96x3.66m within the southern part of the building, the walls were narrower than the external wall at 0.6m wide. The function of this division of space is not clear.

{9} A circular building 48ft in diameter, with four central posts and internal partitions (Fig.14) has been exposed 0.75 mile north-west of Bozeat.

{12} Listed as the possible site of a Roman villa.

{13} Site excavated for the MOW by Mrs G Brown and Mr D Hall who gave the correct site of the excavation at SP 89595998. It has since been covered and at present is under pasture.

<1> Mackreath D.F., 1985, Bozeat, Draft Excavation Report, p.1 (checked) (Draft). SNN43994.

<2> 1966, Roman Britain in 1965, 56/207 (unchecked) (Chapter). SNN28996.

<3> Hall, D N and Nickerson, N, 1970, Circular Roman Building at Bozeat, Northamptonshire, 1964, p.57-65 (unchecked) (Article). SNN112841.

<4> Ordnance Survey, 1950s/1960s, Ordnance Survey Record Cards, SP85NE3 (unchecked) (Index). SNN443.

<5> Miscellaneous documents, Excavations at Bozeat (Report). SNN57469.

<6> Royal Commission on The Historical Monuments of England, 1979, An Inventory of The Historical Monuments in The County of Northampton, p. 3 Site 4 (checked) (Series). SNN77380.

<7> Ordnance Survey, 1950s/1960s, Ordnance Survey Record Cards, SP85NE3 (unchecked) (Index). SNN443.

<8> Meadows I., 1992, Three Roman Sites in Northamptonshire: Bozeat, Higham Ferrers & Great Oakley, P79 (checked) (Article). SNN40363.

<9> 1965, Roman Britain in 1964, p.209-10 (checked) (Extract). SNN197.

<10> Pevsner N.; Cherry B., 1973, The Buildings of England: Northamptonshire, p. 115 (unchecked) (Series). SNN1320.

<11> 1996, Archaeological Assessment of Land at Bozeat, Northamptonshire: Stage 1 (Desk-Based), p.4 (checked) (Report). SNN75776.

<12> Scott, E., 1993, A gazetteer of Roman villas in Britain, p. 140-1/NH15 (Monograph). SNN110345.

<13> Baird, J., 1970, Field investigators comments, F1 JB 08-SEP-70 (Notes). SNN110341.

Record last edited

Jun 22 2023 1:56PM