From the end of the Roman period, little if any activity appears to have taken place on the eastern half of the site: the two ditch groups in the centre of the site were the most easterly features to produce post-Roman finds, although the dating evidence for these features was ambiguous and they are considered more likely to be late Roman. The site appears to have been almost deserted during the Saxon and early medieval period, except for the encroachment of the Minster cemetery into its western edge: although pottery of this date was found in a number of features, it provided definitive dating evidence for none.

The cemetery encountered in two excavation areas was initially ascribed to the late Saxon and/or early medieval phase by analogy with previous discoveries. One of three skeletons partially uncovered during the 2010 evaluation was radiocarbon-dated to 1262±34 BP, giving a likely inhumation date between the late 7th and the mid- to late 8 th century; other areas of the cemetery excavated in the 1950s and 1970s produced no dating evidence but were interpreted as being parts of the graveyard of Southwell Minster, due to their overtly Christian character and their proximity to the minster church.

As with the previously excavated portions of the cemetery, the character of the cemetery was distinctively Christian. Cremation burials were absent from the site: all of the 42 burials encountered were inhumations. Two of the skeletons exposed were so heavily disturbed that they could not be considered to be articulated, and were found within a construction cut and a robber trench respectively, indicating an ad hoc, probably surreptitious reburial after they had been accidentally exhumed during later work. Of the remaining 40 burials, all were supine and extended, and all but two were east to west-aligned with heads to the west, as is typical of Christian burials. The atypical examples were two graves, in which the two skeletons were also supine and extended, but aligned north- west to south-east with head to the north-west and north-east to south-west with head to the south-west respectively; one of the two graves was cut into the surface of the remains of one of the walls of the Roman building. Radiocarbon dating indicated that one skeleton and the neighbouring skeleton were likely to be later than the other burials dated, and it is possible that these three burials represent a later extension to the cemetery and the burial practices of a different generation. The majority of the graves were arrayed in two orderly rows with little intercutting, possibly suggesting that records of the inhumations were kept or markers maintained; the positions of the more widely spread graves suggest that many of these may also have originally been ordered in rows. The presence of disarticulated bone in a number of grave fills may also suggest that the graves were once more closely packed, but that only the deepest burials have survived the construction of the former school. One grave was particularly badly disturbed: the west end had been cut by a medieval pit, but the east end seemed to have been destroyed by site levelling for the construction of the school, leaving nothing but the pelvis of one skeleton and incorporating two sherds of late 18th to 20th- century pottery into the grave fill.

Grave goods were almost wholly absent, also a characteristic of Christian burials. A copper alloy split ring was found in one grave, but it lay within the edge of the grave rather than being directly associated with the skeleton, and may have been residual in the grave fill or a casual loss during the burial process, rather than having been a personal possession of the deceased; the item has similarities with a known example of a 14th-century earring, but it may also have been a functional object such as a suspension loop. Apart from this single find, all artefacts encountered in potential association with skeletons were of iron, and were probably chiefly coffin nails, although a hobnail was retrieved from another grave. A number of the grave fills produced other artefacts, but the majority of these could be identified as redeposited Roman material, including tesserae (mosaic tiles); a total of 10 out of the 33 earthenware tesserae retrieved from the site came from the fills of graves, presumably transported via a process of soil reworking, typical of graveyards in regular use,from the upcast of those graves excavated through the floors of the Roman villa itself. Post- Roman material was found only in one grave, which produced a single sherd of late 10th to 12th-century pottery; another grave, which produced one 10th to 11th-century and one late 10th to 12th-century sherd; a third grave, in which a potsherd was datable only as post-Roman; another grave, from which 3 sherds of 10th to 12th-century pottery was retrieved, and finally, another grave, which produced the only fragment of medieval CBM to derive from any grave fill, although this was identified as 12th to 13th-century and therefore intrusive; all these graves also produced redeposited Roman material.

Almost no stratigraphic dating was possible within the cemetery, as only two pairs of graves intercut one another. One grave cut the south edge of a second grave, but the condition of the skeleton in the second grave was too poor to ascertain how much of the truncation it had suffered was due to the excavation of the first grave. Another grave cut a further grave, but no further dating evidence is available, as the two skeletons lay within the Roman building and were reburied without recording. A programme of radiocarbon dating of sampled skeletal material, however, proved to be in good agreement with the initial assumptions, suggesting that the period of use of the cemetery lay between the 7th and the 9th centuries AD, with a good probability that it lay between AD 685 and 805. Six burials were sampled, in addition to a date already obtained during the evaluation: the date range was relatively close across all samples, although it is notable that the two samples taken furthest to the south lay at the later end of the range with a calibrated date of AD 710-940, possibly suggesting that the cemetery had expanded in this direction. A general north-to- south movement could not be inferred, however, as the earliest of the dated burials, one skeleton with a calibrated date of AD 650-780, lay slightly south of centre of the excavated area. The two intercutting graves were both sampled, but proved to be effectively contemporary, with the earlier skeleton having a calibrated date of AD 660-780, and the later skeleton at AD 680-880. The study indicated that the lifespan of the cemetery, within the potential date range, may have been very short, possibly no more than three generations.

The composition of the cemetery as represented by the articulated burials was biased towards adult males, with fewer women than men, no infants or young children, and only three sub-adults. However, the disarticulated remains displayed a fuller demographic profile, with bones from a neonate or infant, a child and a sub-adult among the 11 individuals identified. The presence of women and children among the burials indicates that the cemetery served a lay rather than an ecclesiastical community. Analysis of the skeletons suggests a population of low economic status, with indications of nutritional deficiencies and heavy physical labour. Some of the breaks and other bone traumas that were observed may have been the result of accidents at work, but two of the burials appeared to have been the victims of violence. One skeleton, a young adult male, had apparently been attacked from behind, with a blade wound to the right shoulder-blade which he had survived for some time, but which had not been fully healed at the time of death, while a mature male skeleton had four broken ribs, all of which had healed, a fracture to a thoracic vertebra and two depressed skull fractures, one of which had completely healed, while the other was incompletely healed at the time of death; injuries of this type frequently derive from hand-to- hand combat, and the variation in the size and shape of the fractures indicates that a different weapon had been used (and therefore, potentially, a different opponent faced) on each occasion, suggesting that this individual’s exposure to violence was long-term and may have been on a professional basis. Environmental samples taken from the pelvic areas of three skeletons proved to contain the ova of intestinal parasites.

Environmental samples from two of the graves were both rich in charcoal flecks and the chaff of spelt and emmer wheat, suggesting that the processing of cereal crops was taking place near the site at the time that the cemetery was in use (and possibly that these burials took place at harvest time). Analysis of micro-debris in dental calculus sampled from ten skeletons identified starch granules and cereal bran derived from the Triticaceae species group (barley, rye and wheat), starch granules from legumes (beans and vetches) and epidermis fragments from leeks and other leafy vegetables, suggesting that all these had formed substantial parts of the population’s diet; the leeks are particularly significant as dating evidence, as leeks are known to have been widely consumed during the Anglo-Saxon period (the Anglo-Saxon word for a vegetable plot was ‘leek-garth’), but are believed to have been replaced by cabbage in the national diet during the medieval period. Wool fibres and fibres of plants provisionally identified as flax and hemp were widely present, but in greater quantities on the teeth of the female skeletons than on the males, suggesting that spinning, weaving and sewing (activities that probably involved biting off threads as well as inhaling loose fibres) were chiefly carried out by women. In contrast, phytoliths (microscopic siliceous structures) from the epidermis of common reed were present to some degree on the teeth of all eight male skeletons, but absent in both females. This division suggested that the reeds were being used in an activity that was carried out only by men (as opposed to sitting or sleeping on floors covered with cut reeds or living in houses thatched with reeds, where evidence would be expected to show up evenly across both sexes). The Potwell Dyke was probably the source of the reeds, and the nature of the remains suggests a process – the reeds were being cut and worked with, rather than simply walked among, as hunters or fishermen might do – but the nature of that process remains speculative. Pollen from trees suggests a wooded environment, while microcharcoal may indicate smoky dwellings as well as the consumption of smoked or burnt food; mineral grit was likely to be derived from the grindstones used to process cereals. These findings are provisional.

The disuse of this portion of the cemetery seems likely to be associated with the substantial rebuilding of Southwell Minster at the beginning of the 11th century, which may well have included a reorganisation of its boundaries and the layout of its grounds. It certainly seems most plausible that the sequence of 10th or 11th-century ditches cut and recut through the former cemetery represent work carried out under the auspices of the ecclesiastical authorities, who need have had no concerns about disturbing so many relatively recent burials on their own consecrated land.

It is possible that the Early Medieval cemtery at Platt's Orchard directly to the north is the same cemetery. See MNT26981

These burials, although not directly associated with any contemporary structural remains, are likely to relate to a predecessor of the existing minster that stands c.150m to the west. See M3148

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