Buildings A and B | Building C | Building D | Building E | Building F
Silty sand and sand levelling deposits with high stone contents were laid down prior to the construction of the building (2077, 2085, 2094, 2097, 2099 and 2104). They were up to 0.2m thick overall, notably over the Period 4.3 quarry pit 2106, but tapered out at the north end of the building (along the south side of the Roman road) and 6.2m south of the south wall. This created a flat construction surface (at 6.7m OD) for the building, with a steady down slope further south.
The west wall (2025/2093), the western part of the south wall (2021) and an internal wall aligned north-south (2023) were observed. The north wall had been robbed, and the eastern part of the building lay beyond the limit of excavation. The observed part of the surviving building was 10.9m long by at least 3.5m wide. If the internal wall was centrally placed, Building A would have been 5.8m wide; but the rooms within it would have been very narrow, and it is more likely that the western room formed a corridor alongside a larger room in a broader building. The only below-ground structural element was a sand- and stone-filled foundation pit (2096) beneath the south-west corner of the walls. Otherwise, the wall footings rested on the ground surface. Pitched limestone fragments formed the bottom course of the footings; they were up to 0.25m long by 0.20m wide, and an average of 60mm thick. The gaps between the upper surfaces of the stones were infilled by a layer of sandy loam and pea gravel, so that the top of the pitched stone course was relatively even. Resting on the pitched stone course of the west wall was a course of flat-laid limestone fragments, which were up to 0.50m long, 0.30m wide and 0.15m thick (2022). The edges of the stones that formed the outer faces of the wall had been shaped to produce an even surface. Presumably, timber sill-beams rested on the flat-laid stones and formed the base of a timber-framed structure.
Gravel layers laid within the building during its construction presumably formed the original floor (2031, 2052, 2065, 2069, 2076, 2078, 2080-1 and 2088). Silty loam occupation deposits accumulated on the floor (2028, 2035, 2046, 2048, 2050-1, 2053-4, 2058, 2061 and 2075). They had a high ash content, but it is not clear whether this represents domestic or industrial activity. There were also patches of limestone fragments, which indicates that the occupation deposits were used as floor surfaces, with occasional stone patching of worn areas. Dating evidence for the construction of the building comprises pottery from the levelling deposit and the walls and floor of the building, of mid/late 3rd century date.
The evidence for the later history of the building is unclear. The north end of the structure had been disturbed and presumably robbed, to the extent that the precise position of the north wall is uncertain. Two post-pits subsequently dug into the infill of the robbed area could be repairs of the building, or much later (possibly modern) structural activity. Any deposits associated with the building that lay above the surviving part of the structure, such as later occupation deposits or demolition debris, would have been removed by medieval ploughing and modern disturbance.
The form and method of construction of Building A has close parallels with buildings found at Dragonby, Lincolnshire, notably Buildings 1, 3 and 8 (May 1996, 77ff; Figs.5.15-16, 5.18-21, 5.29-30). These and other buildings at Dragonby are interpreted as timber-framed, with aisles allowing the structures to reach widths of around 10.0m. The outer walls rested on pitched and flat-laid stone sill walls, and the aisle posts rested on either foundation pits or sill walls. Using this model, the observed part of Building A would have been the western side of an aisled timber building approximately 10.0m square, with the north gable end fronting onto the Roman road. The short internal sill wall along the western line of aisle posts would have acted as a partition between the western aisle and the nave.
A short length of pitched limestone (2019) was observed some 5.5m south-west of Building A. It was 0.7m wide and aligned north-south, and probably represents the east wall of a structure similar to Building A (Building B; Fig.6e). If this structure extended as far north as the Roman road, it would have been at least 14.5m long.
A spread of small stones set in sand and gravel (2020), immediately east of 2019, is almost certainly the southern continuation of the path alongside Building A. Immediately to the west was an area of flat-laid limestone fragments, which measured an average of 0.20m by 0.15m (2018). This is probably a floor within Building B.
There was fragmentary evidence of a timber structure on the east side of Trench 1. It rested on levelling deposits of sand, which sealed the Period 2 features and formed a level construction surface (Group 1.23). The main indication that there was a substantial building in this part of the site was a T-shaped kiln (corn drier). The western half had been destroyed by medieval ploughing, but the east side flue (1502) and the east side of the main flue (1515) survived; originally, the kiln would have measured 3.0m long by 4.0m wide. It was built of clay-bonded limestone slabs, set in a narrow channel (1553). A shallow ovoid stoke hole (1495) lay at the end of the main flue. Both the stoke hole and kiln flues were filled with ash and charcoal.
There was a possible beam slot (1301) 1.5m south-east of the kiln. Immediately south of 1301 was a shallow, irregular cut (1302), which may have been an eaves drip. This was probably the south wall of the building; the extant major east-west ditch lay only 1.0m further south. A small pit (1536) containing a complete neonatal calf inhumation (1556) may have been deliberately positioned beneath the south wall. The north end of Building C could not be identified conclusively amongst a concentration of post-holes, although a short length of possible beam slot (1462) could be the best candidate, in which case the building would have measured about 12.0m long. The western part of the building had been removed by medieval ploughing. Another animal burial (1500) was found in the base of a truncated pit (1522) 2.5m west of the kiln; this could have been located beneath the west wall. The east side of Building C lay beyond the east edge of the trench. The position of two large post-pits (1506 and 1518), immediately east of the T-shaped kiln, suggests they were internal structural elements within Building A. Post-pit 1506 contained a limestone post-pad in its base and limestone packing in the sides; post pipe 1483 indicates that the post had rotted in situ. The probable remains of a post were visible in the other post-pit (1518) as a post pipe (1514).
Although the remains of Building C are fragmentary, they show that it was at least 6.0m wide and incorporated a very large kiln. Such a broad building would almost certainly have required internal roof supports, and was probably aisled. Two points indicate that the outer walls were free-standing. Firstly, if the main posts in the outer walls had been earth-fast, it might have been expected that the bases of the post-pits would have survived beneath the medieval ploughing; no such features were found. Secondly, there is evidence that the north and south outer walls rested on sill-beams. This suggests that earth-fast aisle posts took the weight of the roof, and the outer walls were not load-bearing.
The pottery evidence points to a late 2nd century date for the construction of the building (from the make-up beneath the building), to the early 3rd century for its occupation (floor 1390), and the mid 3rd century for its abandonment (backfill of eaves drip 1302).
Post-pits formed two east-west alignments roughly 5.0m apart in the middle of Trench 1. Concentrations of larger features, at least 1.0m in diameter, lay beyond the two post-pit lines but very few were between them. These features may represent a timber building approximately 12.0m long (east-west) by 5.0m wide, with pits to either side. The lack of occupation deposits and of evidence for an east wall could be due to the insubstantial nature of the structure, or to truncation by medieval ploughing. The small pottery assemblage from these features indicates that this structure was in use during the late 2nd and early 3rd century.
This structure was immediately to the south of the major east-west ditch. The main element was a narrow linear cut, which was 5.8m long and aligned east-west (1605). There was a south return to this feature at its west end (1604). Two post-pits appear to extend the alignment eastwards by another 3.0m (1552 and 1624). Three small post-pits (1606-8) were adjacent to feature 1605. These features may have formed an earlier structure, for 1605 cut post-pit 1606. Approximately 3.0m to the south was an east-west alignment of four post-pits (1619-20, 1625-6), and slightly to the south of these was a single post-pit (1621).
The occurrence of two probable beam slots forming a right-angle suggests that they formed part of a substantial structure, but there is insufficient evidence to be more specific. The post-hole alignment south of the beam slots could form either the south side of a narrow building, or the north aisle posts of an aisled building aligned east-west. Building E cut Period 1.1 features of late 1st century date, but dating evidence from the structure itself is lacking.
Three post-pits appear to form an east-west alignment 8.0m long (3022, 4004-5). Two metres north of the eastern post-pit was a small linear cut, 3.1m long and aligned north-south (3026/3028), which may have been a beam slot. 2.5m east of this feature was another post-pit (3020).
These structural elements could be interpreted as one or more fence-lines. On the other hand, the occurrence of a possible beam slot suggests the presence of a more substantial structure. Judging from the available evidence this might have been a timber building measuring 11.0m long (east-west) by 6.0m wide, with an internal partition resting on a sill-beam.
Dating evidence from the structure itself is lacking. However, it post-dates the Period 4.1 ditch 3021, which suggests a 3rd century date for its use.
© Internet Archaeology
URL: http://intarch.ac.uk/journal/issue9/brough/bldings.html
Last updated: Tue Nov 28 2000