A number of features and feature groups, particularly in the centre of the site, produced only pottery that could not be dated more accurately than ‘Roman’, and either had no stratigraphic relationships or could not be ascribed to a single phase from those that they had.

Feature Details:

Two complexes of small, narrow intercutting linear features appeared to represent beam slots, with possible associated post-holes. The northern complex lay within the north-western angle of the junction of sequential ditch complexes. It had a roughly rectilinear structure, with three approximately north-north-east to south-south- west running beam slots and one perpendicular one. The feature complex had no external stratigraphic relationships and produced no accurately datable finds. It appears to represent a structure, but not all its elements were in place at the same time: one central beam slot cut the adjacent post-hole, but two larger post-holes cut the western beam slot, and this in turn appeared to cut the perpendicular beam slot. It may be significant that two post-holes were closely aligned with the western group of post-holes from the possible early Roman (Phase 1) structure to the north, while the general alignment of the beam slots was comparable to that of the Phase 1 feature group directly to the west, but no evidence can be presented for a connection between the feature groups. One post-hole displayed a quantity of stone post- packing and produced fragments of Roman CBM, while large fragments of Roman tile, probably also post-packing, were retrieved from a pit.

The southern beam slot complex was made up of three feature groups, and lay within the south-western angle of the junction of two Mid-Roman phased feature groups: the area was so disturbed by flooding during the excavation that stratigraphic relationships with one of Mid-Roman ditches could not be established, although the ditch appeared in plan to truncate the small feature complex. The surviving part of the complex consisted of a west-north-west to east-south-east aligned ‘spinal’ slot, with two shorter slots extending perpendicularly from its north side, and one branching at 45° from its south side. Two intercutting post-holes were sited to the north of the main beam slot. A late Roman ditch cut both the northern slots. The majority of the structural elements here appear to have been contemporary, with the exception of aa north-to-south aligned gully, which may in fact have pre-dated the feature complex: it was cut at its south end by possible beam slot, and may have extended northwards past the late Roman ditch and post-medieval ditch complex; the possible terminal of such a feature appears on the site plan, but it was not excavated or recorded, and any extension that might have shown up between the later ditches had been cut away by a pit. Sections excavated at the junctions of the other apparent beam slots failed to establish relationships, and these features were interpreted as being contemporary. The relationship between two other post-holes could also not be ascertained. The eastern end of the possible beam slot intersected a third post-hole on its northern side: although this post-hole was recorded as cutting the possible beam slot, the features had similar fills and may also have been contemporary parts of the structure. Roman CBM was retrieved from three sections, but no other dating evidence was present.

A loosely north-to-south aligned constellation of discrete features in the centre of the site, with no clear structural pattern and little dating evidence, was recorded as pit/post-hole group; eleven features were assigned to this group. Although no pattern could be identified in plan, the features were mostly between 0.90m and 1.30m in diameter, displayed a roughly north to south alignment, and were characterised by the presence of possible post-packing stones (the exception being one pit which was smaller and contained no inclusions identifiable as post-packing). A quantity of wall plaster was retrieved from one fill, the fill of a possible post-pipe in a pit near the north end of the group, but no dating evidence that might have identified a phase of site redevelopment was present; the fill otherwise produced only a sherd of non-specifically Roman pottery, while an environmental sample contained three charred wheat grains, a single charred chickweed seed (a typical weed of arable land) and a moderate amount of charcoal. Another sherd of Roman pottery was retrieved from the upper fill of another pit. Together, the features of this group produced a sizeable mixed group of Roman ceramic building material comprising 68 fragments of imbrex and tegula roof tiles and one piece of brick: the CBM assessment observed that all of the four most common tile fabrics were to be found in the early Roman phase of deposits, suggesting that the material in this pit/post- hole group represents residual or disturbed material from this phase. A further pit, to the south of the group, was initially phased as medieval by the presence of disarticulated human bone, probably derived from the Anglo-Saxon cemetery; the only artefactual material retrieved was undatable CBM. However, since disarticulated fragments of human cranium were also retrieved from a mid-Roman phased pit in another pit group and in a further post-hole in a late Roman phased structure, it seems most likely that the human remains here are also intrusive, and this southern pit has been assigned to this group on the grounds of its position alone.

To the north-west of the post-hole group, remnants of an apparent double ditch, very heavily disturbed by the construction of the school, ran on the same north-north-east to south-south-west alignment as the mid-Roman phased ditches in the east. The two ditches ran very close together, but did not intersect; both features were approximately 0.5m to 0.6m wide and no more than 0.20m deep. Both were cut at their north ends by medieval ditch fragment; one of the features appeared to extend a short distance beyond it, although it was uncertain which one. The undated eastern ditch could be traced for a length of some 12m and appeared in plan to terminate or have been truncated away to the south, but the presence of an apparent small, unrecognised recut in the section of a large post-pit suggests that the ditch may originally have continued to cut the post-pit: one fill in the pit corresponds well to a fill in the southernmost section through the ditch, and this section did appear to have been machined away at the end rather than coming to a terminal. However, this possible relationship does not provide reliable dating evidence, as the pit was itself only speculatively dated to mid-Roman. Directly to the west, the second ditch was also poorly dated, with two sherds of non-specifically dated Roman pottery coming from southernmost section; the interpretation of these ditches as being Roman rests chiefly on the similarity in their alignments to the better-dated ditches to the east.

In the north-west corner of centre, a short length of north-to-south-aligned ditch ran a length of some 5m between the edge of the site and part of the school foundations: at its south end, it intersected the east end of a narrow gully, running westwards to the edge of the site, and may have been contemporary with it, as the feature fills could not confidently be distinguished in section. One section at the junction of the ditch with the gully produced two earthenware tesserae with other small fragments of Roman CBM, while a single fragment of box flue tile was retrieved from the gully: however, it seems most likely that the Roman material in the ditch was redeposited from the gully, and that the larger ditch is in fact of medieval date.

R. D. Savage and J. Sleap, 2015, Proposed Residential Development, Former Minster School Site, Church Street, Southwell, Nottinghamshire: Archaeological Excavation Report (Unpublished document). SNT5953.