Excavations at Eynsham Abbey 1989-92
Oxford Archaeology (South), 1999. https://doi.org/10.5284/1000203. How to cite using this DOI
Data copyright © Oxford Archaeology (South) unless otherwise stated
This work is licensed under the ADS Terms of Use and Access.
Primary contact
Dr
Anne
Dodd
Oxford Archaeology (South)
Janus House
Osney Mead
Oxford
OX2 0ES
UK
Tel: 01865 263800
Fax: 01865 793496
Resource identifiers
- ADS Collection: 277
- DOI:https://doi.org/10.5284/1000203
- How to cite using this DOI
Overview
Project Background
Location
Eynsham is situated on a gravel terrace between the confluences of the Chil Brook and the River Evenlode with the River Thames, some six miles west and upstream of Oxford, and three miles east of Witney. A natural fording point on the upper Thames, now Swinford Bridge, lies immediately east of the town. Thus the town sits on an important control point on the Saxon and medieval route out of Oxford to the west. This natural advantage has been exploited by settlement since the Bronze Age. The medieval Abbey precinct ultimately occupied almost all the land to the south of the historic town centre.
Historical background
Eynsham abbey was founded in AD 1005 under the authority of Æthelmaer, an elder statesman under King Æthelred II. Ælfric, one of the foremost theologians of the time was appointed the first abbott. The new abbey replaced an existing Minster church, which itself was in existence by AD 864, and possibly founded at least a century before that date.
The abbey was abandoned after the Conquest in 1066, and its properties were transferred to Stow, in Lincolnshire, as part of a reordering of the see by the then-Bishop of Lincoln Remegius. His uncompleted plans were changed by his successor Robert Bloet, which lead ultimately to the refoundation of the abbey, confirmed by a charter of Henry I in AD 1109. Thereafter the abbey functioned throughout the medieval period, and although becoming the third richest religious house in the county, it was hampered by periods of economic mismanagement. The surviving parts of the abbey Cartulary are mostly devoted to the landholdings of the abbey and their management.
The abbey was surrendered to the Crown in 1538, and the precinct and home farm passed into the ownership of the Stanley family, who lived on the site for a time. The site was finally acquired by a clothier, Thomas Jordan, in 1657, and soon after the demolition of the buildings of the inner ward of the precinct was completed.
Previous archaeological work
Several limited programmes of archaeological work have taken place within the abbey precinct since 1962. Furthermore, since early this century gravediggers have often encountered masonry - both disturbed and in situ - and tiles and burials in St Leonards and, more recently, St Peter's churchyards, both of which occupy part of the monastic precinct.
In 1963 and 1971 limited excavations were undertaken in the field immediately to the east of the Anglican cemetery. The discovery of part of an early medieval burial ground and evidence of robbed out structural walls ultimately led to the field being scheduled (SAM County No. 118). However, it is worth pointing out that, up until the 1989 excavations, there was no clear understanding of the disposition and character of any of the abbey buildings, or awareness of the survival of earlier medieval or prehistoric remains.
The Excavation
The excavations at Eynsham Abbey took place between 1989 and 1992. The main site incorporated parts of the cemeteries of St Peter's Roman Catholic Church and the adjacent St Leonard's Anglican Church, both of which lay within the known area of the monastic precinct core. The inevitable encroachment of modern graves into this key area prompted the Oxford Archaeological Unit to fund its own evaluation, which demonstrated the presence of significant monastic and pre-monastic archaeological deposits. As a result, English Heritage agreed to fund a full excavation of the area yet to be encroached upon by modern graves. As such the methodology of the excavation was driven to some degree by 'rescue' considerations, albeit within a detailed research framework. The principal elements if this research framework can be summarised as follows:
- To study the evidence for early /mid Saxon settlement and its status
- To examine the evidence for a middle Saxon minster
- To study the late Saxon and medieval abbey
- To provide a dated sequence of mid-late Saxon pottery which might
form a type series for Oxfordshire
- To collate the records and results of previous archaeological work on
the abbey, and to survey the earthworks
The research criteria were augmented after the first season of work, in the light of the results achieved so far, by the following new aims:
- To define the character extent and date of the prehistoric enclosure
- To examine the economy of the site in the Saxon, medieval and
post-medieval periods through the study of artefacts and ecofacts
- To assess any changes in assemblage structure through time
Excavation methodology
Total excavation of the site in a single season was not possible, both because of the lack of space for spoil and the need for St Peters church to have the facility of, and access to, their churchyard for various events during the year. The excavation area was therefore split into two. Trench 1, which lay wholly within the churchyard of St Leonards, was excavated in two stages in 1990. Trench 2, which lay within the grounds of St Peters, was dealt with in three stages in 1991-1992.
Both Trench 1 and Trench 2 were subject to evaluations prior to excavation. Both evaluations used a unique series of context numbers - 1 to 78 inclusive in the 1989 evaluation in Trench 1, and 2000 - 2148 inclusive in the 1990 evaluation in Trench 2. The archaeological deposits encountered and recorded in the evaluations were in almost all instances re-encountered and re-recorded in the main excavation.
In the main excavation the then standard Oxford Archaeological Unit recording system was used for all context, sample and finds recording. This used a system of supplementary numbers for layers within a feature, and letters for cuts within a linear feature. So 251/A/3 = layer number 3 within cut A of ditch 251. Similarly 659/-/6 = layer 6 within feature 659.
In all areas the topsoil, modern and late post-medieval layers were removed by a mechanical excavator, under archaeological supervision. Total depths of this material varied considerably, from as little as 0.20 m over the kitchen and southern part of the cellar, to approximately 1.0 m over parts of the cloister. Inevitably the process of mechanical removal will have reduced the rate of collection of artefacts from these deposits, but care was taken to retrieve as much material as possible.
A small supplementary excavation took place immediately west of the main excavation in 1993. This revealed elements of the western monastic ranges, not encountered in the main excavation. In the formulation of the Structure groups for the main excavation, these additional structures were incorporated. Therefore, while Structures 5200, and 5202-5 are present in the Structure database, they do not contain any contexts from the main excavation, nor are they represented in the plans.
Summary of excavation results
A total of six major phases of occupation were identified, beginning with a large rectangular Bronze Age enclosure ditch, surrounding a single surviving roundhouse and overlain by a probable Roman plough soil. The next phase of occupation was characterised by early-mid Saxon sunken-featured-buildings, five of which were identified. By the 8th century, when the documentary evidence points to the establishment of a minster church, a sequence of post-built buildings were evident, their orientation echoing that of the prehistoric enclosure. Animal bone evidence suggests a relatively abrupt transition from a rural settlement, producing its own animal resources, to a high status consumer community, as would be expected of a proto-monastic community centred on a minster church.
The foundation of the Abbey in 1005 led to a major rebuilding programme in stone. Interpretation of the ground plan has identified the possible south-west corner of the cloister, with a further courtyard to the south associated with a hall, a cellared range and domestic buildings, the whole complex bounded to the south by a perimeter ditch. The Abbey was abandoned after the Norman Conquest, yet the evidence suggests that although the precinct buildings were demolished, the site was not deserted for long, as the documentary evidence asserts. A large late-11th century kitchen was identified, with substantial hearths and nearby rubbish pits. This may indicate that although the Abbey was defunct, the estate was still controlled from a manor on the site.
The re-foundation of the Abbey in 1109 led to a complete rebuilding of the precinct, on a corrected west-east alignment. The layout followed the conventional Benedictine plan. The excavations exposed the southern half of the Cloisters, the Refectory, a large kitchen and an associated walled courtyard and cellared building, and a domestic block with attached latrine block. There was evidence that the first version of the kitchen was intended to have ovens set into three-quarter apses -or turrets- at the south-west and south-east corners, but for some reason the building was never used, and possibly never finished, and was replaced by a more conventional rectangular structure. This rebuilt kitchen survived in use until the second half of the 16th century and contained over 200 accumulated mortar floors, hearths and ashy occupation layers, totalling up to 0.7 m. in depth.
A complicated arrangement of drains and garderobe pits in the southern part of the site suggest that a high status building, possibly the guest quarters, lay immediately south of the excavated area. The stratigraphic evidence indicates that although the claustral buildings unsurprisingly survived to the Dissolution, albeit with some rebuilding suggested by the analysis of window glass and fragments of worked stone, some of the outlying buildings to the south were demolished by the end of the 15th century. Although this might indicate a decline in the Abbey's wealth, the animal bone analysis shows that a consistently high standard of the quality and variety of the diet was maintained throughout the monastic period. By far the largest assemblages of later medieval pottery came from courtyard middens and the infilling of the latrine pits and garderobes in the early 16th century. Prior to this there was clearly a more disciplined system of rubbish disposal, entailing its removal from the vicinity of the excavated area.
Significant collections of pottery, tile, architectural masonry, window and vessel glass and animal bone were recovered. The animal bone from the late Saxon and medieval monasteries represents a particularly valuable collection and has been the subject of a major study by the Faunal Remains Unit. The small finds included a number of items of national significance amongst a general assemblage of late Saxon and medieval building and domestic objects. The most notable of these comprised a fine late Saxon buckle plate in Hiberno-Celtic style, two late Saxon carved ivory fragments from devotional objects, two late Saxon stirrup strap mounts, a miniature vessel possibly used for incense in the late Saxon abbey, a miniature lead window with elaborate tracery, possibly from a lamp, and a fine medieval buckle. Technological analysis has been carried out on selected knives, copper alloy, glass and lead by the Ancient Monuments Laboratory. A selection of mortar and plaster samples representing different periods of construction have been the subject of a programme of analysis at Leicester University, and a number of medieval painted plaster fragments have also been studied.
Immediately following the Dissolution in 1538 the superstructure of the precinct buildings was demolished, with the exception of the kitchen, which continued in use, serving the new owner of the estate, Lord Stanley. During the late 16th/early 17th century, the semi-derelict monastic refectory was used as a burial ground for recusants, with the acquiescence of the Lord Stanley, also a supporter of the old religion. Three of these burials were found in the excavation. The second and final phase of the demolition of the abbey followed the sale of the estate to Thomas Jordan, a clothier, in the late 17th century.
Post-excavation
A limited amount of provisional stratigraphic analysis was undertaken immediately after the excavation, in the Spring of 1992. This included the formulation of the Phasing, and the grouping of contexts into Structures. Funding for the main post-excavation analysis was agreed with English Heritage in 1995, begun the same year, and completed in the summer of 1998.
The phasing
A sequence of stratigraphic phases was developed during the initial post-excavation analysis. All contexts were allotted to a single phase *. All structures were allotted to at least one phase. The dating of the phases was determined by a combination of the stratigraphic, ceramic, and artefactual evidence, refined periodically throughout the post-excavation process. Three 'markers' were provided by the historical events of the foundation of the abbey in AD 1005, the Conquest in AD 1066, and the Dissolution of the abbey in AD 1539.
The phases:
| Phase 1: | Bronze Age | (c.1200 B.C) |
| Phase 2a: | Early Medieval | (c.500 - 650 AD) |
| Phase 2b: | Early Medieval | (c.650 - 750 AD) |
| Phase 2c: | Early Medieval | (c.750 - 850 AD) |
| Phase 2d: | Early Medieval | (c.850 - 950 AD) |
| Phase 2e: | Early Medieval | (c.950 - 1005 AD) |
| Phase 2f: | Early Medieval | (1005 - 1066 AD) |
| Phase 3a: | Medieval | (1066 - c.1120 AD) |
| Phase 3b: | Medieval | (c.1120 - 1200 AD) |
| Phase 3c: | Medieval | (c.1200 - 1330 AD) |
| Phase 3d: | Medieval | (c.1330 - 1450 AD) |
| Phase 3e: | Medieval | (c.1450 - 1539 AD) |
| Phase 4a: | Post-Medieval | (1539 - c 1660) |
* Note: A small number of contexts in Period 2 could not be closely phased and are given as Phase 2x in the databases. Contexts assigned Phase 2x fall somewhere within the range Phase 2a to Phase 2e, but pre-date Phase 2f.
The structures
A sequence of structure numbers was developed during the post-excavation analysis to assist interpretation, each number defining one of the following:
- A distinct building footprint and associated internal/external contexts
- A definable structure, for instance a drain
- A spatially definable area, for instance the kitchen garden
- A spatially and temporally defined area, for instance the outer courtyard in Phase 3c
- A group of functionally linked contexts, for instance the mortar-mixers, or the mid Saxon hearths
Note: some contexts, interpreted as probably being associated with a structure, but not necessarily physically within the footprint of the structure, were assigned that structure number with an `a' suffix in the structure database. This suffix is disregarded in the plan index.
Approximately 83% of all contexts were each allotted to one of the 118 structures or structure groups.
Structure Number sequence:
| 5150-5151 | - | Prehistoric |
| 5160-5176 | - | Early/Middle Saxon |
| 5180-5187 | - | Late Saxon abbey |
| 5190-5199 | - | Early Norman Transition |
| 5200-5250 | - | Medieval Abbey and post-Dissolution activity |
Specialist Summaries
The small finds
The small finds included a number of items of national significance amongst a varied assemblage of late Saxon and medieval building and domestic objects. In general the assemblage reflects the situation of the excavation within the abbey precinct, covering mainly areas of secular activity, but including part of the cloister. Consequently the proportion of objects relating to the religious activities of the monks is modest.
The distribution of finds, particularly of the later medieval period, was greatly affected by the post-Dissolution demolition process. The large latrine pits were used as rubbish pits, and contained quantities of finds clearly originating from elsewhere in the precinct.
The condition of the finds was generally good, primarily due to the uncorrosive geology of the area. On the other hand, the free draining of the gravel natural meant that preservation of organic material by waterlogging was almost totally absent.
The window came
A total of 877 g of lead window came was recovered from the site. The majority of the came was recovered from Phase 4a structures, in particularly the garderobe structure 5220. This mirrors the distribution of window glass and supports the idea that windows were stripped for valuable lead following the Dissolution. Most of the cames are twisted and fragmentary suggesting evidence of dismantling.
Classification of came types is based on the typology established by Barry Knight (1985, 154-156).
The window glass
Quantification
The assemblage comprises a total of 856 fragments. Of these 132 (c.15.5%) were painted. The remainder of the glass was unpainted or had suffered surface loss or undergone complete loss of glassy state.
Methods of examination
Glass was examined by eye, and where necessary with the aid of a hand magnifier (x2). Superficial cleaning of fragments was only undertaken where glass surface and paint were stable, using a scalpel and a soft bristle brush.
Preservation
The majority of the glass was fragmentary and severely weathered, in some cases to the extent of complete breakdown through strain cracking. The term 'strain cracking' is preferred to 'de-vitrification' since the latter refers to the products of glass-making processes rather than depositional weathering processes (Newton and Davison 1989, 4). In may cases different weathering types ocurred together within a single piece of glass. These included iridescence, surface loss and oxide-staining, all of which have contributed to subsequent strain cracking. In some cases, the exterior surface was pitted, a condition caused by atmospheric weathering processes and an indication that some of the windows spent a substantial amount of time exposed to the air, presumably in situ as windows. In some instances, concretion of the surrounding soil matrix has occurred, preventing safe cleaning and diagnostic visual examination.
The roof and floor tile
The total quantity of tile of all types recovered from excavations at Eynsham weighed 708,116 gms. This comprised 92% roof tile by weight, and approximately 7.5% floor-tile, including both decorated and plain. Three distinct decorated floor-tile groups are discernible, including the 'Stabbed Wessex' type and a later two-colour series, well-known in the Oxford region. The third industry produced thick inlaid tiles, but probably only a limited number were supplied. In addition to these industries, there are two probably very early fragments of hand-incised tile, and two types of plain floor-tile. The roof-tile is of some interest and shows a wide diversity of ridge-tile types. Some fabrics are common to both floor- and ridge-tile.
The Post-Roman pottery
Paul Blinkhorn
The Anglo-Saxon and medieval pottery from Eynsham Abbey was recorded in digital form, using a database. The material was grouped by context, then further subdivided into the categories listed below. Wherever possible, numeric rather than alphabetical codes were used, and this considerably eased the later analysis of the data, being more 'computer friendly'.
The Database
The database fields were as follows:
CONTEXT: The context number of the assemblage. As used in the site records.
NO: Number of sherds. Each entry, with the exception of plain body sherds, represented a single vessel. Plain body sherds, were, however, grouped together by simply by fabric type, as they are of limited value analytical value other than as a dating tool.
WT: Weight of sherds. In grammes.
FABRIC: Fabric type. The middle and late Saxon and medieval pottery was sub-divided into a series of fabrics, based on known wares from the Oxford type-series devised by Maureen Mellor (1994), or, if previously unknown in the area, referred to other published examples. A brief description of each is below. The early Anglo-Saxon pottery types were individually defined. Such wares can be paralleled at other sites in the region, but there is no guarantee that they were made by the same potters; there is plenty of evidence that such pottery was made at domestic level i.e. by individuals within a settlement, or even by members of individual households within a settlement. The decorated wares may have been made by specialist potters who perhaps travelled around an area, making pots 'to order'.
RIMF: Rimform. The shape of the pot-rim in cross-section. This has the potential to provide chronological information, as the shape often changed over time, and distinctive types can sometimes be closely dated.
RIMD: Rim diameter. The diameter of the pot rim. Previous work has shown that the size of a pot rim is often directly correlated to the capacity of the vessel when it was complete. Thus, it is possible to analyze the range of pot-sizes in use from the diameters of the rimsherds, which can sometimes allow identification of different functional areas of a site (i.e. storage vs cookery).
EVE: Estimated Vessel Equivalent. This is a measure of the 'completeness' of rimsherds, expressed as a percentage of the original (100%) circumference. It is used as a way of expressing the amount of pottery in an assemblage, and is generally a more accurate measure than number or weight of sherds, as these are affected by the size and wall-thickness of the original pot, whereas rim-sherds are not.
BASE: Base type. This is recorded and used in a similar fashion to RIMF.
BASED: Base Diameter. Similarly, this can be used in a similar fashion to RIMD.
VESSEL: Vessel Type. Usually, it is possible to discern the original vessel type (jar, bowl, jug, bottle etc) by the size and shape of the rim, as well as other factors such as handles, spouts, etc. This information can be useful in identifying the use of different parts of a site, and also how social and economic factors changed over time.
INC: Incised decoration. This can take many forms, such as freehand tooling (eg. wavy lines) or rouletting (bands of decoration formed by a roller with a design cut into it). This can provide information with respect to chronology or vessel function, as utilitarian wares such as cooking pots tend not to be decorated, but tablewares, such as jugs, often are.
APP: Applied decoration. Plastic decoration applied to the outside of a leather-hard pot. This can also take many forms, from simple strips to moulded human and animal figures. It can provide similar information as INC.
STAMP: Stamped Decoration. Decoration made by pressing a carved die-face into the still-wet clay of a pot. Particularly used during the early Saxon period, and can provide similar information as INC and APP.
HANDLE: Handle type. Attached handles, ranging from plain ears ('lugs') through to sophisticated strap and rod types with inlaid and incised decoration. Can again provide chronological and functional information, as some vessel types have very distinctive handles.
SPOUT: Spout type. These, like handles, come in many forms, and can again provide chronological and functional information.
FOOT: Foot type. Some vessels, such as early medieval Oxford ware jugs, have feet applied to the base of the pot. These can provide chronological and functional information.
GLAZECOL: Glaze Colour. Medieval and later vessels often have glazed surfaces. The colour and composition of these can provide chronological and functional information.
SLIPCOL: Slip Colour. Some vessels are decorated with coloured liquid clays (slips) which were applied to the surface of the vessel, and contrast with the colour of the body clay and glazes. The colour and composition of these can provide chronological and functional information.
SLIPPAT: Slip Pattern. Slip can be used in a variety of ways: vessels can have an all-over covering, or be painted in geometric patterns. These can provide chronological and functional information.
CERAMDATE: Context-specific pottery assemblage date. The chronology given to an assemblage based on the collation of all the above data for that context. These are later refined with respect to the site phasing based on stratigraphic relationships, etc, but are a useful initial guide to the site chronology, and can allow the refinement of the chronology of both the site and the pottery.
SITEPHASE: Site phase chronology. The final quantification and analysis of the pottery is based on this, which is the same used for the actual features which the pottery came from.
COMMENTS: A text field which allows notes to be made of significant or unusual features of the pottery in that database field.
The Pottery Fabrics
Early-middle Saxon handmade wares:
F1: Chaff-temper
F2: Oolitic Limestone
F3: Limestone
F4: Coarse Quartz
F5: Fine Quartz
F6: Limestone & sandstone
F7: Coarse quartz and ironstone
F8: Sparse Quartz and ironstone
F9: Iron-rich Quartz
F10: Ironstone
Middle Saxon:
F95: Ipswich Ware. (Hurst 1976; Blinkhorn in press)
Late Saxon and Medieval
Where appropriate, the OX- prefix code of the Oxford type-series (Mellor 1994) are given, otherwise they are referenced to a major publication. Unreferenced wares are unique to Eynsham Abbey.
F100: St. Neots Ware type T1(1) (Oxford Fabric OXR). Denham 1985.
F101: Late Saxon Oxford Ware (OXB)
F200: Late Saxon/Early Medieval West Oxfordshire ware and Early medieval Oxford ware (OXAC)
F201: St. Neots Ware type T1(2). Denham 1985
F202: Early Medieval-Late Medieval East Wiltshire Ware (OXBF)
F205: Stamford Ware (Kilmurry 1980)
F300: Late Saxon/Medieval Oxford Ware (OXY)
F301: Banbury Ware (OX234)
F302: Late Saxon and Early Medieval South-West Oxfordshire Ware (OXAQ)
F303: Ferruginous sandy coarseware.
F350: Nuneaton 'A' Ware (Mayes and Scott 1984)
F352: Brill/Boarstall Ware (OXAM)
F353: Potterspury ware (OX68)
F354: Developed Stamford ware (Kilmurry 1980)
F355: Minety type Ware (OXBB)
F356: Wychwood Ware (OXCX)
F357: Brill/Boarstall types (OXBX)
F362: Abingdon ware (OXAG)
F364: North French Whitewares (Hurst et al. 1986)
F365: Medieval Sandy Whiteware
F401: Red Earthenware I (OX68A)
F402: Raeren Stoneware (Gaimster 1997)
F403: Tudor Green wares (OXBN)
F404: Cistercian wares (Crossley 1990)
F405: Frechen Stoneware (Gaimster 1997)
F406: Farnborough Hill Tudor Green ware (OXBG)
F407: Langewehe Stoneware (Gaimster 1997)
F408: 'Overfired' Brill/Boarstall wares (OXAP)
F410: ?Brill/Boarstall type Tudor Green wares (OXAM)
F411: Late Medieval Transitional ware
F412: Midland Black ware (Crossley 1990)
F413: Westerwald/Cologne stoneware (Gaimster 1997)
F414: Staffs Manganese Wares. (Crossley 1990)
F415: Seville tin-glazed earthenware (Hurst et al. 1986)
F416: Metropolitan Slipware. (Crossley 1990).
F417:English Tin-Glazed Earthenware (Orton 1988).
F418: Creamware (Jennings 1981)
F419: Valencian Lustreware (Hurst et al. 1986)
F420: Martincamp fabric I (Jennings 1981)
F421: Oxford late medieval green glazed ware. (OXAV)
F423: Cistercian type ware (OXDW).
F424: Siegburg Stoneware.(Gaimster 1997)
F425: Red earthenware II.(OXDR)
F426: Red earthenware III
F428: Porcelain.(Crossley 1990)
F429: London Stoneware.(Crossley 1990)
F430: Red earthenware IV (OXDG)
F431: Brill Yellow Wares. (Crossley 1990)
F434: ?French Earthenware (Hurst et al. 1986)
F436: Red earthenware V
F456: Surrey Whitewares (Pearce and Vince 1992).
The Assemblages
The earliest post-Roman pottery at Eynsham Abbey was the hand-made early Saxon wares. They are typical of the pottery of the region, comprising mainly simple, undecorated jar forms, along with a few bowls, and a small number of decorated vessels, most notably a near-complete vessel with stamp decoration and long bosses. It is only possible to date early Saxon pottery by the styles of the decorated vessels, and the analysis of these suggested that whilst there was a small amount of later fifth-century material, the main period of Anglo-Saxon activity dated to the sixth century. This phase probably continued into the seventh century, but as nearly all the pottery of this period was undecorated, it is difficult to identify such groups.
During the eight century, the inhabitants of Oxfordshire appear to have stopped making and using pottery, although small quantities of Ipswich ware were imported, probably as containers for traded goods.
By the mid-ninth century, pottery was again being used in quantity at the site. The wares, mainly St Neots type and Cotswolds type, were mainly simple unglazed jars, used for cookery and other basic functions, although a small number of bowls and lamps were also noted. The pottery remained more or less the same until the later part of the 11th century, when a wider range of types began to be used (such as Wiltshire wares and Oxfordshire sandy coarsewares), and by the end of that period, glazed tripod pitchers of the Oxford ware tradition first came to the site. These were very much the minority, as is the case nationally, and the most common pot types remained the simple jar.
The next major change came in the 13th century, with the introduction of Brill/Boarstall wares. These fine, skilfully-made pots came to dominate the assemblage during the medieval period, particularly the highly decorated jugs with applied decoration, rouletting and coloured slips. By the 14th century, the range of pottery types, in terms of manufacturing centres and vessel forms, began to increase significantly. This again is part of a national pattern, with the increasing sophistication of medieval society, particularly with respect to formal dining and, in the cases of abbeys etc, an increase in mass catering. This is evidenced by the increasing presence in the archaeological record of vessels associated with cookery, such as dripping dishes, which were used to catch the fat dripping from spit-roasted meat for use as the basis for sauces, etc, and also tablewares, such as skillets, probably used for individual servings of the sauces, condiment/sauce bottles, and drinking cups and bowls.
By the Dissolution, a small number of imported vessels had arrived at the Abbey. German stoneware beermugs were noted, although these are common finds throughout Eur Europe, and were imported into Britain by the million. Martincamp flasks, from northern France, which were possibly used to transported Calvados, were also noted, as well as a complete Spanish Albarello, often called 'drug jars', but probably more commonly used for transporting marmalade or sugared fruits, spices, etc.
Two vessels deserve special mention. The first is a green-glazed mug from an unknown, but possibly German source, which has become known as the 'Green Man' pot. It appears to be unique, and has a modelled face on the front with a tree growing from its forehead, and four feet, three of which have faces and the fourth is possibly a representation of a hedgehog. It is incised with shapes which could be flowers or butterflies, and the handle, on the opposite side to the face, resembles a plait of hair.
The second vessel is a horse-and-rider aquamanile. Such vessels were hollow, and would have perhaps been the centrepiece of the late medieval banqueting table. They were designed to hold water for the washing of hands between courses. They are rare but universal finds throughout medieval Britain. The Eynsham example, although incomplete, was of an armoured knight with lance and shield, mounted on a charger. The water was poured out through a spout in the horse's mouth.
These two pots may appear somewhat incongruous at a religious house, but it is known that the Abbot entertained the nobility, and the pots may have been used on such formal occasions, rather than in the everyday cycle of monastic life.
Bibliography
Blinkhorn, PW, in press) The Ipswich Ware Project: Ceramics, Trade and Society in Middle Saxon England Medieval Pottery Research Group Special Paper
Crossley, D, 1990 Post-Medieval Archaeology in Britain Leicester University Press
Denham, V, 1985 The Pottery in JH Williams, M Shaw, and V Denham, Middle Saxon Palaces in Northampton Northampton Development Corp Monog 4
Gaimster, D, 1997 German Stoneware British Museum Publications
Hurst, JG, 1976 The Pottery in Wilson DM (ed) The Archaeology of Anglo-Saxon England, 283-348
Hurst JG, Neal, DS and Van Beuningen, 1986 Pottery Traded in North-Western Europe Rotterdam Papers 6
Jennings, S , 1981 Eighteen centuries of pottery from Norwich East Anglian Archaeol 13
Kilmurry, K, 1980 The Pottery industry of Stamford, Lincs c. 900-1250 British Archaeological Reports British Series 84
Mayes, P and Scott, K, 1984 Pottery Kilns at Chilvers Coton, Nuneaton Soc Medieval Archaeol Monog Ser 10
Mellor, M, 1994 Oxford Pottery: A Synthesis of middle and late Saxon, medieval and early post-medieval pottery in the Oxford Region Oxoniensia 59, 17-217
Orton, C, 1988 Post-Roman Pottery in P Hinton (ed.) Excavations in Southwark 1973-76 and Lambeth !973-79. MoLAS and DGLA Joint Publication 3, 295-364
Pearce, J and Vince, A, 1988 A Dated Type-Series of London Medieval Pottery. Part 4: Surrey Whitewares London and Middlesex Archaeol Soc Special Paper 10
The Oyster Shells
All of the sampled contexts yielded oyster (Ostrea edulis L.) shells but quantities were variable, with context 226 yielding the largest number of shells. Each context shell assemblage was dominated by Ostrea edulis with other species (nonmarine and marine) forming a very small component of the total.
The Assemblage
The shells examined were selected from six contexts:
Table 1
| Ctx | Type | Structure | Period | Phase | Interpretation |
| 394 | Pit | 5164B | 2 | B | Mid 8th century Rubbish Pit |
| 3095 | Layer | 2 | C | 9th/10th century Soil layer | |
| 278 | Pit | 5172 | 2 | C | 10th century Cess Pit |
| 226 | Pit | 3 | B | 11th/12th century Shell dump | |
| 659 | Cellar | 5220 | 4 | A | 13th century Garderobe pit (backfilled) |
| 3522 | Pit | 5241 | 3 | E | 15th century Rubbish pit |
Table 1: Contexts from which shells were selected for examination.
Table 2
| Taxon | 394 | 3095 | 278 | 226 | 659 | 3522 |
| Aegopinella nitidula | 1 | |||||
| Trichia hispida | 10 | |||||
| Candidula intersecta | 2 | |||||
| Cepaea hortensis | 1 | |||||
| Cepaea nemoralis | 53 | |||||
| Helix aspersa | 3 | 3 | ||||
| Littorina littorea | 1 | 1 | ||||
| Hinia reticulata | 1 | |||||
| Buccinum undatum | 1 frag. | 1 | ||||
| Mytilus edulis | 1+ frags | 19+frag | 8 | |||
| Anomia ephippium | 27 | 1 | ||||
| Cerastoderma edule | 1 | |||||
| fossil items | 7 | 1 | ||||
| bone | 1 | 15 | 3 | |||
| fish vertebrae | 1 |
Table 2: Distribution by context of all molluscan (except Ostrea edulis) and bone finds.
Table 3 summarises data for oysters from context 226 only. Numbers of right and left valves in each of the 13 bags are given together with the size range, numbers of shells at upper and lower size limits and modal size of oysters for each bag. The data show that 868 right valves and 843 left valves of Ostrea edulis were collected from the deposit. A mean modal size for right and left valves is given. Work elsewhere (Winder, 1992) has shown that right valves survive in larger numbers and are generally better preserved but morphometric measurements are best carried out on the larger left valve since the right (upper) valve in living oysters is usually inset. The shell dump thus contains a deposit of at least 843 oysters whose left valve size ranges from 92 - 31 mm with an average modal size of 58 mm which in this assemblage is equivalent to an age of 3-5 years.
| Box/bag | Nos. r.v. | size mm | Nos. max. | Nos. min | Mode mm | Nos l.v. | size mm | Nos max | Nos min | Mode mm | |
| S18/1 | 68 | 77-40 | 6 | 6 | 51 | 75 | 80-37 | 8 | 8 | 58 | |
| S18/2 | 100 | 69-30 | 14 | 13 | 55 | 77 | 79-41 | 8 | 8 | 61 | |
| S18/3 | 74 | 82-33 | 7 | 12 | 52 | 61 | 71-48 | 8 | 3 | 62 | |
| S19/1 | 75 | 82-31 | 6 | 7 | 54 | 52 | 82-44 | 8 | 3 | 59 | |
| S19/2 | 61 | 77-37 | 4 | 5 | 54 | 66 | 80-43 | 13 | 6 | 58 | |
| S19/3 | 75 | 69-32 | 9 | 7 | 56 | 68 | 74-40 | 18 | 1 | 61 | |
| S20/1 | 32 | 65-30 | 5 | 3 | 55 | 62 | 78-44 | 3 | 3 | 55 | |
| S20/2 | 62 | 67-34 | 6 | 7 | 52 | 65 | 73-45 | 10 | 10 | 60 | |
| S20/3 | 46 | 67-39 | 5 | 9 | 51 | 59 | 78-46 | 9 | 4 | 52 | |
| S21/1 | 63 | 61-32 | 7 | 10 | 53 | 70 | 67-46 | 10 | 7 | 58 | |
| S21/2 | 76 | 72-36 | 9 | 9 | 46 | 61 | 77-41 | 8 | 9 | 57 | |
| S21/3 | 57 | 65-27 | 9 | 1 | 53 | 63 | 74-34 | 16 | 4 | 60 | |
| SH.2 | 79 | 70-30 | 8 | 3 | 54 | 64 | 92-31 | 5 | 2 | 56 | |
| Total | 868 | mean | = 53 | 843 | mean | = 58 | |||||
Table 3: Numbers of right and left valves of Ostrea edulis collected from context 226 with summarised size data.
Table 4
| CTX | R.V. | L.V. | Comments |
| 394 | 10 | 5 | Generally poor condition, flakiness, some with ferrug. stains |
| 3095 | 11 | 10 | Most shells >70mm, thick and flaky |
| 278 | 4 | 2 | Shells are flaky |
| 659 | 187 | 221 | Majority of shells with ferruginous staining and flaky. size range observed to be similar to shells in context 226 sample. |
| 3522 | 37 | 50 | Contains thick shells, some flaky. |
Table 4: Numbers of oyster shells from contexts other than 226 with observations on condition and size
Environmental sampling
Summary
During the excavations extensive sampling and floatation was undertaken for charred plant remains. Samples were also sieved on site for the retrieval of calcium phosphate mineralised plant remains. The sampling was designed to enable the investigation of changes in economic plants during the history of this high status site. Features sampled cover a long period of occupation from Anglo-Saxon features through to the post-dissolution occupation of the site.
A total of 118 flots were assessed for charred plant remains. A further 13 processed and three unprocessed samples were assessed for mineralised remains. On the basis of this assessment 32 samples were subjected to full analysis. The samples selected were felt to represent the full range of botanical material available for each phase. A further seven samples were selected for the analysis of their charcoal content.
(See text file Cpr.txt)
Select Project Bibliography
Aston, M, 1993 Monasteries, London
Blair, W J, 1994 Anglo-Saxon Oxfordshire, Oxford
Coppack, G, 1990 Abbeys and Priories, London: English Heritage
Gordon, E, 1990 Eynsham Abbey - A Small Window into a Large Room, Chichester
