The Body Mine: A Review of Human Remains within Romano-British Well and Shaft Deposits and Evidence for Multi-stage Mortuary Ritual in First-century a.d. Surrey

Abstract

This paper uses a case study of a shaft in Surrey, England, to illustrate the potential of traditionally overlooked material for increasing understanding of Romano-British minority mortuary rituals. Taphonomic analysis of a substantial assemblage of disarticulated human remains from a first-century a.d. quarry from the Nescot Former Animal Husbandry Centre in Ewell, Surrey, showed evidence of multi-stage post-mortem processing. This included decomposition of the remains within a protected environment followed by removal of certain skeletal elements. The assemblage is then contextualised against the treatment of human remains in other Romano-British shaft and well deposits. Finally, the material is compared to Iron Age examples of post-mortem processing, revealing the Nescot shaft is unlikely to be a direct continuation of pre-conquest practices.

INTRODUCTION

Funerary archaeology faces the difficult task of trying to understand a series of actions and processes while only ever looking at the ‘end’ state: the body and the grave. While textual sources can be useful for bridging this gap, the lack of documentary evidence referring to Romano-British contexts means that archaeological data are the main source of information. As such, methods which focus on deposit formation processes are invaluable for understanding the steps leading to the final deposition of remains. For instance, close examination of the position of cremated remains within urns can reveal the collection process,¹ and patterns of heat damage can indicate the state of the body when it was burned, as well as the intensity of the fire.² Materials found within pyre debris, such as jewellery or oil flasks, may be suggestive of laying out of the body.³ The processes behind inhumations are, in many ways, more difficult to reconstruct than cremations, as they are less likely to leave direct evidence that will survive in the archaeological record. Archaeothanatological methodologies, emphasising the positioning of bodies within mortuary contexts and the taphonomic processes surrounding decay, can give indications of grave furniture which may not have survived, as well as insights about the positioning and post-mortem treatment of bodies.⁴ Material which has traditionally been dismissed as residual disturbed material has proven to be invaluable for recognising rituals which have previously been invisible.⁵ Placement of items with the body has been used to identify possible graveside rituals,⁶ as has the analysis of vessels and animal bones from within grave fills.⁷

The vast majority of these studies have focused on cremation and inhumation burials from formal funerary contexts. Even studies of minority or ‘deviant’ funerary ritual have concentrated largely on prone or decapitation burials,⁸ which occur overwhelmingly within cemeteries rather than other contexts, and comprise mostly articulated individuals.⁹ There is however, a much wider range of evidence for mortuary treatment in Roman Britain. Disarticulated human remains, and remains found within so called ‘structured deposits’, both have the potential to shed light on mortuary rituals which lie outside the formal funerary conventions of Roman Britain, yet these have been largely overlooked.

Disarticulated human remains are not a rarity within the archaeological landscape of Roman Britain. The Roman Rural Settlement Project recorded 570 distinct instances of disarticulated human bone, spanning a wide variety of site types and periods.¹⁰ This is more than the sum of decapitation (250) and prone (221) burials the project recorded.¹¹ While it has been suggested that these remains are likely to represent minority mortuary rituals,¹² this material has not been subject to any detailed analysis. Within site reports, disarticulated remains are often dismissed as either ‘stray’ material, evidence of disturbed graves, or considered as continuation of Iron Age practices with no further analysis. The problem is further compounded by variable recording methods, which often means that, while disarticulated bone is mentioned in reports, few or no details of skeletal element, minimum number of individuals (MNI), or number of identified specimens (NISP) are available. Given the success of studies utilising pre-Roman disarticulated remains to reconstruct mortuary practices,¹³ this seems a missed opportunity.

The Roman Rural Settlement Project also recorded 65 instances of human remains within structured deposits.¹⁴ This represents a relatively small, but still significant, percentage of the structured deposits recorded in the database (14 per cent). This shows manipulation of human remains outside of a formal funerary context, while rare, was occurring. The presence of human remains within such deposits was most commonly observed in the late Iron Age/early Roman period and the mid to late Roman period, with a dip in the early-mid Roman period.¹⁵ This argues against the idea that such deposits are merely a holdover from earlier Iron Age practices. As with disarticulated material, structured deposits containing human remains have been recorded on a wide variety of site types and in a wide variety of features.¹⁶

The present paper examines an especially large and well preserved example of disarticulated human remains from within a structured deposit, recovered from a first-century a.d. quarry shaft in Ewell, Surrey, in order to illustrate the potential of such material for identifying and understanding minority mortuary practices in Roman Britain. Taphonomic and osteological evidence for mortuary treatment, particularly post-mortem manipulation, is discussed, as is the demographic profile of the assemblage. The assemblage is then compared to a wider study of human remains within Romano-British shaft and well deposits, as well as to Iron Age mortuary practices in the area in order to contextualise the deposit further and illustrate the unique nature of the post-mortem processes represented.

THE NESCOT QUARRY SHAFT, EWELL

The Nescot Former Animal Husbandry Centre was excavated by Pre-Construct Archaeology (PCA) in advance of development in 2015.¹⁷ The site was located in Ewell, Surrey, close to both Stane Street and the Roman roadside settlement at Ewell (fig. 1). While current knowledge of the settlement is limited, the area is thought to be a religious or ritual centre during the Roman period due to the large number of shafts containing structured deposits that have been found across the local area.¹⁸ The Nescot site itself was industrial in nature during the Roman period, comprising a large series of quarry pits dating from the first to third centuries a.d., and probably excavated for the extraction of chalk and flint. The quarrying activities appear to have started shortly after the conquest, and the large scale of the pits implies that the material was being exported via Stane Street or the Hogsmill River.¹⁹ The majority of the quarry pits contained deposits interpreted by the excavators as ‘special’ or ritual in nature, mostly comprising articulated and semi-articulated animal burials, particularly dogs, as well as a small number of human infant burials, coins and a variety of other artefacts such as an antler pick and a copper-alloy ear spoon.²⁰ The earliest quarry on the site [280] was backfilled around a.d. 77–100 (see below), and took the form of an oval shaft (fig 2) 12.82 m by 10.85 m across and 4.65 m deep, with a platform present approximately halfway down. The shaft was notable for containing a significant assemblage of human and animal remains, as well as pottery, metalwork, coins and gaming tokens.²¹ The lowest four contexts of the Nescot shaft contained 675 fragments of disarticulated human bone, representing a minimum of 20 individuals, as well as a single, prone, articulated skeleton. The faunal assemblage was also impressive and unusual, consisting of 140 dogs, 40 pigs, 12 cows, 22 horses, 34 sheep and goats, and four birds (almost all of which were, like the human bone, disarticulated). The faunal remains from the shaft are significant in being both one of the largest assemblages from a single shaft/well deposit in the period (NISP = 11,422, MNI = 280) and in the post-mortem treatment of the animals more generally. As such, they have been discussed elsewhere in more detail,²² in order to devote adequate space to their analysis, while this paper focuses primarily on the human remains. Much of the pottery within the shaft was interpreted by the excavators as having been ‘killed’ or intentionally broken at the time of deposition.²³

FIG. 1. Nescot site in relation to surrounding Roman activity, following Haslam and Haslam 2021.

FIG. 2. Section of Nescot quarry shaft.

Other human remains, most of them later in date (third century a.d.), were present on the Nescot site. One individual could be categorised as a ‘formal’ burial: a pair of feet and lower legs were recovered from a heavily truncated grave dated to a.d. 200–275. In addition, three articulated infants, dated to a.d. 250–400, were recovered from the backfill of quarry pits, along with a single adult incisor and isolated ulna. A further four articulated adult individuals were found within boundary ditches to the southwest of the quarry pits, dated broadly to a.d. 40–250.²⁴

Given that Roman Ewell is believed to have been a small roadside settlement,²⁵ it seems likely that the impressive depositions represented at Nescot may have involved a wider community. The proximity of the site to Stane Street would have made it relatively accessible to pilgrims and travellers, and the scale of deposition implies an event more akin to a festival than quiet, private ritual, although no temples have been discovered in the vicinity of the site.

This paper seeks to investigate the treatment of disarticulated human remains on the site, and as such will focus only on the lower four contexts of quarry [280]. A wider analysis of the shaft, its contents and its relationship to the other features at Nescot can be found elsewhere.²⁶

METHODOLOGY

While the meaning of such a deposit is very difficult to address, it is possible to reconstruct the processes and actions which created it. Taphonomic analysis, both at a macroscopic and a microscopic level, can help to identify the processes the bones were subject to, and thus help to identify the post-mortem treatment of the bodies. To that end the degree of erosion and abrasion were noted,²⁷ as was the presence and absence of root etching and animal gnawing, in order to create a general taphonomic profile for each context. Post-mortem breaks were sorted into those which occurred on fresh ‘wet’ bone and those which occurred on dry bone,²⁸ and the bones were all inspected for evidence of anthropomorphic activity such as cut marks. A bone representative index (BRI) was calculated²⁹ in order to investigate which bones were present. This was compared to previously established BRI values from a Roman cemetery in London in order to identify patterns that may be indicative of anthropogenic action as opposed to taphonomic damage. Additionally, eight long bones, seven from the disarticulated remains and one from the articulated individual, were subject to microscopic analysis. Sections of between 100 μ and 50 μ were cut from the midshaft of each bone, then examined at a magnification of ×10, in order to assess the level of bioerosion.³⁰ The level of bioerosion was measured using the Oxford Histological Index (OHI).³¹

RESULTS

The bones, both human and faunal, within the Nescot shaft were exceptionally well preserved at a macroscopic level, with very little abrasion or erosion present in any context. Evidence of weathering³² was observed in less than one per cent of the sample, mostly on animal bone. Evidence of gnawing was even rarer. Taken together, this indicated that the remains likely decomposed in a protected environment, sheltered from both animal scavenging and the elements. Of the 512 human bones which were fragmented, only 12 showed evidence of having been broken when the bone was ‘wet’, indicating that 98 per cent of fragmentation occurred sometime after decomposition. This is similar to the faunal remains from the same contexts (99 per cent of fragmentation occurred after decomposition). Only two cutmarks were observed on the human remains, both on cranial fragments, and no evidence of manual defleshing or dismemberment was present. This again was mirrored in the commingled faunal assemblage, in which only 0.20 per cent of the remains showed cut marks (n = 15).³³

Representation of different body parts

Bone Representative Index (BRI) is a measure of the survival and recovery of individual skeletal elements within an assemblage. The index can be useful for identifying if bodies have been subject to post-mortem manipulation.³⁴ The BRI represents the percentage of elements recovered compared to the number of elements that should be present if the remains were perfectly preserved. Ribs were not included in the BRI calculations, as the level of fragmentation made it difficult to calculate the minimum number of elements (MNE), and as such, no BRI could be determined. The results of the BRI can be seen in fig. 3. In order to establish a baseline for comparison, the BRI from West Tenter Street Cemetery, London³⁵ has been included in the figure as a representation of taphonomic preservation rates in a normative Roman assemblage.

FIG. 3. Bone representative index by context.

Across all four contexts within the quarry shaft at Nescot, bone survival was generally lower than that seen in the West Tenter Street cemetery. Given both the smaller sample size and the disarticulated nature of the Nescot assemblage, this is not surprising. However, the patterns of survival do not mirror the expected taphonomic model. Cranial elements are underrepresented in all four contexts, particularly within [393] and [350]. While in the normative sample from West Tenter Street cranial elements had a BRI of 46.6 per cent,³⁶ the Nescot shaft crania ranged from 29 per cent in [419] to 3 per cent in [393]. Femora and humeri were also under-represented within the shaft. At West Tenter Street the BRI of humeri and femora was 53 and 54 per cent respectively, yet within context [354], only 20 per cent of femora were present, and [350] contained only 12 per cent of humeri. Vertebrae, particularly thoracic vertebrae, were under-represented, with none present within contexts [350] and [354] despite a survival rate of 58 per cent at West Tenter Street. Even context [419], which had the highest recovery of thoracic vertebrae, had a BRI of only 26 per cent. In contrast, patellae and hand and foot phalanges were relatively common within [419] and [393], with only a 7 per cent difference between Nescot and West Tenter Street, although there were considerably fewer within [350] and [354]. These patterns are unlikely to be the result of natural taphonomic processes. Studies of multiple assemblages have shown that the proximal bones of limbs (femora and humeri) have higher survivability than those of the distal limbs (radii, ulnae, tibiae and fibulae), and that survivability of elements correlates with bone density, meaning that the bones of the crania should survive well, while smaller, less dense bones, such as hand and foot phalanges, should not.³⁷ These patterns indicate that the assemblage at Ewell was probably subject to post-mortem manipulation, with some elements being removed from the skeletons. This implies that the selection process was not about what to include within the shaft, but what to remove.

Histology

The level of bioerosion observable on bone at the microscopic level has been shown to correlate with early post-mortem treatment³⁸. While it is not clear if the destruction of the bone microstructure is caused by the visceral bacteria, or by bacteria from the environment, it has been shown that the level of bioerosion is directly correlated to the level of soft tissue decay that the body undergoes. Processes which speed up decomposition, such as dismemberment or subaerial excarnation, therefore result in lower levels of bioerosion. In contrast, bone from individuals who have been buried soon after death and have decomposed slowly have very high levels of bioerosion, often with none of the underlying microstructure surviving.³⁹

The articulated and disarticulated remains from the Nescot shaft displayed different levels of bioerosion, indicating that they were subject to different post-mortem trajectories. The sample taken from the articulated individual displayed very high levels of bioerosion, with none of the bone microstructure visible aside from the Haversian canals (fig. 4). This is consistent with the individual being subject to a lengthy decomposition process,⁴⁰ likely while buried within the shaft.

FIG. 4. Sample of the right humerus of the articulated skeleton at ×10 magnification. High levels of bioerosion have destroyed the microstructure of the bone, resulting in the dark colour and homogenous appearance.

A more mixed pattern of bioerosion was visible in all of the samples taken from the disarticulated material, with pockets of well-preserved bone present between destroyed areas (fig. 5). This implies a less extensive level of bacterial soft tissue decomposition than would occur during burial, but higher than would be expected from subaerial excarnation. This suggests that the bodies were allowed to decompose in a protected environment, closed enough to delay soft tissue loss from invertebrates, but open enough to result in a lower level of bacterial soft tissue decomposition than inhumation.⁴¹ Similar results from a histological study of the remains from Danebury Hillfort and Suddern Farm, in Hampshire, have been interpreted as evidence of decomposition in covered pits,⁴² however decomposition inside an above ground container or building could produce similar results.⁴³

FIG. 5. Sample of a left tibia from context [354] at ×10 magnification. The lighter colour and the presence of identifiable osteons and associated microstructures (blue arrows) in the centre of the sample show a pocket of good histological preservation, while the darker, more homogenous areas on the margin show that some bioerrosion has occurred.

Demographic profile

The age at death profile for the disarticulated remains is presented in table 1. Disarticulated material presents a variety of challenges and makes creating a demographic profile for the assemblage difficult, as each skeletal element must be assessed in isolation. The minimum number of individuals (MNI) for each context was calculated using duplicated elements as well as epiphyseal fusion in order to identify non-adult individuals. Age estimation for non-adults in ideal circumstances relies on looking at epiphyseal fusion in a variety of elements in order to narrow down possible age ranges, however since no two bones could be demonstrated to be from the same individual this was not possible. As such the age ranges are wider than can normally be achieved for human remains. ‘Adult’ has been taken in this case to indicate that the bone was fully fused, rather than to suggest a specific age. Estimations of biological sex were not possible for the majority of the disarticulated material and have not been included here.

Table 1 mni of the nescot shaft by context.

The demographic profile of the Nescot shaft is very similar across the contexts, with each including individuals from across the age profile (adult/adolescent/child/infant). The pattern does not reflect the patterns normally seen in Romano-British funerary populations from this period,⁴⁴ and has an unusually high number of adolescents (20 per cent). This is unlikely to be representative of a natural mortality curve, and implies a degree of selection in terms of who was subject to this post-mortem treatment.

The deposition process

The taphonomic and histological analysis both support the idea that the bodies represented at Nescot were left to decompose in a protected, covered environment before having certain portions removed. It is probable that a wooden or leather covering would have been fitted to the shaft, protecting the remains from both scavengers and the elements, while keeping them accessible. Crania, vertebrae and the bones of the proximal limbs (humeri and femora) appear to have been targeted for removal, although the patterns do vary within each context.

The question then becomes where were the bodies being processed? The BRI can give an indication, as not all of the joints in the body decay and disarticulate at the same rate. The joints of the hands, the toes, and those holding the patellae in place are among the first to decompose and become disarticulated.⁴⁵ The high proportion of such bones in the shaft implies that this is where the bodies were placed to decompose. This is further supported by the excavators’ description of a portion of the remains as ‘semi-articulated’, although the records are not detailed enough to assess the exact level of articulation or how common this was within the assemblage.⁴⁶ A wooden or leather cover could easily have been fitted onto the platform where the shaft narrowed, and would have provided enough protection to explain both the lack of weathering/scavenging and the histological results. The shaft being fitted with a cover is also supported by the small number of ‘pitfall’ faunal remains. ‘Pitfall’ faunal remains are the skeletons of small animals, such as frogs, toads and voles, which have fallen and died within deep features. No single context within the Nescot shaft had more than seven such individuals, indicating the features were unlikely to be open for any significant amount of time. The remains were found primarily clustered around the edges of the shaft, in comingled, disarticulated groups of human and faunal bone. This may imply that the remains were cleared to the side after skeletonisation, possibly to make room for new bodies (fig. 6).

FIG. 6. Intermixed human and animal remains clustered at the edge of the shaft. (© Pre-Construct Archaeology)

Two of the skeletal elements most conspicuously absent from the shaft, femora and crania, are among those most often found in isolation, both in Roman and Iron Age contexts.⁴⁷ It is possible that Nescot represents a step in a longer process, as the isolated skulls and femurs found in other contexts must have originally come from complete bodies.

Given the lack of evidence for manual defleshing, it is likely that the process represented at Nescot took place over an extended period of time, utilising natural decomposition processes before the removal of bones. This, combined with the similar age profile of the individuals within each of the four contexts, suggests a very intentional and controlled set of actions, only a small portion of which we can see archaeologically.

It is also worth considering that such a process would almost certainly produce an unpleasant smell. Given the number, size and backfill dates of the other quarry pits, the Nescot site may have been simultaneously in use as a large-scale mining operation, while the shaft was being used to process bodies. Even if the surrounding quarries were not in use, recovering skeletal elements would have required interacting with the corpse. Both scenarios imply a very different, and much closer, relationship with the dead than is normally envisaged within Roman Britain.

Timescale

There are two important intervals to consider when thinking about the processes occurring within the Nescot shaft: the time it took for the remains to skeletonise, and the time during which the shaft was in use as a place to process human remains.

The length of time it takes for a body to skeletonise is variable and dependent on the decomposition environment,⁴⁸ and as such it is difficult to provide an accurate estimation. Given the temperate climate of the United Kingdom, the process could have ranged from three weeks to several years.⁴⁹ It is worth noting that placement of bodies within a covered context would have resulted in a faster decomposition process than burial, while still protecting the remains from scavengers and weathering, as well as allowing relatively easy access to the remains.

Each context appears to represent a separate infill event, with the exception of context [354], which represents two. The presence of root etching on a series of the bones from each context suggests that they were shallowly buried for a period long enough for plants to grow before the next context was deposited. This also implies that the cover over the shaft was either not constantly utilised, or contained some holes, as plants need sunlight to grow. The presence of root etching on the articulated skeleton (which was located on the base of [354]) implies that a period of time elapsed before the deposition of the rest of [354]. While context [354] is the only context that demonstrably represents more than one deposition event, it should be kept in mind that if the shaft was covered it would prevent natural silting, and the manipulation of bones may likewise help to conceal multiple deposition events within each context.

While it is impossible to estimate the intervals between deposition events, the faunal assemblage may give some indication of seasonality. A large number of foetal, neonatal and juvenile animals were recovered (NISP = 911, MNI = 91), disarticulated and intermixed with the human remains. In particular, at least two horses were present within each context, and all individuals were unusually young, with nine of the ten either foetal or neonatal. Horses naturally breed in the spring and summer,⁵⁰ and Roman textual sources recommend breeding horses between the vernal equinox and summer solstice.⁵¹

The mortuary practices represented by the Nescot shaft appear to have been short-lived. A coin of the Emperor Vespasian dated a.d. 77–78, along with the pottery, indicated that the first episode of backfilling of the shaft took place in the late first century.⁵² The final instance of human remains within the shaft [350] was accompanied by a similar pottery assemblage dating to the late first to early second centuries. These dates are supported by the radiocarbon results, which place the human remains between 34 b.c. and a.d. 118.⁵³ While the curation of human remains cannot be discounted, particularly given the nature of the deposits, it seems from both the artefactual and radiocarbon evidence that the use of the shaft to process human remains stopped in the early second century at the latest.

HUMAN REMAINS IN ROMAN SHAFTS AND WELLS

In order to contextualise the human remains found within the Nescot shaft and the practices they represent, it is necessary to understand the trends visible within shaft and well deposits throughout Britain. The last large-scale study on shaft and well deposits in Roman Britain was published in 1985 by Gerald Wait, and very little work has been done on the human remains from these contexts. To address this problem, a wide-scale literature review was undertaken, searching for sites that listed human remains within shaft or well deposits. The location, date, MNI, state of the remains (articulated, disarticulated, isolated elements), and the demographic profile were recorded where available (table 2). This survey only looked for contexts referred to within their respective reports as shafts or wells; however, there is no set definition of a shaft, and more examples may exist that were categorised as pits or waterholes. A much wider study than was possible in the scope of this project would be needed to identify all those which have been categorised as pits or waterholes within the literature, but the sample provided is still useful for identifying general trends within the data. It is also worth noting that several of these examples were dug before the advent of modern recording and osteological practices, and as such may not accurately present the full assemblage. Unurned cremations in particular are likely to be under-represented due to the difficulty of recognising them without environmental sampling.

Table 2 comparison of romano-british shaft and well deposits.

Shaft and well deposits are a well-known aspect of the Roman ritual landscape in Britain.⁵⁴ While they were considered to be a straightforward continuation of an earlier Iron Age tradition,⁵⁵ more recent reanalysis has disputed this.⁵⁶ The use of shafts and wells within a ritual context was certainly not unknown to the Romans. Classical literature refers to the digging of a mundus, or a deep pit, into which offerings were cast during the initial founding of towns.⁵⁷ While the literature often refers to human remains as being a commonplace inclusion,⁵⁸ there are actually relatively few published examples. The review undertaken for this study identified 53 shafts on 39 sites that contained human remains.

There are several commonalities between the human remains found in these deposits. First, the number of individuals found in each deposit is generally quite small (fig. 7). In 36 of the 53 shafts (68 per cent), only a single individual was represented, and in all but four examples (92 per cent) the number of individuals represented was under five. This serves to highlight how unusual the shaft at Nescot is, which contained 21 individuals.

FIG. 7. Minimum number of individuals per shaft in the reviewed literature.

The most common form of human remains were fully articulated skeletons, present in 21 shafts. This was followed by isolated skeletal elements, present in 34 per cent (n = 18) of shafts. Skulls, both whole and fragmentary, made up 94 per cent (n = 17) of the examples of isolated skeletal elements. While recognition bias may play into this phenomenon, particularly in antiquarian records,⁵⁹ it is important to note that the head likely played an important symbolic role both in pre-conquest British⁶⁰ and Roman cultures.⁶¹ Multiple disarticulated remains were present in seven deposits, although again this should be taken with a degree of caution, as antiquarians may have ignored disarticulated material. The rarest form of remains were cremations, present in only five of the 53 shafts in the survey.

Post-mortem manipulation of remains within shafts is a more complicated issue to address, especially without investigating the material first-hand. Fully articulated bodies are likely to be primary burials which have remained undisturbed until excavation and have not been subject to manipulation. There were five exceptions to this. The first was the body found within the Swan Street well in London, which was missing the left arm, ribs and shoulder. Given the excellent preservation of the rest of the skeleton, this was interpreted by the excavators as ritual manipulation and deposition of the remains.⁶² A similar skeleton was discovered in a waterhole in Cambridgeshire which was missing the lower half of the body, as well as the left humerus and ulna.⁶³ The articulated body within the Nescot shaft was also missing the legs, and the lower portion of the right arm. The two wells in from Gloucester contained only the articulated torsos of the individuals. No cut marks were observed on the bones from any of the sites, which may indicate that the bodies were already in a state of decay when portions were removed,⁶⁴ although it should be noted that zooarchaeological studies have shown that a sufficiently skilled butcher would not necessarily leave cut marks on bone.⁶⁵

The isolated skeletal elements present a different challenge. They have clearly been manipulated, as they have been removed from the rest of the body. However, with skulls there is a question of when this occurred. The burial of a fleshed, albeit decapitated, head is likely to have been a primary deposit, while the presence of a skeletonised head is almost certainly a secondary deposit. While both might look similar archaeologically, the act of depositing a fleshed head would have likely had a very different impact on the people participating in, or watching, the act, than the deposition of dry bone. The presence of the mandible as well as the crania is the best indicator of the state of decomposition, as the temporal-mandibular joint breaks down relatively late in the decomposition process.⁶⁶ While many older site reports are lacking this level of detail, probable examples both of fleshed⁶⁷ and defleshed skulls⁶⁸ have been observed within shaft deposits. Only one example of manual defleshing has been noted, on the skull of a male individual recovered from a second-century shaft at Folly Lane.⁶⁹ In excess of 90 cut marks were observed, along with damage to the base of the cranium, suggesting it may have been mounted on a pole for display before deposition.⁷⁰ This type of violent manipulation is unusual amongst remains recovered from shafts.

Assessing post-mortem manipulation within disarticulated assemblages, to some extent, relies on being able to identify intentional human action. Several non-anthropogenic processes can result in disarticulation, including fluvial movement, animal activity, plant roots and the combined results of decomposition and gravity.⁷¹ Some anthropogenic activity can be easily recognised, such as the cut marks indicating manual defleshing observed on the skull from the Folly Lane shaft.⁷² However, in the majority of cases an in-depth understanding of taphonomic and deposit formation processes is the only way to access the causes of disarticulation, and thus the presence or absence of post-mortem modification. The frequency of bones present, as well as their taphonomic patterns, can provide information on these processes, but caution is required. While the absence of the smaller bones of the hands and feet is often used as an indicator of secondary burial,⁷³ there are a variety of factors which may lead to their loss, including excavation bias and differential preservation.⁷⁴ Indeed, within an articulated, well-preserved set of primary burials in Roman London, less than 10 per cent of the expected number of wrist bones were present, and less than 20 per cent of finger and toe bones.⁷⁵ Patterning within the BRI must be combined with a taphonomic profile in order to distinguish intentional action (such as the removal of certain bones) from incidental patterning due to preservation.⁷⁶ Unfortunately, it is almost impossible to access the evidence for post-mortem manipulation within disarticulated shaft deposits in the literature, as the deposits are rarely reported in the level of detail needed for such analysis.

All age groups are represented within the human remains found in shaft and well deposits, with adults being the most common, and adolescents the least (fig. 8). This roughly matches the cemetery data from the early Roman period in the South of England.⁷⁷ The Nescot shaft is the only deposit that represents all four age categories, and makes up just under half of all examples of adolescents found within shaft and well deposits.

FIG. 8. Age at death of individuals found within shaft and well deposits.

Of all the shaft and well deposits in the literature, a well excavated at the long-term settlement site at Oakridge, Hampshire, is the most directly comparable to Nescot, and has been studied in detail.⁷⁸ It is the only other reported example of a shaft containing over 20 individuals and, like Nescot, it comprised a mixture of articulated and disarticulated remains. As with Nescot, Oakridge was located on chalk, although unlike Nescot the site was primarily a settlement with continuous occupation from the early–middle Iron Age through to the Late Roman period.⁷⁹ The earliest backfill of the well containing human remains was dated to the late third/early fourth century a.d., over two centuries after the well was dug. The Oakridge well assemblage had a MNI of 26 (23 adults and three children). The stratigraphy shows that human remains were deposited over a long time (100+ years) with at least seven discrete depositions of human remains, with the largest deposit having a MNI of 14, and the smallest a MNI of one.⁸⁰ This is a direct contrast to the Nescot shaft, where the human remains were all deposited within a relatively short time period, and where there were relatively equal numbers of individuals within each deposit.⁸¹ Oakridge also differs greatly in that there is no evidence of post-mortem manipulation. The BRI of the Oakridge shaft had a normal distribution of elements, and the disarticulation was judged by the excavators to be the result of gravity and slumping within the deposit.⁸² The Oakridge well also contains evidence indicating that the shaft was left open for significant periods of time, including a large assemblage of ‘pitfall’ faunal remains, as well as the bones of fledgling swallows, indicating birds were nesting in or directly above the open shaft.⁸³ The much smaller assemblage of ‘pitfall’ animals at Nescot indicates that the shaft was likely open for a far shorter period. This is consistent with the histological results which imply that the shaft was covered.

DISCUSSION

The Nescot quarry shaft appears to be unique in three ways: first in the sheer density of the human remains found within the deposit; second, in the type of post-mortem manipulation observed; and third, in the wide range and patterning of age groups represented within the shaft. Such large deposits of human remains are unusual within Romano-British shaft and well deposits, as well as more generally within non-cemetery contexts.⁸⁴ Given the very short time span of deposition, it is possible that the actions represented by the Nescot shaft were in response to a specific event or social stress, rather than a regularly practised rite. This interpretation would be consistent with the overall trends evident among the deposition of human remains within structured deposits in Roman Britain⁸⁵ as well as with more general ideas of sacrifice and appeasement in the Roman world.⁸⁶

The treatment of the remains within the Nescot shaft also appears to be without direct parallel in Roman Britain. The process of controlled decomposition, followed by the removal of select elements, all taking place in the middle of a quarry, is undoubtedly an unusual occurrence in the period. Given the ubiquity of disarticulated material from Roman Britain, however, it is probable that minority rites that utilise post-mortem manipulation and selective deposition of remains are more common than is currently recognised. Further detailed, taphonomic studies of disarticulated assemblages are necessary for investigating this possibility.

The demography of the Nescot shaft also stands out when compared with other shaft and well deposits, being the only example in the literature of a shaft or well deposit containing such a range of age groups. It should also be noted that the Nescot shaft makes up half of all examples of adolescents recovered from shaft and well deposits reviewed for this study, and is the only example of a deposit with more than a single adolescent present. The mirroring of the ages of the individuals within each context is worth consideration, as this does not conform to a normative mortality pattern. It should be noted that all inhumations from this period display a degree of selection, as it was by no means the only or most prolific mortuary treatment at the time. However, the pattern seen at Nescot does not resemble those observed either in formal cemeteries or other shafts.⁸⁷ There are two possibilities to explain the patterning: either the bodies of specific individuals fitting the criteria were selected at death and treated in this manner, or individuals were specifically killed to produce bodies which fit the criteria. Aside from one possible example of blunt force trauma in a parietal fragment from context [354], there was no evidence on the bones indicating violent death in any of the individuals. However, there are a variety of ways to kill which would not be evident on the skeleton. Human sacrifice was prohibited in Britain after the Roman conquest,⁸⁸ although possible cases have been identified within the archaeological record, such as the bog bodies from Lindow Moss and Grewelthorpe Moor.⁸⁹ Sacrifice is almost impossible to demonstrate within the archaeological record, and Nescot does not show the signs of ‘overkill’ that are normally used to identify it,⁹⁰ but given the deeply unusual demography of the deposit, the possibility should be kept in mind.

Given the lack of comparable Roman material, it is easy to explain away the material from Nescot simply as ‘continuity’ of older Iron Age practices. The histological results are similar to the articulated pit burials from Danebury Hillfort and Suddern Farm,⁹¹ and the practice of manipulating human remains is well attested in pre-conquest Britain. Given the diversity and regionality of Iron Age mortuary practices, a more critical examination of the evidence is needed.

There is in fact very little evidence for the mortuary practices of Iron Age Surrey; only a small handful of sites have contained human remains, the majority of which were disarticulated and not reported on in detail.⁹² This is not unusual for Britain; it has been estimated that only between 3 and 6 per cent of the Iron Age population is visible in the archaeological record,⁹³ and as such the majority funerary ritual in the period must be one that leaves little or no trace. Most scholars attribute this phenomenon either to subaerial excarnation⁹⁴ or deposition in water.⁹⁵ It is unlikely that just one rite is responsible for the scarcity of remains from the Iron Age, as recent work has highlighted the diversity of post-mortem treatments bodies were subject to, even within the same site.⁹⁶

Perhaps more relevant to the material from Nescot is the re-use of grain-storage pits as a place to deposit human remains in the Iron Age. The distinction between a shaft, a well and a pit in the literature is somewhat arbitrary,⁹⁷ and it is easy to imagine the deposition of human remains within shafts and wells as a logical evolution of these practices. Indeed, much like Roman shaft deposits, the whole spectrum of human remains, from fully articulated skeletons to selectively redeposited single elements, has been observed within pits.⁹⁸ A thorough examination of a series of Iron Age sites by Justine Tracey⁹⁹ showed that there were no consistent patterns within BRIs of disarticulated assemblages, although a general lack of the smaller bones of the hands and feet was noted. The deposition of single elements was dominated by skulls and femora.¹⁰⁰ The evidence for exposure on these elements, however, was often limited, making subaerial excarnation an unlikely explanation,¹⁰¹ and histological studies from Danebury and Suddern Farm indicate that isolated skeletal elements were buried and allowed to decompose in the ground before being retrieved and redeposited.¹⁰² It is interesting then, that there have been no sites identified as the source of these isolated elements, meaning that either the processing sites have yet to be been found or identified in the archaeological record, or the process happened in such a way as to result in no archaeological traces. This is in direct contrast to the Nescot shaft, where skulls and femora are under-represented and have likely been removed for use or deposition elsewhere.

The sole example of Iron Age pits in the vicinity of Ewell containing disarticulated human bone occur at the Carshalton War Memorial Hospital in Sutton, which contained four such pits.¹⁰³ Three of these pits contained only a single bone, and all of the pits had a MNI of one. Storage pits with ‘structured deposits’ are not uncommon within the area,¹⁰⁴ and the survival of animal bone within these deposits shows that the absence of human material cannot solely be due to taphonomic conditions. Similarly, the scale of development within the south-east makes it unlikely that the lack of evidence is due to a lack of excavation. While it is true that the Nescot shaft shares some features with the Iron Age processing of the dead seen in other parts of the country, there are key differences, both in the skeletal elements present and the level of articulation. While it is very likely that pre-existing local practices influenced the activities at the Nescot shaft, the lack of evidence for similar practices in the area during the pre-conquest period make it unlikely that it was simply a direct continuation of established Iron Age rituals.

CONCLUSION

Studies of Roman Britain have acknowledged the complexity of the merging of Roman and Iron Age cultures in all spheres of archaeology.¹⁰⁵ Surveys of formal funerary sites, including cemeteries, have shown that there are regional nuances visible within the treatment of the dead, and that mortuary treatment is not as simple as ‘Roman’ or ‘Iron Age’ practices being utilised, but a complicated dialogue between the two.¹⁰⁶ However, human remains which fall outside the accepted idea of ‘Roman’ have to some extent been left out of this paradigm shift, and are still often uncritically interpreted as either continuity of Iron Age practice or accidental inclusions within fills. This is particularly evident in the case of disarticulated material, despite it being as common on mid–late Roman sites as it is on late Iron Age–early Roman sites.¹⁰⁷ The Nescot shaft shows the potential of detailed reanalysis of this material, revealing mortuary practices which have not been documented within Romano-British sites previously, as well as highlighting that despite appearances, such practices are far more complicated than the repetition of established Iron Age mortuary rituals.