FROM STUDIUM TO STATION REWLEY ABBEY AND REWLEY ROAD STATION, OXFORD

By Julian Munby, Andy Simmonds, Ric Tyler and David RPWilkinson

With contributions by

Lance Adlam, Martin Allen, John Cotter, Rebecca Devaney, Rosemary Grant, John Higgins, Louise Loe, Fiona Roe, Ian Scott, Edmund Simons, Lena Strid, John Tibbles, and Hugh Willmott

Illustrations by

Hannah Brown, Elizabeth de Gaetano, Lucy Martin, Ric Tyler, and Magdalena Wachnik

Oxford Archaeology Occasional Paper Number 16 2007 From Studium to Station Rewley Abbey and Rewley Road station, Oxford

Edited by Edward Biddulph and Anne Dodd

Front cover: The London, Midland and Scottishrailway station in 1940 No. 26886, collection DY q National Railway Museum

Backcover: View of Rewley Abbey drawn by JBMalchair in 1772

ISBN 978 0904220 40 7 q 2007 Oxford ArchaeologicalUnit

Typeset and printed in Europebythe Alden Group, Oxfordshire Contents

List of Figures ...... iv List of Plates...... v List of Tables ...... vi Acknowledgements ...... vii Summary...... viii

Introduction David RPWilkinson...... 1

Historical Background Julian Munby...... 6

Archaeological Description Andy Simmonds ...... 10

Finds ...... 33 The medieval coin Martin Allen...... 33 Jettons Edmund Simons ...... 33 Medieval andpost-medieval pottery John Cotter ...... 33 Claytobacco pipes DavidHiggins ...... 43 The Ceramic buildingmaterials John Tibbles, with acontribution by John Cotter...... 52 Glass Hugh Willmott ...... 57 Building stone Fiona Roe, with anote on the architectural stone by Julian Munby...... 59 Other finds Metalwork by Ian Scott,worked bone by Rosemary Grant, whetstone by Fiona Roe and flint by RebeccaDevaney...... 59 Previous finds Julian Munby ...... 61

Human Skeletal Remains Georgina Slaterand Louise Loe ...... 61

Animal Bone Lena Strid...... 63

The Rewley Road Station:The Comingofthe Railway Ric Tyler, with contributions by Lance Adlam andDavidHiggins...... 64 Dismantling and re-erecting the station Lance Adlam ...... 93 Amid-19th century groupofclay tobacco pipes DavidHiggins ...... 99

Bibliography ...... 102

iii List of Figures

Figure 1Location of the Study Area ...... 2 Figure 2The site of Rewley Abbey as it appears on RalphAgas’ map of Oxford,1578 ...... 3 Figure 3The site of Rewley Abbey as it appears on David Loggan’s map of Oxford, 1675...... 3 Figure 4Plan of the StudyAreawith trench locations ...... 4 Figure 5Aview of Rewley Abbeyfrom the north, drawn by MBurghers in 1720 ...... 9 Figure 6Trench 1993/1: section through the riverside wall and related deposits...... 13 Figure 7The ruins of Rewley Abbeyseen from the north-west, drawn by WTurner in 1851...... 15 Figure 8Aview of the entrance to Rewley Abbey,drawn by MBurghers in 1720...... 15 Figure 9Plan of the main archaeologicalfeatures in the abbey precinct...... 16 Figure 10 Aview of the ruins of RewleyAbbeyfrom the south, drawn by MBurghers in 1720...... 17 Figure 11 Plan of trenches investigating the abbey church ...... 18 Figure 12 Plan of the chapel/brewhouse in Trenches1986/1 and 1986/2...... 21 Figure 13 Aview of the ruins of RewleyAbbeyfrom the west, drawn by MBurghers in 1720 ...... 22 Figure 14 Plan and section of the western end of the north range, Trench 1994/17...... 24 Figure 15 WTurner’s drawing of the room at the westernend of the north range...... 25 Figure 16 Plan and section of the reredorter in Trench 1994/16 ...... 26 Figure 17 Arepresentation of Rewley Abbey ...... 30 Figure 18 De Gomme’s map of the Civil War defences of Oxford, drawn in 1644...... 31 Figure 19 Section through the Civil War defencesexposed in the 1999 watching brief ...... 32 Figure 20 Pottery ...... 41 Figure 21 Clay pipes,nos 1–18 ...... 49 Figure 22 Clay pipes,nos 19–35 ...... 50 Figure 23 Clay pipes,nos 36–41 ...... 51 Figure 24 Schematic map of rail routes around Oxford at the end of the 19th century ...... 66 Figure 25 Extract fromRSHoggar’s 1850 map of Oxford indicating the proposed L&NWR terminus site...... 67 Figure 26 Interpretative reconstruction of foundation column Fatthe south-west corner of the porte-coche` re, investigated in 1995 ...... 68 Figure 27 Details of verticalcomponents of primarystructure...... 69 Figure 28 Details of horizontal components of primarystructure...... 70 Figure 29 (a) Schematic layout of structural components; (b) plan view of principal walling components, as recorded in 1999...... 71 Figure 30 Details of cast-iron column base mouldings ...... 72 Figure 31 Isometric projection of aC2b column head and extension piece (E), with primary truss T1 and secondary side-beams(B2)...... 73 Figure 32 (a) Front elevation as recorded; (b) front elevation (1914 reconstruction)...... 77 Figure 33 Standard Crystal Palace timber wall panel...... 78 Figure 34 (a) Longitudinal cross-section lookingeast; (b) west elevation as recorded ...... 80 Figure 35 (a) Transversecross-section of concourse and side-wings lookingsouth; (b) northern elevation of trainshedand side-wingslooking south ...... 81 Figure 36 Ordnance Survey First Edition 1876, showing Oxford Station as built (with side-wing extensions)...... 84 Figure 37 DetailofPaxton’s patented wedge-fixing system usedatthe Crystal Palace ...... 85 Figure 38 Details of secondary components related to support of primaryroof ...... 89 Figure 39 Plan stamped ‘Engineers Office,L&NWR, Euston, No. 4038, 28 Sep. 1903’...... 91 Figure 40 Plan of station dated to ‘early 20th century’...... 91 Figure 41 Rewley Road station,extent of restoration ...... 98 Figure 42 Rewley Road station,roof plan showing age of trusses ...... 100 Figure 43 Agroup of mid 19th-century clay tobacco pipes, nos 42–47...... 101

iv List of Plates

Plate 1Countess of Warwick’s foundation stone ...... 8 Plate 2The extant riverside wall andWatergate Arch ...... 12 Plate 3The Watergate Arch in 1912...... 12 Plate 4The western face of the riverside wall ...... 14 Plate 5The south wall of the church with the possible transept wall to the rear, exposed in Trench 1994/18 ...... 19 Plate 6Grave 1104 exposed in Trench 1986/11 ...... 20 Plate 7The end wall of the north range (1702) showing the stone jamb with trefoil moulding exposed in Trench 1994/17 ...... 25 Plate 8The northern wall of the north range exposed in Trench 1994/16...... 27 Plate 9Masonry wall(1632) parallel to the outer wall of the northern range of the cloister...... 28 Plate 10 Pottery...... 42 Plate 11 Medieval decorated floor tiles...... 56 Plate 12 Architecturalstone ...... 59 Plate 13 Whetstone...... 60 Plate 14 The Rewley Road L&NWRstation from the south, 1914 ...... 64 Plate 15 Trainshed, Bay 5(W) before dismantling, 1998 ...... 75 Plate 16 The earliest known view of the station interior looking south from c 1900 ...... 76 Plate 17 Profileofthe original longitudinal ridge-and-furrowside-wing roof preserved in original internal vertical-boarded partition (west wing roof)...... 76 Plate 18 Detailofdecorative spandrels and circle motif in the west elevation of the porte-coche` re...... 79 Plate 19 Photograph of 1883 showing the Oxford Volunteers marching to the GWR station. The porte-coche` re of the L&NWRstation,withoriginal decorative crestsand finial, is clearly visible to the left hand side of the photograph ...... 79 Plate 20 Roof structure of west wing, showing wall plate and stopped lap joints ...... 86 Plate 21 ‘Ghost’ of formercornice in original booking office (east wing) ...... 86 Plate 22 Paxton-esque glazing bars re-usedinthe structure of the secondary roof light of the eastern side-wing ...... 87 Plate 23 Interiorofthe train shed looking north, 1914...... 92 Plate 24 The free-standingbooking office, seen in aphotograph of 1940 ...... 93 Plate 25 Exterior view of station on 1st May 1940...... 94 Plate 26 View of station on 4th October 1951, shortlybefore closure...... 95 Plate 27 View of station in 1959, when it was in use as agoods depot and lodging...... 96 Plate 28 Rewley Road station at the start of dismantling...... 97 Plate 29 The re-erected station (Phase 1) at Quainton...... 99

v List of Tables

Table 1Summaryofarchaeological work and archives...... 5 Table 2Christ Church lessees of Rewley...... 8 Table 3Jettons ...... 33 Table 4Summaryofthe pipes recoveredfromeach phase of work ...... 43 Table 5Ceramicbuilding materials, assemblage quantification...... 52 Table 6Floor tile thickness by quantity...... 54 Table 7Quantificationofsmall finds by functional category ...... 60 Table 8Summaryofprimaryframework components...... 74 Table 9Summaryofprimarywalling components ...... 83 Table 10 Summaryofconstructional componentsrelated to strengthening of primaryroof...... 88

vi Acknowledgements

The authors and editors would like to thank the Saı¨ d maintained on ground reduction associated with the Business School, UniversityofOxford, forfunding construction of the link road between the BotleyRoad the post-excavation work and publication; we are and Hythe Bridge Street by the AOC Archaeology especially indebted to Mathew Davies, Facilities Group(1999).Afull programme of historical and Managerofthe Saı¨ dBusiness School, for his support documentary research was undertaken by Julian during the project. Theproject was monitored by MunbyofOAU in 1996 as part of the pre-dismantling Brian Durham,Oxford City Council archaeologist, study (OAU1998, AppendixA). The following ac- and his comments and assistance are gratefully count is based for the most part uponthese previous acknowledged.Therehas been along and complex studies and the contributions of their respective sequence of fieldwork, but the principal interven- authors are hereby gratefully acknowledged. The edi- tions have been funded by British Rail (subsequen- tors would like to thankthe architect Lance Adlam, tly Railtrack),Knowles and Son, Ove Arup and who oversaw the dismantling and subsequent re- Partners, Oxford City Council, Persimmon Homes, erectionofthe station building, for his contribution to Stanhope Properties, WSAtkins, and the Saı¨ dBusi- this report, not only in terms of his text, but also of his ness School.DavidHiggins extends his thanksto freely-and enthusiastically-given specialist advice. Dr Susie White for preparing the artwork and Figures41and 42, prepared by Jane Phimester, could catalogue of illustrations for the clay pipes section not have beenproduced withouthis help. as well as for inserting the figure numbers and The following individuals and institutions that carrying out the final checkingofthat text. have suppliedimagesare gratefully acknowledged: The station building and its site were the subject of Bodleian Library, UniversityofOxford (Fig. 7; aseries of archaeological investigationsinadvance Plate 19), Oxfordshire County Council(Figs 5, 8, 10 of, during and after the dismantling process. Reports and 13; Plates1,2,14and 23), National Railway on the foundation structures were prepared by Rob Museum (Plates24and 25), Dr GWait of Gifford Kinchin-Smith,Julian Munby and Jonathan Hiller Archaeology (Plate 20), and Lance Adlam (Plates 28 of OAU (OAU 1995a and b) and afull analytical and 29). The editors would also like acknowledge the survey was undertaken prior to dismantling by Rob Lens of SuttonAssociation (Plate 19), JDEdwards Kinchin-Smith of OAUin1997–98 (OAU1998). A (Plate 27), and the copyright holder or administrator furtherprogramme of surveyand recording work of the Dr Parkes Collection (Plate 26), although it has was undertaken by Stephen Dean of Gifford Archae- not been possible to contact them. ology during the dismantling process itself (Gifford The post-excavation programme was managed by and Partners 2000). Following the removal of the Edward Biddulph.The report was edited by Edward station building fromthe site, awatching brief was Biddulph and Anne Dodd.

vii Summary

This reportpresents the results of over 40 years of studying at the university,and can therefore claim to be excavation, historic building survey anddocumentary one of Oxford’s earliest colleges. The railwaystation that researchthat have been carried out by Oxford Archae- subsequently occupied the site in 1851 followed the ology andothersatthe site of the Cistercian house design of the Crystal Palace, built for the Great Exhibi- of Rewley, achantry founded in 1281. It became an tion, andwas the last surviving representative of that abbey and studium providing accommodation for monks internationally important building.

viii INTRODUCTION northern end of the Business School overlies much of David RPWilkinson the preservedabbey remains, and the garden has been laid out to reflect the medieval plan. Part of the The Cistercian house of Rewley, on the western extreme south of the site now lies underthe widened outskirts of medieval Oxford, was founded in 1280 and realigned BotleyRoad as it runs towards Hythe as achantry, and becamebothanabbey and a Bridge Street. Rewley Road, after passing the west studium.The latter statusenabled monks to live side of the firestation,now loops acrossthe east side there whilestudying at the university, giving Rewley of the abbey complex, and runs north into the areasonable claimtobeone of Oxford’s early housing development. Ground levels on the site, colleges. Apparently never very successful as aplace prior to the redevelopment, were generally between of study, Rewley Abbey survived until the Dissol- 58 and 58.5 mOD, falling to 57.5 mODinthe south- ution, after which some of its buildings remained west corner. Theunderlying geology consists of standing and were put to secular use.Wealso know the gravels of the Thames floodplain, overlain by that some parts of Oxford’s western Civil War alluvium, the upper surface of which canbefrom defencescrossed the site.All of this was swept 1.5 to 2.5 mbelow ground level. Where investigated, away or buriedwith the coming of the railway to the alluviumwas up to 0.25 mthick. Oxford in the mid 19th century.The railway station that occupied the site from 1851 was, by chance, a remarkable building as atwin sister to the Crystal Project background Palace,and the last surviving representativeofthat This section concentrates on the below- and above- renowned building of international importance. groundinterventions that have led to the current Over the last 40 years, aseries of developments on report. This report has beenwritteninthe 300th the old station yard site have prompted excavation, anniversary year of known research intoRewley historic building surveyand documentary research. Abbey. In 1706, Thomas Hearne exposed and de- This report brings together the results from all of that scribedparts of the abbey complex, on asite which work, the bulk of which was carried out by Oxford Agas (in 1578, Fig. 2) andLoggan (in 1675, Fig. 3) Archaeology (OA, formerlyOxford Archaeological depicted as mainly opengroundwith only afew Unit), although resultsare included herefrom investi- buildings, some field or garden walls, and anumber gationsbyother bodies and privateindividuals. of ponds.Much of the layoutofthe watercourses was In addition to the abbey complex, which is a recognisable, but the site as known by Hearne was Scheduled Monument (Oxon. No. 80), the site also some 1to1.5 mlower than now, due to deliberate included the formerLMS station building, con- raising of the ground for railway construction during structed in 1851 and aListedBuilding (Grade II*), the second half of the 19th century. Rewley Abbey and afurther Scheduled Monument,the railway streamwas filled in as partofthe sameoperations. swingbridge over Rewley AbbeyStream. The station The complex story of modern archaeological inter- building has now been removed from the site, and its ventionsatRewley is set out in detailinTable 1. It historyforms partofthis report, whilethe swing- began in 1967–69 when trenches were dug by Oxford bridge remains in situ and has not yet been the subject Archaeological Excavation Committee on the site of, of adetailedsurvey. and justnorth of, the proposed firestation. These Many different bodies have funded investigations trencheswere not deep enough to reach any in situ and reports, but it seems fittingthat the last or- archaeology from the abbey period. Asingle trench in ganisation to do so, and to fund thispublication, is 1981 by Oxford ArchaeologicalUnit (OAU,now OA) the Saı¨ dBusiness School (Oxford University),which cut across amedieval wallonthe north edge of Rewley was founded in 1996 and completed its new building Abbeystream. Since then there have been three main on the Rewley Abbeysite in 2001. evaluations using trial trenches, all by Oxford Archaeology: on the abbey complex in 1986, for British Rail and Oxford City Council; at Rewley Road, Geology and topography on the south-eastern corner of the site, in 1993, for The site is at NGR SP 5056 0645 (centred) and OxfordshireCounty Council; and over almostthe is roughly the shape of anarrow triangle pointing entiresite in 1994, including the area north of Rewley north-west, with the tip being cut off by Rewley Abbeystream, for Stanhope Properties PLC (Fig. 4). AbbeyStream (Fig. 1). The investigated areawas The 1994 evaluation extended beyond the abbey boundedtothe east and north-east by the Castle Mill complex and explored aspects of the surrounding stream; to the south by ParkEnd Street, on the east landscape, including the Civil War defences.These by Rewley Road and the fire station,and to the west three phases of work produced most of the below- by the current Oxford Station and associated railway grounddata described and discussed in this report. land. The derelict rail yards that occupied the site 40 Other below-ground fieldwork has consisted yearsago have nowbeen infilled or covered by mainly of archaeologicalmonitoringonthe various housing in the northern part of the site,student phases of construction work since 1993 (not de- accommodation to the east, the Oxford Station scribedhere in detail –see Table 1). In all, there have forecourt to the west, and the Saı¨ dBusiness School been nine such interventions.Thisworkwas part of to the central south. TheFellows’ Gardenatthe amitigation processbypreservation in situ,which

1 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 1Location of the Study Area.

2 OA Occasional Paper No: 16

Figure 2The site of Rewley Abbey as it appears on Ralph Agas’map of Oxford, 1578 (south is to the top).

was applied to the student accommodation built at work was undertaken by machine stripping within Rewley Road, the housing developmentofthe the trench areas only, followed by hand excavation. In northern partofthe site, the construction of the Saı¨ d all these cases, the main purpose of the investigations Business School and the new station forecourt. was to gain abetter understanding of the layoutof The former LMS Station, listed Grade II* was the abbey, andtocharacterise the surviving archae- surveyed by OA in 1995, and again in more detailin ologicaldeposits. Workingtoguidelines set by 1997–98 when it was becoming clear that the building EnglishHeritage as conditions of Scheduled Monu- was to be moved to make room for the Saı¨ dBusiness ment Consent (from 1993 onwards these guidelines School.Furtherrecording work was undertaken by were formalised in written briefs set by Oxford City Gifford Archaeology during the dismantling process Council Planning Department in conjunction with and the building’s reconstruction at the Bucking- EnglishHeritage),excavationnormally either stopped hamshire Railway Centre at Quainton. at the top of significant archaeological deposits, or a small sample of the deposits was excavated, allowing artefactual and palaeoenvironmental evidence to be Methodology assessed. Exceptions to this approach were graves, The trialtrenches for the 1967–9 and 1981 evaluations some of which were partially excavated in 1986, and were excavatedentirely by hand. In 1986, 1.8 mwide robber trenches,which were normally excavated to areas of the site were stripped down to the 1850 clarify wallalignments. The layout of the trenches was groundlevel by machine, and narrower strip trenches in all cases chosen following study of the available were then excavated by hand, while the 1993 and 1994 historic maps and views of the site.Recording metho- dology was as set out in the Oxford Archaeology Field Manual (Wilkinson 1991). Although the 1993 and 1994 trencheswere slightly wider than those dug in 1986, the emphasis on narrow trenchescontinued. With hindsight, amore varied approach including some wider excavation areas mighthave givenbetter results, although such an approach might not have been granted Scheduled Monument Consent. Becauseofits statusasaScheduled Monument and the archaeologicalsignificance of the abbey site, confirmed by the evaluation work just described, the policy of preserving the archaeologicalremains in situ was continued during the ensuing developments. Archaeologists from Oxford Archaeology worked with EnglishHeritage andOxford City Council, and Figure 3The site of Rewley Abbey as it appearson with developers, architects and engineers, to ensure David Loggan’s map of Oxford, 1675 (south is to the top). that construction work impactedeither outside the

3 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 4Plan of the StudyArea with trench locations.

central zone of preserved archaeology, or into the EnglishHeritage and/or Oxford City Council. The overlying post-1850 deposits. Once construction work general lack of archaeology from this monitoring began it was monitored by Oxford Archaeology (and shows that the preservation in situ was successful, in one case by AOC Archaeology Group), again although in the case of the student accommodation, according to written specifications approved by construction of piles over (and in some cases through

4 OA Occasional Paper No: 16 the walls of) the east end of the abbey church from Trench 1, resulting in the replication of trench undoubtedly caused some damagetothe monument. numbers between investigations. For clarity in the narrative that follows each trench number is prefixed The report by the year of excavation, thusTrench 1ofthe 1986 evaluation becomesTrench 1986/1and so on. The overallaim of this report is to bring together Anumber of environmental samples taken dur- and discuss all classesofevidence for the history of ing the various stages of fieldwork were initially the RewleyAbbeysite, from before the abbey to the believedtobesufficiently well-preservedand sui- presentday. tably coherent in terms of site stratigraphy to allow furtheranalysis work during the post-excavation Phasing and dating project (OA 2005). However, doubt over the integrity of one seemingly ideal sample and the inability to Matriceswere preparedfor each trench, and links locate others meant that further work would not between trencheswere definedwhere possible. Dates have been justified,and therefore no analysis work derivedfrom the pottery and other finds were applied was carried out. to the matrices, and theywere divided intopre-abbey, abbey and post-abbey phases. Given the limited nature of the trenched evaluations (see above), and The archive the extensive post-dissolution robbing of the site, it Archiveshave been assembled for each of the inter- was rarely possible to further refinethe phasing. ventionsthat took place at the site, and these contain all the detailed supporting information from which this report has been written, including the original Structureofthe report site records, records made during post-excavation The narrative derivedfrom the above work has been analysis, and previous reports. Deposition of these expanded using the documentary and historic build- archiveswill be at either the Oxfordshire County ing evidence, andthis forms the core of the report. Museum, or the AshmoleanMuseum, depending on Specialist reports are referred to in these chapters as the land ownership, and the archivesare or will be appropriate, and are presentedseparately following availablefor referenceand research at these institu- the main narrativesections. In each of the phases of tions. Table 1listswhich archiveswill be deposited fieldwork the trenches were numbered sequentially with which museum.

Table 1Summary of archaeological workand archives.Fieldwork type: Excavation (Ex), evaluation (Eval), watching brief (WB)and building survey (BS). Unit: Oxford Archaeological Excavation Committee (OAEC), Oxford Archaeology (OA), AOC Archaeology Group(AOC).

Year Site code Description Fieldwork Unit Client Accession Receiving type code museum

1967–9 –Station Yard Ex OAEC –1986.161 OXCMS 1981 OXRA81 Rewley Abbey Ex OA –1986.161 OXCMS 1986 OXRA86 Rewley Abbey Ex OA British Rail/Oxford City 1986.248 OXCMS Council 1993 OXRR93 Rewley Road Eval OA Oxford City Council 1993.103 OXCMS 1994 OXRR2 94 Rewley Road WB OA Knowles and Son 1993.103 OXCMS 1994 OXSTRA94 Oxford Station Rewley Eval OA Stanhope Properties1984.11 OXCMS Abbey 1994 OXRR3 95 Rewley Road (Pumping WB OA Knowles and Son 1993.103 OXCMS station) 1995 OXLMSBS Former LMS station BS OA Stanhope Properties– – 1995 OXLMSBS Former LMS station BS OA Railtrack and Stanhope –– Properties 1997/8 OXSTRA97 Oxford Station Rewley WB OA Persimmon Homes 1984.11 OXCMS Abbey 1997 OXSBS97 Saı¨ dBusiness School WB OA –1997.92 OXCMS 1998 OXLMSBS Former LMS station BS OA Railtrack –– 1998 OXFIST98 Rewley Road Fire Station WB OA WSAtkins 1997.109 OXCMS 1998/9 OXSAID98 Saı¨ dBusiness School WB OA Ove Arup &Partners 1998.112 Ashmolean 1999 OXPES99 Old LMS station WB AOC Oxford County Council –– 2000/1 OXSAID00 Saı¨ dBusiness School WB OA Saı¨ dBusiness School 1998.112 Ashmolean 2001 OXRAW01 Rewley Abbey Wall WB OA Oxford City Council 2001.142 OXCMS

5 OA Occasional Paper No: 16

Evidence for agricultural activity pre-dating the the trenches,with the notableexception of Trenches establishment of the abbey was identified in four 1994/16 and 1994/17, which may indicate the trenches. To the west of the moat that encircled the locationofanatural rise in the original topography abbey aseries of layers of loamy soil(1946, 1952, of thispartofthe site.Variations in the composition 1959) and possiblywater-lainsilt (1980, 1995, 1947– of this made ground across the site may indicate 9) were recorded in the base of Trench 1994/19. successive phases of levelling. Theinitial phase of These layers were sealed beneath deposits associated levelling may have concentrated on the central part with abarn contemporary with the abbey and of the site and is represented by dumpsofloamy soil contained considerable quantities of pottery dating recorded in Trenches 1994/18 and 199425 in the area from between the mid-11th and mid-13th centuries. of the church, Trenches 1986/1and 1986/2ashort Similar deposits were also recorded in Trenches distancetothe west and Trench 1986/5 in the areaof 1994/27 and 1994/28, immediately to the north and the eastern range of the cloister.The soil used for this south of Trench1994/19, although here the very levelling contained muchpottery and may have limited samples excavated did not yield any datable originatedasmidden material imported from the material, and similar silt layers in Trench 1994/21 nearbycity, which would explain the presence of furthertothe south produced asinglesherd of late muchofthe earlier pottery found on the site. These 11th to mid 13th-century pottery. Dumpsofloamy deposits also containedthe only coin from the soil were also identified in the vicinity of the abbey excavations, an Irish silver farthing of Edward I(SF church in Trench 1994/18 and contained asimilar 8) dating from 1280/1–1295, and an English jetton assemblage of pottery. The loamy soils are likely to (SF 5) datingfrom1302–7, bothrecoveredfrom be formertopsoil layers with an admixture of pot- Trench 1994/18. tery resulting from manuring with middenmaterial, The made ground recorded in the eastern part of whilethe silts were likely to have derivedfrom the site in Trench 1986/4 and allfourtrenches of the episodes of floodingfrom the nearby channels of 1993 investigation (102, 107, 224, 326, 422) contrasted the River Thames. with that in the central partofthe site in consisting of dumpsoforange clay and gravel. These layers contained pottery of slightlylater dateinaddition The Abbey to early material similar to that seen in the central Preparing the ground:reclamation and levelling area, and is likely to represent aslightly later phase of deposits levelling on account of its relationship with the precinct wall (below).Similardumps (1637–1639) Rewley Abbey occupied the northern part of Oseney were also recorded immediately beyond the northern island,anareaoflow-lying groundbetween two part of the moat in Trench 1994/16. channels of the River Thames ashort distance north west of Oxford. The area is likely to have been Enclosure and access: theprecinct wall, moat and susceptible to flooding, as demonstrated by alluvial entrance road deposits which formedthe natural substrateinmany of the trenchesand the silts in Trenches1994/19, Mapsand views of the abbey drawn during the post- 1994/27 and 1994/28 referred to above. The first medieval period show that the complexwas stage of construction therefore entailedaconsider- enclosed by aprecinct wall (Figs 2and 3), the able engineering operation to drain and consolidate construction of which was licensed by Edward Iin the site before building could begin. 1299. The wall on the eastern side, facing onto the This comprised the digging of aseries of drainage river, is still extant, including an archeddoorway of channels which effectivelydivided the areainto a 14th-century design (Plates 2and 3), and the series of smaller islands, as is shown on the historic relationship of this walltoadjacent archaeological maps(Figs 2and 3). Oneofthese islands, which deposits was examined in Trench 1993/1. fronted onto the eastern channel of the Thames, was The wall was set withinafoundation trench (110) chosen for the site of the abbey complex itself, while dug intothe underlying alluvial silt (131) immediately the remainingislands are likelytohave been used as adjacent to the edge of the river channel (Fig. 6). The fields and pasture, possibly benefiting from irriga- fill of the foundation trench (130) contained asingle tion by seasonal flooding. The groundlevel within piece of glazed floor tile of late 13th- or early 14th- the site of the abbey complex was raised above flood centurytype which would be consistent with con- level by the dumpingoflayers of made ground to struction around the time the abbey was founded, and create aplatform on which the abbey buildings the presence of such material may be indicative of could be built. All this groundwork did not pass otherconstruction work being carried on at the same withoutcomment and it is recorded that Oseney time. That the wall and madeground were con- Abbey, located ashort distancedownstream from temporaneousisdemonstrated by the construction of Rewley, agreed amoney settlement in casethese the wall, with facedstones beginning at the original changestothe floodplain should cause flooding groundlevel on the river side but considerably higher affecting its land. on the inner side, which is butted by 0.8 mofdumped The layers of dumpedsoilusedtoraise the ground clay and gravel(102, 107) to makethe ground up to level were the earliest deposits recorded in most of this level (Plate 4).

11 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 2The extant riverside walland Watergate Arch q Oxfordshire County Council.

The natural silt of the river bank was overlainbya dump of gravel(129) against the foot of the wall, upon which rested athick layerofrubbleand mortar (127) that spilled downthe bank and into the river channel. This layer contained large faced limestone blocks and is likely to be the result of apartial collapse of the wall in this area, the repairing of which was identifiable as an areaofcruder construct- ion in the upper part of the wall. No evidencewas recoveredtoprovideadate forthis event, but it may have occurred fairly late in the history of the site as it was overlain by aseries of flood silts (120–125), the earliest of whichwas dated by pottery to the late 18th-early 19th century. The drainagechannel around the north and west sides of the main abbey complexwas formalisedinto amoat, the sides of which were revetted by stone walls which were exposed in Trenches 1994/17, 1994/19 and 1994/21. In all three trenches only the top of the surviving part of the wall was exposed as the depth of the trenchesand the instability of the overburden precluded excavation to agreater depth. The revetment wall on the inner sidewas presumably continued upward to form the precinct wallonthese sides, as Loggan’s map of 1675 clearlyshows the wall facing directly onto the moat (Fig. 3).The historic depictions of the abbey also show that along the north side of the complexthe place of the precinct wall was taken by the north range of the cloister. Plate 3The Watergate Arch in 1912.

12 OA Occasional Paper No: 16 deposits. related and wall riverside the through section 1993/1: rench 6T Figure

13 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

1994/20. Most of the fills encountered withinthe moat consisted of the clay backfill used to level the site for the construction of the Rewley Road railway station during the 19th century. Possibly earlier fills were recorded in Trench 1994/21, in the form of a homogeneouslayer of silt (2124) more than 0.6 m deep, and in Trench 1994/16 where aseries of deposits dumpedagainst the inner revetment wall (1657, 1658, 1662–4) may have been part of the 18th- centurybackfilling but could equally have started earlier.The latter scenario would be consistent with drawings of the northern frontageofthe abbey, which show trees and plants growing on aslightbank against the wall (Fig. 7). There was some evidence that the moat had been re-defined or cleanedduring the 18th or 19th century, particularly in Trench 1994/21 where an apparent re-cut (2106) truncated alayer (2121) containing pottery of this date. There is some evidence that the main abbey com- plex may have been further defined by an east-west ditch dividing it from the southern part of the island on which it stood, and perhaps intended to separate the monastic buildings from the parts of the abbey estatewithmore secular uses. This ditch (216) was recorded at the southern ends of Trenches1986/1and 1986/2investigation, where it curved around the south-westerncornerofthe abbey complexand exten- ded eastward beyond the trenches. It was cut into a layer containing early 15th-century pottery and had been re-cut on at least one occasion. Although it approached close to the westernpartofthe moat, the fills in the base of the ditch showed no indication of being waterlain and so the ditch is likely to have been Plate 4The westernface of the riverside wall. Note the dry and nottohave formedacontinuation of the moat unfaced stonework of the lower part of the wall, which was system. The evidence from pottery within its fills abutted by dumps of made-ground. and fromthe overlying soil layer suggests that it was filled in during the 16th century, which would be consistent with its omission from Agas’map (Fig. 2). A stone footing that was approximately square in shape The revetmentwalls were generally 0.5–0.7 m (217) was set deep into the fills of the ditch, and may wide, and neatly constructed from limestoneblocks. have been apier for abridge carrying the entrance Trenches 1994/17 and 1994/19 showedthat both the road across it, although no correspondingabutments inner and outer walls were not constructed flush were observed at the edges. with the cut of the moat, but stood proud of it by as The abbey grounds were entered from Hythe muchas0.55 m, with the intervening gap filled in Bridge Street to the south, where amedieval arched with deposits that containedpottery of late 13th- to gateway was still standing in 1720 (Fig. 8). An en- 14th-century date. The existence of asimilar gap tranceroad extended northward from the gate to the in Trench 1994/21 was less clear, as stone robbing, main abbey complexalong the line now followed which had affected the wall in all of the trenches,had by Rewley Road. The entranceroad was sectioned disturbed this sideofthe wall. where it approached the main abbey complex in In Trench 1994/19 the wall revetting the outerside Trench 1986/3. The road comprised asurface of of the moat (1957) was found to be overlainby limestone cobbles(302) beside which was ashallow successive layers of silt (1928–1931) that contained north-south ditch (303), presumably oneofapair of pottery datingfrom the late 13th or 14th century. drainage ditches flankingthe road. The ditch had These layers may have been laid downeither by over- been backfilled with gravelthat may previously bank floodingofthe river and moat, or representup- have been usedasmetallingfor the road. Pottery cast from periodic cleaning out of the moat. from the road surface dated from between the 13th The moat itself measured between 6mand 8m and 15th centuries. wide. Duetothe difficulty in excavating safely to any Two successive gravellayers (213, 209) pre-dating greatdepth in narrow evaluation trenchesitwas not the ditch in Trenches1986/1 and 1986/2and two possible to establish the full depth of the moat, despite later similar layers overlyingthe earlier cut of the excavating to adepth of more than 3min Trench ditch may indicate the presence of apath branching

14 OA Occasional Paper No: 16

Figure 7The ruins of Rewley Abbey seen from the north-west, drawn by WTurner in 1851. Minn Coll. Neg. 10/44 q Bodleian Library, University of Oxford. off the entrance road to extendaroundthe outside orientated ENE-WSW, with asmall cloister on the of the south-west corner of the cloister,possibly north side. This is the reverse of the design nor- extending acrossabridgeover the moat shown in this mally adopted by the Cistercian order, which nor- locationonhistoric maps. mally located the cloister to the south of the church, and may have beendictated by adesire to place the The church and cloister (Fig. 9) church away fromthe edge of the moat to avoid the risk of subsidence. Accessinto the complex from The results of the excavations at Rewley have dem- the entranceroadmay have been through an arch at onstratedthat the abbey complexcomprised achurch the westernend of the church, which was shown in

Figure 8Aview of the entrance to Rewley Abbey,drawnbyMBurghers in 1720. HenryWTaunt Coll. (HT12905) q Oxfordshire County Council.

15 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford precinct. abbey the in features archaeological main the of lan 9P Figure

16 OA Occasional Paper No: 16

Figure 10 Aview of the ruins of Rewley Abbey from the south, drawn by MBurghers in 1720. HenryWTaunt Coll. (HT12905) q Oxfordshire CountyCouncil.

Burghers’ engraving of 1720 (Fig. 10), although it is was constructed of large, roughly shaped limestone also possiblethat this featurehad beenfrom another blocks bonded with ayellow sandymortar. On its locationafterthe Dissolution and that access intothe southern, external side was the foundation of abadly cloister was originally gained at some otherpoint. robbed piece of masonry (1852) 1.4 mwide and of Although the west, east and north rangeshave been identical build. This feature extended beyond the located by excavation, none was investigated exten- southern edge of the trench and is likely to be either sively enough to establish the functions of each range awall of the south transept or abuttress. It lay on the or identify specific rooms, but it seems most likely that same alignment as arobber trench (801) recorded in the standard layout of aCistercian monastery was Trench 1986/8and has beenviewedasbeing part of simply reversed due to the specific circumstance of the the east range of the cloister, suggesting that the abbey’slocation. Limitedinvestigation of the areas formerinterpretation is more likely. If masonry 1852 outside the cloister has also enabled some details of is indeed the west wall of the transept, then arobber the use of these areas to be filled in. trench recorded in Trench 1986/10 is likelyto indicate the position of the corresponding east wall. The line of the south wallofthe church identified in The church (Fig. 11) Trench 1994/18 was continued in this trench by a Theprecise locationofthe abbeychurchhad been robber trench (1006) from which areturn(1007) much debateddue to an absenceofhistoricalrecords, branched off to the south c 5.2 meast of wall 1852. as it wasapparentlydemolishedshortly afterthe Dis- Thereturnextendedfor 1.2mbeforecontinuing solution andnotrace remained abovegroundbythe underabaulkwithinthe trench.The northern wall time theearliesthistoricmapsweredrawn.The issue of thechurchwas identifiedasarobber trench 1.5m waspartlyresolvedbythe 1986 investigation, when wide extendingthrough thenorthernendsofTrenches partsofits remainswereexposed in Trenches 1986/ 1986/9 and1986/11a(906, 1156), butthe stonework 9–1986/11a.Its identificationwas confirmed in 1994 by at itsbasewas notinvestigatedinany detail. Trenches 1994/18and 1994/25, butthere is stillmuch Within the interior of the church four stone piers uncertaintyoverits exactdimensionsand layout,par- were identified, which would have formedpart of ticularlyasithas notbeenpossible to satisfactorily apair of arcades dividing the nave from two narrow identify either theeastand west ends or thetransepts. aisles only c 1.7 mwide. The piers were constructed The stonework of the church had beensubstan- from blocks of limestone and,likethe walls, had tially robbed, with the outer walls reduceddownto been affected by stone robbing, though to aslightly the level of the water table, but it was possible to lesser and more variable extent. Although none of trace the line of the north and south walls from the the pierswas completely exposed within the limited alignment of the robbing cuts. This indicatedthat the area of the trenches,they appear to have been main body of the church had an internal dimension rectangular in shape with their longeraxes parallel of 12.5 mand was at least 40 mlong. The most with that of the church. ThepiersinTrenches 1986/9 substantial piece of surviving masonry from either and 1986/10 bothmeasured c 2m· 1.6m,whilethe wall was partofthe foundation of the southern wall largestexample,the pier in Trench 1986/11, meas- (1851) exposed in Trench 1994/18 (Plate 5). This was ured at least 2.15 mwide, its longer dimension 1.5 mwideand survived to aheight of 0.5–0.6 mand extending beyond the edge of the trench.

17 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

N Church. Abbey the investigating trenches of Plan 11 Figure

18 OA Occasional Paper No: 16

exposed within Trench 1986/11 were excavated (Plate 6), but those elements lying beyond the limits of the trench were left in situ.Sufficient was exposed of the skeletons in graves1157 and 2523 to confirm the identification of the features as graves, but the remainswere left in situ.Unusually, the body in grave 2523 lay with its headatthe east end, the reverse of the orientation of the other burials. The body from grave 1104 was identified as female, but the others lacked diagnostic features to establish their sex. The backfill of grave 1103 contained two sherds of pottery dating from the mid12th–14th centurywhileahandful of residual pottery recovered from grave 1104 dated from the 13th–14th centuries. Evidence for at least three otherburialswas also recovered. In addition to its intended occupant, grave 1103 also contained adisarticulated leg, pre- sumablydisturbed from an earlier burial.Robber trench 900, excavated in Trench 1986/9, contained the disturbed remainsofarelatively complete burial, which likethat in grave1104 was female, while an apparent robber cut (1100) within the nave in Trench 1986/11 contained the remains of amorepartial skeleton. Asmall assemblage of unstratified human bone was also recovered from Trench 1986/11. Outside the south wall of the church the made ground(1858) was overlainbyaskim of mortar (1853) that may have beenassociatedwiththe construction of the church, abovewhich was amake up layer (1850) forahard gravelsurface (1827, 1848). This surface also sealed aseries of dumped deposits that seem to have formed part of amidden (1828, 1829, 1831, 1834–8) c 6mfrom the wall of the church, Plate 5The south wallofthe church with the possible containing large quantitiesofpottery, charcoal and transept wall to the rear, exposed in Trench 1994/18. bone. Alayer of soil (1810/1847) that had accumu- lated over the gravelsurface containedpottery datingfromthe late 13th–15th century. The limited amountoftrenching carried out revealed Alarge roughly circular pit (2538) within the nave few details of the interior of the church. Thefloor in Trench 1994/25 was found to have many large comprised aseries of make-up layers (1863–5) pieces of limestone in its base. The backfill of the pit overlaininmostofthe trenches by ahard gravel (2537) containedpottery of 17th- to 18th-century surface (1823),none of which containedany dating date, and it is likely that the feature represents evidence. In Trench 1986/9 asequence of three furtherrobbing of stonework, although it is unclear mortar floors was encountered (928–930), but these whatelement of the church couldhave beenlocated seem to be localised to thistrench. The gravel would here. Alarge pit(918) was partly excavatedwithin have formedanunusually crude surface for the the nave at the south end of Trench 1986/9. The insideofthe church, and so is likely to have been a pottery from this feature was 13th- to 15th-century in beddinglayer forabetter quality surface that does date and included ajug base with an inner surface not survive. Enough glazed floor tileswere recovered heavilystained with adark red-brown substance from residual contexts to indicate that at least partof which may be red lead, acommon medieval paint. the church was tiled, although this is unlikely to The jug may have been usedfor storing paint, or the applytothis partofthe floorasthe gravel would not broken base re-used as apalette.This pit was cut by be an appropriate bedding for such asurface. asecond, shallower pit (902) that was dated by Four graveswere identified within the church, all pottery within its fill to the 13th–15th century. This of which sharedthe same orientation as the building. featuremay have extended into Trench 1986/10, One grave (1103) was located in the middleofthe where pit 1002 was in an analogous position and nave, one (1157) in the north aisle, and two (1104, was cut by the robber trench (1016) for one of the 2523) in the south aisle. The inhumationswere piers of the southern aisle. placedinanextended supine position in shallow Afurther indicationofthe appearanceofthe sub-rectangular cuts withoutcoffins,although it church was provided by the discovery of pieces of is possiblethat they were wrapped in shrouds. medieval window glass painted with designsof The parts of the skeletons in graves 1103 and 1104 foliageand cross-hatched in-filling, in astyle typical

19 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 6Grave 1104 exposedinTrench 1986/11. of the 13th century. One such fragment was re- on its north side by the west range of the cloister. covered from apit within the nave of the church Againstthe westernedge of Trench 1986/2arebate (902) andone from asoillayer immediately outside in the north wall indicatedthe original position of a the south wall (1810),while athird piece came from doorway openinginto the west range,which had adisturbed context in Trench 1986/1. subsequently been blocked (236). Within the build- ing aseries of layers of gravelly soildeposited to raise the groundlevel in advance of construction The chapel/brewhouse (29/2–4) were sealed beneath agravelfloor surface Arectangularbuilding at the south-western corner of (7, 29/1). Arobber trench (259) inside the building, the complex was uncovered at the west end of the branching off at aright-angle from the north wall church in Trenches1986/1 and 1986/2(Fig. 12) and may indicate the position of aformer internal wall, furtherinvestigated in Trenches1994/18 and 1994/ and contained aFrench dolphin jetton (SF 88) dating 25. This building appears to have had apredecessor, to c 1375–1415, in additiontoarange of medieval as part of awall(31) was exposed running parallel to pottery sherds. the west wall of the building. Apossible return of this Burghers’ engraving of the abbey in 1720 shows wall (28) was cut by the foundations of the building. this building as having apair of arched windows in The building lay on the sameENE-WSW orienta- its western gable end, indicating that it was medieval tion as the church and measured c 10 mwide and in date and had been constructed as partofthe 12.5 mlong. The west wall (3) was verysubstantial abbey (Fig. 13). Alikely interpretation is that it was with an offset plinth on its outer face, abattered originally achapel, although on presentevidence it inner face and afooting at least 1.0 mdeep. The is not possible to say whether it was afree-standing south wall (201/202) was of similar proportions,but building or was integral to the west end of the the north wall(218) was narrower and also differed church. Astack of barrels shown outside its entrance in being bonded with amore friable gravelly mortar. suggests that by this time it had been converted into The eastern sideofthe building was represented by a abrewhouse (Fig. 10), and this identification was returnfrom the north wall at the north-eastern end confirmed in the excavationbythe presence of a of the trench. The variation in the dimensions of the large featureinterpreted as asteeping vat (6) cut walls is likely to be due to the south and west walls into the floor of the building. This vatwas sub- being external walls whilethe north wall was butted rectangular in shape and measured 5m· 3m.Itwas

20 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

HISTORICAL BACKGROUND with its own abbot, where amasswould be said daily Julian Munby for his father’s soul.3 West Oxford, Oseney and Rewley OxfordSchools and Colleges Westward from the city beyond Hith Bridge stands the ruins of an abbey called ‘Rewley’, in Oxford was by no meansacollegiate university at the Latineevidences stiled ‘locus regalis in North end of the 13th century, and most students lived in Ousney’. It was sometimes seated with plea- academic halls under the care of amaster. Withthe sant groves and invironed with clear streams, arrival of the friars (Blackfriars in 1221, Franciscans in remotefrom the city, and for its solitariness 1224, Carmelites in 1256 and Austins in 1266) schools agreeable to the disposition of the inhabitants, were associated with their houses and linkedtothe who (as it hath allwayes bin observed) made university.4 Three secular halls had been endowed to choice of such placestoexercisetheir devotion become ‘colleges’: University (1249), Balliol (1263), in. [Wood City of Oxford ii,290] and Merton (1264),but theyonly provided foratotal of 48 scholars. The Cistercians had founded aplace of It may be asked why aCistercian Abbeywas built in study in Paris in 1244, and the foundation of Rewley the fields of West Oxford. The westernsuburbs, in 1281 gave themanEnglishschool to which other mostly consisting of hay meadows and pasturesin Englishabbeys couldsend their monks for study. the manor of Oseney, were probably an ancientpoint Other colleges for Benedictine monks at Oxford were of entry to Oxford fromthe ancient(?Roman) route soon to follow: Gloucester College in 1282 and cominginfrom Ferry Hinksey and acrossthe fields, Durham College in 1286. but there was not amajor road out of town before the 16th-century building of acauseway to Botley and thence acrossWytham Hill to Eynsham and The Foundation beyond.Neither was there adirect route: St Thomas High Streetprovided alower route that turned up Rewley Abbeywas dedicated in December 1281 by Hollybush Row to reach the Botleyroad and Hythe Robert Burnell, Bishop of Bathand Wells (himself an Bridge Street, formed by OseneyAbbeyaround intended college founder at Burnell’sInn), with the 1200, also had to turn adog-leg to reach the Botley first six monks from Thame Abbey being responsible road. 1 The manor or island of Oseney, like the fields for the chantry masses.Itwas from the beginning an on the north of Oxford, was divided between the two abbey with aschool. Its seal showedamaster with Norman lordsofOxford,Robert d’Oilly and Roger fourteenpupils, and it is evident from later records d’Ivri.2 South Oseneyincluded muchofthe new that the building(s) of the studium were somehow suburb of St Thomas’s and the site of Oseney Abbey. distinct from the abbey, just as the monk scholars North Oseney comprised the north side of St were perhaps distinct from the othermonks, Thomas’s street and the fields as far as Rewley, supporting themselves at the expense of their own 5 where there may have been amanorial centre. houses. D’Ivri’s holding in North Oseney descended to the Honour of St Valery which was granted by the crown to RichardEarl of Cornwall in 1227. Abbeylands The foundation was not without controversy: Ose- ney Abbey received apayment in 1281 to withdraw The Earls of Cornwall and theCistercians their objection to the loss of tithes, and the flooding Richard(1209–72), founder of the Cistercian Hailes that they imaginedwould be caused by raising the Abbeyin1246, was the youngerbrother of King land forthe abbey.6 Not long after this, in ahearing Henry III and in the course of avaried public life was before the itinerantjustices in 1285 the Abbot of crowned King of the Romans (i.e. Germany) but did Rewley was fined forraising an outer chamber not go on to become Emperor. It was his son Edmund ( cameram forinsecam)over the water of the Thames, who brought the famous relic of the Holy Blood to 12 foot long by 6foot wide, ‘by which boatsbringing Hailesin1270, and who thenfounded Rewley Abbey victualstoOxford are impeded, to the greatdamage in memory of his parents (its name regalis loco ‘the of the borough’. 7 The abbot was fined and the Sheriff royal place’ referringtoRichard’s rank). They had was to see to the demolitionofwhatever was been buried at Hailes Abbey, but Richard’s heart damaging. At the same Eyre the Abbey of Eynsham was interred before the high altar of the Franciscan claimedthe manor of Yarnton from Rewley, and the church in Oxford, and he had intended to found a AbbotofRewley called upon Edmund Earlof college or chantry. This was taken up by his son who Cornwall to give warranty, which was given.8 The petitioned the general chapter of the Cistercian order charter of Edmund EarlofCornwall in 1294 in 1280 to build at his own expense ahouse of study confirmed to the abbot and fifteen monks of St Mary ( studium )atOxford for English Cistercians. Their of Rewley the manors of North Oseney and Yarnton, response was to licence a studium,built by the earl his wood in Nettlebed in two parks, and land and to be subject to the Abbot of Thame,but this was in Mixbury, together with achurch in Cornwall.9 altered,atthe earl’s request, to be aproper abbey In 1354 Bishop Grandison granted Rewley the

6 OA Occasional Paper No: 16 appropriation of St Stithian’s in Cornwall(of which anew building for students by raising alevy from Edward the Black Prince had previously given them Cistercian houses in England and Wales. Repeated the advowson).10 efforts were madein1411 and 1425, but it was not Rewley had obtained the north end of Warham until 1438 with the foundation of St Bernard’s Bank nexttotheir cemetery in 1286, after an College (now St John’s) that apermanent home was inquisition decided that it would not cause damage found; its foundation charter specifically stated that so long as they did not breach it to the damage of the the Cistercians had no one place of abode in Oxford castle mill.11 Concerns for the control of the river and but dwelt in various places [ in diversis hospiciis sive mill streams was aconstant concern fortheir locis].21 neighbours, and in acomplex agreement with Oseney Abbeyin1297 about tithe payments the Abbotof Rewley agreed nottodoanything in the abbey close Last days of the abbey that would impede water to Oseney mill.12 Disputes Rewley continued as an ordinaryabbey, occurring on tithes in the valuable hay meadowscontinued to occasionally in records of the order (rather unusu- occur at later dates,13 while there were also quarrels ally in 1470 whenamonk wrote inquiring about the with the city in the 1330s on the abbey’srights to hold production of gold by alchemy),22 and as a pleas in its manor.14 neighbour paying tithes to Oseney and in dispute with the town in 1515 over enclosing and water The Studium management.23 Theabbey was sufficiently prosper- ous in the 15th century to engage in some building Something of the arrangementsfor the studies of the activity, and benefactions were made to works on monksatOxford is indicatedbythe records of the the church in 1461 and the rebuilding of the choir in Cistercian General Chapter. In 1289 the places of 1488.24 The income of the abbey was reckoned to be study were Paris, Oxford,Montpellier, Toulouse and £107 in 1526, and £188 15s 11d (£174 3s clear) in Compostella,15 and in 1292 the chapter commanded 1535.25 The king seems to have considered retaining that every monastery with 20 monks should main- Rewley as agrammar school, and in 1536 Abbot tain one at University. Following this the Abbotof Austen offered Cromwell £100 to savethe mon- Citeaux ordered that Cistercian houses in the Prov- asterybyconverting it to aCollege, but in the event ince of Canterbury shouldsend students to Oxford, the abbey was suppressed in that year and the and each student to have an allowance of 60 s.16 abbot was pensioned off and retired to Cambridge.26 In 1335 Pope Benedict XIIruled that Cistercian The site was soldtoDrGeorge Owen in 1541, students of England, Scotland, Ireland and Wales together with the lands of Godstow Nunnery, though were to use the University of Oxford bothfor he later released them to Henry VIII forthe foun- primaryeducation ( primitive scientie )and for theol- dation of Christ Church in 1545.27 ogy.17 That the studium was active in the first halfof the 14th century is shownbyaformulary of model letters of c 1340 containing references to Cistercian Post-dissolution history students at Oxford, and otherreferences.18 However, the system of financial support for the monk The church and cloister were quickly removed, and students doesnot seemtohave worked, and by by chance we knowthat the windows from the 1381 the buildings and possessions of the studium abbey were taken to Hampton Court in the autumn were forfeited to crown before being returned to the of 1537, apparently for the king’s new bowling abbey. The crown’s intervention was on the ground alley. Atotal of 40 windows (20 from the church that there was abreach of the Statute of Mortmain, clerestory and 20 from the cloisters) were used, and but it was observed that there were ‘buildings within lead from the roof was used to cover the queen’s the bounds of the abbey, built at the joint expense of new gallery.28 The remaining buildings were con- certain abbeys of the order in Englandfor the abode verted to domestic use, and leased out by Christ of monks of the order studying in the University of Church.Aslandlords,Christ Church took on the Oxford’; theyhad been forfeited to the king and now role of protagonist with the city over the control of belonged to the abbey of Rewley by the king’s gift. water, and in particular Rewley Lock, over the next Saltersuggested that the Earl of Cornwall’s build- decades; this included the city closingthe water- ings had been superseded by ones built by Cistercian course ‘goingtothe howse of offyce of Ruley’ in abbeys,and that these were all now givento August 1576.29 Monastic sites were often converted Rewley.19 for industrial purposes,and in 1555 aleasewas Acomplaint was made in 1398tothe general drawn up for asite at Rewley for afulling mill, but chapterfrom Cistercian scholarsatOxford about the the main building was leased as afarmhouse with provision of buildings, maintenance and books. adjacent landtoNicholas Woodson, yeoman from They were evidently not living at Rewley, and 1553 to 1580.30 The large number of leases at Christ about then we hearofCistercians renting rooms at Church is repetitive and none provides much University College, andlivingatTrillocksInn (New topographical information, nor indicatesthe extent Inn Hall).20 An attempt was made in 1400 by of sub-letting,which was usual. They are here adelegation from the general chaptertoorganise summarised (Table 2).31

7 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Table 2Christ Church lessees of Rewley.

Date Lessee Occupation

23 Eliz (1580) Francis Willis PresidentofStJohn’s College 1618 Katherine Willis Widow of Francis 18 Charles I(1642–3) John Rouse ‘of the University’ No date William Hutchenson – 1650 Sylvester Hutchenson Daughter of above 1659, 1667, 1674 John Hicks Gentleman (of Gloucestershire) 1684 Unton Croke Of Inner Temple, London 1695, 1702 Robert Horne Of the University, M.A. 1716, 1720, 1723 Timothy Bourne Gentleman 1737, 1744 John Saunders Gentleman of the parish 1758 Revd Benjamin Blayney Worcester College 1765, 1772 Dr David Durrell Principal of Hertford, D.D. 1779, 1786 Thomas Spencer Of St Clement’s,horsedealer 1793 Richard Godfree &wife Of London,upholder 1800, 1807, 1839 James Rowland – 1851 Acquired by Bucks. Railway –

Antiquarian discoveries ‘That part of the Abbey of Rewleynow stand- ing which seems to have beenpart of the In the early 18th century Thomas Hearne made a Chapell to some did notatall belong to the systematic study of Oxford antiquities, publishing Chapell as appears from the whole Circumfer- notesinthe appendices to his numerous publica- ence of the Chapell lately discovered, whenthe tions, and often illustrated by the accurate engrav- Foundation of the Walls were dug up. It was 32 ings of DavidLoggan’s pupil,Michael Burghers. in the Garden on the East side as you enter His diariesinclude many references to his investi- into the House, &did not cometothe said 33 gationsatRewley, and discoveries made there. part nowstandingbyagreatmany Yards. Ican- His major discovery was the Countess of Warwick’s not learn that any coffins or Bodys have been foundation stone in 1705 (Plate 1), which he bought ever dug up; so that it should seem their bury- for halfacrown (see below), and he found coins ing place was at St. Thomas’s Church,orelse 34 there in February 1706. He madeanimportant at Oseney, &perhaps thismay be one reason observation about the standing north range later why Ela Longspee(that built Rewley Chapell) that year referringtoanotherwiseunrecorded was buriedinthe Chapell of Oseney, as ap- excavation: pears from Evidences cited by Mr. Leland’.35

Plate 1Countess of Warwick’s foundation stone q Oxfordshire CountyCouncil.

8 OA Occasional Paper No: 16

On avisit in May 1706 he notedthe arms at the gate plan made in 1775 showed the north range still (see below) and some surviving glass: standing, but when surveyedin1829 it had been reducedtoarump at the west end.42 In May 1849 an ‘At Rewley, just as you enter intothe House now agreement was madewiththe Great Western Rail- standingare the Armsofthe Earls of Cornwall way for the purchase of five acresinoccupationof &c., andinaCloset on the south side are the James Rowland(i.e. the present station site), with a Images in Glass of some of the Apostles, as St. subsequent valuation, and an agreement that the Andrew,St. Thomas &c. And as you go upstairs money would be invested for Rowland’s benefit.43 in the window are these Arms, viz. Azure aBend 36 At the final sale to the Buckinghamshire Railway Sinister argent, &parted per Pale Azure &or’. Company in January 1851 the valuation stated that Odd notes about Rewley occur at various points in the property consisted of Rewley gardens, dwelling the diary, and in 1717 he noted, ‘There are two house,malthouse and sundry buildings (12 ac., 1 1 FishPonds remaining at Rewly that belong’d to ye rood, 27 ⁄4 perches), boundedwest by the intended Monksthere’.37 In his usual manner, Hearne GWR station, and south by the turnpike road. It was included an engraving of the foundation stonein the freehold of the Dean and Chapter, with avalue of his 1718 publicationofLeland’s Itinerary, and in £12,355 (in relation to the price of land in the vicinity, 1719 he remarkedonthe ownership of Mr Bourne, a which exceeded £1000 per acre,and the scarcity of brewerinStThomas’s parish, who now wished to freehold landinthe areabecauseofcollege owner- sell.38 Ayear later Hearne publishedhis edition of ship, and the expansion of that part of Oxford in the the Textus Roffensis, which included the Burghers previous 30 years).44 view of Rewley (Fig. 5, 8, 10 and 13), notable for showing brewing in progress. And in March 1725: Buildings of Rewley Abbey ‘This day walking to Rewly Isaw many Altera- The buildings of Rewley Abbey are known from a tions makingonthe North sideofthe Remains series of pictorial representationsdating from the of that House,now turned intoaBrew House, 16th to 19th centuries, supplemented by anumber of particularly apassage madethroughthe Wall later plans and descriptions. The birds-eye view on for better convenience of the Place,asitisnow Ralph Agas’s map (Fig. 2) of 1588 is instructive, but aBrewHouse.’39 must in part be interpreted by referencetoDavid Loggan’s carefulmapping of 1675 (Fig. 3). This Finally in 1732 Mr Bourne, described as asadler and shows the extent of the precinct, the outline of the maltster, died at ‘Rewly House’and was buriedat moat, and the fishponds. St Thomas’s: The key illustration is Burghers’ view of 1720, ‘Tis pretty certain he did not thrive at Rewly, which is adetailed archaeological recordofthe which was areligious House, as ‘tislikewise standing remains at that date, viewed from the south, prettycertain that never anyone thrived at it west and north, and including the outer gate. The since the Dissolution. Mr Bourne used to love brewhouse (Fig. 10, a) was taken to have been the to talk of Antiquities’.40 ChapterHouse (which it was not), and the 1574 datestoneonthe chimney(c) with initialsFW, KW. Final years LW &RW(not shown) must relate to the Woodson family, though LC 1711 does not relate to aknown The later historyofthe estate, recorded in Christ lessee.The most interesting part is the depiction of Church records, included surveys in 1775 and 1829, the north range as atwo-storeyed building, possibly and valuations madein1807, 1828, and 1842.41 The with medieval windows on the first floor (Fig. 5).

Figure 5Aview of Rewley Abbey from the north, drawn by MBurghers in 1720. HenryWTaunt Coll. (HT12905) q Oxfordshire County Council.

9 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

The heavy offset or corbel table along the build- lived, as it were the Lay Brothers of the standard ing may on the otherhand indicate the topofthe Cistercian plan), and only minimalevidence forthe medieval walls andthe level at which rebuilding east range and its cloister walk in one trench. From took place in the 16th century or later. At the east the cloister came atotal of 20 windows in 1537, that (left-hand)end there are two projections, one with a would be five to each walk. For the north range, door and one with an arch over the stream (presu- the limited trenching was able to corroborate some of mably the reredorter) and amakeshift bridge emer- the evidence of topographical views. The windowre- ging fromahole in the wall. Later drawings confirm cess at the west end for atriple splayed opening was the general layoutofthis drawing (apart from one located with the details show by Grimm and Turner, very misleading view publishedbySkelton which is perhaps as aground-floorwindowbeneath the single fore-shortened). Malchairdrewasplendid if unin- windowshown on the first floor by Burghers and Tur- formative view of the ruins with the trees reflected in ner. The early views show that the groundfloor was the water of the moat(see back cover),whileBuckler’s well fenestrated on the north side, though quite how view of the buildingsshows that the north-west the monastic refectory and kitchen were fitted in, let cornerhad been rebuilt with its roof running north- alone any provision forthe monk students,remains south instead of east-west as shown by Burghers. It quite uncertain. Grimm’sillustration of alarge late was this cornerbuilding that survivedtothe end, and medieval fireplace (itsmantel decorated with apanel one finalview of the interior shows the fireplace and of quatrefoils) is likely to have come from this range.46 windowopening that was recoveredinthe excava- With the arch over the moat, of which more substan- tion. The ruins were frequently shown by topogra- tial remainswere found, this is very likely to be the phical artists,e.g. William Turner ‘of Oxford’ (Fig. 7), common lavatory or reredorter, which shouldbeat though notnecessarily with much informative detail; the north end of the east cloister range (leadingout of the views by S.H. Grimmin1773 are valuable for the dormitory). Apost-medieval phase of modifica- corroborating the details of the Burghers views.45 tions to this structure seems to relate to its use as a bridge. The excavations have clearlyshown the potential Excavation results of the site fordiscovering more about this unusual The excavations, more fully described in the next Cistercian Abbey, while leaving many questions section, for the first timelocated the principal build- unanswered, and aboveall justhow the buildings ings of RewleyAbbey, whileleaving many remain- were divided between the two communities. ing problemsabout the details of the monastic plan. The preparation of the site was clearly revealedinthe dumping of materials to raise the ground level, and ARCHAEOLOGICAL DESCRIPTION the creation of aseries of ditches –virtually amoat – Andy Simmonds aroundthe abbey site. Finds dating from before the Before the Abbey late 13th-century were perhaps brought in with spoil for dumping. The precinct wall was constructed as the Documentary sourcesrecords that in 1281 Edmund, internal ground level was raised.The church was Earl of Cornwall, granted the site of his manor at located,asdescribed by Hearne, though the walls had North Oseneytothe Cistercian order for the foun- been severely robbed. The south wall of the church dation of RewleyAbbey. However, these sources was well defined, and the probable junction with the are less informative as regards the previous use of south transept was identified, but only asmall part of this land. In particular it is unclear whether amanor the north wall was seen. Remnants of four pier bases house or othersettlement already existed on the site allowedageneral reconstruction of the plan, but little on which the Abbeywas to stand. was found at the east end, and nothing at the west Thevarious episodes of excavation carried outatthe end. This gives an approximate plan area of 165 by 50 site have yielded tantalisingresultsregarding the pos- ft, comparable with the 13th-centuryparish church of sibilityofearlier occupation but have notbeen able to St Giles(115 by 55 ft) or the 15th-century rebuilding of resolve this problem conclusively.Althoughasmuchas St Mary the Virgin (175 by 60 ft). It is half to two-thirds onethird of the potteryrecovered from theexcavations the size of Hailes Abbey. It must be recalled that atotal could pre-date theestablishmentofthe abbeythis of 20 clerestorey windows were removed from the material wasrecovered predominantly from layers church in 1537 (St Mary’shas 22 windows). The only whose origin is uncertain andcouldderivefrom difficulty is with the very narrow aisles, but then the dumping associated with theconstructionofthe church is not likely to have had much of aparochial abbey. In Trenches 1986/1 and1986/5, forexample, function. the earliest layers associated with the occupation of the More substantial remains were encountered of the abbeyoverlay asuccession of gravelly loam deposits ‘brewery’ building west of the church, with remains thatyielded pottery that need notdatefromlater than of abrewing vat, but no indication of its function (it the mid-13th century.Similarly,duringthe 1993 seems to have had adoortothe cloister), or the investigation potterydating from the12th–13th cen- possibility that it may have been aseparate chapel for turieswas recoveredfromTrenches1993/3and 1993/4, the monk students. Only slight traces were found of although in theformeritwas associated with asmaller the west cloister range (where the students may have quantity of potterycontemporarywiththe abbey.

10 OA Occasional Paper No: 16

Figure 12 Plan of the chapel/brewhouse in Trenches 1986/1 and 1986/2. more than 0.5 mdeepbut couldnot be excavated to pipes with diagnosticbowl forms indicating that the its full depth due to the depth of the trench and the featurewas filled during the early part of the 18th instability of the overburden.Aneffluent culvert (10) century. It is uncertain whether this meansthat the that drained westwardfrom the vat and cut through building fell out of use at this time, or whether it the wall of the building contained pieces of clay continued with adifferent function until the final

21 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure13Aview of theruins of Rewley Abbeyfrom the west,drawn by MBurghers in 1720. Thegable endofthe chapel/brewhouse canbeseenatthe right of thepicture.Henry WTaunt Coll.(HT12905) q OxfordshireCountyCouncil. clearing of the site for construction of the railway in (801, 810) aligned NNW-SSEwereexposed, which the 19th century. Sherds of medieval windowglass were likely to be all that remained of the range.The recoveredfrom the steeping vat and the floor surface removal of stonework had beenparticularly severe may have derived from the windows shown by in this partofthe site, and stone survivedonly at the Burghers. deepest level of robber trench 801. Robbertrench 801 lay on the same alignment as wall 1852, interpreted as being the west side of the south transept, and The west range shouldtherefore have been the west wall of the east The west range of the abbey cloister extended north range,inwhich case robber trench 810 represented from the formerchapel. The only part of this range that the wall of the cloister walk. Alarge cut feature at has been recorded was astub of wall (220) that butted the westernend of Trench 1986/8 may have been onto the north wall of the chapel and was of the same part of adrain within the cloister. Thecloister walk build. The masonry of this wall extended for only 0.6 was flooredwith mortar (812) bedded on alayer of m, beyond which its line was continued by arobber sandyloam (813), whilethe surface insidethe trench (233). Adoorway was identified between the cloister itself was ayellowish white mortar (821) stub of the walland the robber trench, through which with two gravelmake-up layers (822, 824) to the afragmentary tiledfloor (238) extended from the west of robber trench 810. An areaofpitched stones westernside, andamortar surface with tile impres- (815) to the east of robber trench 801 may be the floor sions (265) was exposed on the eastern side. Thetiles of of aroomwithin the range. floor 238 were plain and likely to be 16th-centuryin Trench 1986/5was excavated in an area that date, representing are-surfacing of the floor late in the shouldbeinternal within the east range, but was abbey’sexistence.Within ahollow that had formerly unable to locate either walls or floorsurfaces. Instead, held adoorpost, the cornerofanearlier stone the natural alluvium (506, 508) and made ground threshold was exposed some 0.15 mdeeper, indicating (502–505) were overlainbyalayer interpreted as that earlier floor surfaces may be preserved beneath either amiddenoragarden soil manured with those recorded in the evaluation. middenmaterial (500, 501). The finds recovered from These remains were sealed by athinlayerofrubble this layerincluded the lower portion of the base of a (222), possibly associated with the demolition of the hanging lamp as well as aconcentrationofsherds range at the Dissolution, which was cut by the from Brill/Boarstall ware bottles, the latter including construction of alater wall(234) that is likely to have one almost complete example.These vesselsmay been partofanew range constructed during the use of have servedascontainers forrefilling oil lamps and it the formerabbey as adomestic house during the post- is tempting to associate them with lighting for the medieval period. This range was notpresent in 1578 monks’ dormitory, which was usually locatedover when Agas drew his map of Oxford (Fig. 2), and was the east range, although they could have been first recorded on Loggan’s map of 1675 (Fig. 3). dumped hereafterbeing used elsewhere in the complex. Indeed, one possible interpretation for this depositisthat it is refuse dumpedhere as amake-up The east range layer for afloor surface that has not survived. Other Trench 1986/8 was targeted on the assumed location examples of these vesselswere foundlargelyinthe of the east range of the cloister.Two robber trenches vicinity of the church, particularly in the midden

22 OA Occasional Paper No: 16 outside the south wall, where they mayhave been windows (Fig. 15). Within the northern part of dumped after being used to refill lamps lighting the Trench 1994/17 the north wall of the range was church. identified as arobber trench 1.2 mwide (1701). No east range is shown on any of the historic In the interior of thispartofthe rangetwo thin depictions of the abbey, indicating that thisrange make-uplayerswereoverlainbyapatchy tiled floor was completely demolished at or shortly afterthe (1720) on amortarbedding(1721). Thelevel of this Dissolution. This is confirmed by sherds of Brill/ floor matchedthe base of the recess in thewall, and Boarstall ware ( c 1225–1600),Cistercian-type ware wasclearlythe originalfloor level. Abovethis wasa ( c 1475–1650) and Surrey/Hampshire whiteware layerofaccumulatedsoil(1724)thatcontained ajetton ( c 1550–1700) found in robber trench 801, although dating fromthe mid16thcentury (SF14).Itisunlikely more recent pottery recovered fromrobber trench thatsuchalayerwouldhavebeen allowed to 810 indicates that further robbing took place, accumulate while the buildingswereinuse, so it possibly when the railway was built. seems most likely thatthisdeposit datesfromafterthe abandonmentofthe abbey. Re-occupation of thenorth rangeasadomestic residence is represented here by The north range subsequentre-surfacing of thefloor withsuccessive The north range, which on analogy with other mortar layers (1736, 1734)and an intervening make- Cistercian housesislikely to have housedthe up layer(1735), none of whichyielded datablefinds. refectory, was investigated by two trenchesexcavated The1850drawing of theinterior of thisbuilding during the 1994 evaluation, Trenches1994/16 and showsthatithad abrick floor at thistime, whichmay 1994/17. have been bedded on theupper mortar layer. Trench 1994/17 (Fig. 14) was positioned to Trench 1994/16 was excavatedacross the north examinethe western end of the range.The earliest side of the north range,targeted on the locationofa featurerecorded in this trench was awall that pre- structure over the moat depicted ashort distanceto dated the construction of the north range. Thewall the east of the norh range on Loggan’s map (Fig. 3) (1703) was recorded nearthe north-western cornerof and on Burghers’ view of the abbey (Fig. 5, extreme the abbey precinct and lay on aNW-SE alignment. left). The structureisshown adjacent to adoorway in Only the foundation survived, 1.1 mwide and asection of the outer wall of the north range that has constructed rather roughly from small piecesof been stepped out towards the moat, but its precise limestone. It was set into the natural alluvium and nature is unclear. Much hangs upon the date of the was presumably trench built as no construction cut structure, which is uncertain.Agas’ map,drawn in could be identified. The wallwas overlain by alayer 1578, does not show thisstructure, but doesdepict a of demolition rubble (1739/1744) contained pottery structure projecting over the moat from the rear of datingits destruction to the late 13th–14th century. the north range. If this is in fact the samestructure Although thisdatewould fit with thisbeing an and Agas had simply mis-located it, then it is earlier building demolished to clear the groundfor probablymedieval in date and is most likely to be the construction of the abbey, this cannotbethe case areredorter associated with the abbey. However it is as the demolition layer overlay the construction cut also possiblethat the structure shown by Agas had for the precinct wall, which is unlikely to have been been demolishedby1675, when Loggan drew his built before 1299. No evidence was foundtoindicate map,and anew structure, possiblywithadifferent the form or function of the building of which the function, had beenbuilt over the moat furthertothe wall had formed apart. Its location suggests that it east. Burghers’ drawing, dating from 1720, seems to could have been an early and short-lived predecessor show it being usedasabridge across the moat, with of the north range of the cloister, but it could equally afigure crossing it, but on Loggan’s map the north have been afree-standingbuilding. The demolition side of the structure appears to be asolid wall. It rubble was directlyoverlainbypost-medieval would therefore appear that the structure originally garden soils, and so presumably remained as an had some otherfunction and that the opening area of hardstanding throughout the occupation of throughwhich Burghers’ figure crossesthe moat the abbey. was broken through between the drawing of the two The eastern end of the demolition layer was cut by images. the foundation trench (1750) for the west wall of the The north wall of the north range (1612) was north range.Anoff-set foundation of small un- identified extending intothe trench from the south shaped stone supported awall 1.2 mwide and well- west on the samealignment as the robber trench in constructed from roughly squared limestoneblocks Trench 1994/17 (Plate 8). Ashort distanceinto the (1702).One course (0.18 m) above the foundation trench the wall underwent achange of alignment the wall had a0.6 mdeeprecess flanked by the (Fig. 16), stepping out by adistance of 2.25 mtothe lower partofasculpted triangular stone jamb with north west,asindicatedonthe historic views. The adecorated roll moulding at its apex (Plate 7). The formerlocationofadoorway through this stepped- jamb has been dated on stylistic grounds to the mid out section of the wall was indicated by large shaped 14th century.Adrawing of the samewall from 1850 stones marking the jambs and avoid where the sill shows several of these jambs defining threewide had formerly been(1620) To the east of this the recesses along the west wall containing doors and stonework of the wall had beencompletely robbed

23 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford 1994/17. Trench range, north the of end western the of section and Plan 14 Figure

24 OA Occasional Paper No: 16

Figure 15 WTurner’s drawing of the room at the westernend of the north range, drawn in 1850, showing three recessesflanked by jambs with roll mouldings. out, although its line was indicated by arobber masonry wallset into the northern bank of the moat trench (1651) that continued beyond the eastern end and running parallel to the outer wall of the northern of the trench. range of the cloister (Plate 9). This abutment must An abutment(1632) exposed on the north side of correspond with the northern side of the structure the moat was interpreted as being partofthe possible shown by Loggan and Burghers and is likely to reredorter. This comprised asubstantial piece of be the wall exposed in the trench excavated in 1981.

Plate 7The end wall of the north range (1702) showing the stone jambwith roll moulding exposedinTrench 1994/ 17. Asequence of medieval and post-medieval floor surfaces can be seen at the foot of the section to the left of the frame.

25 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford 1994/16. Trench in reredorter the of section and Plan 16 Figure

26 OA Occasional Paper No: 16

Plate 8The northern wall of the north range exposedinTrench 1994/16. The doorway (1620) can be seen to the left of the frame.

The wall was 1mwide, constructedfrom large blocks the moat, but this part of the trench was not excavated of masonry neatly faced on its exposed southern to sufficient depth to locate it. On the southern side side. The wallwas not flushwith the edge of the moat, of the wall 1612, withinthe north range,were the but had been constructed 0.1–0.2 mproudofthe cut, remainsofafloor comprised of amortar bedding with the gap behind it having been filled in with layer (1617) bearing the impressions of tileswhich gravel(1660) that contained three sherds of pottery had evidently been removed. Part of areturnwall datingfrom the 12th–13th century. This datewould (1615) was exposed extending to the south east from be consistent with the structure having been partof the point at which the north wall changed alignment, the original construction of the abbey, and thus being and thisislikely to have beenaninternal dividing the same as the reredorterdrawn by Agas. wall. These features are likely to date to the occupa- Loggan’s depiction of the structure as triple- tion of the abbey, as Loggan shows that by 1675 the arched was confirmed by the discovery of afurther range had been shortened and no longerextended abutment (1666) set withinthe moat. This feature this far east. was uncovered in asondage excavatedwithin the fills of the moat c 1msouth of the northern Outbuilding east of the cloister abutment, and would have been one of the piers supporting the structure’s central arch. Thepier was The foundations of two walls (405, 416) that are 0.9 mwide and extended beyond the trench to north- likely to form partofabuilding were uncovered to east and south-west. Excavation was not deep the east of the cloister in Trench 1986/4, the eastern enoughtoinvestigate its base, the earliest layer wall extending into Trench 1993/2. The foundations encountered being alayerofdark grey silt (1667) were trench built and cut into dumpsofclay and that had built up against the southern side of the gravelused to raise the ground level. Both had been pier. This was overlain by layers of gravel (1665, substantially robbed but were clearly the remains of 1659) that overlay the abutment, the latter containing abuilding or room measuring 4mwide and at least 19th-century pottery and an incompletebone comb 6.5 mlong and alignedNW-SE. This alignment was (SF 33) dating from the 17th century or later. somewhat askew from that of the nearby northern The otherabutment of the central arch should be lo- wall of the abbey precinct and indicated that the cated between abutment 1666 and the south side of building was free-standingrather than buttingonto

27 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 9Masonrywall(1632) parallel to the outer wall of the northern range of the cloister. this wall. It was, however, parallel with the riverside The garden soil in Trench 1993/4 was overlain wall that definesthe eastern side of the precinct. No by amortar surface (412) associated with aNW-SE floor surfaces were identified within the building, alignedwall of which only the lowest course and this absence of formal flooring suggests that survived(408). Asimilar wall (116) crossed Trench the building was utilitarian in nature.Agravel 1993/1onacomparable alignment. Neither wall surface (407) lay on the western side of wall405 and possessed foundations, so theywere unlikely to be could have beenapath or the floor of an adjacent structural and are more likely to be garden walls. In room.The remainsofthe building were overlain by a the area behind the precinct wall in Trench 1993/1 layer of rubble (400) presumably resulting from its the absence of medieval deposits stratified between demolition. No evidence was found to provide a the made ground associated with the construction of date for eitherthe construction or the demolitionof the abbey and the overlying post-medievalgarden the building. Alarge cut feature(213/234) was dug soils may indicate that these soillayers were also through this layer and had dug away the south- being cultivatedthroughout the medieval period. eastern end of the eastern wall of the building. Although only part of this feature was exposed The barn within Trench 1993/2, it appeared to be rectangular in form and is likely to have been one of the ponds Trenches 1994/19, 1994/27 and 1994/28 were shown in thispartofthe site on the historic maps. targetedtoinvestigate the site of abarn shown on Sherds of 17th-centurypottery recovered from one of Agas’ map of 1578 acrossthe moat west of the main its fills (206) confirms this as being apost-medieval cloister (Fig. 2). The earliest features cutting into the featureassociated with the conversion of this part of pre-abbeyagricultural layers discussedabove were a the site into aformal garden. pit (1985) and two ditches (1979, 1999) that contained mainly earlier pottery but with afew sherds dating to the 13th and 14th centuries. Thewest wall of the Garden soils barn was identified as aroughly built stone founda- In Trench 1986/5and Trench 1993/4 soil layers (501, tion (1938) 1.1 mwide andextending acrossTrench 407) were identified that contained large quantities 1994/19 on aNE-SW alignment parallel to the edge of pottery and bonewhich are likely to indicate of the moat. The foundation was trench built and areas of garden soils fertilised with middenmaterial. cut into the fill of ditch 1999. On its internal side

28 OA Occasional Paper No: 16 was amortar surface (1926),the make-up layerfor of records relating to the abbey, and the existence which (1951) contained pottery contemporary with of such an installation would be consistent with the the construction of the abbey. Partofarobbed-out agricultural basis of the monastic community as well wall (2708) recorded in Trench 1994/27 is likelyto as with other Cistercian houses. have beenthe northern end of the building, where Adrawing showing arepresentation of Rewley apossiblebuttress (2710) was also discovered. A Abbeyinits original state is shown on Figure 17. short length of stone foundation (2815) was exposed to the south in Trench 1994/28 but its interpretation After the Abbey is uncertain as it was substantially cut away by later features. Nevertheless, it lay on the same alignment The abbey was dissolved in 1536 and shortly after- as foundation 1938 and so was likely to be acontinu- wards records indicate that stone and timber was ation of that wall. No evidencewas found for the being sold off. This was likely to have beenthe cause of eastern wallofthe building, which had presumably muchofthe stonerobbing recorded in the excava- been completely destroyed by later groundwork, tions, particularlyinthe church and the east and west but the existing evidence confirms the presence of a ranges, no trace of which survivedin1578 when Agas substantial buttressed barn measuring at least17.5 m was drawing his map of Oxford (Fig. 2). The in length. accumulationofsoil (1724) on the partially robbed All the stone elements of the barn had been subject tile floor of the north range noted in Trench 1994/17 to stone robbing, and in all three trenches the robber may indicate aperiod of abandonment before the site trenchesweresealed by alayer of demolitionrubble was re-occupied as adomesticresidence, and this (1909, 2703,2808) the latest pottery from which was would account for the absence of diagnostic 16th- Surrey/Hampshire whiteware dating from c 1550– centurypottery. 1700. This was overlain by alayer of loamy soilthat Remarkably little evidence was uncovered for was likelytohave resulted fromcultivation of this post-medieval activity at the site,although it is area until it was buriedbeneath the thick dump levels known to have remained in occupation into the 19th associated with the construction of the railway. century. Historic depictions of the site show that the north range was retained as the main residence, and this occupationisevidenced by the sequence of Northofthe Abbey complex mortar floors recorded in Trench 1994/17. The finds Part of afeature interpreted as afishpond (1645) from this period are predominantlydomestic, includ- was discovered to the north of the moat in Trench ing asmall assemblage of glass vessels representa- 1994/16. Theshape and dimensions of the feature tive of areasonably wealthy household, although the could not be ascertained as only the north edge discovery of sherds of ceramic ‘chicken feeders’ from was identified within asondage at the end of the agravelsurface on the north side of the moat in the trench, but it was more than 0.9 mdeep. It was 1981 excavationserves as areminder that the site filled by aseries of silt deposits (1641–2,1668–9) was aworking farm as well as adomesticresidence. that contained no useful datingevidence, but it was The main evidence discovered for new building likely to be the sameasalinearpond shown extend- during the post-medieval period was the construction ing along the north side of the moat on Agas’map of anew west range shown on Loggan’s map (Fig. 3), (Fig. 2). Thefishpond did not appear on Loggan’s one wall of which (234) was recorded in Trench map (Fig. 3), by which timealine of trees is shown 1986/2. Theadjacent chapel may have remained in extending along this area, and so it was presumably use as abrewhouse until the final destruction of the filled in during the late 16th or early 17th century. site in the 19th century.After aperiod of disuse during Furthernorth, Trenches 1994/11–1994/14,located which asoillayercontaining 17th century pottery north of Rewley AbbeyStream, and Trench 1994/23 accumulated,the gravelpathcurving aroundthe and 1994/24, locatedinthe area of the island north south-west side of the brewhouse was superseded by of the main abbey complex, generallyuncovered aseries of superimposed metalledsurfaces creating a only natural alluvium and 19th-century made raised track leadingtothe bridge across the moat to ground. This would be consistent with the agricul- the west. tural use of these areas during the medieval and As mentioned above, there is doubt as to the post-medieval periods. However, asondageatthe datingand function of the structure over the north north west end of Trench 1994/24 revealedaNW-SE part of the moat shownbyLoggan and Burghers. alignedditch (2420) sealed beneath adepth of more Although thismay be the remains of areredorter than 1mofalluvial clay. In addition to pottery asociatedwiththe abbey, it is also possible that it is a contemporary with the abbey, the fill of the ditch later structure, built after the drawing of Agas’map contained waterlogged and charred plant remains in 1578. Burghers shows that by 1720 it was being that indicated that the drying of grain prior to used as abridge,but thisisnot consistent with milling and of malted grain had taken place nearby. Loggan’s earlier depiction of aclosed structure, This environmental evidence may hint at the pre- suggesting that it originally had adifferent, currently sence of amill in this area. Although no mention unknown, function. Ahard-packed surface of gravel of amill survivesinthe documentary evidence, this (1627) and stone (1633, 1629) excavated on the north could be simply aproduct of the general paucity side of the moat may have been associatedwith

29 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford Abbey. Rewley of epresentation Ar 17 Figure

30 OA Occasional Paper No: 16 the track shown beyond the moat on Burghers’ outer embankment was thrownupbeyond the engraving, extending to the north from the bridge. medieval city walls, and de Gomme’s map,drawn The surface contained pottery dating from the 18th in 1644, clearlyshows that the site of Rewley Abbey centuryand was likely to have been the same deposit was included within this, with the north and west as layer 11 recorded in the 1981 trench. arms of the moat and the precinct wall being Much of the grounds of the formerabbey were incorporated intothe defences(Fig. 18). Partofthis converted intoaformal garden,asindicated by the rampart was recorded to the south of the site of garden soils that overlay the medieval remains in the abbey in awatchingbrief conducted by AOC most of the trenches. The probable pond which Archaeology during road widening between Botley truncated the southern end of the outbuilding in Road and Hythe Bridge Street(Fig. 19). Ground Trench 1993/2was likely to have been one element reduction exposed a20mlength of the defences, of the landscaping that accompanied the creation of comprising asubstantial ditch and dumped rampart. this garden.Alarge cut feature (139) recorded at the The ditch (16) was 12 mwide and ran north-south, south-westernend of Trench 1993/1 could have turning slightlytowards the west at its south end. been the end of asecond such pond.One of its fills Only the upper 1mofthe ditch was excavatedasthe (140) contained pottery dating from the 16th or 17th remainder extended below formation level, but this century. The feature had subsequentlybeen cut by was sufficient to establish that the western, outer the trench for astone-lineddrain (135, Fig. 6). The edge had quite agradualslope while the east side drain may have been associated with the drainage or dropped almost vertically. No datable finds were irrigation of the garden, although how it functioned recoveredfrom the fills of the ditch,but pottery, is unclear as terminated within the trench without glass and agreatmany pieces of clay pipe were connecting with any otherdrainage features. recoveredfrom the groundsurface into which it was cut (30). Thelatestofthese dated fromthe early-mid 17th century, confirming the identification of the Civil war defences earthwork as aCivil War defence.Aberm 2.5 m Oxford was the Royalist capital for most of the Civil wide separated the ditch from abank c 8mwide that War and as such was strongly defended. Anew survivedtoaheight of 0.5 m. Thebank was composed of asequenceofredepositedsubsoil and natural gravel(40–44),presumably spoil excavated during the digging of the ditch. Adeposit of limestone rubble,some of it worked, lay on the berm and although no structural form was apparent it is possiblethat this material was used to revet the face of the rampart. It is also possible that some or all of this material may have been obtained from the ruins of Rewley Abbey. Layers 36, 38 and 39 to the rearofthe bank may be derived from the levelling of the upper part of the earthwork. The barn to the west of the moat may have been demolished when the defenceswerebuilt, as the latest pottery in the layer of demolition rubble overlying the barn was Surrey/Hampshire whiteware datingfrom c 1550–1700. The demolition of any buildings outside the defensive line,defined in thisareabythe moat, would have beenapractical course of action to create a clear fieldoffire for the defenders while simulta- neously gaining stone that could be used in the defences. The demolition of the barn at this time would also explain why there is no structure in this area on Loggan’s map (Fig. 3).

Possible sconce Possible evidence for another defensive installation datingfrom the Civil War was uncovered in Trench 1994/24, in the areanorth of the site of the abbey. The natural alluvium was overlain by alayerof gravelly sand(2416) from which asingle sherd of mid 11th- to late 12th-centurypottery was recovered. This layer was cut by aditch 5mwide. The ditch Figure 18 De Gomme’s map of the Civil War defences of (2406) couldonly be excavatedtoadepth of 0.5 m Oxford, drawn in 1644. due to the high water table, and the fills encountered

31 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford 09 15 16 50 32 brief. 30 42 44 watching 33 1999 the 31 41 in 42 exposed defences 40 War 36 Civil the 38 39 through Section 19 Figure

32 OA Occasional Paper No: 16 contained no finds. Asingle course of roughly laid c 1294, but Jeffrey North (1997, 11–13) has suggested limestone pieces(2412) ran along the northern edge that it remained open from 1280/1to1295. North of the ditch and may originally have revetted abank, (1997, 17) tentatively dates the farthings of the Water- since removed. Althoughnodatingevidence was ford mint and similar coins from the Dublin mint to foundtoconfirm such atheory, it is tempting to c 1280–1284. Thefarthing from Rewley Abbeycan be identify the ditch with that surrounding asconce dated to 1280/1–1295, with aprobabledate in the depicted to the north of the site of RewleyAbbeyon early 1280s. de Gomme’s map (Fig. 18). Areview of post-1279 Irish coins of Edward Iincoin hoards(Dolley 1968, 277–9) has shown that they appear in Englishhoardsdepositedasearly as c 1285, Constructionofthe railway and it canbeassumedthat they were quickly inte- The remaining buildings were demolished during grated into the Englishcurrency as exact equivalents the mid 19th centurytomakeway for the construc- of the English penny, halfpennyand farthing. tion of the Rewley Road station.The absence of The relatively unworn conditionofthe Rewley demolition layers associated with this event indi- Abbeycoin suggests that it mayhave beendeposited cates that the building materials recovered from the on the site between the 1280s and the mid 14th dismantling of these structures were taken off site century. and presumably re-used or sold off. The moat and otherdrainage channels were filled in and the Jettons groundmadeupbythe dumping of 1–1.5 mofclay, identified by Edmund Simons (see table below) as can be seenonthe section drawings above (Fig. 6, 14, 16, and 19). This material cannot have been Medieval and post-medieval pottery sourcedatthe site and was presumably broughtin John Cotter from elsewhere by rail. In Trenches1994/16 and Introduction 1994/24 this made groundalso included deepdepo- sits of refuse containing much pottery and glass, The pottery reported on here comes from four which may have been brought from the city. separate sets of excavations that have taken place within the abbey precincts over the years. These are coded OXRA 81, OXRA86, OXRR93and OXSTRA FINDS 94. Togetherthese comprise 3044 sherds of pottery weighing 48.885 kg. The 1986 and 1994 excavations The medieval coin produced the highest numbers of sherds (1238 and Martin Allen 1239 sherds respectively) but with 1986 producing The medieval coin (8) from context 1843 is an Irish the greatest weight (17.963 kg). All of this material silver farthingofEdward I(1272–1307), belonging to is of post-Roman date (mainly 11th–19th century) Withers’ type 1(2005, 47), and issued by the Water- exceptfor asinglesmall sherd of Romanfine grey ford mint (Spink 2003, 127,no. 6268). The inscrip- ware (2nd–4thcentury AD) which was residual in a tions are partly illegible,becausethe coin was medieval context. The pottery was generally very struckoff-centre, but they can be reconstructed as fragmentary but not, on the whole, excessivelyabra- ERA/NG/[LIE](E[dwardus] R[ex]Anglie) on the ded. Asmall number of complete or nearly-complete three sides of the triangle around the crowned bust medieval vessel profiles have survived, however, of Edward Ionthe obverse,and [C]IVI/TAS/ or are capable of reconstruction (generally the VATE/[RFOR]inthe four quarters of the cross on smaller more robust forms) as well as alarger num- the reverse.The coin weighs 0.33 g, which is within ber of post-medieval forms –although the latter are the normal range forcoins of this type (the official mostly late in date and well-known from other weightatissue was c 0.36g). Englishsites. MichaelDolley(1972, 11–14) argued that the Water- The following accountaims to summarise the ford mintwas operational during the reign of Edward main trends of pottery supplyand usage to the site Iintwo separate periods, from 1280/1to1282, and not only during its monastic phase (1281–1536), but

Table 3Jettons.

SF Number IdentificationDate Range Comments

88 French Dolphin Jetton c 1375–1415 From Viennne, Paris or Tournai.Tressure around dolphin on Obv is not atype collected by Mitchiner or Barnard 14 Nuremberg Rose and Orb type c 1550 Illiterate inscription? Possibly Hans Schultes (1553–1584)? 5English Jetton contemporary with 1302–7 Similar (but not identical) to Mitchiner type 183 Type 10 Stirling Jettons 28 Fragmentary worn jetton ?Unknown type (worn). Flowers visible on Obv are similar to Mitchiner type 577 ( c 1450?)

33 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford also during the less well-understood pre-monastic OXBF: Late Saxon-EarlyMedieval South-West Oxfordshire phase and the better documented post-monastic ware (flint- and sand-tempered), c 875–1250 (mainly c 1050–1250 at Oxford). (12 sh.). phase up to and including the construction of the MISC.1: Fine grey micaceous sandy ware with sparse flint, railway station in 1851. Although the range of ?Hampshire (similar to Winchester Fabric MBK), medieval and post-medieval pottery recovered from ? c 975–1200. (4 sh.). the site is fairly typical of Oxford sites generally, the OXAD: Andenne-type ware, c 1050–1225. Import, East Bel- increased frequency of certain vessel forms (such as gium. (1 sh.). OXAK: Developed Stamford ware, c 1150–1250. Lincolnshire. medieval ceramic bottles and jugs)islikely to be a (3 sh.). reflection of specific activities relating to lifestyle of OXAG: Late Saxon-MedievalAbingdon ware (sand-tempered), the site’s Cistercian inhabitants. In addition, the pot- c 1050–1400.Berkshire. (47 sh.). tery gives us insights into some aspects of medieval OXY: Late Saxon-Medieval Oxford ware (sand-tempered), c 1075–1300. Oxfordshire. (399 sh.). pottery production and supplyaswell as pottery OXBK: Medieval shelly coarseware, c 1100–1350. Including ‘consumption’.Averylimited number of pieces have Northamptonshire shelly ware (9 sh.). been illustrated, mainly becausethe types of pottery OXAQ: Early-Late Medieval East Wiltshire ware (flint and recoveredhave been treatedinsome detail in other algal limestone), c 1150–1425. (183 sh.). reports of Oxford sites (see below). OXAH: ?Nuneaton ware, c 1175–1250. Warwickshire.(1sh.). OXAS: ?Early Surrey Border ware, c 1175–1250. Surrey/ Hampshire.(17 sh.). OXAW: Early Brill/Boarstall ware, c 1175–1400. Buckin- Methodology ghamshire. (77 sh.). WORC: Worcester-type sandy glazed ware, c 1175–1400. Anumber of pottery specialists have worked on this Worcestershire. (1 sh). pottery assemblage in succession over the years. OXAM: Brill/Boarstall ware, c 1225–1625. Buckinghamshire. Most of the original spot-dating and cataloguing was (1038 sh.). undertaken by Catherine Underwood-Keevill and OXBG: Coarse Border ware, c 1350–1500. Surrey/Hamp- shire. (4 sh.). Lucy Whittingham, bothofwhom produced interim OXBX: Late medieval Brill/Boarstall ware, c 1450–1625. summaries which are freely drawn upon in the Buckinghamshire. (27 sh.). presentsummary. The original catalogueshave TUDG: Tudor Green ware, c 1375–1550 (mainly c 1450– subsequently beencomputerised and updatedwhere 1550). Surrey/Hampshire. (12 sh.). CIST: Cistercian-type ware, c 1475–1650. Midlands.(9sh.). necessary but notfundamentally changed.Although FREC: Frechen stoneware, c 1525–1750. Import, Germany. the original catalogues providesherd countsfor each (28 sh.). pottery fabric type per context, they vary in the OLIV: Spanish olive jars, c 1550–1750. Import, Spain. (1 sh.). methodofrecording weights. Sometimes weights BORDG: Surrey/Hampshire white Border ware, green-glazed, were recorded perfabric, in the caseofselected c 1550–1700. (69 sh.). BORDY: Surrey/Hampshire white Border ware, yellow- groupsorsites,orsometimesweightwas recorded as glazed, c 1550–1700. (22 sh.). acontext sub-total, so although an overallweight BORDB: Surrey/Hampshire white Border ware, brown- total canbeobtained foreach site, it is not possible to glazed, c 1650–1700. (2 sh.). compareindividual pottery types by using weight a REW: Post-medieval red earthenwares, c 1550–1900. Local, including Brill (Bucks.). (348 sh.). meansofquantification. For this reason the only MLTG: Montelupo maiolica, c 1575–1650. Import,Tuscany, methodofquantification usedinthis report is sherd Italy. Early type. (1 sh.). counts. The original catalogues also containcom- WEST: Westerwald stoneware, c 1590–1750. Import,Ger- ments on the range of vessel forms presentand in many. (3 sh.). some cases pencil sketches of rim profiles, particu- ENGTN: English tin-glazed earthenware, c 1575–1825. Lon- don, Bristol etc. (27 sh.) larly for the late Saxon/early medieval wares. All the TN: Tin-glazed earthenware, source uncertain, c 1300– catalogues and lists of spot-dates remain in archive 1900 (mainly c 1550–1825). (2 sh.). although some details and summaries are availablein CHIPO: Chinese porcelain, c 1600–1900þ (mainly c 1725– the interim site reports. 1900). Import, China. (3 sh.). BLREW: Black-glazed red earthenware (‘Blackware’), c 1600– 1800. Midlands etc. (8 sh.). Fabrics BRSL: Red earthenware slipware, Brill-type, c 1650–1800. Buckinghamshire. (14 sh.). Pottery fabrics were recorded using the system of REWSL: Red earthenware slipware, ‘Metropolitan’-type, codes developed for the OxfordshireCounty type c 1650–1800. Various sources. (18 sh.). EST: English stoneware, c 1675–1900. (60 sh.). series (Mellor 1984;1994) and extended in more NOTTS: Nottingham stoneware, c 1670–1900. Nottingham recent yearsbyCaroleWheeler(2004, unpublished). and Derbyshire. (12 sh.). The types and quantities occurring at Rewley Abbey MBBEW: Staffs-type mottled brown-glazed earthenware, are listed below in roughly chronological order. c 1675–1800. (8 sh.). STSLBEW: Staffordshire-type combed slipware, c 1675–1900 R10: Roman fine grey ware, 2ndCþ .(1sh.). (mainly 18thC.). (15 sh.). MISC: Miscellaneous or unidentified, c 400–1900. (7 sh.). WHSG: Staffordshire-type white salt-glazed stoneware, OXR: St Neot’s-type ware, c 850–1100 (mainly c 950–1075 c 1720–1780. (36 sh.). at Oxford). South-east Midlands.(6sh.). JACK: Jackfield-typeware, c 1740–1780, Staffordshire, OXAC: Early Medieval Oxford ware (‘Cotswold’-type calcar- Shropshire. (4 sh.). eous gravel-tempered), c 875–1250 (mainly c 1050– EPO: English porcelain: c 1745–1925þ .(13 sh.). 1225 in Oxford). Central and north-west Oxfordshire, CRM: Creamware, c 1750–1830. Staffordshire etc. (227 sh.). Gloucestershire. (80 sh.). PW: Pearlware, c 1780–1830. Staffordshire etc. (48 sh).

34 OA Occasional Paper No: 16

BLB: Black Basalt ware, c 1770–1900. Staffordshire etc. although these comprise pottery types with a (2 sh.). currency from the late Saxon period through to the CEW: Coloured earthenwares, c 1800–1925þ .Stafford- shire-type. (6 sh.). 12th or 13th century (OXR, OXAC, OXBF, MISC.1). WHEW: Mass-produced white earthenwares, c 1825–1925þ . These possible late Saxon types mostly comprise Staffordshire etc. (119 sh.). handmadecoarseware cookings pots or jars plus a few bowls. Thecommonest type here is early Chronological Overview medieval Oxford ware (OXAC) which, despite its name,was probably from several sources to the The main trends of pottery supply to the site down north and west of Oxford as well as the Cotswold throughthe centuries are considered here; this is of area to the west outside the county. This type necessity asimplistic and subjective overview rely- however was notcommon in Oxford until just a ing almostentirely on the pottery to tell, in asense, few decades before the Norman Conquest and so it is its own story rather than relying on the details of site likely that that most of the assemblage here is post- phasing. Pottery, as usual, is the main tool for dating Conquest in date. Another handmade ware (OXBF) archaeological contexts on the site and this, in turn, was probablyfrom asource in the Savernake Forest affects the dating assigned to individual site phases. in east Wiltshire. Afew sherds including acooking Scattered trench excavation as employed on this site pot rim occur in St Neots’s-typeware (OXR) –a rarely yields large or deeply stratified pottery seq- wheel-thrownlate Saxon shelly ware fromthe uencessuitable for statistical analysis, and no sys- Northamptonshire/Bedfordshire area. tematicdetailed recording of vessel forms has been Pottery types arriving on the site after the Norman carried out. Subjectively, however, it is possibleto Conquest of 1066 but (mostly) pre-dating the divide the assemblage into three broad ceramic foundation of Rewley Abbeyin1281 include all phases,based on pottery currency dates, and which those in the fabrics list from OXAD to WORC, roughly correspond with three major phases or mainly early medieval types with some fully-fledged periods in the historyofthe site (Roman and miscel- medieval types. Although coarseware cooking ves- laneouspottery excluded). These are: apre-abbey sels are stillpredominant the range of vessel types phase (28% of pottery);anabbey phase (35%), and a has widenedtoinclude vessels for carrying and post-abbeyorpost-Dissolution phase (37%). Some serving drinks, such as spouted pitchers –a pottery types herewill obviously overlap in date and forerunner of the jug –basically ajar-shaped vessel could arguably belong to one or otherphase. Many with ashort tubular spout attached to the shoulder, unknown factors,such as breakage rates, modes of and also true jugs with atallernarrower neck,a rubbishdisposal and residuality as well as occupa- pouringlip pulled from the rim,and ahandle. tion intensity will have an affect on how these These vessels, though mostly still handmade, mark figuresshould be interpreted. the appearance of glazing and the frequent use of various typesofapplied,paintedand incised Pre-Abbey or early medieval phase decoration. The main pottery industry represented in (?10th Century –1281) the wholeofthisphase is medieval Oxford ware (OXY, 48%ofphase) which is thought to have been This comprises all those fabrics in the list above from produced somewhere to the north-east of the city. OXR to WORC (total 840 sherds or 28% of site This pottery type flourished in the 12th century. assemblage). This phase may be slightly over- By c 1250 it represented only 15% of the pottery represented due to the inclusion of early medieval assemblage at the Dominican Priory in Oxford Brill/Boarstall ware (OXAW, c 1175–1400) although suggesting it was in declinebythisdate. While also it is likely that most of this is 13th-century in date used for cooking vessels,and some bowls, medieval and therefore mostly pre-abbey. Likewise the inclu- Oxford ware also occurs as distinctive yellow-glazed sion of early-late medieval east Wiltshire ware spouted pitchers with avarietyofsimple applied (OXAQ, c 1150–1425). It has been observed, how- and incised decoration appropriate forserving ves- ever, that OXAQ nearly always occurs in association sels usedatthe table. Somespouted pitchers also on this site with medieval Oxford ware (OXY, c 1075– have characteristic tripod feet and strap handles with 1300) and one researcher (LW) has therefore sug- braidedclay strips applied along the spine of the gested anew daterange of c 1075–1300for thisfabric handle–examples of all these features are present type. This may well be the case, although this in the sherds from Rewley Abbey. Some jugs occur suggestion requiresfurthertesting. This phase is in this fabrictoo, as doesarim from ashallow thought to include agricultural activity on the site unglazedcresset lamp with heavy internal sooting. prior to the construction of the abbey, though much Jugs are generallycommoner in early medieval of the pottery may well have arrivedinloads of soil Brill/Boarstall ware (OXAW) –the third most broughttothe site from elsewhere in town to common fabricinthis phase and which marks improve the quality of its marshy soils or to provide Oxford’s long connection with this prolific pottery- afirmer foundation for the projectedabbey. producing area of west Buckinghamshire (see There is no early Anglo-Saxon pottery from the OXAMbelow). site but there is almostcertainly some late Saxon Medieval east Wiltshire ware (OXAQ), also (like (10th/early 11th-century) pottery in the assemblage OXBF)from the Savernake Forest area, is the second

35 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford commonest pottery type in this phase and is a becauseitwas so easily available from the markets in common find from Oxford sitesofthis period. This Oxford. Brill/Boarstall ware is asmooth fine sandy flint- and limestone-temperedcoarseware is mainly ware ranging in colour from cream to light orange or presentascooking pots/jars and acouple of bowls buff. Vesselswerewell-pottedonapotter’s wheel (Fig. 20.1). Medieval Abingdon ware (OXAG), and most commonly glazed (Mellor1994, 111–140). asand-temperedtradition from Berkshire, has a The industry is best-known for its wide variety of moderatepresence on the site although 34 of the 47 jug forms which were often highly decorated with sherds recovered come from just one crushed 13th- designsinred and white applied strips and pads of centuryjug with characteristicdecoration in painted clay, usuallyinlinear geometric schemes but not white slip (Mellor1994, fig. 27.3). Other wares have a infrequently in exuberant floralorfoliageschemes lesser presence but these includesmall sherds sometimes accompanied by stamped designs(Plate from one or two wheel-thrown green-glazed jugs 10.4). Applied zoomorphic and anthropomorphic in Developed Stamford ware from Lincolnshire, decoration (Plate 10.5) also occasionally occur and ayellow-glazed jug possibly fromNuneaton in incised or combed decoration is common. Vessels Warwickshire,and asinglesherd from ayellow- were either clear glazed(yellow or orange) or green glazed vessel, almost certainly aspouted pitcher, glazed.With the exception of zoomorphic decora- from the Andenne kilns in east Belgium –ahigh tion, examples of nearly all these types of jug qualitytableware which is rarely foundonsitesas decoration are present at RewleyAbbey, though all far inland as Oxford. in afragmentary state. Aprobable13th-century jug rim from this phase Although detailed quantification of vessel types has arare incised potter’smarkonthe neck(Fig. was not routinely carried out on thisassemblage, it is 20.2, Plate 10.2). This in incompletebut includes an possible to form areasonably clear picture of the apparently abstract arrangement of horizontal and Brill/Boarstall ware assemblage from the comments verticallines forming abox with adiagonal line recorded in the existing catalogues. The frequency of within. These marks were made before the jug was each form type is briefly considered first, followed fired and were almostcertainly madebythe potter – by aconsideration of likely function.One vessel perhaps as abatch-mark for aparticular customer. category, Brill/Boarstall ware bottles, has been This sandycream-coloured fabric(OXAS) has yet to singled-out for more detailed treatment (see below). be accurately sourcedalthough it may be aSurrey/ Jugs are easily the predominant vessel form present Hampshire or perhaps evenanearly Brill product. A and –although their numberscannot be calculated sherd from ajug in Worcester-typesandy glazed with any accuracy –jug sherds occur in virtually ware (WORC, Fig. 20.3, Plate 10.3) is probably of every one of the 183 contexts containing Brill/ 13th- or early 14th-centurydateand is the first Boarstall ware (OXAM). Remarkably only four example of this ware to have been recognised from instances of cooking pot/jarforms were noted, from Oxford although oneortwo other possible sherds as many contexts –even fewerthan bowls or dishes have been identified from Witneyafew miles west (8 vessels from 8contexts). There is also asingle of the city. sherd from the stem of adouble-shelled oil lamp. Bottles, usually afairly rare form in thisware, are surprisingly common here –occurring in 32 contexts The abbey phase (1281–1536) and representingaminimum of 17 individual vessels This phase comprises amuchnarrower range of (see below). Fine-walled green-glazedcups copying pottery fabrics (OXAM to TUDG)but still a TudorGreen ware, around4examples, are known considerable volume of pottery (1081 sherds,35% from 2contexts and there is asingle sherd from a of site assemblage). Some of the ‘early medieval’ dripping pan. wares from the previous phase howevercontinued to be supplied to the site during the earlier part of Brill/Boarstall ware bottles: acase study this phase. Themain suppliers of pottery to Oxford, and most of Oxfordshire, during thisphase were the The minimum of 17 individual bottles from Rewley Brill/Boarstall kilns of west Buckinghamshire,only Abbeyrepresents the largestnumber of these vessels twelve milesfrom the city. Brill/Boarstall ware recoveredfrom asingle site in Oxford, as pointed out (OXAM) comprises some 96% of pottery fabrics by Maureen Mellor some yearsago (1994, 118). safely attributable to this phase and whilethe true Because of this remarkableconcentration and the percentage is probablyslightlylowerthan this uniquecontribution that Rewley Abbey makes owing to the overlap in dateofearlier types towards the study of this particular vessel form, a (including BrillOXAW, some OXY, OXAGand separate more detailed catalogue of this type was OXAQ), it is still, overwhelmingly, the main pottery compiled and their contexts examined more closely type in use during the lifetime of the abbey. This to see if any patterns occur in their distribution could implythat the abbey –ashas been suggested acrossthe site. Their presence here raisesanumber for some medieval colleges in the city –ordered its of interesting questions. Foremost of these are: what pottery supplies directly from the potters at Brill. It is were they usedfor,and is there aspecial connection rather more likely howeverthat Brill ware is the between this vessel form and monastic sites?We commonest type of pottery from the abbey simply mightnot be able to provide definite answers to all

36 OA Occasional Paper No: 16 these questionsatthe moment but by considering Brill/Boarstall ware bottles are considered to have the evidence availableweshould come closer to the amainly later 13th-century to 14th-century dating truth. although production continued into the 15th century Brill/Boarstall bottles are slender,slightly balus- (Maureen Mellor, pers. comm.). However, agroup of ter-shaped vessels with aflat base and aplain rim. at leastseven bottles, including several complete and They show no evidence of handles or otherapplied near-complete examples, are known from the Dis- features and are plain except for athinwash of solution backfill of alatrinepit at Eynsham Abbey, glaze on the shoulder (Fig. 20.6–7; Plate 10.6, 8; west Oxfordshire, there assignedtoPhase 4a dated Mellor 1994, fig. 55.1–9, 13–16). Suggestions as to 1538–1660(Blinkhorn 2003, fig. 7.18.137–43). This their use include containers for oils and saucesfor seems surprisinglylate but all the othervessels culinary use,orperhaps measures, although there illustrated from the latrine pit, including an almost is little definite evidence to support these sugges- complete 15th-century Spanish lustrewarejar (ibid., tions. At least165 sherds from the site were fig. 7.19.156) appear to be late medieval forms. A identified as coming from bottles.Bottles thus pair of bottles were among adeposit of vessels comprise 15% of all sherds belonging the monastic buried c 1290 in the foundation of the walls of Trinity phase (or 16% of all Brill/BoarstallOXAM sherds). College, Oxford (Mellor 1997, fig. 83). The1281–1536 These range fromthree almost complete examples lifespan of Rewley Abbeyprovides only broad which lack only their rims, to several half-complete confirmation for their dating.Unfortunately, in examples (or halfprofiles), rightdown to small almost all of the monasticcontexts, Brill/Boarstall thick-walled body sherds that seemquite probably ware is the only datingevidence availabledue to its from bottles. The more robust bases generally predominance on the site.This often means that survive better than the delicate rims(17 bases contexts containing bottle sherds cannot usually be representing the samenumber of vessels as dated muchcloser than acouple of centuries and opposedto12rim sherds representing just 9 broad context dates of the 13th–14th or 14th–15th vessels). Rim diameters from the site are in the centuryare common. 30–50 mm range,but 38–40 mm (5 vessels) was Examination of their contexts and locations does clearly the preferredsize.Mostrims are of plain reveal an interesting concentration of bottle sherds tapering slightly evertedform but oneortwo from the areaofthe church but also, curiously, from examples are thicker with aslightexternal furrow the east side of the cloister. Thereishowever athin or hollow. Basesrange in diameter from42–62 mm scatterofbottle sherds from acrossalmost the whole (two larger bases of 80 mm and 160 mm can site although mainly fromtrencheswithin the probablybediscounted as jugsand are excluded precinct walls only. from these figures). Thereisaclear size preference The highest concentrationofbottle sherds comes for bases of 47–53 mm (12 vessels; mainly 50– from Trench 1986/5located within the east side of 52 mm). Bases are flat but with asplayed and the cloister walk. Here asequence of threeextensive pedestal-like foot, often with aslightridge or layers interpreted as amiddenordump (500, 501, moulding abovethe outeredge which sometimes 502) produced 86 bottle sherds –more than half the createsaslight pad reminiscent of some metal total number recovered from the whole site. Base vessels of thisperiod. Most bases have asingle fragments suggest aminimum of 6vessels here, central dimple on their underside causedbythe although there are probablymore than thisnumber. potter’sthumb, and the lower wall of several Pottery finds were concentrated at the northern end examples has an untidy scar or nick possibly of the trench. Associated jug sherds in Brill/ causedbythe potter’s thumbnail when carelessly Boarstall ware includelate 13th- to 14th-century removing the vessel from the wheel. types but also ajug with areeded or grooved Size analysis points, perhaps, to the production of shoulder band andasherd from aBrill/Boarstall two sizesofbottle –asmaller commoner type (12 TudorGreen-style vessel, boththe latter probably vessels:rims 38–40 mm, with bases 47–53 mm), and datingtothe late 14th to 15th centurysuggesting alarger rarer type (3 vessels: rims 43–50 mm, with the context couldbeofmixed date. The bottle bases 60–62 mm), with the occasional production of sherds recoveredinclude two of the most complete bottles outside this vessel range. Examination of the bottles from the site –one of which is complete but capacities of complete examples in museum collec- has lost its rim (Fig. 20.6; Plate 10.6). Curiously, tions mightshow whether these variations relate to although many of the sherds are obviously fresh,a known medieval units of capacity. few bottle sherds, including threeseparate bases in Most bottles have abroad band of green copper- asofter unglazed fabric, show considerable abra- fleckedorspeckled glaze aroundthe shoulder area sion or weathering suggesting they were already only. Only one example has abright yellow (clear) old when deposited (or redeposited) which again glaze. Oneortwo have athinclear olive-green points to material of mixed date. Bottle sherds from glaze –inone case (perhaps accidental) extending the church area (see below) also include afew all over the lowerwalls and underside of the base. softer weathered sherds suggesting that these The interiors are alwaysunglazedand show no deposits too contain some residual earlier material visible evidence of residues which might help and thus perhaps were depositedsome decades, at explain their function. least, after the 1281 foundation.

37 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

The Trench 1986/5 ‘midden’ layers also contain lamps to commemorate the dead. The lamps, which large pieces of charcoal, oysters shell and fragments are sometimes shown in medieval illuminations, of stone roofingshingles. The excavators have were probably of metal and usually suspended suggested that these layers represent domestic abovethe image of afavourite saint. The same rubbishdumped frombuildings to the east of the documents sometimes mention the purchase of cloister. Whateverthe source it seems alittle odd to ‘pottles’ofoil whichmay be areferencetoceramic dump domestic rubbish in the middle of acloister bottles or containers, although it was also amedieval walkway –unless perhaps it was usedaslevelling unit of fluid measurement –equivalent to halfa prior to aresurfacing of the cloister.Afew bottle gallonand thus too large to relate to the bottles here. sherds also occur in verysimilar layers in Trench Many medieval monasteries had their own pharma- 1986/4several metresaway to the north-east in the cies or dispensaries where medicines and herbal garden area. preparations were kept and sometimes prepared. The other main concentration of bottle sherds is These sometimesincluded stills for the production of from the group of trenches dug across the central alcohol –aprecious commodity that may, perhaps, part of the church (Trenches, 1986/9, 1986/10, 1986/ have beenstored or dispensedinsmall containers 11a and 1994/18). These collectively produced 52 such as ceramic bottles –but this suggestion does not bottle sherds from aminimum of eight vessels (base explain why they might have been needed in a counts) –nearly athird of all bottle sherds from the church. Of course,itisquite possible that Brill/ site (but more than half the minimum vessel count). Boarstall bottles were multi-purpose containers – These,onthe whole, are more fragmentary and less though presumably forsmall amounts of afairly concentrated than thoseinthe eastern cloister precious liquid commodity –whatever that may (Trench 1986/5), although they do include an almost have been. Until areasonable sample of bottle sherds complete (reconstructable) bottle from Trench 1986/ has been scientifically analysed to determine what 11a (presumably residual in apost-Dissolution they mighthave contained everythingelse, for the robbing layerover the north wall of the church).A moment, remains speculation. few bottle sherds come from contexts within the church (3 sherds from arectangular pit, context 908, to the south of apier in Trench 1986/9) and afew Other vessel forms andtheirfunction in others come from post-Dissolution wall-robbing the monastic phase features, but most (31 sherds) are from just outside the south wallofthe church at the southern end of The fact that only one Brill/Boarstall oil lamp was Trench 1994/18. The bottle sherds here were from foundinthe excavations (outside the ancillary asequence of quite deeply-stratified dump layers building, Trench1986/1) is slightlyproblematical almost certainly laid down during the lifetimeofthe given the obvious need forlighting in the church and abbey. monastic buildings and the suggestion above that Two bottles, including acomplete (reconstruct- ceramic bottles may have been use as lamp-fillers. able) example, lacking only its rim, come from The ceramic lamps, which have asimilar datingto alayerabove disturbed alluvium in Trench 1993/3 the bottles, were alittle like candlesticks in form with several metres to the south-east of the church. This adrip-tray or saucer(‘shell’) and astem supporting a context also produced a13th- to 14th-century dec- small shallowdish which contained the oil and a oratedjug sherd but also aprobablelate 14th- to wick (Mellor 1994, fig. 54.18–22). They were de- 15th-century sherd copying TudorGreen ware. signedtobeplaced on aflat surface. Obitorshrine Unfortunately one cannot deducefrom any of lamps,assuggested above, were probablysus- these contexts exactlywhere in the abbey the bottles pendedfrom the ceiling and were of metal or even would originally have been used. However, the glass. Wax candles, stone cressetlamps and rush concentration of bottle sherds in and around the lamps may have been used to illuminate other parts central partofthe church and the eastern side of of the church and monasticbuildings, although the cloister to the north of thisisundeniable and candles may have beenreserved for the church. unlikely to be pure coincidence. Some college buildings in Oxford, and sitesalong the The possible use of ceramic bottles within the main roads, have produced fairly large numbersof church is slightlysurprising given their supposed Brill/Boarstall oil lamps –atleast seven were found culinary function. An earlier suggestion, now out of during recent excavations at Merton College – favour, is that ceramic bottles may have servedas suggesting alikely connection with students.The altar cruets for holy wine or oil. However, they seem low number of lamps from Rewley Abbey is there- rather crude for this purpose and it is more likely fore alittlepuzzling given that the abbey had a that metal cruets were preferred.Another and studium and aresident student populationofits own. perhaps more plausible suggestion, giventheir Some form of lighting, other perhaps than oil lamps, context, is that they couldhave been used as would have beenneeded, or perhaps the areaof containers for refilling oil lamps which may have the studium has not yet beendiscovered (but see illuminated parts of the church (see below). Medie- Roe below for apossible stone candle holder). val church records and wills sometimes mention The removal of rubbish from the site during the donations of money for the oil usedtoburn in obit medieval period, as well as the consequences of the

38 OA Occasional Paper No: 16

Dissolution, is likely howevertohave left us with a the scarcityofthis form on the site, one might very incompleteceramic and artefactual record. supposethat metal equivalents were preferred. The very few bowls or dishesidentified in Brill/ Definitepurpose-made drinking vessels are Boarstall ware (eight examples) are of small size uncommon in the monastic phase. Averysmall (ibid., fig. 53.5) and are likely to have beenused for number of cups and possiblydrinking jugs are table condimentsrather than serving vessels. Cera- represented by 12 sherds of TudorGreen ware – mic table dishes or plates as we now knowthemare ahigh quality green-glazedwhiteware from the largelyapost-medieval invention. Medieval people Surrey/Hampshire border produced between the would most probablyhave eaten from wooden late 14th and early 16th centuries.Tothese must be trenchers or bowls. The even smaller number of jars added three or four copies made in Brill/Boarstall in Brill/Boarstall ware (four examples) presumably ware. One or two black glazedCistercian ware cups served as cooking vessels or storage jars. To the may also belong to this phase although these have small number of jarsand bowls in Brill/Boarstall been assigned to the following phase as most ware, however, we must also add an unknown examples look later. number in medieval east Wiltshire ware (OXAQ Although some of the more decorative Brill/ c 1150–1425) which overlap in datewith the mon- Boarstall ware jugs in the monastic phase canbe astic phase although most of these probably date regarded as high qualitytablewares, there is little to the pre-monastic phase. Even so, the number else in the monastic assemblage that speaks of of bowls and especially jars from the site is luxury.All the pottery is functional and utilitarian. undoubtedly low. This may be areflection of the Thereare no decorative imported wares of the monastic or institutional nature of the site –with period such as Spanish lustrewares or highly dec- large-scalecatering taking place in acentralised oratedFrench waresalthough these are very rare on kitchen and bulk cooking in just afew large metal Oxford sites generally and probably reflect the city’s cauldrons rather than many smaller, more fragile, inland location. High class tablewares, of course, ceramic pots. Increasing competition from metal could have been of metal or glassrather than pottery vessels,which becamecheaper and commoner as the but these materials were usually recycled and tend to medieval period wore on,had agradual effect on the be rare on monastic sites, particular those of the number of ceramic cooking vessels in general Cistercian order, whose monksfollowed alife of self- circulation. sufficiency and severe austerity. Conversely, the relatively high number of jugs on the site is also likely to be areflection of the Post-Dissolution phase ( c 1536–1900) institutional nature of the site. Jugs were usedfor the fetching and serving of drink and smaller jugs, This, by asmall margin, comprises the largest particularly in the late medieval period, were proportion of site pottery –all those fabrics from increasingly used as drinking vessels –primarily CIST to WHEW (1115 sherds, or 37% of site perhaps for beer or, rather, small beer, which was assemblage). It almostcertainly comprises the bulk regarded as asafer optionthan untreated water. This of pottery weightfromthe site. Thepost-medieval high ratio of jugs to cooking vessels and othervessel assemblage is not particularlyremarkable or forms has been observed on anumber of monastic well-preserved and is typical of numerous post- sites in Englandand is probably areflectionofthe medieval assemblages from Oxford and southern fact that cooking, in monasteries, was aspecialised Englandgenerally. Consequently it will notbe activitycarried out by the few, whereas drinking considered in much detail (see fabrics list). After the was carried out by the many (Cotter 2001). Occa- Dissolution in 1536 it is known that muchofthe sionally jugs could be used or re-usedfor less abbey grounds was given over to gardens while obvious purposes. Ajug base found in an early pit muchofthe stonework was carted away. There insidethe church (Trench 1986/9; Plate 10.9), is seems to be very little in the way of diagnostic16th- heavilystained internally with adarkred-brown centurywares which suggests, perhaps, aperiod of substancewhich may be red lead (an oxide of lead), near-abandonment. Some late, plain, Brill/Boarstall acommon medieval pigment or paint. Perhaps the products (OXAM, OXBX) may have continued to jug was usedfor storing paint or was originally a reach the site as late as c 1600 but there is notmuch broken jug base re-used as apaint pot or palette.A evidence for this. Similarly, avery small number of small but very similar sherd from aseparate jug, also thin-walled black-glazed Cistercian-typeware drink- with red pigment, was foundinapost-Dissolution ing vessels from the Midlands may have reached the layer outside andtothe south of the church (Trench site during the 16th century –some perhaps even 1994/18). before the Dissolution –although this generic term Dripping pans, of which asingle fragment was includes some vessels of later date. Some local red identified, are long oval or sub-rectangular dishes earthenwares (REW)and Surrey/Hampshire white- often with apouring lip at each end and apair of wares (BORD) were also in circulation by the mid side handles. These were used in the kitchen to 16th century but there is little to distinguishthese collect fat from spit-roasts. The basal/wall fragment rather plain vessel forms from their successors in the identified from the site is characteristically thick following two centuries. Asinglefairly small sherd of walled and heavily sootedexternally. Again, given Montelupo maiolica, aluxury tin-glazed tableware

39 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford importedfrom Tuscany is, however, most likely to The18th-century assemblage here, as elsewhere, date to the late 16th or early 17th century. The sherd witnessesthe gradualriseofthe Staffordshireand is from just below the rim area of abowl-likefruit Midlands pottery industries–slipwaresand stonewares stand or tazza with polychrome fruit or floral at first –and mass-producedcream andwhite table- decoration insideand out (Plate 10.10). Montelupo waresbythe late 18th andearly 19th centuries. Thehigh maiolica is rare on British sites generallyand even number of Creamwaredishand plate sherds fromthe rarer from sitesasfar inland as Oxford,although a site (mainly c 1780–1830)isnoteworthy.Someofthe couple of Ligurianmaiolicadishes of similar date are latest Staffordshire-type products (WHEW)andEnglish known from the city (Mellor1997;see below also). stonewaresofthe 19th centurymay have been In the main, the post-medieval assemblage ap- associatedwiththe RewleyRoadstation constructed pears to be predominantly of 17th- to 19th-century on the site in 1851,although thereisnothing particularly date. The17th- to 18th-century assemblage is dom- noteworthy or obviouslyconnected to therailwayin inated by common red earthenwares (REW), mostly thislatestassemblage.Several cylindricalwhiteware glazed.These are the general crockery of the period, jars (perhaps fromthe stationcafeteria?) have inscrip- mostly in the form of large bowls or dishes and tionsshowing theycontained ‘Frank Cooper’sOxford storagejars. An unusual item is an unglazed ‘chicken Seville Marmalade ... by RoyalAppointmenttoHis feeder’(Plate 10.11) found in amid 18th-century Majesty King George V’.The Oxford marmalade factory context along with otherdomestic pottery and clay layonly ashort distance southofthe station–although pipes.These flattish dish-like vessels had two or the jars themselvesweremadetoorder by Malling of three added concentric rings of clay. Similar dishes Newcastle. are known elsewhere from 19th-centuryfarmyard contexts, so the traditional identification as chicken feedersisprobably correct and implies that the 18th- Pottery illustration catalogue (Fig. 20; Plate 10) centuryinhabitants of the abbey site kept poultry. Some or perhaps most of this redware was made at 1. Early-late medieval east Wiltshire ware (OXAQ). Brill (see also medieval wares)and perhaps at other Bowl. 12–13C. OXSTRA 94 (1739).Fig. 20 fairly local potteries too. Likewise the more decora- 2. ?Early Surrey/Hampshire border ware (OXAS) tive slipware dishes in red earthenware with their or unidentified fine cream sandy ware from same trailed decoration in white slip (Plate 10.12; BRSL, general source. Jug rim with incised (pre-firing) REWSL). The Brill slipwares, here almost entirely potter’s mark on neck. Large rounded off-white 18th-century,occasionally have green glaze high- finesandstone inclusion (8 mm across) protrud- lights giving apolychrome effect. ing fromsurface. Large patch of clear yellow The Germanstonewares present, all common glaze on shoulder to rightofmark. 13C? imports of the period, includeafew plain Frechen OXSTRA 94 (1622).Fig. 20; Plate 10 drinking jugs of the late 16th to early 17th centuries, 3. Worcester-type sandy glazed ware (WORC). and 17th-centurytype ‘bellarmines’ or Bartmann Jug sherd with stamped decoration. Identified jugs with applied bearded masks on the neck. One by Alan Vince.13C-early 14C. OXSTRA 94 othermoreunusual tin-glazed vessel (TN) of uncer- (1952).Trench 1994/19. Probably 13C dump tain continental or English origindeserves closer layercut by ‘barn’wall. Fig. 20; Plate 10 consideration. This is represented by arim sherd 4. Two decorated Brill/Boarstall ware (OXAM) jug from asmallish jar (Fig. 20.13; Plate 10.13), possibly a sherds. 13–14C. Max. length 68 mm. Left OXRA wet drug or syrup jar,with an off-white or 86 (910); rightOXRA86810/A/2. Plate 10 discoloured palegrey tin glaze with external blue- 5. Brill/Boarstall ware (OXAM)jug rim with painted floraldecoration outlined in placeswith applied face pad. 13C. Height of sherd 43 mm. black –adesign ultimately derived fromChinese Diameter 160 mm (160–180mm?). Slight possi- porcelain and probablydatingtothe 17th or early bility an unusual bowl becauseoflarge diameter 18th century. This may be Portuguese, or possibly and internal and external green glaze?OXRR Ligurian (north-westItaly), but withoutscientific 93 (228). Linear feature cutting retaining analysis of the clay it is best described, forthe wall(226) in Trench 1993/2, courtyard/garden. moment, as apossible import. It may have come Plate 10 from an apothecary’s shop, or have been apurely 6. Brill/Boarstall ware (OXAM). Bottle. Complete decorative piece from aprivatehouse. Asherd from except forrim.Zone of speckled copper green an early 17th-centuryEnglish tin-glazed drug jar glaze aroundshoulder area. 13–14C. OXRA 86 with polychrome geometric decoration was also re- (501). Fig. 20; Plate 10 covered as well as plain white glazeddrug jars and 7. Brill/Boarstall ware (OXAM). Bottle rim. Over- ointment jarsofthe 17th–18th centuries.Asingle fired. Unglazedexcept for clear specks. 13–14C. sherd from aSpanish olive jar (OLIV), arare type OXSTRA 94 (1829).Fig. 20 in Oxford, was foundinamid 18th-century context. 8. Agroup of four Brill/Boarstall ware (OXAM) These coarseware containers, shaped like Roman bottles. 13–14C.Height of tallest to left 165 mm. amphorae, were acommonproductofSeville and Various contexts. Plate 10 were usedtoexport luxury Mediterranean foodstuffs 9. Brill/Boarstall ware (OXAM) jug base (flat) such as olives, oliveoil, honeyand dried fruit. containing tracesofred pigment. 13–14C. Height

40 OA Occasional Paper No: 16

Figure 20 Pottery.

45 mm. Base diameter130 mm. OXRA 86 (904). 18C context. OXRA 81 1/6/1 &1/11. Second of Plate 10 tip layers N. of wall10.Plate 10 10. ItalianMontelupomaiolica(MLTG). Dish sherd 12. Fine red earthenware with trailedwhite slip (tazza). External view showing traces of poly- decoration (REWSL). Possibly ?Brill slipware. chrome decoration. Late 16-early 17C. Max. Internal detail of dish sherd with curvilinear slip length 42 mm. OXRA 86 226/A/1. Plate 10 decoration with dot fillers. Late 17–18C.Rim 11. Post-medieval red earthenware (REW). Two diameter 420 mm. OXSTRA 94 (2821).Plate 10 fragments from a‘chickenfeeder’. Possibly Brill 13. Portuguese or ?Ligurian maiolicadrug jar (TN). (Bucks.)? Diameter of outer ring c. 170mm. Mid 17–18C. OXSTRA 94 (310). Fig. 20; Plate 10

41 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

2 3 4

5

8 0250 mm

1:4

6

9 10 11

13

0100 mm

12 1:2

Plate 10 Pottery.

42 OA Occasional Paper No: 16

Clay TobaccoPipes would be verysurprising if pipemakers did not David Higgins establish themselves in the areaatanearly period. The mid 17th-centuryforms closely follow London The six phases of work on this site produced atotal fashions but the presumption must be that most of of 623 fragments of clay tobacco pipe from81 the pipesinuse by this datewere being produced different contexts, with oneunstratified group. A locally. Both spur and heel forms were commonly summary of the pipes recoveredfrom the site as used but, if anything, it is the spur forms that appear awholeisprovided in Table 4. This shows the to have been slightlymore common. Thepipes are numbers of bowl(B), stem (S) and mouthpiece (M), generally well made and neatly finished, and pro- as well as the total number of pipe fragmentsthat duced using fine almostinclusion free clay. There this represents, from each phase of work. The are anumber of individual pipes datingfrom this markedpipes are then listed, as well as the number period, but the best group was recoveredfrom of decorated pieces, before the figure numbers of beneath acollapsedwall (OXRA 81 [0011];Fig. 21.7– the illustrated pieces and any general comments. A 13). This does notappear to be asingle-phase group fuller record of the pipes by phase of work and of pipes, but ratheracollection of material that built context, which has beenused to informthe dating up over aperiod of some 30–40 years. and interpretation of the archaeological features Towards the end of the 17th century the pipe and deposits on the site, can be found in the context assemblages exhibit anumber of changes, in parti- summary in the project archive. cular with regard to form and fabric. The London forms that had dominated production until around 1680 were replaced by more distinctive local styles. Pipes of the 17th and 18th centuries In particular, large and rather sharply curved spur Some of the early pipes are very neatly made and forms appeared, which seemtodraw on West Coun- finished, for example OXRA 86 [0006/C], which dates try styles for their inspiration as muchasLondon. from c 1620–1650(Fig. 21.1). This piece is fullymilled This style emergedfromthe middle of the century and finely burnished and the pipe is of agood form, onwards and was noted by Oswald as his ‘Type B’ neatly finished.These early bowls are rarely marked in the St Ebbe’s report (Oswald 1984, fig. 51). An and madeofafine, inclusion free fabric so that it is early example of this style, datingfrom c 1660–90, impossible to tell whether they were made in London was recovered from OXRR93[0204] (Fig. 21.14). or produced locally in aLondon style. Given the size The author considers that this form developed and significance of Oxford as an urbancentre,it into arather ‘top-heavy’ and heavily built spur bowl

Table 4Summary of the pipes recovered from each phase of work.

Site BSMTot Marks Dec General Comments

OXRA 81 34 63 4101 All C17th and C18th material. Two good mid-C17th groups account for 78% of the assemblage OXRA 86 35 120 1156 ED 2or3 Almost all C17th and C18th material with just afew C19th BEASTEN, fragments. Includes some good early C18th material AM, WT (including 2stamped stem marks),adecoratedDutch stem and an early C19th mouldedmark OXRR 93 13 40 154**Mainly rather small groups with some battered looking fragments. Most material dates from C17th and C18th but with some C19th groups. Context 204 includes two interesting transitional forms. The only mark is part of a C19th moulded spur mark (stars) OXSTRA 94 48 243 6297 RICH/ 1Most of the groups fairly small and of mixed C17th to ARD/ C19th dates. The majority of the finds (207 pieces) come CUTTS; ID from apit group of c1690–1710 [2821], which provides a well dated and important reference point for the study of Oxford pipes OXSTRA 96 12 12 BH · 1; 9Single group of C19th pipes from alevelling deposit ** · 2 including aspur mark of Benjamin Huggins,working 1841–75. May well be from the construction of the railway terminus in about 1850 OXSAID 01 33 Single context group [302] containingthree pieces of C17th or early C18th stem

TOTALS 142 469 12 623 912or13

43 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford with an even more pronounced curve to its profile. of the‘barrel shaped’17th-centurytradition andwhich This type has beenencountered at other sites in wouldcertainly be expected to be presentinsome Oxford, forexample the NewChemistry Labora- numbersuntil at leastthe 1680s. Furthermore,there is tories (Higgins2005, fig. 11.2) as well as at thisone. onestemmarkofRichard CuttsfromEastWoodhay in An example was recovered from OXRA 81 [0011], Hampshire(Fig. 22.22),who wasmarried in 1693 where it seems out of placeinsize and style amongst (Cannon1991, 22)and so wouldhavebeenunlikelyto the predominantly mid 17th-century forms that have been making pipesofhis ownmuchbefore1690. made up the rest of the group, suggesting that it is At theother end, thereare none of theveryupright intrusive. Two examples were presentinapit group formsthatappearedfromearly in the18thcentury and of c 1690–1710(OXSTRA 94 [2821];Fig. 22.20) and a none of thepipes hascut rims,another characteristic similar example was also recovered from OXRA86 that appeared at aboutthistime. [0010/A/2] (Fig. 22.31), adeposit dominated by The close datingafforded by this group allows early 18th-century forms. it to be used to characterise the types of pipe that The occurrence of three examplesofthis large spur were being made and used withinOxford at this form in good late 17th- and early 18th-century date, and it provides an importantbenchmark for groupsclearlysuggeststhat it is acontemporary the region. In terms of style, there are 29 identifiable style and one that can be seen as alocal competitor to forms represented in the group. Three of these are the West Country spurforms of the same date. The spur types (10%)and 26 are heel types (90%). This large version of this particular form seems to fit dominance of heel types is in markedcontrast muchbetter in these later contexts, where its size is with the mid 17th-century groups from this site, more in keeping and it can be seenasanestablished where spur types appear to have beenslightly design being produced alongside newervarieties. more common than heel types. Of the three spur Being an established or ‘old fashioned’ variety, forms represented, two are of the heavy local style it was produced with abottered and milled rim, discussed above (Fig. 22.20) and oneisofarather techniques that had beendropped with the newer more elegant style with afiner spur(Fig. 22.21). styles (cf. Higgins 1987, 257, Types 5c and 5f). It is The two local forms are so similar that theymay well suggested that the large Oxfordshire spur form was have been produced in the same mould and both introduced around 1680 and that some examples have one quarter of the rim facing the smoker milled, may have continued in production until around1720 the only pieces in this group to be so treated. The or 1730 alongsidethe newerforms. This is arather finer form is of astyle characteristic of Wiltshire and later and longerrange than that suggested for an Hampshire and this piece maywell have been example from St Ebbe’s (Oswald 1984, fig. 52.7) and importedfrom outside of Oxford (but in the Thames it also has implications for the dating of the RP stem Valley area as it appears to be made of alocal fabric). mark that is associatedwith it. It is suggested that It may well have had astem stamp,likethe incuse the use of large unbordered incuse initial stem marks RichardCutts example found in this group (Fig. (Oswald 1984, figs 52.7 and 52.10), some of which 22.22), which would have been on asimilar bowl were clearly associated with this bowl form, dates to form. The Oxfordshire makersdonot appear to have c 1680–1710. An incuse mark of thistype, but on a used this style of full name incuse mark on their heel form of c 1690–1720, was recovered from these pipes and they rarely seemtohave used any marks excavations (Fig. 21.16). at all until the nineteenth century.Noneofthe other heel or spur pipes from thisgroup appears to have been marked or decorated. Apit group of 1690–1720 and changing pipe styles There are 22 heel bowls where the bowl form can At theend of the17thcentury thebarrel-shaped spur be determined. About one third of these (seven andheelforms that haddominated pipe styles forthe examples) have arather straight-sided and forward- previous centurysuddenly startedtofallfromfashion leaning form and at leastfive different types are andthe pipemakers experimented with new‘transi- represented (Fig. 22.23–27). These tendtohave ra- tional’styles. Avariety of newforms were introduced ther large heels, cut close to the stem and often not around 1680–1710, from whichafairly upright very well differentiated from the bowl. In one ins- cylindricalbowlformemerged anditwas this form tance, afragmentary example has quite amarked that went on to dominate themarkets formostofthe curve at the front of the heel, which is also slightly 18th century. Apit fill(OXSTRA 94 [2821])associated flared to each side (Fig. 22.27). All of these types are with thedemolitionofthe barn to thewestofthe abbey represented by singleexamples except for one type moat produced 207pipefragments (40bowl, 162stem (Fig. 22.24), of which there are three examples. The and5mouthpieces), allofwhich appear to have been remaining two-thirds of the heel bowls (15 exam- depositedwithinashortspace of time.The bowl forms ples) have arather more upright and cylindrical presentare alloftypes that canbebroadly datedto bowl form with alongerand more sharply dif- around 1680–1720, butadate of c 1690–1710seems ferentiated heel (Fig. 22.28–30). This is the forerun- much more likely forthisgroup,withdepositionof ner of the style that went on to dominatepipe around 1700 seemingmostprobable. This is because, assemblages in this area for most of the 18th century. at theearlier end, thereare none of theforward Threebasic types are represented; those with large, leaningbut rather curved formsthatrepresent theend medium and small heels. Those with large heels

44 OA Occasional Paper No: 16 tend to have slightly taller and broader bowls and Early 18th-century groups of c 1710–1730 those with the small heelssmaller ones. These differences may be relatedtostem length,which in Around 1700, arather upright heel bowl with a turn was related to the cost of the pipe (longerstems cylindrical body was introduced that went on to being more expensive). Thereare six examplesof dominatelocal production for the next 70–80 years. bowls with large heels (Fig. 22.28), five examples This form is directly comparable to the London Type with medium heels (Fig. 22.29) and four with small 25 form (Atkinson and Oswald 1969), but the Oxford heels (Fig. 22.30). versiontends to be alittle shorterand more heavily Every singlebowl fragment is burnished, show- modelled aroundthe heel area. TheOxford heels ing that this was an absolutely standard finish in are rarely marked and the bowl is often set at asharp, Oxford around 1700. Theburnish is usually of upright angle to the stem. Furthermore, the bowls average to good quality, although often it is quite are often burnished and, in particular, theyare often lightly applied.Thereare afew finelyburnished made of adistinctive fine sandy fabric, quite dif- pieces.Almost all the stems are burnished too, ferenttothat used in London. Another characteristic although there are one or two that do not seemto of these pipes is that they often have an unusually 00 be, suggesting that notall of the bowls that were small stem bore for the period, often around4/64 . burnishedhad burnished stems as well (or that the This small boreseems to be particularly associated burnishing stopped part-way along the stem). As with the new bowl forms and late 17th-century types, notedabove, all the surviving rim fragmentshave some of which were probably still being produced been bottered, that is, shaped and smoothed by at the same time, invariably have much larger stem twistingafinishing tool over the rim, and most of bores. The large Oxfordshire spurtypesare agood the rims have been internally knife trimmedas exampleofthis. Individually many of the differences well. Only two rimshave been milled, both on the from London pipes are slightbut taken together,they locally produced spurbowls, where it occurs on one define adistinctive local style of pipe that was being quarter facing the smoker. The use of milling seems produced in the Oxford area. to be specifically associated with this form where The early 18th-centuryforms produced in Oxford it was almost certainly seen as part of the ‘style’ of are well represented by OXRA 86 [0010] and [0010/ the pipe. A/2], two contexts associated with the fill of a The stem bores are generallyquite large, around steeping vat culvert in abrewhouse. The finds from 6/6400 or 7/6400 with only oneortwo pieces smaller these two contexts (20 bowl and 47 stem fragments) than this. This is an important characteristic to note, form avery homogeneous looking group, but one and another one that placesthis group at about 1700, that appears to be slightlylater than the pit fill since by the 1710s or 1720s the Oxford makers (OXSTRA 94 [2821])discussedabove.The pipes from appear to have suddenly adopted the use of much the brewhouse culvert not only exhibitmoreupright, smaller wires, often producing unusuallysmall bores cylindrical forms, but they also have later finishing for this period of around4/6400.The exact date at characteristics,such as fewer bottered rimsand which thischange occurred needs to be pinned smaller stem bores. These differences clearlyindicate down more closely, but it shouldprovide an alater date for the culvert group than the pit group, importantwatershedindetermining the date of that is, after c 1690–1710. On the other hand, some early 18th-centurygroups. The stems themselves bottered rims still occur and there is one transitional taper to quite long, thin tips but, once again, their spur form with amilled rim, both of which are large bores provides aclue as to dating. characteristics showing that this group does notdate The final characteristic of this group is the use of from too far into the 18th century.For these reasons, afine sandy fabric that was usedtoproduce the the culvert group has been dated to c 1710–1730. This pipes. This is most evidentunder ax10 lens, when group provides an ideal opportunity to compareand numerous closely spaced sandgrains are visible on contrast the pipes of this period with the slightly the surface of the pipesand the broken edges earlier pit group of c 1690–1710. exhibitafine, granulartexture. Almost all of these The culvert group is rather smaller than the pit pipes are made of this type of fabric, which seems group and only 11 bowl styles can be identified. One to have been exploited from around1680–1750. of these is alate spur form of local style (Fig. 22.31) Similar fabrics are found at Reading (Higgins, but the otherten are all heel types,asimilar forthcoming) and the clay must have been obtained proportion to the pit group. Thespur form is once somewhere in this region of the Thames valley. It is again the only milled example, in thisinstance with interesting to note that the fine spur form (Fig. half the rim milled. This piece could be residual, 22.21) appears to be made of this type of fabric, but although it is perhaps more likely that it represents not the RichardCutts stem from East Woodhay the late production of on ‘old fashioned’ style for (Fig. 22.22). This suggests that the fine sandyclay is which there was still asmall market. In contrast to from adifferent source to that used by the East the pit group, however, none of the heel forms is Woodhay industry and that the fine spur form was particularly forward leaning and it is clear that a importedtoOxford from somewhere in the Thames more upright bowlform has become standard. One valleyarea, rather than from the heartland of this form, represented by at leasttwo examplesfrom the form in the Wiltshire/Hampshire area. same mould, still has quite asimple bowl form and

45 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford chunkyheel (Fig. 22.32), but the majority of the pipes discussed below, is rather scrappy. Such as it is, have smaller, more clearlydefinedheel areas (Fig. the evidence suggests (as would be expected) that 22.33–35,Fig. 23.36–37). In some instances, the bowl moulded decoration is adopted fromthe early 19th is extremelysharply angled back towards the stem centuryonwards. It also seems that the 19th-century (Fig. 23.36). One or two of the bowl forms overlap makersquite frequently adopted the use of moulded with those from the pit group, showing that both initialsorsymbolmarks on the heelsorspurs of deposits are quite close in date, but ageneral shift their pipes. towards new forms is clear. Asimilar change can be seen in the form of the Marked Pipes heels from the culvert groups, which are all either medium or small in size, showing that the use of Very few marked pipes were recoveredfrom the large heels had gone out of fashion. There is also excavations, which is typical of pipe assemblages more of atendency forthe heels to be slightlyflared, from the Oxford area. Becauseofthe small numbers especially when viewed from the ends of the pipe. involved, the stamped and moulded marks will be The pipes are still madeofafinesandy fabricbut the considered together.The markedpipes from the finishing techniques have changednotably over excavations are listed and described in alphabetical time. Although it is not the most immediately order below, with symbol marks at the end of the apparent difference between the two groups, one of list. the most marked changes is in the rim finish. There ED BEASTENAlong (99 mm) fragment of poorly burnished stem are 16 complete or partial rims from the culvert with alarge incuse mark reading ED/BEAS/TEN stamped across group out of which only six may have had their rims it was found in OXRA 86 [0810/A/2](Fig. 22.19). This has astem bottered (and some of these are dubious) as opposed bore 6/6400 and the style of the mark is typical of that used on the to every single rim fragment from the pit group stems of spur pipes by manufacturers in the Wiltshire/Hampshire area around 1690–1730. Despite being well known from his marks, being so treated. Furthermore, only oneofthe rims this maker has proved elusive in the documentary records. has been internally knife trimmed, acommon feature Cannon (1991, 22) dates this mark to c 1700–1730 and lists in the pit group. The majority of the culvert rims, 10 examples of it from Marlborough, Salisbury, Old Swindon, examples, have simply been formed with asingle Overton, Fulham, Stroud (31), Winchester (30), Newbury(25, 29), Littlecote, Swindon, Highworth, Cricklade (26), Hannington knife cut across the bowlaxistoproduce aquite Wick (70), Cirencester (78), Brinkworth and East Woodhay. In sharply squaredrim. This was aquicker production addition to some of these places Oswald (1991) has noted methodand one that was clearly taken up very other examples of this mark from Oxford and Faccombe Nether- rapidlyonce it had been introduced.Asimilar ton in north-west Hampshire.Cannon says that the greatest change towards quicker production methods is concentration of these marks comes from Newbury and suggests either Newburyorthe important pipemaking centre of East evidentinthe burnishing, which not only tends to Woodhay in north-west Hampshire as the source, althoughthe be of slightlypoorer qualitybut, most significantly, large number of examples that he notes from Hannington Wick in now only occurs on half of the bowl fragments north-east Wiltshire would seem to argue against this. The new recoveredasopposedtoall of them. The final change find adds to the known list of examples that were made by this prolific maker. to note is stem boresize,which is not related to the RICH/ARD/CVTTS Astem stamped with the incuse mark speed of manufacture, other than that it is slightly of Richard Cutts was found in OXSTRA 94 [2821] (Fig. 22.22). easier to produce asmall hole, especially towards Richard Cutts worked at East Woodhay in Hampshire, which was the tip of the pipe. Although one or two larger bores aregionallyimportant pipemaking centre. He was married there in are still present, the majorityofthe fragments from 1693 and baptised five children between 1694 and 1702, before 00 00 his own death in 1731 (Cannon 1991, 22). Two of his pipes have the culvert group have stem bores of 6/64 or 5/64 been recovered from Reading (Higgins, forthcoming)and others 00 with afew being as small as 4/64 .Aswith the are known from Chilton Foliat, Coleshill, East Woodhay,High- pit group, none of the pipes has an internal bowl worth, Littlecote, London, Marlborough, Newbury, Poulton (near cross and none of the culvert pipes is markedor Marlborough), Ramsbury, Southampton, Swindon and Winchester decorated. (Cannon 1991, 22). As with the Beasten mark (above), this shows the very wide market area that some makers were able to achieve, Although none of the individual characteristics with Cutts finding aregular market within a35mile radius and listed above can be reliably used in isolation, they with some pieces travelling as far as 60 miles from their place of clearlyprovide aseries of clearly definedtrends that, manufacture. This example is the first recorded from Oxford and taken together, provide an important meansof adds to the known distribution pattern for this maker. ID Aplain pipe bowl of c 1820–1860 with the relief moulded placing local pipe groups within abroad chronolo- initials ID on its spur was recovered from OXSTRA 94 [1641] (Fig. gical framework. It is also important to note how 23.40). The initials are quite large and simply executed and the distinct the two groups are from oneanother, despite pipe is not particularly well finished. Oswald (1984, 261–2) does their similarity in date. This underlines howsensitive not list any Oxfordshire makers with these initials and this piece is likely to represent apreviously unrecorded local maker. pipes can be for datingpurposes, since their forms AM Aheel bowl of c 1690–1720 made of afine sandy fabric and and finishing characteristics changedrapidly over with an internally trimmed and bottered rim (but no milling) was time while their short life expectancy often means found in OXRA 86 [0906/4].The pipe has astem bore of 6/6400 that they are the most recently produced artefact and both the bowl and stem have an average burnish. About type to be discardedinany given deposit. 26 mm from the bowl aband of milling has been impressed around the stem and then the incuse letters AM have been Unfortunately, there are no good examples of impressed 32 mm from the bowl. The use of unbordered incuse mid- to late 18th-century pipes from this site and stamped initials in this way is very unusual nationally but the the 19th-centuryevidence, other than the group style appears to have used by afew of the Oxford area

46 OA Occasional Paper No: 16 manufacturers from around 1680–1720. Oswald (1984, figs 52.7 borders were becoming fashionable at thistime and and 10) illustrates late 17th-century bowl forms with the initials it is possible that this maker was using an ordinary RP and MC stamped on their stems in this manner. He dates these types from as early as 1660 but the present author believes milling tool in conjunction with his initials to mimic the bowl forms to date from around 1680 or later, which would this style. also fit better with the date attributedtothis new mark. None of From the late 18th century pipes were often mould these three sets of initials (MC, AM or RP) can be matched with decorated but there are very few of these types known makers, but the style of the bowls is certainly local as is represented from the site, since not many late 18th- the use of afine sandy fabric for the new piece. AlthoughnoAM maker is currently known,itisworth noting that aJohn or 19th-century deposits were excavated. One frag- Matthews was recorded as apipemaker in Oxford in 1764 mentary bowlof c 1820–70 was recoveredfrom (Oswald 1984, 262) and so it is possible that AM was an earlier OXSTRA 94 [1629] with short, narrow, fluted around member of the same family. the lower halfofthe bowland hanging swags above WT Relief-moulded initials on the heel from apipe dating from c 1790–1830 (Fig. 23.38). The bowl is missing, but slight traces at (Fig. 23.39). The spur, which may have been marked, the broken edge suggest that it was probably mould decorated is missing. originally. The base of the heel has not been trimmed, acharac- Apart from aheel marked WT that may have had teristic of pipes produced after about 1800. AWilliam Tuckwell is adecorated bowl (Fig. 23.38), the only other deco- recorded at Wallingford in 1796 (Oswald 1984, 262) and he could rated material was recovered from amid 19th- have been the maker of this pipe, although the large number of excavated pipes from Reading, which is about the same distance centurylevelling layer, OXSTRA 96 [0010]; see from Wallingford as Oxford is, do not include any WT marks below. (Higgins, forthcoming). There are anumber of other pipes with these initials known from Oxford (eg Oswald 1984, figs 55.29a; 55.29b and 56.36) and so it is perhaps more likely that this maker worked in or near Oxford itself. All of the other known examples Conclusions are of asimilar date and form, but with plain bowls. ** Aplain 19th-century bowl, probably dating from c 1830–1870, The excavations at Rewley Abbeyhave produced an was recovered from OXRR 93 [0123] (Fig. 23.41). The spur is interesting range of pipes that help refineour largely chipped away but enough survives on the right hand side understanding of the Oxford pipe making industry. to show that it had acrudely executed star on it. There would have been another one on the other side. Symbol marks such as this Although there are none of the earliest pipe styles, were quite common during the 19th century and they are not the forms from around1640 through to the middle related to any specific maker. Two further examples of pipes with of the 18th centuryare fairly well represented and star marks were recovered from OXSTRA 96 [0010] (see below; there are afew 19th-century groups. The17th- and Fig. 43.43). 18th-century styles are similar to thosethat have been identified from Reading (Higgins, forthcom- Decorated Pipes ing) and there arecertainly similarities between the two areas, for example in the exploitation of Very few of the excavated pipes are decorated. Those presumably local fine sandy fabrics.Despite this, that were recovered are described and discussed subtle differences in the bowl forms and finishing below. techniques clearlyshow that the two towns both An elaborately decorated stem, almost certainly had their own local pipemaking industries that from aDutch pipe of c 1660–1740, was recovered supplied the majority of their needs.One particular from OXRA86[0810/1] (Fig. 21.18). Immediately difference can be seen in the use of moulded behind the bowl junction are five identical roll- makers’ initials during the 18th century. This type stamped borders decorated with geometric patterns of mark was typical of London and many of the and dots. Just overlapping these astemtwist has surrounding counties and it was certainly adopted then been applied, followedbyfourlinesof at Reading. But the use of moulded initialswas decorative milling, one of which clearly runs over extremely rare in Oxford before the 19th century, the borders.There are also traces of burnishing showing that marking fashions at the two centres running back from the bowl. This was clearlyawell- evolved independently. made and verydecorative pipe, most likely pro- The local marks that occur also reflect the degree duced around 1660–1740. Dutch pipes of this date of self-sufficiency in the local pipemaking industry, are not often foundinEnglandand, when theyare, whilethe small number of importedmarks reflect they tendtohave acoastal distribution, suggest- the areas fromwhich pipes (and by extension other ing that they were primarily being carried as the goods and services) were being drawn. The locally personal possessions of sailors rather than being produced pipes at Oxford dominate the marketand tradedobjects in their own right. This is an inte- suggest that many more local manufacturers remain resting and unusual object for an inland site andone to be identified from documentary sources. Where that perhaps reflects the international links and marks do occur, such as the AM stem mark, they are cosmopolitan nature of Oxford itself. often in local styles and seem likely to represent Aband of milled stem decoration was usedin as yet unidentified local makers.Afew ‘imported’ conjunction with incuse maker’s initials (AM)ona marks occur, such as the RichardCutts mark from pipe from OXRA 86 [0906/4] (Fig. 21.16). This piece East Woodhay. These show asmall number of pipes probablydates from c 1690–1720and the use of arrivingfrom the Wiltshire/Hampshire area, per- milling on the stem is interesting since it was not haps reflecting individual travellers and/or trading used on the bowl rim. Decorative roll-stamped stem links. Asmall number of pipes fromthese areas have

47 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford also beennoted at Reading and London, showing 8. Heel bowl datingfrom c 1640–1670; notburn- this to be partofawider pattern of movements head- ished; rim bottered and fullymilled;stem bore ing towards the capital. The Dutch stem is interest- 7/64 00.(OXRA81 Context: 11). ing in that it represents amore distant source, and 9. Heel bowl datingfrom c 1640–1670; notburn- one that rarely reachedthe inland areasofEngland. ished; rim bottered and fullymilled;stem bore Some of the deposits from Rewley have produced 7/64 00.(OXRA81 Context: 11). very coherentand tightly datable groups of pipes. 10. Heel bowl datingfrom c 1640–1670; notburn- The pit group of c 1700 (OXSTRA 94 [2821])and ished; rim bottered and fullymilled;stem bore the early 18th-century pipes from the brew-house 7/64 00.(OXRA81 Context: 11). culvert (OXRA86[0010] &[0010/A/2]) bothprovide 11. Spur bowl dating from c 1640–1680; not burn- importantreference groupsthat not only characterise ished; rim bottered and three-quarters milled; the local pipes at these period but also provide stem bore7/6400.(OXRA81 Context:11). benchmarks forfuturework.The later pipes are not 12. Spur bowldating from c 1640–1680; average so well represented but the mid 19th-century group burnish; rim bottered and three-quarters milled; hints at the more decorative designsthat would have stem bore7/6400.(OXRA81 Context:11). been currentatthat time as well as showing that local 13. Spur bowldating from c 1680–1700; average manufacturers were still providing for the town until burnish; rim internally trimmed, bottered and at least the middle of the century. three-quartersmilled; stem bore7/6400.(OXRA81 Context: 11). 14. Spur bowldating from c 1660–1690; average List of illustrations (Figs 21–23) burnish; rim bottered and one-quarter milled; 00 All of the pipes are illustrated at 1:1 with burnished stem bore7/64 .(OXRA93 Context:204). surfaces being shown with abroken line on the surface 15. Largespurbowldatingfrom c 1690–1720; average burnish;rim botteredand half milled; of the pipe. The illustrations have beenarranged in 00 roughly chronological order to show the evolution of stem bore9/64 .(OXSTRA94 Context: 2134). types as represented amongst the excavated assem- 16. Heel bowl dating from c 1690–1720; average burnish; rim internally trimmedand bottered blage.Four particularly good groups of material are 00 represented from this site as follows: - but with no milling; stem bore 6/64 .About 26 mm from the bowl aband of milling has been Nos 7–13 OXRA 81 [0011]Accumulated deposit impressed around the stem and thenthe incuse over aperiod of c 1650–80. letters AM have been impressed. Themaker is Nos 20–30 OXSTRA 94 [2821] Pit group of c 1690– unknown but this is alocal style of mark. 1710 (tight group –probably c 1700). (OXRA86 Context: 906/4). Nos 31–37 OXRA 86 [0010] Brewhouse culvert fill 17. Heel bowl dating from c 1680–1710; good of c 1710–1730 (good group) and [0010/ burnish; rim is cut and internally trimmed. A/2] (OXRA93 Context: 204). Nos 42–47 OXSTRA 96 [0010] Levellinglayer, pro- 18. Stem fragment dating from c 1660–1740; tra- bably c 1850 for station construction ces of burnishing;stem bore 6/6400.This elab- (see below; Fig. 43). orately decorated stem is almost certainly Dutch. Immediately behind the bowlare 1. Heel bowl dating from c 1620–1650; finelyburn- five identical roll-stamped borders decorated ished; riminternally trimmed, botteredand fully with geometric patterns and dots. Just over- milled; stem bore 7/6400.(OXRA86 Context:6/C). lapping these astem twist has thenbeen 2. Heel bowldatingfrom c 1640–1660; notburn- applied, followedbyfour linesofdecorative ished; riminternallytrimmed, botteredand fully milling, one of which clearlyruns over the milled;stembore8/64 00.(OXRA81 Context: 6/1). borders. 3. Heel bowl datingfrom c 1640–1670; notburn- 19. Stem fragment dating from c 1690–1730; poorly ished; rim bottered and fullymilled;stem bore burnished;stem bore 6/64 00.The stem has alarge 7/64 00.(OXRA81 Context: 11/1). incuse mark reading ED/BEAS/TEN stamped 4. Heart-shaped heel fragment datingfrom c 1640– across is. Alarge number of these marks have 1670;not burnished; stem bore 8/64 00.(OXRA81 beenrecovered from sites in the Wiltshire/ Context: 11/1). Hampshire area although the maker has not 5. Spur bowl dating from c 1650–1670; not burn- yet been traced in the documentary records. ished; rim bottered and fullymilled;stem bore (OXRA86 Context: 810/A/2). 7/64 00.(OXRR93 Context:140). 20. Spur bowldating from c 1690–1710; average 6. Spur bowl dating from c 1650–1670; not burn- burnish; rim bottered and internally trimmed ished; rim bottered and half milled; stem bore but not milled; stem bore7/6400.(OXSTRA94 7/64 00.(OXRR93 Context:140) Context: 2821). 7. Heel bowl datingfrom c 1640–1670; notburn- 21. Spur bowldating from c 1690–1710; finely ished; rim bottered and fullymilled;stem bore burnished;rim bottered but not milled; stem bore 7/64 00.(OXRA81 Context: 11). 6/64 00.(OXSTRA94 Context: 2821).

48 OA Occasional Paper No: 16

Figure 21 Clay pipes,nos 1–18. Drawn by Dr Susie White.

49 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 22 Clay pipes, nos 19–35. Drawn by Dr Susie White.

50 OA Occasional Paper No: 16

Figure 23 Clay pipes,nos 36–41. Drawn by Dr Susie White.

22. Stem fragment dating from c 1690–1710; poorly maker who married in 1693 and diedin1731. burnished;stem bore6/6400.The stem has the (OXSTRA94 Context: 2821). incuse lettering RIH/ARD/CUTTS stamped 23. Heel bowl dating from c 1690–1710; average across it. RichardCutts was an East Woodhay burnish; rim bottered and internally trimmed

51 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

but not milled; stem bore7/6400.(OXSTRA94 swags hanging above. (OXSTRA94 Context Context: 2821). 1629). 24. Heel bowl dating from c 1690–1710; average 40. Spur bowldating from c 1820–1860; not bur- burnish; rim bottered and internally trimmed nished; cut rim with no milling; stem bore 4/6400. but not milled; stem bore5/6400.(OXSTRA94 On either sideofthe spurare the relief moulded Context: 2821). initials ID. Thereisnoknown maker with these 25. Heel bowl dating from c 1690–1710; good initials, but it is most likely to be alocal product. burnish; rim bottered and internally trimmed (OXSTRA94 Context: 1641). but not milled; stem bore6/6400.(OXSTRA94 41. Spur bowldating from c 1830–1870; not bur- Context: 2821). nished; cut rim with no milling; stem bore 4/6400. 26. Heel bowl dating from c 1690–1710; average Thespurisnow missing.There are traces of astar burnish; rim bottered and internally trimmed mark on oneside of the spur, and it would almost but not milled. (OXSTRA94 Context: 2821). certainly have been the sameonboth sides. 27. Fragment of aheel bowl datingfrom c 1690– (OXRR93 Context:123). 1710;average burnish; no rim surviving; stem bore7/64 00.(OXSTRA94 Context: 2821). 28. Heel bowl dating from c 1690–1710; average The Ceramic Building Materials burnish; rim bottered but not milled; stem bore John Tibbles,with acontribution by John Cotter 6/64 00.(OXSTRA94 Context: 2821). Introduction and methodology 29. Heel bowl dating from c 1690–1710; average burnish; rim bottered and internally trimmed Atotal of 1380 fragmentsofceramicbuildingmaterial but not milled; stem bore6/6400.(OXSTRA94 weighing84349 grams, and72fragments of stoneroof Context: 2821). tile weighing13007grams weresubmitted forexami- 30. Heel bowl dating from c 1690–1710; average nation.All thefragments wereretrieved from 233 burnish; rim bottered and internally trimmed contexts andwerevisibly examined using a15x-magni- but not milled; stem bore6/6400.(OXSTRA94 ficationlens. Information regarding the dimensions, Context: 2821). shapeand fabric of thematerialwas recorded and 31. Spur bowldating from c 1690–1720; average catalogued accordingly. It shouldbenoted that the burnish; rim bottered and half milled;stem bore diversityofsizeand colour within thetile causeddur- 7/64 00.(OXRA86 Context: 10/A/2). ingthe manufacturingprocess must be takeninto 32. Heel bowldating from c 1700–1730; one of two considerationwhencomparingexamples within col- bowls from the samemould; not burnished; lectedassemblagesand local typologies. Thevarying rim cut; stem bore5/64 00.(OXRA86 Context: 10/ sizesand colours can be attributed to thevariation in A/2). theclays used, shrinkageduringdrying, firingwithin 33. Heel bowl datingfrom c 1700–1750; poorly thekiln or clampand thelocationofthe tile within burnished;rim bottered and internally trimmed thekiln.Fabric descriptions canbefoundinthe site but not milled; stem bore5/6400.(OXRA86 archive. Context: 10/A/2). 34. Heel bowl dating from c 1700–1750; good The assemblage burnish; rim cut and notmilled; stem bore 5/64 00.(OXRA86 Context: 10/A/2). Roof tiles 35. Heel bowldating from c 1700–1750; not bur- Thereisclear evidence to show that clay roof-tiles nished; rim bottered but not milled; stem bore were in use within England by the late 13th century 00 5/64 .(OXRA86 Context: 10/A/2). (Drury 1981) and continued throughtothe 18th 36. Heel bowl dating from c 1700–1730; average century. Non-diagnostic or fragments of flat tile with burnish; cut rim with no milling; stem bore4/ no suspension or otheridentifiable characteristics can 00 64 .Bowl is sharply angled back towards the therefore only be attributedwithin this date range. stem. (OXRA86 Context:10/A/2). 37. Heel bowl datingfrom c 1700–1750; notburn- ished; cut rim with no milling; stem bore 5/6400. Table 5Ceramic building materials, assemblage quanti- (OXRA86 Context: 10/A/2). fication. 38. Fragment of aheel bowl datingfrom c 1790– 1830;not burnished; rim bottered but not milled; Form No. of fragments Weight (g) stem bore 5/6400.The bowlismissing, but slight tracesatthe broken edge suggest that it was Brick 84 10199 probably mould decorated originally. On either Flat roof tile 689 32360 sideofthe heel are the relief moulded initials WT. Floor tile 505 34210 (OXRA86 Context: 10/A/2). Ridge tile 102 7580 39. Fragment of aspurbowl datingfrom c 1820– Stone tile 72 13007 1870;not burnished; cut rim;stem bore 4/6400. Totals 1452 97356 Mould-decorated with short narrow flutes with

52 OA Occasional Paper No: 16

The contextual deposition of the ceramic tile were recorded of which 58 fragments (57%)were assemblage is often of limited interpretative value, of F11. although it doesreflect avariety of forms and use within construction. Thematerial provides evidence Brick for the architecture of buildings that may have had flat-tiledroofs cappedwith ridge tiles. Themajority Approximately 7.5% of the ceramic building material of the assemblage was of good condition. assemblage comprised of medieval bricks, although none was complete or near complete. Dating of bricks is highly contentious due to their Flat roof tile re-use nature as avaluablebuilding commodity. At 00 00 1 00 Representing 48% of the total assemblage, 689 York in 1505 bricks were standardised at 10 · 5 · 2 ⁄2 ; fragments of flat roof tile were identified, of which Parliament in 1571 decreed that the size of abrick 00 1 00 1 00 sixty-three fragmentsboresuspension holes. The shouldbe9 · 4 ⁄2 · 2 ⁄4 ,revising that in 1725 to 00 1 00 00 holes ranged between 10 mm–17 mm in diameter 9 · 4 ⁄2 · 2 and approximately 50% showed adiameter of bet- Of the eighty-four fragmentsofbrick within the ween 14 mm–15 mm. Twojoining fragments from assemblage none were complete or near-complete. context 250 formed afull width of atile 175 mm Two part bricks from context 1610 andone from acrosscontaining two suspension holes 85 mm context 1658 displayedbothwidth and thickness apart. Two further joining fragmentsfrom context (110 mm · 56 mm and 100 mm · 50 mm respec- 1152 also showedtwo suspension holes 50 mm tively). The remainder of the assemblage displayed apart. This tile type can be equated with near- thickness only ranging between 40 mm–60 mm 5 00 3 00 parallels recorded at Oxford Castle(Tibbles 2005a). (1 ⁄8 –2 ⁄8 )all comparable with medieval manufac- The majority of the identifiable flat tiles were of this ture. Sixteen fabrics were recorded of which F16 type [Type 28] (Tibbles in prep.), which represented was the most common(33%). (9%) of the tile assemblage. The remaining frag- Asinglefragment fromcontext 818/A may be of ments displaying thickness only were unable to be modern origin. In addition to the traditional finished classified. brick the assemblage containedexamplesof‘samels’ Thickness of the tiles rangedbetween 7mm– (underfired bricks). Only one fragment showed 18 mm with 391 tiles (57%)within a14mm–16 mm evidence of light post-breakage burningthat may range.Twenty-three fabricswere provisionally iden- suggest demolitionburning rather than secondary tified of which F2, F4 and F5 were most common. usage withinhearth make-up. Thirty-nine flat tiles (6%) were eitherglazed or dis- played splashes which ranged between adark olive The floor tile green to adark reddish brown. Diagnostic qualities included the varying methods Five-hundred and five fragmentsoffloor tile of suspension, length, widthand thickness. How- representing 35% of the total assemblage was ever, thickness alone suggested multiple possibilities examined. The majority were eitherdecorated or within the flat roof tile typologyand were therefore plain glazed of which 69 boredesignsorpartthereof. impractical to attemptidentification.Arabbits paw Atotal of onehundred and eighty-one(36%)showed impression ( Oryctolagus cuniculus)was noted within bevels between 8 – and 10– and sixty-six (13%)dis- atile fragment from context 126 andagrey slip/ played ‘stabbed’ keying on the underside. Most of wash over the tile surface from context 140. the stabbing appeared random although several tiles exhibited an ‘oblique’ stabbingpattern. Asubstantial proportion of the tilesdisplayed Ridge tiles heavy wear obliterating many designsand glazes. At One hundred and two fragments of ridge tile least two glazedtilesexhibited afresh crisp appea- representing 7% of the assemblage were identified rance suggesting an original location close to awall within the assemblage, of which 27 fragmentswere or hidden out of general access. Glazes ranged glazed.Two types of tile crests were recorded: pin- between reddish brown, dark/light olivegreen and ched/thumbed crest, possibly 13th–14th century,and yellow with variations between. The majority of the knife-cut/dog-tooth. Theformer was most common glazed tiles bore partialdesignsorremnantsofwhite with 27 fragments recorded whilst the latter only two slip and residual glazes.Partial designshave been fragments recorded. Six fragments from context 2512 provisionally identified as floral(1003, 501, 261), were found to join forming approximately 68% of a eagle wings (12), fleur-de-lis (301/1), animal head complete tile. Alarge fragment of the crested type (204) and stag (702, 1003). Tiles from contexts 1818, 1 from withincontexts 1716 had its crestsremoved by 1010, 401 and 1153 show ⁄4 tile designs. knife and re-used as arepair (?) piece. Plain yellow,green and dark green/black glazes Two fragmentsofridge tile from contexts 107 and were also recorded on five square tilesand twelve 208 displayedthe residual elements of finial sockets. triangular tiles. The latter tiles having been scored Thickness of ridge tile varied between 7mm–20 mm prior to firing to facilitate post-firing division. with 92 fragments(90%)displaying athickness of Seven decorated tilesfrom contexts 500, 501 and between 10 mm–16 mm.Eleven different fabrics 1003 displayed heavy mortar adhesions on all or

53 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Table 6Floor tile thicknessbyquantity. Decoratedmedieval floor tiles in the Oxfordshire and Buckinghamshire region have beenthe study Thickness Range Quantity of two detailed and, by now, classic reports: the first 10 mm–17 mm 31 of these by Loyd Haberly (1937), and the second 18 mm–22 mm 242 by ChristopherHohler (1942).Referencestotiles 23 mm–26 mm 63 in Haberly’s catalogue are usually abbreviated ‘LH’ 27 mm–34 mm 39 followed by the catalogue number in Romannum- erals (eg LH XLIX).Other importantassemblages of Total 375 ceramic building materials from Oxford, including decorated floor tiles, include those fromthe Domin- ican Priory (Blackfriars) (Lambrick and Mellor 1985), St Frideswide’s Priory (Oxford Cathedral) (Green most surfaces suggesting re-usedascoursing level- 1988), St Ebbe’s (Greyfriars) (Mellor 1989), and lers or internal rubblefilling. Other tiles displayed more recently from Beaumont Palace (Whitefriars) mortar over breaks suggesting their use as repair (Atherton and Mitchell 2001). Amedieval tile fabric pieces.Atleast three tilesfrom contexts 401, 731/2 series forthe Oxford regionwas first devised for and 1909 displayedglaze over breakage and/ The Hamel site, Oxford (Robinson 1980) and subse- or mis-firing suggesting their use as seconds. Tile quentlyextended in some of the reports listed above. thickness rangedbetween 10 mm–34 mm (See The 18 floor tile fabrics from Rewley Abbeyiden- Table 6) with 65% withinasize range of between tified by JTibbles (above) are almost certainly vari- 18 mm–22 mm. ants of thoseinthe existing Oxford tile fabric reference collection, although no attempthas been Item of intrinsic interest made here to correlatethese. Loyd Haberly (1937, 76) lists the site of Rewley Afragment of acarveddisc was recovered from Abbeyasone of the many placeshevisited in within context 226/A/1 manufactured from re-used gathering data for his book, although he offers few glazed floortile.Itdisplayedadiameter of c 45 mm, details of what he mighthave seenthere.Hemay, in athickness of 17 mm and weighed 40 g. The frag- fact, have been referring to acollection of floor tiles in ment had been cuttocreate amulti-faceted edge. the Ashmolean Museum which, perhaps, included Flat roof tile was generallythe accepted raw tiles from the site,ormore likely to sketchbooks of material possiblybecauseofits standard thickness of tile designsbylocal Victorian antiquarians such as between 12 mm–16 mm and couldbeeasily chipped Percy Manning and the Parker-Hoare family. His to the desired diameter. Occasionally an alternate comments are sometimes alittle obscure and eccen- raw material was utilisedsuch as stone. tric to modern eyes but it is evident that he (and Discs shaped from potsherds also appear with Hohler) must have seen tiles, or drawings,from the assemblages but are generally much smaller in Rewley Abbey site. Haberly (196–8) illustrates four diameter. Althoughdiscs have been recorded within ‘printed’ floor tile designs fromMerton College Romancontexts theygenerallyappear from the Library(LH CIL, CL-II) showing poorly executed thirteenth centuryand continue throughtothe post- or jumbled letters of the alphabetinterspersed with medieval period where they are likely to be of a hares or rabbitswithinalarger roundel (each partof residual nature.Objects of asimilar nature have been afour tile pattern). He describes these as belonging recorded at Oxford (Tibbles 2005a), Beverley (Tibbles to apuzzling and incomplete series and mocks that 2005b), and Hull (Tibbles forthcoming). Their exact they are unfit foranacademic institution such as use is still arguable but previously they have been MertonCollege: ‘‘Possibly the 17th-century builder recorded generally under the generic terms of pot who laid the library floor had inherited them from a lids, counters or tally markers. They are more likely, 16th-century ancestor who bought the spoils of however, to have beenusedasgaming counters for Rewley Abbey!’’. He adds that one of the tiles (LH the game of TabulaorTables. CL) ‘‘was drawn by Manning as aRewley tile’’. Haberly’s alphabettiles are listed under his ‘printed Acommentary on the source and local context and unkeyed’ tiles, implying, as seems evidentfrom of the floor tiles the drawings,that they were products of the Penn or John Cotter Chilternfactories and therefore would now be dated to c 1330–1380. However, no examplesofthese Although the quantity and material characteristics of ‘alphabet’ tileswere recognised from the excavations the Rewley Abbey floortileshas been dealtwith in discussed here. An eagle tile from Rewley Abbey, some detailabove, afew comments on their local originally drawn by Manning, is also illustrated by context and significance can usefully be offered here. Haberly (LHLXVI). Examples of the latter were Afew of the best preserved tiles from the excavations – foundinthe assemblage discussed here (see below). some of which can be paralleled on other sitesinthe Hohler (1942) also has scattered references to region –are illustrated here (Plate 11), although, as Rewley Abbey tiles. pointedout earlier,most are very worn and frag- The illustrated floor tiles from the recent excava- mentary and the designsmostly unrecognisable. tions that canbesafely paralleledelsewhere in the

54 OA Occasional Paper No: 16 city or the region are shown in Plate 11 (see illus- is only paralled elsewhere in Oxford at Carfax tration list for details),although the list of parallels is church (N Mitchell, in notes).This design, in origin, by no means exhaustive. Most of the tiles with may hark back to Richardthe Lionheart doing battle recognisabledesignsare fairly common types either against Saladin –adesign that occurs in several tile- in Oxford itself or the nearby region, or they are close makingtraditions including the celebrated Chertsey variants of these. Lesscommon types arementioned tiles (Eames1968, 7–8) and on those from Hailes below. AbbeyinGloucestershire –founded by RichardEarl of Cornwall (father of Edmund)in1251 (examples in British Museum). Description Two other ‘stabbed Wessex’ tile fragmentsfrom Decorated medieval floortiles from Oxford are Rewley (OXRA 81 (12), OXSTRA 94 (1818),not generally of two main types, or traditions, reflecting illustrated) are from tilesshowing an eagle tile with two main source areas and two successive chron- outstretched wings.This is one of the Rewley Abbey ologicalperiods. Designs from the earlier tradition tiles illustrated by Haberly, also known from Carfax were inherited by the later tradition but were Church (LH LXVI) and both Blackfriars and Grey- increasingly debased with the passing of time. Also friars (N Mitchell, in notes). Arelateddesign the the technology of tile manufacture changed from double headed eagle (LH LXI though not from ‘inlaid’ to ‘printed’ designs in ways that are still not Rewley) was, in origin, the badge of the Holy Roman fully understood. Floortilesofthe earlier tradition emperors and also occurs at Hailes Abbeywhere it are known as the ‘stabbedWessex’ type. These have may be areference to RichardofCornwall’s claim to afairly uniform orange or brick-redsandyfabricof be king of Jerusalem.However, the design is also relatively fine-medium coarseness,usually with a known fromOxford Cathedral and OsneyAbbey. By grey core (Oxford tile Fabric IIIB). They derive part the later 13th century many tile designssuch as of their name from the distinctive deepcircular these, which originally may have had more specific stabbing or keying on their undersides –although and heraldic significance,had become part of the not every example displays this feature. Compared general repertoire of many tile factories and may with later tilesthe designs in contrasting white slip have lost much their original significance.Ifthere are unusually deeply inlaid(up to 5mmdeep) into a were an unusually high number of these particular previously stamped design or pattern and the whole designsfrom Rewley Abbey there mighthave been a design is generally more detailed andmore carefully case for connecting them with Edmund of Corn- executed than thoseoflater tiles. Many other, more wall’s foundation and dynastic claims,but as there subtle, differencesalso distinguishthe earlier from are so few this connection, tantalizing though it the later tiles (size,thickness, pattern subject matter seems,appears rather tenuous. etc.), but these distinctionsare alittle more sub- Floor tiles of the later tradition found in Oxford jective.The main production period of ‘stabbed are described as ‘printed’ owing to the thinness of Wessex’ tiles is thought to have been c 1280–1330 the white slip designswhich wear offmuch more (arguably from c 1250) (Lambrick and Mellor 1985, easily than the earlier ‘inlaid’ tiles. They are also un- 185–6). No kilns producing the ‘Stabbed Wessex’ stabbed or unkeyedonthe reverse. ‘Printed’ tiles tiles have yet been discovered in the regionbut the tend to have apaler-firing fabricthan the ‘stabbed similarity of the sandybrick-red fabric to medieval Wessex’ tiles. This is generallylight orange or pottery believedtohave been produced in the pink-buff in colour and has more visible coarse Newbury-Reading areaofBerkshire strongly sug- inclusions –mostly hard red-brown clay pellets or gests asource in this area (ibid., 186). ‘grog’, red iron oxide and sometimes thin streaks of Tiles of the ‘stabbedWessex’ tradition tend to cream or pinkish marlyclay –but relatively little form the majorityofdecorated floor tiles excavated quartz sandgiving asmoother matrix texture. Oxford from monastic sites in Oxford,most probably Fabric IIIC is ahard grog-tempered fabricthought to reflecting the fact that the late 13th and early 14th originate from the Nettlebed area of south-east centurywas amajor period of ecclesiastical building Oxfordshire, while Fabric IVB is sandier and probably or renewal in the city. The tiles from RewleyAbbey comes from the Penn or ChilternfactoriesinBuck- conform to this patternwith the c 1281 foundation inghamshire (Lambrick and Mellor 1985, 178, 186). date for the abbey providinguseful additional Printed tiles are mainly datable to c 1330–1380(Green confirmation forthe introductionofthe ‘stabbed 1988, 109) with some limited outputaslate as the 15th Wessex’ type. Most of the tiles illustrated from the century. They are commonfrom the Rewley Abbey excavations are of thistype (Plate 11. T1–4, T6).Tiles site, though less so than the earlier ‘stabbed Wessex’ T1 and T2 with amixture of stylized floraland type. Their presence in quantity suggeststhat new geometric designs are common types but T3 (qua- tiled floors were laidinthe Abbeyaroundthe mid- trefoilsinsquares) is much less common. Acut dle of the 14th century, or that earlier floors were triangular tile fragment is decorated with stylized extensively repaired. Recognisable designsinclude fleur de lys which can be paralleled on several other several examples of the four tile pattern shown in Oxford sites (LH XXXI, OXSTRA 94 1804, not Plate 11.T5 (probably aNettlebed fabric) with a illustrated). There is asinglefairly definite example procession of four-legged monstroushuman-headed of aknight on horseback (T6) which, it would seem, creatures within acusped roundel and aquatrefoil in

55 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford the centre. Another recognisable four tile pattern Afew fragmentsofmuch thicker and much (OXRA 86 1003, not illustrated) also has alarge larger tiles with afine silty orangefabric are likely roundel enclosingfleur de lys and stylized vegetation to be late medieval imports from Flanders (mainly within cusped triangles outside the roundel (LH 14th–15th century). These are either plain with a CLXXXVIII, also at Dorchester Abbey). clear glaze, or an almost black glaze, or covered

Plate 11 Medieval decoratedfloortiles.

56 OA Occasional Paper No: 16 with white slip under aclear (yellow) or agreen in outstretched arm and shield in other. Very glaze. One or two examples are pre-cutfor worn. Stabbed underside. 20 mm thick. Design divisioninto two triangular tiles. Flemish tiles also known fromCarfax Church,Oxford, were often laid in chequerboard patterns or used Eynsham Abbeyand Leicester Abbey. here and there for repairs in older floors.Some of the earlier plain tiles from the ‘stabbedWessex’ and printed tile sourceswere probablyused in a Glass similar way. Both the latter include examplesof Hugh Willmott pre-cut tilesfor areas against the wall and else- Areasonable sizedassemblage, consisting of around where where small triangular tileswere neededas 430 fragmentsofglass from 78 different contexts, fillers (Plate 11.T4). was recovered from the excavations at Rewley None of the Rewley tileswas recoveredfrom an Abbey. The glass ranges in date from the 13th to in situ tiledfloor, although an areaofmortar the 20th centuries, the majority being from the latter beddingshowing tile scars was foundinthe area end of this span. This report is principally concerned of the reredorter (Trench 1994/16) and alittle with glass that relates to the period of use of Rewley furtherwest insidethe northern boundary wall Abbey, and the immediate post-Dissolution period. (Trench1994/17). Theveryworn state of most of Alist of all the glass is included in the project the tiles indicated that they had been in use for long archive.The glass notfully catalogued and discussed period of time. Many had evidently been reused as mainly consists of wine and otherbottle fragments, levelling or hardcore. It is likely that many of the as well as ordinaryplain windowglass, common better tiles were carted off for reuse elsewhere at the glass finds on almost all post-medieval sites, and Dissolution. therefore not particularlydiagnostic.

List of illustrated floor tiles (Plate 11) Glass dating to pre-1536 T1. OXSTRA 94 (1828).‘Stabbed Wessex’ type. Matches LH XLIX (Loyd Haberly 1937). Size The medieval glass can be divided into two broad 132 mm · 130mm · 21 mm thick.Stabbed un- categories, vessel and window. In total seven derside. Bevellededges. Design also known fragments of glass from aminimum of four vessels from OXRA 86 (1003) andthe Greyfriars (St were recovered. All are highly weatheredand in an Ebbe’s), Oxford (Mellor1989, fig. 78.6). advance stageofdevitrification,asisusual with T2. OXSTRA 94 (1817).‘Stabbed Wessex’ type. glass of this date. The most characteristic, GL1, is Matches LH I. Length130 mm · 22 mm thick. asmall portion of pushed-in base from asmall squat Stabbedunderside. Bevellededges. Design also beakerortumbler. Madeinaclear soda-rich glass known from Oxford Cathedraland several this is the only imported tableware present in the otherOxford sites including the Blackfriars medieval assemblage. This form, known as amoiolo (Dominican Priory) (Lambrick and Mellor 1985, in Italy where it was produced, was used for fig. 20.9). Also at Dorchester Abbey. drinking wine and probably dates to the later 15th T3. OXSTRA 94 (1818).‘Stabbed Wessex’ type. century(Charleston 1984, 43). The other medieval VariantofLHL(LH has apiercing in each vessel forms are more ordinaryand typical English quatrefoil). The surviving white slip square products (Willmott 2005, 60). Onefragment, GL2, is measures 59 mm · 56 mm. 25 mm thick. Dee- the very lower portion of astub base from ahanging ply stabbed underside. Bevelled edges. Design lamp. These are finds commonly associated with also known from Oxford Cathedral (St Frides- monastic houses, such as ninerather more complete wides)(Green1988, fig. 49.12). examples excavated at Battle Abbey (Charleston T4. OXSTRA 94 (1818).‘Stabbed Wessex’ type? The 1985, 140). Although lamps are most commonin surviving quarter measures 70 mm square. the 13th to 15th centuries, theyare occasionally 20 mm thick Original tile pre-cutinto 8small foundin16th-century contexts. Theremaining two triangular tiles. All over white slip under clear fragments, GL3–4, are portions of evertedrim and glaze with copper-green flecks. neck from ordinarypotash glass flasks, perhaps the T5. OXRA 86 (1003).Printed, probablyPenn-type. most common medieval vessel and typically of late Two tile fragmentsfrom afour tile schemewith 13th- to 15th-century date. repeating pattern of monstrousormythologi- The majorityofthe medieval glasscomes from cal creatures within acusped roundel. Length windows ratherthan vessels, and three examplesare 145 mm. 20 and 24 mm thick. Related to Hohler painted on their interior surface. All are in a13th- W25–6 and LH CXLV-VI. More fragmentsof centurygrisaille style,apattern that, given its use of this design from OXSTRA 94 (1833) and (1844). naturalistic and foliage design,was particularly Similar from Oxford Cathedral(St Frideswides) favoured by Cistercian houses. Two are from sections (Green1988, fig. 49.9). of stickwork borders,GL5 having the remainsofa T6. OXRA 86 (1003).‘Stabbed Wessex’ type. foliagedesign,and GL6 being made up from aseries Matches LH LVIII. Corner fragment showing of compacted roundels. The final painted piece,GL7 knightonhorseback holding upright sword has asection of tight cross-hatching, the typical infill

57 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford used between foliage patterns in grisaille work. Catalogue (selected pieces) Fragments of plain windowglass were also recov- ered from six other contexts, GL8–13, and although GL1. 1fragment of pushed-in base with pointed difficult to date accurately, they areprobably of kick from aplain squat beaker or moiolo. similar date to the painted examples. Clear fine glass with heavy weathering.Base diameter 35 mm. 15th-early 16th century. OXRA86(507). Glass post-dating 1536 GL2. 1fragment of stub base with external pontil Fragments from sevendifferent vessels of early post- mark from ahanging lamp. Green potasso- medieval date were recovered. Some,ifnot all, relate calcic glass with extremeweathering, to the immediate post-Dissolution use of the site. The totally devitrified. 13th–15th century. earliest fragments, GL14–15, are portions of rim and OXRA86(501). foldedbase from 16th-century pedestal goblets. GL3. 1fragment of everted thick rim from aflask. These are made in atinted, but clear glass, and are Green potasso-calcic glass with extreme decorated with optic-blown moulding. Pedestal weathering, totally devitrified. 13th–15th gobletssuch these are relatively uncommon, and century. OXRA94.11(1830). tend to date from the early to mid 16th century GL4. 4fragmentsofthick body, probablyfrom (Willmott 2002, 69). Of slightly later date and rather aflask. Green potasso-calcic glasswithex- more common are the fragmentsofbowl and base treme weathering, totally devitrified. 13th– from aclear glass goblet madewith aseparate knop 15th century. OXRA86(264). stem, GL16. Although the stem is now missing, so GL5. 1fragment of window glass. Paintedwith its precise form cannot be reconstructed, the style stickwork border and foliage design of suggests it is late 16th century in date. uncertain type. Green potasso-calcic glass Two beakers were also recovered from the exca- with heavy weathering.Late13th–14th cen- vations, of slightlydifferent forms. The first, GL17, tury. OXRA86(5/A/2) is aportion of foldedfoot-ring from atall pedes- GL6. 4joining fragmentsofwindow glass. tal variety and is late 16th-century in date. The Painted with avery fine stickwork border second,GL18, dates from the 17th century and is a containing scratched roundels. Green potas- squat tumbler shape, decorated with optic-blown so-calcic glass with heavy weathering. 13th– bosses. This type, madeinagood clear soda-rich 14th century. OXRA94(1810). glass, is commonly thought to be an Italianimport, GL7. 1fragment of window glass. Paintedwith and there is no reason for this not to be the casewith aborder and sectionofgrisaille cross- the Rewleyexample (Willmott 2002, 44). hatching. Green potasso-calcic glass with The final two vessels are both containers. The first, some weathering. 13th–14th century. GL19, is the evertedrim from asmall jar decorated OXRA86(902) with optic-blown wrythen ribbing, atype well- GL8. 5fragmentsofplain window glass.Green known on late 16th- and early 17th-centurysites. potasso-calcic glass with heavy weathering. The second,GL20, is the base from asmall octagonal 12th-early 16th century.OXRA86 (5) bottle, atypical Englishproduct of the late 16th GL9. 1fragment of plain windowglass.Green century. It would have been usedfor the semi- potasso-calcic glass with heavy weathering. permanent storage of liquids or medicines (Willmott 12th-early 16th century.OXRA86 (6) 2002, 87–9). GL10. 1fragment of plain windowglass.Green potasso-calcic glass with heavy weathering. 12th-early 16th century.OXRA86 (6/A/1) GL11. 1fragment of plain windowglass.Green Conclusions potasso-calcic glass with heavy weathering. Although only relatively few fragmentswere re- 12th-early 16th century.OXRA86 (7/A/1) covered that relate to the monastic phase of the site, GL12. 1fragment of plain windowglass.Green they are typical of what might be expectedforamonas- potasso-calcic glass with heavy weathering. tic house.The vessel glass with its range of table and 12th-early 16th century.OXRA86 (47) storagewares reflects large assemblages from GL13. 7fragmentsofplain window glass.Green sites such as BattleAbbey. Likewise the grisaille win- potasso-calcic glass with heavy weathering. dow glass, which seems to be entirely 13th century in 12th-early 16th century.OXRA86 (260) date, is atypical glazing schemefor aCistercian house GL14. 1fragment of vertical rim from apedestal that was wealthy enoughtopurchase such luxuries. goblet,decorated with optic-blown mould- The early post-medieval assemblage, although again ing. Clear glass with very little weathering. small,isstandard forthis type of site. Whether the Rim diameter uncertain. 16th century. material derivedfrom 16th-century occupation on or OXRA81I(6/1). close to the excavated area, is uncertain.However, GL15. 1fragment of foldedbase-ring from a the range of bothtable and storagewares is repre- pedestal goblet,decorated with optic-blown sentative of areasonably wealthy household. moulding. Clear glass with littleweathering.

58 OA Occasional Paper No: 16

Base diameter uncertain.16th century. OXRA86(18/A/1). GL16. 3fragmentsofflaring base and plain lower bowl from agoblet. Clear glass with medium weathering. Base diameter58mm. Late 16th- early 17th century. OXRA81 I(11). GL17. 1fragment of foldedbase-ring from aplain pedestal beaker. Green potasso-calcic glass with heavy weathering. Base diameteruncer- tain. Late16th century. OXSTRA94.11 (2512). GL18. 1fragment of pushed-in base from asquat beaker, decorated with optic-blown bosses. Clear glass with very little weathering. Base diameter 55 mm. 17th century. OXRA81 Plate 12 Architectural stone. I(11/1). GL19. 1fragment of everted rim and upper should from ajar decorated with optic-blown wry- Three fragments of building stoneappear to be then ribbing. Green clear glass with quite Corallian limestone, another material that was widely heavy weathering. Rim diameter85mm. used for building in medieval Oxford. It is known that Late 16th-early 17th century. OXRA86 between 1290 and 1377 it was an important source of (814/A). supplytoMerton College (Arkell 1947, 37), and it GL20. 1fragment of base fromanoctagonal bottle. seems likely that limestone from Wheatley would Green glass with quite heavy weathering. have beenused foratleast some of the abbey Base diameter50mm. Late 16th-early 17th buildings. However, by the end of the 14th century century. OXRA86 216/B/1. Headington stone had come into use as the main building stone for the Oxford colleges (Arkell 1947, 39), so this too may have been used at the abbey for Building stone any later medieval construction work. The three frag- ments of limestone that have survived may have come Fiona Roe, with anote on the architectural stone from either of these sources. Oolitic limestone is by Julian Munby suitable for use on architectural details such as All eleven pieces of building stoneare fragmentary mouldings and the one workedfragment of this and most have also been burnt. Good quality build- (500) is an angled piece that retains traces of diagonal ing stonefromthe abbey would have beenre-used tooling. Another fragment of oolitic limestone is and after the dissolution of the abbey in 1536 the discoloured fromburning (1006).Acoarser-grained new owners,Christ Church,began to sell the stone- and more shelly varietyoflimestone was re-used for a work for new buildings (see above). Although parts pierced, circular object (324) which couldperhaps of the abbey were still standing in the 19th century, have beenusedasacandle holder. these too have now gone. The excavatedpieces of Two further burnt fragmentsoffiner-grained, shell stone are all poor quality fragmentsunsuitable for fragmental limestoneare probablyfragments of re-use.Noneisclosely dated but it is probably a roofingtile (326) and these may have comefrom a reasonable assumption that they originate from the quarried areaknown as the Slat Pit, nearBuckland abbey buildings,since the area was not subsequently (Arkell 1947, 150). They came from alayer of 14th- to built over to any greatextent. Two fragments(232 15th-century date and so providemore certain and 500) are in fact from possiblepost-abbeycul- evidence for the medieval use of stone roofingtiles. tivation.Thereare three different varieties of build- Few architectural stone fragmentsofany signifi- ing stone, but all are from local Corallian beds. cance were recovered from the excavations. The single Six fragments, all burnt, can be identified as Lower item that has beenillustrated (Plate 12) is alarge Calcareous Grit, aJurassic sandstone. This calcareous limestone springerfor an arch or window, with roll sandstone had been used in Oxford as abuilding mouldings and chamfer. Probably13th–14th century. material since at least Norman times and can be found, together with the overlyingshell bed and Coral Rag, in the Castle Tower and the City Wall (Arkell 1947, 33). It Other finds has been suggested that the Lower Calcareous Grit Metalwork by Ian Scott, worked bone by Rosemary Grant, was obtained in North Hinksey (Arkell 1947, 35), only whetstone by Fiona Roe and flintbyRebecca Devaney about 8kmfrom Rewley Abbeyand with access by river as well as probably by road.This therefore would The assemblage of finds from this site is limited in have beenagood optionasabuilding stone. It seems number and range of objects (Table 7). Not one likely that the samecombination of materials as those object in the assemblage could be identified as used for the Castle Tower and the City Wall could certainly medieval, with the possible exception of a have beenusedfor the construction of the abbey. flesh hook fragment; asmall number are probably

59 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Table 7Quantification of small finds by functional category.

Material Function Total

Tools Household Personal Token Window StructuralNails Misc Unident Waste

Copper alloy 2416215 Iron 25 2262 12 5288 Iron and copper alloy 11 Lead 17 2111 Stone 11 Bone 11 2

TOTAL 38 6172262 20 63316

early post-medieval,but most of the objects that can thicker than the other. Theknife handle is very be dated are more recent. In terms of numbers the smooth and highly polished and has arectangular nails dominatethe ironwork assemblage. The mini- section. Thetang extends along the entirelength of mum number of nails is 262, but in total there are at the handlesuggesting that it is amedieval or post- least 354 fragments. All the nails are hand made, but medieval example (Goodall 1993, fig. 87, no 768). otherwise not closely datable. There are two tools, a Comb. Incomplete. Bone.Two fragments of double knife with acurved antler handle(OXRA 86 Tr VII sided simple comb. Theremaining end has aconvex ctx 732 sf 96), and asmall file (OXRA86TrVIII ctx curve. The uncut piece in the middle is wide but tapers 810/2 sf 95), bothprobably late post-medieval in towards the broken end. Theteeth are slightly thicker date. The five household objects comprise acast iron on oneside of the comb than on the otherbut are all saucepan (OXSTRA 94 ctx 1614 sf 9) of 19th- or very fine. OXSTRA 96 Cxt:1659, SF: 33, L: 44 mm. early 20th-century date, two post-medieval table or Handle: Incomplete. Bone. Whittle tang knife dessert knives (OXRA86TrIIctx 241/Asf93& handle. Rectangular section. Appears that the tang OXSTRA 94 ctx 1614 sf 10), awhittle tang knifeof extends along almost the entire length of the handle, uncertain form (OXSTRA 94 ctx 1970 sf 19) and a suggesting thismight be alate example.OXSTRA 96 possible flesh hook (OXRA86TrIctx 9sf111). Cxt: 2414 SF: 34 L: 42 mm. The latter could be medieval or post-medieval.The The whetstone (OXSTRA 94 ctx 1970 sf 35, structural items otherthan nails are aT-staple and a Plate 13) was retrieved from adeposit of 13th- to modern chainand fixing bolt. Theremaining objects are miscellaneous fragmentsand unidentifiable objects of uncertain function (n ¼ 5). The copper alloy objects include two household items, aprob- able drape ring (OXRR 93 ctx 134sf1)and amodern decorative lion head roundel (OXSTRA 94 ctx 2512 sf 26). Thereare four personal items:two pins (OXRA 86 Tr VIII ctx 813 sf 98 &OXSTRA 94 ctx 2514 sf 20), alace tag fragment (OXRA 86 Tr VIII ctx 813 sf 104) and acast button (OXRA 86 Tr VIII ctx 813 sf 105). The pins and button are later post-medieval and the lace tag early post-medieval. There is also a possible token (OXRA 86 Tr Xctx 1010 sf 100), again post-medieval. The lead comprises apossible lead stylus (OXSTRA 94 ctx 1836 sf 6), and seven pieces of windowcame. There is also part of atoy umbrella made from two pieces of bonejoined by athreaded iron pin (OXSTRA 96 ctx 10). The latter is of 19th- centuryorlater date. Two workedbone items, an incomplete comb (sf 33) and an incompleteknifehandle(sf 34) were recovered. The bone comb is of the double sided simple form. This form dates back to the Roman period and becamepopular again in the medieval period. This example has aslight convex curve on its remaining end suggesting its date as 17th-century or later (Galloway 1990, fig. 185, no. 2185). Both sets of teeth are very fine with one side being slightly Plate 13 Whetstone.

60 OA Occasional Paper No: 16

14th-century date in the fill of ditch 1999 associated Basset, Justiciar of England (d.1271). She was buried with the construction of the barn to the west of the at Oseneyon9February 1298 beneath astone abbey moat. One end is nowmissingfrom this with amemorial brass.51 Withher second husband whetstone but it has two well worn, concave she was abenefactress of Merton College, and when surfaces. Macroscopic examination shows that it she eventually retired to Godstow Nunnery we find was made fromafairly fine-grained sandstone the college building achamber for her and sending a speckled with small areas of iron staining. The stone boat to Godstow with preserved gingerfor her. 52 In may have beenobtained from local Corallianrocks returnshe gave the college money fordrinks on the such as the Lower Calcareous Grit or perhaps other day theysaid amassfor her.She also founded aloan fine-grained Jurassic sandstone like that occurringin chest for poor students (the Warwick Chest), and the Pusey Flags. Whetstonesmade from comparable helped the Friarsofthe Sack get established in sandstone have not often been recorded, though one Oxford.53 The Rewley inscription showsthat she built came from ageneralised Anglo-Saxon context at achapel at Rewley, and according to Leland she also Eynsham Abbey (Roe 2003, 291). Perhaps they were gave land to the abbey.54 used eitherasasubstitute, whenwhetstones of importedmaterials were not available, or just for specific purposes. ARMORIAL SHIELDS Asmall assemblage of flint consists of 13 frag- The armorial shields from the inner gate at Rewley, ments of burnt, unworked flint and two flakes. One and shown in situ on MichaelBurghers engraving of these may have beenablade removal when of 1720, were saved at the time of the coming of complete. Neither flake is datable. the railway and taken to Christ Church Cathedral where they are nowmounted on the walls in the ChapterHouse. They are in the form of sunk panels Previous Finds with angels holding shields of (a) the imperial Julian Munby double-headed eagle, and (b) the lion of Richardof 55 Anumber of important antiquities relating to the Cornwall. abbey and recovered from previous episodes of acti- vity have been preserved, and are here described. THE ABBEY SEAL The abbey seal was engraved for William Huddes- DEDICATION STONE (Plate 1) ford’s Life of Hearne, from asingleimpression of In July 1705 ThomasHearne notedthat an inscrip- 1431.56 It shows ascene of the ?annunciation in the tion had been found at Rewley, 47 and he later gave central panel, above aschoolroomscene of an abbot an accountofthe ‘stone dug up at Rewley in the with fourteen scholars; in the sides are the arms of the Ground, where the Chappel of the Abbeyformerly Empire and Cornwall, as above. The inscription reads: was. This Stone Ipurchas’d of Mr. Coxwho lives SIGILL(UM)COM(M)UNE. ABBATIS.ET. now in the house which is partofthe Abbey, and I CONVENT(US)DEREGALI. LOCO. have since put it intothe Anatomy Schoole Adjoyn- ing to the Publick Library.’48 This was one of several public collections of antiquities in the University, THE YARNTON GLASS housedinwhatisnow the Lower Reading Room of Afragment of medieval glass in the great collection the BodleianLibrary. Fr. JeromeBertram has provi- 49 at Yarnton church includes two early fifteenth- ded atranscription and translation. centuryportraits of Cistercian monks, which may Building inscription, now in Ashmolean Museum. reasonably be associated with Rewley Abbeyaslord 35 · 42 cm;brown sandstone; Lombardic/Roman- of the manor.57 esquelettering. Inscription in roughly cut lettering. þ ELA:LONGESP(EE)./COMIT(ISSA) WAREW (IKIENSIS)/H(A)NC: CAPELLAM: / HUMAN SKELETAL REMAINS FECIT: C(UIUS):P(RE)MIV(M): SIT: XP(ISTU)C: Georgina Slater and Louise Loe IN: GL(ORI)A: AM(EN)/þ Two skeletons were excavatedfromthe abbey Ella Longspe´ e, Countess of Warwick, built this church, one from grave 1103 in the middle of the chapel;may Christ be her reward in glory, Amen. nave, and the other from the south aisle. Disarticu- lated material, believed to be from an earlier burial, Rewley Abbeywas dedicated on 3December 1281, was also recovered from grave 1103. Additional which seems the most plausible datefor this inscrip- disarticulated material includes the remains of tion. Theslab was found on the Rewley site in 1705, one relatively complete disturbed burial and one and rescuedbyThomas Hearne.50 disturbed partialskeleton, both recovered from post- The connexion of the Countess with Rewley is not medieval robber trenches (900 and 1100). Limited withoutinterest. Ela Longspee was adaughter of the disarticulated material derives from Trench 1986/11 Earl of Salisbury, andwife of Thomas Earl of and is of unknown date. The preservation of all of the Warwick (d.1242), and then in 1254 marriedPhilip skeletal material was good; the outer surfaces of

61 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford the bones and joint surfaces had survivedintact and represents an inflammatory response, although the there was limited post-mortem erosion.However, all cause is unknown. Inflammation involving the of the material was fragmentary. Standard anthro- sphenoid bones may be as aresult of scurvy (Ortner pological and palaeopathological examination was and Ericksen1997). Scurvy, or vitamin Cdeficiency, undertaken in accordance with published guidelines causeshaemorrhage of the skin, gums, joint spaces (Brickley and Mckinley 2004), and afuller version of and subperiostealregions (Zimmerman and Kelley this summary report can be found in the project 1982). This can trigger askeletal responseinwhich archive. sub-periostealnew boneisformed and periodontal disease (a condition whereby the alveolar boneis resorped and the tooth roots are exposed) is common Results due to bleeding gums. In the presentexample, Skeletal remains from grave 1104 (Plate 6) although ante mortem tooth loss is also present, the changes are notenoughontheir own to confirm This contained the remainsofaskeleton buried in an adiagnosisofscurvy. extended, supine position within the south aisle of the church. Other evidence forfunerary treatment, Skeletal remains from grave 1103 such as acoffin, was not identified. The skeleton was approximately 15% complete This includes onediscrete skeleton and disarticu- and was represented by fragmentsofcranium, lated remains. The discrete skeleton was excavated mandible and ribs, and the first two vertebrae (atlas from ashallow grave that lay partially beyond the and axis) of the spine. The missing elements of this limit of the excavation. No associated burial rites skeleton had beenleft in situ owing to the fact were identified, other than the fact that the indivi- that they lay beyond the limitsofthe excavation. dual had been placedinthe grave in an extended The remainswere undoubtedly thoseofanadult, supine position. although no indicators that would allowamore pre- The skeleton was approximately 20% complete cise age estimationhad survived. Basedonthe and was represented by bones from the forearm, morphology of certain landmarks of the skull and right hand, left leg and foot, all of varying degrees of the shape of the mandible, it was concluded that the completeness.Itwas not possible to estimate the remainsrepresent thoseofafemale. preciseage of these remains, other than to say that Dental conditions include aslight deposit of they representanadult. The surviving bones did not calculus (tartar) on one mandibular canine and a permitthe estimation of biological sex. carious cavitythat involved amaxillary premolar. In Changes that are consistent with non-specific bone addition four teeth, one premolar and three molars, inflammation(periostitis) were present on the distal had been lost before death.Calculus is formedby third of the left femur. Measuring 17.7 mm wide and the mineralisation of organic material and bacteria 25 mm long, the lesion appeared as awell demar- and, as such, reflects the lack of importance (or per- cated, smooth deposit of new bone. Trauma or haps inability owing to illness) given to maintaining infection are likely to have caused these changes. healthyteeth.The prevalence of calculus in the The remainsofanother left leg bone(distal femur) medieval period was higher than otherperiods and were also present and comprise the disarticulated its occurrence was associated with asoft diet (Stroud material from this context. These represent an adult and Kemp 1993). Caries, an infectiousand stagnating of unknown sex. disease,results in the destruction of the enamel, dentine andcement of the tooth. It ariseswhen Skeletal remains from robber trench 900 bacteria harbouring calculus convert carbohydrates into acid that subsequentlyundermines the mineral The remainsfrom this context comprise one dis- matrix of the tooth. Rates of caries in the past show articulated individual, approximately 65% complete asharp rise in the later medieval period, which and represented by differentially preservedbones coincides with the introduction of sugar into the diet of the axial skeleton, the upper and lower appen- (Moore and Corbett 1971). During the early medieval dages and the jaw. The wide angleofthe sciatic period cane and beet sugar were unknown, honey notch of the pelvis, backed up by the size of the being the only sweetening agent at thistime (Roberts femoral and humeral heads, the shape of the man- and Cox 2003,190). dible and the overall gracile appearance of the skele- Ante-mortem loss may arise as aresult of several ton, strongly indicatedthat the remains were those factors including abscess developmentsecondary to of afemale. The appearanceofthe auricularsurfaces caries, periodontaldisease secondary to calculus of the pelvis suggests that she was between 30 and formation, and pulp exposure and abscess formation 40 years of ageand, based on the maximum length secondary to severe attrition. Any oneofthese factors of the ulna, it was estimatedthat she was 157 cm tall may have resulted in the tooth loss observedinthe (5 ft 15 inches). presentskeleton, although the absence of avisible Hypoplasia, in the form of linear grooves of abscess suggests that this is an unlikely explanation. depressed enamel, was present and involvedtwo Smooth remodelled new bonewas presentona teeth,one canine and one premolar. These changes fragment of left sphenoid boneofthe cranium.This would have arisen during the individual’s childhood

62 OA Occasional Paper No: 16 when the enamel was developing as aresult of non- height was notcompromisedbychildhoodhealth specific health stress. Factors that may cause thiscan stress as this falls within the expectedrange for this broadly be classified as those that relate to childhood period (Robertsand Cox 2003). The changesob- illness (eg measles) and thosethat relate to nutri- served on the remainsfrom grave 1104 are perhaps tionaldeficiency. the most intriguing since they may have been caused Slight calculus and carious cavitieswerealso by scurvy. This need not, however, necessarily have present, the latter involving one premolar. Ante- been the cause of the changesobservedonthis mortemtooth losswas extensive and involvedseven individual, whoseremains are too incomplete to ex- teeth.Incomplete resorption of alveolar bone, plore this further. Owingtothese factors this, and coupledwithspiculated, active new bone, indicated otherpoints raised here, should be regarded with that these teeth had been lostnot long before death caution. and that this had resulted in severe inflammation, possibly infection (although no sinuses were pre- ANIMAL BONE sent). Thecause is unknown, although abscess Lena Strid formation is unlikely(seeabove). Sinusitisisanother pathological condition that In accordance with the agreed research aims for was observed on thisskeleton. Sinusitisisbelieved to post-excavation (OA 2005, 6), study of the animal result from inflammationofthe mucous membrane bone assemblages was undertaken in order to shed and is diagnosed by the presence of new boneonthe light on diet and agriculture at the medieval abbey. maxillary antra (Boocock et al.1995). Upper respira- In the event, the quantity of data available has tory tract infections, poorliving conditions, environ- proved insufficient to support detailed analysis. The mental pollution,congenital abnormalities, dental full report, by Lena Strid, can be foundinthe pro- disease and specific infectiousdiseasessuch as ject archive.The following account provides agene- tuberculosis and leprosy are among the aetiological ral summary of the material that was present in factors associated with thiscondition (Lewis 2002, the pre-abbey and abbey phases of occupation of the 21). In the present example,the changes were pre- site. The limited size of these assemblages undoub- sent as smooth, non active deposits of new boneon tedly reflects the type of features on which the the left maxillary sinus. Sinusitisarising from an excavation strategy was targeted. However, Cister- abscess involving the maxillary molars can be ruled cian monks were forbiddentoeat meat unless they out here since this skeleton showednoevidence for were ill, and theywere expected to followalargely adental abscess. vegetarian diet, with fish and eggs as occasional treats. The assemblage from the pre-abbeyand abbey Other human remains from robber trench 1100 phases consistedof549 fragments, of which 154 The disturbed remains of incomplete leg, foot and (34.5%)couldbedeterminedtospecies.Ofthese, 446 arm bones were present in robber trench 1100. The bones were from contexts predating the foundation remainsrepresent at least one adultwith no diagnos- of the abbey, of which some 25% could be identi- tic features that would allow amore preciseage fied to species.Unfortunately the provenance of this or sex to be estimated.Noskeletal pathology was material is veryuncertain.While some of it may identified. The remains from Trench1986/11 were have derived from activitiesatthe site itself,it very limited, including fragments of rib only, prob- is likely that much of the bone was brought in ably adult. Neither amore accurate age nor sex with dumpedsoil to raise the ground level for cons- could be estimated and no skeletal pathology was truction and couldhave originated more or less present. anywhere in the medieval town. Allthree major domesticates, cattle, sheep/goat and pig,were pre- sent, with sheep predominant according to number Comment of bones. The19identifiable bird bones included Asmall, fragmentary assemblage such as thisoffers fowl, goose,duck and two bones possiblyofcoot, a limited potential in terms of what can be learnt about speciesthat would have been living on the river the formerpopulation associatedwith the abbey. All nearby. of these individualshad survivedinto adulthood, The abbey period assemblage was even smaller, although the presence of enamel hypoplasia suggests and comprised 93 bones, of which 37% were identifi- that, forsome,thiswas not without the experience of able to species. As before,the three main domesticates health stress during childhood.Sinusitis, identified were identified. Both meat-rich and meat-poor ele- here as bone inflammation, is associatedwithpoor ments of cattle and sheep/goat were present, suggest- environmental conditions and was also present. This ing that wholeanimals were butchered on the abbey also implies non-specific health stress. Theimpact of site itself,although not necessarily slaughtered there health stress on the population may not, however, as complete carcasses could have been acquired from have beengreat, if the estimated stature is anything the town. Onesheep bonehad skinning cutmarks. to go by. This stature was estimated by employing One horsebonewas presentinthe assemblage, and the measurementofone of the least reliable long four bird bones, which included both domesticfowl bones (Trotter, 1970). It indicates that final attained and goose.

63 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

THE REWLEY ROAD STATION:THE COMING the Crystal Palace dismantledand disposed of as its OF THE RAILWAY owners saw fit. The popularity which the building Ric Tyler, with contributions by Lance Adlam and had experienced, however, encouraged Paxton to set David Higgins up the Crystal Palace Company and secure over £500,000 of privatefunding to finance the purchase, Introduction dismantling and re-erection of the exhibition build- The historical importance of the L&NWR(subse- ing on amuch-enlargedscale at anew site at quentlyLMS) railway station at Rewley Road (Plate Sydenham Hill in south-east London. Here it remai- 14), based upon its close association with Joseph ned for the next82years until it was sadly destroyed Paxton’s ‘Crystal Palace’structure built in HydePark by fireon30November 1936. Since its destruction, for the ‘Great Exhibition’ of 1851, has long been recog- the Rewley Road station at Oxford, built ‘on the nised.The Crystal Palace was aremarkable example plan’ of the exhibition building and using asimilar of contemporary architecture in glass and iron, acele- range of constructional elements, has provided the brationofthe achievements of British innovation and last representativeexample of the technology asso- technological prowess.Designedand erected in only ciated with this highly significant building.In nine months, it was characterised by its use of pre- recognition of this importance, the station was fabricated modular construction, and thus occupies included in 1985 as Grade II* on the List of Buildings an important place in the historyofarchitectural and of Special Architectural or Historic Interest. technological developmentinthe mid 19th century. The dismantling and removal of the station build- Its standardised, modularconstruction represented a ing from its Rewley Road site has provided, for the pioneer model for later ‘system building’, whilethe first time, the opportunityfor adetailed study of the sheer scale of the undertaking and the industrialised structure to be made and for the nature of the rela- nature of the mass production of its structural com- tionship with the Crystal Palace to be more properly ponents servedtoheighten its significance.58 assessed. An examination of the records of the With the closureofthe Great Exhibition aftersix Buckinghamshire Railway has further illuminated months in October1851, it had been agreed that the connection between the two buildings, long Hyde Park was to be returned to its original state, acknowledged but never fully understood. Areview

Plate 14 The Rewley RoadL&NWR station from the south, 1914. HenryWTaunt Coll. (CC54/00324) q Oxfordshire CountyCouncil.

64 OA Occasional Paper No: 16 of archive plans and historical photographsofthe was agreed at ameeting of the Board of the station has proved useful in tracing later alterations to Buckingham Railway in July 1850 and Robert the building.However, as will be seen in the following BensonDockray,principal engineer with the discussion, the sometimes ambiguous datingofthese L&NWR, was invited to draw up draft plans for sourceshas madethe assigning of exact dates to the station buildings, comprising locomotive shed, certain,specific interventions unfeasible and we are goods shed, watertower andweigh officeinconsul- left with aloose chronology of change as opposed to a tationwith the L&NWR’s general manager, Captain closely defined historical development. Mark Huish. At ameeting of the Boardon7November 1850, Dockray’sdesigns, along conventional lines in an Historical Background eclectic style combining gothic and classical ele- The coming of the railway to Oxford presents a ments,were presented. They were accepted and it somewhat complexhistory, reflecting contemporary was ordered that detailed contractplans be drawn rivalriesbetween anumber of competing indepen- up based upon them, with certain modifications dent railway companies.The railway first reached including the transposition of the passenger and the city in 1844, when the Great Western Railway goods sheds so that the formerlay nearer to the (GWR) completed abroad-gauge line from Didcot to town. Asum of £7,000 was allocatedfor the atemporarystation located south of the Thamesin building, which, it was resolved, should ‘be made Western Road, Grandpont.59 TheGWR soon began and placedsoastobecapable of Extension without work on extending this line via Rugby to Birming- any demolitions’.60 ham. However, in direct competition with the GWR, Dockray’sfinalplans were presentedtothe Board the Londonand Birmingham Company together on 12 December 1850, and it was agreed that ‘ ... with its allies,the GrandJunction Railway and the tenders be obtained for the Oxford Station on the Manchester and Birmingham Railway, which mer- plans proposed for the alternatives of Wood and ged in 1847 to form the Londonand North Western Stone for the front buildings’.61 Interestingly, at this Railway (L&NWR), proposed the creation of the late stageinthe process,the Board also invited alter- Buckinghamshire Railway. This locally-backed native plans to be submitted by the engineering part- schemecomprised aseries of linesthroughout that nership of Fox,Henderson and Company ‘for the countyserving Aylesbury, Banbury, Bicester, and whole erection on the plan of the exhibition build- Buckingham and linking Oxford to the L&B Euston ing, in all respects, as informationfor the Board’.62 to Birmingham line at Bletchley. Theproposal was Given this proposed association with the Great authorised by the Buckingham Railway Act of 1847 Exhibition, it is notaltogether surprising that Charles (Fig. 24). Fox and John Henderson were invited to tender for From the start, it would appear that the company the Oxford terminus building. They had been, from intended to associate its new line with the opening of the earliest tendering stage, heavily involved in the the Great Exhibition at the Crystal Palace in Hyde construction of the Crystal Palace itself,generally Park,Londonon1May 1851, aproject which had considered ajoint venture between Fox and Joseph been the subject of muchpublicity and had captured Paxton, whileFox (a formerpupil of Robert the imagination of the general public. It was pro- Stevenson on the Londonand Birmingham Railway) posed to synchronise the openingdates and to run could lay claim to an extensive portfolio of work for cheap excursions to the royal opening of the exhi- the L&NWRincluding the roof of the original Euston bition from the new terminus. terminus in 1837. The L&NWRitself had been invol- While work on the new Oxford line progressed ved in the transportation of materials from foundries apace, reaching Islip by October 1850, and, with in Birmingham to the capital for the construction of atemporarystation openingat‘Oxford Road’ the Crystal Palace. (Banbury Road, Oxford) in December of the same Atotal of 18 tenders for the work were submitted year,the establishment of apermanent terminal based uponthe Dockray design, and were examined proved to be arather more involvedprocess. Various by the Board at ameeting on 9th January 1851. proposals were put forward for the final approach of Tenders rangedincost from £6,904 to £12,330 against the line,within the limits of deviation defined by the asurveyor’s estimate of £8,000. In the event it was Fox Act. However, giventhe existence of the GWR line, Henderson’s tender, on the altogether different de- options were restricted and, in the end, the Buck- sign, that provedtobethe lowest at £6,552, inclusive of inghamRailway had littlechoice but to run into 12 months maintenance and of apparatus to warm the Oxford on the east side of the GWR’s designated offices with hot water. TheBoardcame to an imme- route. This solution entailedcrossing the Sheepwash diate decision and ‘resolvedthat the tender of Fox Channel, anavigable link between the River Thames Henderson and Coy be accepted as follows:- the and the Oxford Canal, at alow level, necessitating whole works specifiedtobedone for £6,552 including the construction of atwo rail swing-bridge (rebuilt in 12 months maintenance or if corrugated iron be used 1890, still extant),and locating the terminus some- £31 more. Any alterations as to the strength or detail, where nearthe formersite of Rewley Abbeyinthe which Mr Dockray may consider necessary to be low-lying fields of North Osneyfronting onto Park made at his request withoutincreased cost to the End Street in the west of the city (Fig. 25). The site Company. The whole to be completed ready for use in

65 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 24 Schematicmap of rail routes around Oxford at the end of the 19th century (after Waters 1986).

3months from January 16th instant.’63 Regarding the The Rewley Road Station was officially opened three preciselocationofthe building ‘ ...CaptainHuish was weeks later on Tuesday 20 May, 1851 amongstmuch requested to confer with the Engineer and on the celebration as recorded in Jackson’s Oxford Journal groundtoset out the Station in such asituation as may on 31 May: ‘The new station, which is constructed in be most expedient.’64 asimilar manner to the Crystal Palace in Hyde Park, There seems to have been some delay in obtaining was decorated with flags, the bells of St Thomas’s the land for the station, as in February the powers to Church poured forth amerry peal, and the rattling of concludethe purchases were delegated and, as late the various carriages, and the great concourse of as 10 April 1851, the Board learnedthat Christ peopleassembled at and aroundthe station,created Church ‘declined to part with the property at ascene of bustle and excitement which has been Oxford’.65 At the samemeeting, the Directors were hitherto unknown in thislocality.’ informed of the probability that the line would not be ready for traffic on 1May and, although they ordered‘every exertion be madebythe engineers to The station as built, 1851 secure its being openedonthat date’,itwas notto Preparation of the site be. On 1May the L&NWRwas forced to run its ‘CheapExcursion’ for the royal opening of the Prior to the construction of the railway, the site Great Exhibition from the Oxford Road Station. comprised relatively low-lying,undeveloped alluvial

66 OA Occasional Paper No: 16

Figure 25 Extract from RSHoggar’s 1850 map of Oxford indicating the proposed L&NWR terminus site. grassland and water meadow,atalevel of around The first components of the station building to be 56m OD, and was pronetoflooding. To the east, the erected were the c 2mlong, 8-inch (203 mm) dia- site was bounded by the remains of the earthen bank meter circular, cast-iron foundation columns (F) of the civil war defences,which would have stood no (Fig. 26). As theywere also to serve as downpipes more than c 0.5 mhigh at thistime. In order to avoid for the evacuation of rainwater from the finished any potentialproblemsofflooding, it was thus station roof,the columns were furnished with necessary to build the groundlevel up around the integrally-cast outflowpipes just abovetheir base proposed station site. The Geological Survey of for connection to ceramic drainage pipes feedinginto Great Britain (England and Wales) Sheet 236 (1982) brick-built culverts to the south and east of the site. clearly shows a100–200 mbroad swatheofmade With the below-ground ironwork and drainage groundbeginning c 900 mnorth of the station site at in place, foundation walls were constructed, using Medley Weir and continuing to the south of Park grey engineering brick laid in English bond, up End Street. An archaeological watching brief under- to proposed groundlevel on the linesofthe principal taken following the removal of the station building walls of the building, the retainingwalls for the train has confirmed the pre-station ground surface lay at a lines and the concourse. level of some 2mbelow the level of the finished Once the subterranean ironwork, drainage pipes station concourse. The necessityofraising the and brick footings were complete, the ground within groundlevel resulted in an interesting sequence of and aroundthe station could be raised to its required construction for the station building. final level. Without, the surface was raised by the Initial works involvedthe laying out of agrid dumping of c 1.5 mofimportedblue-grey clay, of foundation pads to support the principal uprights followed by layers of sand and limestone rubble, and of the iron framework. Aseries of pits, c 0.65 m finally asand beddingfor the granite cobble setts square · 0.4 mdeep, were cut intothe alluvial surface of the station yard. Within the area of the concourse, and were filled with concrete to form solid founda- the ground level was madeupwith dumped build- tions for the verticalelements of the station structure. ing rubble material priortothe laying of aYork

67 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 26 Interpretative reconstruction of foundation column (F) at the south-west corner of the porte-coche` re, investigated in 1995. (J. Hiller, Oxford Archaeology)

flagstonefloor. Thetimber floors of the east and west concourse/trainshed had aspan of 48 ft (14.64 m), 1 flankingwings were suspendedover substantial stood 17 ft 10 ⁄2 in. (5.45 m) tall to eaves level, and voids,supported uponpiers of Oxfordshire red originally comprised 13 bays, each 24ft (7.32 m) in brick. width with afurther, single-bay, two-part porte- coche` re to the front elevation. 67 The main structure thus had an overalllength of 336ft (102.4 m) The ironworksuperstructure66 and an overall width of 96ft (29.26 m) including the Thestation as built comprised acentral,single-spancon- side wings, which each measured48ft (N–S) · 24ft course/trainshed alignedapproximately north-south, (E–W). Thebuilding’s superstructure was of modu- flanked to east andwest by single-storey wings. The lar construction,utilisingboth cast-iron and timber

68 C2b; and C2a columns (c) C1-C3; columns (b) F; column foundation (a) structure: primary of E. components piece vertical of extension (e) Details C4; 27 column Figure (d)

69 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford T1. truss (b) B1; side-beam (a) structure: primary of components horizontal of Details 28 Figure

70 Figure 29 (a) Schematic layout of structuralcomponents; (b) plan view of principal walling components, as recorded in 1999. Based on asurvey by Colin Wide. From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 30 Details of cast-iron column base mouldings.

3 structural elements, summarised below in tabulated columnsection, and with a ⁄4 -inch thick upper and form (Table 8onp.74) and illustrated in Figures27– 2-inch thick lower flange. The union details between 28. Aschematic layout of the primaryelements is columnand extensionpiece varied. Within the shown in Figure 29. trainshed (Bays 4–13), the connection was relatively All cast-iron columns of the framework (C1–3) simple,mirroring the bottomdetail, with four bolts were madeessentially to the same basic design, passingthrough integrally-cast projecting lugs in the though varying in shaft-length and upper union column(C1) and the square flangeatthe base of the details according to their location within the building extension piece. Within the concourse (Bays 2–3), (Fig. 27b and c; Table 8). They were circular in cross- however, the detail was complicated by the presence section, 8-inch (203 mm) diameter, with four raised of large sockets with projecting flanges in the lateral square fillets. Theflat faces presented by these fillets faces of the columns (C2a and 2b), necessitating the provided adecorative featuretothe uprights, but elongation of the projecting lugs on the column to they also reflected the profileofthe upper extension some 16 in. (0.40 m) in length (Fig. 27c and 31). The pieces (Fig. 27e) where flat faceswere afunctional flanged sockets relate to aprimary, deep-flow valley necessity, allowing flush boltedconnectionsfor both gutter draining the original roofsofthe flanking longitudinal beams (B1) and transverse trusses (T1). wings by channelling water intothe columndown- The main column sections were joined to the sub- pipes,asystem which had become redundant by surface foundation columns (F)byboltedunions, the time of the reinforcement of the concourse roof in bolts running throughfour 2-inch deep, integrally c 1888 (see below). cast projecting lugs at the base of the column and at It is likely that the station would have been the top of the foundation column. Thebases of the constructed one bay at atime, most probablyfrom columns were visually enhanced, and the bolt-heads the south. Each pair of columns would be raised and hidden, by the use of decorative cast-iron bases securedtotheir corresponding foundation column, (Fig. 30). These base mouldings were of two designs; the extension pieces fitted and the primarytruss (T1) externally, they were square in plan, while in the insertedbetween them. The trusses were 47 ft 4in. 1 arcadesseparating the station concourse from (14.42 m) long · 3ft ⁄2 in. (0.93 m) deepand were the side wings, asix-sided form was employed. formedofpairedwrought-iron angles and flats, 1 Bases were also usedfor the intermediary timber riveted together with additional ⁄4 -inch flange plates, columns (C4). top and bottom(Fig. 28b). The trusses were divided At the head of each columnwithin the concourse into six panels along their length by elegantly and trainshed, an extension piece (Fig. 25e) was barrelled, cast iron, cruciform-profile uprights. Each 1 attached. These were standard castings, 3ft ⁄2 in. panel containedasingletensionbrace, angled (0.93 m) long, of identical cross-section to the main downwards towards the centreline, forming a‘Pratt’

72 OA Occasional Paper No: 16

Figure 31 Isometric projection of aC2b column head and extension piece (E), with primary truss T1 and secondary side-beams (B2). (Note redundant deep-valley gutter-flanges areblocked by wooden blocks.)

type truss. Theend standards of the trusses were furrow structure, on the model of the Crystal Palace, securedtothe columnextensionpieces by paired 1- comprising six 8ft(2.44 m) wide bays, running inch diameteriron bolts.Longitudinal bracing to the perpendicular to the primarytrusses (N–S), the structure was supplied by cast-iron, lattice edge- valleys aligning both with the verticalmembers of beams (B1) spanningbetween the principal columns. the trusses and with the iron and timber columns of 1 These beams were 23 ft 4in. (7.11 m) long · 3ft ⁄2 in. the south elevation. The form of this original roof is (0.93 m) deep, divided into three panels along their recorded in an early interior photograph of c 1900 length by cruciform-section uprights, each panel (Plate 16 on p. 76) and is evidenced structurally by containing integrally-cast, cruciform-section diagonal the presence of regularly-spaced,grouped bolt-holes bracing (Fig. 28a). As with the transverse trusses,the in the upper chords of primarytrusses(T1), recorded edge-beams were secured to the column extension during dismantling, representing connection points pieces by paired iron bolts, and were furnished with for the valleys. externally mounted, louvred ventilation panels The side wings were originally roofed with asimi- (V1–2:see Plate 15).68 lar glazed, ridge-and-furrow style structurethough The two-bay long side wings were similarly the alignment herewas transverse(E–W) as opposed framed using the shortercolumn C3, herewithout to longitudinal (N–S). These roof structures were the use of upper extension pieces, the upper flange of replaced in their entirety at an early stage (see the columns being bolted directlytoatimber wall below), though evidence for the original arrange- plate. ment was recorded during the pre-dismantling survey in the form of the tongue-and-groovecladd- ing of aprimaryinternal partition wall, retaining Primaryroofs the characteristic ridge-and-furrow profile, survi- The original roof of the main concourse/trainshed ving within the roof space of the west wing comprised alongitudinal, fullyglazedridge-and- (Plate 17).69

73 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Table 8Summary of primary framework components (see Figures27–28).

Ref Type Material Description

Structural Elements FFoundation Cast Iron 8in. (203 mm) diam. hollow,circular section foundation column associated Column with columns C1–3. Integrally-castdrainage offshoot. Upper flange with four projecting lugs for bolted connection to main column section. C1 Column Cast Iron 8in. (203 mm) diam. hollow, circular section column with square fillets, c 19 ft (5.79 m) long. Four 2-inch deep, integrally cast projecting lugs to top and bottom for connection to sections Fand E. Originally used throughout length of trainshed, four examples survive in Bay 5(Nand S). C2 Column Cast Iron 8in. (203 mm) diam., hollow, circular section column with square fillets (as C1); 14 ft 10 in. (4.52 m) long. As C1 except for shorter shaft length. Originally used throughout the porte-cochere, only one complete example survives at the south-west corner of the reduced structure (that to the south- east corner has been replaced).Central column of south elevation (truncated) was also originally of this type. C2a Column Cast Iron Variant of column C2; similar except for the inclusion of alarge square hole in the lateral face with 16 in. (0.4 m) deep integrally-cast flange. This detail originally housed aprimary valley gutter (removed) for the draining of the flanking wing roof, though it is now obsolete. Asimilar opening is present in the external face of the column, set some 4in. above the existing valley. Four examples survive, used at junction of concourse/trainshed (Bay 4, N) and concourse/porte-cochere (Bay 2, S). C2b Column Cast Iron Variant of column C2; identical to C2a except for additional lateral openings and deep gutter flanges in Nand Sfaces as well as that in the lateral face. As the columns no longer serve as downpipes, the lateral and transverse openings have been sealed by wooden blocks. Two examples survive in Bay 2, N. C3 Column Cast Iron 8in. (203 mm) diam, hollow,circular section column with square fillets (as C1); 14 ft 4in. (4.37 m) long. Essentially similar to C1/C2 though of shorter length. Used in exterior walls of east and west flanking wings. Six examples survive. C4 Column Timber Semi-circular section with square fillets; imitation of primary cast-iron columns columns (C2 &C3). Used @8ft centres between columns of the concoursearcades and the external walls, 27 examples survive though most of the exterior examples have been truncated by the introduction of brickwork plinths. EExtension Piece Cast Iron Standard casting, 8in. (203 mm) diam. hollow, circular section column with 1 square fillets · 3ft ⁄2 in. (0.93 m) long, used to extend columns C1 &2and provide connection for trusses and side beams, which are bolted to the extension pieces. Square flanges to top (3/4 in. thick) and bottom (2 in. thick) for connection to wooden gutter/cornice and main column section respectively. 1 T1 Truss Composite Cast/ 47 ft 4in. (14.42 m) long · 3ft ⁄2 in. (0.93 m) deep. Wrought-iron angles Wrought Iron and flats riveted together;cast-iron cruciform, barelled uprightsdivide truss into six panels. ‘Pratt’ type truss with single diagonal tension braces in each panel. Span concourse/trainshed between primary columns, four examples survive (Bays 2–5, N). 1 B1 Edge-Beam Cast Iron 23 ft 4in. (7.11 m) long · 3ft ⁄2 in. (7.11 m · 0.93 m) deep; trussed edge- beam spanning between primary columns. Divided into 3panels by four cruciform section uprights, each panel including integrally-cast diagonal cross-braces. Four examples survive, in bays 4and 5.

Decorative Elements A1 Decorative arch/ Cast Iron Cast iron, arched spandrels with integrally-cast decorative scrolls at spandrel connection points. Used in pairs to form arched decorationinthe south elevation of the concourseand in the sides of the shorter bay of the porte- coche` re (that to east has been replaced), spanning 7ft4in. (2.24 m) between adjacent iron and wooden columns, fixed with bolts. An additional integrally- cast decorative moulding serves to hide the join between paired castings.

74 OA Occasional Paper No: 16

Table 8(Continued)

Ref Type Material Description

1 P1 Decorative Panel Cast Iron Rectangular cast-iron panel, 7ft4in. (2.24 m) long · 3ft ⁄2 in. (0.93 m) deep with integrally-cast, central circular motif. Used above the southern concourseelevation and in the shorter bay of the side elevations of the porte-coche` re (that to the east elevation has been replaced). A2 Decorative arch/ Cast Iron Cast-ironarched spandrel, originally used singularly in the southern, wider spandrel bay of the porte-coche` re (removed); similar in most respects to A1, though lacking the integrally-cast mouldinghiding the join between paired elements, redundant in this location. 1 P2 Decorative Panel Cast Iron Rectangular cast-iron panel, 15 ft 4in. long · 3ft ⁄2 in. deep with integrally-cast, paired circular motifs. Originally used in the front and side elevationsofthe southern, wider bay of the porte-coche` re (removed).

Decorative elements the recreation of the Crystal Palace fac¸ade panel was almost complete (Fig. 33). The front elevation of the station concourse and the Early photographic sources(eg Plate 19) show that projecting porte-coche` re were visually enhancedby the eavesofthe station were decorated with dis- the use of cast-iron spandrels and decorative panels tinctive crests and the corners enhanced with tall with integral circle-motif at high level (Fig. 32b; Plate finials.These were again reproductionsofdecorative 18), representing exact copies of elements used elements employed at the Crystal Palace,and served extensively in the Crystal Palace. In the concourse to hide the structure of the roof. It is known that the elevation and the northern sectionofthe porte- crestsatthe Crystal Palacewere made of zinc and it coche` re, where spandrels were pairedtoformsemi- is probablethat the samematerial was usedat circulararches with the circular-motif panel above, Oxford, while it is possible that the station finials

Plate 15 Trainshed, Bay 5(W) before dismantling, 1998. Note inserted type C5 rail-built column, also the surviving type V2 louvre panel.

75 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 16 The earliestknown view of the station interior looking south from c 1900. The original longitudinal glazed roof is still in-situ, though the intermediary supports(1988) havebeen added. Note the oval booking office and the original low platform level. Photo: Lens of Sutton Collection.

Plate 17 Profileofthe original longitudinal ridge-and-furrow side-wing roof preserved in original internal vertical- boarded partition (west wing roof).

76 OA Occasional Paper No: 16 Wide. Colin by urvey as on Based reconstruction). (1914 elevation front b) ;( recorded as elevation Front (a) 32 Figure

77 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 33 Standard Crystal Palacetimber wall panel. (Downes and Cowper) were adapted fromgallery handrail balusters from Buckinghamshire Railway stations on both the the Crystal Palace (OA 1998, 18). Oxford and Banbury branch lines. Eachpanel span- ned between adjacent cast-iron columns,towhich they were affixedusing cast-iron ‘butterfly’fixings, Primarywalling and incorporated two intermediate,evenly spaced, At the timeofsurvey, the exterior walls of the timber columns (C4).Internally, the flankingwings primarystructurewere formed of studwork panels were linedwithvertical, reeded, tongue-and-groove (W1–3: see Table 9), which appearedtohave been boarding, whilethe concourse arcadesand the fabricated on site, clad externally with horizontal southern wall, facing onto the station concourse, ship-lap boarding, of atype familiar from many were clad in corrugated, galvanisediron sheets. Each

78 OA Occasional Paper No: 16

Plate 18 Detail of decorative spandrels and circle motif in the west elevation of the porte-coche` re, photographed prior to dismantling, 1998.

Plate 19 Photographof1883 showing the Oxford Volunteers marching to the GWR station. The porte-coche` re of the L&NWR station, with original decorative crests andfinial, is clearlyvisibletothe left hand side of the photograph. MS. Top.Oxon. d. 505, no.189 (Minn Coll. Neg. 6/122a) q Bodleian Library, University of Oxford.

79 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford Wide. Colin by urvey 0m as on Based 1:200 recorded. as elevation 01 west (b) east; looking cross-section Longitudinal (a) 34 Figure

80 OA Occasional Paper No: 16 by urvey as on Based south. looking side-wings and trainshed of elevation northern (b) south; looking side-wings and concourse of cross-section Transverse (a) 35 Wide. Figure Colin

81 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford bay of the concourse arcades originally included Similarities to and differences from the Crystal three round-headed double-doors with wooden arch- Palace structure itravesand glazedsemi-circular fanlights (Fig. 34), with the exception of Bay 3(W), which only Many elements of the primarystation building are included two. At least some of these openings essentially similar to the components illustrated in would appear always to have been blocked. contemporary published drawings of the Crystal While the flankingwalls of the concourse (Fig. 35) Palace (Downes and Cowper1852). Many of the are primary, close archaeological investigation has columncastings are the same in all but shaft length revealed that the exterior wallpanels had been whilethe edge-beams appear almost identical to the subject to major alterations contemporary with the standard 24 ft Crystal Palace floor beams. Apparent re-roofing of the side wings (see below) and it structural differences between the primary station appears that, in its original form, the station building trussesand their Crystal Palace counterparts, how- was fully clad externally with vertical ship-lap ever, have proven on closer examination to be boarding. This original cladding survivesonlyin illusory(Sutherland 1975, 69). Despite the outward isolated areas, where it was preserved behind later similarity of the components usedinthe two panelling. Where original internal partition walls buildings and the obvious use at Oxford of the survive,these are also clad in verticaltongue-and characteristic structural and, in particular, the grooveboarding. decorative motifs of the Crystal Palace,close inspec- tion has revealedthat very few members of the station building represent exact matches. While a Interior arrangements small number of elements may have beencast from In ascertainingthe original layoutand arrange- the Crystal Palace moulds, or may even represent mentsofthe interior of thestation,archaeological ‘spares’ from the HydePark structure, it is apparent evidenceisaugmented by thesurvivalofseveral that considerable reworking of the foundry patterns, plansofthe building from the19thand early if not acompletely new second set, would have been 20thcenturies, andanumber of earlyphotographs. necessary to suit the particular form of the Rewley TheFirst EditionOrdnance Survey1:500planof Road station.Certain components appear unique to thestation of 1876 (Fig.36) would appear to show the station,for example columns C2a, C2b and C3 thebuildingmuchasbuiltwiththe exception of the with their complex upper flangedetailing relatedto northern extensions to thesidewings (see below). the primaryroof drainagesystem. They were unlike Theopenstation concourse wasconfined to Bays 2 anything in the publishedCrystal Palace drawings and3,the trainshedoccupying Bays 4to14, with and would have required entirely new casting thetracksthemselves terminatinginbuffers within moulds. thesouthern half of Bay 4. Theislandplatform was One of the principal differencesbetween the two bisected by atransversesidingwithwagon turn- structures lies in the method of connecting the tables within Bay6.Bay 1formedthe bi-partite principal verticaland horizontal components of porte-coche` re providing coveredloading/unload- the framework. At the Crystal Palace, asystemof ingspace in frontofthe main entrance to the wedge fixings, patented by Paxton, was used station. Offices andpassenger facilities were accom- (Fig. 37) while at Oxford, trussesand beams were modated within the flankingwings, accessed via securedtocolumns by paired1inch (25 mm) bolts. theconcourse arcade.The plan is detailedenough The reasons forthese differencesare not entirely to show internal room divisionsand anumberof clear, though Kinchin-Smith (OAU 1998, 25)has room arelabelled. Theeastwingcontained alarge suggested it may have been away forFox and booking office, occupying Bay3and thenorthern Henderson to avoid breaking Paxton’s patent. thirdofBay 2, awaiting room (possibly3rd class) Alternatively, it may simply be that the bolted andstore,withthe StationMaster’srooms located connections were designed to be more durable, in thesecondary northern extension. Thewestwing better suited to apermanent building with alonger contains twowaiting rooms(presumably1st and envisaged lifespan than the Crystal Palace, which 2ndclass),acloak room andattached toiletfacilities was from the start planned as atemporary alignedalong thewestwall. Thenorthern extension structure. Patent avoidance may also have been a houses aporter’sroomand lamp room.Within deciding factorinthe decision not to use Paxton’s theroof-space of thewest wing tongue-and-groove Crystal Palace trussed gutter beams in the station boarding,preserving the profile of theprimary roof, resulting in persistent structural problems ridge-and-furrow roof,was recorded cladding the which led eventually to the need to re-roof the rear brickworkofthe chimneyindicatingthatthe whole station building at an early date (see below). fireplaceheating thewaiting room wasprimary,or Though financial and timeconsiderations must at least averyearly addition to theplanasitmust have played apart in the Board’s decision to con- date to before thereplacement of theprimary roof tract Fox and Henderson, and the rapidity of the (see below).Evidenceregarding thechimney in the construction process tested at the Crystal Palace east wing suggeststhatitisaninsertion,possibly clearlylent itself commendably to the circumstances contemporary withthe re-roofing of theflanking of the station project, their involvement should wings. not be reduced simply to the level of economics.

82 OA Occasional Paper No: 16

Table 9Summary of primary walling components.

Ref Type Material Description

W1 Wall Panel Timber Tri-partite wall panel used in external walls of flanking- wings, spanning between primary cast-iron columns (C2a/C3) and incorporating · 2timber columns (C4) per 1 panel. Formed of 4 ⁄2 in. · 2in. (115 · 50 mm) softwood studwork. Clad externallywith horizontalship-lap board- 3 ing, 6 ⁄4 in. (170 mm) · 7/8 in. (22 mm), and internally with 3 vertical, reeded tongue-and-groove boarding, 6 ⁄4 in. · 7/8 in. (170 mm · 22 mm). Panels secured to iron columns using cast-iron‘butterfly’ fixings. Manufactured in situ. W2 Wall Panel Timber/corrugated-iron Tri-partite wall panel used in the arcade between the station concourse and the flanking wings, spanning bet- ween primary cast-ironcolumns (C2a/b) and incorp- orating · 2timber columns (C4) per panel. Formed of 4in. · 2in. (100 · 50 mm) softwood studwork, clad to the flanking wing side in vertical, reeded tongue-and- 3 groove boarding, 6 ⁄4 in. · 7/8 in. (170 mm · 22 mm) and to the concourse side in corrugated iron. Panels mostly include three integral,round-arched door openings with glazed fan-lights (one example contains only two doors). Panels are secured to iron columns using cast-iron ‘butterfly’ fixings. Manufactured in situ. W3 Wall Panel Timber/corrugated-iron Tri-partitepanel wall panelusedinsouth elevationofstation concourse, spanning betweenprimary cast-ironcolumns (C2/C2a) to whichtheyare securedtoironcolumns using cast-iron‘butterfly’fixings. Formed of 5in. · 2in. (127 mm · 50 mm)studwork, clad externally with standard 3 horizontal ship-lap boarding,6⁄4 in.(170mm) · 7/8in. (22mm),internallywithcorrugatediron. W5 Wall Panel Timber Wall panel originally used throughout the full length of the trainshed (Bays 4–13). Most were replaced prior to 1914, though examples survive within Bay 4, where they were preservedbythe low extensions built against them. 3 They comprise asingle skin of 8 ⁄4 in. reeded, vertical tongue-and-groove boarding, faced to the exterior, fixed to the outer face of 3-in. deep, chamfered horizontal rails. 3 These rails are 6inches thick at the centre narrowing to 1 ⁄4 in. where they are secured to cast-iron columns (C1/C2a) using cast-iron ‘butterfly’ fixings. Extend to bottom flange of edge-beam B1. 1 V1 Ventilation Panel Timber 7ft6in. long · 2ft10 ⁄4 in. deep, louvred panel. Used in external walls of trainshed below eaves level, fitted between the upper and lower flanges of the edge-beams (B1). Divided into two panels, each containingsix horizontallouvres. Used in conjunction with V2, one of each per bay, the shorter module being located to the north of the longer module in each case. 1 V2 Ventilation Panel Timber 15 ft 10 in. long · 2ft10 ⁄4 in. deep louvred panel. Used in external walls of trainshed below eaves level, fitted between the upper and lower flanges of the edge-beams (B1). Divided into four panels, each containingsix horizontallouvres. Used in conjunction with V1, one of each per bay, the shorter module being located to the south of the shorter module in each case.

83 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Figure 36 Ordnance Survey FirstEdition 1876,showing Oxford Stationasbuilt (withside-wing extensions).

The choice of engineers and design was unquestion- with it, to capturethe imagination of the people ably influenced by the Board’s desiretocapitalise of Oxford with the aim of encouraging them to on the publicity surrounding the Hyde Park exhibi- use their more circuitous route to the capital. The tion building, and by associating themselvesclosely use of the structural and decorative motifs of the

84 OA Occasional Paper No: 16

and southern walls of the westernextension were in brick, laidinFlemish bond, that to the south containing asubstantial fireplace serving the newly createdroom. The exact phasing of these extensions is proble- matical; that theyincluded standard profile columns at their outer corners would seemtoimplycon- temporaneity with the main station building. How- ever, it is clearthat that the columns were not custom made in this context, and taken together with the evidence of differing foundation andfloor structure details, it seems reasonable to suggest that the extensions represented an on-site adaptation to the original plan as opposed to aseparate building phase per se.They were certainly in place by the time of the 1876 First Edition Ordnance Survey plan (Fig. 36) on which they are shown; the eastern ex- tensionhoused the ‘Station Master’s Rooms’while that to the west housed the ‘Porter’s Room’ and ‘Lamp Room’.Interestingly, during dismantling of the westernextension, asmall information poster (L&NWR No.1229) pertaining to the ‘Instructions as to Trimming Lamps’ issued by the ChiefTraffic Managers Office at Euston on 3April1875, was recorded.

The re-roofing of the flanking wings, pre-1888 At an early date, the transverseridge-and-furrow roofs of the flanking wings were replacedintheir entirety by fully hipped roofs comprising five substantial, timber queen-post trusses (Plate 20), clad with softwood roof-boards and slate,subse- quentlyoverlaid with bitumen. In both wings, the Figure 37 Detail of Paxton’s patented wedge-fixing outer ends of the trusses were carried on hardwood system used at the Crystal Palace. (Downes and Cowper) wall-plates running around the external walls. Where the flanking wings abutted the central Crystal Palace was conspicuous, possibly down concourse arcade, the details varied; in the eastern to the level of reproducing the original paintwork wing, the trusses were supported on iron brackets or scheme. cleats bolted to the outer facesofthe primary columns (both iron and timber) of the concourse arcade whileinthe westernwing, the trusses were Alterations to the PrimaryStructure carried by atimber plate supported on aseries of The extensionstothe flankingwings, pre-1876 secondary, stop-chamferedsoftwood posts. No evidence exists to allow adefinite date to be assigned At apoint early in the life of the station,two small, for this work. However, it must have been carried low extensions were appended to the north of the out prior to or contemporary with the introduction east and west flanking wings, abuttingthe sidewalls of aseries of intermediary trussestosupport the of the trainshed. Eachmeasured one full bay (7.32 m) primaryroofstructure over the concourse in c 1888 N–S · 2/3 bay(4.88 m) E–W and were built with (see below) as the valleygutters serving the original double,longitudinal low-hipped roofs, partly hidden side wing roofs, feedinginto the hollow arcade behind the exterior walls seemingly in imitation of columns via the flanged openings described above, the original concourse/trainshed roof.Atthe outer would have to have beendismantled forthe corners, cast-ironcolumns of identical section to the insertion of the fish-bellied edge-beams associated primaryuprights were used, although theywere with this phase of work. The principal roomsofthe here crudely truncated to match the low level of the wings were ceiled over at this stage and incorpo- extensions and, though serving as downpipes to rated,inplaces, elaborate moulded cornices (Plate drain the roofs, were notconnected to sub-surface 21), though the less prestigious rooms, such as the foundation columns as in the primarystructure. The toilet facilities in the west wing, remained open to outer walls were mainly of studworkconstruction, the underside of the roof. clad externally with verticalship-lap boarding and With the removal of the glazed roof,the funda- with no intermediate timber columns. The northern mental problem of lighting the side wings arose. The

85 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 20 Roof structure of westwing,showing wall plate and stoppedlap joints. Photo: Gifford Archaeology. solution to this problem involved the total renewalof the exterior studwork walls and the introduction of new panels including square-headed, double sash- windows within the west, south and east walls of the flankingwings. Theexterior elevations of the new panelswere finishedinhorizontal ship-lap boarding, whileinternally the original verticalcladding was re- used though, in places, ratherawkwardly applied. Each side wing roof, as recorded,was additionally furnished with two light wells. In the east wing,these served the main booking officeand the ladies’ waiting room;inthe west wing, the northern of the two waiting rooms and the adjacent toilet facilities. The light wells have been assumed, for the most part, to be contemporary(though see below), based upon the fact that the structure of at leastone example in the eastern wing was found to incorporatereused ‘Paxton-esque’ glazing bars, presumably originat- ing from the dismantled primaryroof structure (Plate 22). Further, evidence recorded within the roof space (OA 1998, 24) indicatesthat the original room plan was maintained after the replacementofthe glazed ridge-and-furrow roof, and it is apparent from an examination of the 1876 plan that certain of the roomswould have been left with no source of natural light if the roof lights do not represent contemporary features. Evidence from the dismantling surveyis, however, somewhat contradictory in this respectand small sockets recorded withinthe purlin of the west Plate 21 ‘Ghost’ of former cornice in original booking wing roof suggest that the outer side at least of this office (east wing).

86 OA Occasional Paper No: 16

Plate 22 Paxton-esque glazing bars re-used in the structure of the secondary roof light of the eastern side-wing. roof may originally have been glazed. This would principal trusses throughout the full length of implythat the roof lights of the west wing may the concourse, trainshedand porte-coche` re, halving represent aslightly later adaptation. the distancethat the roof had to span between Archaeologicalevidence, recorded afterdisman- supports. Structural componentsrelated to this tling, suggests that the brick-built chimney stack of phase of work are summarised below in Table 10 the eastern flanking wing is secondary to the station and illustrated in Figure 38. It is worth noting in structure, and it seems reasonable to associate the this respectthat the roof of the locomotive shed, introduction with the re-roofing of the wings. evidently built to asimilar design,was recorded as having been repaired in 1877 and 1879 before being completely renewedin1883 (Waters 1986, 43). The reinforcement of the primaryroof, c 1888 The inserted trusses were of composite steel Weaknessesinherentwithin the original longitudi- and wrought-iron construction and were somewhat nal roof structure becameevident at an early stage deeperthan the primarytrusses T1, thoughagain of and aprogramme of remedial works was necessary ‘Pratt’ type. Within the trainshed, the inserted trusses to provide additional support. The flaw was (T2) were supported on simple, free-standing‘A’- associated with the wide span between the primary frames (C5) formedofpaired sections of wrought trusses, exacerbatedbythe fact that the trussed iron, double-headed rail, set approximately 1ft. in gutter beam, patented and usedbyPaxtoninthis from the exterior walls, to which theywere braced context at the Crystal Palace, was apparently not with diagonal struts. Within the area of the concourse employed in the Oxford building, most probably to (Bays 2and 3) the introduction of the new trusses avoid breaking the terms of the patent. To redress was alittle more involved. Here the trusses (T3), the problem,aseries of additional, intermediate rather than being carried on inserted ‘A’-frames, trusses(T2–4) were inserted midway between the which would have disrupted circulation between

87 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Table 10 Summary of constructional components related to strengthening of primary roof.

Ref Type Material Description

C5 ‘Column’ wroughtiron ‘A’-frame constructed of paired lengths of wrought iron, double-headedrail, fixed horizontally with bolts and spacing pieces at c 4ft(1.22 m) centres. Set c 1ft(0.30 m) in from outer wall. Introduced (together with trusses T2) midway between principal columns/trusses to provide additional support for the primary, longitudinal ridge-and-furrow roof structure within the trainshed section of the building. Four examples survive, located within Bays 4and 5. T2 Truss steel/wrought iron c 46 ft (14 m) long · 4ft2in. (1.27 m) deep. ‘Pratt’ type truss divided into six panels with single diagonal tension brace per panel. Braced to ‘A’-frame columns (C5) by raking struts. Introduced to provide additional support to primary, long- itudinal ridge-and-furrow roof structure within the trainshed section of the building. Two examples survive, located within Bays 4and 5. B2 Side-Beam steel/wrought iron 23 ft 4in. (7.12 m) long · 3ftin. (0.93 m) deep at fixing/4 ft 2 in. (1.27 m) deep for central 2/3 of span). Span longitudinally between arcade columns of concourse and support inserted transverse trusses T3 at their mid-point. Introduced to provide additional support to primary, longitudinal ridge- and-furrow roof structure within the concoursesection of the building. Four examples survive, within Bays 2and 3where they replace original B1 side-beams, reusing primary bolt connections. T3 Truss steel/wrought iron 48 ft (14.63 m) long · 4ft2in. (1.27 m) deep. Essentially similar to T2, they were introduced to provide additional support to primary, longitudinal ridge-and-furrow roof structure within the concoursesection of the building. In order to avoid using ‘A’-frame columns C5, which would disrupt circulationwithin the concourse/flanking wings, the trusses are here carried by inserted, fish-bellied side beams B2. Two examples survive, located within Bays 2and 3. T4 Truss steel/wrought iron Essentially similar to T2/T3, though here with ‘fish-bellied’ profile: 48 ft (14.63 m) long · 3ftin. (0.93 m) deep at fixing/ 4ft2in. (1.27 m) deep for central 2/3 of span. Only one example of this type exists, located within the stub of the porte-cochere (Bay 1). the concourse and flankingwings, were carried evidence demonstrates this nottobethe case, clearly through to the line of the main north-south walls. showing the insertedsupportsinconjunction with However, as the new trusses were some 0.35 mdeeper the primaryroof(Plate 16). Further, the grouped than the primarymembers, their introduction neces- bolt-holes recorded in the upper chords of the sitated the replacement of the original cast-iron edge primarytrusses (T1) during dismantling, were also beams (B1) with fish-bellied beams (B2), deep enough notedtobepresentinthe inserted trusses T2/T3, at the centre to accommodate the new trusses. Within proving that these trusseswere inserted for the the porte-coche` re, avariant truss (T4), of fish-bellied support of the original roof. form, was introduced. As noted above, the introduc- No evidence survives to allow an exact datetobe tion of the deeper, fish-bellied edge-beams (B2) within assignedfor the work on the station roof, though the concourse made the valley gutters of the side wing adated plan (Drg No. 66792, L&NWROxford roofs redundant,implying that the roof structures of Passenger and Goods Station: July 1888), seen by the flankingwingshad already been replacedbythis Sutherland in 1975 in preparationofhis article point (Fig. 38). ‘Oxford Midland Station and The Crystal Palace’, Although it would at first appear logical to markedthe insertedsupports forthe intermediate associate the introduction of these intermediary trussesas‘new’ at this date. Unfortunately, thisplan trusseswiththe replacementofthe original longi- has since been lost. The last dated rail rolling mark tudinal roof by the transverse ridge-and-furrow roof is 1872, and the rail had seenservice on the track which respects them (see below), photographic evidenced by compressions where it had sat on cast

88 OA Occasional Paper No: 16

Figure 38 Details of secondary components related to support of primary roof: (a) side-beam B2; (b) truss T3/4; (c) truss T2. iron chairs,while the location of the ‘A’-frame Surveyplanhad gone out of use by this time. supportsfor the inserted trusses withinthe trainshed Indeed, Plate 16 clearlyshows one of the inserted demonstrates that the transverse sidingand asso- ‘A’-frames obstructing the large wooden sliding ciated wagonturntables visible on the 1876 Ordnance doors in the trainshedside screens.

89 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

The replacementofthe primaryroof, 1901–1906 The new booking officeisnot shown on aplan of 1903 (Fig. 39), which showsthe layout of the station Despite the reinforcement work undertaken on the facilities muchasitwas in the 1876 Ordnance Survey concourse and trainshed roof in c 1888,itisclear that plan, with the additional inserted trusses of the problemswiththis elementofthe building persisted concourse/trainshed roof. The original context of and eventually the entire roof had to be replacedto this plan is, however, unclear and the similarity of an alternative design. The new roof maintainedthe the internal layouttothe 1876 plan has led to hipped, ridge-and-furrow pattern, but was aligned speculation that these details may simply have been transversely (E–W) with valleys set wider than the traced from the earlier plan, bringing intodoubt its primaryroof, at 12 ft (3.66 m) centres, respecting value as areliable historical document of the layout with the primary(T1) and insertedtrusses(T2–4). A of the station at any specific date. furthertwo bays, of 8ft(2.44m) and 16 ft (4.88 m), With the relocationofthe booking officetothe protectedthe porte-coche` re. centre of the concourse, the arrangements of the side Assigning aprecise date for the introduction of wings could be improved. The 1906 plan (see above) this roof has not proved possible, although two indicatesthat at that date the east wing housed, from sourcesallow aclose estimate to be proposed. As south to north, aladies’ waiting room,slightly notedabove, the earliest known photograph of the enlargedfrom the 1876 plan, with attached toilet station interior (Plate 16) shows the longitudinal facilities, ageneral waiting room and aparcels office roof still in place along with the inserted trusses, and porters’ office extending into the northern low confirming that the intermediate trusses were in- extension. The new general waiting room and parcels serted for the support of the original structure and officewerecreated by the subdivision of the former were not contemporary with the new, transverse large booking office. The west wing containeda roof. Although this photograph is undated, the second ladies’ waiting room (again with attached character of the rollingstock is suggestive of adate toilets), an enlargedgentlemen’s waiting room and 70 in the early yearsofthe 20th century, and it toilet block and,within the northern extension, the provides aloose terminus post quem for the introduc- station master’s office, fireman’s officeand store. Each tion of the new roof. Adated drawing of 11th July room was separately accessedfrom the concourse, 71 1906, illustrating the ‘renewal of the roof etc.’ with no interconnectivity between rooms. Afurther shows the transverse ridge and furrow arrangement plan of c 1920 (Fig. 40) shows an essentially similar in place at this date, and details the ‘proposed layout, though the station master’s officehad been glazing’ of thisstructure, though it includes few moved to occupy the northern halfofthe central ticket details of the roof construction itself. It is unlikely office, whilethe porters’ room has been moved to the that aroofintroduced post-1901 would require re- west wing extension and the parcels office in the east glazing within five years and it seemsreasonable to extension enlarged. assume that the surviving drawing represents one of an original series detailing amore extensive pro- gramme of works to the roof. Adate range of The raising of the platform,1906 between c 1901 and 1906 can thus be proposed for The 1906 plan referred to abovespecifically details the the introduction of the transverseroof,with the raising of the island platform by c 1ftalong with likelihood that it falls towards the latter end of this associated alterations to the ramp between the range. concourse and the new platform level and the relocation of the buffer stops one-and-a-half bays to Internal reorganisation the north. The reasons for these works were based on both safetyand loading/unloading efficiency con- An early20th-centuryphotographshows afreestand- cerns.Inthe early years of the railway in Britain, the ing, apsidal-endedbooking office(Plate16),located train/station interface was designed on the basis that centrallywithinBay 3ofthe concoursearea. This struc- passengers steppedupinto the train from alow-level ture wasintroducedearly in thelifeofthe stationto platform,oreven straight from the ground, via a replaceanoriginaloffice in theeastside-wing and system of running boardsfixedtothe outside of the survived untilthe closureofthe stationtopassenger train running gear below the carriagebodywork (see traffic in 1951 andwas subsequently removed. A Plate 16). From the turn of the 20th century,however, photograph from c 1940 (see Plate24) showsits form in it was deemed desirable that platform levels be raised some detail,while its‘footprint’ waspreserved within closer to the height of the trainfloor, with aview to theYorkstone slabsofthe concoursefloor. Located ensuring speedier and safer embarkation. The new 1 centrallywithinBay 3, thestructure was16ft9in.(5.1 platform at Oxford,raisedfrom 1ft 10 ⁄2 in. to 3ft, is m) wide by 26 ft 3in. (8 m) long with apsidalends. It first seen in aphotograph of 1914 (Plate 23). wasdivided into vertical andhorizontalpanels, used forinformation andadvertising display, by timber Rationalisation of the rail network, 1921–1951 railsand columnsabove adadooffielded timber panelling. Thetop of thestructure wasfurnished with Throughout the early years of the 20th century, a decorative ‘Crystal Palace’crests, mirroringthe eaves slow rationalisation of the national rail network took leveldecorationofthe stationexterior. place.During the First World War, the network had

90 OA Occasional Paper No: 16

Figure 39 Plan stamped ‘Engineers Office, L&NWR, Euston, No. 4038, 28 Sep. 1903’, though annotated‘copied from (?) plan dated 18(?)7.CLR’. The inserted intermediary roof supports are clearlymarked, while the booking office is still shown as being located in the east wing. operatedunder central, state control andthis Midland and ScottishRailway (LMSR) (Plate 25), continued until 1921. Underthe Railways Act of the Londonand North Eastern Railway (LNER)and 1921, which came intoeffect on NewYear’sDay the Southern Railway (SR). 1923, more than 120 individual companies were Furtherrationalisationfollowed and when the regrouped to form four major companies, each entirerail network was nationalizedunder the dominating its own geographic area. These were Transport Act of 1947, coming into effect on 1 the Great Western Railway (GWR), the London, January 1948, the London, Midland and Scottish

Figure 40 Plan of station dated to ‘early 20th century’. Note introduction of free-standing booking office.

91 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 23 Interior of the train shed looking north, 1914. Note the secondary, transverse ridge-and-furrow roof and the raised platform level. Note that the upper ventilation panels (V1 and V2) have been removed by this date. Henry WTaunt Coll. (CC54/00325) q Oxfordshire CountyCouncil.

Railway was merged into the London Midland in the lower parts of timber columns and wall Region of British Railways. As apartofthe panels(Plate 26). The windows would have been rationalisationofthe railway network, from 1951 supported and the external cladding removed, on all passenger trains on the formerLMS branch leaving the internal vertical cladding in situ,up line were divertedtothe former GWR station.The against whichthe brick plinths were constructed.The LMS tracks and sidings continued in use,for goods introduction of these plinths involved the truncation traffic only, in ever diminishing volumes whilethe of allbut one of the external,intermediate timber station building itself found temporaryuse as a columns. At some point between 1951 and 1959 (see hostel for goods train crews. Plate 27), the southernmost 16 ft, outer bay of the Modifications to the station building during this porte-coche` re was removed, the columns being period appear to have been fairly limited. Small-scale truncated at groundlevel (the foundation columns internal reorganisations appear to have been under- of the port-coche` re were recorded during post- taken,for example the subdivision of the general dismantling works). waiting room in the east wing, and were traceable as wall and ceiling scars within the surviving Post-railway use fabric.Larger scale works included the creation of asmall lobby in the south elevation of the eastern The station finally moved out of railway use in 1969, flanking-wing, undertaken at some point between when it was leased by the British Rail Property 1920 and 1936 and involving the removal of the Boardfor commercial and light industrial use. It central window of the elevation.Between 1949 and was at this timethat the northern nine bays of 1951, substantial blue-brick plinths were introduced the trainshedstructure (Bays 6–14) were removed to the exterior walls of the flankingwings, presum- and abreeze-block wallerected between bays4 ably in an attempttoalleviateproblems of wet rot and 5. Recognising the historical importance of the

92 OA Occasional Paper No: 16

Plate 24 The free-standing bookingoffice, seen in aphotograph of 1940. 26888, collection DY, courtesy National Railway Museum. structure, two complete bays of the dismantled University and accepted the building for use as a trainshed structure, including the inserted ‘A’-frame new visitor centre for the Buckinghamshire Railway supports, were acquired by the National Science Centre at Quainton Road Station, together with fund- Museum. The goods shed, locomotive shed and ing to move, refurbish and re-erect the building. water tank were demolishedand the island platform Thedismantling andpre-erectioninspection and levelled. In its final yearsinOxford, the station testingprocess forthe Rewley Road stationindicated building saw use as atyre and exhaust depot and car anumberofsignificantstructuralweaknesses and rental centre.Internally, anumber of partitions were defects,and while therewas thedesiretopreserve and removed and an ephemeral officestructure formed reuse as much of theoriginalbuildingfabricaspossible within the area of the concourse; otherwise the early in thereconstruction, it wasnecessarytoerect the 20th-century layoutofthe concourse and offices buildinginawaythatwouldmeetlong-term conserva- survivedrelatively intact up until the time of the tion needs andrequirethe lowest possible maintenance archaeological survey. in itsnew function.Asthe structurewas dismantled (Plate 28), themetalwork wascleaned backtothe bare base,and bothvisualand non-destructivetestingwere Dismantling and re-erecting the station carriedout to identify anyworks of renewalorrepair Lance Adlam that werenecessarybeforere-erection. Introduction The main superstructure (Fig. 41; Plate 29) Since the Rewley Road station building could not be considered withinthe plans for the new Saı¨ dBusi- Thecast-iron hollow foundationcolumns (see Fig. 26) ness Schooland road scheme, an alternative site were foundondismantling to have areservoir of that could accommodate the station was sought. waterand detritus within them,and therewas no Quainton Railway Society was approached by Oxford wayofemptyingorclearingthe bases, or of rodding

93 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

Plate 25 Exterior view of station on 1st May 1940 clearly showing the form of secondary train shed and side-wing roofs. Note the lobby, formed in the south elevation of the eastern side-wing. 26886, collection DY, courtesy National Railway Museum. thecolumns whichformedthe rainwaterpipes forthe fundamental as roof supports, all the head sections building.Additionally, anumberofthe column bases were recast for the reconstruction. Theoriginal heads were broken,cracked or rotted through, andunusable were re-used as balusters to support the platform forthe reconstruction.Afewalsohad damage to the balustrade so that theycouldbeinspected. base plates,and wouldnot thereforebeabletospread The four cast-iron trusses taken from the original thecolumnloads evenly across thecasting.Asthe building appeared to be in reasonable condition, as ditchwater levels into whichrainwater discharged they had been little more than decorative features were considerably higher at Quainton than at Oxford, holding the column heads apartand supporting the it wasnecessary to recast thebases at alength gutters. Therewas, however, some serious deteriora- appropriatetothe newsiteinorder to maintain the tion of the iron and timber trusses across the main original appearance of thebuilding. Some movement hall and trainshed where the guttersabovefailed. of thefoundations hadbeendetected at Oxford,and For the re-erected building, some of the truss ends to avoidthishappeningatQuainton,aconcrete-filled were drilled out and replacedwith new castings, foundationtrenchwas dug. along with afew of the intermediate struts. The Of the 22 structural columns broughtfrom Oxford, rivets from the first diagonal tie barfromeach two were beyond use,and afurther three were columnheadwere drilled out and the fixing changed foundonstructural testing to be seriously degraded. to alargersectionmodern bolt. In the reconstruction, aplastic rainwater pipe was The vents underneaththe original ground surface insertedthrough the core of the original and recast had been blocked over the years, and the floors columns to avoid rainwater degradation. Some of had rotted out almost completely. As the station the columns were turned in the reconstruction from was to become aclosed building at Quainton, the their original orientation so that theycouldbeseen. reconstructed groundfloor was laid with reinforced The majority of the column headsections were concrete throughout, preventing excessive mois- damaged throughstructural changes, poor work- ture entering the building. The main hall of the re- manshipinthe original castings, or weathering,and erected building was flooredwith pre-cast concrete most of bolt fixingshad rottedaway. As they were simulated York Stone pavings; sadly, there were

94 OA Occasional Paper No: 16

Plate 26 View of station on 4th October1951, shortly before closure. Note the expose circular-motif decoration of the port-coche` re, and the brick plinthwalls introduced in the side-wings. Photo: Dr Parkes Collection. very few original pavings salvableondismantling. gated iron galvanisedsheeting. The sheets were Little was also left of the platform;the reconstructed foundtobeinasgood condition as when they were platformswere built with brick longitudinal walls first erected, with the exception of some damage supporting pre-cast,pre-stressed concrete planks where fixings were added over the years. All were with apavedsurface as used for the concourse, but re-used, although some additional modern sheets with modern pre-cast concrete tactile platform had to be used in one location where an existing edges, and built to the same height as the final sheet was re-cut for use over the cafe´ servery Oxford platform levels. entrance. Theflue to the west wing front room was The extant internal walls generallyfell apart as reconstructed as aworking chimneyfor acoal fire, they were removed. The walls were rebuilt with and to the cafe´ foragas fire. The other flues to the timber studs and plywood forming astiff box to east side becamedummies forappearance purposes which the facings were applied, and the variety of only, and were built solid aboveroof level to avoid temporaryinfill panelsusedatOxford were scrap- weathering problems. ped and replacedwith glazed doors and panels. The external windowed walls were taken down It was also necessary to removeany lead-based where possibleascomplete panels, generallywith paint from any of the re-used materials. Most of the the external face and window frames left intact. The architraves and trims and the semi-circular win- panelling was removed from the inside-wing wall dows survivedand were incorporated in the new faces to allowthe panels to be removed. As before, it design.Noneofthe original material in the brick was mostly tongue-and-grooved board and so des- or block walls was salvable. Much of the panel- troyed when it was removed. The boarding below ling was tongue-and-grooved boarding which was cill level was generally rotten, probablydue to con- destroyed on removal where fixingshad been pin- densation on the brick backing. Thepanels were ned through the tongues. re-erected and the timber given aspray or hand- The internal decoration set at high level in the applied timber treatment. The timber was replaced concourse area was provided by the use of corru- only where severely damaged or rotted.All the

95 Plate 27 View of station in 1959, when it was in use as agoods depot and lodging. Note that the southern section of the port-coche` re has been removed by this date. Photo: JDEdwards. OA Occasional Paper No: 16

Plate 28 Rewley Road station at the start of dismantling. Photo: Lance Adlam. panelshad new cills, and rotted timber column ends The roofs (Fig. 42; Plate 29) and some boardswere replaced. The brick elements 1 The gutter edge/roof beams allrequired replace- were rebuilt in ⁄2 brick Staffordshire Bluebrickwork externally from cill to finished floor level, and solid 1 ment; these were built up using two sections, since brick thick below (in hindsight, these plinth walls suitable wood matching the original construction shouldhave been built as halfbrick thick,but with was not available. The roof adoptedafter c 1888 snappedheaders to resembleEnglish Bond, notas placedanunacceptably high bending moment on the stretcherbond.) top flanges of the trusses,requiring astiff block at The vertically-mounted, ship-lap boarding was the tie bar location. The pads at the re-erected foundbehind the later panelling was in poor con- structure were therefore increased in number from ditionand did not survive the removal process. The five to sevenper truss to sit in the original positions circle-motif decorated panels from the front eleva- used whenthe roof was first constructed, centred tion were all severely damaged bar one. Several over the vertical strutsinthe trusses. The roof to the panelsappear to have been repaired at the foundry front canopywas clad in patent glazing to all four prior to despatch to Oxford in 1851, and most of facets.The main roof to the main halland trainshed those suffered further breaks while in situ.Never- had the north faces clad in patent glazing, and the theless, the original panels could be re-used, since otherthree faces were clad in Terne-coated stainless they were notnow subject to the roof loadsthey had steel sheet, with insulated construction below. previously carried. The semi-circular headed brack- The angle of the high level roofscovering the east ets were reused, except for broken sections, which and west wings was tooshallow, giventhe size of were recast. slates removed during dismantling, andconse- Thewalls of therearwingblockswereinsuch quentlythe roofs leaked. The slateswere ‘Turnerised’ apoorstate that they hadpartiallycollapsed.The (coatedinabituminous paint), which causedmost of twoblockswerereconstructed with timber facings, the slatestodisintegrate,aswell as some rot pro- butwithout windowsinthe walls. Timber louvres blems in the timbers below, due to trapped con- were re-located at Quainton on thesoutherly side densation. The central purlin was removed on the of thebuildingastheyhelpedtoreducethe direct re-erected building, and new timber rafters spanning sunlight on theinternalartefacts,and glareonbright the gutter to the rooflight edge were fitted, thus days. enabling the sameinsulated roof construction as for

97 Adlam. Lance by supplied drawings from thanks with Reproduced restoration. of extent station, Road Rewley 41 Figure

98 OA Occasional Paper No: 16

spurlesstypes. Therewas one plain spurless form that appeared to have had along stem (Fig. 43.44) and afurthersix similar forms, all from the same mould,decorated with asimple rib seam (Fig. 43.45). One seam had both oak leavesand acorns on it, a design certainly current by the middle of the century since it occurred on pipes celebrating the Great Exhi- bition of 1851. Another decorated form has looped panelsofdecoration on the stem and lowerpartof the bowlwithplant motifs above, including the rose, thistle and shamrock motifs for England, Scotland and Ireland (Fig. 43.46). This was apopular theme on Plate 29 The re-erected station (Phase 1) at Quainton. pipes datingfrom the midtolate 19th century. Photo: Lance Adlam. Finally there is amuch thicker plain bowl with a forward leaning form (Fig. 43.47). This style was known as a‘Miner’ or ‘Woodstock’ pattern and is usu- the main roof to be used. Thecentral valleyfacesof ally associated with late 19th- or early 20th-century the roofscovering the rearblockswere glazedwith groups. It is possible, however, that this form was patent glazing, while the four perimeter faceswere introduced earlier and that it was contemporary with the same roofingsystemasfor the other roofs. the other designs. These all fit with the 1841–76 range provided by the Huggins pipe and it could well be The porte-coche ` re and trainshed that thislevelling layer dates from around1850 and that it is associated with the London Midland and The porte-coche` re was damaged by aheavy vehicle ScottishRailway terminus of 1851. If this is the case, in the 1950s and, for the safetyofthe railway staff, then the group provides am important benchmark removed as scrap, leaving only the canopy over the for mid 19th-centurypipe styles in Oxford. front entrance to the station.Itwas reconstructed in asecond phase of works at Quainton using pho- tographs and the known measurements of the deco- Marked and decorated pipes rative ironwork. The circle-motif patterns were BH One spur bowl with the moulded initials BH on the spur was modified, and the panels castintwo equal halves recovered from OXSTRA 96 [0010] (Fig. 43.42).This can be attributed with acentre joint; anew sectionwas cast to provide to the Oxford pipemaker Benjamin Huggins, recorded working from 1841–76 (Oswald 1984, 262) and this particular example comes the link between the two quarter-circle column brac- from alevelling deposit that may well be related to the construction kets. Theroof was constructed to the samepattern as of the railway terminus in c 1850. Other pipes made by Huggins, the canopy,and was fully glazed with clear patent some of which also have circular or shield shaped bowl stamps or glazing. leaf decoratedseams, are also known (Oswald 1984, fig 56). Only four bays survivedfrom the original train- Decorated material was additionally recovered from amid 19th- century levelling layer, OXSTRA 96 [0010]. This probably dates shed. There-erected trainshedwas extended as part from c 1850 and shows the styles in production at that period. Most of the Phase 2work by afurther five bays to the rear of these pipes all seem to have had long stems althoughseveral of to allow the feel of the former station to be recreated. the bowl forms are of types that were later produced with short This work emulatedthe original structural appear- stems. The decoration was fairly restrained for the main part. There were just two examples with leaf-decorated seams, both from the ance but used modern materials and all-welded same mould (Fig. 43.43). Had alarger sample of 19th-century pipes construction for the trusses. been recovered there would, no doubt, have been amuch larger range of decorative designs represented. Amid 19th-century group of clay tobaccopipes David Higgins List of illustrations (Fig. 43) Agroup of 12 pipe fragmentswere recovered from a 42. Spur bowl dating from c 1840–1875; not burn- 19th-century levelling layer(OXSTRA 96/97 [0010]). ished; cut rim with no milling; stem bore 4/6400. It is unfortunate that notmore material was recove- On either sideofthe spurare the relief moulded red from this layer since it was clearlyafresh and initials BH. This is most likely to be Benjamin little disturbed deposit, the number of duplicate Huggins of Oxford, who was recorded working forms presentsuggesting anarrow date range.All 1841–1876. (OXSTRA96 Context:10). the pieces were complete bowls with up to 125 mm 43. Spur bowl dating from c 1840–1875; not burn- of surviving stem. Therewas onespur bowl with the ished, cut rim with no milling; stem bore 4/6400. moulded initials BH on the spur (Fig. 43.42). This can Oneoftwo identical bowls with leaf and acorn be attributedtothe Oxford pipemakerBenjamin decorated seams and star marks on the spur, Huggins, recorded working from 1841–76 (Oswald which may have been cut over earlier initials. 1984, 262). There were two identical bowls with leaf (OXSTRA96 Context: 10). and acorn decorated seams and star marks on the 44. Spurless bowldating from c 1840–1875; not spur, whichmay have beencut over earlier initials burnished;cut rim with no milling; stem bore (Fig. 43.43). The remaining nine bowls were all 4/64 00.(OXSTRA96 Context: 10).

99 Adlam. Lance by supplied drawings from thanks with Reproduced trusses. of age showing plan roof station, Road Rewley 42 Figure

100 OA Occasional Paper No: 16

Figure 43 Agroup of mid 19th-century clay tobacco pipes, nos 42–47. DrawnbyDrSusie White.

101 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

45. Spurless bowldating from c 1840–1875; not Cotter, JP,2001 ThePottery in St Gregory’sPriory, burnished;cut rim with no milling; stem bore Northgate: Canterbury excavations 1988–1991 4/64 00.One of six similar forms,all from the same (eds MHicks and AHicks), The Archaeology of mould, decorated with asimple rib seam. Canterbury NewSeries II, 231–266 (OXSTRA96 Context: 10). Dolley, RHM, 1968 The Irishmints of Edward Iin 46. Spurless bowl datingfrom c 1840–187; not the light of the coin-hoardsfrom Ireland and burnished;rim cut with no milling; stem bore Great Britain, Proceedings of the Royal IrishAcademy 4/64 00.Mould decorated with looped panels on 66,section Cno. 3, 235–97 the stem and lower partofthe bowl with plant Dolley, [R H]M., 1972 Medieval Anglo-Irish coins, motifs above, including the rose, thistle and Seaby,London shamrock motifs forEngland, Scotland and Downes Cand CowperC,1852 The building erected in Ireland. (OXSTRA96 Context: 10). Hyde Park for the Great Exhibition of the Works of 47. Spurless bowldating from c 1840–1875; not Industryofall Nations, 1851 (reproduced by Victoria burnished,rim cut with no milling; stem bore and Albert Museum,1971) 4/64 00.This thick-walled, forward leaning form is Drury PJ,1981 The production of brick and tile in astyle that was known as a‘Miner’ or ‘Wood- medieval England, in Medieval industries (D W stock’ pattern. (OXSTRA96 Context: 10). Crossley), CBA Res Rep 40 Eames, ES,1968 Medieval tiles: ahandbook ,British Museum, London BIBLIOGRAPHY Galloway, P, 1990, Toilet equipment: combs of bone, antler and ivory, in Object andeconomy in medieval Aksa` di, Gand Nemeske´ ri, J, 1970 History of human Winchester (M Biddle), Oxford University Press, life span and mortality, Akade´ miai Kiado´ ,Budapest Oxford Atherton, K, and Mitchell, N, 2001 Ceramicbuilding Gifford and Partners Ltd, 2000 Relocation of Rewley material’, in Beaumont Palace and the Whitefriars: Road station, Oxford: Report on the archaeological Excavations at the Sackler Library, Beaumont Street, recording work,unpubl. report no. B1436A/R09 Oxford (D Poore and DRPWilkinson), OAU Goodhall, IH,1993 Implements,in Norwich house- Occasional Paper 9 ,69–74 holds,medieval and post-medieval finds from Norwich AOC Archaeology Group, 1999 An archaeological Survey excavations 1971–78 (S Margeson),The watching brief at the site of the old LMS Station, Norwich Survey/Norfolk Museum services Oxford, unpubl. report Green,J,1988 Medieval floor-tiles from St Frides- Arkell WJ1947 Oxford stone,Faber and Faber, wide’s Priory, Oxoniensia 53,103–114 London Haberly, L, 1937 Mediaeval Englishpavingtiles, Atkinson, DRand Oswald, A, 1969, London clay Shakespeare Head Press, Oxford tobaccopipes, JBrit Archaeol Ass 32,171–227 Hardy, A, Dodd,A,and Keevill, GD,2003 Aelfric’s Bass, WM,1987 Human osteology. Laboratory field Abbey:Excavations at Eynsham Abbey,Oxfordshire, manual (3 ed.), Special Publication 2 ,Missouri 1989–92,Oxford Archaeology ThamesValley Archaeological Society, Columbia Landscapes Monogr 16,Oxford Blinkhorn, PW,2003 Thepottery, in Hardy et al. Higgins,DA, 1987 The interpretation and regional 2003, 159–206 study of clay tobacco pipes: acase study of the Broseley Boocock, P, Roberts, Cand Manchester, K, 1995 District,Doctoral thesis submittedtothe Univer- Maxillary sinustitis in medieval Chichester, Eng- sity of Liverpool land, American Journal of Physical Anthropology 98, Higgins,DA, 2005 Clay tobacco pipes,inPrehistoric 483–496 and Romanactivity and aCivil War ditch: Brickley, Mand McKinley, J, 2004 Guidelines to the Excavations at the Chemistry Research Labora- standards for recording human remains,IFA Paper 7 , tory, 2–4 SouthParks Road, Oxford (P Bradley British Association for Biological Anthropology et al.), Oxoniensia 70,172–174 and Osteoarchaeology and the Institute of Field Higgins,DA, forthcoming Clay tobacco pipes from Archaeologists excavations at the Oracle Site, Reading,Berkshire Buikstra, JEand Ubelaker, DH(eds), 1994 Standards (REOR 96–97 and REORM 98), Oxford Archaeo- for data collection from human skeletal remains, logy ThamesValley Landscapes Monographseries Arkansas Archaeological Survey Research Series Hohler, C, 1942 Medieval paving tiles in Buckin- 44,Arkansas ghamshire, Rec Buckinghamshire 14, parts 1and Cannon, P, 1991, Evidence of tobacco pipe making in 2, 1–49; 99–131 East Woodhay and District, Transactions of the Lambrick, G, and Mellor, M, 1985 The tiles, in Newburyand District Field Club 14 (1),16–27 Furtherexcavations on the second site of the Charleston, R, 1984 Englishglass and the glass used in Dominican Priory, Oxford (G Lambrick), Oxonien- England, 400–1900,Allen and Unwin, London sia 50,177–87 Charleston, R, 1985 The vessel glass, in Battle Abbey: Lewis, ME,2002 Urbanisation and child health in The eastern range and excavations of 1978–80 medieval and post-medieval England,BAR Brit. Ser. (J Hare), HBMC ArchaeologicalReport, 139–46 339,Oxford

102 OA Occasional Paper No: 16

Lovejoy, CO,Meindl, RS,Pryzbeck, TRand Roe, F, 2003 Whetstones, querns and other non- Mensforth, RP,1985 Chronological metamorpho- structural workedstone, in Hardy et al. 2003, 290–7 sis of the auricular surface of the ilium: anew Roe, F, 2005 Assessment of stone from Oxford Castle, methodfor the determinationofadultskeletal age unpublished assessment report for Oxford at death, American Journal of Physical Anthropology Archaeology 68,15–28 Scheuer, Land Black,S,2000 Developmental juvenile Mellor, M, 1984 Asummary of the key assemblages. osteology,Academic Press,London Astudy of pottery, clay pipes,glass and other SimpsonW,1981 Oxford to Cambridge Railway Volume finds from fourteen pits, dating from the 16th to 1: Oxford to Bletchley ,Oxford the 19th century, in Excavations at St Ebbe’s (T G Smith, TP,1998–9 London’searliest medieval Hassall et al.), Oxoniensia 49,181–219 roofingtiles: acomparative study, Medieval Cera- Mellor, M, 1989 Tiles,inExcavations in St Ebbe’s, mics 22–23,66–71 Oxford, 1967–1976: Part I: Late Saxon and medi- Spink, 2003 Coins of ScotlandIreland and the Islands eval domestic occupationand tenements, and the (Jersey, Guernsey, Man &Lundy). Pre-DecimalIssues, medieval Greyfriars (T GHassall, CEHalpin and (2 ed.), Spink, London MMellor), Oxoniensia 54,248–255 Stroud,Gand Kemp, R, 1993 Cemeteries of St Andrew, Mellor, M, 1994 OxfordshirePottery: asynthesis of Fishergate,The Archaeology of York: The medieval middleand late Saxon,medieval and early post- Cemeteries 12/2,CBA Res Rep for York Archae- medieval pottery in the Oxford Region, Oxoniensia ologicalTrust, York 59,17–217 Sutherland, RJM1975 Oxford Midland Station and Mellor, M, 1997 Pots and people ,Ashmolean Museum, the Crystal Palace, The Structural Engineer 2 ,53, Oxford 69–72 Moore, WJand Corbett, C, 1971 Distribution of dental Tibbles,J,2005a The Ceramic and other building mate- caries in ancientBritish populations: Anglo-Saxon rials from Oxford Castle,unpublished report period, Caries Research 5 ,151–68 no. 2076 North,JJ, 1997 The Anglo-Irish halfpence, farthings Tibbles,J,2005b The ceramic building materials, in and post-1290 pence of Edward Iand III, British Archaeological investigations at MortonLane, Bever- Numismatic Journal 67,11–19 ley, East Yorkshire (R McNaught and AWebb) OA 2005 Rewley Abbey and the old LMSstation, Oxford: Tibbles,J,forthcoming The ceramic building materi- post-excavation assessmentand project designfor al, in Excavations at Blanket Row, Hull (J Lee) publication,Oxford Archaeology unpubl. post- Trotter, M, 1970 Estimation of stature from intact excavation assessment and updated project de- long limb bones, in Personal Identification in Mass sign, revised version November 2005 Disasters (ed. TDStewart), National Museum of Ortner, DJand Erikson, MF,1997 Bone changesin Natural History, Smithsonian Institute, Washing- the human skull probably resulting from scurvy ton DC, 71–83 in infancy and childhood, International Journal of Waters,L,1986 Rail Centres:Oxford,Ian Allen, London Osteoarchaeology 7 ,212–220 Withers, Pand BR,2005 Anglo-Irish coins John – Oswald, A, 1984, Clay Pipes,inExcavations in EdwardVI, SmallChange VI,Llanfyllin,Galata St Ebbe’s Oxford, 1967–76: Part 2: Post-medieval Willmott, H, 2002 Early post-medieval vessel glass in domestic tenements and the post-dissolutionsite England, 1500–1670 ,CBA Res.Rep. 132,York of the Greyfriars (T GHassall, CEHalpin and Willmott, H, 2005 Ahistory of English glassmaking, MMellor), Oxoniensia 49,153–266 AD 43–1800,TempusPublishing, Stroud Oswald A, 1991, Pipe stamp index, unpublishedmanu- Wilkinson, D(ed.), 1991 OAU fieldwork manual, script (4 vols), copy held at the National Clay Oxford Archaeological Unit, Oxford Tobacco Pipe Archive, Liverpool Zimmerman, MRand Kelley,MA, 1982 Atlas of OAU, 1995a Oxford Rewley Road (LMS) Station: Human Paleopathology,New York Investigation of FoundationStructure,Oxford Archae- ologicalUnitunpubl. report END NOTES OAU, 1995b Oxford Rewley Road (LMS) Station: Investigation of Foundation Structure to Station Porte 1. Cart. Osen .ii, 449 (1055). Coche` re,Oxford Archaeological Unit unpubl. 2. VCH Oxon iv, 279. report 3. Thehistory of theCistercians at Oxford is to be OAU, 1998 Former LMS Station, Rewley Road, Oxford: foundin VCHOxonii ,83; WHStevenson&HE Archaeological Investigation,Oxford Archaeological Salter, TheEarly HistoryofStJohn’sCollege [here- Unit unpubl. report inafter EHSJ]OHS NS 1(1939); H.G. Richardson Roberts,Cand Cox, M, 2003 Health and disease in ‘CistercianFormularies’, in H.E. Salter et al. Britain from prehistory to the presentday,Sutton Oxford Formularieswhich Bear on thehistory of Publishing, Stroud Oxford II O.H.S. NS v(1942), 281–327. Robinson, S, 1980 The tile,inABeakerburial 4. MWSheehan, ‘The Religious Orders 1220–1370’, and medieval tenements in the Hamel, Oxford (N in J. Catto (ed.) The History of the University of Palmer), Oxoniensia 45,196, microfiche 2.D09–D14 Oxford I: The Early Oxford Schools (1984).

103 From Studium to Station Rewley Abbey and Rewley RoadStation, Oxford

5. VCH Oxon ii,83; see below for seal. 39. Hearne Collections viii,349. 6. Cart. Osen .ii, 450–9 (105–68). 40. Hearne Collections xi,134. 7. JEThorold Rogers, Oxford City Documents OHS 41. Christ Church Archive, MS Estates 77, pp.117– 18 (1891),206–7 aa48 &53(from Bodl. MS. Twyne 118 (items 395 to 433). IV,617 et seq.). 42. Badcock’s plan and schedule reproduced in Cart. 8. Cart. Eynsham i ,308–9 (460–1). Osen. ii, 624–7. 9. VCH Oxon ii,82; Dugdale, Monasticon v, 699–700 43. MS Estates 77, items 416, 419 and 429. from Inspeximus,Pat. 22 Edw.I, m.16. 44. MS Estates 77, items 421 and 425. 10. VCH Oxon ii,82; Reg. Grandison, 1132. 45. British Library, Additional MS 15,546, fos 52–57 11. Cart. Osen. ii, 470–3 (1066) from MS. Twyne (online imagesatwww.collectbritain.co.uk). In xxii, 327. general see the Bodleian Library card index of 12. Cart. Osen. ii,459–60 (1059). Oxford Views. 13. Cart. Osen. ii,460–1(1060). 46. British Library, Additional MS 15,546, f. 54. 14. Salter, Mun. Civ. Oxon.,86–7 (98), 101 (114). 47. Hearne’s Collections iO.H.S. 2(1884),8. 15. EHSJ ,6;Denifle, Cartularium ii, 38. 48. Ibid., 47. 16. VCH Oxon ii,82; E.H.S.J.,6–7; AG Little in Eng. 49. ‘Medieval Epigraphy in the City &University’, Hist. Rev .viii (Jan. 1893), 84, from BMRoyal MS. no.7. 7Dxv f.64. 50. T. Squires, In West Oxford (1928),112–5; plate 17. EHSJ ;C.Henriquez, Regula Cisterciensis (1680), lxix. Clark Wood’s City of Oxford II, 300; Hearne 102–3. Collections I, 47, 70. 18. Richardson, ‘Cistercian Formularies’,in For- 51. GEC, Complete Peerage 12/2, 365. mularies ii,281–3 (from MS at Longleat). Oxford 52. Early Rolls Merton,27, 254, 261–2, 445–9. Theologians O.H.S., 225 n.4. 53. History of the University of Oxford I, 276; VCH 19. EHSJ ,10; Cal. Close Rolls, 17. Oxon ii, 150. 20. EHSJ ,11–12, and 62–6. 54. LTSmith, Leland’sItinerary i. (1906),125. 21. EHSJ ,13; C.H. Talbot, Letters from the Eng. Abbots 55. RCHM Inventory, Oxford (1939),47, pl. 108. to the Chapter at Cıˆ teaux 1442–1521 Camden4th 56. The Lives of those eminent Antiquaries John Leland, Ser. iv (1967),12ff forhistoryofcollege. Thomas Hearnand Anthony Wood (1772) I, pl. 22. Talbot 44–6 (§6), 144(§71), 198, 228 (§110). at 138–9; also reproduced by Skelton, Oxonia 23. Cart.Osen. vi,244; W.H. Turner, Records of the Antiqua (1823) opp.Pl. 117. City of Oxford (1880),12(Twyne ix). 57. PANewton, The CountyofOxford. ACatalogue of 24. VCH Oxon iv,365, quoting Some Oxon. Wills the Medieval StainedGlass (1979),225 (Yarnton (Oxon. Record Soc. 39), 40. A1–2). 25. Valor Ecclesiasticus ii,255. 58. It is beyond the scopeofthe current report to 26. VCH Oxon ii,82–3; L. and P. Hen VIIIx,387; xiii provide adetailed description of the Crystal (i),1520 (38) and xii (2), 1320; also Emden Palace building, asubject for which acopious BRUO. literature exists. 27. Turner Records,163; L&P Hen VIII xxi (2),pp157, 59. In the earlyyears of thedevelopment of the rail 334. network in Britain, oneofthe keypoints of 28. HMColvin et al.(eds), History of the King’s Works contention wasthatofrailgauge –the distance IV (1485–1660) (1982),132, quoting works betweenthe tworunning rails of thetrack.Two accounts (PRO,E36/244). distinct gaugesoftrack were commonly em- 1 29. Turner, Records,177,382, 384–7, 432, 434; ployed, broadgauge at 7ft ⁄4 in andstandard 1 Oxford City Council Acts 1583–1626 (OHS87, gauge at 4ft8⁄2 in preference beingbased up- 1927)256, 281, 325. on criteriaofperformance andcost-efficiency 30. Cart. Osen. iii,99–22. respectively. ARoyalCommission set up in 1845 31. Information gathered by Nicholas Doggett in to establishanorm decidedinfavourofthe use 1984 from Christ Church Archive, original leases of standard gaugeoverbroad gauge as the then (Oxford: Rewley), Estate Registers (C XX, 1–36), existing networkusing theformerwas signifi- and Book of Evidences (I C.2). cantlylargerthanthatusing the wider gauge 32. SeeT.Harmsen, Antiquarianism in the Augustan (confined principallytothe GWR).The GaugeAct Age:Thomas Hearne 1678–1735 (2000). of 1846 ruledthatnew tracks shouldbelaidto 33. Remarks and CollectionsofThomas Hearne 11 vols standard gauge,but allowedfor broadgauge (OHS2–72, 1884–1918). companiestocontinuetoexpandtheir networks. 34. Hearne Collections i ,8,47, 184. This resulted in the usefor aperiod of time of 35. Hearne Collections i ,241; LTSmith, Leland’s mixed-gauge line comprising threerunning railsto Itinerary i. (1906),124. accommodatebothwideand standard gauge 36. Hearne Collections i ,244. rollingstock.The GWRfinally convertedits whole 37. Hearne Collectionsii ,216, 387; iii, 43, 67, 314, 395; network to standard gauge in 1892. iv, 347; v, 110; vi, 18. 60. BuckinghamshireRailway Board of Directors Mi- 38. Hearne Collections vii,21; ix,356. nute Book1847–74 (PRO,Rail86/3),Minute§499.

104 OA Occasional Paper No: 16

61. Ibid., Minute §499. The tender documents forthe there was abuilt-up gutter,probably on the lines proposed station have by chancesurvived, and of the main roof gutters. The lower gutter would one(partial) drawing indicates abrick building have spanned between the columns, and sup- with iron elements, in usual Gothic/Italianate portedthe edge of the ridge-and-furrow roof railway style: Oxon Record Office 0159/F/1 adjacent to the main building. Reference to the (costs) and 0159/P/1 (drawing). Crystal Palace drawings shows that most of the 62. Ibid., Minute §507. roofs spanned with the main building axis. If the 63. Ibid., Minute §514. roof spans were at right angles to the main 64. Ibid. building, with lead flat gutters, would the buil- 65. Ibid., Minute §548. ders nothave simplified the construction to stop 66. In the following description of the structure, all the wing block glazing against the main building measurements will be first given in imperial sidewall, and directedall the rainwater to the units as this was the system usedinits construc- perimeter gutters, and saved the cost of the tion. Metric conversions are given for reference. expensive and non-standard column heads? This Bays have been numbered from 1(porte-coche` re) was acheap building being constructedona to 14, from south to north. fixed price as adesign-and-build contractwhere 67. Thenorthern ninebays of the trainshed structure the contractor took all the risk ’. were dismantled in 1969. 70. Thestock shown is aBletchley District32ft 68. These ventilation panels had been removed, at 0in. · 7ft9in. five-coach set with one 30 ft 1in. · leastwithin the areaofthe trainshed, by 1914 8ft0in. at the rearasastrengthener. Thecoa- (see Plate 19), thoughithas been suggested they chesstill have ogee-ended headstocks and the were reintroduced during WWII as a‘black-out’ old type of buffer, andthe rings along the cor- measure. nicefor the Harrison alarmcordare prominent. 69. Lance Adlamviews the roof differently and These carriage sets were in use on the Oxford writes: ‘The wing block roofs, if constructed branch line until the introduction of the new parallel to the main building, would have had Bletchley Districtbogiesets in 1906. These six- beam structures through the centre of the block, wheeler sets were notfinally eliminated until spanning onto the cast iron columns, and the April 1909. If the Harrison cord is notinuse, as drainage of the rainwater for the most part of the would appear to be the case, then the photograph roof would have been via these gutters and also would date to after 1901.Info: PhilipMillard, downthe threeoutercolumns to the wing LNWR Society. blocks. The internal columns have flanged out- 71. Reproducedin British Railway Journal vol. 8, no. lets to the lower-level outletswhich suggests that 72, 294–5.

105