1 3 MAR 2012 PURCHASED Natural History Museum Library THE NATURAL HISTORY MUSEUM j 1 3 MAR 2012 F 'U hOHASED GENERAL LIBRARY The Wiltshire Archaeological and Natural History Magazine Volume 105 2012 Published by The Wiltshire Archaeological and Natural History Society 41 Long Street, Devizes, Wilts. SN10 INS Telephone 01380 727369 Fax 01380 722150 Email: wanhs@wiltshireheritage.org.uk Website: http://www.wiltshireheritage.org.uk/ Founded 1853 Company No. 3885649 Registered with Charity Commission No. 1080096 VAT No. 140 2791 91 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE VOLUME 105 (2012) ISSN 0262 6608 © Wiltshire Archaeological and Natural History Society and authors 2011 Hon. Editor: Andrew Reynolds, BA, PhD, FSA, FSA Scot, FRHistS, FRGS. Hon. Assistant Editor: Stuart Brookes, BA (Hons), PGDTiLL, MA, PhD, FSA. Hon. Local History Editor: James Thomas, BA, PhD, FRHistS. Hon. Natural History Editor: Michael Darby, PhD, FRES Hon. Reviews Editor: Robert Clarke BA (Hons), Cert Ed, MlfL. Hon. Excavation and Fieldwork in Wiltshire Editor: Simon Draper, BA (Hons), MA, PhD (Dunelm.). Finds Liaison Officer: Katie Hinds, BA (Hons), MA. Editorial Assistant: Lorna Haycock, BA (Hons), PhD, FRHistS, Dip.ELH, Cert Ed. We acknowledge with thanks grants towards the cost of publishing specific papers in this volume from the following bodies: the National Trust, for ‘Geophysical survey of the northeast and southeast quadrants of Avebury henge’ and ‘Evaluation excavation at the Bury, Bishopstrow, Warminster;’ and Foundations Archaeology, for ‘Prehistoric and Roman archaeology at Mill Lane, Swindon. The journals issued to volume 69 as parts of The Wiltshire Archaeological and Natural History Magazine (Part A Natural History; Part B Archaeology and Local History) were from volumes 70 to 75 published under separate titles as The Wiltshire Natural History Magazine and The Wiltshire Archaeological Magazine. With volume 76 the magazine reverted to its combined form and title. The cover title ‘Wiltshire Heritage Studies’ (volume 93) and ‘Wiltshire Studies’ (volume 94 onwards) should not be used in citations. The title of the journal, The Wiltshire Archaeological and Natural History Magazine , remains unchanged. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Society and authors. Cover illustration: The Savernake Horn, image courtesy of Graham Bathe Typeset in Aldine by Stuart Brookes and produced for the Society by Information Press, Eynsham, Oxford, 0X29 4JB Printed in Great Britain Contents NATURAL HISTORY, ARCHAEOLOGY and LOCAL HISTORY East of Avebury: tracing prehistoric activity and environmental change in the environs of Avebury henge (excavations at Rough Leaze 2007), by Joshua Pollard, Michael Allen, Rosamund Cleal, Nick Snashall, Jim Gunter, Vaughan Roberts and David Robinson 1 Geophysical survey of the northeast and southeast quadrants of Avebury henge, by Martin Papworth 21 Prehistoric and Roman archaeology at Mill Lane, Swindon, by Andrew Hood 43 An Early Anglo-Saxon cemetery at Aldbourne, Wiltshire, by Nick Stoodley, Anthea Boylston and Jacqueline I. McKinley 58 Why does the north-western boundary of Wiltshire ignore the river Avon?, by Hannah Whittock 96 A Perambulation of the Hundred and Parish of Westbury, by Alison J. Maddock 105 Early to middle Anglo-Saxon settlement, a lost medieval church rediscovered and an early post-medieval cemetery in Wilton, by Rob DeAthe 117 The Romanesque doorway at Little Langford, by Rita Wood 145 The miracle of the horseshoe: a 15 th century wall painting at High worth Church, Wiltshire, by Ellie Pridgeon and Roger Rosewell 157 The Savernake Horn, by Graham Bathe 168 Forging alliances: the Life of Edward Seymour, Earl of Hertford, and his commissioning of the Great Illuminated Roll Pedigree of the Seymours and monumental tombs in Wiltshire and Westminster, by Graham Bathe and Andrew Douglas 1 82 John Buckler (1770-1851): the topographical artist in Wiltshire, by Robert Moody 219 The Heytesbury Collection: visitors to William Cunnington’s Moss House, by Lorna Haycock 225 The country-house social system in the 19th and 20th centuries, by F. M. L. Thompson 231 NOTES and SHORTER CONTRIBUTIONS Evaluation excavation at The Bury, Bishopstrow, Warminster, by Martin Papworth 240 Celts, the Sem, and Semington Brook, by Andrew Breeze 242 Books and pamphlets in the library of the Wiltshire Archaeological and Natural History Society printed before 1700, by Robert Moody 244 WANHS Archaeology Field Group: recent activities and future plans, by Jim Gunter 245 REVIEWS, edited by Robert Clarke 247 EXCAVATION and FIELDWORK in WILTSHIRE 2010, compiled by Simon Draper 253 Highlights from the Portable Antiquities Scheme (PAS) in Wiltshire in 2010, recorded by Katie Hinds 260 INDEX, by Philip Aslett 266 The Wiltshire Archaeological and Natural History Society The Society was founded in 1853. Its activities include the promotion of the study of archaeology (including industrial archaeology), history, natural history and architecture within the county; the issue of a Magazine, and other publications, and the maintenance of a Museum, Library, and Art Gallery. There is a programme of lectures and excursions to places of archaeological, historical and scientific interest, and an archaology field group for members who wish to become actively involved in the archaeological scene in the county. The Society’s Museum contains important collections relating to the history of man in Wiltshire from earliest times to the present day, as well as the geology and natural history of the county. It is particularly well known for its prehistoric collections. The Library houses a comprehensive collection of books, articles, pictures, prints, drawings and photographs relating to Wiltshire. The Society welcomes the gift of local objects, printed material, paintings and photographs to add to the collections. The Wiltshire Archaeological and Natural History Magazine is the annual journal of the Society and is issued free to its members. For information about the availability of back numbers and other publications of the Society, enquiry should be made to the Museum Shop Manager. Publication by the Wiltshire Archaeological and Natural History Society does not imply that the Society endorses the views expressed; the factual content and the opinions presented herein remain the responsibility of the authors. Notes for Contributors Contributions for the Magazine should be on subjects related to the archaeology, history or natural history of Wiltshire. While there is no fixed length, papers should ideally be under 7,000 words, though longer papers will be considered if of sufficient importance. Shorter, note length, contributions are also welcome. All contributions should be typed/ word processed, with text on one side of a page only, with good margins and double spacing. Language should be clear and comprehensible. Contributions of article length should be accompanied by a summary of about 100 words. Please submit two copies of the text (with computer disk if possible) and clear photocopies of any illustrations to the editors at the Museum, 41 Long Street, Devizes, Wiltshire, SN10 INS. A further copy should be retained by the author. The editors will be pleased to advise and discuss with intending contributors at any stage during the preparation of their work. When submitting text or graphics on disk, Word or Rich Text Format files are preferred for text, jpeg or tiff format for graphics. Contributors are encouraged to seek funding from grant-making bodies towards the Society’s publication costs wherever possible. Referencing: The Harvard System of referencing (author, date and page, in parentheses within the text) is preferred: e.g. £... one sheep and one dog lay close together (Clay 1925, 69)’. References in footnotes should be avoided if at all possible. Only give references which are directly applicable, repeating as little as possible. All references cited in the paper should be listed in the bibliography using the following style, with the journal name spelled in full, and the place and publisher of books/ monographs given : For a paper: PITTS, M. W. and WHITTLE, A. 1992. The development and date of Avebury. Proceedings of the Prehistoric Society 58, 203-12 (Note that in citations Wiltshire Archaeological and Natural History Magazine is abbreviated to WANHM) For a book or monograph: SMITH, I.F., 1965, Windmill Hill and Avebury: Excavations by Alexander Keiller, 1925-39. Oxford: Clarendon Press For a paper in a book or monograph: FITZPATRICK, A., 1984, ‘The deposition of La Tene metalwork in watery contexts in Southern England’, in B. Cunliffe and D. Miles (eds). Aspects of the Iron Age in Central Southern Britain, 178-90. Oxford: University Committee for Archaeology Endnotes can be used for specific information that cannot otherwise be comfortably incorporated in the main body of the text. Illustrations need to be clear and easily reproducible, the format and proportions following that of the Magazine. If possible, all original artwork should not exceed A3 before reduction. If not supplied as computer graphic files, drawings should be produced on drafting film or high quality white paper using black ink. Detail and lettering should not be so small that it will become lost in reduction. Mechanical lettering (dry transfer or computer generated) is preferred over hand lettering. Photographs should be supplied as good quality black and white prints, and transparencies and colour prints avoided wherever possible. Original illustrations and photographs should only be sent once a contribution has been accepted. Digital files (in jpeg or tiff format) should be of at least 300 d.p.i. resolution. Offprints: Offprints will no longer be supplied, but authors (or the first named author where there are more than one) will receive an electronic pdf. file of the article as printed. WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY SOCIETY (appointments as at 1 January 2012) PATRONS J.B. Bush, OBE, JP E Harding, DUniv, FSA, MIFA D. Inshaw BOARD OF TRUSTEES Chairman N.B. Harte, DSc(Econ), FSA, FRHistS Hon Treasurer E Glasson, FCA Other Elected Trustees D.J.M. Carson M.G. Pope C. R. Chippindale, BA, PhD, MIFA, FSA Prof A. Reynolds, BA, PhD, FSA, FSA Scot., FRHistS D. J. Field, PhD, FSA, FSA Scot D.L. Roseaman, BSc (Eng), C Eng, MI Mech E Ms A.M. Hemmins, BEd, ACE, CertEd P R. Saunders, BA, FSA, FMA, FRSA Mrs S.P Ingram Hill, BA (Hons), FIH, JP Co-opted Trustee W.A. Perry, MSc Nominated Trustees Cllr. C. Callow, BSc, ARCS, MBCS (Member Devizes Town Council) Cllr. C. Cochrane, BSc (Hons), PhD, DMS (Member Wiltshire Council) A.X.T. Green, BA, MA, AMA (Director, Salisbury and South Wiltshire Museum) Cllr. J.G. Ody (Member Wiltshire Council) In attendance: Ms V. Barlow, BA (Hons), PG Dip Museum Studies (Team Manager Conservation and Heritage, Wiltshire Council) STAFF Director Curator Education Officer Librarian & Archivist Assistant Curator Assistant Curator & Volunteer Co-ordinator Marketing & Administration Manager David Dawson, BA (Hons), FSA, AMA Lisa Brown, BA (Hons), MA Ali Rushent, BA (Hons), MA Kerry Nickels, BA (Hons) Heather Ault, BA (Hons) Vacant & Company Secretary Membership Services Officer Shop Manager & Visitor Services Officer Finance Officer General Assistant Karen Jones, BA (Hons) Katrin Bailey Diann Barnett, BA Malcolm Church Jacqui Davies Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 1-20 East of Avebury: tracing prehistoric activity and environmental change in the environs of Avebury henge (excavations at Rough Leaze 2007) by Joshua Pollard1, Michael Allen2, Rosamund Cleal3, Nick Snashall3, Jim Gunter4, Vaughan Roberts4 and David Robinson5 The results of a programme of geophysical survey, test pitting and excavation at Rough Leaze, immediately to the east of the Avebury henge, are here described. Intended to examine evidence for settlement and other activities pre-dating or contemporary with the henge, the fieldwork revealed a moderate density scatter of mostly Neolithic flintwork, colluvial build-up against the henge bank, stake-holes, and a series of Neolithic tree-related features, one of which had been modified. Molluscan analysis indicates that activity here during the early and middle Neolithic took place within a woodland setting. Other evidence relating to the pre-henge settlement history and environment is reviewed. Introduction We know relatively little about the extent and scale of Neolithic and earlier Bronze Age settlement and other kinds of non-monumental activity in the immediate environs of the Avebury henge. Were there areas of settlement linked to gatherings for construction and ceremony around the outside of the monument, or was this a zone of reserved space devoid of contemporary activity? Recent work at the later Neolithic henge of Durrington Walls - a monument that in its scale and multi- entrance format shares a number of similarities with Avebury (Wainwright and Longworth 1971, 198-203) - has revealed traces of an extensive late Neolithic settlement at the eastern entrance and under the henge bank around its eastern and southern circuit. Generated by episodic large-scale gatherings connected to the construction and ceremonial use of timber monuments within Durrington Walls, and even Stonehenge in its principal megalithic phase, the settlement traces comprise numerous houses with chalk marl floors and central hearths surrounded by unusually dense midden deposits, all dated to a period of several decades around 2500 cal BC (Parker Pearson 2007; Parker Pearson et al. 2006). These discoveries have highlighted the need to investigate the spaces around, rather than just within, the great henges, and to think of the broader context of congregation, residence and resource mobilisation associated with their creation. Evidence recovered to date suggests the immediate environs of the Avebury henge may well lack the kind of settlement traces found at Durrington Walls. Those parts of the monument that have been subject to excavation, within the western and south-eastern sectors, have not 1. Archaeology, University of Southampton, Southampton, S017 1BF; 2. Allen Environmental Archaeology, Redroof, Codford, Warminster, BA12 0NW; 3. Alexander Keiller Museum, Avebury, SN8 1RF; 4. TALITS, The Crown Inn, 60 Wilcot Road, Pewsey, SN9 5EL; 5. Archaeology, University of Central Lancashire, Preston, PR1 2HE 2 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE produced obvious concentrations of later Neolithic material culture (Gray 1935; Smith 1965, 224-9). Low density scatters of early and middle Neolithic ceramics and lithics are known from the interior and under the henge bank, but almost certainly relate to pre-henge activities; though the potential exists for some overlap with early phases of megalith erection (Gillings et al. 2008, 203). Whether or not contemporary settlement was present adjacent to the Avebury monument is a matter of more concern than just a search for the homes of the henge builders. It may additionally say much about the varied ontological status of these monuments: for example, whether they were identified as places connected to the celebrations of the living or to those of the ancestral dead (Parker Pearson and Ramilisonina 1998); or about how engagement with sacred architecture was mediated through structures of symbolic control (notions of ‘taboo’ space). Identifying similarities or differences in the range of activities around Wessex henge enclosures has the potential to provide useful information about their utilisation, ontological and cosmological status, which in turn should feed back into how we think of these sites as either inter-linked or distinct architectural categories. Here, we report on a minimally intrusive programme of fieldwork that aimed to address these issues. The work was undertaken during May and June 2007 in the small field (Rough Leaze) to the south of Green Street and immediately to the east of the Avebury earthwork and its eastern entrance (NGR SU 106700). Currently under pasture, the field occupies part of the northern slope of a slight SW-NE ridge running off the end of Waden Hill (Figure 1). The ground falls gently by c. 3m across the investigated area, off the ridge and towards the east entrance of the henge and Green Street. The solid geology comprises the junction between Lower and Middle Chalk. A combination of geophysical survey, test-pitting and targeted excavation was employed in order to identify areas of prehistoric activity. Specifically, we wished to know whether: traces of Neolithic and/or earlier Bronze Age settlement and other activity lay immediately adjacent to the henge earthwork; whether there were prehistoric features and deposits present within this zone; and how far beyond the Scheduled area at the base of the bank spreads of bank material and areas of protected buried soil might extend. A further concern related to how histories of pre-henge activity might have structured subsequent phases of monumentalisation. We were also cognisant of management issues relating to this area, given proposals to utilise this field as a temporary campsite for solstice gatherings at Avebury. If such usage was to be considered in the future, there was a real need to understand the impact that vehicle access and camping activity might have on any archaeological deposits. Between the Monuments This work is the first component of a larger project under the banner of Between the Monuments (Pollard et al. 2011) designed to enhance our understanding of the landscapes of Neolithic and earlier Bronze Age settlement, routine and environment within the Avebury region. It builds upon the work undertaken by John Evans and Alasdair Whittle on the region’s post-glacial environmental history and Neolithic archaeology (Evans et al. 1993; Whittle 1993; 1994; 1997; Whittle et al. 1993; 1999; 2000), and on that of the Longstones Project (Gillings et al. 2008) and the Stonehenge Riverside Project (Parker Pearson et al. 2006). The project aims to identify the range of practices that constituted routine and ceremonial life in the region during this period, their role in shaping social relations, and their relationship to monument construction and to natural constituents of the landscape. It is hoped it will act as a vehicle for new ways of theorising and interpreting landscape inhabitation, routine and non-monumental architectures during the period. Knowledge of settlement/residence and other non-monument-focussed practices during the region’s Neolithic and early Bronze Age remains poor ‘and the context in which the monuments developed and were used remains largely unknown’ (Cleal 2001, 41). The region has not been subject to the concerted programmes of extensive surface collection that have taken place in the environs of Stonehenge (Richards 1990) , the eastern Dorset Ridgeway (Woodward 1991) and around Thornborough (Harding 2000), for example. Systematic surface collection has only covered a small percentage of the ‘core’ area. The most ambitious programme of work remains that undertaken by Robin Holgate and Julian Thomas during the mid-1980s (Holgate 1987). To this can be added the work of Ian Dennis and Alasdair Whittle on the southern slopes of Windmill Hill in 1992-1993 (Whittle et al. 2000), and on an as-needs basis by Wessex Archaeology and the National Trust in advance of the conversion of areas of arable to EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 3 Fig. 1 Location of Rough Leaze and the area of fieldwork 4 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 2 Resistance survey results with selected anomalies indicated pasture. All of these programmes of surface collection involved sampling on an ‘extensive’ basis (i.e. collection along wide-spaced transects), providing a coarse-grained view of the extent of activity, but little specific detail on within-scatter compositional or chronological variability. Possessing even poorer spatial control, there also exist sizeable historic collections of lithics accumulated during the early 20th century (Holgate 1988) and more recent private collections held by members of the local community. Both refined method and sophisticated interpretation are fundamental to deal with this often recalcitrant data; and as Whittle rightly states the hard task may in fact be to ‘re-think both our expectations and our interpretations’ (Whittle et al. 2000, 177). Reflecting such concerns to re-think the fundamental questions, there have been recent and highly productive engagements with ‘difficult’ lithic scatter evidence that we wish to follow, ones which have sought to provide accounts of shifting residential choices and composition, the tempo of landscape occupation and the role of history and memory in structuring routine activity (cf. Snashall 2002; Edmonds et al. 1999). Fieldwork Methodology and Results Geophysical Survey In advance of test-pitting and targeted excavation, a resistance survey was undertaken across the area of the field in May 2007 by Jim Gunter and Vaughan EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 5 Fig. 3 Location of test pits (left) and trenches (right) Roberts of TALITS (see Gunter and Roberts 2007 for the detailed report from which this edited version derives). The work formed part of an on-going programme of geophysical survey in the Avebury landscape. The survey grid was first plotted using a Topcon Total Station 226. Resistance survey was undertaken over an area of 0.8ha (20 grids, each 20m x 20m) using a Geoscan RM15 twin probe array, employing a one metre zigzag traverse, with sampling at one-metre intervals. The results of the resistivity survey were analysed using Geoplot 3.00s software. Several features of possible archaeological origin were identified (Figure 2). The most striking is a broad, linear area of low resistance approximately 10m wide running roughly parallel to the line of the henge bank. Across the northern half of the survey area are at least six small, low resistance anomalies approximately 2m-3m in width, five of which looked to form a broken arc. It was thought these might relate to pits or other artificial features such as large post- or stone-holes. An unusual high resistance feature close to the centre of surveyed area forms a pattern that resembles a starfish (i.e. five linears each about 8-9m in length converging to a point), its regularity suggesting the feature might be artificial rather than natural. Finally, a series of high resistance linears running SW-NE across the area are most probably either geological or a product of agricultural activity. Test pitting and excavation In order to evaluate topsoil artefact densities, as well as the presence of archaeological features and colluvial deposits, a series of lxlm test-pits was hand-dug on a regular 20m interval following the grid established for the geophysical survey (Figure 3). All soil from the test-pits was carefully hand- sorted for artefacts and a 20% sample sieved in order to provide a control on recovery. 24 test pits were excavated (numbers 1-3 in the northern part of field and 27 in the SE corner were not excavated). Combined topsoil/subsoil depths ranged from 0.19m in the SE corner (TP23), to 0.84m 6 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 4 Surface topography (left: contour heights arbitrary ) and depth of soil above chalk (right) against the western field edge (TP9). The depth of brown earth colluvium under the topsoil increased dramatically across a 20m-wide band against the field edge, where soil had accumulated against the henge bank (Figure 4). This explains the seeming disparity in the height of the henge bank as observed from Rough Leaze (where it appears a relatively modest earthwork) as opposed to the view from the interior, where colluvial masking of its base has not occurred. The build-up of colluvium also contributes in part to the hollow-way effect of Green Street as it runs from the east entrance of the henge. Certain and possible stake-holes were observed cut into the chalk in TP 18, TP22 (three), TP23 and TP26 (five). 161 pieces of worked flint were recovered from the test pits (Figure 5). Lithic densities ranged from 0 to 15 pieces per test pit excluding mirco-debitage (pieces under 15mm in maximum dimension), and 0 to 21 pieces including micro-debitage. The test pit average was 5.7 pieces; a little over 80% producing between three and nine pieces. The greatest density of worked flint occurred in TP8 in the north-eastern part of the field. Otherwise densities were moderate to high (5+) in the western-most test pits where colluvial build up was also greatest, suggesting that density here was closely correlated with soil volume, and in the south-eastern corner. Burnt unworked flint came from TPs 6-8, 13, 14, 16-18 and 26; individual test pit densities only being within the range of one to three pieces. One small sherd of prehistoric pottery was recovered from the lower soil in TP24. Among the lithics were two scrapers, two flake knives, a piercer and a retouched flake. In total, tools account for a little over 5% of the assemblage, excluding micro-debitage. The majority of the debitage comprises hard-hammer struck flakes typical of 3rd and 2nd-millennium BC assemblages. A small number of bladelets from TPs 8 and 13 look to be later Mesolithic or earlier Neolithic. Guided by the results of the geophysical survey and identified topsoil artefact concentrations, five trenches were hand-excavated (Figure 3). Trench 1 (10x2m) was sited to investigate the broad low resistance anomaly and colluvial build-up against EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 7 o Stake hole V Direction of stake hole slope 0 1 1 1 1 1 2m _j 1 Stake Hole Dimensions Diameter (mm) Depth (mm) Diameter (mm) Depth (mm) 1 60 10 12 60 90 2 80 170 13 80 200 3 50 130 14 60 100 4 40 130 15 70 120 5 50 80 16 60 140 6 60 90 17 50 80 7 50 160 18 40 180 8 50 150 19 70 80 9 40 100 20 60 70 10 50 80 21 40 110 11 50 70 22 60 80 23 50 60 Left: Fig. 5 Distribution of worked flint from test pits Right: Fig. 6 Stake-hole cluster in Trench 4 the henge bank. Trenches 2 (5x5m) and 5 (5x2m) were intended to investigate pit-like low resistance anomalies in the northern half of the field; and Trench 3 (5x2m) the high resistance ‘starfish’ anomaly in the centre of the area. Trench 4 (5x5m) incorporated TP22, where the combination of a moderate density of lithics and several stake- holes suggested that features relating to Neolithic occupation might be present. No archaeological features or deposits were present in Trench 3, the geophysical anomaly seeming to correspond with an area of higher and more solid chalk. In Trench 4, 23 certain or probable stake-holes were revealed, distributed relatively even over all but the north- west quarter of the trench (Figure 6). There was no coherent pattern in their distribution, suggesting they did not form part of a single coherent structure, but five groups of three to five stake-holes look to describe shallow arcs that could relate to erstwhile shelters or insubstantial houses. The broad, linear low resistance anomaly in Trench 1 was found to be the product of substantial soil accumulation. Under the 0.2m deep topsoil (001) was a thick deposit of sorted orange-brown silty clay colluvium (002). This comprised upper (paler) and lower components; in total 0.3m deep at the east end of the trench, thickening to 0.65m at the west where it was close to the tail of the henge bank (Figure 7). Among the lithics from the base of this layer was a barbed-and-tanged arrowhead and broken fabricator. Most likely deriving from erosion of the bank, a thin (<0.1m thick) lens of chalk rubble (033) was present at the base of this deposit at the west end of the trench, extending for a distance of c. 3m. A large tree-throw pit, FI, was discovered in the eastern third of Trench 1, sealed by the colluvium (Figure 8). With a north-south axis, both ends of this feature lay outside the confines of the trench. The pit [024] was 1.9m wide and up to 0.7m deep, the profile dished and slightly asymmetric. The lower fill (023) was of compact chalk rubble within a pale silty clay. Above this, and largely confined to the sides, was a pale beige silty clay (022); in turn overlain by a friable orange-brown silty clay with THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE KEY Dark brown clay loam Orange-brown silty clay Molluscan column Chalk Silt TV TV chalk (003). (010) was a thin layer of brown silty clay with moderate quantities of small chalk, restricted to the top of the fill, and present across all but the eastern edge. Three small sherds of pottery, one of Peterborough Ware, another grog-tempered, came from (003) and (010), along with 17 pieces of worked flint (Table 1). Other tree-related features were present in Trenches 2 and 5, in both instances corresponding to geophysical anomalies. F2 in Trench 2 was a sub-oval pit [027], 3mx2m in extent and up to 0.45m deep, with sloping sides and an irregular base, probably representing a tree hollow; i.e. the root-weathered hollow. It was fully excavated. The primary chalk- dominated rubble (032) was largely restricted to the west side. The main fill (011) of silty clay with sparse chalk was much disturbed by root and animal action in the upper profile. From this came over 30 pieces of sarsen (up to 0.2m across), many fractured and some burnt, and loosely concentrated against the feature edge, possibly representing a displaced hearth setting. The same context also contained 28 pieces of flint debitage, mostly characterised by hard-hammer struck flakes and micro-debitage, but also including a fragment of blade core and one possible Mesolithic bladelet (Table 1). Seven small West East FI, north facing section Fig. 7 Trench 1 following excavation showing depth of colluvium against henge hank and tree-throw pit FI Fig. 8 North facing section of FI showing position of mollusc column EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 9 West F1 1 , south facing section East TV (028) KEY Topsoil Dark brown clay loam Clay loam and chalk Brown clay loam ip sin ■ mi J Molluscan sample [I£°| Chalk Disturbance 0 2m TV Section + Bos prim, bone 0 2m Fig. 9 Section and plan of tree-throw pit FI 1 in Trench 5 sherds of Peterborough Ware were present within (Oil), another coming from an initial surface clean of the feature. Fll in Trench 5, another tree-throw pit, was similar in morphology to FI (Figure 9). An elongated cut [026] extended beyond the sides of the trench. Up to 2.1m wide and 0.7m deep, its profile was asymmetric. A very regular circular cut in the base, 0.15m in diameter and 0.1m deep, was interpreted as a small post-hole [030] filled with a brown loam (031). The level from which this was cut was not entirely clear, although it was only recognised when the bulk of the lower fill had been removed. Situated on the base immediately adjacent to, if not within the fill of this, was a substantial fragment from the distal end of a right humerus of Bos primigenius. No other bone was recovered from the primary fills of this or the other tree-throw pits, and the most 10 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 1. Worked flint from feature fills Flakes Rejuvenation flakes Cores Chips (<15mm) Retouched Total FI (003) 10 2 - 3 1 retouched 16 (010) - 1 - - - 1 F2 (Oil) 1 10 2 1 r^n 15 I - 1 28 Fll (025)/(034) 9 - 1 - 2 utilised 12 Total 29 5 2 18 3 57 Table 2. Worked flint from non-test pit soil contexts Flakes Prep. flakes Rejuv. flakes Cores Chips Retouched Total Tr.l (001) 1 - . - - - 1 (002) 8 - - - 4 5 (1 barbed-and-tanged arrowhead, 1 fabricator, 1 notched blade, 2 retouched) 17 Tr.2 (001) 8 1 1 1 - 1 (leaf arrowhead frag.) 12 (002) 4 1 - - 6 1 (notched flake) 12 Tr.3 (001) 2 - - - - 3(1 scraper, 2 notched blades) 5 (002) 5 . - - 2 2 (1 knife, 1 notched blade) 9 Tr.4 (001) 40 7 3 1 20 8 (3 scrapers, 2 microdenticulates, 2 retouched flakes, 1 chisel arrowhead) 79 Tr.5 (001) 1 - - 1 - - 2 Total 69 9 4 3 32 20 137 parsimonious explanation is that the bone was set within the post-hole fill. A radiocarbon date of 4230- 3660 cal BC (95.4%) or 4058-3969 cal BC at 90.4% probability (5227±25BP; NZA-37435) was obtained from the Bos bone. The lower fill (028) was of a pale brown loam with only limited amounts of small chalk towards the base and sides, especially on the east; this being material that had fallen from the primary root plate. Within this, small chalk lenses and some small sarsen were present. The main fill (025), a mixture of the B horizon (subsoil) and soil eroded and developed in situ , was a soft and friable silty clay loam with chalk, more frequent on the west side of the feature. This was succeeded by a darker soil on the east side (034), which had been subject to limited animal disturbance. From (025) and (034) came a small quantity of sarsen, a fresh sherd of earlier Neolithic Plain Bowl pottery and 12 pieces of worked flint (Table 1). The lithics included two utilised flakes, a core fragment and several thin and narrow debitage flakes in very fresh condition, some probably removed by soft hammer. All may relate to a brief episode of earlier Neolithic activity that was confined to the hollow created by the tree-throw. 137 pieces of worked flint came from the topsoil and lower ploughsoil (Table 2). Although recovery was not as systematic as that applied to the test pits, there was notable variation in density, with only two pieces of worked flint (or 0.2 per m2) from Trench 5, between 14-24 pieces (or 1.0- 1.8 per m2) in Trenches 1, 2 and 3, and a high of 79 (or 3.3 per m2) in Trench 4. Diagnostic pieces span the early 4th to early 2nd millennia BC and include the barbed-and-tanged arrowhead from Trench 1, a fragment of leaf-shaped arrowhead from Trench 2, and a chisel arrowhead and flake from a discoidal ‘Levallois’ core from Trench 4. The flake debitage from the latter included numerous hard hammer struck pieces typical of later Neolithic industries. A small single platform core with scars from narrow flake removals from Trench 2 is likely to be of 4th- millennium BC date. Comprising 14.6% of the total topsoil/lower ploughsoil assemblage, the percentage of tools and utilised pieces is high and perhaps indicative of occupation-related activities. This is quite distinct from the figure of 5.2% from the feature fill assemblages, heavily influenced by the quantity of debitage in F2. EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 11 Auger testing Other geophysical anomalies scattered across the northern half of the field (‘a’-‘d’) were investigated by auger (Figure 2). The base of anomaly ‘a’ was reached at a depth of 0.75m (i.e. cut c. 0.5m into the chalk), the feature being filled with a pale silty clay, increasingly chalky in its lower profile. Both ‘b’ and ‘c’ contained homogenous deposits of pale silty clay with chalk; the base of ‘b’ being reached at 0.65m (cut c. 0.4m into natural), while a block of sarsen or flint was encountered at 0.9m in ‘c’. No cut feature was identified in the area of anomaly ‘d\ A further small scale programme of auger transect and probablistic augering was undertaken immediately to the west of Rough Leaze and south of the henge bank (Allen and Snashall 2009a; 2009b). Augering demonstrated the presence of shallow (c. 0.5m) colluvium against the henge bank. Close to the bank colluvium up to 0.65m depth was encountered. At 0.4m depth was a layer of compacted cemented chalk marl with a number of fine chalky pieces, which contained sub-microscopic pieces of what looked like pottery. The deposit strongly resembles a chalk made surface that has been preserved because of the depth of protecting hillwash. Although this could be a recent or older chalk trackway, the similarities of the matrix of the material to, and its location just outside the henge, is strongly reminiscent of the late Neolithic house floors found at Durrington Walls and Marden; both close to the henge bank where protection has been afforded by hillwash or other cultural deposits. Land-use and environment: mollusc evidence from the tree -related features A series of seven samples were taken during the excavation for land snail analysis from the tree- throw pits (FI 1 and FI) and one shallow pit feature (F2). One is of early Neolithic date with Plain Bowl pottery and a radiocarbon date of 4230-3660 cal BC (Fll), while two others (FI and F2) have Peterborough Ware pottery in their upper fills. Possible Mesolithic flintwork considered to be Avebury; Rough Leaze ■ III I I I I I I woodland/ open woodland F11 tree hollow ° =°25_ c 181 ■ an Mil J325 \&° 381 ■ ■ ■ IIII 0 10 0 30 0 30 0 0 10 0 30 0 1 — 1 i_ 1 i . i_ 1 — I I + + + □ + + + 0 Fig. 10 Rough Leaze: Land snail histogram of relatively abundance from three throw pits and subsoil hollow 12 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 3. Land snails from tree-related features from Rough Leaze (ARL 07) Tree throw pit | Tree throw pit | Tree-hollow Trench 5 5 Feature Fll Fll Context 025 025 Depth (cm) spot 1 spot 2 Wt (g) 1000 1000 MOLLUSCA Pomatias elegans (Muller) 57 19 Carychium tridentatum (Risso) 68 22 Carychium spp. 22 5 Cochlicopa cf. lubrica (Muller) - 1 Cochlicopa spp. 6 4 Columella edentula (Draparnaud) Vertigo pusilla Muller Vertigo pygmaea (Draparnaud) - 1 Vertigo cf. alpestris Alder Vertigo spp. - 2 Pupilla muscorum (Linnaeus) - 5 Lauria cylindracea (Da Costa) cf. 2 cf. 1 Vallonia costata (Muller) 5 6 Vallonia cf. excentrica Sterki - 6 Vallonia spp. 1 1 Acanthinula aculeata (Muller) 24 4 Ena montana (Draparnaud) 2 Merdigera obscura (Muller) 2 1 Punctum pygmaeum (Draparnaud) 1 Discus rotundatus (Muller) 69 51 Vitrina pellucida (Muller) - 1 Vitrea crystallina (Muller) 5 Vitrea contracta (Westerlund) 11 8 Nesovitrea hammonis (Strom) - 1 Aegopinella pura (Alder) 21 4 Aegopinella nitidula (Draparnaud) 15 11 Oxychilus cellarius (Muller) 11 6 Limacidae 3 6 Euconulus cf. fulvus (Muller) Cecilioides acicula (Muller) 18 33 Cochlodina laminata (Montagu) 2 Macrogastra rolphii (Turton) Clausilia bidentata (Strom) 17 5 Candidula intersecta (Poiret) - 1 Candidula gigaxii (L. Pfeiffer) [1] Cemuella virgata (Da Costa) 1 Helicella itala (Linnaeus) 1 2 Trochulus striolatus (C. Pfeiffer) 1 Trochulus hispidus (Linnaeus) 23 4 Arianta arbustorum (Linnaeus) + Helicigona lapicida (Linnaeus) Cepaea nemoralis (Linnaeus) Cepaea hortensis (Muller) 1 Cepaea spp. - 3 Cepaea/ Arianta spp. 9 Ostracod valves Taxa 26 23 TOTAL 381 181 residual was found in F2. All features were sealed by colluvium or a colluvial brown earth. Samples taken by the excavators were spot samples (F2 and Fll) or at crude (20cm) contiguous intervals through the deeper tree-throw pit FI (see Figures 8 and 9). Samples of 1000-1500g were processed following 1 1 1 1 2 FI FI FI FI F2 023 003 003 010 Oil 60-73 40-60 20-40 0-20 Basal 10cm 1500 1500 1490 1500 1500 20 62 42 29 28 18 151 91 19 36 2 18 23 1 7 - - - - 5 - 7 1 7 - 7 _ _ _ 3 - 2 1 - 1 1 2 4 6 7 2 - cf. + cf. 1 - - - 5 3 3 15 - 1 4 7 6 - - 1 - 3 3 6 5 5 3 - - 2 1 1 - 1 2 - - 1 2 - - - 8 138 92 49 72 - 2 - - - - 15 11 3 1 2 7 5 1 5 _ 8 16 5 9 1 16 8 9 1 1 16 17 3 11 - 3 2 - 2 1 7 _ 2 8 13 1 - 5 5 1 - - cf. 1 - 2 - 38 26 8 27 - 2 2 2 5 - 1 - - 4 - 12 5 7 23 - + - + 1 1 12 8 4 5 1 11 24 24 18 26 60 530 386 168 288 standard methods (Evans 1972); nomenclature follows Anderson (2005) with ecological groupings after Evans (1984). The results are present in Table 3 and as histograms of relative abundance in Figure 10, where other shade-living species include Vertigo pusilla, Vertigo cf. alpestris, Ena montana, Merdigera EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 13 Table 4. Occurrence of rarer woodland species from earlier Neolithic features from Rough Leaze and other subsoil hollows and Neolithic pits Feature Source Date cal BC Molluscs per kg Ena montana Vertigo alpestris Vertigo pusilla Columella edentula Lauria cylindracea 2005: Avenue pit Allen & Davis 2009 3300-2900 30-65 Y - - - - 1982: tree hollows Evans et al. 1985 pre-2600 232-720 Y Y Y - - 1969: A\C and subsoil hollow Evans 1972 pre-2600 42-604 Y Y Y Y Y 2007 : tree hollows This paper 4230-3660 to 3400 40-381 Y Y Y Y Y obscura , Acanthinula aculeata, Euconulus fulvus, Columella edentula, Trochulus striolatus and Helicigona lapicida and other catholic species include Punctum pygmaeum, Vitrina pellucida, Nesovitrea hammonis, Limacidae, Cochlicopa spp., Cepaea spp. Shell numbers were moderate with 40-381 shells per kilogram. All seven assemblages were dominated by shade-loving species (65-79%) and open country species only ever achieve a maximum of 11%. They are dominated by Carychium tridentatum, Zonitidae, Discus rotundatus and Trochulus hispidus. The assemblages are species-rich and ecologically diverse, with a number of rare woodland species present (Tables 3 and 4). The anthropophobe Ena montana , typical of ancient undisturbed woodland, was present in all features. Although the rare Vertigo species V alpestris and V. pusilla typical of ancient woodland are present in subsoil hollows and horizons below the buried soil under the henge bank in the 1969 and 1982 excavations (Evans 1972; Evans etal. 1985), they were only present in tree hollow F2, but Lauria cylindracea, a rupestral species, was present in all features at Rough Leaze except F2. These species are noteworthy as their demise is a result of removal or absence of woodland, but they do persist in low numbers in the later Neolithic and occasionally later prehistory in Wiltshire. Overall this provides evidence of deciduous woodland with leaf litter and possibly fallen logs. Although no evidence of open country conditions was present - a contemporaneous or successive turfline was not present - there are hints of slightly more open conditions in the upper samples of all three features. This is represented by a moderate increase in Vallonia costata and Pupilla muscorum suggesting perhaps more open woodland conditions. The results here concur well with the tree hollows examined by John Evans from the cuttings through the henge bank in 1969 (Evans 1972, 268-74) and 1982 (Evans et al. 1985), as well as the nearby pit from the West Kennet Avenue (Allen and Davis 2009). Where our sequences lack the upper profile containing a ‘clearance phase’ and buried soil with an established grassland, they are unique in being associated with Mesolithic, earlier Neolithic and middle Neolithic activity. This then begs the question of in what environment were these activities taking place, and when was the woodland cleared? If the artefacts are contemporary with the deposits and the mollusc assemblages as we assume, then this indicates prehistoric activity occurred within a shady, essentially woodland, environment prior to clearance. This is perhaps initially counter-intuitive, but the fact there is an internal chronology and stratigraphy (e.g. FI 1), and that some variation and possible modification of the woodland environment is suggested in three-throw pit Fll, certainly indicates the woodland conditions continued to prevail during the episodes of human activity from the Mesolithic to middle Neolithic at least. If earlier activity had been associated with clearance we might expect the assemblages, particularly in FI, to reflect this, and this to be noted between the earlier and later features (e.g. FI and F2). We can only conclude that the limited, but significant, prehistoric activity here occurred within woodlands prior to the later locally extensive clearance required for the construction of the henge at about 3000-2600 cal BC. There is no reason why a whole range of localised and repeated activities cannot have occurred within woodlands and be reflected in artefact scatters, nor indeed to have concentrated within natural tree-throw pit hollows which may afford shelter, access to flints, and so forth (see Evans et al. 1999; Allen and Gardiner 2009). Indeed the tree-throw pit itself indicates the presence of Pnaturally fallen trees and some slightly more open conditions within the woodland that need not be reflected in the molluscan assemblages due the ‘ecological island’ effect and lack local of source environments from which they could migrate (Thomas 1985). The felling of the woodland must have been planned and relatively prolonged (there is no evidence of fire) to enable the removal of the trees, the timber, and of the main stumps too, as well as 14 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE the establishment of pre-monument ungrazed, but established grassland (and its invasion by more open country snail species). Estimating the date of woodland clearance is difficult, but through examination of the previous data we can see a clear, but undated, clearance phase above the subsoil hollows and beneath the turfline buried by the Avebury henge bank (Evans 1972; Evans et al. 1985). Open woodland undoubtedly existed to 4000 BC and probably c.3350 cal BC during the currency of Peterborough Ware c. 3400-2900 BC. The snail evidence from the pit on the line of the West Kennet Avenue dated to the latter part of the calibrated range 3090-2910 cal BC (Allen and Davis 2009), however, indicates the presence of established dry ungrazed grassland existing in a mosaic of local ecological habitats in, probably, relatively recently cleared woodland. We can tentatively postulate clearance of woodland in the south-eastern areas of the Avebury henge circuit after about 3400 cal BC, but before 2900 cal BC. We have yet, however, to prove the full occurrence, distribution and nature of woodland, here noting the presence of natural open areas that encouraged pre-monument building activity in Cranborne Chase, Dorchester and Stonehenge (Allen and Gardiner 2009). Discussion The fieldwork identified scatters of worked flint that included material of possible late Mesolithic and certain early Neolithic to early Bronze Age date, a series of early Holocene tree-throw pits containing small quantities of artefactual material within their upper fills, and one location where there exists a concentration of stake-holes likely associated with prehistoric activity. Ephemeral sets of signatures of this kind can be seen to constitute the norm in terms of the register of Neolithic and early Bronze Age residence both on and off the chalklands of southern England. It would be easy to write-off as unremarkable the results of this work. But it is the very typicality of such traces, combined with the difficulty we seem to encounter in making sense of them, that should drive us to work harder and to start to tease out what they might represent; and this is what we attempt to achieve here. These are, after all, the very traces of routine lives and landscape dwelling. The distribution of worked flint recovered from the test pits and topsoil within the trenches is by no means even. Some downslope movement of material in the ploughsoil has probably taken place and may account for concentrations against the western edge of the field, but localised high densities of 7+ pieces of worked flint per test pit in the east and south of the field (e.g. TPs 8, 18, 23 and 26) probably do reflect a sense of original distribution. This spatially varied pattern, along with the chronological range of the material, indicates that the scatter is unlikely to have resulted from a single phase or episode of activity. It is therefore necessary to attempt to disentangle its individual components. The earliest activity may be marked by bladelets from TPs 8 and 13 in the east of the field. In terms of their technology of production - soft hammer struck and displaying careful platform preparation - there is a strong possibility they relate to late Mesolithic activity; which is rare but nonetheless present within the region (Whittle 1990). It is likely that this material was generated by short- stay events such as hunting forays {ibid.). An early Neolithic presence is indicated by the leaf-shaped arrowhead fragment from Trench 2, and the plain bowl sherd and flintwork from the tree-throw pit Fll in Trench 5. Small sherds of middle Neolithic Peterborough Ware came from a restricted zone in the north-west of the field, notably from pit F2 in Trench 2. The chisel arrowhead and Levallois-style core from Trench 4 could be contemporary with the Peterborough Ware related activity or belong later: both types are well represented at the middle Neolithic West Kennet Avenue occupation site (Smith 1965, 233-43), while Grooved Ware and chisel arrowhead associations occur at the Sanctuary (Cunnington 1931; Pollard 1992). The barbed-and- tanged arrowhead from Trench 1 hints at an early Bronze Age presence, though on an uncertain scale since no other flintwork can be confidently assigned to this period. Of the other tool types, items like the fabricator from Trench 1 and microdenticulates from Trench 4 could belong anywhere in the 4th and 3rd millennia BC (Smith 1965, 91, 93, 239, 241). In relation to the questions that structured this fieldwork, traces of Neolithic and/or earlier Bronze Age activity have been located within close proximity to the henge earthwork. They do not, however, constitute evidence of the kind and scale of settlement found at Durrington Walls, even if at least some of the traces are the residue of settlement/ residential activity. Some at least of this activity, and perhaps a good proportion of that in the north-west corner of the field, pre-dated the creation of the henge earthworks. If there is settlement related to the construction and use of the henge it might best be EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 15 represented by the flintwork and stake-hole scatter on the higher ground to the south-east around the area of Trench 4 and TPs 22 and 26, but without a precise handle on the chronology of events this remains ambiguous. Features included a number of tree-throw pits and a tree hollow. Although produced by natural agencies (i.e. the former by tree fall through storm activity and/or root instability (FI and Fll) and the latter by in situ rooting and weathering (F2)), the presence of soil material containing artefacts within the tops of those excavated demonstrates their presence as visible hollows and landscape features into the 4th, 3rd and 2nd millennia BC. Simply because we might label these features as ‘natural’ we should not ignore their capacity to structure in various ways human activity; acting as foci for deposition, as landmarks, and even as architectural affordances that could be modified to provide dwellings and shelters for fires and other activities (Evans et al. 1999; Lamdin-Whymark 2008, 80-100; Pollard 2006). In the case of FI 1 there is evidence for modification, with a post-hole cut into the base soon after the feature formed, into which was probably incorporated a split Bos primigenius humerus. The date of this at 4058-3969 cal BC at 90.4% would place it very early in the region’s Neolithic, if not the latest Mesolithic (Whittle et al 2011). Unfortunately, the identity of those responsible for the modification to the tree-throw pit and deposition of the bone is unknown, and there will always remain a possibility that the bone was curated. A small quantity of worked flint and early Neolithic pottery was later deposited in the top of this feature. That this activity was specific to the tree-throw is confirmed by the virtual absence of worked flint from topsoil contexts around it. In the case of tree hollow F2, Peterborough Ware sherds in the disturbed upper fill were associated with worked flint and a scatter of fractured and burnt sarsen tentatively interpreted as hearth stones. The lithics were all pieces of debitage, and included a good quantity of micro-debitage indicative of in situ working. It is unfortunate that the fills were so disturbed since here the hollow may well have acted as the base for a shelter. Past ploughing has resulted in a considerable accumulation of colluvium in the dip against the western side of the field formed by the base of slope and the henge bank. The inception of this is uncertain, but it evidently post-dates the creation of the bank, and given the way the northern edge to the accumulation is defined by the southern side of Green Street, is most likely medieval or early post-medieval in date. Fowler has suggested that ditches off-alignment to the medieval plots visible on a geophysical survey within the Avebury henge could represent fragments of prehistoric fieldsystem (Fowler 2004, 134-6), but this remains to be verified and no trace of such was encountered in Rough Leaze. Activity Density Our initial perception of the test-pitting results was one of a light prehistoric presence. Lithic densities ranged from 0 to 15 pieces per test pit (including micro-debitage, up to 21 pieces), which at first sight hardly seemed to register as a dense concentration. However, this may be deceptive. Only one of the test pits failed to produce worked flint, suggesting a relatively continuous trace of prehistoric activity across this area. Comparison with other scatters within the region is also instructive. Only 5% of the test pits on the southern slopes of Windmill Hill, often taken as an area with one of the richest scatters in this landscape, had more than 15 pieces of worked flint (Whittle et al. 2000, 151). Rough Leaze may therefore be characterised in its local context as a medium-high density scatter. Despite wide spatial coverage, test pitting covered under 0.25% of the area. Scaling up, with 132/161 pieces (excluding/including micro-debitage) of worked flint recovered the topsoil and lower ploughsoil, the total lithic population within the field could easily fall within the range of 50,000-70,000 pieces. The scatter here suddenly looks much more significant. This is, of course, a cumulative register - a product of multiple episodes of activity over two millennia or more that individually could have varied greatly in duration and intensity, from single, task-specific events to settlement episodes of a few years’ duration. The density of worked flint is nowhere close to that at Durrington Walls or some of the excavated later Neolithic scatter sites within the Stonehenge landscape such as Wilsford Down and Coneybury (with values over 50 pieces per square metre in places: Richards 1990, figs 95 & 116). Within the Durrington settlement midden worked flint values occur within the range of 1 13-149 pieces per square metre square, and regularly in excess of 42 (Chan 2010); this from an occupation that lasted for under half a century. Clearly the character of residential practices or the intensity of flintworking varied considerably between the two regions at points during the Neolithic. There is one known location within the environs of Rough Leaze where lithic densities start to come 16 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE close to those at Durrington Walls, Wilsford Down and Coneybury. While difficult to quantify because figures are only given for implements rather than totals that include debitage, densities within the West Kennet Avenue occupation site midden, c. 800m to the south, were very high, perhaps reaching over 40 pieces per square metre if implements formed a nominal 5% of the assemblage (Smith 1965, 237). While the material here had probably accumulated over a considerable period (Pollard 2005), both ceramics (including Fengate vessels) and distinctive lithic types such as chisel arrowheads and edge-polished implements point to a concentration of activity in the middle Neolithic (Smith 1965, 233-43). What might the Rough Leaze assemblage represent in terms of duration and scale of activity? In speculative fashion, one could hypothesize that a single residential group (a ‘household’ or equivalent) might produce around 50-100 pieces of worked flint per month, equating to the working of two or three cores and the production, use, discard and replacement of a range of light tools. The estimated field total would then relate to c. 1000 months of occupation, or a little under a century of activity. Given that the chronological range of material recovered is wide, being spread over at least two millennia, that activity cannot have been continuous, and we must envisage intermittent occupation and use of the area. Conceivably, there were periods during the Neolithic when very little was going on here, and other times of relatively busyness. There are, of course, caveats to such modelling of the evidence. We should acknowledge that the size and composition of residential groups would fluctuate; and it may be that more expedient later Neolithic flintworking strategies would have generated greater quantities of worked flint per episode, thus pulling down the figure for total duration. What is significant about this activity is that it seems to have been confined to woodland, albeit with highly localised openings not even large-enough to be called glades. The fact people were producing and using flints may in itself suggest that the woodland canopy was open enough to allow light, and the tree stands were not so dense as to prevent or restrict activities. Whether the activities were woodland- specific, or even whether other more open grassland coeval with this existed is yet to be determined. The local context Other scatters of worked flint and ceramics, some associated with small cut features, are known within the immediate environs of Rough Leaze, particularly those areas of the Avebury henge and northern section of the West Kennet Avenue which have been subject to excavation (Figure 11). The West Kennet Avenue occupation site has already been mentioned. In addition to the dense midden spread of largely middle Neolithic artefacts, two later Neolithic pits and several possible post-holes were encountered during the course of Keiller’s excavations here (Smith 1965, 212-6). To the north, the feature identified as stone-hole 15b of the Avenue may be a further late Neolithic pit (Gillings et al. 2008, 132); while a small pit containing sherds of Mortlake bowl was encountered close to this during cable work near stone 16a (Allen and Davis 2009). The latter was dated to 3090-2910 cal BC; with mollusca indicating a predominantly open yet still mosaic environment. The two pits occupy the head of a ridge which extends from the north-east of Waden Hill to the edge of Rough Leaze, and it would be useful through future work to establish whether these features and the concentration of middle/laie Neolithic flintwork in the south-east corner of Rough Leaze are on the periphery of an extensive connecting scatter. Though limited in scale, investigations of the buried soil under the bank of the Avebury henge have consistently produced traces of pre-monument activity, though varying in date, density and kind. The closest investigation to Rough Leaze was undertaken by Henry Meux in 1894 (Smith 1965, 184). There is little detail relating to this work, but surviving finds are presumed to be from the buried soil and include a chisel arrowhead, scraper and utilised flake (ibid., fig. 76). Gray’s Trench X bank excavations c. 200m to the south-west of Rough Leaze and on a similar contour, produced cattle bone, charcoal of hawthorn, chestnut and hazel in ‘some quantity’, over 40 sherds of Peterborough Ware and nine possible sherds of Grooved Ware (1935, 140, 158, 160; Smith 1965, 224). Among the 75 pieces of worked flint recovered (a density of 1.1 per square metre) were five scrapers, two serrated flakes, one combined scraper/serrated flake and one fragmentary bifacially flaked piece (Gray 1935, tables XVII and XVIII). The extent of this scatter of material is currently impossible to determine; and while an assemblage of this kind might have been produced by two or more brief episodes of occupation, there is possibility that the Grooved Ware may have been deposited as part of practices and gatherings linked to the construction of the henge earthwork. Radiocarbon dates on the charcoal produced ranges of 3050-2450 and 3340-2870 cal BC EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 17 Fig. 11 Identified traces of pre-monument activity and vegetational conditions at Avebury ( sources : Gray 1935; Smith 1965; Evans 1972; Evans et al. 1985; Allen & Davis 2009; this paper) (Pitts and Whittle 1992), consistent with a middle into the earlier part of the late Neolithic range indicated by the ceramics and lithics. Following the circuit of the bank round, Keiller’s limited cutting in the SW Sector behind stone 14 recovered 19 sherds of earlier Neolithic Plain Bowl from several vessels. To the immediate north-west of this in the area of levelled bank south of the west henge entrance came 27 sherds and several crumbs of Plain Bowl (Smith 1965, 224). The limited range of lithics from this area comprises a retouched blade, knife, chisel arrowhead, 87 flakes and two cores (Smith 1965, 224-6). Once again, there is no reason to think of this as other than settlement- 18 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE generated refuse. In this instance predominantly earlier Neolithic occupation might be linked to a phase of ard cultivation, exposed in a narrow trench through the bank at the Avebury School Site (Evans 1972, 273). Both at this location and under the bank in the NW Sector the molluscan sequence showed a subsequent reversion to lightly grazed impoverished grassland (Evans 1972, 274; Evans et al. 1985). A further low- density artefact scatter is recorded from the interior of the henge in the area of the Southern Inner Circle (Smith 1965, 226-8; records in Alexander Keiller Museum). 315 pieces of worked flint were recovered from a zone 90x50m in extent, principally around the Obelisk and stone settings i-xii. This could relate to depositional activity contemporary with the stone settings, but the presence of blades and narrow flakes, and a broken polished flint axe, along with 38 sherds of Peterborough Ware and Plain Bowl from the stone- holes (Smith 1965, 226) implies a 4th-millennium BC date range for this material; one far earlier than is normally countenanced for the megaliths. Indicative of the enactment of varied productive tasks at this location, a range of other retouched pieces are present among the flintwork, including scrapers and a knife, while the debitage includes several flakes from the thinning of axes or other large bifacial implements. Nowhere across the area later occupied by the henge is Neolithic activity particularly intense (at least activity of a kind that leaves a resilient material trace), yet it is both recurrently present and indicative of a place returned to. The picture is one of short-lived and probably small-scale settlement and visits during the early and middle Neolithic, some associated with cultivation. A demonstrably pre- henge late Neolithic presence is only firmly attested under the bank in the SE Sector, with evidence for the establishment by this date of lightly-grazed grassland to the west. At the medium scale, we can see more earlier (early-mid 4th millennium BC) activity in the zone to the west, and extending close to the Winterbourne within 300m of the henge (Evans et al. 1993, 151-3). By the later centuries of the 4th millennium BC focus has shifted to slightly higher ground to the east, including the area later covered by the SE Sector of the henge, Rough Leaze and to the south along the foot of Waden Hill. A picture of shifting residence areas and variation in intensity and scale of activity is hazily emerging, set within a context of increasing monumentalisation of this landscape. However, we believe that understanding of residence and landscape inhabitation can be much refined through further targeted fieldwork, more systematic application of radiocarbon dating to pit and buried soil material, the production of high- resolution environmental sequences, and a bolder approach to what is often characterised as difficult and ephemeral data. Acknowledgements The fieldwork team comprised the authors, undergraduate and postgraduate students from the University of Bristol, along with Mark Gillings (University of Leicester), Amanda Jones and local volunteers. The National Trust and their tenant farmer, Andy Orton, kindly granted permission for the fieldwork to take place. The site archive and artefactual material will be deposited in the Alexander Keiller Museum, Avebury. Jacqui Mulville (Cardiff University) kindly identified the cattle bone from Ell. The mollusc histogram was drawn by Robert Read and other figures by Anne Leaver. The molluscan analysis and reporting was funded by AEA: Allen Environmental Archaeology. References ALLEN, M. J. and DAVIS, B. 2009. A Middle Neolithic Pit in the West Kennet (stone) Avenue, Avebury, 2005. WANHM 102, 12-21 ALLEN, M. J. and GARDINER, J., 2009, ‘If you go down to the woods today; a re-evaluation of the chalkand postglacial woodland; implications for prehistoric communities, in M. J. Allen, N. Sharpies and T. P O’Connor (eds), Land and People; papers in memory of John G. Evans , 49-66. London: Prehistoric Society Research Paper 2 ALLEN, M. J. and SNASHALL, N. 2009. New features at Avebury from hand coring: geoarchaeology in action! PAST 63, 12-13 ANDERSON, R. 2005. An annotated list of the non-marine Mollusca of Britain and Ireland, Journal of Conchology 38(6), 607-37 CHAN, B. T-Y. 2010. Durrington Walls then and now: the description, interepretation and meaning of a monstrous Assemblage. Lithics 31, 44-54 CLEAL, R. M. J., 2001, ‘Neolithic and Early Bronze Age’, in A. Chadburn and M. Pomeroy-Kellinger (eds). Archaeological Research Agenda for the Avebury World Heritage Site , 63-7. Salisbury: English Heritage/ AAHRG CUNNINGTON, M. E. 1931. The ‘Sanctuary’ on Overton Hill, near Avebury. WANHM 45, 300-35 EAST OF AVEBURY: EXCAVATIONS AT ROUGH LEAZE 2007 19 EDMONDS, M., EVANS, C. and GIBSON, D. 1999. Assembly and collection - lithic complexes in the Cambridgeshire Fenlands. Proceedings of the Prehistoric Society 65, 47-82 EVANS, C., POLLARD, J. and KNIGHT, M. 1999. Life in woods: tree-throws, ‘settlement’ and forest cognition. Oxford Journal of Archaeology 18(3), 241-54 EVANS, J. G., 1972, Land Snails in Archaeology. London: Seminar Press EVANS J. G. 1984. Stonehenge - the environment in the Late Neolithic and Early Bronze Age and a Beaker burial. WANHM 78, 7-30 EVANS, J. G., LIMBREY, S., MATE, I. and MOUNT, R. 1993. An environmental history of the upper Kennet valley, Wiltshire, for the last 10,000 years. Proceedings of the Prehistoric Society 59, 139-95 EVANS, J. G., PITTS, M. W. and WILLIAMS, D. 1985. An excavation at Avebury, Wiltshire, 1982. Proceedings of the Prehistoric Society 51, 305-10 FOWLER, P, 2004, ‘East of Avebury: ancient fields in a local context’, In R. Cleal and J. Pollard (eds), Monuments and Material Culture. Papers in honour of an Avebury archaeologist: Isobel Smith, 130-8. East Knoyle: Hobnob Press GILLINGS, M., POLLARD, J., WHEATLEY, D. and PETERSON, R., 2008, Landscape of the Megaliths: excavation and fieldwork on the Avebury monuments, 1997-2003. Oxford: Oxbow Books GRAY,H. StG. 1935. The Avebury excavations, 1908-1922. Archaeologia 84, 99-162 GUNTER, J. and ROBERTS, V., 2007, Rough Leaze, Avebury, Wiltshire: geophysical survey report. Unpublished report, TALITS HARDING, J., 2000, ‘Ceremonial centres, cosmology and pilgrimage: the monument complex of Thornborough, North Yorkshire’, in A. Ritchie (ed.). Neolithic Orkney in its European Context, 31-46. Cambridge: McDonald Institute for Archaeological Research HOLGATE, R. 1987. Neolithic settlement patterns at Avebury, Wiltshire. Antiquity 62, 349-53 HOLGATE, R., 1988, Neolithic Settlement of the Thames Basin. Oxford: British Archaeological Reports, British Series 194 LAMDIN-WHYMARK, H. 2008. The Residue of Ritualised Action: Neolithic deposition practices in the Middle Thames Valley. Oxford: British Archaeological Reports, British Series 466 PARKER PEARSON, M., 2007, ‘The Stonehenge Riverside project: excavations at the east entrance of Durrington Walls’, In M. Larsson and M. Parker Pearson (eds). From Stonehenge to the Baltic: living with cultural diversity in the third millennium BC, 125-144. Oxford: British Archaeological Reports, International Series 1692 PARKER PEARSON, M., POLLARD, J., RICHARDS, C., THOMAS, J., TILLEY, C., WELHAM, K. and ALBARELLA, U. 2006. Materializing Stonehenge: the Stonehenge Riverside Project and new discoveries. Journal of Material Culture 11, 227-61 PARKER PEARSON, M. and RAMILISONINA. 1998. Stonehenge for the ancestors: the stones pass on the message. Antiquity 72, 308-26 PITTS, M. and WHITTLE, A. 1992. The development and date of Avebury. Proceedings of the Prehistoric Society 58, 203-12 POLLARD, J. 1992. The Sanctuary, Overton Hill, Wiltshire: a re-examination, Proceedings of the Prehistoric Society 58, 213-26 POLLARD, J., 2005, ‘Memory, monuments and middens in the Neolithic Landscape’, in G. Brown, D. Field and D. McOmish (eds). The Avebury Landscape: aspects of the field archaeology of the Marlborough Downs, 103-14. Oxford: Oxbow Books POLLARD, J. 2006. Solid nature and ephemeral architecture? Understanding the architecture of earlier Neolithic occupation in southern Britain from a dwelling perspective. Journal of Iberian Archaeology 8,41-51 POLLARD J., ALLEN, M., CLEAL, R., FRENCH, C., GARDINER, J., GILLINGS, M., McFADYEN, L. and SNASHALL, N. 2011. Between the Monuments: new fieldwork at Avebury. Past 68, 10-11 RICHARDS, J., 1990, The Stonehenge Environs Project. London: English Heritage SMITH, I. F., 1965, Windmill Hill and Avebury: Excavations by Alexander Keiller, 1925-1939. Oxford: Clarendon Press SNASHALL, N., 2002, The Idea of Residence in the Neolithic Cotswolds. Unpublished University of Sheffield PhD thesis THOMAS, K. D., 1985, ‘Land snail analysis in archaeology: theory and practice’, in N. R. J. Fieller, D. D. Gilbertson and N. G. A. Ralph (eds), Palaeoenvironmental Investigations: research design, methods and data analysis, 131-57. Oxford: British Archaeological Reports, International Series 258 WAINRIGHT, G. and LONGWORTH, I. H., 1971, Durrington Walls: excavations 1966-1968. London: Society of Antiquaries of London WHITTLE, A. 1990. A model for the Mesolithic-Neolithic transition in the upper Kennet valley, north Wiltshire, Proceedings of the Prehistoric Society 56, 101-10 WHITTLE, A. 1993. The Neolithic of the Avebury area: sequence, environment, settlement and monuments. Oxford Journal of Archaeology 12, 29-53 WHITTLE, A. 1994. Excavations at Millbarrow chambered tomb, Winterbourne Monkton, north Wiltshire. Wiltshire Archaeological & Natural History Magazine 87, 1-53 WHITTLE, A., 1997, Sacred Mound, Holy Rings. Silbury Hill and the West Kennet Palisade Enclosures: a Later Neolithic Complex in North Wiltshire. Oxford: Oxbow Books WHITTLE, A., DAVIES, J., DENNIS, I., FAIRBAIRN, A. and HAMILTON, M. 2000. Neolithic activity and occupation outside Windmill Hill causewayed enclosure, Wiltshire: survey and excavation 1992-93. 20 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE WANHM 93, 131-180 WHITTLE, A., HEALY, F. and BAYLISS, A. 2011. Gathering Time: dating the Early Neolithic enclosures of Southern Britain and Ireland. Oxford: Oxbow Books WHITTLE, A., POLLARD, J. and GRIGSON, C. 1999. Harmony of Symbols: the Windmill Hill Causewayed Enclosure, Wiltshire. Oxford: Oxbow Books WOODWARD, EJ. 1991. The South Dorset Ridgeway: survey and excavations 1977-1984. Dorchester: Dorset Natural History and Archaeological Society Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 21-42 Geophysical survey of the northeast and southeast quadrants of Avebury henge by Martin Papworth Geophysical survey work within the Avebury henge revealed details of buried stones in the northeast and southeast quadrants of the outer circle. The exceptionally dry conditions for the resistivity survey revealed additional stones not previously detected. By combining this survey with previous records, the archaeological resource of the circle is estimated. The resistivity survey also detected anomalies interpreted as the footings of pre- 18th century buildings and at least two phases of linear boundaries. Introduction This report is an analysis of a geophysical survey carried out by the National Trust in October and December 2003 (Figure 1). The work took place to provide additional data to enhance the research programme designed by Mike Pitts (2003), which included a proposal to erect a fallen megalith (Smith 1965, stone no. 78) within the SE quadrant of the outer circle at Avebury. Raising a stone would follow total excavation of the stone socket to obtain environmental and dating evidence for the outer stone circle to compare with evidence from the henge ditch and bank and other features within the enclosure. Publication of the surveys presented here has been delayed until now because of an (unfulfilled) intention to completely survey the northeast quadrant. In addition, the dryness of the | soil in October 2003 produced unusual results that | are not easily repeatable. The dating of the original erection of the outer stone circle within the henge is unclear and based on few reliable radiocarbon dates with a wide date range ( 2900-1700 BC). The evidence is summarised by Joshua Pollard and Andrew Reynolds (2002, 90- 91), while Pollard and Cleal (2004 126-7) argue that anomalous radiocarbon dates from Smith’s (1965) stone holes (44) and (8) (date range 770 BC-AD 150), previously dismissed as intrusive, may indicate re- erection of fallen stones and therefore the potential development and maintenance of the stone circle into the Iron Age and Roman periods. It was clear that further work was needed to improve the understanding of the development of the stone circle. In June 2003 discussions took place between National Trust and English Heritage staff concerning Mike Pitts’ review of the current knowledge of the stones of the outer circle. As a World Heritage Site and Scheduled Monument of international archaeological significance, it was considered particularly important to determine the relative significance of the proposed excavation site in comparison with the level of survival of other stones and stone sockets within the outer stone circle of Avebury henge. How would the total excavation of stone hole 78 diminish the remaining surviving stone hole stratigraphy, given that about 40% of this had been removed during the excavation and reconstruction work carried out by Alexander Keiller in the 1930s? The research design had considered all available National Trust South West, Eastleigh Court, Bishopstrow, Warminster, BA 12 9HW 22 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Reproduced from the Ordnance Survey mapping with the permission of the Controller of HMSO (c) Crown copyright. Unauthorised reproduction infringes Crown copyright and may lead to prosecution or civil proceedings. The National Trust AL 100018591 2003 Avebury, Wiltshire Geophysical Survey November 2003 Fig. 1 Extent of survey October 2003. Resistivity survey plot overlain on map of Avebury henge information on relative stratigraphic survival in the outer circle but the lack of fieldwork in the eastern quadrants meant that the survival of in situ prehistoric stone packing within stone holes in these areas could only be estimated. The 2003 geophysical survey provided much of the information required to understand the potential dateable resource of the outer circle stone settings. It has enabled most of the stone locations to be sorted into a hierarchy of potential for prehistoric stratigraphic survival. In this report the stone settings have been divided into five categories (A-E) based on the levels of survival estimated from Alexander Keiller’s 1937- 9 excavation records (Smith 1965) and more recently from the work carried out on the Beckhampton Avenue (Pollard and Reynolds 2002, 102; Gillings et al 2008, 252-364). Categories of stone settings A : Stones erected and still standing in their original stone hole, as far as can be determined without excavation. These sites contain the greatest archaeological potential with c. 80-100% of the original in situ stone hole filling. Burnt upright stones where a stone stump was left behind after the bulk of the stone was removed will have meant that most of the original stone packing remained in situ thus enabling a higher survival rate estimated as c. 40-80%. These sites have the greatest research potential after ‘A’ sites where they can be identified from the geophysical survey. These sites are categorised ‘AD’. B : Stones that have fallen. The fall of a stone will have disturbed the stone hole filling, but an estimated c. 30-50% of the archaeological potential of the filling is likely to have survived. C: Stones of either A or B categories that were deliberately buried. Excavated evidence has revealed that at least some of the stones were buried in the medieval period (Jope 1999, 67; Pollard and Reynolds 2002, 240) but a more detailed analysis of all the available evidence (Gillings et al 2008, 252-290) has demonstrated that this practice took place over a wide time GEOPHYSICAL SURVEY OF AVEBURY HENGE 23 span covering at least four centuries. The felling of a stone and excavation of its burial pit will leave an estimated c. 10-40% of in situ prehistoric stone hole filling. D: Stones of either A, B or C categories that were broken up, often by burning and splitting of the stone on the site of its original stone hole. Historical and archaeological evidence suggests that this destruction took place mainly in the 17th and 18th centuries (Ucko et al. 1991, 177). The felling, burning and breaking of a stone on site will have disturbed the stone hole to leave an estimated survival of c. 10-40% of in situ prehistoric stone hole filling. E: Stones of either A, B, C or D which had their stone socket fillings archaeologically excavated in the 1930s under the direction of Alexander Keiller. For category E c. 0-5% of in situ stone hole filling is likely to have survived (Smith 1965). Relative survival may depend on variables such as whether a burial pit was cut into the stone hole or excavated away from the site. Alternatively, a burial pit may have been re-excavated and then turned into a burning pit, the stone broken up and taken away. Various researchers have estimated the number and survival of stones within the outer circle at Avebury using the evidence available. These include Aubrey 1650s; Stukeley 1720s; Lukis and Smith, 1882; Smith, 1965, Ucko et al ., 1991, Bewley et al. 1996. However, the lack of geophysical survey of the eastern quadrants of the henge had been highlighted as a research requirement in the most recent reports (Ucko 1991 ; Bewley 1996) and within the Archaeological Research Agenda for the Avebury World Heritage Site (David 2001, 74). Therefore, with this in mind, Mike Pitts’ research proposals and the availability of the National Trust’s geophysical survey equipment, the opportunity was taken to carry out the 2003 geophysical survey. Previous geophysical survey by staff of the Ancient Monuments Laboratory (AML)1975-96 has revealed that within the henge resistivity produces good results, but magnetometry less so. The survey results of the north central and south central parts of the henge are published in Ucko et al (1991,212- 227) and the survey of the northwest quadrant is published in Bewley et al (1996, 640-646). The 2003 work took place following discussions with Andrew David of English Heritage who had directed the previous geophysical surveys within the henge and was best qualified to advise on previous results and techniques used. Method of geophysical survey Following the granting of a Section 42 licence by English Heritage, the period October to December 2003 was available to carry out the work. Resistivity and gradiometry took place in the northeast and southeast quadrants of the henge (from 27-31 October), with further survey in the northeast quadrant (3-5 December). The results from the two surveys were very different. In October, the ground was very dry following a late summer drought. On arrival at the site, a test area was surveyed to see whether it was worth using resistivity in such conditions. The test survey revealed the site of a buried stone, the contrast was such (70 ohms for the area around the stone and 500 ohms for the centre of the stone) that it became clear that the dry conditions were advantageous for detecting buried stones or walling. However, low resistance features did not show up well e.g. the ditches of field boundaries. Even higher resistance features such as banks of chalk rubble or brick lined drains were poorly defined on the October resistivity plots. By December there had been significant rain. As the ground was surveyed, the readings seemed to be revealing very little of the buried archaeology. The typical variation along a 20m traverse only fluctuated between 79 and 80 ohms. However, this apparent lack of detail in the waterlogged soils produced excellent definition of low resistance features when the data was downloaded on to the computer. In December, field boundary banks and ditches showed up very clearly in sharp contrast to the October results. The survey grid The base line for the survey was the field bank within the southeast quadrant that divides Stukeley’s (Figure 15) ‘Pasture VII’ and ‘Pasture VIII’ (NGR SU10341 69926 to NGR SU10379 69860). This was the approximate boundary of the south central survey of the henge (David in Ucko 1991, 261), which stopped at the north end of the boundary bank, but at the south end continued beyond the east edge of the bank by 20m. The grid was laid out at 20m intervals east of the bank and was continued across the road into the northeast quadrant. A total of 32 grids and part grids were surveyed in the southeast quadrant using resistivity and within the same grid, 19 grids were surveyed using a fluxgate gradiometer in the southwest 60% of the survey area. 24 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE In the northeast quadrant, 32 grids or part grids were surveyed using resistivity in October and a further 19 grids were added to the west side of this survey area in December. The December survey overlapped part of the north central survey (David in Ucko 1991, 223). Two October grids - R61 and R62 - were resurveyed in December as R72 and R73. The difference in results from the same area of ground surveyed five weeks apart highlighted the need to resurvey the whole area in wetter conditions. In the northeast quadrant, four grids of gradiometry were surveyed above the hollows for stone holes - grids G20-23 - in the positions of resistivity grids R43, R44, R52 and R53 to see whether burning pits showed up where resistivity indicated no buried stones. The gradiometer did not reveal anomalies where hollows were visible on the ground. Fluxgate gradiometer details A Geoscan FM36 fluxgate gradiometer was used. Readings were taken every 0.5m on 1.0m traverses beginning at the northwestern corner of each 20m grid walking along parallel traverses west to east. Therefore 800 readings were taken per 20m grid. Resistivity meter details A Geoscan RM15 resistivity meter was used. Readings were taken every 0.5m on 1.0m traverses beginning at the northwest corner of each 20m grid and walking zig-zag traverses first west to east then east to west and so on with 800 readings taken per 20m grid. The mobile probes were set 0.5m apart. Originally it had been planned to survey the area at 0.25m intervals on 0.25m traverses to try to increase detail of anomalies within the henge. However, the dry conditions and the amount of time available led to the decision to take readings at wider intervals, to cover the area rather than examine a small part of it in great detail. Results The data were processed using Geoplot 3.0 and each grid was examined individually and then combined. Using this analysis the drawn plots (Figures 4 and 7) were created. These results can be compared with the earthwork evidence (Figures 3 and 6) which was surveyed on the same grid and enhanced using the English Heritage survey (Bewley et al. 1996). In the following description and interpretation of □ 0 Fig. 2 Gradiometry anomalies SE quadrant the results, each anomaly is given a unique number and considered in relation to evidence depicted on historic maps and earlier surveys. For stone positions the numbering system of Smith (1965) is used and these stone numbers are enclosed in brackets e.g. (77). An attempt at correlating evidence for stones forming the outer circle is given in Table 1 below. Southeast quadrant (Figure 5 (except 1; 2 & 5 shown on Figure 2)) 1: Linear positive gradiometry anomaly aligned north-south and about 35m long. Its south end is on a line with 5 (see below) which extends from the east-west field boundary corner 4m to the west. The north end joins the southwest corner of building-like anomaly 6 (see below). It also runs parallel with anomaly 2 (see below) 12m to the east. 2: Linear positive gradiometry anomaly aligned north-south and about 35m long. Its south end is on a line with 5 (see below) which extends from east- west field boundary corner 16m westwards. The north end continues to the modern road edge between building-like anomalies 6 and 7. It also runs parallel with linear anomaly 1 12m GEOPHYSICAL SURVEY OF AVEBURY HENGE to the west. Anomalies 1 and 2 may be interpreted as land boundaries continuing a pattern east of that visible on Stukeley 1724 (Figure 15). They divide building plots and gardens fronting the road. One of them may be that shown on maps between 1845-1900. These anomalies are not visible on the resistivity plot. 3: A faint linear lower resistance anomaly visible as an earthwork ditch aligned north-south, measuring 65m long, 2-3m wide and 0.3m deep. 4: A faint higher resistance anomaly visible on the ground as an earthwork bank aligned north- south immediately west of ditch 3. It measures 65m long 3m wide and 0.3m high. Anomalies 3 and 4 are the field boundary between pastures VII and VIII on Stukeley’s plan of Avebury (Figure 15) where it is shown as having a line of trees along it. These boundaries within the henge are thought to date from the medieval period (Pollard and Reynolds 2002, 240). Although a clear earthwork, the bank and ditch are barely visible on the plot because of the dry October conditions. In comparison, the clear definition of a contemporary boundary 66- 68, surveyed in December, indicates the need to re-survey the SE quadrant when the soil has a much higher moisture level. 5: Traces of a linear east-west anomaly shown on the gradiometer plot. This alignment is a continuation of the hedge boundary east of the field corner that forms the northwest boundary of the survey. The line is confirmed by the southern extent of linear anomalies 1 and 2, which end on the alignment of 5. 6: A discrete rectangular block of high resistance readings. This coincides with an area of positive gradiometer readings (see Figure 2). This area of high resistivity is interpreted as a building and measures 10m east to west and is 6m wide. It lies immediately adjacent to the road edge on a terrace cut into the northern edge of the field. It shares this roadside terrace with 7 and 8 (see below) east of 6. 7 : A discrete rectangular block of high resistance readings, which coincides with an area of positive gradiometer readings (see Figure 2). The area of high resistivity is interpreted as a building and measures 12m east to west and is 5m wide. It lies 2m south of the road edge, lm west of 8 (see below) and 4m east of 6. It lies on a roadside terrace cut into the field edge. A large block of 25 sarsen stone is visible protruding from the grass on the site of this anomaly. 8: A discrete rectangular block of high resistance readings is interpreted as a building measuring 12m east to west and 5m wide. It lies 2m south of the road edge lm east of 7. It lies on a roadside terrace cut into a field edge. A large block of sarsen stone is visible protruding from the grass on the site of this anomaly. 9: A discrete rectangular block of high resistance readings is interpreted as a building aligned north-south measuring 12m by 6m. It lies 15m south of the road and immediately east of anomaly 21 which is interpreted as a north-south track. 10: A discrete rectangular block of higher resistance readings, although not as strong as anomalies 6- 8. It lies on the same roadside terrace as 6-8 and 1 1 (see below). This anomaly is interpreted as a building and lies 25m east of 8 and 25m west of 1 1 and measures 20m east to west by 5m wide. None of the buildings 6-10 are shown on Stukeley’s 1724 map (Figure 15) or on later maps of 1794 (WRO E/A 95), 1845 (WRO T/A Avebury) 1886, 1900 and 1924 (OS editions). They were abandoned before 1724 and are evidence that the roadside settlement pattern, that flanks the High Street east of St James Church, once continued along the north side of the SE quadrant as far as 1 1 (see below) sited on the eastern causeway of the henge. 1 1 : A rectilinear block of higher resistance readings immediately south of the road measuring 15m east to west by 5m wide. Stone footings of a building are visible here and a building is shown on maps between 1724 (Stukeley) until 1900 (OS 2nd edition). The site still lies within its own fenced and hedged enclosure. 12: A linear high resistance reading aligned east- west and 15m in length. Extending from the fenced boundary at the western edge of the survey as far as brick and concrete structure 13 (see below). Probably a service trench linked to the house on the western side of the survey area. 13: Brick and concrete structure 5m east to west by 2m wdie which shows as a high magnetic response on gradiometer plot (Figure 2). Probably the site of a reservoir or soakaway for the house to the west. 14: Linear higher resistance anomaly aligned east to west and traceable along the top of the terrace above building-like anomalies 6-8. A boundary line is shown in this position on the 1886-1924 26 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE OS editions. 15: Linear higher resistance anomaly aligned east to west and extending from the southeast corner of building-like anomaly 8 traceable for 10m: parallel and 5m north of 14. 16: Hook-shaped high resistance anomaly aligned north-south and traceable for 8m with a 3m long extension to the west at the south end. Possibly indicating the footings of a boundary wall or building. 17: Faint curvilinear higher resistance reading north of 18 (see below). Possibly a trace of an early boundary, but it does not align or seem to be associated with any other anomaly. 18: Curvilinear anomaly touching the north side of building-like anomaly 9, which branches from the northwest corner for 10m and from the northeast corner of 9 for 10m: perhaps boundary walls. The arms of this anomaly surround the semi-circle 10m in diameter of higher resistance 17. 19: A semi-circular arc of high resistance 2m wide and 20m in diameter south of building-like anomaly 10. 20: A spread of higher resistance readings on the south side of an east-west aligned hedge boundary. Possibly building rubble; some of this high resistance spread may be attributed to the roots of a large tree that grows from the bank. The earthwork enclosure containing an apparent building platform (Figure 3) is respected by 9 (Figure 4), which lies on its western edge but overlapped and ignored by 20 (Figure 5). 21 : A broad higher resistance linear anomaly, aligned northeast- southwest, 5m wide and traceable for 20m. Its course links it with a curvilinear low earthwork to the southwest (Figure 3) and takes it along the west side of building-like anomaly 9. Both anomaly and earthwork indicate the alignment of a track associated with 9 continuing southwest to the edge of the henge ditch. It does not quite align with a gap in the boundary bank 4 where Stukeley shows a gate on his perspective of 1724 (Figure 15). Anomaly 2 1 also forms a boundary between a speckled area of high and low resistance readings and 97 (see below), a distinct area of consistent low resistance readings, 22: An apparent pattern of scattered high resistance readings forming a corridor c. 8m wide and aligned northwest-southeast measuring about 70m long from the west field boundary at its northwest end to buried stones 29-31 at its southeast end. Beside the field boundary at the northwest end an area of higher resistance contains some larger spots of higher resistance 95 and 96 (see below). A water trough and fragments of modern rubble and rubbish lay against the hedge and some of the high resistance may GEOPHYSICAL SURVEY OF AVEBURY HENGE 27 Fig. 5 Southeast quadrant resistivity plot with numbered anomalies be due to modern dumping and disturbance. However, of potential significance are spots of higher resistance, which can be linked to form a circle c. 30m in diameter 96 (see below). 23: A faint linear anomaly aligned northwest- southeast traceable for 20m, possibly a boundary that would show up more clearly in wetter conditions. It ignores other alignments visible on the survey plot and lies within the lower resistance zone 97 (see below). 24: A curvilinear higher resistance anomaly aligned northeast-southwest. This follows the southeast perimeter of the survey area on the lip of the henge ditch. It is a low hedge bank visible as an earthwork 2m-3m wide and 0.3m high. There was a hedge above the bank in Stukeley’s 1724 survey (Figure 15) but there are now only occasional trees and bushes along its length. Similar to 98 (see below) in the northeast quadrant, although 24 does not have a visible ditch. 25: A semi-circular block of higher resistance, measuring 20m northeast-southwest by 5m wide. It lies immediately east and between fallen stones 38 and 39 and follows the edge of the henge ditch. 28 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Stone positions A number is given for each stone position. Numbers in rounded brackets (89) equate to numbers allocated by Isobel Smith (1965) following Alexander Keiller’s excavations in 1937-1939. Numerals and letters in square brackets [3] are those allocated by Rev A. C. Smith following probing and excavations by W. C. Lukis and Smith in 1881-2 (see Table 1). Very few, if any stone holes can be identified from the geophysical survey. The dry conditions enabled the stones to show up clearly but stone holes do not, probably because they were low resistance features difficult to detect without more moisture in the soil. 26 (90): There is an area of higher resistance around this point. It lies 5m east of the survey baseline and boundary bank. A few very high resistance readings are scattered east of this point. The AML survey (David in Uckol991, 219) records an area of higher resistance readings here and attributes the site to (89)(see below). An additional stone position can be found in Table 1 based on Smith’s (1885. 141) records for the southeast quadrant and thus numbering has been altered in this report. Readings for this area indicate that the stone has been broken up and removed. Stukeley records a fallen stone near here and it is shown on his 1724 map east of the boundary bank between pastures VII and VIII. There is an 88ft gap between Smith’s (1885) buried stones [3] and [4] either side of the bank. His first buried stone east of the bank is [4] which he measured as lying 53ft (16.1m) east of the bank. His distance between stones is on average 34-36ft (10.5-1 1.0m). The interpretation of anomaly 26 is a stone broken up and taken away between 1724 and 1882. 27 (89) [4]: A very high resistance anomaly interpreted as a buried stone measuring 3m east- west by 2m. Smith (1885, 141) found a large stone here by probing 16.1m east of the field bank. 28 (88) [5]: A very high resistance anomaly interpreted as a buried stone measuring 4m north-south by 2m wide. Smith (1885, 141) found a large stone here by probing 12.2m northeast of [4]. 29 (87) [6]: A very high resistance anomaly interpreted as a buried stone measuring 4m northwest- southeast by 1.5m wide. Smith (1885, 141) found a large stone here by probing 10.1m northeast of [5]. 30 (86): A very high resistance anomaly interpreted as a buried stone measuring 4m north-south by 2m wide. Smith failed to locate this stone being 5m east of [6] and 6m south-west of [7]. It falls outside the regular 10m-12m spacing of stones of the outer circle. As buried stones have been shifted away from their stone holes in many instances, the original regular spacing of stones in the southeast quadrant is still a possibility. 31 (85) [7]: A very high resistance anomaly interpreted as a buried stone measuring 2.5m east-west by 2m wide. Smith (1885, 141) found a large stone here by probing 11.6m northeast of [6]. 32 (84) [8]: A very high resistance anomaly interpreted as a buried stone measuring 3.5m east-west by 2.5m wide. Smith (1885, 141) found a large stone here by probing 11m northeast of [7]. 33 (83) [9]: A very high resistance anomaly interpreted as a buried stone measuring 3m east-west by 2.5m wide. Smith (1885, 141) found a large stone here by probing 10.1m northeast of [8]. 34 (82): A very high resistance anomaly interpreted as a buried stone measuring 3m north-south by 1.5m. Smith did not find this stone, which is 12m northeast of 33 and 8m southwest of 35. 35 (81): A very high resistance anomaly interpreted as a buried stone measuring 3m east-west by 2m wide. Smith did not find this stone, which is 10m northeast of 34 and 11m southwest of 36. 36 (80) [10]: A very high resistance anomaly interpreted as a buried stone measuring 3m north-south by 1.5m wide. Smith (1885, 141) found a large stone here by probing 35m northeast of [9]. 37 (79) [11]: A very high resistance anomaly interpreted as a buried stone measuring 4m north-south by 3.5m wide. Smith (1885, 141) found a large stone here by probing 10.7m northeast of [10] and 9.1m southwest of [IV]. 38 (78) [IV]: A fallen stone, 4m long and 2m wide. This is the stone that was to be re-erected as part of Mike Pitts’ research proposals. Standing in 1724, it fell later in the 18th century. 39 (77) [V]: A fallen stone, mostly hidden beneath the turf, its southeast corner is currently exposed. The resistivity survey indicates that it measures 3m northwest to southeast by 2.5m wide. 40 (76): Predicted position of a stone 10-1 1m northeast of 39. There is the edge of a higher GEOPHYSICAL SURVEY OF AVEBURY HENGE Fig. 6 Northeast quadrant earthworks and parch marks with resistance anomaly here, but a hedge bank crosses the north side of it and the fence line and thick vegetation prevented access to the bank top. This hedge bank existed in 1724. 41 : William Stukeley’s predicted position of a stone 8m- 10m northeast of 40 and just west of site 1 1, the building shown on maps between 1724-1900. Sarsen footings of a building are visible amongst the vegetation. Perhaps the stone was broken up when the building was constructed but equally it may never have existed. The distance between fallen stones 39 and 44 indicates that there is insufficient space for four intervening stones, thus one of the sites 40-43 may not have been that of a stone. 42 (75) and 43 (74) are predicted stone positions. Two fallen stones are shown on the 1724 map (Figure 15) flanking the road to Marlborough. The visibility of these stones in the road verges in 1724 is questioned by Ucko (1991, 208-9), particularly as they are not recorded by Aubrey in 1663. NE Quadrant (Figures 8 and 9) 44 (73) [VI]: A Fallen stone, called the Pond Stone 29 grids by Smith (1885, 141) presumably because of the pond that existed within the henge ditch a few metres to the northeast and which is shown on the 1886-1924 OS editions; the earthworks of the pond are still clear. 45: Predicted position of a stone setting. A faint higher resistance anomaly is visible at this location measuring 1.5m by 1.5m. It lies 6m north of 44 and 10m south of 46. 46 (72) [12]: A very high resistance anomaly interpreted as a buried stone measuring 4m east to west by 2m. Smith (1885, 141) found this stone by probing 15.5m north of [VI]. 47 (71): A high resistance anomaly interpreted as a buried stone measuring 1.5m by 1.5m. Smith did not locate this stone presumably because of its relatively small size. This is possibly a stump with the rest of the stone having been taken away: probably an AD category site. 48 (70) [13]: A very high resistance anomaly interpreted as a buried stone measuring 2.5m east-west by 2m. Smith (1885, 141) found this stone by probing 23.8m north of [12]. 49 (69) [14]: A very high resistance anomaly interpreted as a buried stone measuring 3m 30 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 8 Northeast quadrant resistivity survey (east area) numbered anomalies October 2003 north-south by 2m. Smith (1885, 141) found this stone by probing 9.8m north of [13]. 50 (68) [D]: A standing stone shown on the plans of Avebury drawn by John Aubrey and William Stukeley. 51: A higher resistance reading interpreted as a buried stone measuring 1.5m by 1.5m. This site was not located by Smith and lies 6m north of 50. Both 49 and 51 seem too close to standing stone 50 to reflect the regular spacing of stones found elsewhere in the circle. 5 1 may be the stump of a stone left in an original socket. This is possibly an AD category site. 52 (67): A stone position coincident with a spread of higher resistance with some very high resistance spikes indicative of a scatter of stone fragments (as with 26) indicating debris from breaking up Fig. 9 Northeast quadrant anomalies (west area) December 2003 of a sarsen. 53 (66): A stone position coincident with a spread of higher resistance with some very high resistance spikes indicative of a scatter of stone fragments. 54 (65): A stone position. Some local high resistance spikes but no clear evidence of an anomaly in the vicinity. 55 (64) [1]: A large earthwork depression coincident with a trace of higher resistance possibly the remains of a stone setting measuring 1.5m north- south by lm. This is probably Smith’s (1885, 141) [1] which lay 42.7m from standing stone 50 [D]. This distance allows for three stone positions not taking into account the close spacing of GEOPHYSICAL SURVEY OF AVEBURY HENGE 31 anomaly 51 to 50. Therefore four stones have been assumed between 50 and 55. The five earthwork hollows recorded by RCHM 1990 are matched with Smith [l]-[p]. Stukeley’s stone positions drawn on his 1724 map are difficult to match exactly with Smith’s (1965) stone positions (68)-(58) suggested here. However, he recorded three surviving fallen stones in 1724 that existed somewhere between (66)-(60). They had been removed by the time of Smith’s survey in 1882. 56 (63) [m]: An earthwork hollow close to 55 with evidence of an anomaly in the vicinity. This is probably Smith’s (1885, 141) [m], which lay 5.8m from [1]. 57 (62) [n] : An earthwork hollow coincident with an area of higher resistance 2.5m by 2.5m with a spike of high resistance 0.5m wide at its centre suggesting robbed megalith debris or packing for a stone hole. This is probably Smith’s (1885, 141) [n], which lay 11m from [m]. 58 (61) [o]: An earthwork hollow with no evidence of an anomaly in the vicinity. This is probably Smith’s (1885, 141) [o], which lay 12.5m from [n] . 59 (60) [p]: An earthwork hollow with no evidence of an anomaly in the vicinity. This is probably Smith’s (1885, 141) [p], which lay 11m from [o] . 60 (59) [15]: A very high resistance anomaly interpreted as a buried stone measuring 3m northeast-southwest by 2m. This is Smith’s (1885,141) [15], recorded as a large buried stone found by probing 10.4m from [p]. 61 (58) [16]: A very high resistance anomaly interpreted as a buried stone measuring 3m north-south by 2m. This is Smith’s (1885,141) [16], recorded as a large buried stone found by probing 11.6m from [15]. Stukeley records two fallen stones close to the kink in the field bank, which can be attributed to 60 and 61. This may be a rare example of an 18th century fallen stone becoming buried. Another would be (77) which now has only a corner visible above the grass. 62 (57) [q]: A stone position that cannot be linked with a geophysical anomaly. This is probably Smith’s (1885, 141) [q], recorded as a pit 12.2m from [16]. 63 (56) [r]: A stone position that cannot be linked with a geophysical anomaly. This is probably Smith’s (1885, 141) [r], recorded as a pit 10.1m from [q] (Figure 8). 64 [s]: A possible stone position that cannot be linked with a geophysical anomaly. Smith (1885, 141) recorded a pit 5.2m from [r] (Figure 8). 65 (55): A stone position that might be linked to an area of high resistance measuring 4m northwest- southeast by 3m immediately south of the lip of the henge ditch (Figure 8). Anomalies within the outer circle Avebury northeast quadrant (Figure 9, 66-82, except 71) 66: Linear higher resistance anomaly 135m long aligned northeast-southwest. A bank associated with 67 and 68. Together these features form a boundary which was abandoned by 1724 (Figure 15) but which formed part of a pattern of parallel boundaries, probably planned land allotment of the medieval period. 67 : A linear low resistance anomaly visible as a ditch on the ground 135m long and 2m wide and part of the same complex as 66 and 68. 68: A linear high resistance anomaly visible as a bank on the ground 135m long and 2m wide and part of the same complex as 66 and 68. 69: A circular low resistance area measuring 12m in diameter, apparently cut by 70 on the southwest side, defined by higher resistance areas to the northeast and northwest. A possible circular feature was identified during the AML geophysical survey of this area (David in Ucko 1991) but this lay to the southeast, crossed by the line of 66-68. This anomaly did not show on either the October or December 2003 surveys. 70: A linear high resistance anomaly aligned northwest-southeast and possibly associated with 69. It is traceable from the western edge of the survey area for 15m. It follows a similar alignment to 71, 72, 74, 75 and 76 (Figure 8). 71: A linear low resistance anomaly aligned northwest- southeast and traceable for about 70m. Probably a ditch following a similar alignment to 70, 72, 74, 75 and 76. 72: A linear low resistance anomaly aligned northwest- southeast and traceable for about 50m. Probably a ditch following a similar alignment to 70,71,74, 75 and 76. 73: A linear low resistance anomaly aligned east- west and traceable for about 30m. Probably a ditch, it 32 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE crosses the line of 66-68. 74: A linear low resistance anomaly aligned northwest-southeast and traceable for about 30m. Probably a ditch, it forms a junction with 73 near the western boundary of the survey area and crosses the line of 66-68 following a similar alignment to 70, 71, 72, 75 and 76. 75: A linear low resistance anomaly aligned northwest-southeast and traceable for about 20m. Probably a ditch, it crosses the line of 66- 68, following a similar alignment to 70, 71, 72, 74 and 76. 76: A linear higher resistance anomaly aligned northwest- southeast and traceable for about 10m in the southwest corner of the northeast quadrant of the survey area. Possibly a ditch similar to 70, 71,72, 74 and 75. 77: A linear high resistance anomaly aligned east-west and traceable for about 10m in the southwest corner of the survey area. Possibly a wall footing. 78: A rectilinear enclosure with the southwest side formed by 75. The other sides are faint higher resistance anomalies and measure about 14m long. Possibly cut on the south side by the quarry ditch for the prominent rectangular mound 89. 79: A circular low resistance anomaly about 3m in diameter. Traces of a brick and concrete structure are visible here in the grass and it marks the position of a modern drain/soakaway for the neighbouring cottage. 80: An irregular higher resistance anomaly about 20m long and 8m wide. 81 : A distinctive linear high resistance anomaly 1.5m wide aligned northwest-southeast measuring about 6m long and forming a junction with 82 at its southeast end. Possibly wall footings or a stone-lined drain. 82: A distinctive ‘L’-shaped high resistance anomaly aligned north-south for about 20m. At its northern end it forms a junction with 81, turning east-west at a right angle and continuing for about 8m. It measures 1.5m wide and becomes fainter at the eastern edge of the December 2003 resistivity survey area. The differing soil moisture conditions of October 2003 may explain why this anomaly was not detected further east. The anomaly follows the upper slope of the west and north sides of a prominent rectangular mound. Anomaly 82 may be a wall footing indicating that the mound is a building platform with its gable end-on to the road. The south end of the mound has been truncated by the garden on the east side of an adjoining National Trust holiday cottage. Low resistance anomaly 89 marks the position of the earthwork quarry ditch for the mound (Figure 8). 83: A disjointed linear high resistance anomaly aligned east-west and measuring about 15m in length with a 5m long north-south aligned return at its east end. This anomaly is tentatively interpreted as a building footings aligned with the proposed trackway 84. 84: A linear high resistance band of readings over 60m long and aligned east-west. At its east end the anomaly diverges from the road on the south side of the Pond Stone 44 where it is 2m wide. It then widens to 5m on the west side of the stone and runs along the bottom of an earthwork terrace as far as the lawn of a cottage at the western edge of the survey area. The northern edge of the anomaly aligns with a fence shown on the 1886 OS edition. While the survey was in progress it was assumed that the high resistance readings were associated with a line of demolished roadside buildings similar to those found on the south side of the road. However, a grid-by-grid analysis of 84 (R42; R50; R59) showed no gaps in the high resistance until the west side of R42. This anomaly is interpreted as a road or trackway predating the construction of the garden terrace. Two buildings are shown on this alignment on the 1886-1924 editions of the OS map. One building, probably a barn aligned east-west, lay immediately west of the Pond Stone 44. Another, probably a cottage, also aligned east-west, lay against the east side of the existing cottage garden. Remains of these buildings cannot be identified within the band of high resistance readings. They were presumably demolished as part of the scheme to remove village buildings from the henge begun by Alexander Keiller and continued by the National Trust in the early years of its acquisition of the property. The width and high resistance of the track- like anomaly indicates substantial metalling and the road may have been associated with the rectangular earthwork mound bordered by ditch 89 and wall-like anomaly 82. 85: A linear higher resistance anomaly aligned north-south and traceable for about 25m varies between lm-2m wide and joins the north side of 84. Linear 86 (see below) forms a junction with GEOPHYSICAL SURVEY OF AVEBURY HENGE Fig. 10 High contrast resistivity plot southeast quadrant 85 on its east side. This feature is interpreted as a boundary bank or ditch possibly associated with building like anomalies 87 and 88. 86: A lm wide linear higher resistance anomaly aligned east-west and traceable for about 8m from the northwest corner of building-like anomaly 87 continuing as far as its junction with 85: it appears to form a junction with 90. 87: A rectilinear high resistance linear anomaly aligned north-south measuring 6m long and 4m wide, which corresponds with an earthwork hollow, The width of the outline of the anomaly measured between 0.5m-2m wide. The footings of a mortared stone wall can be seen eroding from a visitor footpath that crosses this feature. Probably a building not shown on any known maps. 88: A hook-shaped high resistance anomaly aligned north-south at the south end of 87 and perhaps a 0.5m- 1.0m wide wall footing associated with 87. It measures about 4m long with a 2m westward extension at its northern end. 89: A 3m-5m wide band of low resistance readings running for 12m along the northern side and 10m along the western side of the rectangular mound associated with 82. The low resistance follows an earthwork ditch. There is no visible quarry ditch on the northern side of the mound which, on the east and west sides is cut into by terracing for the cottage garden on its south side. 90-93: Traces of roughly parallel alignments of higher resistance features aligned northeast- 33 southwest extending from the henge ditch southwest between 20m-30m. Definition might be enhanced if surveyed under wetter ground conditions. 90: A higher resistance anomaly measuring lm-2m wide and about 15m long. From the henge ditch it passes between buried stones 46 and 47 and continues to the junction of linear feature 86 and building-like anomaly 87. 91: Faint traces of low resistance about lm wide. From the henge ditch it passes to the northwest of buried stone 51 and is traceable for about 20m. 92: A 3m-4m wide band of higher resistance crossing linear 71 and traceable for about 25m. Lies to the northwest of stone position 57. 93: A faint higher resistance anomaly extending southwest from buried stone 60 for about 20m. Anomalies 90-93, with 70-72 and 74-76, may be field boundaries. 94: Linear higher resistance anomaly aligned northwest-southeast and extending to the edge of the survey area from the east side of building- like anomaly 87. About 10m long and 0.5m wide, this feature is interpreted as a drain. SE Quadrant (Figures 5 and 10) 95: A 3m long and 2m wide area of high resistance readings aligned east-west against the western edge of the survey area. Interpreted as a block of buried masonry, possibly a buried megalith. 96: A scatter of six high resistance spikes at the western edge of the survey area south of 95. This anomaly might be a scatter of buried sarsen fragments. A ring of high resistance features 30m in diameter can be seen in an area where modern debris is eroding from the turf. However, the geophysics suggest archaeological potential, possibly a ring of stone settings. 97: An area of low resistance with few spikes of higher resistance measuring about 50m northeast-southwest by 30m. Bounded to the west by trackway 21 and building-like anomaly 9, and to the east by the henge ditch. To the north, 97 is bounded by a hedge and higher resistance area 20. Overall, area of 97 has a distinctive geophysical signature in comparison with its surroundings indicating a different land use. NE Quadrant (Figures 8 and 9) 98: A linear high resistance anomaly running along the east and north edges of the survey area. 34 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Visible on the ground as a bank 2m-3m wide and up to 0.6m high, this feature is a continuation of 24 in the southeast quadrant. Running along the edge of the henge ditch from the road north as far as stone position 58, it turns to pass buried stones 60 and 61 and continues westwards 10m- 12m south of the lip of the henge ditch. Ditch 99 and bank 100 are part of this boundary, which is shown as a hedgeline on the 1724 map (Figure 15). 99: A linear low resistance anomaly traceable from stone position 58 where it diverges from the henge ditch edge on the north side of 98. Measuring 3m wide and 0.4m deep, this feature is the quarry ditch for 98. 100: A linear high resistance anomaly traceable from stone position 58, where it diverges from the henge ditch edge on the north side of 99. It represents a counterscarp bank not visible on the ground. Conclusions The geophysical survey has increased our knowledge of the outer circle of stones at Avebury. The especially dry conditions revealed 18 buried stones as distinctive high resistance readings. With the exception of the gap caused by robbed stone 26, an arc of 15 buried or fallen stones is now known to exist in the southeast quadrant. Seven buried stones were detected in the northeast quadrant, but it is clear from the survey that most of these stones have been broken up and taken away. In many cases very little has been left behind, but occasional traces of higher resistance indicate stone fragments and stone-hole packing. The potential for dating a stone hole, without moving a buried stone (C) or fallen stone (B) to reach the stone pit, lies in the surviving silts of the burning pit/removed stone (D) sites. The present survey has identified many such sites e.g. 52-59 in the northeast quadrant, but they are unlikely to have more than 40% of in situ material within them. The highest potential sites for dating - containing the most in situ material and without affecting one of the four (A) standing stones in the outer circle - are those where a stone was burnt while upright. In such a situation, the stone might be broken off at ground level. If the upper stone was taken away, the lower stump might be left in the ground with its intact original packing material. These are here Fig. 11 Selected stone sites with numbers e.g. Smith 1965, (78); Smith 1885, [IV]; Geophysics 2003, 38 termed (AD) sites, potentially retaining 80%-100% of their stone-hole filling. Potential AD sites have not been reported in historical records nor found during excavation, but the geophysical survey suggests that they may exist. Resistivity results from the northeast quadrant confirm that this area is much more disturbed by stone breaking than the southeast quadrant and with the surviving buried stone anomalies tending to be more fragmented. The best candidates for AD category sites are found in the northeast quadrant. They are identified by shape and intensity of the resistance in comparison with the buried stone anomalies of the southeast quadrant. They are: 45 a faintly traceable higher resistance site indicating some stone packing in situ ; 47 a small square buried stone anomaly, perhaps a vertical stump rather than a buried stone and a good AD candidate; anomaly 51 is similar to 47, a strong AD candidate; while 57 is a faintly traceable higher resistance site that may represent stone packing and the bottom of a stone in situ. Table 1 is a correlation of information from various sources that describe the presence and condition of the megaliths of the outer stone circle of Avebury henge. Column 1 matches the A-E categories used in this report following Isobel Smith’s (1965) numbering of the stones and noting the condition of stone positions as found in the excavations by Alexander Keiller 1937-9. A GEOPHYSICAL SURVEY OF AVEBURY HENGE 35 Table 1 : Correlation of records for stones of the outer Avebury stone circle. Smith I. 1965 Keiller A. 1930s Numbering 1-98 A-E 12003) Smith A.C. 1885 Stukeley W. 1724 plan E Aubrey 1663 plan A Geophysics Reports & Parch Marks 1990-2003 A1 standing B standing Standing Stone E2 Position E3 burning pit Depression Stone E4 buried between E5 buried 3 & 6 E6 buried Depression E7 fallen III fallen Fallen 8 stones E8 standing A standing Standing between E9 buried Position 6 & 23 E10 buried Position Ell burning pit f) pit found nothing Fallen E12 fallen II fallen Fallen E13 burning pit e) pit sarsen fragments Fallen E14 fallen I fallen Standing E15 burning pit Depression E16 buried Position E17 burning pit Depression E18 burning pit Depression E19 buried & burnt d) pit nothing Depression E20 buried & burnt 1) large buried stone Depression E21 buried & burnt c) pit found nothing Fallen E22 buried & burnt b) pit found nothing Position E23 buried & burnt a) pit sarsen fragments Position E24 buried Fallen ?D25 Position ?D26 Fallen ?D27 Fallen ?D28 Position ?D29 Standing E30 buried Depression Area from stones E31 buried Depression 30-45 surveyed in 1996 but E32 standing I standing Standing These stone positions had E33 standing H standing Standing Been excavated and stones E34 stump X fallen Fallen Re-erected or marked with E35 stump IX fallen Fallen Concrete posts in the 1930s E36 stump in f.b. s) Removed between 1819 & 1857 Position Therefore this resistivity E37 burning pit Standing 3 stones Not significant for this E38 burning pit Standing between Table E39 burning pit Standing 40 & 29 E40 stump VIII fallen Fallen E41 stump VII fallen Fallen E42 stump in f.b. v) Removed between 1819 & 1857 Fallen E43 burning pit Standing Stone E44 standing G standing Standing Stone E45 burning pit Standing Stone E46 standing F standing Standing Stone D47 Fallen D48 e)Removed between 1819 & 1857 Fallen Stone PC95) D49 a)36ft west of E man v sarsen frags. Fallen A50 standing E standing Standing Stone D51 v) pit 29ft from u 47ft from E Depression D52 u) pit 73ft from t Depression D53 D54 t) pit 60ft from s Depression P(95) Low res. Area (90) D55 High res. Area (03) 65 D s) pit, not clear 17ft from r Depression 64 36 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 1: continued Smith 1. 1965 Keiller A. 1930s Numbering 1-98 A-E 12003) Smith A.C. 1885 Stukeley W. 1724 plan E Aubrey 1663 plan A Geophysics Reports & Parch Marks 1990-2003 D56 r) pit, not clear 33ft from q 2 stones 63 D57 q) pit 40ft from 16 Depression Between 62 C58 16) large stone buried 38ft from 15 Fallen 56 & 67 P(95) Buried stone (03) 61 C59 15) large stone buried 34ft from p Fallen P(95) Buried stone (03) 60 D60 p) pit 36ft from o Depression 59 D61 o) pit 41ft from n Position P(95) 58 D62 n) pit 36ft from m Fallen Trace (03) 57 D63 m) pit 19ft from 1 Fallen 56 D64 1) pit 140ft from D Fallen Trace (03) 55 D65 Depression High res. Area (03) 54 D66 Depression P (95) High res.Area (03) 53 D67 2 Positions 52 C Position P(95) Buried stone (03) 51 A68 standing D standing Standing Stone 50 C69 1 14) large stone buried 32ft from 13 I Position | |P(95); Buried stone (03) 49 | C70 13) large stone buried 78ft from 12 Position P(95); Buried stone (03) 48 C71 Position Buried stone (03) 47 C72 12) large stone buried 53ft from VI Position P(95); Buried stone (03) 46 D Trace (03) 45 B73 fallen VI Pond Stone fallen Fallen Stone 44 ?D74 Fallen? Road Road unsurveyed (03) 43 ?D75 Fallen? Road Road unsurveyed (03) 42 ?D Position Unlikely Building footings (03) 41 ?D76 Position Hedgeline unsurveyed (03) 40 B77 fallen V fallen Fallen Stone Fallen stone (03) 39 B78 fallen IV fallen Standing Stone 38 C79 11) large stone buried 35ft from 10 Position Buried stone (03) 37 C80 10) large stone buried 1 15ft from 9 Buried stone (03) 36 C81 P(95); Buried stone (03) 35 C82 Position P(95); Buried stone (03) 34 C83 9) large stone buried 33ft from 8 Position P(95); Buried stone (03) 33 C84 8) large stone buried 36ft from 7 Position Buried stone (03) 32 C85 7) large stone buried 38ft from 6 Position P(90); Buried stone (03) 31 C86 Position Buried stone (03) 30 C87 6) large stone buried 33ft from 5 Position P(90) Buried stone (03) 29 C88 5) large stone buried 40ft from 4 Position P(90) Hi res(90); Buried Stone (03) 28 C89 4) large stone buried 53ft from bank Position P(95) Hi res(90); Buried Stone (03) 27 D90 Fallen High res (90); Frags (03) 26 C91 3) large stone buried 25ft from 2 Depression High resistance (90) C92 2) buried 9ft large stone 1 18ft from k Position High resistance (90) D Smith’s distance between k and 2 Fallen D93 Suggests 2 stone positions Position Plan shows 4 Stones be- tween 92-97 D94 k) pit found nothing 17ft from I Position D95 i) pit found nothing 21ft from h Depression P (95) D96 h) pit sarsen fragments 30ft from g Position D97 g) pit many sarsen chips 36ft from C Standing Stone P (95) High res (90) A98 standing C standing Standing Stone GEOPHYSICAL SURVEY OF AVEBURY HENGE questionable allocation of A-E where the records of stone condition are unclear is preceded by ? e.g. stone positions (25)-(29). Column 2 matches Smith’s (1885) letters and numbers, allocated after probing and excavating stones in 1882. A strong correlation can be observed with the 2003 geophysical survey owing to Smith’s careful measurements between each of the sites that he identified (Figure 16). Column 3 is a best fit of stone positions recorded by William Stukeley in his plan E dated 1724 (Figure 15). He produced a number of plans (see Ucko 1991 for discussion) and the details of some contradict others. Occasionally the matching in Table 1 is questionable, particularly at a stone position where a fallen stone recorded by Stukeley is matched against a buried stone recorded by Smith. The general pattern is for 18th century fallen stones to be broken up and taken away rather than buried, although (58) and (59) are exceptions as they lie within a field bank. Column 4 is a best fit based on stone positions recorded by John Aubrey from a plane table survey he carried out in the mid 17th century (plan A: Figure 14). Ucko (1991) discusses other plans surveyed by Aubrey that show different arrangements of stones, but those in the northeast and southeast quadrants can be matched with surviving stones or those broken up and taken away. For example, Aubrey’s seven stones in the south central area can be closely matched following the 2003 survey that revealed (79)-(92) to be buried stones (with the exception of (90) which was broken up after 1724. Aubrey’s two stones in Stukeley’s pasture VII are the two surviving stones (77) and (78). The five in pastures VIII and IX lie west of buried stone (92). There are six vacant stone positions in these two fields. Aubrey does not show any of these stones. Four (93)-(96) had been broken up and removed by 1724, but (97) was still standing although it was demolished by the end of the 18th century. Aubrey’s two stones beside the road in pasture X are (98) and (1) and are still standing. Column 5 notes recent discoveries by aerial photography and geophysical survey. P (90) and P (95) are parch marks indicating buried stones recorded by RCHME in 1990 and 1995. High Res (91) refers to high resistivity anomalies over stone positions recorded by the AML survey (David in Ucko 1991). Buried stone (03) refers to the distinct high resistance readings recorded in 2003 followed by the survey number e.g. 27 for the anomaly used in this report. This comparison of different sources of evidence 37 has led to five stone positions to add to the 98 proposed by Isobel Smith (1965). Based on the combined evidence, there were originally at least 100 stones erected in the outer circle at Avebury. Most stones are spaced at lOm-llm intervals, but the geophysical survey has revealed two areas where stones occur 5m-6m apart: stones 29-31 (85)-(87) in the southeast quadrant and the buried stones either side of standing stone 50 (68) in the northeast quadrant. In the western part of the henge, Keiller did not excavate stone positions in gardens either side of the High Street (25) - (29), where Stukeley shows a standing stone and two fallen stones no longer visible. These stones have been classified as ?D because they have been assumed removed - future geophysical survey may show buried stones there. This analysis has probably produced one too many stone positions at the road crossing between the SE and NE quadrants. Another difficult area to match available evidence to stone positions is the arc of the outer circle between the surviving standing stones (50) and (68) as all but 3 of the stones have gone from this section. An additional stone may also have been introduced here. However, bearing in mind the probability that up to 3 of the stone positions overlap other stone positions and that the D category assigned to (25) - (29) is questionable the following totals have been calculated. Table 2: Pie chart showing proportions of categories A-E within outer circle □ A Standing ■ B Fallen □ C Buried □ D Broken up ■ E Ex cavated Stone Hole Filling estimated in situ Neolithic survival Category Quantity Standing (80-100%) A 4 Fallen (30-50%) B 3 Buried (10-40%) C 20 Broken up (10-40%) D 36 Excavated (0-5%) E 40 C sites possible Stumps (40-80%) AD 2 38 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Left: Fig. 12 Holworth, Chaldon Herring (RCHM 1970, 36) © crown copyright. Right: Fig. 13 Philipston ( Taylor 1994, 215) © crown copyright Prehistoric Enclosures? It was hoped that the survey of the henge interior would reveal details of prehistoric enclosures. Rings of post-holes of the type found during excavation and geophysical survey at Durrington Walls (Wainwright 1971; David and Payne 1997) were not identified at Avebury, although such might be found if the area was resurveyed in wetter conditions. The possible ring of anomalies tentatively identified in the north quadrant during the AML survey (David in Ucko 1991, 228) was not recognised when the area was resurveyed in dry October or wet December 2003. A circular low resistance anomaly, about 12m in diameter, was identified near the north edge of the henge 69. Faint rectilinear anomalies that ignore other alignments (e.g. 78) may indicate the position of prehistoric features. The band of high resistance spikes 22 in the southeast quadrant and the c. 30m diameter arc of higher resistance anomalies 96 may be significant, but the level of definition is insufficient for a clear interpretation. Close to 96, anomaly 95 is of particular interest with high resistance readings over an area that matches that of buried megaliths. However, it lies 95m west of the nearest megalith in the outer circle and, if 95 represents a buried megalith, it must be part of an inner feature of the henge that may continue to the west. The regularity of this feature in plan when compared with the plots of the buried megalith anomalies may mean that it is modern. Early Field System? A group of low resistance anomalies in the northeast quadrant may be a field system as indicated by their common orientation. Aligned northwest-southeast they ignore the north-south alignments of medieval banks and ditches. For example, two of these linear features (70 and 75) cross the ditches and banks 66-68. They are not visible on the ground and are presumably ditches levelled by the medieval ploughing which may have ceased within the henge in the later medieval period (Pollard and Reynolds 2002, 246). Trace anomalies aligned northeast- southwest 90-93 suggest that the northeast part of the henge was once divided into a rectilinear field system, possibly prehistoric in origin. Medieval Boundaries The interior of the henge is divided up into a series GEOPHYSICAL SURVEY OF AVEBURY HENGE 39 Fig. 14 John Aubrey 1663 survey Plan A ( Bodleian Library MS. Top. Gen. c. 24) © Bodleian Library of north-south aligned plots, many of which are shown as hedges in 1724 (Figure 15), e.g. 3-4. Others are visible as banks and ditches 66-68, although they had fallen out of use by the 18th century. The gradiometer survey revealed two further north- south divisions 1 and 2 continuing from the road to the east-west hedgeline 5 in the southeast quadrant. In the northeast quadrant (Figure 6) the RCHME earthwork and parchmark survey (Bewley et al. 1996, 642) recorded further north-south divisions not visible on the plot, probably because of the dry conditions during the October resistivity survey. These boundaries are interpreted as planned land allotment. The roadside enclosures were probably tofts for houses and gardens and the longer strips behind were crofts for arable and pasture. Despite alteration of property boundaries alongside the road, the original pattern is still discernible and has been dated to the 12th- 13th centuries on the basis of pottery found within excavated ditches and banks (Jope 1999, 68). This planned land division may reflect a change of ownership, perhaps the granting of Avebury manor to Cirencester Abbey in the mid- nth century. Jope (1999) summarises the medieval history of Avebury and suggests from the excavated evidence that settlement within the henge did not take place until the medieval period (ibid., 68). Regular tofts and crofts are common to medieval settlements in Dorset and south Wiltshire (Figures 12 and 13). Evidence for planned settlement has long been understood as reflecting the allotment of land to tenants by manorial lords during the period of population growth in the 12th- 1 3th centuries. Many such allotments were abandoned during the population decline of the late 14th and 15th centuries (Taylor, 1994, 218). One of the best excavated examples is Holworth, Chaldon Herring where an abandoned toft revealed occupation and building evidence between the 12th-15th centuries (Rahtz, 1958, 105). Buildings At Avebury, five building-like high resistance anomalies 6-8, 10 and 1 1 were found beside the road in the southeast quadrant. They occupy a terrace with their long axes aligned with the road. Another rectangular building-like anomaly 9 is south of the others and aligned north-south beside a trackway 40 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 15 William Stukeley 1724 Plan E ( Frontispiece in Stukeley’s Abury 1743) 21 that branches from the road. Earthworks within this quadrant (Figure 3) may indicate the positions of buildings constructed of materials other than stone. On the east side of 9, for example, there is a roughly square earthwork enclosure 20m wide with an apparent building platform on its north side. The date of the stone buildings is unknown but their layout suggests episodes of building rather than a single construction phase. Only building 11 was occupied in 1724, 6-10 are not shown by Stukeley (Figure 15), who depicts this field as pasture. Subsequent maps show no buildings in this area. Building 1 1 remained within its own hedged enclosure and was finally demolished between the 1900 and 1924 editions of the Ordnance Survey. Anomalies 6-10 might occupy medieval sites, although buildings may have been constructed and subsequently demolished in the 16th and 17th centuries. The spacing of these buildings and the available boundary evidence does not reveal a pattern of dwelling house and outbuildings grouped in individual tofts (although building 6 lies between linears 1 and 2). John Aubrey’s plan C indicates buildings across this area to the east entrance of the henge on both sides of the road. This suggests that buildings were still standing in 1663 (Ucko 1991, 21), but the plan is rather stylised and cannot be relied upon. On the north side of the road, building-like anomalies are not present, although two buildings are shown on 1886-1924 Ordnance Survey maps (demolished in the mid 20th century). A 5m wide band of high resistance 84 branches from the road at the Pond Stone 44 and then runs parallel and 5m north of it. The anomaly may include the position of the demolished 19th century building, but it seems instead to represent a metalled track. The route of the track ends just east of the prominent rectangular GEOPHYSICAL SURVEY OF AVEBURY HENGE 41 Left: Fig. 16 Rev A.C Smith 1885, stones discovered by probing © Wiltshire Heritage Museum Archive & Library. Right: Fig. 1 7 Ordnance Survey 25 inches to one mile map, 1 900 mound on the east side of a National Trust holiday cottage. A garden is shown here in 1724 and terracing presumably took place before this and would have removed any relationship between the mound and the suggested trackway. The mound has been interpreted as a possible Neolithic long barrow (Mike Pitts pers. comm.), but this seems unlikely based on the geophysical evidence. Its rectangular plan and anomaly 82 - which appears to be a wall defining the summit of the mound - suggest a platform for a substantial building of pre-1724 date. Traces of other likely building sites 83, 87, 88, on the north side of 84 are indicated by rectilinear high resistance anomalies. 83 may be associated with the demolished 19th century cottage, but 87 and 88 are probably of pre-1724 date occupying a hollow levelled out of the slope. The two 19th-century buildings on the north side of the road were demolished in the 20th century as part of a deliberate clearance policy. The removal of the buildings in the southeast quadrant before 1724 may also have been part of a planned clearance after the amalgamation of tenements by a single farmer. This area is shown as one field on Aubrey’s 1663 plan A and was Stukeley’s pasture VII in 1724, presumably held by Tom Robinson who Stukeley names as the person responsible for demolishing the stones there in 1700. This claim was clearly incorrect as Aubrey’s plan shows only the stones 77 and 78 still visible within pasture VII. The other stones in this field, except 90, are now known to have been buried and it is likely that they have been buried since the medieval period. The story told to Stukeley about Robinson’s demolition work must have applied to other megaliths or perhaps to the buildings within pasture VII. Brian Edwards (2001, 54) describes Thomas Robinson as a leading non-conformist who encouraged immigrant dissenters to settle within the henge. The chapel on the south side of Green Street was built in 1670 after the Five Mile Act (1665) had forced non-conformists to worship beyond a five mile exclusion zone around towns. The number of Avebury non-conformists rose from 25 in 1670 to 130 in 1715 and documentary evidence suggests that this population increase speeded the destruction of the megaliths for building stone (ibid. 54). Aubrey’s plan A, however, indicates that many of the stones within the 2003 survey area had been buried or removed before 1663 and that the stone for the dissenters’ chapel, and most of their houses, paths and walls came from other megaliths within the henge. The enclosure map of 1794 (WRO E/A 95) confirms the amalgamation of tenanted plots continuing a process of rationalisation of smallholdings into larger farms that had begun 42 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE in the late medieval period and accelerated in the 17th and 18th centuries. The northeast survey area lies within a field called ‘Gumm’s Close’ and the southeast survey area is held by Robert Nalder, therefore they are fields within farms by that time and devoid of buildings. The land seems to have remained as pasture ever since. The results of the geophysical survey have revealed a certain amount of useful detail concerning occupation and settlement at Avebury. Resurveying the area when the soil moisture conditions are favourable should enhance the existing data and enable low resistance anomalies to be detected. Resurvey of selected areas taking readings at 0.25m intervals would enhance the survey considerably. The resistivity plot of the buried stones has raised the question of whether the megaliths should be excavated and re-erected. From the perspective of the conservation of archaeological deposits, however, this is a controversial proposal. Developing technologies for ground probing radar will enable increasingly accurate 3-D images of buried stones to be generated and it is hoped, therefore, that future research at Avebury will include GPR survey. Virtual re-erection of stones recorded by GPR could, as computer images, form part of an interpretative display of the henge. Acknowledgements The survey was initiated with the help and advice of Mike Pitts and Andrew David and carried out by Geoffrey Brown and Chris Green for the National Trust. Thanks are also due to Brian Edwards for information on post-medieval Avebury, National Trust property staff Rob Mimmack and Hilary Makin for their hospitality and support, and to Ros Cleal who provided advice on the history of research at Avebury References Avebury Archaeological and Historical Research Group, 2001, Archaeological Research Agenda for the World Heritage Site: Salisbury: Wessex Archaeology BEWLEY, R., COLE, M., DAVID, A., FEATHERSTONE, R., PAYNE, A. and SMALL, F. 1996. New features within the henge at Avebury, Wiltshire: aerial and geophysical evidence. A ntiquity 70, 639-46 DAVID, A. and PAYNE, A., 1997, ‘Geophysical Surveys within the Stonehenge Landscape: a review of past endeavour and future potential’, in C. Renfrew and B. Cunliffe, B. (eds), Science and Stonehenge, 73-113. London: Proceedings of the British Academy 92 EDWARDS, B. 2001. Changing Avebury. 3rd Stone 47, 54-58 GILLINGS, M., POLLARD, J., WHEATLEY, D. and PETERSON, R., 2008, Landscape of the Megaliths, Excavation and Fieldwork on the Avebury Monuments 1997-2003. Oxford: Oxbow JOPE, E. M. 1999. The Saxon and Medieval Pottery from Alexander Keiller’s Excavations at Avebury. WANHM 92, 60-91 MALONE, C., 1989, English Heritage Book of Avebury. London: Batsford POLLARD, J. and REYNOLDS, A., 2002, Avebury, the Biography of a Landscape. Stroud and Charleston SC: Tempus POLLARD, J. and CLEAL, R. M. J., 2004, ‘Dating Avebury’ in R. Cleal and J. Pollard (eds). Monuments and Material Culture, papeis in honour of an Avebury archaeologist Isobel Smith, 120-9. East Knoyle: Hobnob Press PITTS, M., 2003, Renewed Archaeological Excavations at Avebury. Unpublished project design RAHTZ, E 1958. Excavations at the Medieval Village of Holworth. Proceedings of the Dorset Natural History and Archaeological Society 80, 103-5 RCHM, \91ti, Dorset, South-East, Volume II, pt. 1. London: HMSO SMITH, Rev. A. C., 1885, Guide to the British and Roman Antiquities of the North Wiltshire Down in a hundred square miles around Abury, 2nd edn. Devizes: WANHS SMITH, I. F., 1965, Windmill Hill and Avebury. Excavations by Alexander Keiller, 1925-39, Oxford: Clarendon Press TAYLOR, C., 1994, ‘The Regular Village Plan: Dorset revisited and revised’ in M. Aston and C. Lewis (eds), The Medieval Landscape of Wessex, 213-218. Oxford: Oxbow Monograph 46 UCKO, P, HUNTER, M., CLARK, A. and DAVID, A., 1991, Avebury Reconsidered. London: Unwin Hyman WAIN WRIGHT, G. and LONGWORTH, I. H., 1971, Durrington Walls: Excavations 1966-1968. London: Reports of the Committee of the Society of Antiquaries of London 29 Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 43-57 Prehistoric and Roman archaeology at Mill Lane, Swindon by Andrew Hood A programme of archaeological monitoring and subsequent excavation was undertaken in advance of road construction. The majority of the monitored groundworks were devoid of archaeological finds or features; however ; numerous archaeological deposits were present on a clay ridge immediately to the east of Mill Lane. The earliest on-site evidence comprised a small assemblage of worked flint datable to the Mesolithic and Neolithic /Bronze Age periods. A small amount of later prehistoric pottery was also recovered. The remains of a 2nd-4th century AD Roman rural settlement , which comprised ditches /enclosures, pits and possible structures was also identified with at least two phases of settlement activity /layout and evidence for landscape management in the form of ditch re-cuts. A substantial number of quarry pits for clay were present along the route of the proposed road for at least 300m. These possibly represent significant medium/large scale industrial activity, perhaps pottery production. The pits are poorly dated, although later than the Roman features and are perhaps associated with the adjacent deserted medieval settlement of West Leaze. Introduction Between July 2008 and June 2009 Foundations Archaeology undertook archaeological monitoring and excavation in advance of, and during, the construction of roads associated with the Swindon Southern Development (Wichelstowe). The work was commissioned by CgMs Consulting Ltd. Topography and geology The site was located at NGR SU 1374 8275, approximately 2km southwest of Old Town, Swindon (Figure 1). The study area was situated on relatively low ground between the chalk escarpment of the Marlborough Downs to the south and the sandstones and limestones of Old Town hill to the north. The excavation areas were located on a roughly circular ridge (105m OD), which overlooked the River Ray, approximately 400m to the northeast. The Scheduled Ancient Monument of West Leaze Medieval settlement (SM 28962/01-03) is situated immediately north of the excavation area. The underlying geology comprised Kimmeridge Clay, overlaid by clay soils. At the time of the investigation the land was former pasture. Archaeological background The study area is situated to the south of the former West Leaze farm, within the parish of Wroughton. Historic records indicate that West Leaze formed part of the Westlecot manor or estate. Westlecot is listed in the Domesday survey as Wichelstote , and is described as containing arable, pasture and meadow Foundations Archaeology, Pembroke Centre, Cheney Manor, Swindon, SN2 2PQ 44 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE land, along with a mill. The Charterhouse map of 1880 (Figure 2) indicates that, at this date, the study area lay within the fields south of West Leaze farm. Later historic maps show that the landscape remained essentially unchanged into the modern period. The earliest systematic investigation of the area comprised a ‘walk over’, undertaken as part of the South Swindon Survey (Pattison 1986). This work identified extensive earthworks, immediately to the north of the current study area, thought to represent the deserted medieval settlement of West Leaze. The earthworks were subsequently designated a Scheduled Ancient Monument and Thamesdown Archaeological Unit conducted limited trial trenching within the area of the earthworks, which yielded a mix of Roman and medieval pottery (Digby 1988). A programme of desk-based assessment and archaeological evaluation prior to the current development is summarised in the archive Post- excavation Assessment Report (Hood 2009, 2.2). This work related to the wider landscape and identified a generally low level of rural/agricultural PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 45 Left: Fig. 2 Charterhouse Map, 1880 Om 1000m « - Iron Age finds/activity I ~~ 1^— 0 -Iron Age pit Right: Fig. 3 HER Data ; Prehistoric Left: Fig. 4 HER Data; Roman Right: Fig. 5 HER Data; Post Roman 46 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 6 Site Plan, Roman Features activity, predominately of the Roman and Medieval periods. Methodology A road centre-line-strip watching brief indicated the presence of archaeological features on the clay ridge to the east of Mill Lane, directly south of the West Leaze medieval settlement. In order to mitigate the archaeological deposits, the centre-line-strip was widened to full road width in the areas containing features. The majority of features were investigated in accordance with the project methodology, whilst a sample of numerous amorphous pits was excavated, prior to the remainder being planned under watching brief conditions (ibid.). Summary of stratigraphic evidence A full stratigraphic description of all archaeological features present within the site is given in the post-excavation assessment report (ibid.). Natural Kimmeridge Clay deposits were encountered at an average depth of 0.37m below the modern ground surface. These were intermittently overlain by a mid grey clay silt subsoil, up to 0.20m thick. The natural deposits and intermittent subsoil were sealed by brown clay silt topsoil, up to 0.40m thick. Numerous features were visible, cut into the top of the natural substrates (Figures 6 and 8). Prehistoric activity Limited prehistoric activity was present in the form of a small number of Mesolithic and Neolithic/ Bronze Age worked flints and a small assemblage of later prehistoric pottery. These finds were generally unstratified or residual material in later features. Roman features Numerous ditches, gullies, pits and postholes datable to the Roman period (Figure 6) by pottery finds suggest activity spanning 2nd to 4th century AD. The Roman features were dispersed across the site. Some ‘pockets’ of stratigraphy were present, indicating that the Roman activity included at least two phases, with the occurrence of pits at a location PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 47 Fig. 7 Roman Features, Selected Sections of former ditches representing a significant change in landscape layout. Several ditches had been re-cut on at least one occasion and, as such, provided evidence for long term landscape management. The majority of ditches appear to represent field boundaries, set out on a co-axial northeast-southwest - northwest- southeast alignment. Ditches [6000]/[6006] and [3012] may have been part of smaller enclosures. The only direct evidence for structural remains within the study area comprised postholes [4010], [4013] and [4016]. Further evidence for Roman structures was provided by ceramic building material (CBM), including tegulae and possible hypocaust fragments. A cluster of pits at the northwestern end of the site contained assorted artefacts, including pottery, fired clay, animal bone and iron nail fragments, along with occasional to frequent charcoal flecks. The composition of the assemblage from these features is highly suggestive of domestic refuse. Amorphous pits A large number of amorphous pits were present across the study area (Figure 8 and Plate 1). The features varied in size but were always relatively shallow, up to a maximum depth of 0.55m. The pits were consistently stratigraphically later than the Roman deposits, with pits [1010], [1013] and 48 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 4 Jyioio] * 4^»'^nos5] ' &>V V C ■%Vj^ ^ = PLANNED UNDER EXCAVATION CONDITIONS Q- = PLANNED UNDER WATCHING BRIEF CONDITIONS OpoVC va MAXIMUM EXTENT OF WATCHING BRIEF Fig. 8 Site Plan, Pit Features Plate 1 General photograph of watching brief strip, showing extensive area of amorphous pits. Looking northwest. Copyright: Foundations Archaeology. Photographer: Andrew Hood. [1014] demonstrably later than ditch fill (1009), which contained Roman and Anglo-Saxon pottery. They were filled with a light orange/grey/brown mottled clay silt, distinct from the darker grey fills associated with Roman features (Plate 2). A mixed Plate 2 Roman enclosure [6000]/ [6006] cut by amorphous pits, in turn, cut by Post-medieval boundary ditch. Looking northeast. Copyright: Foundations Archaeology. Photographer: Andrew Hood. pottery assemblage from the pit fills comprised residual Prehistoric and Roman pottery, along with twelve sherds of medieval pottery. On the whole, the volume of material associated with the pits was relatively low, with an average pottery sherd- PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 49 NORTHWEST FACING SECTION PIT [2006] WEST FACING SECTION PIT [5049] NORTHEAST FACING SECTION [3007] SOUTHWEST FACING SECTION DITCH [1007] and PIT [1010] (1005) [1007] NORTHEAST FACING SECTION DITCH [1007], PIT [1013] and PIT [1014] SOUTHWEST FACING SECTION DITCHES [1050] and [1051] and PIT [1055] 105,80m LAND (1053) DRAIN NWA ASE NORTHEAST FACING SECTION DITCH [6006] and PIT [6010] ITSOTT fNW [1050] 0m EAST FACING SECTION DITCH [6000] and PIT [6002] POTTERY Fig. 9 Pit Features, Selected Sections per-feature count of 5.5, along with generally low amounts of animal bone, CBM, slag and rare to occasional charcoal flecks. Worked flint, by Hugo Lamdin-Whymark The worked flint assemblage comprised two flint flakes, a chert flake, a flint blade, a flint end and side scraper on a flake and an edge-retouched flint flake. The blade was parallel sided and exhibited dorsal blade scars indicating that it was struck from a well prepared blade core; this flint probably dates from the Mesolithic. The flake debitage and other artefacts are not intrinsically datable, but technologically and morphologically they bear affinities with neolithic and Bronze Age industries. All pieces exhibited heavy post-depositional edge-damage and were retrieved from Roman or later archaeological features or as unstratified finds and, as such, they indicate no more than an early Prehistoric presence in the locale. Pottery, by Jane Timby The archaeological work resulted in the recovery of 2115 sherds of pottery, weighing approximately 18.5kg. The assemblage largely dates to the Roman period with smaller amounts of prehistoric, Anglo- Saxon, medieval and post-medieval pottery. The latter, comprising just two sherds, is not discussed further other than for dating purposes. 50 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE The Prehistoric assemblage was sorted into fabrics following the PCRG (1997) guidelines. Roman sherds were coded using the national Roman reference fabric codes (Tomber and Dore 1998), or where not classified, codes based on these. The assemblage was quantified by sherd count and weight and the data entered onto an MS Excel spreadsheet, a copy of which is deposited with the site archive. Very small crumbs were counted and weighed but not sorted into fabrics. The pottery is in quite a fragmented state with an overall average sherd size of only 8.7g, despite the emphasis on Roman material which tends to be better fired and more robust. This might indicate a high level of disturbance of deposits or adverse soil conditions. Pottery was recovered from 54 recorded cuts with further material from 35 other deposits. Some 357 sherds, 17% of the assemblage, came from unstratified collection. In the following report the individual fabrics and forms for each main chronological period are briefly described. The final section considers the assemblage in its local and regional setting. Prehistoric pottery A small group of 28 sherds none of which are featured and most appear to be residual finds in Roman or later contexts. Two fabrics can be defined, a calcined flint-tempered ware and a medium-fine sandy ware. The latter occurs as five sherds and is likely to date to the Iron Age. One sherd was residual in linear [2002]; four sherds were the only ceramic finds in pit [2006]. The flint-tempered sherds do not feature alongside the sandy wares and are less easy to date. All the sherds are residual with the highest incidence, 13 sherds in pit [5034] accompanied by a single Roman sherd. The character of the material might indicate a later Bronze Age date. Roman pottery Continental imports Samian: total 23 sherds of Central Gaulish samian were recovered, all plain and many extremely small and abraded. Recognisable forms include dishes Drag 31 and cups Drag 33. One basesherd from linear [1024] has a broken stamp reading ]ANIM. Overall the samian contributes approximately 1% count and weight to the overall assemblage. Amphora: a single small and abraded sherd of possible amphora was recovered from pit [1029]. Regional coarsewares: Dorset black burnished ware (DOR BB1) (Tomber and Dore 1998, 127) accounts for 11.4% by count of the total assemblage and is the commonest of the regional imports present. The standard repertoire of jars (Figure 10.11), bowls and dishes is present with jars accounting for 54.8% and bowls/dishes 45.2%. Of the latter there are just single examples of the flat-rimmed and flanged-rim forms (Figure 10.19) with most examples being plain-rimmed dishes. The lattice decoration on the jars ranges from acute to just oblique with few sherds of very oblique suggesting a date range from the mid- later 2nd century through to the later 3rd/early 4th century. New Forest wares (NFO CC) (ibid., 141): a small group of five colour-coated sherds, all from beakers, was recovered from linear [2024] and pits [1035 and 2025] suggesting these are later Roman in date. Overwey whiteware (OWY WH) (ibid. 146): two jar sherds from pits [2033] and [5024] suggesting a date in the 4th century. Oxfordshire wares (ibid. 173-6): collectively products from the Oxfordshire industries account for just 1.6% of the total assemblage. The wares include three pieces of white ware mortaria (OXF WH) and 30 sherds of colour-coated (OXF RS). Amongst the colour-coated ware rims are four mortaria (Young 1977, forms C99 and C100); one flanged bowl (ibid., form C51); two dishes (ibid., forms C45 and C47) (Figure 10.15), two bowls (ibid., C75) typical of the 4th-century and a beaker. Local coarsewares: Savernake ware (SAV GT) (ibid., 191): although only 4.5% by count, this ware accounts for 14.7% by weight, a reflection of the fact that most of the sherds derive from large storage jars (Figure 10.1). South-west wares (SOW WS; SOW CC) (ibid., 192): although only the white-slipped ware is defined in Tomber and Dore 1998, the same fabric is present with a colour-coated finish. This fabric first appears in the later 2nd century continuing into the 3rd century and largely features as small flagons. Only a single sherd is present of each ware. North Wiltshire wares: this group of wares not surprisingly accounts for just over 80% of the total assemblage by count, 70.3% by weight. Of this 65.2% (count) comprise reduced grey wares (WIL RE) and 11% the oxidised sandy wares (WIL OX). The remaining 4% consist of black sandy wares (WIL BW), imitations of DOR BB1 forms (WIL BB), colour-coated wares PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 51 Table 1 : Quantified summary of Roman pottery Fabric code Description No No % Wt Wt% EVE EVE % Imports LEZ SA Central Gaulish samian 23 1.1 229.5 1.3 21 1.4 ?AMP amphora? 1 0.0 5 0.0 0 0.0 Regional DOR BB1 Dorset black burnished ware 234 11.4 1910 10.5 219 14.3 NFO CC New Forest colour-coated 5 0.2 139 0.8 0 0.0 OVY WH Overwey white ware 2 0.1 11 0.1 0 0.0 OXF RS Oxon colour-coated ware 30 1.5 306 1.7 73 4.8 OXF WHM Oxon white ware mortaria 3 0.1 23 0.1 0 0.0 Local BWMIC micaceous black ware 1 0.0 9 0.0 0 0.0 GREY misc grey sandy ware 2 0.1 14 0.1 0 0.0 GYWS white-slipped grey ware 1 0.0 36 0.2 0 0.0 OXID miscellaneous oxidised 8 0.4 16 0.1 0 0.0 SAVGT Savernake ware 92 4.5 2681 14.7 22 1.4 SHELL shelly ware 1 0.0 10 0.1 0 0.0 SOW CC South-west colour-coat 1 0.0 2 0.0 0 0.0 SOW WS South-west white-slipped 1 0.0 1 0.0 0 0.0 WIL BW Wilts black sandy ware 48 2.3 261 1.4 51 3.3 WIL BWF finer black ware 4 0.2 28 0.2 0 0.0 WIL BB1 Wilts BB1 copies 10 0.5 112 0.6 22 1.4 WIL BWGR Wilts black grog-tempered 6 0.3 26 0.1 0 0.0 WIL CC Wilts colour-coated ware 6 0.3 87 0.5 0 0.0 WIL GL Wilts glazed ware 1 0.0 9 0.0 0 0.0 WIL GR Wilts grog-tempered 2 0.1 2 0.0 0 0.0 WILOX Wilts oxidised sandy 225 11.0 905 5.0 100 6.5 WIL OXGR Wilts oxidised grog-tempered 5 0.2 23 0.1 5 0.3 WIL RE Wilts grey sandy ware 1336 65.2 11363.5 62.4 1016 66.4 WIL REF Wilts fine grey ware 1 0.0 6 0.0 0 0.0 WW white ware 1 0.0 1 0.0 0 0.0 TOTAL 2050 100.0 18216 100.0 1529 100.0 (WIL CC), oxidised or black grogged wares (WIL OX/BWGR), other grog-tempered wares (WIL GR) and a single glazed ware (WIL GL). Kilns have been documented at Whitehill Farm, Toothill Farm and Purton to the west and north- west of Swindon (Anderson 1979). Production of the sandy wares appears to date from the later lst-early 2nd century; the grog-tempered sandy wares may be slightly earlier. It is unclear whether the later grey wares, which account for most of the wares here, come from the same or nearby sources. The grey wares (WIL RE/BB): a fine to medium sandy fabric. Several variants have been subsumed into this single category belonging both to the earlier and later facets of this industry. The repertoire Table 2: Summary of forms (based on rim estimated vessel equivalents) Form EVE % EVE jar 1210 78.9 dish/bowl 215 14.0 jug 19 1.2 beaker 27 1.8 cup 15 1.0 tankard 2 0.1 mortaria 21 1.4 lid 25 1.6 TOTAL 1534 100.0 52 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 3: Comparison of Mill Lane pottery with other local sites (expressed as % no.) Swindon Groundwell Blunsdon Atworth Hermitage IMPORTS Samian 1.1 1.9 2.5 3 0.7 Other fineware np 0.1 np ★ np amphorae ★ 0.7 0.1 ★ np REGIONAL Dorset/south-west BB1 11.4 12.2 2.1 13 3.8 Lower Nene Valiev colour-coat np ★ * ★ np Lower Nene vallev whiteware np np 0.1 np np Mancetter-Hartshill mortaria np np 0.1 np np New Forest colour-coated ware 0.2 0.3 np ★ np Overwev whiteware 0.1 0.2 np ★ np Oxford black surface whiteware np ★ np np np Oxon red- slipped ware 1.5 3 np 1 1.8 Oxon whiteware mortaria 0.1 0.8 np ★ np Oxon white-slipped mortaria np 0.2 np ★ 0.1 late Roman shellv ware np 1.9 np ★ 1.0 Severn Valiev ware np 0.1 1.4 * np Verulamium-tvpe whiteware np np np np np LOCAL Savernake ware 4.5 2.6 4.9 1.5 41.6 SW oxidised/colour-coated ware ★ 0.7 np 8 np Wiltshire black burnished ware 0.5 0.3 np ★ np Wiltshire black sandv ware 2.5 7.3 10.4 * np Wiltshire colour-coated ware 0.3 0.3 np ★ 0.1 Wiltshire grog-tempered ware 0.4 1.5 22.2 ★ np Wiltshire mica-slipped np ★ np np np Wiltshire glazed ware ★ np np np 0.1 Wiltshire oxidised sandv ware 11 6.8 9.4 10 0.9 Wiltshire reduced sandy ware 65.2 46.8 36.5 45 47.4 UNKNOWN unknown/other 0.7 11.8 11.1 c 10 2.5 TOTAL 99.5 99.5 100.0 100.0 100.0 is dominated by jars (94.4% by EVE) mainly everted rim, expanded rim and hooked rim forms (Figures 10.2, 10.4, 10.5, 10.9-10, 10.12-13 and 10.16). One vessel (Figure 10.4) is overfired with blisters suggesting a second. The remaining 5.6% comprise dishes/bowls (Figure 10.3, 10.18) tankard, jugs (Figure 10.14) and lids. Bowls/ dishes include examples with flat and flanged rims and plain-rimmed dishes (Figure 10.6-7). A small number of black sandy wares (WIL BW): these may reflect an earlier facet of the Wiltshire industry. Vessels include beaded rim and necked jars, lids, plain rimmed dishes and a small ‘S’- shaped bowl (Figure 10.17). The oxidised wares (WIL OX) show a similar range of forms (Figure 10.10). The six colour-coated and single glazed sherd are products of a local fine ware industry previously identified from the assemblage at Wanborough (Anderson 1978) dating to the early 2nd century. Unknown wares: few other wares are present mainly as unfeatured bodysherds. These include a black micaceous ware (BWMIC); a fine shell-tempered ware (SHELL); a white ware (WW) and miscellaneous grey and oxidised sandy wares. Anglo-Saxon pottery Four handmade sherds with an organic temper are present. One sherd was recovered from ditch [1007] and one from linear [1011]; the remaining two sherds were unstratified finds. This tradition is quite a long-lived one and thus the sherds could PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 53 date from anywhere between the 6th and 8th/9th centuries AD. Medieval pottery Twenty-two sherds of medieval date comprise mainly a mix of Minety wares from North Wiltshire and Kennet Valley plain jars/cooking pots. Ten sherds came from unstratified deposits and eight from quarry [1004]. The remaining four sherds came from feature [5082]; pit [2008]; pit [3009] and context (110)B. The local and regional setting The recovered assemblage demonstrates low level activity in the area from later Prehistoric times. The main focus of settlement activity appears to have been in the Roman period, probably initially in the later 1st and early 2nd centuries. Sporadic activity continues at the site through the 2nd and 3rd centuries. The latest Roman pottery dates to the first half of the 4th century. The absence of any typical later Roman products might suggest abandonment from around the mid 4th century or earlier. A sparse scatter of Anglo-Saxon sherds suggests occupation of this date somewhere in the neighbourhood, although these sherds are probably casual losses in agricultural land. Similarly the small number of medieval sherds may derive from field manuring. North Wiltshire was the location of quite an extensive pottery industry in the Roman and later periods, with numerous pottery and tile kilns documented, for example, Purton, Toothill, Whitehill, Minety and Ashton Keynes, which demonstrate the suitability of the local clays for such purposes. It should, however, be noted that apart from one possible warped second, none of the pottery suggested it was waster material. Indeed, there was no evidence for large-scale pottery production within the site. In terms of quality and range, the calibre of the material is not high, with very few imports suggesting that this is a typical rural assemblage and not of a high-status. In terms of the overall vessel profile (Table 2) jars dominate at 78.9% EVE followed by dishes/ bowls at 14%. All other forms contribute less than 2% and include a small number of drinking vessels, mortaria, jugs and lids. Table 3 places the assemblage from the site alongside other quantified assemblages from the immediate locality; Groundwell Ridge, Blunsdon (Timby forthcoming), Lower Widhill Farm, Blunsdon (Timby 2007), Atworth R.oman villa (Timby 2008) and The Hermitage, Swindon (Butterworth and Seager Smith 1997). The current assemblage falls into the general pattern for the area in terms of the limited range of continental and regional imports and the dominance of the local grey wares, although this is noticeably higher for this site compared to the others. To some extent this is a reflection of the chronology of the sites, Lower Widhill Farm, with the lowest, has a higher incidence of earlier Roman material. The percentage of samian is slightly lower, although higher than The Hermitage, but at all the sites other continental imports are low or absent. The proportion of DOR BB1 is similar to that from Groundwell Farm and Atworth, both sites with a more pronounced later Roman component. Catalogue of illustrated pottery 1 . Handmade storage jar in Savernake ware. Fabric: SAV GT. Feature [1004] (1005). 2. Narrow necked jar. Fabric: WIL RE. Feature [1004] (1005). 3. Small bowl with a flat rim. Fabric: WIL RE. Feature [1007] (1009). 4. Necked jar. Well-fired with firing blisters. Fabric: WIL RE. Feature [1024] (1025). 5. Wide-mouthed, necked jar. Fabric: WIL RE. Feature [1024] (1025). 6. Plain-rimmed dish. Fabric: WIL BB. Feature [1024] (1025). 7. Flanged bowl. Fabric: WIL RE. Feature [1024] (1025). 8. Hook-rimmed dish. Young 1977 form C47. Fabric: OXF RS. Feature [1024] (1025). 9. Flared rim jar. Fabric: WIL RE. Feature [1024] (1026). 10. Hook-rimmed jar/ Fabric: WIL OX. Feature [1024] (1026). 11. Flared rim jar. Fabric: DOR BB1. Feature [1024] (1026). 12. Everted rim jar. Fabric: WIL RE. Feature [1050] (1052). 13. Cavetto-rimmed jar. Fabric: WIL RE. Feature [1051] (1056). 14. Strap-handled jug (one handle extant). Fabric: WIL RE. Feature [1045] (1060). 15. Beaded rim dish. Young 1977, form C45. Fabric: OXF RS. Feature [3012] (3014). 16. Everted rim jar. Fabric: WIL RE. Feature [3012] (3014). 17. Small ‘S’ shaped bowl. Fabric: BWSY. Context (4024). 18. Small dish with flaring walls. Fabric: WIL RE. Context (119). 19. Flanged bowl. Fabric: DOR BB1. Context (122). 54 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 10 Pottery Illustrations PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 55 Fired clay and ceramic building material Eight amorphous fragments of fired clay were recovered from five contexts, with no indication of their original use. Fourteen pieces of ceramic building material were also recovered from twelve contexts. Most pieces were small and fragmentary, but at least three pieces of tegulae (roofing tile) could be identified, one with a perforation (nail hole) on the fracture. In addition, one piece from quarry [1004] had traces of combing suggesting it was a fragment of hypocaust. Animal bone5 by Sylvia Warman A total of 293 bone fragments from 239 bones, weighing 2.2kg, were recovered. Of these, 57 bones were identifiable to species and 4 epiphyses were present. The animal bone was largely recovered from ditch fills, of broadly Roman date, with some intrusive Medieval material. The complete assemblage is listed by feature, context and spot- date in the Post-excavation Assessment (Hood 2009). This small assemblage, from which less than a quarter could be identified to species, provided limited information on animal husbandry at the site. Horse, cattle and sheep/goat were present and dog can probably be added, based on the presence of gnaw marks. A bias towards teeth is likely to result from differential preservation and the lack of typically less robust sheep/goat bones is consistent with this. Low levels of gnawing and weathering are consistent with rapid burial of the animal bone at the time of deposition. With such a small assemblage comparison and further interpretation is not appropriate. Charred plant remains, by Wendy Smith Analysis of archaeobotanical samples from the site established that generally well-preserved charred plant remains were present. Where plant macrofossils were identifiable to species level, the assemblages primarily contained cereal grain and chaff (glume bases) of spelt ( Triticum spelta L.). The abundance of cereal remains is unlikely to reflect an absence of cultivation of other crops. Instead, it is more likely that this reflects the regular processing of cereal crops and/ or disposal of cereal crop processing waste; both of which are activities likely to involve heat at many stages (van der Veen et al. 2007). Discussion, by Andrew Hood The road centre-line-strip extended for approximately 1.5 km and was largely cut through waterlogged soils onto the underlying Kimmeridge Clay. The majority of the road strip on the low lying generally waterlogged ground, yielded negligible archaeological remains. Significant deposits were only present on the relatively dry clay ridge, to the east of Mill Lane. The relative paucity of prehistoric evidence is likely to be partly a product of truncation by later activity. There is a known cluster of neolithic and Bronze Age features, including inhumations, on the western brow of Old Town hill, approximately 900m to the northeast of the study area (Figure 3). It is uncertain if the prehistoric material from the site represents evidence for a focus of activity on the clay ridge, but further prehistoric evidence may be present in the area around the site. Roman evidence comprised a dispersed set of features, mainly boundary and enclosure ditches, along with a small number of pits and postholes. The pottery assemblage largely consisted of low status material, dominated by local north Wiltshire black and grey wares. It is likely that the site was part of a larger rural settlement, which spanned the 2nd to 4th centuries. Tegulae and possible hypocaust fragments suggest higher status buildings, although none was present within the excavation areas. Charred cereal grains, along with horse, cattle and sheep/goat bones, in some of the Roman features, suggested that the economic basis of the settlement was mixed agriculture. Although a small amount of slag and one possible fragment of ceramic waster material were present, there was no evidence for a significant industrial element to the settlement. It is likely that activity at the Mill Lane site was distinct from that at the pottery manufacturing site at Toothill (Figure 4), approximately 1.5 km northwest (SAS 1975, 135). The focus of the settlement is unknown, although the recovery of over two thousand sherds of pottery from the site suggested that it was close by. The top of the clay ridge is the most likely area, with the group 56 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE of artefact rich pits at the northwest end perhaps representing an outlying part of the settlement core. It is a distinct possibility that the focus of Roman activity may correspond with the site of the medieval settlement. The recovery of Roman pottery from test pits excavated within the area of the medieval earthworks supports this supposition (Digby 1988, 4-8). Given the limited nature of the investigation, it is not possible to suggest a settlement morphology (Taylor 2007, 19-20). A Roman building, possibly a villa, is situated on the southwest facing slope of Old Town hill, approximately 1km northeast (Passmore 1898). In light of the longevity of settlement activity within the study area, the two sites were contemporary and possibly inter-visible. A total of four sherds of Anglo-Saxon pottery present limited evidence of activity in the vicinity of the site. The small number of sherds indicates casual loss, rather than manuring practice. The numerous pits are difficult to interpret. Although mainly amorphous in plan and profile, they contained consistently homogenous fills, suggesting that they were not the result of rooting/ burrowing activity. The generally shallow profiles of the pits may reflect avoidance of the water table. Poorly dated on the whole, a limited number of stratigraphic associations suggest that they post-date the Roman period. Given the general lack of post-medieval evidence from the site, it is most likely that they represent medieval in-filled clay quarry pits. This interpretation is consistent with their amorphous appearance, the low finds count, and their occurrence over a considerable distance. The purpose of the clay extraction is not entirely clear; there was no significant evidence for pottery production from within the study area and there are no known medieval pottery production sites in the immediate vicinity. It is possible that the clay was utilised as building material, potentially for the earthworks at the adjacent settlement, although this is untested. Possible medieval clay quarry pits were present within evaluation trenches near South Leaze farm, just over 1km to the west of the current study area (Michaels 2003, 5.6.2). The area of pitting within the trench was c. 20m wide and 0.35m deep and is comparable to the evidence from the present site: interpretation is hampered by the limited nature of the fieldwork. Given the proximity between the excavation areas and West Leaze, the paucity of medieval artefacts is surprising. The excavation areas were, however, located to the south of the ditch-and-bank enclosure around settlement earthworks, outside the settlement and away from the focus of domestic activity. The project reported upon here indicates that the clay ridge to the east of Mill Lane was settled in the Roman and Medieval periods, possibly earlier. The poorly drained and frequently waterlogged nature of the surrounding land, suggest that other areas of higher ground in this locale may also contain settlement deposits. Acknowledgements Foundations Archaeology would like to thank Melanie Pomeroy- Kellinger of Wiltshire County Council and Myk Flitcroft of CgMs for their help during the course of the project. The fieldwork was undertaken by Andrew Hood, Roy King, Diana King, Jack Crennell, James Vessey, Phillip Matthews, David Lang and Rodrey Baines. References ANDERSON, A. S., 1978, ‘Wiltshire finewares’, in E Arthur and G. Marsh (eds), Early fine wares in Roman Britain, 373-92. Oxford: British Archaeological Reports 57 ANDERSON, A. S., 1979, The Roman pottery industry in North Wiltshire. Swindon: Archaeological Society Report 2 BUTTERWORTH, C. A. and SEAGER SMITH, R. 1997. Excavations at The Hermitage, Old Town, Swindon. WANHM 90, 55-76 DIGBY, N., 1988, Trial Trenching at West Leaze Deserted Medieval Village. Unpublished report HOOD, A., 2009, Land at Mill Lane, Swindon, Wiltshire: Archaeological Excavation and Associated Watching Briefs (Phase 3); Post- excavation Assessment. Unpublished report: Foundations Archaeology MICHAELS, T., 2003, Swindon Southern Development Area: Archaeological Evaluation. Unpublished typescript report: Foundations Archaeology PASSMORE, A. D., 1898. Notes on a Roman Building, and Interments, lately discovered at Swindon. WANHM 30, 217-221 PATTISON, E, 1986, South Swindon Survey. Unpublished report: Swindon Archaeological Society PCRG, 1997, The study of later prehistoric pottery: general policies and guidelines for publication. Gloucester: Prehistoric Ceramics Research Group: Occasional Papers 1 and 2 (revised) SAS. 1975. Toothill Farm, Lydiard Tregoze. WANHM PREHISTORIC AND ROMAN ARCHAEOLOGY AT MILL LANE, SWINDON 57 70-1, 135 TAYLOR, J., 2007, An Atlas of Roman Rural Settlement in England. York: CBA Research Report 151 TIMBY, J., 2007, Pottery assessment, Lower Widhill Farm, Blunsdon. Unpublished report: Cotswold Archaeology TIMBY, J., 2008, ‘Pottery coarsewares’, in J. Erskine and P Ellis, Excavations at Atworth Roman villa, Wiltshire 1970-1975, WANHM 101, 100-108 TIMBY, J., forthcoming, ‘The pottery from Groundwell Ridge, Blunsdon’. English Heritage. TOMBER, R. and DORE, J., 1998, The National Roman Fabric Collection - A Handbook. London: Museum of London Archaeology Service VAN DER VEEN, M., LIVARDA, A. and HILL, A. 2007. The archaeobotany of Roman Britain: Current state and identification of research priorities. Britannia 38, 181-210 YOUNG, C. J., 1977, Oxfordshire Roman Pottery. Oxford: British Archaeological Reports 43 Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 58-95 An Early Anglo-Saxon cemetery at Aldbourne, Wiltshire by Nick Stoodley, Anthea Boylston and Jacqueline I. McKinley with a contribution by Jorn Schuster, and illustrations by Rob Goller ; S.E. James and Ken Lymer In 2007 Wessex Archaeology excavated parts of an Anglo-Saxon cemetery on the south-west margins of the village of Aldbourne in north-east Wiltshire. The presence of a cemetery had been highlighted by the discovery of six skeletons of unknown date in 1960. The excavations revealed all or parts of 26 inhumation graves in two neighbouring areas. Most graves lay on a south-west-north-east alignment and the majority of burials had been made supine and extended , but one adult male had been laid prone and one grave held the remains of two juveniles laid facing one another. The demographic profile suggests this represents the remains of a ‘domestic’ population. One of the earliest cases of leprosy was recorded in one adult male and evidence for tuberculosis was seen in the remains of an elderly female. One of the children in the double grave had suffered peri-mortem sharp-weapon trauma to the skull. Grave goods were relatively sparse (eight graves) and limited in form ( predominantly knives) indicating a Ith-early 8th century date for the cemetery. Slight variations in grave orientation and distribution, together with frequency of grave goods, suggest a possible temporal or possibly socially hierarchical variation in the use of this apparently Final Phase cemetery. Introduction The site lay on the south-west margins of the village of Aldbourne on the north Wessex Downs c. 10km north-east of Marlborough (NGR 426260 175308; Figure 1). The area investigated was the site of a former food factory, the construction of which had required some levelling of the natural slope across parts of the plot. Bounded on the south-east side by the Marlborough Road, the land falls steeply to the north-west and the east, the ground having been reduced on the latter side during previous house construction. The site lay between the 150m and 155m contours, and the ground level in the areas subject to excavation was between 151.40m (Area 1) and 152.90m (Area 2) aOD. Set on the edge of one of the many small intersecting spurs characteristic of the chalk downland, the site lay close to the upper edge of the south-west to north-east ridge (Middle Chalk) overlooking the village in the valley bottom (River Valley Gravels: Geological Survey Sheet 267; Figure 2). Aldbourne is set at the confluence of six steep-sided valleys, all dry except for that leading to the south-east which carries a stream rising in the north of the village c. 5km downstream to the River Kennet. The downland around Aldbourne, particularly to the west (Aldbourne Chase) and east, is rich in prehistoric earthworks and field systems (Sewell 1988; Crowley et al. 1983): Membury Ring lies c. Wessex Archaeology, Portway House, Old Sarum Park, Salisbury SP4 6EB AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 59 Fig. 1 Site and trench location plan 60 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 2 View of Area 1 from the south, looking downslope to the village to the north and the downland beyond 3.5km to the east, The Ridgeway passes c. 5km to the west, and the major Roman route Ermin Street 3km to the north-east. Prehistoric finds have been made in the vicinity (e.g. SMR ref. SU27NE056; SU276SE110; SU27SE203; WANHM 1986, 240- 1). There is abundant evidence for Romano-British activity in the area including the recovery of coins from several find-spots within the village (e.g. SMR ref. SU27N317). There is no evidence for Early Anglo-Saxon settlement in the vicinity, although one was certainly established by the 10th century (recorded as Aldincbuman in AD 970; Gover etal. 1970; Crowleys al. 1983). A coin of Athelred II was recovered c. 1km south of the village centre (SMR ref. SU27SE418). There clearly was an earlier Anglo-Saxon (possibly mid-5th-6th century) presence in the area, however, as supported by the recovery of several fragments of saucer and small long brooches and other items of metalwork from various find spots on the east and south side of the village (WANHM 1986, 243; 1990, 148; SMR refs. SU27SE420, SU27SE409 and SU27SE 415). The nearest evidence for an earlier settlement comes from Ramsbury (Aldbourne parish) in the Kennet Valley c. 4km to the south. Here, excavations conducted in the 1970s in the High Street revealed Middle Anglo-Saxon (8th-9th centuries AD) iron smelting (Haslam 1980, 1-68). The primary focus of settlement at Aldbourne is likely to have been on the north-east side of the village in the vicinity of the source of the stream and the church (Crowley et al 1983) of which 12th century fabric survives. In 1960, during construction of an extension to a former factory, remains of about six burials (graves cut to c. 0.45-0.60m below ground level) were disturbed (Meyrick 1961). No artefacts were recovered and, on the basis of the good state of bone preservation, the remains were thought to be post-medieval date; a Civil War context was postulated in view of the skirmishes believed to have occurred in the neighbourhood at that time (SMR ref. SU27NW526). The precise location of these graves was not recorded, though additional discoveries appear to have been made at both the south-west and north-east ends of the main factory building. The current whereabouts of the remains are unknown. There is no record of the recovery of human remains during the construction of the original factory building in 1949 (Crowley etal. 1983) but this does not necessarily preclude such - the absence of artefactual remains may simply reflect the circumstance of discovery. Following acquisition of the site by Infinity Homes Ltd., Wessex Archaeology was commissioned to undertake a programme of archaeological investigations in advance of housing development in accordance with planning conditions applied by Kennet District Council and to a brief set by Wiltshire County Council’s Assistant Archaeologist. The site was initially designated for a watching brief but included provision for excavation in the event of the discovery of significant archaeological remains. The watching brief was undertaken in May 2007, with subsequent excavation in two areas of the southern part of the site between August and November 2007. The site archive and finds will be deposited with Wiltshire Heritage Museum under the site code 65080. Investigation Methods The down-slope, north-eastern block of the development area, previously levelled, built up and terraced, revealed no archaeological features or deposits. Land fronting the Marlborough Road had been substantially disturbed, presumably during construction of the factory buildings, and no archaeological remains survived here. Two groups of graves were revealed in the footprints of the building plots; Area 1 (c. 22 x 17m) and Area 2 (c. 21 x 9m)(Figure 3). The area in between was not affected by the development. AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 61 Fig. 3 Areas 1 and 2 showing graves and burial positions 62 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Evidence Archaeological Features and Deposits No Romano-British or earlier features were identified. A number of flints were recovered from the old ground surface/subsoil layer (context 2), together with a few sherds of Romano-British and earlier pottery from the backfills of several Anglo- Saxon graves. The Anglo-Saxon graves were cut into the old ground surface/subsoil, which had been truncated to varying degrees by later construction/ demolition activity. A few sherds of medieval pottery from layer 2 probably indicate bioturbation. Two of the three natural features identified as tree-throw holes were cut by Anglo-Saxon graves. Inhumation Graves The only archaeological features revealed comprised 26 Anglo-Saxon graves and a discrete bundle of disarticulated bone (Object Number (ON) 4). Partial remains of one other burial (context 83) were disturbed by workmen prior to the excavations and it is unclear whether this grave lay in Area 1 or 2. A description of each grave and its contents is presented in the Grave Catalogue (see below). Figures 9-19 show the location of grave goods with the items presented in schematic form and selected body positions shown in detail; all body positions are shown in Figure 3. The dimensions of the adult graves showed a relatively broad range: length 1.49-2.60m, average 1.98m; width 0.47-1. 12m, average 0.78m; with surviving depths of 0.11 -0.56m and an average of 0.21m. There was no obvious spatial patterning (e.g. between Areas 1 and 2) associated with variations in any of the three dimensions. Unsurprisingly, given the small number of immature individuals identified, their grave depths showed a more restricted range - 0.07-0. 16m - much shallower than those of the adults. All graves were broadly sub-rectangular in plan with rounded ends. Most had concave sides with either flat or shallow concave bases. Seven had steep straight sides and three had irregular bases. None showed the presence of any integral features. The majority of the graves lay on a south-west to north-east alignment, with six showing slight deviation; four west-south-west to east-north- east and two east to west. The general alignment respected the natural contour of the ridge. The graves in Area 1 were regularly aligned and appear to have formed several (?five) north-west to south- east rows. Grave 43 appeared spatially dislocated from the rest in this area though it is possible that further graves had existed in the disturbed area to the north-east. Indeed, this area forms one of the two possible two locations for the graves found in 1960 (see above), which were described as being ‘roughly in a line head to foot’ (Meyrick 1961). The graves in Area 2 had a less regular distribution than those in Area 1 (Figure 3). There was no intercutting of graves but six had been subject to modern disturbance associated with the factory. A further grave was truncated during machine stripping, while bone in two others was displaced during machining. There was no evidence for coffins in the form of either stains or furniture. The generally ‘tight’ posture of most of the skeletal remains suggests either the bodies were shrouded/wrapped or that grave fill minimised post-depositional movement. The majority of burials (19) were supine and extended (Figures 3, 9-19). In three graves the skeleton lay on one side (two right, one left) with the legs showing varying degrees of flexion: one other was supine but with the legs flexed to the right. One burial was extended but prone, whilst the burial position was unclear in two cases due to disturbance/truncation. Grave 43 (Figure 12) contained the remains of two juveniles laid facing one another. Grave goods accompanied eight of the inhumation burials, all but one of which lay in Area 1 (Figure 5). An iron knife (ON 3), similar to those from several other graves, was found with a bundle of disarticulated human bone (ON 4) in context 2 (see Grave Catalogue ). Object types were limited to iron knives in five graves, iron buckles in three, singular items of iron (e.g. seax) or worked antler (double- sided composite comb) in six other graves. Most had single items only, but two male graves had three items and one other two items. In most cases objects appear to have been worn or laid on the body at the time of burial. Each grave had a single backfill comprising a slightly reddish grey brown silty clay with occasional flint nodules and chalk pieces, i.e. the redeposited former land surface/subsoil (context 2). AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 63 Human Remains, by Anthea Boy Is ton The remains of 26 Early Anglo-Saxon inhumation burials and a small quantity of disarticulated bone were examined. Methods A skeletal inventory was made following standard procedures (Mays et al. 2002) and skeletal recovery from individual graves was quantified on a five- point scale. The sex of the adults was assessed from the sexual dimorphic traits of the pelvis and the cranium (Bass 1987; Buikstra and Ubelaker 1994; Phenice 1969). The age of the immature individuals was assessed using long bone regression formulae (Scheuer et al. 1980), dental development and long bone length (Hillson 1986, 194; Scheuer and Black 2000; Ubelaker 1989). The age of the adults was assessed from the level of dental wear (Smith 1991) and pelvic changes (Katz and Suchey 1986; Lovejoy et al 1985). Stature was estimated following Trotter (1970). Cranial and post-cranial metrics were obtained using the measurements outlined in Buikstra and Ubelaker (1994). Non-metric traits were recorded in accordance with Berry and Berry (1967) and Finnegan (1973). Dental calculus was graded according with Brothwell (1981) and Hillson (2000). Abscesses and granulomas were recorded by the method of Dias and Tayles (1997) and periodontal disease by the method of Ogden (2008). Diagnosis of degenerative joint disease and osteoarthritis was based on the criteria of Rogers and Waldron (1995). Enthesopathies, the bony outgrowths at the insertions of tendons and ligaments were graded on an interval scale according to their size (Hawkey and Merbs 1995). Bone preservation Grave cuts were difficult to identify and most burials suffered disturbance in machine stripping of the site. Skulls were particularly badly affected and it was not possible to take any cranial measurements. The calcareous burial environment led to good visual preservation of bone. Bone preservation is closely related to bone mass which is at optimal levels in young adults. A fragile spine will degrade quickly in the burial environment and osteoporotic bones are liable to fragment when they are lifted since they are very brittle (Walker 1995). The percentage skeletal recovery (see Grave Catalogue ) is, therefore, partly a reflection of burial conditions/disturbance and partly related to the demographic profile of this small assemblage. Almost 40% of the skeletons are over 90% complete (10) and a further 8 (31%) are more than 50% complete, and most of the dentitions are present. No cranial elements were recovered from seven graves (see Grave Catalogue ). The bone from most graves (11: 42.3%) is poorly preserved, though that from 30.8% (8) is in good condition and in 26.9% (7) it is fair. Demography The in situ remains of twenty-one adults (> 18 yrs) and five juveniles (< 12 yrs) were recovered (Figure 4). In addition, fragments of disarticulated bone from a juvenile ofc. 1-5 yrs and two or three further adults were represented amongst the redeposited bone. Sex estimation The adults comprised 1 1 males or probable males, nine females or probable females, and one individual of indeterminate sex; in this latter case only the lower legs and feet were present. Sex estimation was sometimes challenging due to poor skeletal recovery and limited sexual dimorphism, particularly with regard to the mandibular characteristics. This presented a dilemma with burial 35 where the pelvis is poorly preserved, general skeletal robusticity is borderline and the very gracile mandible suggested the individual was female yet the grave good comprised an unusual seax. Saxon females are, however, sometimes found with such weapons (Buckberry,pm comm.). There can also be a problem of sex estimation in older individuals, with females tending to acquire more masculine characteristics as they age (Walker 1995). Age estimation The cemetery was not completely excavated, which probably explains the absence of adolescents (13-18 yrs) and young adults (18-25 yr). However, several facts emerge: women lived longer than men, despite the rigours of childbirth; only four individuals (both sexes) died between 26-35 years of age; and most men died before the age of 45 (Table 1). The foetus and neonate may have been stillborn or died shortly after birth. Their graves (10 and 68) lay among adult burial but not close to any particular one. This is a departure from the apparently common 64 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE practice of burying infants with unrelated adults in order to avoid their spirit returning to haunt the living (Crawford 1991). Individuals under the age of five are often under-represented in cemeteries of this date, though there are exceptions, e.g. at Collingbourne Ducis where c. 16% of population fell in this category (Egging Dinwiddy forthcoming). Two juveniles were recovered from the same grave (43)(Figure 15) and may have died at the same time. Epidemics of infectious disease would have been the most common cause of death in this age group. However, one of them (49) showed evidence of a more sinister reason for their demise (see trauma). Table 1 : Demographic profile (in situ remains only) Age Male ?Male Female PFemale Indeterminate Total Foetus?full term - - - - 1 1 Neonate 0-6 mth - - - - 1 1 Juvenile 5-12 yr - - - 3 3 Young middle adult c. 26-35 yr 2 - 2 1 - 5 Old middle adult c. 36-45 yr 5 - - - - 5 Mature adult 46+ yr 1 1 2 1 - 5 Old adult 60+ yr - 1 2 1 - 4 Adult >18 yr - 1 - - 1 2 Total 8 3 6 3 6 26 AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 65 Metrics Individuals of both sexes were tall and well built. The average height of the males is 1.72m, with a maximum of 1.78m (burial 17) and a minimum of 1.64m (burial 20). This latter figure is the same as the female average, with a range of 1.57m (burial 57) to 1.69 (burial 38). The males are close in height to those from Anglo-Saxon cemeteries elsewhere e.g. Edix Hill, Barrington, Cambridgeshire and Worthy Park, Kingsworthy, Hants, (mean 1.73m; Duhig 1998; Wells et al. 2003) and Castledyke South, Lincolnshire (mean 1.72m Boylston et al. 1998). The female average is consistently higher than that of the females from the same sites (1.60- 1.63m). Up to the time of the industrial revolution in Britain height, particularly that of males, was remarkably uniform. While growth curves in children usually fall well below modern values, in most cases even an individual who had suffered bouts of famine simply kept growing into his twenties until his genetically determined height had been achieved or children who died were shorter than those who survived to adulthood (McKern 1970; Mays 2008). The platymeric index (demonstrating the degree of anterior-posterior flattening of the proximal femur) is related to the biomechanics of the lower limb and has been used to distinguish between groups using different subsistence strategies. Sedentary people have a higher index than nomadic hunter-gatherers (Ruff et al. 1984). Platymeria in females appears to have different biomechanical properties from that seen in males and to be more related to the local terrain. Alterations in biomechanics due to disease can also affect platymeria. Although the index was higher in the males from Aldbourne there appeared to be no marked difference between the sexes, but there were large variations in the values. The minimum value of 67.6 was recorded for an elderly ?male (burial 60) with scoliosis of the spine, and the maximum of 100 from a middle adult male (burial 14) with mushroom-shaped femoral heads and shallow acetabulae; this long-standing condition might have had an influence on femoral shaft shape. Non-metric traits Minor variations occur in both the cranial and post- cranial skeleton. In the former, these are often extra bones located near the cranial sutures, variations in the form or number of foramina for small blood vessels or retention of open sutures into adulthood, e.g. the metopic suture which normally closes between one and two years of age. Twenty-eight cranial and 16 infracranial traits were recorded; a few individuals demonstrating multiple traits (see Grave Catalogue ). The meaning of these traits is still a subject for debate since their mode of inheritance is unclear and it is not easy to infer relatedness between individuals who share the same trait (Saunders 1989). The post-cranial supracondyloid process observed in burial 23 is most likely to be biomechanical in origin. Pathology A summary of the pathological lesions observed is presented in the Grave Catalogue. Dental health and disease Teeth interact with the environment more than the other hard tissues of the body (Ogden 2008). They bear the marks of dental trauma and lifetime stress and reflect use as tools in daily activities. Dental pathology is influenced by diet and wear patterns indicate longevity. A total of 306 permanent teeth were recovered from eight male and nine female dentitions (Table 2). Thirty-one deciduous and permanent teeth were recorded from three juvenile dentitions. Caries Cavities in the teeth may result from consumption Table 2: Dental disease - teeth affected (includes probable males/females) All males All females Juvenile permanent Juvenile deciduous N % N % N % N % No. teeth 128 65.0 178 77.7 10 100 19 90.5 No. sockets 197 - 229 - 10 21 - Ante mortem tooth loss 40 20.3 39 17.0 0 0 - - Caries 10 7.8 13 7.3 0 - 0 - Enamel hypoplasia 2 1.6 2 1.1 0 - 0 - Abscesses & granulomas 6 2.5 6 2.6 0 - 0 - 66 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE of carbohydrates such as starchy foods and honey - the only sweetening agent used in Anglo-Saxon times. In archaeological populations, cavities are rare in children but increase in frequency with age. Four male and five female dentitions had evidence of dental caries with an overall rate (TPR - true prevalence rate) of 7.4% (Table 2); no cavities were seen in the juvenile’s teeth. Caries frequencies for 51 Anglo-Saxon sites were almost all below 10% with a mean of 4.2% of teeth affected (Roberts and Cox 2003, 191). At Worthy Park 3.33% of teeth were carious (Wells et al. 2003), at Castledyke South the prevalence was 6.6% (Boylston et al. 1998) and at Norton, Cleveland 3.4% (Marlow 1992). The frequency at Aldbourne is higher than the average for the period because of the large proportion of older people in the sample. Ante mortem tooth loss Resorption of the alveolar bone surrounding the tooth sockets occurs as a result of recurrent inflammation of the gums (periodontal disease) and may lead to loosening and eventual loss of the tooth. In addition, extraction of painful carious teeth has been practised over millenia (Lukacs 1995). Eight male and six female dentitions had ante mortem tooth loss of between one and ten (burial 14) teeth with an overall rate of 18.1% (permanent dentitions). The TPR is more than twice the average of 8% for the early medieval period (Roberts and Cox 2003: 193). A missing maxillary incisor from one male dentition (burial 23) may have been related to the facial changes of leprosy. Abscesses and granulomas An attempt has been made to distinguish between abscesses and granulomas: the first is represented by a rounded hole in the jaw for the release of pus and the second has sharper margins and is a manifestation of chronic inflammation. Only two true abscesses were detected, most of the remaining lesions being granulomas. Five of these were found in the males and six in the females, affecting four individuals of each sex. In one case (male, burial 8) an oro-antral fistula had perforated the maxillary sinus. One male (burial 20) had two granulomas above the right maxillary teeth (Figure 5) and another (burial 60) had three in the mandible (one right, two left side). This is about average for the period. Calculus All the dentitions had deposits of mineralised plaque on most of the remaining tooth crowns and where the Fig. 5 Burial 20: granulomas in the right maxilla with hypercementosis of one maxillary molar root was exposed there were also thin deposits in the subgingival position. Deposits were most severe in the dentitions of four adults (male and female) over 36 years of age (most >45 yrs) Their frequency and severity suggest a diet containing porridge or some other soft substance rich in carbohydrates. One adult female (burial 35) had calculus covering the occlusal surfaces of the right maxillary molars and premolars suggesting that these teeth had been little used. Tooth wear Teeth may become worn as a result of attrition, abrasion and erosion. In burial 26 there was severe attrition of all the molar teeth and in burial 35 the pulp chamber of both mandibular first molars was completely exposed. In response to such attrition the teeth continue to erupt to maintain correct occlusion. Root exposure is the inevitable result of this process. Uneven wear to a maxillary canine from burial 57, creating a pointed tooth, may have been the result of dental trauma or the abrasive use of this tooth as a tool. Two males (14 and 23) had wear facets on the anterior surfaces of the mandibular teeth suggesting an exaggerated degree of overbite. Periodontal disease Severe periodontal disease (see tooth loss) in one or both parts of the jaw was recorded in three male dentitions (14, 20 and 32) with less pronounced lesions in three others (17, 23 and 41). Only three of the nine female dentitions were affected (26, 57 and 66). Enamel hypoplasia Interruptions in the development of tooth enamel during childhood occur as a result of episodes of famine or severe childhood illness and leave AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 67 distinctive and permanent pits or grooves. Only two dentitions (female (5) and male (23)) had slight linear grooves - on the mandibular canines in the former and the maxillary central incisors in the latter. This suggests that the environment of Anglo-Saxon Aldbourne was a relatively benign one. Developmental anomalies of the teeth It is quite common for an individual to have either additional or congenitally missing teeth, known as hyper- or hypodontia. One male (23) had the former condition in the form of a small socket for a fifth mandibular incisor. Hypodontia is quite common, most frequently affecting the third molar; the condition was seen in three male and two female dentitions (see Grave Catalogue ). One individual had 30° rotation of the right maxillary central incisor (burial 20); burial 20 had a ‘peg tooth’ in the right third maxillary molar position; and an unusual three-rooted mandibular third molar was seen in burial 32. When space in the jaw is constrained, the third molar may develop but fail to erupt normally. This was observed in two dentitions (burials 38 and 20). Infection Infections in bone may be caused by bacteria, viruses or fungi (Roberts and Manchester 2005). The former can be either non-specific, if it is not possible to identify the species of bacteria that caused them, or the bony changes may be typical of a disease such as leprosy or tuberculosis. Non-specific infections or inflammation are normally seen as deposition of new bone (periostitis), or a more severe and complex suite of events as in osteitis or osteomyelitis, representing inflammation of the cortex and bone marrow respectively. Periostitis was seen in one (burial 41) or both (burial 20) tibiae of two males (TPR 10.8% of tibiae). The condition can be caused by trauma to the shins but where the fibulae are involved and there is bilateral involvement it is more likely to be triggered by an infection. Severe osteitis and periostitis were seen in the left tibia and fibula from burial 63, which were also much thicker than normal. These changes most likely represent a localised infection possibly due to penetrating soft tissue injury. Prevalence rates of 8% were recorded at Norton (Birkett 1992) and Raunds, Northamptonshire (late Anglo-Saxon; Powell 1996). Non-specific rib infection is represented by plaques of new bone on the internal surface of the shaft, which is often thicker and more porous close to the neck. Four mid-thoracic left ribs of an adult female (burial 26) were affected by finely porous new bone formation. These lesions are caused by a chronic chest infection and, from studies of populations with a known cause of death (Kelley and Micozzi 1984), the most likely candidate is tuberculosis. Rib lesions are quite common in British populations from all periods and are often associated with spinal tuberculosis; e.g. the four cases from the late Anglo-Saxon churchyard site at Addingham, Yorkshire (Boylston and Roberts 1996; Roberts etal. 1998). Chronic inflammation of the maxillary sinuses can also produce new bone formation. Caused by dental disease or by recurrent sinus infections it can be exacerbated by pollution or by allergies (Lewis et al. 1995). The condition was observed in two male (burials 8 and 14) and one female (burial 38) maxillae. Both males cases were associated with dental infections; in one example (8) the infection from a tooth socket had spread to the sinus via a fistula. A mean prevalence of 4.7% was recorded for 14 Anglo-Saxon sites by Roberts and Cox (2003, 174); a rate of 5.9% being recorded at Worthy Park (Wells et al 2003). Leprosy is a specific infection caused by Mycobacterium leprae which, unlike other bacteria, attacks the sensory and motor nerves causing a lack of sensation in the hands and feet. As a result, the individual does not recognize pain, wounds do not get treated and other bacteria infect the bones of the extremities. Manifested as a skin condition in its early stages, it is only months or years later that bone becomes infected. The type of disease depends on an individual’s immune response. The lepromatous kind, typical of a lower immune status, affects the nasal region and the hard palate, and symmetrically involves the hands and feet. Tuberculoid leprosy tends to be unilateral and affects those at the higher end of the immune spectrum (Jopling 1984). The 26-35 year old male from grave 22 has minor facial changes: however, the right hand and foot were clearly affected with septic arthritis of several joints (Figures 6 and 7) and both lower legs were thickened as a result of infective changes. In addition, the disruption to the bones of the right foot had placed a strain on the tendons of the femur. The biomechanics of an entire leg are changed when the gait is altered by such a severe, chronic infection of the foot. Eighteen individuals were diagnosed with leprosy among the 1677 recorded from Anglo-Saxon cemeteries dispersed across England (Roberts and 68 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 6 Burial 23: right hand bones illustrating the septic arthritis with ankylosis of an interphalangeal joint ( centre ) and a volar groove on the distal end of the upper proximal phalanx; changes indicative of claw hand deformity Cox 2003, 218), including cases from examples in Cannington, Somerset (Brothwell et al. 2000), Beckford, Gloucestershire (Wells 1996), and Edix Hill, Cambridgeshire (Duhig 1998). The individual with leprosy from Aldbourne is particularly important to the history of the disease in Britain, however, since there are few others where the changes are so distinctive. Developmental Major congenital or developmental abnormalities affect about 1% of the population (Smith 1981) but none was observed in this assemblage. Minor abnormalities of development occur most commonly in the spine and ribcage; for example, transitional vertebrae or cervical ribs as seen in the adult female from grave 37. The same individual probably had complete spina bifida occulta , although the arches were obscured in places by post mortem damage, and there seemed to be only four sacral vertebrae. Often anomalies of the spine are accompanied by an abnormality of the sacrum which develops from the same field at the embryonic stage (Barnes 1994). Os acromiale , non-fusion of the acromion tip on the scapula, was observed in the right scapulae of two females (57 and 75) and one male (60). In the latter case, the distal articular surface of the right humerus was much larger than normal and there was a very pronounced muscle insertion for m. anconeus , which is an extensor of the elbow, with marginal lipping surrounding the joint surface. These lesions, together with the os acromiale , suggest some sort of activity commenced at a young age, possibly archery, where ability to hyperextend the Fig. 7 Burial 23: right foot bones showing the destructive changes to the metatarsophalangeal joints and remodeling of the metatarsals indicative of leprosy arm at the shoulder would be an advantage and to which it has been linked elsewhere (Stirland 2000). They closely resemble the changes found in the elbow of one of the soldiers from the mass grave at Towton, Yorkshire (Knusel 2007). Burial 60 also had a crushed lunate bone in his wrist as a result of a fall on his outstretched arm (see below). Joint disease This category of disease covers both the minor changes that occur as a result of normal ageing of the joints and also more severe abnormalities, which are indicative of osteoarthritis. Degenerative joint disease (DJD) is the most common pathology recorded on the skeleton and is represented by marginal lipping or osteophytosis at the rim of an articular surface. The pattern of change throughout the skeleton is important since spinal changes are normally associated with deterioration of cartilage around the other major joints. Spinal DJD or spondylosis was found in all the male spines, and was particularly extensive in that of one 36-45-year old male (burial 14) where the lumbar vertebrae were most severely affected. Osteophytosis of the vertebral bodies was found in 35% of male but only 19% of female vertebrae. A similar prevalence (26% male and 19% of female vertebrae affected) was found by Wells (2003) at Worthy Park. Such changes are related not only to age but also stress on the lower part of the spine from carrying heavy weights and occupations such as mining (Lawrence 1961). The mature adult male from grave 59 had a severe scoliosis with left-sided compression of two mid-thoracic vertebrae and osteophytosis of the affected vertebral bodies. The lumbo-sacral joint was AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 69 affected in burial 17; he was the tallest individual in the group which would have put additional stress on his spine. Schmorl’s nodes are roughly circular lesions in the vertebral body surfaces related to pressure on the intervertebral disks due to spinal stress. Very few lesions were observed in the female spines (0.95%) compared with the males (8.3%); however, the small numbers may reflect loss of evidence due to both taphonomic changes and senile osteoporosis. The condition affected 11% of individuals from Castledyke (Boylston et al. 1998). Osteoarthritis is a systemic disease characterised by patchy destruction of cartilage on joint surfaces and, where severe, bone-to-bone contact is indicated by polishing (eburnation). Lesions were seen in the remains of five older (>45 years) individuals, mostly female (one male). One group (e.g. left wrist and hand, burial 29) or pair (e.g. hips, burial 66) of joints were generally affected, but in the case of one female (burial 75) the spine, both shoulders and both hands were involved, constituting generalised osteoarthritis which is thought to have a genetic basis (Stecher 1948; Spector et al. 1996). Sometimes a small piece of bone, or joint mouse, sloughs off the joint surface and is absorbed by the synovial fluid, leaving a porous ‘mouse bed’ or cortical defect termed osteochondritis dissecans. The condition was found in a foot bone from two male burials. In burial 38 it was associated with an unusual change indicated by porosity at the base of two metatarsals. Enthesopathies Bone is highly malleable and will remodel in response to physical stress (Kniisel 2000). This adaptation to environmental influences enables inferences to be drawn about activities practised by people in the past. An individual’s bone mass is greatest in early adulthood but regular intensive exercise can cause an increase in bone mass even in later life. Enthesopathies (changes tendon and ligament insertions) can give some indication of the type of activities an individual was performing in his/her daily life. Enthesopathies were seen in all but one (incomplete skeleton) of the male skeletons (91%) but only four of the female (44.4%). Three males and one female had rotator cuff enthesopathies, some in association with cortical defects for the costoclavicular ligaments (burial 14) or biceps enthesopathy (burial 20). Such lesions indicate intensive use of the arm bone, presumably in some kind of subsistence activity. Gluteus maximus enthesopathies were recorded in six individuals, affecting both sexes equally. This muscle is important in assisting in such activities as walking uphill. Achilles tendon enthesopathies were seen in three individuals; the one female affected (burial 66) also had strongly developed muscle markings on her tibiae and fibulae; and one of the males (burial 29) also had lesions in the patellae and soleus insertion on the right tibia, all testament to considerable physical activity performed by this individual. In old age the diameter of the cortical bone in the limbs decreases and there is an increasing risk of osteoporosis and loss of muscle mass. This could be the reason why there are so few enthesopathies in the females from Aldbourne: they are either young adults or very old. There was also a lot of fragmentation of long bones in older individuals which make these markings more difficult to discern. Trauma The elderly male (burial 60) with a crushed lunate (carpal) was the only example of healed trauma. This particularly injury suggested a fall on the outstretched right hand which had compressed the joint between the bones of the wrist and the distal radius. This individual had several pathological changes in his upper limb bones suggesting he might have been involved in interpersonal conflict of one sort or another. The most unusual of these was the increased size of his right elbow joint. A surprising finding was the presence of unhealed cranial trauma in one of the juvenile skulls (burial 49). There is a probable perimortal cut which transects the occipital bone obliquely; linear (51mm), with a slight curve, it appears to have been caused by a sharp weapon, the blade being arrested close to the lambdoid suture. There is a further probable unhealed cut mark above the left orbital roof where the blade had passed through the bone at a very shallow angle with bevelling of the cut surfaces. The cut mark on the frontal bone measures 27mm and on the fragment with the zygomatic process 38mm. A small flake had been removed from the internal surface of the frontal bone and there is a superficial score mark from a sharp weapon just above the zygomatic process. It is more difficult to identify cut marks on the bones of children than on those of adults, particularly when they are badly degraded, but the linearity of these marks together with incised wounds makes it probable that they are indeed perimortal injuries. The question remains as to why this child should have met such a violent end 70 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE at the hands of a probable adult aggressor. He/she was found in a double grave (43) facing the remains of a younger child (burial 44). Metabolic disease This category of disease is mainly represented in bone by signs of vitamin and mineral deficiency or osteoporosis. Peak bone mass is attained between the ages of 18 and 35 years and is greater in males both because of the size of their bones and their later growth spurt (Brickley and Ives 2008). Loss of bone mass occurs in two ways. In women, it decreases after the menopause when the balance between bone formation and resorption is affected, with the latter occurring at a higher rate than the former. However, in some cases it is extreme and dietary or genetic factors may be important. Age-related bone loss occurs in individuals of both sexes and in women exacerbates any problems that may have been produced by the menopause. In archaeological populations it is quite easy to detect osteopenia, the precursor of osteoporosis. Cortical bone grows thinner and forms a lower percentage of the bone diameter on X-radiograph. Often the weakened element is susceptible to breakage and osteoporotic bone characteristically causes fractures of either the distal radius, a vertebral body or femoral neck. By these criteria, the elderly female from grave 25 was osteopenic; most of her vertebrae had been destroyed by taphonomic processes. The probable male from grave 59 was osteoporotic with a compression fracture of a mid- thoracic vertebra accompanied by scoliosis of the spine. Possible evidence for residual rickets deformity, due to a lack of vitamin D in early childhood, was seen in the curved ulna of the female from grave 34. However, it is always more difficult to be definite about the cause of such a curvature in a forearm than in a bone from the lower limbs. Porosity of the orbital roofs ( cribra orbitalia) is recorded as a stress indicator and is thought to be linked with anaemia, normally resulting from iron deficiency. The condition was seen in one adult female from grave 25 (TPR 5.9%). Discussion An assemblage of only 26 burials is too small to be able to draw extensive conclusions about the health of the population or for valid comparisons to be made with contemporaneous groups. The assemblage appears to reflect a ‘normal’ population (range of ages and both sexes), but the age profile is somewhat skewed, possibly due to the partial excavation of a larger cemetery. About 20% did not survive to reach adulthood, a lower percentage than normal for a pre-industrial society, however, children are often under-represented in such archaeological populations. Two infants died in that dangerous period around the time of birth. One of the older children suffered major peri-mortem head wounds involving at least three separate blade injuries. The cut to the frontal bone had been delivered at a shallow angle suggesting that the blow was made by a right-handed individual facing the child. Stature was average for the Anglo-Saxon period with women being in general slightly shorter than men but with less dimorphism than in Romano- British or late medieval periods. The post-cranial metrics supported the finding of biomechanical changes related to an active lifestyle in individuals from both sexes. Dental health was generally good. Heavy calculus indicated a diet that was rich in carbohydrates but generally soft in nature. There was little enamel hypoplasia to suggest serious childhood illness or episodes of undernutrition. The probable case of tuberculosis (burial 26) would have been contracted in one of two ways: either by inhaling the bacterium as a result of coughing or spitting by a person with the disease, or from living in close proximity with diseased cattle who also cough up infected sputum. In a rural situation the latter route could be considered the most likely means of transmission. There was limited evidence for other forms of infection which was often the case in the early medieval period before larger town size increased the level of density- dependent infectious diseases. Metabolic conditions indicative of vitamin and mineral deficiency in early life were also rare at Aldbourne. However, several cases of osteoporosis testified to the high proportion of older adults recovered. Enthesopathies indicated that all the males had an active life, particularly the elderly male from grave 28. The changes recorded to the right elbow and scapula of the adult male from grave 59 are similar to those seen in the upper limbs of archers from the late medieval period. These lesions and the various forms of joint disease are an indication of an active lifestyle and are more frequent and prominent in those practising a rural economy than in sedentary individuals. The most unexpected pathology was a case of AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 71 leprosy (burial 23), with unilateral changes to hand and foot bones and bi-lateral lower leg lesions. This symmetry may suggest that the disease was at the lepromatous end of the immune spectrum although the rhinomaxillary changes, if present, were not severe. Segregation of those suffering from the disease does not seem to have been a feature of the early medieval burial rite, unlike in the later medieval period when individuals with obvious leprosy were normally confined to hospitals and buried within their precincts. In England leprosy was rarely seen before the Norman Conquest and most of the hospitals for those suffering from the disease had closed by the beginning of the 15th century. Leprosy is considered an index of civilisation because once a society rises above a certain poverty level the disease tends to disappear of its own accord. In Britain, although it was most prevalent during the late medieval period, a few cases have been recorded in England from the Romano-British period onwards. These earliest instances are from Poundbury in Dorset (Molleson 1993), Cirencester in Gloucestershire (Manchester and Roberts unpublished) and a possible case in a mature male from West Thurrock, Essex (McKinley 2009). Artefacts, by Nick Stoodley with a contribution by Jorn Schuster The majority of the artefactual material comprised grave goods and formed an intrinsic part of the burial rite. Incidental finds from the grave fills included 18 sherds of pottery, mostly Romano-British but with a few fragments of Beaker and Late Bronze Age date. Worked flint (35 pieces) consisted entirely of chronologically indistinct waste flakes and the burnt flint (13 pieces) was all unworked. Small quantities of the same and similarly dated archaeological components were recovered from the machine- stripped subsoil. Grave goods Eight burials (31%) included grave goods (Figure 5). In addition, an iron knife (ON 3) was found with a discrete deposit of redeposited disarticulated bone in the subsoil/old ground surface (context 2), which yielded several other iron objects. Compared to the majority of contemporary cemeteries from a wider study area (Berkshire, Dorset, Hampshire, Table 3: Types and frequencies of grave goods Grave Knife Seax Buckle Comb Pin Misc 7 • • nail 13 • curved strip 28 • 31 • • 34 • 37 • nail 43 • 74 • clip Context 2 • ON 3 fragments Table 4: percentage of burials inclusive of grave goods from the study area Cemetery No./% with grave goods Jewellery Weapons Aldbourne 8/31 Possibly (?pin) Yes Bargates 18/60 No Yes Bradford Peverell 11/61 Yes Yes Burghfield) 33/66 Yes Yes Hambledon Hill 2/15 No No Didcot 12/71 Yes Yes Long Wittenham II 6/55 Yes ? Monkton Deverill 1/7 No No Portsdown I 3/12 No Yes Portway West 12/50 Yes Yes Purton ‘The Fox’ 6/50 Yes Yes Snell’s Corner 27/82 Yes Yes Trumpet Major 3/38 Yes No Winnall II 26/55 Yes Yes Oxfordshire and Wiltshire), this figure is low, although it may be a function of the excavation of a part of the cemetery (Area 2) where deposition of grave goods was uncommon (see below). None of the assemblages can be described as wealthy and object types and materials are restricted in range: apart for the composite antler comb from grave 31, all are iron (Table 3). This disparity becomes clear when the site is compared with others from the wider area. At Didcot, Oxfordshire (Boyle et al. 1995), for example, in addition to the ubiquitous knives and simple iron buckles, a much wider range of object types was recovered that include shears, gaming pieces, a spindle-whorl, finger rings, necklaces, a spear, work box and pin suite. Likewise at Winnall II, Hampshire (Meaney and Hawkes 1970), the finds include bead and ring necklets, combs, brooches, a pin suite plus silver and copper alloy pins, spindle-whorls and animal bone. A similar situation is found in other cemeteries. Aldbourne notably lacks jewellery and dress accessories and has affinities with Bargates, Dorset (Jarvis 1983) and Portsdown I, Hampshire 72 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 8 Distribution of grave goods (Corney et al. 1969), sites with a preponderance of iron artefacts (Table 4). Detailed descriptions of the grave goods can be found in the Grave Catalogue. Knives Knives are the most common artefact (Table 3), which is unsurprising considering that this all- purpose implement was the most numerous grave good throughout the 5th-8th centuries (Stoodley 1999a, 30; Geake 1997, 102). At Winnall II, 20 interments had knives, while the next most popular artefact was the buckle (n=5); at Snell’s Corner, Hampshire (Knocker 1956) 18 burials had knives and 10 had buckles, and the knife was by far the most popular item at Burghfield, Berkshire (Butterworth and Lobb 1992) having been discovered with 28 of 33 individuals buried with objects. All the knives are of the whittle tang form, and all except for ON 3 (context 2), which was found with a disarticulated collection of human bones, were recovered from graves. Most, if not all, have a straight cutting edge, or one that curves up to the tip of the blade, and a back angled down to the tip (Figures 10-11, 15, 17, 18). Blades of this shape belong to Bohner’s Type C (1958) which, although first appearing in the 6th century, was very popular during the 7th-8th centuries, often outnumbering all other types at individual sites: for example, 59% of knives at Burghfield, 44% at Snell’s Corner and 57% at Winnall II. Three of the knives (graves 7, 13 and 31) have evidence for wooden(?) handles. The knives were found on the left side of the body at waist or slightly above waist level and may have AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 73 been fastened to a belt. No buckle was present in any case, although a belt may have consisted of knotted length of rope or similar. Moreover the blades are aligned in various directions, for example in grave 7 the blade is laid in towards the body with the point towards the head, and this may indicate that some knives had been placed on the body rather than attached to clothing or accessories. Grave 13 is particularly interesting; here the knife (ON 5), which was possibly contained within a sheath (see ‘buckles’ below), lay on the base of the grave to the north (head end) of the skull. There was no evidence of disturbance and this, the longest grave excavated (2.60m), may have been constructed this way in order to accommodate the knife and other grave goods. At Bradford Peverell, Dorset (Keen 1979, 133), a knife was found in a similar position, and it seems that in these examples it was a separate offering - not an element of the deceased’s burial costume. Buckles During the 7th and 8th centuries, buckles ranked as the second most popular artefact (Geake 1997, 79) and this is reflected at Aldbourne. Two buckles are complete and one is fragmentary, all can be described as ‘simple’ (Figures 10-11, 14). Such forms were popular during the 5th-6th centuries (Marzinzik 2003, 30, 46) and the most common in the 7th-8th centuries (Geake 1997, 79). Both intact buckles are simple D-shaped iron fasteners with their pins anchored around the frame. The large buckle from grave 28 (ON 10) is similar in shape and size to the one from grave 4 at Bargates, Dorset. Both the Aldbourne buckles lack a plate that would have secured the strap; the belt must have been attached directly to the loop. The fragmentary buckle from grave 7 (ON 1) is represented by part of a pin and a fragment of an attachment plate which would have been folded around the end of the strap and secured by several rivets. There are traces of mineral preserved textiles on one side of the artefact, which may have come from clothing or a belt. At Winnall II an iron loop and plate was found in grave 4, while at Burghfield mineralised textile was present between the plate of the buckle in grave 323 demonstrating that it was originally riveted to a leather belt. The buckle in grave 13 (ON 9) was found with the knife to the north of the skeletal remains (see above). The small size of the artefact, and its association with a knife, suggest that it was used to secure the strap of a sheath. In grave 7, the buckle lay adjacent to the knife at waist height and presumably had fastened a belt. The buckle in grave 28 was lateral to the left thigh and the alignment of the buckle suggests that the belt had not been worn but had been laid out alongside the body. Sea: c Grave 34 produced the only example of a weapon: an iron seax (ON 13). In contrast to the position of the knives it was discovered on the right side of the pelvis (Figure 16). It has a straight cutting edge and back that is angled down at an obtuse angle to the point. In length it can be classified as a small narrow seax, a type that appears in the later 6th century but which was predominantly deposited in graves during the following century. In width it is more typical of a large narrow seax (the width of a small narrow seax is typically 180-300mm), a type restricted to the 7th century. Given its shape and dimensions it may be correct to view it as a transitional type that bridges the categories of small and large narrow seaxes. A much shorter example of a small narrow seax was found across the left pelvis in Winnall II grave 39 (length 188mm), while another, described as a knife, was found at Portsdown I ‘near the left hand’ (Corney et al. 1969, 23) of the adult male in grave 3 (length 209mm). Both these are classified as seaxes on the basis of their form, not their length. ?Pin Three fragments of an iron shaft were excavated from grave 74 (ON 20). When reconstructed the object tapers down to a point and is a probable pin. The object is missing its head, however, and the possibility that it was a nail should also be considered. Iron pins are relatively rare in 7th-8th century contexts, possibly because of poor preservation. The majority are small copper alloy types (Geake 1997, 66), although more elaborate copper alloy or silver pins, often in pairs and linked by a chain, are type fossils of the period, e.g. at Winnall II (grave 8). Iron examples are known: from Winnall II grave 7, where a fragment of a pin came from under the female’s chin; while at Didcot, an example with a perforated and flattened head was found near the infant’s skull in grave 12 (Boyle et al. 1995, 224). Geake’s (1997, 66-7) analysis has revealed that in the majority of cases the pins were found around the area of the neck. The only other item recovered from grave 74 was an iron staple (ON 19, see below) and the pin may have served to fasten a shroud in an unclothed burial. 74 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Nails , by Jorn Schuster Two iron nails were found in graves (7, 37; Figures 5, 10, 17), both lay next to a knife in the left waist area, and in grave 7 there was also a buckle (ON 1). The provision of a single nail suggests that these are unlikely to have been used in the construction of coffins or boards, although the example from grave 7 had a bent tip and thus had clearly been used prior to deposition. The small number of nails alone should not exclude the possibility of coffins having been present at the time of burial as these could have been constructed without the use of nails, in keeping with the general picture of Anglo-Saxon carpentry and timber working (cf. Evison 1987, 100). Single or small groups of nails are known from graves in other Anglo-Saxon cemeteries (e.g. St Mary’s Stadium, Birkbeck 2005, 38 fig. 21 ;41; Pewsey, Wiltshire, Annable and Eagles 2010, 264 fig. 72; 266; Twyford School, near Winchester, Schuster forthcoming; Eton Rowing Course, Boveney, Boyle et al. 2002, 33 fig. 3.5; Boss Hall, Ipswich, Scull 2009, 111-2). Individual nails may have had ritual or magicical properties as apotropaic amulets to ward off evil. A number of suggested early to late Roman examples have been identified (e.g. Black 1986,223; Dungworth 1998, 153; McKinley and Egging Dinwiddy 2009, 45; Schuster 2009; 2011). It is likely that the examples from Aldbourne are further evidence of this practice in an Anglo-Saxon context. Comb A double-sided composite antler comb with iron rivets was recovered from grave 31 (ON 1 1). It is in a very poor state of preservation, but was constructed from a rectangular plate of antler reinforced along the centre on either side by plates and fastened together by iron rivets. No decoration is visible. This is a very long-lasting type of comb, in use from the 3rd- 13th centuries (MacGregor 1985, 92). The Aldbourne example was found on the left side of the body at waist level (Figure 15), which appears to be a standard location, and according to Geake (1997, 63), indicates that it would have been placed on the body during the funeral, rather than being part of the burial dress. Several examples from Winnall II (graves 5, 11 and 15) were constructed in a similar manner to the Aldbourne example, but each exhibits simple decoration (Meaney and Hawkes 1970, 10, fig. 8). Miscellaneous An iron lug (ON 8) formed out of a strip that was bent into a ring was found with the right finger bones in grave 13. A comparable, but slightly larger object, from Snell’s Corner (grave 28) was found by the left tibia (Knocker 1956, 145, figs 8 & 15), and at Shudy Camps, Cambridgeshire (grave 31; Lethbridge 1936) a similar object was associated with a length of chain. In each case the examples were associated with the lower axial area of the body. An iron clip or staple (ON 19) was found by the right elbow in grave 74 and is possibly a repair to a wooden vessel. Vessels made of organic materials must have been more prevalent than the surviving evidence indicates and such furnishings would probably have contained a food substance that may have derived from the funeral and perhaps represented an offering. In addition to the knife (ON 3) discussed above, several other items of probable Anglo-Saxon date were recovered from the subsoil (context 2) and probably derived from graves either within or external to the excavated areas. A fragment of an iron plate (ON 6) is rectangular but narrows towards one end, while at the other (widest) it is slightly curved; it is possibly part of a buckle plate. Other items comprise triangular- shaped iron fragment (ON 15) and a nail (ON 16). Discussion, by Nick Stoodley Chronology Without stratigraphic relationships between the graves, dating relies on the grave goods. The seax (grave 34) and the assortment of Bohner (1958) Type C knives provide the best indication of date: late 6th and 7th century (-early 8th century) and identifies the cemetery as belonging to a distinctive class - the so-called Final Phase. Aldbourne displays characteristics typical of such cemeteries, including a lack of cremation burials, graves placed in rough rows with a fairly standard orientation, low frequency of grave goods and unaccompanied burials (Hyslop 1963). None of the interments included chronologically distinctive artefacts, although the lack of later 6th and 7th century objects is significant. No objects typical of the later 7th century are present and a broad 7th-early 8th century date is likely for the cemetery. As the cemetery was only partly investigated, it is possible that dating evidence has already been destroyed, or remains to be discovered. It is interesting, however, that Area 1 contained all AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 75 but one of the accompanied burials (see below), which begs the question whether these pre-dated the burials in Area 2, i.e. that there may have been a sequence of burial extending along the ridge from east to west. Meaney and Hawkes (1970, 29) suggested that the unfurnished graves at Winnall II post-dated the furnished ones, but admitted (at the time of writing) that there was no way of proving this. Although grave goods do become scarcer over time, recent work, using the latest developments in radiocarbon dating at the Buttermarket cemetery in Ipswich, Suffolk, shows that an absence of grave goods cannot be interpreted in chronological terms (Scull and Bayliss 1999, 86). Apart from chronology, there are other possible reasons for unaccompanied burials (see below). The orderliness of the graves in Area 1 may indicate that the location of each grave was known, implying a relatively short-lived period of burial, especially in view of a lack of evidence for grave markers (see below). Burial Rite Grave Structure Anglo-Saxon graves of the 5th and 6th centuries had relatively simple grave structures. With the appearance of distinctly high status burials in the 7th century, large graves, more complex mortuary structures, and forms of external marking can be observed. These developments filtered down the social ladder, with structural changes also found in 7th century community cemeteries (Hogarth 1973), for example at Long Crichel, Monkton Deverill,Winall II, and Bargates. There is no evidence for internal or external grave structures at Aldbourne, however, this absence needs to be viewed with caution due to the incomplete excavation of the cemetery and truncation of graves. The orderliness of Area 1 suggests that the graves were visible, probably owing to mounded backfilling of grave. The simplicity of the graves seems to reflect an impoverished community. Evidence from the human remains suggests a relatively well-nourished population, however, and grave construction may not have been an aspect of mortuary practice that the community wished to invest in. The shape and size of graves may have been determined by social or cultural values. In particular the greater effort required to dig a large and regularly-shaped grave may reflect the social standing of an individual or their family (Tainter 1975). At Aldbourne the majority of the graves were sub-rectangular and slightly irregular, i.e. one or more of the sides were not straight. In contrast, over 40% of graves at Bargates and 30% Winnall II had straight sides. The nature of the bedrock does not seem to have been a factor, because although Bargates lay over sand and gravels, at Winnall II the graves were dug into the chalk of the Itchen Valley. It appears that expending energy on the creation of regular graves was not a priority at Aldbourne. Adult graves were generally shorter (average 1.92m, though see grave 13 above) than those at Winnall II (average 2.05m) and Bargates (average 2.13m); the great length of those from the latter possibly a reflection of the softer underlying geology. In contrast, the Aldbourne graves were, on average, wider (0.79m) than those at the other two sites (Winnall II 0.41m, Bargates 0.35m). At 1.02m, grave 16 was particularly wide, but there were several over 0.90m (see Grave Catalogue ). Intra-cemetery comparison of grave depth is unreliable because of differing truncation of the ground surface. It is only a useful form of analysis within individual cemeteries, especially comparing depth against other variables, such as burial wealth (Sherlock and Welch 1992, 91). At Aldbourne there is no correlation between the presence or absence of goods and grave depth and there is no appreciable difference in depth between Areas 1 and 2. Orientation Standard orientation is typical of the 7th-8th centuries and contrasts with 5th-6th centuries where a much greater degree of variation existed (Stoodley 1999a, 64-5). At Aldbourne, orientation was fairly standardised: the majority of graves/burials (c. 76%) lay in a south-west to north-east direction (Figure 3). There are, however, subtle variations: graves 52 and 80 were aligned west-east, while graves 19, 53, 59 and 65 were west-south-west to east-north-east. Except for grave 19, they were all in Area 2, and apart from grave 80 they were near the south-west corner of the site. This variance indicates that orientation was not consistently enforced and that topographical or structural factors, for example, influenced grave alignment. A similar situation is found throughout the study area (Table 5), although several examples demonstrate that such constraint was not always adhered to. For example, at Bargates and Didcot there were two main alignments, north- south and west-east. Although orientation became more regularised during the 7th century, it was not an aspect of mortuary behaviour that was rigidly 76 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 5: grave (and burial) orientation Cemetery Principle orientation In rows Bargates west-east; north-south majority not Bradford Peverell west-east yes Burghfield west-east; north-south majority yes Hambledon Hill west-east unknown Didcot west-east; north-south no Long Crichel south-west-north-east yes Long Wittenham II various no Monkton Deverill west-east-south-west; north-east yes Portsdown I west-east probably Portway West south-north; south- west-north-east majority yes Snell’s Corner south-north probably Trumpet Major south-west-north-east; Psouth-north probably Winnall II west-east yes implemented by all communities. Reasons for the layout of a cemetery are seldom obvious, although it seems probable that graves were aligned according to anthropogenic or natural features. The former seems probable when graves have been arranged in rows. For example, at Winnall II a cemetery boundary may have served as a reference point, and at Burghfield the prehistoric earthwork that the cemetery clustered around formed a focus for the majority of the graves. At Aldbourne it is probable that the graves followed the line of the ridge on which the cemetery was sited. Burial position Extended supine burial was the norm throughout the Anglo-Saxon period, either with the legs straight or slightly flexed to one side. Departures from the standard position (crouched, prone and flexed on one side) are persistent features of the early phase, but with increasingly standardised burial in the 7th century, this variation might be expected to decline. At Aldbourne, a range of minority positions is evident (17% of all known positions) and all but one of these were found in Area 1. The prone adult male (grave 16) is the most unusual burial. Overall the very small number (1.46% of 2196 burials in this writer’s database) of these burials in Early Anglo-Saxon cemeteries strongly suggests that the individual had been singled out for special treatment. Until recently it was believed that the position was associated with sinister connotations: the individual had committed a crime or been involved in an act that had brought disgrace, either upon themselves or others, resulting in the administering of particularly stern punishment (Hirst 1985, 36-7). Prone burial may also have been reserved for those that were feared and were seen as posing a threat: undertaken to keep the spirit in the ground and deter it from returning to haunt the living (Harman et al. 1981, 168; Hirst 1985, 37). Examples can exhibit other characteristics that appear to reinforce this belief, such as the 6th-century woman from Sewerby, East Yorkshire (burial 41) who was buried face down in the upper levels of a grave containing the remains of a wealthy female. It was argued that the body’s position indicated that the woman was struggling to free herself from the grave (Hirst 1985, 39); although alternative interpretations have subsequently been suggested (see Lucy 2000, 78-80). Other cases feature an unusual orientation or a covering of stones (Wilson 1992, 83-4). A prone burial (grave 1) was excavated from the 6th century cemetery at Market Lavington, Wiltshire. This young woman was not accompanied by grave goods, and had been placed in a small irregular cut that was spatially isolated from the rest of the excavated graves (Williams and Newman 2006, 35). An adult male (grave 71) was found in a prone position at Blacknall Field, Pewsey, Wiltshire (Stoodley 2010, 95). This individual was also unaccompanied, and it was found that his left arm had been amputated before death and that both his lower legs and feet were amputated at, or shortly after, death (Stuckert 2010). The fact that these examples differ in important ways to the other burials in their cemeteries supports the notion that the individual was considered different or special. Unfortunately, at Aldbourne the grave was disturbed in antiquity and only the legs remained. The burial did not display any other unusual practices nor any clues as to why this individual was buried in such a manner. Because very few prone burials have been excavated under controlled conditions in Wiltshire, it is difficult to gain an overview of the practice against which the Aldbourne case can be compared and from which a possible understanding of it might emerge. Three burials were made with the body laid on one or other side, with varying degrees of flexion. The latter was most pronounced in grave 19, where the adult male’s legs were flexed 90° at the hip and knee (Figure 12). Flexed burials are found throughout the country (Lucy 2000, 81) and are represented in Wiltshire at Blacknall Field and Market Lavington, although as at Aldbourne they are in a minority and, in common with a prone attitude, it may have been a sign that the individuals were in some way viewed differently. Grave 19 was, however, the shortest grave in the site, and the position may have been necessary AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 77 in order to accommodate the body. At Horton Kirby, Kent (Lucy 2000, 81), several bodies had been bent at the head and feet in order to fit them into the grave. It is possible that a short grave may have been dug in anticipation of the adopted burial attitude and thus an alternative position was deliberately chosen to mark this individual out. The analysis of the skeletal remains showed that this individual was the shortest male within the assemblage, had relatively poor dental health, and had experienced an active and physically strenuous life. His physical condition, or circumstances in his life that resulted in such a condition, may have resulted in the use of this position. The two children in grave 43 had been buried on their sides, facing each other, with their legs slightly flexed (Figure 18). Once again such positions may have been necessary to accommodate them in a grave that was no wider than the average adult grave (grave 43 width 0.77m; average adult grave 0.79m). Alternatively it might be seen as a sentimental response by the mourners to the death of two children that had died prematurely at the same time. Multiple burial Grave 43 is the only example of a grave that contained the remains of two individuals, but it is also unusual because the children were interred at the same time (contemporaneous multiple burial). In a national survey, only 6% of multiple burials are of this type (Stoodley 2002, table 5). The grave lay on the eastern edge of the excavated area spatially separate from the others within Area 1, and although the ground immediately to the east fell outside the area of investigation, it is possible that it had been placed in isolation. Other sites within the study area have produced multiple burials including Winnall II, where an 8 year old child and a 15-16 year old were buried simultaneaously in grave 24. More common are examples where a grave had been reopened to allow an additional burial to take place (consecutive multiple), such as at Meonstoke, Hampshire, where grave 13 originally contained a young infant (12-8 months), but was subsequently enlarged to also take a c. 3 year old infant (Stoodley and Stedman 2001, 156). At Monkton Deverill, Wiltshire , grave 736, enclosed within an annular ditch, contained the remains of an adult male over which a lining of stone blocks had been laid, perhaps to support a lid (Rawlings 1995, 36). Above this structure rested the remains of a child of 7-9 years whose head was at variance to the rest of the burials (north-east as opposed to south-west), and was also the only individual from the cemetery to have a grave good (knife). The Aldbourne example is the only contemporaneous multiple burial from the study area that contained juveniles. Moreover, the older child (burial 49) had perimortem blade injuries to the head. Overall, it is a particularly rare and intriguing case: only 11% of all contemporaneous multiple burials contained a pairing of non-adults (Stoodley 2002, 112), and the figure suggests that it was not common practice to bury juveniles together. This author has argued that multiple burial was a reaction to the stress caused by simultaneous deaths in small agricultural communities, basically a superstitious mechanism aimed at preventing further fatalities (Stoodley 2002, 119-21). The unusual character of the burial may suggest that a similar situation was triggered at Aldbourne, especially by the manner in which one of the juveniles had died. Cemetery layout The construction project gave archaeological access to two areas of what must have comprised a larger cemetery. Consequently, any suggestions regarding the layout and development of the burial ground have to be viewed with caution because they may not be representative of the whole. Graves were disturbed during an extension to the former factory but their precise location is unknown (Meyrick 1961). Photographic evidence of the factory prior to demolition indicates that extensions were added to either end of the original building. It was noted that skeletons, ‘buried roughly in a line head to foot’, were found during the construction of a foundation trench, suggesting that the graves were orientated in an orderly fashion in the same way as in Area 1. The southern extent of the cemetery was established in Area 2 , but graves probably extended further to the north. The northern boundary in Area 1 appears to have been identified, but it is less certain in the southern part of the site because of previous development and it is possible that the 1961 discoveries were made in this part of the site. In all probability the cemetery originally consisted of a linear arrangement of graves strung along the edge of the ridge and may have continued beyond the limits of the excavation to the west, with possible additional outliers to the east. Area 1 was the more carefully organised of the two. There was some attempt to place the graves in rows, especially 78 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE graves 40, 13 and 22 (Figure 3), and several others may have been placed in a similar juxtaposition. Row-grave cemeteries are characteristic of the 7 th- 8th centuries and, compared to earlier practices, their ordered nature may have derived from an ideological shift about how the dead should be laid to rest. The majority of the sites within the study area can be unequivocally classified as row-grave cemeteries, for example Winnall II, or have evidence that strongly indicates that this is the principle on which the cemetery was laid out, for example at Monkton Deverill or Portsdown I (Table 5). In contrast some burial grounds display an unordered arrangement of graves, for example, Didcot and Long Wittenham II, Oxfordshire (Meaney 1964, 54). Aldbourne is unusual because it appears to have had zones of both organised and unorganised graves, although they may originally have been part a continuous band. It is unclear whether there was a gradual transition from one type of layout to the other or if the change was more abrupt. Each scenario would have different implications for an understanding of the cemetery’s development and any social or cultural inferences that are derived from it. The fact that the investigation did not encompass the whole cemetery means that a detailed analysis of the layout, which may provide important chronological and social information about its development, is of limited value. The analysis of more completely excavated cemeteries has provided valuable insights. For example, in Kent at Dover Buckland (Evison 1987) and Finglesham (Hawkes and Grainger 2006) both cemeteries grew organically and exhibit horizontal stratigraphy. Elsewhere, clusters of graves may have been the burial plots of individual households (Harke 1997, 138-9; Stoodley 1999a, 131-35), such as at the 6th century cemeteries of Pewsey and Collingbourne Ducis, Wiltshire (Stoodley 2010; forthcoming a). This type of structure ends with the appearance of row-grave cemeteries in the 7th and 8th centuries. It is now found that individual rows often included a mix of different sexes and ages, for example at Winnall II and Monkton Deverill, and it is possible that the row had now become the burial locale of the household. Despite the fragmentary nature of the evidence, some tentative remarks about the internal structuring of Aldbourne can be made. The graves in Area 1 had been ordered into a number of north-west to south-east aligned rows and included males, females and juveniles (Figure 4). The rows contained too few graves to belong to households, although it is possible that evidence for the continuation of these rows in a southerly direction has been destroyed. The row of three graves on the eastern edge were all those of adult males. It is possible that this whole area was occupied by one household that had organised the graves of its members into rows. Area 2 is completely different: the graves had been placed more randomly and, with one exception, lacked grave goods (Figure 5). The absence of grave wealth may indicate that the interments in this area are later than their counterparts in Area 1 and that by this time the control that had previously governed the laying out of the cemetery had lapsed. Alternative interpretations for this spatial difference are considered in the following section. Social structure and community identity Gender and age During the 5th and 6th centuries the placing of grave goods was determined to a large extent by an individual’s gender and age. In particular, weapons were very strongly confined to males, while jewellery demonstrates a strong association with females (Brush 1993; Stoodley 1999a). Compared to younger individuals, i.e. those under the age of 10- 12 years (where an important age-related threshold demarcating cultural childhood and adulthood appears to have been placed), adults were given a greater number of objects and wealthier assemblages (Crawford 1993; Gowland 2006; Stoodley 2000). With the decline of the grave good rite, the signalling of an individual’s gender and age was affected and the expression of these identities was no longer a major priority. Before these practices ceased, they underwent something of a transformation. Gender still controlled the placing of weapons and jewellery, but there is some evidence to suggest that the rigid division of the previous two centuries was beginning to weaken (Stoodley 1999a, 79). For example, at St Mary’s Stadium, Hampshire, two weapon burials were identified as possible females, of which grave 5488 contained a spearhead and a seax (Birbeck 2005, 43-45). It is notable, therefore, that at Aldbourne the seax was also found with a possible female (grave 34). In a similar way, the age-related thresholds that controlled the deposition of jewellery weaken (Stoodley 1999b, 103). For example, at Didcot, Winnall II, Snell’s Corner and Bradford Peverell (Keen 1977, 120) some of the wealthiest jewellery AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 79 was found with children below 12 years of age. Finally, throughout the study area it is noted that the majority of weapon burials have been found in separate burial locales away from the community cemeteries, either under barrows, or in a type of cemetery that has an unusually high number of males (Stoodley 1999b). Weapons had become a symbol of elite male status rather than a general symbol of masculinity. This might help to explain the very low number of weapon burials at Aldbourne. There is some subtle evidence for gendered behaviour at Aldbourne. Some of the males in Area 1 were found in close proximity and, compared to the females, they had a greater number and range of objects (Figures 4-5). They also displayed a wider variety of burial positions. Gender may still have been an important structuring principle, but instead of the traditional weapon:jewellery dichotomy, a range of more subtle practices was used. The almost total lack of jewellery at Aldbourne requires a comment because as the 7th century progressed women tended to become more visible through the types of jewellery deposited with them (Geake 2002, 147). Jewellery occurs in modest quantities in several Wessex cemeteries, e.g. Winnall II and Didcot. Given the way the Aldbourne cemetery is organised (see above) it is unlikely that there is a group of women adorned with jewellery still waiting to be unearthed. It seems more likely that this was a community with limited access to such accoutrements, or which had decreed that jewellery was no longer to be consigned to the grave. The remains of one full-term foetus, one neonate and three juveniles were recovered (mostly from Area 1) and it is clear that they were treated differently from the adults, for example, burial together in the same grave (grave 43), and none of the juveniles had been placed in an extended supine position (a position which naturally cannot be readily adopted by foetal/neonate individuals). Furthermore, apart from the eldest (9-1 1 year old; grave 43), they had not been given grave goods. The differential treatment of children is recognised throughout the Early Anglo- Saxon period (Stoodley 2000) as a method through which their age and age-related status was signalled. The Aldbourne cemetery is important in showing that this practice continued into the 7th century. Social hierarchy Variation in the quantity and quality of grave goods during the 5th and 6th centuries is often interpreted as signifying differences in social status within local communities (Harke 1997, 138-9; Sherlock and Welch 1992, 102). The reduction in grave goods during the 7th century results in the evidence for social stratification becoming more blurred. The lack of a social hierarchy at Aldbourne, as expressed through portable wealth, is therefore not that surprising. Some of the contemporaneous cemeteries within the study area did, however, contain burials with relatively wealthy assemblages of jewellery, in addition to investing above average amounts of labour and raw materials in the structure of the grave. During the 7th century particularly rich burials emerge that are interpreted as evidence for a regional elite. They are often located away from community burial grounds, usually in elevated locations commanding impressive view sheds (Semple 2003) and they tend to be associated with prehistoric earthworks. Wiltshire has produced ample evidence (ibid.’, Stoodley forthcoming b), for example the barrow burial with weapons at Ford and the female laid on a bed under a barrow at Swallowcliffe Down. This is a rank that used portable wealth, monumentality and topography to signify and symbolise their newfound standing in society. This transformation in burial can be viewed as part of the social change that accompanied the establishment of the West Saxon kingdom (Arnold 1997, 201-210; Yorke 1995, 52-93). By now society had become more clearly ranked, and Aldbourne’s place was on the lower rungs of the ladder. If the lack of burial wealth is anything to go by, it was probably lower than other contemporary communities, such as Winnall II. The Aldbourne cemetery seems to have served a rural community with little disposal of wealth and a flat social hierarchy. It is possible that changes to social structure that took place in the 7th century directly influenced the composition of the cemetery at Aldbourne. As Morris (1983, 55) suggested, richer burials may have been removed from the community to be made in barrows. With the relatively large numbers of barrow burials in Wiltshire this is a possibility. The implication is that inferences regarding the social hierarchy of the living based solely on burial evidence may be misleading. The variation in burial, especially between Areas 1 and 2, may have resulted from social differences within the community. Area 1 contained most of the accompanied burials and the greatest range of burial positions. It also produced most of the juveniles which may indicate that it was a household burial plot. There were also twice the number of males as females which may reflect a genuine imbalance in the 80 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE sexes or result from the excavated sample being non- representative of the cemetery as a whole. Therefore the variation evident in this area may have been part of suite of rites intended to signal the status of this group in opposition to those interred in Area 2. It is possible that Aldbourne had a complex social structure and that different parts of the cemetery served different social groups. The implication is that in Area 2 the burials belonged to members of the extended household or were servants or retainers. The latter notion is not supported by the osteological evidence. For example, evidence for trauma is rare, and all the males had led a physically active life. The few cases of osteoarthritis were found in both areas of the cemetery as were the two cases of infection (graves 28 and 59). The former idea may, however, be supported by the fact that a greater proportion of the adults here were over 45 years of age (60% v 33%): a generationally extended household with segregation, at least in death, of the elders. The two groups may have had equal standing within the community, but used divergent practices as symbols of each group’s identity. As in the 5th-6th centuries, burial was a public affair under the jurisdiction of individual households and the community (Williams 2010) and was invested with social meaning. For example, in grave 13 the knife was not part of the individual’s mortuary costume; the same applies to the comb in grave 31 (both graves were in Area 1). It is possible that both these artefacts were placed in the grave after the interment of the body and were separate to it. The decision whether to deposit objects with the dead, and how they were placed, can be interpreted as a statement, not just about the personal identity of the deceased, but about the wider kin group, household and their social memories. The greater variation in Area 1 may indicate that a broad range of identities and allegiances were being signalled. In addition to marking difference within the community, burial rites may have signalled allegiance with groups from outside. It may be pertinent, therefore, that not all 7th and 8th century cemeteries displayed such standardisation in respect of their layout (see above). Did the variation in layout at Aldbourne have cultural significance? Was it a way in which external links were maintained and reinforced? Landscape The cemetery was located on the steep slopes of a north-west facing ridge of the North Wessex Downs, affording stunning views across the valley below. This spot must have been deliberately chosen and factors influential in this decision will be considered. The cemetery was very close to the edge of the ridge and would have been visible from the valley. Analysis of cemetery location (Lucy 2000, 128) has found that the Early Anglo-Saxons situated their burial grounds around sites with pre-existing funeral associations and earlier settlements, but also on natural ridges and mounds. Such features would have helped to mark out the position of a cemetery, perhaps calling and directing people to it for ceremonies of interment and remembrance (Williams 1997, 25). The ridge at Aldbourne may have functioned in a similar way, encouraging and facilitating social contact, and its role both as a topographic marker and catalyst of social interaction has to be considered as one of the motives influencing its location. The question of the inter-visibility of the cemetery and its view- shed raises questions about how Aldbourne may have been encountered in more general terms by the early medieval people who inhabited and travelled through the area. Engagement with burial sites at particular points may have served to integrate ancestors into the fabric and routine of daily life. This seems to be the case at Overton Hill, West Overton, Wiltshire, where a small group of Early Anglo-Saxon burials were made in several barrows of Romano-British and Bronze Age date (Eagles 1986; Semple 2003). The barrows were located on downland close to the intersection of two major routes, the Roman road from Bath to Mildenhall and the Ridgeway. The decision to place the dead in these barrows must have been intentional and influenced by factors other than the ready availability of a pre-existing monument. In fact a considerable number of Wiltshire’s Anglo-Saxon burials and cemeteries have direct associations with prehistoric monuments (Meaney 1964, 18). It was a practice that could take place in the 6th century, for example at Barrow Clump (Stoodley forthcoming b), but was more prevalent in the 7th century. At Aldbourne, although there are several monuments to the north-west which would have been visible from the cemetery there was no direct spatial relationship and there are no known prehistoric barrows in the immediate vicinity. It may be that the location of the cemetery was made sufficiently obvious by the ridge itself and no additional marker was required. If monument reuse was a means by which groups could legitimise AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 81 claims to territory and resources by linking with a mythical past (Williams 1997, 26), then perhaps it indicates that at Aldbourne the community felt sufficiently secure about their ownership of the land (McKinley pers. comm.)- The location of contemporary settlement would also have been an important element within the landscape and may have determined the siting of the cemetery. Lucy’s (1998) work on the Anglo-Saxon cemeteries of East Yorkshire has shown that the different spatial relationships between the settlements (in valleys) and the cemeteries have a chronological dimension, and that by the 7th century the dead were being placed high up on the Yorkshire Wolds. That the dead were further away from the living might indicate an increasing marginalisation of ancestors (Lucy 1998, 99). No positive evidence for an Early or Middle Saxon settlement has been found in the area around Aldbourne, but documentary sources attest to a 10th century church (Wiltshire SMR) and an associated lay presence is probable. A number of fragmentary pieces of metalwork point to an Anglo-Saxon presence in the 5th and 6th centuries (WANHM 1986, 243; 1990, 148; SMR refs. SU27SE420, SU27SE409 and SU27SE 415), yet the contexts from which these finds originated is unknown and furthermore they are not close enough to the cemetery to have derived from an associated site. It is possible that an earlier Saxon settlement had been established in the valley below the cemetery, as demonstrated elsewhere in country: e.g. at Collingbourne Ducis, settlement evidence was found beside the Upper Bourne, about 150m below the excavated cemetery (Pine 2001, 8-117). A pit containing Saxon pottery was discovered close to the river Avon, 800m to the west of the cemetery at Petersfinger, while at Winterbourne Gunner, pottery was recovered 400m north of the cemetery, close to the river Bourne (Eagles 2001, 206). A cautionary note is sounded by excavations at Grove Farm, Market Lavington, which revealed settlement evidence in close proximity to the cemetery (Williams and Newman 2006, 171-3). Overall there appears to be a range of different spatial relationships in Wiltshire, though in the majority of cases the living and dead were kept separate. The situation as it applied to Aldbourne is presently unknown, yet the spatial relationship between the living and dead must have played an important decision in the creation of the mortuary landscape. Aldbourne: a typical Final Phase cemetery in Wiltshire? During the 7th century, the mortuary landscape of England witnessed the appearance of several new and distinctive types of burial ground. Aldbourne appears to belong to the category of Final Phase cemetery. Examples have been found throughout Wessex, but in Wiltshire few cases have been investigated under modern excavation conditions. Final Phase cemeteries were in use for about a century, from the early 7th-early 8th century (Geake 1992, 84-5), and are distinguished by a number of defining features (Hyslop 1963; see above), but it is the changes that occurred in the provision of grave goods which is key to their identification: objects are fewer in number and are of a different type to those found in the 5th-6th centuries. In particular, brooches and weapons are much rarer. Although roughly half of the interments could be unaccompanied, for example at Winnall II 45% lacked objects, many had a simple assemblage such as a knife and buckle. Yet, a tiny proportion of burials had richer assemblages, such as necklaces that included precious materials, e.g. Didcot grave 7 with a silver pendant, or weapons with elaborate and ornate fittings, for example several of the burials at St Mary’s Stadium, Southampton. Lethbridge (1936) was the first scholar to recognise these cemeteries when excavating in Cambridgeshire, and on the basis of what he considered to be ‘pagan’ objects, e.g. weapons, interpreted them as belonging to newly converted Christians who had not fully dispatched with the practice of depositing grave goods. Leeds (1936, 96-114), who was actually responsible for coining the term ‘Final Phase’, took the opposing stance: he highlighted the presence of grave goods, especially wealthy assemblages, which contrasted with the later unfurnished practice, and argued that such burials were the final expression of a pagan lifestyle. Influential work by Hyslop (1963) and Meaney and Hawkes (1970) reasserted the case for Christianity seeing the decline in burial wealth, the character of some of the objects, west-east orientation and the establishment of a new cemetery away from an earlier and ‘pagan’ predecessor, as resulting from the adoption of a new religious ideology. Recently, scholars have been more cautious about explaining the Final Phase in religious terms, mainly because 82 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE the church appears to have had little influence over burial (Morris 1983), and it has been argued that the changes to mortuary practice resulted from a combination of factors ‘deriving from landscape, social, economic and religious change’ (Boddington 1990, 196). Can Aldbourne be described as a Final Phase cemetery? Can the evidence from it contribute to an understanding of these sites? It displays many of the characteristics; the only omission is above ground structures, which may be a result of disturbance. But on closer inspection a more complex picture emerges: it is the eastern group of interments that are Final Phase in character. In contrast, the burials excavated in Area 2 were largely unfurnished and had a more irregular layout. Importantly there are two different forms of burial that may be spatially separate. In fact, Geake draws attention to a category of cemetery during the 7th and earlier 8th century largely made up of unfurnished extended, supine interments, generally orientated west-east (Geake 1992, 86-7). Because these were contemporary with furnished burials, the relationship between the two categories is complex and cannot be explained in chronological terms, i.e. both traditions existed at the same time. Examples include the large, but yet unpublished, site at Staple Gardens (Winchester, Hampshire) and several of the burial grounds in Hamwic (Southampton, SOU 32, and SOU 862; Cherryson 2010) and possibly in Wiltshire at Monkton Deverill. As a cemetery, Aldbourne exhibits both traditions and is therefore important because it adds to a growing corpus of evidence that demonstrates the variety of burial practices in the 7th and early 8th century. It shows that burial practice was considerably more complex than previously believed both in the study area, but also elsewhere, such as in East Anglia (Penn 2000, 97-98). The important point is that by taking the form of burial, rather than cemetery, as the basis of classification (Geake 1992, 89), it has been possible to highlight a range of practices and to suggest several reasons for this diversity. Overall, Aldbourne has provided evidence for a variety of practices that demonstrate the complexities of mortuary behaviour within a Middle Anglo-Saxon community in Wiltshire. It is only by similar detailed examinations of the county’s 7th- 8th century cemeteries that a better knowledge of mortuary behaviour will be provided and from this an understanding of whether Aldbourne is typical of a Wiltshire ‘type’ of cemetery, or resulted from social and cultural conditions unique to that community. Acknowledgements The archaeological investigations were commissioned and funded by Infinity Homes, and thanks are due to Chris Halligan and Alan Bessell for their support. Thanks are also extended to Sue Farr, former Assistant Archaeologist for Wiltshire County Council, for her advice during the fieldwork and to Melanie Pomeroy- Kellinger, County Archaeologist for Wiltshire County Council, for agreeing to provide part funding with Wessex Archaeology to enable the completion and publication of the project. The project was managed on behalf of Wessex Archaeology by Caroline Budd. The watching brief was undertaken by Steve Beach and the excavations were directed by Rebecca Fitzpatrick. Anthea Boylston would like to thank Dr Jo Buckberry for providing literature on Anglo- Saxon burial practices and for commenting on her manuscript. Grave Catalogue compiled by Kirsten Egging Dimviddy and Jacqueline I. McKinley ( with contributions by Jorn Schuster ) Grave 4 (burial 5) SW-NE sub-rectangular cut with rounded ends, concave sides & flat base. 2.10 x 0.78m, 0.20m deep (base at 151.69 m aOD). Extended, supine with arms across abdomen. Single backfill (6). Human bone: >90%; adult c. 26-35 yr. female Pathology summary: dental caries (1); calculus; enamel hypoplasia; osteophytosis - 4 right & 3 left rib tubercles. Backfill finds: Pottery (Beaker) Grave 7 (burial 8) SW-NE sub-rectangular cut with rounded ends, concave sides & base. 2.19 x 0.85m, 0.20m deep (base at 151.72 m aOD). Extended, supine with hands on left hip. Single backfill (9). Human bone: c. 90%; adult c. 36-45 yr. male Pathology summary: ante mortem tooth loss (2); calculus, granuloma; oro-antral fistula; enthesopathies - 4 rib tubercles; exostoses - occipital, atlas. Grave goods: ON 1 : Folded iron sheet plate with remains of probable buckle frame; no slot for pin visible in centre of folded tube. Rectangular- shaped plate, outer edge probably fragmented, c. 7th-8th century. 4g. Organic materials: Mineral replaced textile remains on almost entire AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 83 84 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE upper surface, mineral replaced organic remains (Pleather) remaining on outer edge of inner surface. The textile is clearly a tabby weave. ON 2: Iron nail, head missing, tip bent, lg (not illus.). PApotropaic ON 7 : Iron knife blade. Whittle tang form, Bohner Type C. 6th to 8th century. 40g Organics: mineralised Pwood on handle, Pminerised leather on blade Backfill finds: Burnt flint Grave 10 (burial 11) not illus. SW-NE sub-rectangular cut with rounded ends, moderately sloping sides & flat base. 0.40 x 0.25m, 0.07m deep (base at 151.94 m aOD). Position unclear, disturbed and incomplete. Single backfill (12). Human bone: c. 50%; foetus/neonate (38 wks gestation ± 2 wks) Grave 13 (burial 14) SW-SW sub-rectangular cut with rounded ends, concave sides & base. 2.60 x 0.68m, 0.26m deep (base at 151.68 m aOD). Extended, supine, with left arm across abdomen & right extended. Single backfill (15). Human bone: c. 95%; adult c. 36-45 yr. male Pathology summary: ante mortem tooth loss (10); periodontal disease; infection & sinusitis - maxilla; Schmorl’s node - T7-10; osteophytosis - cervical, thoracic, lumbar; enthesopathy - 4 right & 6 left ribs, humeri, patellae, calcanei; cortical defect -clavicles; mushroom-shaped femoral heads & shallow acetabulae. Grave goods: ON 5: Iron knife blade. Whittle tang form, Bohner Type C. 6th to 8th century. 36g. Organics: mineralised Pwood on handle ON 8: Iron object, tapering strip, bent. 12g AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 85 ON 9: Iron buckle (part of a sheath strap), small, D-shaped. c. 7th-8th century. 8g Backfill finds: flint Grave 16 (burial 17) not Ulus. NE-SW sub-rectangular cut with rounded ends, shallow sides & flat base. 2.33 x 1.02m, 0.23m deep (base at 151.75 m aOD). Extended, prone. Only legs in situ. Single backfill (18). Human bone: c. 50%; adult c. 26-35 yr. male Pathology summary: dental calculus; periodontal disease; osteophytosis - 1 cervical, 1 thoracic & 1 lumbar vertebra; porosity - lumbo-sacral junction; enthesopathies - femora. Backfill finds: Pottery (Roman), flint Grave 19 (burial 20) WSW-ENE sub-rectangular cut with rounded ends, concave sides & base. 1.49 x 0.84m, 0.20m deep (base at 151.82 m aOD). Flexed, on right side with hands in front of skull. Single backfill (21). Human bone: c. 98%; adult c. 36-45 yr. male Pathology summary: ante mortem tooth loss (8); calculus; granulomas; hypercementosis; impaction - left maxillary M3; periostitis - tibiae, fibulae; enthesopathies - humeri, right radius, left patella; Schmorl’s nodes - Tll-12; osteophytosis - 2 right rib tubercles, 3 cervical, 7 thoracic & 4 lumbar vertebrae Grave 22 (burial 23) SW-NE sub-rectangular cut with rounded ends, irregular sides & concave base. 2.28 x 0.91m, 0.34m deep (base at 151.83 m aOD). Supine with legs slightly flexed to the right; right hand at neck, left across pelvis. Single backfill (24). Human bone: c. 98%; adult c. 26-35 yr. male Pathology summary: ante mortem tooth loss (1); calculus; periodontal disease; 3rd molar agenesis; fracture - 3rd metatarsal; leprosy - rhinomaxillary syndrome, volar grooves, thinning, destruction & ankylosis, collar stud deformity (right hand); osteitis & periostitis - right tibia & fibula; tarsal bars (both feet); septic arthritis - right metatarso-phalangeal joints; enthesopathy - right femur; spina bifida occulta ; cortical defect - right clavicle Grave 25 (burial 26) not illus. SW-NE sub-rectangular cut with rounded ends, concave sides St base. 1.84 x 0.62m, 0.29m deep (base at 151.77 m aOD). Extended, supine with arms by sides. Single backfill (27). Human bone: >90%; adult 46+ yr. female Pathology summary: ante mortem tooth loss (6); abscess/ granuloma; calculus; periodontal disease; infection (periostitis) - 4 left ribs; cribra orbitalia ; osteoporosis - thoracic & lumbar vertebrae; spondylosis - 3 cervical; osteophytosis - 4 cervical vertebrae & frags, thoracic & lumbar, hips; porosity - left clavicle; enthesopathy - left ischium, left humerus, left femur Grave 28 (burial 29) SW-NE sub-rectangular cut with rounded ends, steep sides & undulating base. 1.62 x 0.47m, 0.25m deep (base at 151.76 m aOD). Extended, supine with hands at left hip; disturbed/truncated above legs. Single backfill (30). Disturbed by machining, some displacement of bones. Human bone: c. 50%; adult 46+ yr. ? male 86 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 14 Plan of grave 28 AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 87 Pathology summary: osteoarthritis - left wrist & hand; degenerative joint disease - right hand, left sacro-iliac, left first metatarso-phalangeal; enthesopathy - femora, patellae, right tibia, fibulae, calcanei; exostoses - right humerus. Grave goods: ON 10: Iron buckle (belt), large D-shaped simple, c. 7th- 8th century. 20g Backfill finds: worked & burnt flint Grave 31 (burial 32) SW-NE sub-rectangular cut with rounded ends, concave sides & base. 2.30 x 0.98m, 0.17m deep (base at 151.73 m aOD). Extended, supine with hands over pelvis. Single backfill (33). Human bone: c. 75%; adult 46+ yr. male Pathology summary: dental caries (2); ante mortem tooth loss (6); calculus; periodontal disease; osteochondritis dissecans - 1st proximal phalanx (foot); erosive changes - right 2nd rib; osteophytosis & contour change - right clavicle; degenerative joint disease - 4 cervical, right glenoid, hips, left patella; enthesopathy - left proximal humerus Grave goods: ON 11: Antler comb, double-sided composite, iron rivets, in very poor condition, no visible decoration. 42 88 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE fragments, 24g ON 12: Iron knife blade. Whittle tang form, Bohner Type C. 6th-8th century. 21g. Organics: mineralised ?wood on handle Grave 34 (burial 35) SW-NE sub-rectangular cut with rounded ends, concave sides & base. 1.72 x 0.82m, 0.14m deep (base at 151.80 m aOD). Extended, supine with hands over pelvis. Single backfill (36). Human bone: c. 50%; adult c. 26-35 yr. ? female Pathology summary: dental caries (4); calculus; 3rd molar agenesis; rickets - right ulna Grave goods: ON 13: Iron seax. Straight cutting edge and back. Merovingian (transitional between small and large narrow) type. Evidence for bone/horn handle and possible sheath, c. 7th century. 285mm x 37mm, 199g Grave 37 (burial 38) SW-NE sub-rectangular cut with rounded ends, concave sides & base. 1.98 x 0.71m, 0.20m deep (base at 151.79 m aOD). Extended, supine. Single backfill (39). Disturbed in machining, some displacement of bones. Human bone: c. 98%; adult c. 26-35 yr. female Pathology summary: calculus; impacted mandibular AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 89 90 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE C. 6th to 8th century. 15g ON 21: Flat-headed iron nail, sub-rectangular shank. PApotropaic Backfill finds: Animal bone (frog/toad) Grave 40 (burial 41) not illus. SW-NE sub -rectangular cut with rounded ends, concave sides & base. 2.36 x 0.80m, 0.18m deep (base at 151.64 m aOD). Extended, supine. Single backfill (42). Human bone: c. 80%; adult 46+ yr. male Pathology summary: ante mortem tooth loss (8); caries (1); calculus; periodontal disease; osteitis - left tibia; periostitis - right tibia; osteophytosis - left rib; porosity - left lateral clavicle; enthesopathy - calcanei Backfill finds: Pottery (Roman), worked & burnt flint Grave 43 (burials 44 & 49) SW-NE sub -rectangular cut with rounded ends, concave sides & base. 1.70 x 0.77m, 0.19m deep (base at 151.50 m aOD). Sk 44 slightly flexed on left side, Sk 49 slightly flexed on right side with hands possibly by chest; positioned face to face with skulls & legs touching. Single backfill (45). Human bone: Sk 44: c. 40%; juvenile c. 5-6 yr. Sk 49:c. 70%; juvenile c. 9-11 yr. redep. (probably 44 or 49) >10 bone fragments Pathology summary: Sk 49: sharp weapon trauma - occipital, left orbit & zygomatic process; porosity - mandible Grave goods (Sk 49): ON 17: Iron knife blade. Whittle tang form, Bohner Type C. 6th to 8th century. 14g Backfill finds: Pottery (Roman) including ON 18; flint Grave 46 (burial 47) not illus. SW-NE sub-rectangular cut with rounded ends, concave sides & base. 1.07 x 0.56m, 0.15m deep (base at 151.71 m aOD). Extended, supine; only legs remain. Single backfill (48). Human bone: c. 25%; juvenile c. 5-10 yr. Grave 52 (burial 51) not illus. W-E sub-rectangular cut with rounded ends, moderately sloping sides & irregular base. 1.53 x 0.58m, 0.56m deep (base at 1 5 1 .49 m aOD). Extended, supine with hands over right hip. Single backfill (50). Human bone: c. 50%; adult c. 46+ yr. (probably 60+ yr.) female Pathology summary: ante mortem tooth loss (1); abscesses/ granulomas; caries (2), calculus; osteoarthritis - 2nd- 5th cervical facet joints; spondylosis - 3 cervical; osteophytosis - right glenoid, 1st metacarpals & phalanges (bilateral). Backfill finds: Pottery (Late Bronze Age) Grave 53 (burial 54) not illus. WSW-ENE sub-rectangular cut with rounded ends, moderately sloping sides & concave base. 1.72 x 0.90m, 0.14m deep (base at 152.08 m aOD). Position uncertain; heavily disturbed in antiquity. Single backfill (55). Human bone: <25%; adult c. 46+ yr. female Pathology summary: ante mortem tooth loss (all 16 mandibular teeth). Backfill finds: Pottery (Late Bronze Age), worked & burnt flint Grave 56 (burial 57) not illus. SW-NE sub-rectangular cut with rounded ends, concave sides & flat base. 2.22 x 0.86m, 0.17m deep (base at 152.19 m aOD). Extended, supine. Single backfill (58). Human bone: c. 90%; adult c. 46+ yr. female Pathology summary: ante mortem tooth loss (3); calculus; periodontal disease; 3rd molar agenesis; os acromiale; cysts - acetabulae (subchondral), left hand phalanx (solitary); intervertebral osteochondrosis - 10th thoracic Grave goods (possible): Iron pyrite Backfill finds: Pottery (Late Bronze Age & Roman), worked & burnt flint Grave 59 (burial 60) not illus. WSW-ENE sub-rectangular cut with rounded ends, steep sides & flat base. 2.02 x 1.12m, 0.12m deep (base at 151.92 m aOD). Extended, supine. Single backfill (61). Human bone: c. 85%; adult c. 46+ yr. ? male Pathology summary: ante mortem tooth loss (3); caries (2); granulomas; trauma - crushed right scaphoid; os acromiale - right; enlarged right humeral head; scoliosis - 2 mid-thoracic; osteophytosis - 1 cervical vertebra & thoracic frags.; osteoporosis - frags, thoracic & lumbar vertebrae, sacrum, pelvis; inflammatory joint disease - left distal femur; enthesopathy (inflammatory) - right clavicle, right distal humerus, right proximal ulna, calcanei Backfill finds: flint Grave 62 (burial 63) not illus. SW-NE probably sub-rectangular cut, shallow sides & concave base. 2.15 x 0.58m, 0.11m deep (base at 151.99 m aOD). Extended, supine; only legs remain. Single backfill (64). Human bone: <25%; adult >25 yr. unsexed Pathology summary: osteitis & periostitis - left tibia & fibula; periosteal new bone - right fibula Backfill finds: pottery (Late Bronze Age & Roman), sheep/ goat tooth Grave 65 (burial 66) not illus. WSW-ENE sub-rectangular cut with rounded ends, concave sides & flat base. 1.52 x 0.60m, 0.21m deep (base at 152.18 m aOD). Extended, supine; only legs remain. Single backfill (67). Human bone: c. 50%; adult c. 46+ yr. ? female Pathology summary: ante mortem tooth loss (8); caries (1); calculus; granulomas; subchondral cysts - right acetabulum; osteoarthritis - hips; osteophytosis - right distal radius & proximal ulna (Ptrauma), right AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 91 Fig. 19 Plan of grave 74 acetabulum, left proximal femur; enthesopathy - right proximal humerus, radii, right pelvis, right fibula, calcanei Backfill finds: burnt flint Grave 68 (burial 69) not illus. SW-NE sub -rectangular cut with rounded ends, steep sides & flat base. 0.96 x 0.44m, 0.16m deep (base at 151.95 m aOD). Supine, legs slightly flexed apart & left arm extended. Single backfill (70). Human bone: >75%; neonate (40 wks gestation ± 2 wks) Grave 71 (burial 72) not illus. SW-NE sub-rectangular cut with rounded ends, moderate to steep sides & flat base. 2.25 x 0.37m, 0.17m deep (base at 151.85 m aOD). Supine, probably extended. Single backfill (73). Human bone: <25%; adult > 25 yr. ? male Grave 74 (burial 75) SW-NE sub-rectangular cut with rounded ends, shallow to moderately sloping sides & concave base. 1.80 x 0.76m, 0.15m deep (base at 152.05 m aOD). Extended, supine, left leg slightly flexed. Single backfill (76). Human bone: >90%; adult c. 46+ yr. ? female Pathology summary: ante mortem tooth loss (4); caries (3); calculus; granuloma; osteoarthritis - 2 cervical & 2 thoracic vertebrae & frags, lumbar, acromio-claviculars, 1R & 1L hand joint; os acromiale - right; spondylosis - 5 cervical & 4 thoracic vertebrae & lumbar frags. ; intervertebral osteochondrosis - 4 cervical vertebrae; osteophytosis - left acetabulum; degenerative joint disease - left rib, manubrio-clavicular joints, right glenoid Grave goods: ON 19: Iron clip or staple, possible repair for wooden vessel. 75g ON 20: Iron shaft. Probable pin or nail. Three fragments, no head, lg Backfill finds: flint & animal bone (frog/toad) Grave 80 (burial 81) not illus. W-E sub-rectangular cut with rounded ends, shallow to moderately sloping sides & concave base. 0.80 x 0.27m, 0.19m deep (base at 151.84 m aOD). Supine; heavily disturbed & truncated (longitudinally) by drainage pipe. Single backfill (82). Human bone: c. 10%; adult c. 26-35 yr. male Pathology summary: Schmorl’s nodes - 2 thoracic vertebrae; osteophytosis - 2 thoracic vertebrae; enthesopathy - left patella Context 2 (ON4) Subsoil/old ground surface. Bundle of disarticulated and redeposited human bone deposited in a discrete c. 0.45 x 0.28m area. Disturbed prior to recent development. Includes fragments of skeletal elements from skull and axial skeleton, predominantly elements of upper limb. MNI: three adults >18 yrs. ON 3: Iron knife, recovered as part of the bundle. Context 83 Additional number allocated in post-excavation to bone recovered by workmen prior to an archaeological presence on the site. Probably in situ at time of discovery; left arm & some axial skeletal elements only. NB. Only recorded in the catalogue; not included in Anthea Boylston’s human bone 92 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE report, above (recorded by J. I. McKinley) . Human bone: c. 10%; adult c. 25-35 yr. male References ANNABLE, F. K. and EAGLES, B. N., 2010, The Anglo- Saxon Cemetery at Blacknall Field, Pewsey, Wiltshire. Devizes: WANHS Monograph 4 ARNOLD, C. J., 1997, An Archaeology of the Early Anglo- Saxon Kingdoms, 2nd edn. Guildford: Routledge BARNES, E., 1994, Developmental Defects of the Axial Skeleton, Washington: Smithsonian Institution Press BASS, W. M., 1987 , Human Osteology: a Laboratory and Field Manual, 3rd edn. Columbia: Missouri Archaeological Society BERRY, A. C. and BERRY, R. J. 1967. Epigenetic variation in the human cranium. Journal of Anatomy 101, 361 - 79 BIRBECK, V., with SMITH, R. J. C., ANDREWS, P and STOODLEY, N., 2005, The Origins of Mid- Saxon Southampton: excavations at the Friends Provident St Mary’s Stadium, 1998-2000. Salisbury: Wessex Archaeology Report 20 BIRKETT, D., 1992, ‘Skeletal pathology’, in S. J. Sherlock and M. G. Welch, An Anglo-Saxon Cemetery at Norton, Cleveland, 118-120. London: Council for British Archaeology Research Report 82 BLACK, E. W. 1986. Romano-British burial customs and religious beliefs in south-east England. Archaeological Journal 143, 201-39 BODDINGTON, A., 1990, ‘Models of burial, settlement and worship: the final phase reviewed’ in E. South worth (ed.), Anglo-Saxon Cemeteries: a reappraisal, 177-199. Stroud: Alan Sutton BOHNER, K., 1958, Die Frankischen Altertiimer des Trierer Landes Berlin: Germanische Denkmaler der Volkerwanderungszeit ser B, 1, 1 and 2 BOYLE, A., DODD, A., MILES, D. and MUDD, A., 1995, Two Oxfordshire Anglo-Saxon Cemeteries: Berinsfield and Didcot, Oxford: Thames Valley Landscapes Monograph 8. BOYLE, A., HACKING, P, ALLEN, T. and AMBERS, J., 2002, ‘The Anglo-Saxon grave from Boveney’, in Foreman, S., Hiller, J., and Petts, D., Gathering the people, settling the land: the archaeology of a middle Thames Landscape, Anglo-Saxon to post-medieval, 28-34. Oxford: Oxford Archaeology BOYLSTON, A. and ROBERTS, C. A., 1996, ‘The human remains’, in M. Adams, Excavation of a Preconquest cemetery at Addingham, West Yorkshire’, 24-29. Medieval Archaeology 40, 1-40 BOYLSTON, A., WIGGINS, R. and ROBERTS, C., 1998, ‘Human skeletal remains’, in G. Drinkall and M. Foreman, The Anglo-Saxon Cemetery at Castledyke South, Barton-on-Humber, 21-36. Sheffield: Sheffield Excavation Reports 6 BRICKLEY, M. and IVES, R., 2008, The Bioarchaeology of Metabolic Bone Disease, London: Academic Press BROTHWELL, D., 1981, Digging up Bones, 2nd edn. London: British Museum Publications BROTHWELL, D. R., POWERS, R. AND HIRST, S. M., 2001, ‘The pathology’, in P Rahtz, S. M. Hirst and S. M. Wright, Cannington Cemetery: Excavations 1962-3 of Prehistoric, Roman, Post-Roman, and Later Features at Cannington Park Quarry, near Bridgewater, Somerset, 195-256. London: Britannia Monograph Series 17 BRUSH, K. A., 1993, Adorning the Dead: The Social Significance of Early Anglo-Saxon Funerary Dress in England ( Fifth to Seventh Centuries AD). Unpublished University of Cambridge PhD dissertation BUIKSTRA, J. E. and UBELAKER, D. H. (eds.), 1994, Standards for Data Collection from Human Skeletal Remains. Fayetteville: Arkansas Archeological Survey Research Series 44 BUTTERWORTH, C. A. and LOBB, S. J., 1992, Excavations in the Burghfield Area, Berkshire. Developments in the Bronze Age and Saxon Landscapes. Salisbury: Wessex Archaeology Report 1 CHERRYSON, A. K., 2010, “‘Such a resting-place as is necessary for us in God’s sight and fitting in the eyes of the world”: Saxon Southampton and the development of churchyard burial’, in J. Buckberry and A. Cherryson (eds), Burial in Later Anglo-Saxon England c. 650-1100 AD, 54-72. Oxford: Oxbow Books CORNEY, A., ASHBEE, P, EVISON, V. I. and BROTHWELL, D. 1969. A prehistoric and Anglo- Saxon burial ground, Ports Down, Portsmouth. Proceedings of the Hampshire Field Club and Archaeological Society 24 (1967), 20-41 CRAWFORD, S. 1991. When do Anglo-Saxon children count? Journal of Theoretical Archaeology 2, 17-24 CRAWFORD, S. 1993. Children, Death and the Afterlife in Anglo-Saxon England. Anglo-Saxon Studies in Archaeology and History 6, 83-91 CROWLEY, D. A., (ed.), BAGGS, A. P, FREEMAN, J., and STEVENSON, J. H., 1983 ‘Parishes: Aldbourne’, A History of the County of Wiltshire: Volume 12: Ramsbury and Selkley hundreds; the borough of Marlborough 67- 86 URL: http://www.british-history.ac.uk/report. aspx?compid=66518 DIAS, G. and TAYLES, N. 1997. Abscess cavity - a misnomer. International Journal of Osteoarchaeology 7, 548-54 DUHIG, C., 1998, ‘The human skeletal material’, in T. Malim and J. Hines, The Anglo-Saxon Cemetery at Edix Hill ( Barrington A), Cambridgeshire, 154-199. York: Council for British Archaeology Research Report 112 DUNGWORTH, D., 1998, ‘Mystifying Roman nails: clavus annalis, defixiones and minkisi ’, in C. Forcey, J. Hawthorne and R. Witcher (eds), TRAC 97: Proc. Seventh Annual Theoretical Roman Archaeology Conference, 148-59. Oxford: Oxbow Books EAGLES, B. N. 1986. Pagan Anglo-Saxon burials at West AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 93 Overton. WANHM 80, 103-119 EAGLES, B. N., 2001, ‘Anglo-Saxon presence and culture in Wiltshire 450-675’, in E Ellis (ed.), Roman Wiltshire and After, 199-203. Devizes: WANHS EGGING DINWIDDY, K., forthcoming, ‘The human remains’ in K. Egging Dinwiddy and N. Stoodley, The Anglo-Saxon cemetery at Collingboume Ducis, Wiltshire. Salisbury: Wessex Archaeology EVISON, V. I., 1987, Dover: The Buckland Anglo-Saxon Cemetery, London: Historic Buildings and Monuments Commission for England Archaeological Report 3 FINNEGAN, M. 1978. Non metric variation of the infracranial skeleton. Journal of Anatomy 125, 23-37 GEAKE, H., 1992, ‘Burial practices in seventh- and eighth- century England’, in M. Carver (ed.), The Age of Sutton Hoo, 83-94. Woodbridge: Boydell GEAKE, H., 1997, The Use of Grave Goods in Conversion Period England, c.600-850. Oxford: British Archaeological Reports 261 GEAKE, H., 2002, ‘Persistent problems in the study of Conversion-Period burials in England’, in: S. Lucy and A. Reynolds (eds.), Burial in Early Medieval England, 144-155. London, Leeds: Society for Medieval Archaeoogy Monograph 17 GOWLAND, R., 2006, ‘Ageing the Past: Examining Age Identity from Funerary Evidence’, in R. Gowland and C. Kniisel (eds.), Social Archaeology of Funerary Remains, 143-54. Oxford: Oxbow Books GOVER, J. E. B., MAWER, A. and STENTON, F. M., 1970, Place-Names of Wiltshire. Cambridge: Cambridge University Press HARKE, H., 1997, ‘Early Anglo-Saxon Social Structure’, in J. Hines (ed.). The Anglo Saxons: Towards an Ethnography, 125-170. Woodbridge: Boydell HARMAN, M., MOLLESON, T. I. C. and PRICE, J. P L. 1981. Burials, bodies and beheadings in Romano- British and Anglo-Saxon cemeteries. Bulletin of the British Museum Natural History ( Geology ) 35(3), 145-88 HASLAM, J. 1980. A Middle Saxon Iron Smelting Site at Ramsbury, Wiltshire. Medieval Archaeology 24, 1-68 HAWKES, S. C. and GRAINGER, G., 2006, The Anglo- Saxon Cemetery at Finglesham, Kent. Oxford: Oxford University School of Archaeology Monograph 64 HAWKEY, D. E. and MERBS, C. F. 1995. Activity-induced musculoskeletal stress markers (MSM) and subsistence strategy changes among ancient Hudson Bay Eskimos. International Journal of Osteoarchaeology 5, 324-338 HILLSON, S., 1986, Teeth. Cambridge: Cambridge University Press HILLSON, S., 2000, ‘Dental pathology’, in M. A. Katzenberg and S. R. Saunders (eds), Biological Anthropology of the Human Skeleton, 249-285. New York: Wiley-Liss HIRST, S. M., 1985, An Anglo-Saxon Inhumation Cemetery at Sewerby, East Yorkshire. York: York University Archaeological Publications 4 HOGARTH, A. C. 1973. Structural features in Anglo- Saxon graves. Archaeological Journal 130, 104-119 HYSLOP, M. 1963. Two Anglo-Saxon cemeteries at Chamberlains’s Barn, Leighton Buzzard, Bedfordshire. Archaeological Journal 120, 161-200 JARVIS, K., 1983, Excavation in Christchurch, 1969-1980. Dorchester: Dorset Natural History and Archaeology Society Monograph 5 JOPLING, W. H., 1984, A Handbook of Leprosy. 3rd edn. Oxford: Butterworth-Heinemann KATZ, D. and SUCHEY, J. M. 1986. Age determination of the male os pubis. American Journal of Physical Anthropology 69, 427-435 KEEN, L. 1977. Dorset archaeology in 1977. Proceedings of the Dorset Natural History and Archaeological Society 99, 120 KEEN, L. 1979. Dorset archaeology in 1979. Proceedings of the Dorset Natural History and Archaeological Society 101, 133 KELLEY, M. and MICOZZI, M. 1984. Rib lesions in chronic pulmonary Tuberculosis. American Journal of Physical Anthropology 65, 381-387 KNOCKOR, G. M. 1956. Early burials and an Anglo- Saxon cemetery at Snell’s Corner, near Horndean, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 19, 117-171 KNUSEL, C. J., 2000, ‘Bone adaptation and its relationship to physical activity in the past’, in M. Cox and S. Mays (eds), Human Osteology in Archaeology and Forensic Science, 381-402. London: Greenwich Medical Media KNUSEL, C., 2007, ‘Activity-related skeletal change’, in V. Fiorato, A. Boylston, and C. Kniisel (eds), Blood Red Roses: the Archaeology of a Mass Grave from the Battle of Towton , 2nd edn., 103-118. Oxford: Oxbow Books LAWRENCE, J. S. 1961. Rheumatism in coal miners. Part III: occupational factors. British Journal of Industrial Medicine 110, 270-6 LEEDS, E. T., 1936, Early Anglo-Saxon Art and Archaeology. Oxford: Oxford University Press LETHBRIDGE, T. C., 1936,^4 Cemetery at Shudy Camps Cambridgeshire, Cambridge: Cambridge Antiquarian Society Publications Quarto Series, New Series 5 LEWIS, M. E., ROBERTS, C. A. and MANCHESTER, K. 1995. Comparative study of the prevalence of maxillary sinusitis in later medieval urban and rural populations in northern England. American Journal of Physical Anthropology 98, 497-506 LOVEJOY, C. O., MEINDL, R. S., PRYZBECK, T. R. and MENSFORTH, R. P. 1985. Chronological metamorphosis of the auricular surface of the ilium: a new method for the determination of age at death. American Journal of Physical Anthropology 68, 15-28 LUCY, S., 1998, The Early Anglo-Saxon cemeteries of East Yorkshire. Oxford: British Archaeological Reports 212 LUCY, S., 2000, The Anglo-Saxon Way of Death. Stroud: Alan Sutton LUKACS, J. R. 1995. The ‘caries correction factor’: a new 94 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE method of calibrating dental caries rates to compensate for antemortem loss of teeth. International Journal of Osteoarchaeology 5, 151-156 MacGREGOR, A., 1985, Bone, Antler, Ivory and Horn: the technology of skeletal materials since the Roman period. Oxford: Barnes and Noble MANCHESTER, K. and ROBERTS, C. A., unpublished, Palaeopathological Evidence of Leprosy and Tuberculosis in Britain , University of Bradford: Science and Environment Council Report MARLOW, M., 1992, ‘The human remains: the population’, in S. J. Sherlock and M. G. Welch, An Anglo-Saxon Cemetery at Norton, Cleveland, 107-118, London: Council for British Archaeology Research Report 82 MARZINZIK, S., 2003, Early Anglo-Saxon Belt Buckles (late fifth to early eighth centuries A. D.). Oxford: British Archaeological Reports 357 MAYS, S., 2008, ‘The human remains’, in S. Mays, C. Harding and C. Heighway, The Churchyard, 77-192. York: York University Archaeological Publications 13 MAYS, S., BRICKLEY, M. and DODWELL, N., 2002, Guidelines for the Production of Assessments and Analytical Reports on Human Skeletal Remains Excavated from Archaeological Sites. London: English Heritage MEANEY, A. L., 1964, A Gazetteer of early Anglo-Saxon burial sites. London: George Allen and Unwin MEANEY, A. L. and HAWKES, S. C., 1970, Two Anglo-Saxon Cemeteries at Winnall, Winchester, Hampshire. London: Society for Medieval Archaeology Monograph 4 MEYRICK, O. 1961. Skeletons at Aldbourne. WANHM 58, 48 McKERN, T. W. 1970. Estimation of skeletal age: from puberty to about 30 years of age’, in T. D. Stewart (ed.), Personal Identification in Mass Disasters, 41-56. Washington: Smithsonian Institution Press McKINLEY, J. I., 2009, ‘Human bone’, 20-32, in P. Andrews, ‘West Thurrock: Late prehistoric settlement, Roman burials and the medieval manor house. Chanel Tunnel Rail Link Excavations 2002. Transactions of the Essex Society for Archaeoogy and History 40, 1-77 McKINLEY, J. I. and EGGING DINWIDDY, K. 2009. ‘Deviant’ burials from a late Romano-British cemetery at Little Keep, Dorchester. Proceedings of the Dorset Natural History and Archaeological Society 1 30, 43-62 MOLLESON, T. L., 1993, ‘The human remains’, in D. E. Farwell and T. L. Molleson, Excavations at Poundbury 1966-80. Volume II: The Cemeteries, 142-214. Dorchester: Dorset Natural History and Archaeological Society MORRIS, R., 1983, The Church in British Archaeology. London: Council for British Archaeology Research Report 47 OGDEN, A., 2008, ‘Advances in the palaeopathology of the teeth and jaws’, in S. Mays and R. Pinhasi (eds), Advances in Human Palaeopathology, 283-307. Chichester: John Wiley PENN, K., 2000, Norwich Southern Bypass, Part II: The Anglo-Saxon Cemetery at Harford Farm, Caister St Edmund. Gressenhall: East Anglian Archaeology 92 PHENICE, T. 1969. A newly developed visual method of sexing in the os pubis. American Journal of Physical Anthropology 30, 297-301 PINE, J. 2001. The excavation of a Saxon settlement at Cadley Road, Collingbourne Ducis, Wiltshire. WANHM 94, 88-117 POWELL, F., 1996, ‘The human remains’, in A. Boddington, Raunds Furnells: the Anglo-Saxon Church and Churchyard, 113-124. London: English Heritage Archaeological Report 7 RAWLINGS, M. 1995. Archaeological sites along the Wiltshire section of the Codford-Ilchester Water Pipeline. WANHM 88, 26-49 ROBERTS, C. A., BOYLSTON, A., BUCKLEY, L., CHAMBERLAIN, A. E. and MURPHY, E. M. 1998. Rib lesions and tuberculosis: the palaeopathological Evidence. Tubercle and Lung Disease 1, 55-60 ROBERTS, C. and COX, M., 2003, Health and Disease in Britain from Prehistory to the Present Day. Stroud: Alan Sutton ROBERTS, C. and MANCHESTER, K., 2005, The Archaeology of Disease, 3rd edn. Stroud: Alan Sutton ROGERS, J. and WALDRON, T., 1995, A Field Guide to Joint Disease in Archaeology. Chichester: John Wiley RUFF, C. B., LARSEN, C. S. and HAYES, W. C. 1984. Structural changes in the femur with the transition to agriculture on the Georgia coast. American Journal of Physical Anthropology 64, 125-36 SAUNDERS, S. R., 1989, ‘Nonmetric skeletal variation’, in M. Y. Iscan and K. A. R. Kennedy (eds ), Reconstruction of Life from the Skeleton, 95-108. New York: Alan Liss SCHEUER, L. and BLACK, S., 2000, Developmental Juvenile Osteology. London: Academic Press SCHEUER, J. L., MUSGRAVE, J. H. and EVANS, S. P 1980. The estimation of late fetal and perinatal age from limb bone length by linear and logarithmic regression. Annals of Human Biology 7, 257-65 SCHUSTER, J., 2009, ‘Metalwork’, 32-4, in Andrews, P, West Thurrock; late prehistoric settlement, Romano- British burials and the medieval manor house, Channel Tunnel Rail Link excavations 2002. Essex Archaeology and History 40, 1-77 SCHUSTER, J., 2011, ‘Springhead metalwork’, in Biddulph, E., Seager Smith, R. and J. Schuster, Settling the Ebbsfleet Valley: CTRL excavations at Springhead and Northfleet, Kent - the late Iron Age, Roman, Saxon and medieval landscape Vol. 2: Late Iron Age and Roman finds reports, 189-291. Salisbury: Oxford Wessex Archaeology SCHU STER, J., forthcoming , ‘Grave goods’, in K. Egging Dinwiddy, An Anglo-Saxon cemetery at Twyford, near Winchester. Hampshire Studies 65, 95-146 SCULL, C., 2009, Early Medieval ( late Sth-early 8th centuries AD) cemeteries at Boss Hall and Buttermarket, Ipswich, Suffolk. Leeds: Society for Medieval Archaeology Monograph 27 AN EARLY ANGLO-SAXON CEMETERY AT ALDBOURNE, WILTSHIRE 95 SCULL, C. and BAYLISS, A., 1999, ‘Dating burials of the seventh and eighth centuries: a case study from Ipswich, Suffolk’, in J. Hines, K. H. Nielsen and F. Siegmund (eds). The Pace of Change: Studies in Early Medieval Chronology, 80-88. Oxford: Oxbow Books SEMPLE, S. 2003. Burials and political boundaries in the Avebury Region, North Wiltshire. Anglo-Saxon Studies in Archaeology and History 12, 72-91 SEWELL, A., 1988, The prehistory of the Aldboume Basin. Privately published SHERLOCK, J. S. and WELCH, M., 1992, An Anglo-Saxon Cemetery at Norton, Cleveland. York: Council for British Archaeology Research Report 82 SMITH, D. W., 1981, Recognizable Patterns of Human Deformation: Identification and Management of Mechanical Effects on Morphogenesis. Philadelphia: W.B. Saunders SMITH, H., 1991, ‘Standards of human tooth formation and dental age assessment’, in M. A. Kelley and C. S. Larsen (eds), Advances in Dental Anthropology , 143-168. New York: Wiley- Liss SPECTOR, T. D., CICUTTINI, F., BAKER, J. and HART, D. 1996. Genetic influences on osteoarthritis in women: a twin study. British Medical Journal 312, 940-944 STECHER, R. M. 1948. Heberden’s nodes: the clinical characteristic of osteoarthritis of the fingers. Annals of the Rheumatic Diseases 7, 1-10 STIRLAND, A. J., 2000, Raising the Dead: The Skeleton Crew of King Henry VIII’s Great Ship, the Mary Rose. Chichester: John Wiley STOODLEY, N., 1999a, The Spindle and the Spear: A Critical Enquiry into the Construction and the Meaning of Gender in the Early Anglo-Saxon Burial Rite. Oxford: British Archaeological Reports 288 STOODLEY, N. 1999b. Burial Rites, Gender and the Creation of Kingdoms: the Evidence from Seventh Century Wessex. Anglo-Saxon Studies in Archaeology and History 10, 101-109 STOODLEY, N. 2000. From the Cradle to the Grave: Age Organisation and the Early Anglo-Saxon Burial Rite. World Archaeology 31/3, 456-72 STOODLEY, N., 2002, ‘Multiple Burials, Multiple Meanings? Interpreting the early Anglo-Saxon multiple interment’, in S. Lucy and A. Reynolds (eds), Burial in Early Medieval England, 103-121. Leeds: Society for Medieval Archaeoogy Monograph 17 STOODLEY, N., 2010, ‘The social structure’, in B. Eagles The Anglo-Saxon Cemetery at Blacknall Field, Pewsey, 87-100. Devizes: WANHS Monograph 14 STOODLEY, N., forthcoming a, ‘Discussion of burial practices’, in K. Egging Dinwiddy and N. Stoodley forthcoming, The Anglo-Saxon Cemetery at Collingboume Ducis. Salisbury: Wessex Archaeology STOODLEY, N. forthcoming b. Barrow Clump, Figheldean, Wiltshire. Unpublished English Heritage report STOODLEY, N. and STEADMAN, M. 2001. Excavations at Shavards Farm, Meonstoke: the Anglo-Saxon cemetery. Proceedings of the Hampshire Field Club and Archaeological Society 55, 129-169 STUCKERT, C., 2010, ‘The human remains’ in B. Eagles The Anglo-Saxon Cemetery at Blacknall Field, Pewsey, 111-138. Devizes: WANHS Monograph 14 TAINTER, J. R. 1975. Social inference and mortuary practices: an experiment in numerical classification. World Archaeology 1, 1-15 TROTTER, M., 1970, ‘Estimation of stature from intact limb bones’, in T. D. Stewart (ed.), Personal Identification in Mass Disasters, 71-83. Washington: Smithsonian Institution Press UBELAKER, D. H., 1989, Human Skeletal Remains: Excavation, Analysis, Interpretation, 2nd edn. Washington: Taraxacum Press WILTSHIRE ARCHAEOLOGICAL MAGAZINE. 1986. Wiltshire Archaeological Register for 1984. WANHM 80, 240-244 WALKER, P L., 1995, ‘Biases in preservation and sexism in sexing: some lessons from historical collections for paleodemographers’, in S. R. Saunders and A. Herring (eds), Grave Reflections: Portraying the Past through Cemetery Studies, 31-47. Toronto: Canadian Scholars Press WELLS, C., 1996, ‘Human burials’, in V. I. Evison and P Hill, Two Anglo-Saxon Cemeteries at Beckford, Hereford and Worcester, 41-66. York: Council for British Archaeology Research Report 103 WELLS, C., DENSTON, B. and CHADWICK HAWKES, S., 2003, ‘The inhumations and cremations’, in S. Chadwick Hawkes, and G. Grainger, The Anglo-Saxon Cemetery at Worthy Park, Kingsworthy near Winchester, Hampshire, 153-188. Oxford: Oxford University School of Archaeology Monograph 59 WILLIAMS, H. M. R. 1997. Ancient landscapes and the dead: the reuse of Roman and prehistoric monuments as early Anglo-Saxon burial sites. Medieval Archaeology 41, 1-32 WILLIAMS, H. M. R., 2010, ‘Engendered bodies and objects of memorty in final phase graves’, in J. Buckberry and A. Cherryson (eds), Burial in later Anglo-Saxon England c. 650-1100 AD, 26-37 . Oxford: Oxbow Books WILLIAMS, P and NEWMAN, R., 2006, Market Lavington, Wiltshire: An Anglo-Saxon Cemetery and Settlement. Salisbury: Wessex Archaeological Report 19 WILSON, D., 1992, Anglo-Saxon Paganism. London: Routledge YORKE, B. A. E., 1995, Wessex in the Early Middle Ages. London: Leicester University Press Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 96-104 Why does the north-western boundary of Wiltshire ignore the river Avon? by Hannah Whittock The north-western boundary of Wiltshire ignores the river Avon and a range of evidence suggests that this characteristic is based on Late Iron Age political arrangements that were reinforced in the post-Roman and Anglo-Saxon periods. This is in striking contrast with the use of the Avon as a boundary to the west of the Limpley Stoke Valley. [Note: Anglo-Saxon charters are identified in the references by their * Sawyer number ’ (egSSl) and the published work in which these charters can be found. * Sawyer numbers’ act as a common reference system for Anglo-Saxon charters.] An anomaly requiring an explanation That river systems were used, in a number of significant instances, as boundaries in Anglo-Saxon England is clear from the available evidence. The river Humber, as Bede records in the 730s, clearly demarcated a long-lasting political boundary between the Northumbrians and their southern - Southumbrian - neighbours (Sherley-Price 1968, 56), since, as Bede recorded, it formed ‘the boundary between the north and south Angles’ (Sherley-Price 1968, 68). In a similar way, Bede described the East Saxons (Essex) as being ‘separated from Kent by the river Thames’ (Sherley-Price 1968, 104). The longevity of this particular riverine boundary meant that, even when Essex was overshadowed by Kentish imperium , London was still regarded as an East Saxon city and this continued to affect the spheres of influence of successor states when Mercian power eclipsed that of the East Saxons north of the Thames and West Saxon power eclipsed that of Kent south of the river. So clear was this boundary, that any extension of political power south of it was worthy of note and so Surrey was the ‘ Suprige ’, the ‘southern district’ (Ekwall 1960, 453), in the Anglo- Saxon Chronicle’s annal for 722. Similarly, analysis of patterns of artefact distribution in East Anglia has led to the interpretation that the patterns that can be seen in the sixth and seventh centuries were related to ancient cultural areas, as the distribution of objects north and east of the rivers Lark and Gipping correlates with artifact distribution associated with the pre-Roman Iceni (Williamson 2008, 140). The Anglo-Saxon Chronicle makes specific references to rivers marking the extent of emerging kingdoms. Its annal for 485 (the date itself is dubious) records how the South Saxon ‘ALlle fought against the Britons near the bank of Mearcredesbuma ( Whitelock 1979, 154). The name means something like ‘the stream of the agreed frontier’ (Myres 1986, 136). Another annal, for 658, refers to the West Saxon king,Cenwealh, putting the Britons ‘to flight as far as the [river] Parret’ (Whitelock 1979, 165) and that for 716, in manuscript C, which records the death of Osred, king of the Northumbrians is extended, in manuscripts D and E, with the words ‘south of the border’ (Whitelock 1979, 171); almost certainly a reference to the river Humber. This boundary character of rivers is seen throughout the period. The Alfred- Guthrum Treaty, of the 880s, famously Pembroke College, Cambridge CB2 1RF THE NORTH-WESTERN BOUNDARY OF WILTSHIRE 97 stated: ‘First about our boundaries [ landgemcera ]: up the Thames, and then up the Lea, and along the Lea to its source, then straight to Bedford, then up the Ouse to Watling Street’ (Lavelle 2010, 326). It is noteworthy that, of the six stretches of boundary [landgemcera] identified in this treaty, four are along rivers. In 927, Athelstan fixed the border with the West Welsh (Cornish) on the river Tamar and with the Welsh at the river Wye (Bradbury 2004, 4). In the 1080s the rivers Tees and Ribble still marked the northern extent of the local government taxation system characteristic of England to the south. And where, in the north-west, the shire structure itself ceased, the Domesday commissioners specifically I referred to the area in question as ‘ Inter Ripam et Mersham’ (‘Between the Ribble and the Mersey’); in which the rivers marked the boundaries of an area which, though unshired, still represented a distinct English region (Campbell 2000, 49). Clearly then, even this limited examination of the evidence suggests that it is not unreasonable to consider rivers as boundaries in the Anglo-Saxon period. Whilst rivers did not inevitably have this function - many major rivers did so. An examination of the Bristol Avon reveals that it too had a boundary function, although the situation in this case was complex and one that had a clearly anomalous character. One of the striking characteristics of the north- western county boundary of Wiltshire is that it ignores the course of the river Avon. On its own this might not seem worthy of note, since a river was not inevitably used as a political or cultural boundary, despite the examples cited above. However, this is an anomaly because this is historically not the case with regard to boundaries and the course of the river west of the Limpley Stoke Valley. This raises the obvious question why this distinct functional division in the use (and non-use) of the river as a boundary should have arisen. A recent exploration of the history of the Avon valley as a frontier from the fourth to the eleventh century AD reveals that this anomaly is of long-standing and arose from cultural and political boundaries first discernible in the Late Iron Age; these continued to influence boundaries throughout the Anglo-Saxon period and are fossilised in modern administrative boundaries in the twenty-first century. In short, there is an anomaly in the use of the river Avon as a boundary and this anomaly can be explained. The evidence of the ‘western Avon’ In order to fully understand the distinct character of the north-western Wiltshire county boundary and its relationship with the ‘eastern (Wiltshire) Avon’ it is necessary to briefly outline the contrasting character of the ‘western Avon’ as a frontier and boundary (Figure 1). West of the Limpley Stoke Valley, the river Avon has been argued to have marked an internal subdivision within Dobunnic territory in the Late Iron Age. Coins carrying the name Corio and Catti are found throughout Dobunnic territory but the distribution south of the Avon is in contrast to few finds in the area around Cirencester (which appears to separate areas of Corio/Catti distribution). In contrast, coins of Bodvoc are found almost entirely north-east of the river Avon (Webster 2003, 60). These latter issues show a noticeable concentration in the Cirencester area and to the north-east of this location (Leins 2008, Map 5a). There is also some similarity between the distribution of Bodvoc and the so-called ‘Western IJ’ types (Leins 2008, Map 5b). Questions have been raised over the significance of this pattern of finds (Todd 1999, 37), and alternative interpretations advanced which reject the connection made between coin distribution patterns and tribal territories (Leins 2011, pers. comm.). However, a number of other researchers have suggested that this distribution may indicate that the ‘western Avon’ was possibly an internal boundary differentiating a sub-unit of the Dobunni from the Dobunnic heartland (Cunliffe 2005, 189-190; Laycock 2008, 34). Clearly, this matter remains an area of contention, with the significance of the distribution open to question. However, reading back from later evidence may strengthen the interpretation that there was something distinctive about the area south of the river Avon in contrast to the situation north and north-east of the river. And, if so, this putative boundary-function appears to have influenced later developments. The correlation between the putative riverine border of this Dobunnic sub-unit and later boundaries (e.g. West Wansdyke) suggests that this was not simply a temporary Dobunnic extension of cultural influence southwards; and even if it did not precede the time of the minting of these coins, was to prove long lasting after this point. How this affected political arrangements in the Roman period is more difficult to discern since the evidence for civitas boundaries is slight and most 98 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 1 The Rive Avon and key features referred to in the text mappings of these boundaries are highly conjectural. In some reconstructions the Avon loses all boundary functions along its entire course, with the territory of the Dobunni (with its civitas capital at Corinium/ Cirencester) stretching as far south as the Mendips and south-eastward onto Salisbury Plain (Jones 2004, 163, Fig. 10.1). A variant suggestion is that, while northern Somerset may have been included within Dobunnic territory, that of the Belgae may have stretched as far west as Bath (Yorke 1995, 3). This would mean that the Avon was not regarded as a boundary by either the Dobunni or the Belgae. However, closer analysis may suggest that the matter was not so simple. If indeed Dobunnic territory stretched as far south as the Mendips this may simply represent administrative control over the sub-unit of Dobunnic territory suggested by the pre-Roman distribution of the coins of Corio/Catti. In which case the ‘western Avon’ may still have functioned as an internal - if temporarily suspended - boundary of a sub-group of the Dobunni and the later post- Roman evidence of West Wansdyke strongly suggests that this was indeed a reality (which later reasserted itself in the fifth century), even if it was temporarily subsumed within Roman arrangements which emphasised the integrity of Greater Dobunnic territory, including that which lay between the ‘western Avon’ and the Mendips. On the other hand, the possibility of Belgic control of Bath strongly suggests that the ‘eastern Avon’ wasstill straddled by political arrangements which ignored it as a boundary in both the Late Iron Age and the period of Roman rule (see below for an examination of the evidence for this in Ptolemy’s Geography ) . In the post- Roman period imported late-fifth to sixth-century Mediterranean pottery has been found in Somerset, Devon and Cornwall (Costen 201 1, 17- 19), but not in Gloucestershire (Campbell & Bowles 2009, 309, Fig. 20.5). This absence suggests a cultural boundary between the two areas (Rahtz 2003, 27); a characteristic reinforced by the West Wansdyke. THE NORTH-WESTERN BOUNDARY OF WILTSHIRE 99 Recent exploration tends to suggest that it was a fifth century construction (Erskine, 2007, 102-3), although it may have incorporated pre-historic features which point to an earlier boundary following the same approximate line. The matter, though, is subject to wide ranging debate and it has also been strongly argued that its building was contemporary with East Wansdyke and, as such, the scale and the military organisation necessary to implement it negates a fifth century date and suggests that these two named earthworks formed a much later unified frontier arrangement (Reynolds 2006, 140). Indeed, the fluctuating nature of West Saxon/Mercian relations leaves the possibility open that the whole Wansdyke frontier was an unfinished project (Reynolds & Langlands 2006, 35-6). However, whether it was a fifth century boundary between British polities or a Middle Saxon frontier between Wessex and Mercia its western length at least confirms the ‘western Avon’ as demarcating a frontier. The fifth-sixth century division of pottery types, outlined above, can also be seen with regard to elite habitation, which again seems to show a cultural division along the line of the ‘western Avon’. The post-Roman fifth-century elites to the south of the Avon seem to have based their centres of power on hillforts, whereas hillforts to the north - despite recent evidence for some level of re-occupation at Crickley Hill, south of Cheltenham (Dixon 2005, 394) - do not seem to have been re-occupied in anything like the same scale, or number (Rahtz 2003, 25). Corroborative evidence lies in the fact that, while eastern Mediterranean pottery has been found at ‘beach market’ site such as Bantham and Mothecombe, in Devon, and Tregurthy and Gwithian, in Cornwall (Costen 2011, 18), a large number of the West Country sites showing concentrations of imported Mediterranean ceramics are ‘hilltop fortified elite residences’ (Campbell & Bowles 2009, 301). The evidence of the ‘eastern (Wiltshire) Avon’ East of the Limpley Stoke Valley the ‘eastern (Wiltshire) Avon’ does not appear to have had a frontier role in the Late Iron Age and detailed study of coins and pottery, which may have served as tribal- identifiers, suggests that Dobunnic territory ran to a probable border with the Durotriges near the source of the river Wylye (Webster 2003, 61 ; Pap worth 2007 and pers. comm.) and with the Belgae in the vicinity of Salisbury Plain (Cunliffe 2005, 190; Haselgrove 2011, personal comment). However, under Roman rule, Belgic hegemony may have extended further to the west. A recent review of the evidence of Late Iron Age coin distribution has offered the more radical interpretation that the coins usually ascribed to the ‘Dobunni’ were probably actually issued by rulers of independent polities, who were not formed into a political unit (the Civitas Dobunnorum ) until the Roman period and then only as a direct result of Roman intervention (Leins 2011, pers. comm.). However, even if this is correct, the distribution of find-spots of the, so-called, ‘Western coinage’ (Leins 2008, Map 5b), which has commonly been classified by other researchers as ‘Dobunnic’, still suggests that a cultural frontier lay on the general line of the river Wylye and Salisbury Plain. South and south- east of this line the number of ‘Western coinage’ finds drops away dramatically. This is consistent with the earlier interpretations concerning the extent of general ‘Dobunnic’ influence, even if the concentration of so-called ‘Dobunnic coins’ north and north-west of Salisbury Plain (and beyond that into the Cotswolds and south-west Midlands) actually represents a number of inter-related tribal groups, rather than the extent of one distinct Late Iron Age polity. The relevance for this paper is that either of these interpretations still corroborates the idea that a Cotswolds-based culture (represented in some form through the ‘Western Coinage’) extended well to the south-east of the course of the ‘eastern Avon’. However, it seems likely that the extent of this ‘Dobunnic’ political/cultural influence was curtailed under Roman rule (even, perhaps, as the political unit of the Civitas Dobunnorum was being formed). The assertion, referred to earlier, that Bath in the Roman period was included in Belgic territory is found in Ptolemy’s Geography , written c.AD 150. He identified Bath by the nam e Aquae Calidae (‘hot springs’) and claimed it as a polis of the Belgae (Rivet & Smith 1979, 255-6). Ptolemy’s association of Bath with the Belgae - when all Late Iron Age coin evidence points to it being within ‘Dobunnic’territory - has led to suggestions that the Geography is in error at this point. However, two arguments that are very relevant to the role of the ‘eastern Avon’ may suggest that the identification should not be so easily dismissed. The first is a textual one from the Geography itself. Ptolemy states: ‘Below the Dobunni lie the Belgae and the cities: 100 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Ischalis, Aquae Calidae, Venta' (Rivet & Smith 1979, 144). Ischalis is almost certainly Charterhouse on Mendip and so the overall reference encompasses the southern Cotswolds, the Mendips and the Winchester region; within which western Wiltshire would also lie. This suggests a distinct political entity and the association of Ischalis with Aquae Calidae , suggests that the latter was not just a stray reference, out of place, and therefore erroneously linked to the Belgic centre further east. It looks as if there was a political arrangement that brought this region within Belgic control. More importantly, from the perspective of western Wiltshire, the reference strongly suggests that the lands straddling the ‘eastern Avon’ were linked by a common authority (which, for a time, looked towards f^wta/Winchester as the seat of its political power). Consequently, despite an initial appearance of confusion in Ptolemy’s account, this is actually entirely consistent with the evidence of Late Iron Age coinage which suggests ‘Dobunnic’ territory ran to Salisbury Plain. An explanation for Ptolemy’s assertion (which clarifies the situation)is that this south-eastern region of the territory of the Dobunni,or a group of related tribal polities (straddling the ‘eastern Avon’), had been granted to the Belgae in the Roman period for reasons now lost to us. What is important is that this evidence is consistent with the claim that the ‘eastern Avon’ lacked a boundary function in the Roman period as in the Late Iron Age. This is a particularly valuable clue, given the otherwise rarity of any evidence for internal political/administrative boundaries in the Roman period. However, in the Late Roman and immediately post- Roman period, this Belgic control of the area (as claimed by Ptolemy) may have been challenged by the Dobunni once Roman authority waned (Laycock 2009, 234). Given this lack of a frontier-role, the continuation of the Avon as an internal feature within the successor kingdoms of the sixth and seventh centuries and beyond should come as no surprise. As a result, the ‘eastern Avon’ and land beyond it to the west were more likely to be absorbed into the emerging kingdom of Wessex, rather than be used as a barrier between Wessex and the Hwicce/Mercia. The campaigns of West Saxon warbands north of Bath - at Dyrham, c.577, and as far as Cirencester, c. 628 - appear to confirm this. If there was any previous frontier in this area then it was probably one that faced north rather than east/west across the Avon. This is given some tentative support from the indication of possible pre-historic elements incorporated into the post-Roman, or Anglo-Saxon, construction of East Wansdyke (Eagles and Allen 2011, 147-155). This situation may have been given particular emphasis by the West Saxons’ emerging role as successors to Romano-British authority based on E£wta/Winchester. In which case, the drive beyond the ‘eastern Avon’ and into the Cotswolds may have been an attempt toreimpose the rule of Venta / Winchester on this region, which had once more reverted to Dobunnic control (and was consequently probably ruled by post-Roman British authorities based in Bath, or Cirencester). It is instructive that the West Saxon campaign of 577, that must have driven through western or north-western Wiltshire, killed rulers of Bath and Cirencester (and also Gloucester) who are designated as ‘ cyninges ’ (‘kings’) in the account of the battle of Dyrham found in the Anglo-Saxon Chronicle (Whitelock 1979, 157). In which case, north-western Wiltshire was almost certainly a border region of one of these British kingdoms and the implication of the Chronicle's account (regardless of arguments over its dating of this event) is that Wessex was seizing the territories dependent on these places (Yorke 1990,135). It has recently been suggested that north-western Wiltshire may have been part of a polity dependent on Old Sarum (Costen 2011, 15) but the context of the Chronicle's account of Dyrham suggests that British-controlled Bath, or Cirencester, are more likely candidates; especially in the light of the previous examination of Dobunnic influence in this region. And in this early phase of West Saxon expansion this whole region seemed set to fall to Wessex. We may perhaps extend Chris Wickham’s comment that ‘the upper Thames was West Saxon before it was Mercian’ (Wickham 2005, 345) to suggest that so was the southern Cotswolds. In the second half of the sixth and the first half of the seventh century the attention of the West Saxons (as recorded by the ninth century compilers of the Chronicle ) was very much on this strategic triangle of north Wiltshire, north Somerset and southern Gloucestershire (Yorke 1990, 136). As a result, we find battles recorded (with debateable dates) as having occurred at Searoburh (Old Sarum, Wiltshire) in 552, Barbury (Barbury Castle, Wiltshire) in 556, at Dyrham(Gloucestershire) in 577, at ‘Woden’s Barrow’ (Adam’s Grave near Alton Priors, Wiltshire) in 592, at Cirencester (Gloucestershire) in 628, at Bradford on Avon (Wiltshire) in 652, and at Peonnan (probably Penselwood, Somerset) in 658 (Whitelock 1979, 156-165). And, even when checked by the Mercians, at the battle of Cirencester, it is clear THE NORTH-WESTERN BOUNDARY OF WILTSHIRE 101 that West Saxon ambitions continued to focus on extending the frontiers of Wessex in this direction whenever they could (Yorke 1995, 57) and never settled for a frontier as far back as the river Avon. This contested nature of the region can be clearly demonstrated from evidence concerning the seventh-century monastery at Malmesbury. The first recorded abbot, Aldhelm, is reported by William of Malmesbury to have been a member of the West Saxon royal house and this would support the idea that the monastery was within West Saxon territory (Eagles 2001, 223). However, the contested nature of its hinterland is revealed in the charters which show us that Aldhelm was successful in securing patronage from both the Mercians and the West Saxons. The Mercian comes Cenfrith (SI 166, in Kelly 2005, no.2) and the kings ALthelred (S71, in Kelly 2005, no.3 and S73, in Kelly 2005, no.4) and Berhtwald (SI 169, in Kelly 2005, no.5) all made land grants to Malmesbury during the 680s. In the same period, Csedwalla of Wessex was also making grants to the abbey (S231 and S234, in Kelly 2005, nos.6 and 7), as were Baldred, a West Saxon sub-king of Somerset (SI 170, in Kelly 2005, no.8) and Leuthere, the bishop of the West Saxons (S 1245, in Kelly 2005, no.l). The contested Malmesbury area is revealed in the fact that the Mercian Berhtwald was, as nephew of the king of Mercia, clearly charged with holding a sensitive border area in the 680s.In the charters he is described as rex, or sub-regulus and this implies that he was a regional ruler, acting in the capacity of a sub -king for the overall ruler of Mercia. Clearly, this was part of a wider policy whereby ALthelred of Mercia secured disputed border areas (Kirby 2000, 98). In the same decade, Berhtwald’s counterpart on the West Saxon side was Baldred, who granted land in the same area (SI 170, in Kelly 2005, no. 8).Witnessing this charter as rex , he is generally considered to have also been a sub-king of a border area, as was Berhtwald. The charter evidence reveals that Malmesbury was very much part of a contested frontier-zone and one in which it had strong links with both Wessex and Mercia. This is vividly revealed in what was probably the last charter granted by ALthelbald of Mercia in 757. It concerned land at Tockenham (Wiltshire) and was granted to Malmesbury but the grant was witnessed by Cynewulf of Wessex. Clearly, both men took a keen interest in this zone of interaction between the two kingdoms in north-west Wiltshire. The competitive nature of this relationship is revealed by the fact that, on ALthelbald’s death later that same year, Cynewulf reasserted West Saxon control in the upper Thames valley (Williams 1999, 27). Clearly this expansion of territorial control would have also had implications for the north-west Wiltshire region where it bordered Mercia and it is significant that there is no evidence that Cynewulf ever acknowledged the overlordship of ALthelbald’s eventual successor, Offa. This, despite evidence for Offa’s active intervention at Bath, in 781, to compel Bishop Heathored of Worcester to relinquish the minster there to Offa and Mercian royal control and to give up land earlier bought from Cynewulf of Wessex south of the river Avon. This effectively reversed efforts by Cynewulf to gain influence at Bath through his grant of land to the minster there, in 758, following the death of ALthelbald (Williams 1999, 27-8). It is surely significant that no such assertion of Mercian power occurred at Malmesbury. Between the death of ALthelbald of Mercia in 757 and the end of an independent Mercian monarchy in 879, only one land grant to Malmesbury was made by a Mercian ruler (Offa’s short-lived successor, Ecgfrith, in 796). All other grants were made by kings of Wessex.This, despite the fact that, following Cynewulf’s death in 786, Offa actively intervened in other areas of Wessex, including northern Wiltshire, until his death in 796. Clearly, the Mercian hegemony, witnessed at Bath, was not repeated at Malmesbury. It therefore seems fair to assert that, despite suggestions to the contrary (Yorke 1995, 61), the monastery on the Avon at Malmesbury was not on the front line in the way that the minster at Bath clearly was. We may surmise from the later evidence that this was because the perceived front line lay further to the west. In contrast, the enigmatic evidence of East Wansdyke may suggest that this earthwork was an attempt to create a frontier demarcation running east-west during this period, when Wessex and Mercia contested control of northern Wiltshire from the seventh to the early ninth centuries (Reynolds 2006, 140). If so, its orientation suggests a north/south diplomatic face-off, in contrast to the orientation of the course of the Avon further west; although the apparently unfinished nature of the Wansdyke-frontier indicates that it quickly became redundant (Reynolds & Langlands 2006, 35). In time, the Thames frontier succeeded the line of Wansdyke, once firmer boundaries between the two competing kingdoms became established (Reynolds & Langlands 2006, 44). If this interpretation of the dating of this earthwork is correct, then its name - far from indicating a period of construction in the pagan Anglo-Saxon past - is more likely to have been 102 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE a fabrication of tenth-century Christian propaganda rather than evidence of pre-Christian construction (Reynolds & Langlands, 2006, 33) This view of East Wansdyke’s political role is in contrast to earlier interpretations which, while considering a range of date-options, viewed it as more likely representing a sixth-century boundary within an embryonic Wessex, and formed prior to the fusion of the Saxons of Wiltshire with those of the Thames valley (Fisher 1973, 35-6); or an even earlier frontier between a British authority in Wiltshire facing a northern threat posed by Saxons from the upper Thames valley (Myres 1986, 156). This, it has been suggested, constituted a ‘short-lived emergency’ which was soon superseded by other events (Fowler 2001, 196). However, even if this earlier dating is correct, it would still have represented a northern frontier for either a threatened British community, or alternatively one of the foci around which Wessex formed; and such a northern frontier, as has been noted earlier, was on a different orientation from the potential boundary that the Avon would have formed had it been chosen as a demarcation line. Clearly the main threat, as seen from the perspective of East Wansdyke, was perceived as emanating from the north rather than the north-west when the dyke was constructed. This is even more persuasive if East Wansdyke did not, in fact, form a continuous frontier with West Wansdyke. This is the interpretation favoured in this paper and is based on the very different scale of the two dykes and the lack of categorical evidence for continuation of the East Wansdyke west of Morgan’s Hill. This is compounded by the lack of Old English references to Wansdyke in this ‘gap’, despite the existence of Old English charter evidence for this area of north-west Wiltshire. This contrast with the occurrence of the name of the dyke in charter boundaries to the west and east of the ‘gap’ is surely significant. References to Wansdyke in this ‘gap’ on the second edition OS 25 inch map (1901) seems no more than speculation based on earthworks on the western side of Bathford Hill and are certainly no substitute for the silence from the Old English sources. Even a reference, in the fifteenth century copy of Survey A of the Cartulary of Shaftesbury Abbey, to ‘ Wadenesdich ’ near the otherwise unidentified settlement of Tortelee (Harvey 1998, 79) need be no more than antiquarian speculation, probably inspired by the visible line of the course of the Roman road (Margary 53) leading from Bath to Mildenhall, and which is intercepted, further east on Morgan’s Hill, by the earthwork of EastWansdyke. The nature of the record in the cartulary is so vague that it could easily refer to land, held north of Bradford-on- Avon, in either Wraxall or Atworth (both owned by Shaftesbury Abbey) whose northern boundaries followed this Roman road. This suggestion that the Roman road was mistakenly identified as Wansdyke in the ‘gap’ is corroborated by the Perambulation of Melksham Forest, c. 1 300 which described two sections of the road as ‘ Wodenes Ditch ’ (Grundy 1939, 576-9). Once again, this is no substitute for Old English documentary references. Significantly, where an Old English source does clearly refer to the line of the Roman road in a charter (ALthelred II’s grant of Bradford-on-Avon to Shaftesbury Abbey, in 1001) it is described as ‘pe kinges imare’ (‘the king’s boundary’) and ‘ alfgaresi mare ’ (‘Alfgar’s boundary’) and there is no hint of the name Wansdyke at any point in the charter’s bounds (S899, in Kelly 1996, 116). In short, it is probably best to conclude that the frontier represented by the two dykes was either unfinished or (in the view of this paper) always discontinuous, though sharing a common name. This would then be consistent with the view that Late Iron Age, Roman and post-Roman polities based in Wiltshire considered that their north-western spheres of influence extended into the southern Cotswolds, to the west of the ‘eastern Avon’, until blocked by the annexation of the Hwicce into Mercia during the late seventh century. And even then, this did not force a retreat to the line of the Avon for the frontiers of Wessex, Wiltshire and Ramsbury (see below). Even if it could be proved that the two dykes were in fact continuous this would not undermine these conclusions since the ‘Wansdyke frontier’ proved to be a transient border despite its impressive size and the effort expended on its construction. This was the case whether it was built in response to a fifth century ‘non-event’ (Fowler 2001) and was soon rendered redundant by changing strategic circumstances, or was a product of the later and better attested rivalry between Wessex and Mercia. Other borders would replace it and, in north-west Wiltshire, these were much more persistent over time. The later kingdom, shire and diocesan boundaries which put Wessex, Wiltshire and Ramsbury well to the west of the ‘eastern Avon’ were more closely orientated on the general line of the Fosseway. Indeed, if there is any feature that seems to have influenced the course of the county’s north-western border it is the general line of the Roman road. Even that though had only approximate influence, THE NORTH-WESTERN BOUNDARY OF WILTSHIRE 103 since mapping of the land holdings included within Wiltshire at Domesday Book (1086) reveals thatThorngrove Hundred straddled the road, Dunlow Hundred lay west of the road, as did the portion of Chedglow Hundred between Ashley and Long Newnton (both now in Gloucestershire); though in this hundred there is a correlation between the boundary and the road west of Brokenborough and again west of Kemble (Thorn 1979). However, at least one could suggest that the Roman road was an influencing factor, although clearly not a defining one. In contrast to this, the earlier Viking Wars, of the late ninth/early tenth century, briefly saw the river Avon acting as a frontier along its entire length (linked to the Thames), with the defended burghal system running: (possibly Bristol) Bath (now absorbed into Wessex), Malmesbury, Cricklade, Oxford, Wallingford, Sashes, and Southwark. However, this frontier along the whole course of the Avon was not permanent. During the tenth century the old frontier to the west of Bath formed the shire boundary between Somerset and Gloucestershire and the diocesan boundary between Wells (after it was carved out of the diocese of Sherborne) and Worcester; whereas to the east of Bath the shire and diocesan boundaries (that of Ramsbury, 909- 1058 and Worcester) ignored the Avon as earlier boundaries had done since the Late Iron Age. The close correlation between the diocesan boundary of Worcester, as mapped c.1291, and the extent of land granted within the kingdom of the Hwicce to 821 (Bassett 1989, 9, Fig. 1.2) clearly indicates that this boundary (and that of north-western Wiltshire) dated from long before the ninth century. Conclusion The north-western boundary of Wiltshire is the product of a long tradition whereby political units claimed land either side of the ‘eastern Avon’. Crossed by the lands of the Dobunni, by a temporary Belgic expansion and by competing post- Roman kingdoms, it is therefore not surprising that the Anglo-Saxon shire and diocesan boundaries eventually lay well to the west of the river. What is remarkable is the fact that this relationship (or rather non-relationship) between the river and frontiers was clearly over a millennium old by the end of the Anglo-Saxon period and this characteristic continued to influence administrative arrangements and is consequently fossilised in both secular and ecclesiastical boundaries in the twenty-first century. This paper draws on some of the issues explored and conclusions reached in a dissertation in the department of Anglo-Saxon, Norse and Celtic (ASNaC), University of Cambridge, 2010. Acknowledgements A number of people provided valuable advice, which assisted in the preparation of this paper; without them, of course, carrying any responsibility for the interpretations offered by the paper. Martin Pap worth, National Trust Regional Archaeologist (Wessex Region), shared his ideas regarding Dobunnic and Durotrigan culture zones. Especially helpful was: Pap worth, M., 2007, Deconstructing the Durotriges , Reading University: unpublished PhD thesis. This was later published as: PAPWORTH, M., 2008, Deconstructing the Durotriges. A Definition of Iron Age Communities within the Dorset Environs British Archaeological Reports, British Series 462, Oxford: Archaeopress. Professor Colin Haselgrove, Professor of Archaeology, University of Leicester, assisted with comments regarding the distribution of Late Iron Age coinage of the Dobunni. Dr Ian Leins, Curator of Iron Age and Roman Coins, Department of Coins and Medals, at the British Museum, provided detailed information regarding his interpretation of the significance of the distribution of Late Iron Age coins.The advice offered by Professor Andrew Reynolds, Professor of Medieval Archaeology, University College London and by Dr Bruce Eagles, Visiting Research Fellow, Bournemouth University, also greatly assisted in the writing of this paper and in extending the scope of a number of the issues covered. Peter Bull, of Peter Bull Art Studio, produced the map with a speed and flexibility that was much appreciated. MartynWhittock assisted in the abridging of material from a much longer original dissertation and in preparing it for WANHM. References BASSETT, S., 1989, ‘In search of the origins of Anglo- Saxon kingdoms’, in S. Bassett (ed), The Origins of Anglo-Saxon Kingdoms , 3-27. Leicester: Leicester University Press BRADBURY, J., 2004, TheRoutledge Companion to Medieval Warfare. London: Routledge 104 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE CAMPBELL, E. and BOWLES, C., 2009, ‘Byzantine trade to the edge of the world: Mediterranean pottery imports to Atlantic Britain in the 6th century’, in M. Mundell Mango (ed), Byzantine Trade, 4th- 12th Centuries: the Archaeology of Local, Regional and International Exchange, 297-314. Farnham: Ashgate Publishing CAMPBELL, J., 2000, The Anglo-Saxon State. London: Hambledon COSTEN, M., 201 1, Anglo-Saxon Somerset. Oxford: Oxbow Books CUNLIFFE, B.,2005, Iron Age Communities in Britain: An Account of England, Scotland and Wales from the Seventh Century BC until the Roman Conquest. London: Routledge DIXON, P, 2005, Thirty-five Years at Crickley. Current Archaeology, 200, 390-395 EAGLES, B. N., 2001, ‘Anglo-Saxon Presence and Culture in Wiltshire c.450-675’, in P Ellis (ed), Roman Wiltshire and After: Papers in Honour of Ken Annable, 199-233. Devizes: Wiltshire Archaeological and Natural History Society EAGLES, B.N. and ALLEN, M., 2011, ‘A reconsideration of East Wansdyke: its construction and date - a preliminary note’, in S. Brookes, S. Harrington and A. Reynolds (eds), Studies in Early Anglo-Saxon Art and Archaeology: Papers in Honour of Martin G. Welch, British Archaeological Reports, British Series 527, 147-55. Oxford: Archaeopress EKWALL, E., 1960, The Concise Oxford Dictionary of English Place-Names. Oxford: Oxford University Press ERSKINE, J., 2007, The West Wansdyke: an appraisal of the dating, dimensions and construction techniques in the light of excavated evidence. The Archaeological Journal, 164, 80-108 FISHER, D.J.V., 1973, The Anglo-Saxon Age, c.400-1042. Harlow: Longman FOWLER, P, 2001, ‘Wansdyke in the Woods: An unfinished Roman military earthwork for a non-event’, in R Ellis (ed), Roman Wiltshire and After: Papers in Honour of Ken Annable, 179-198. Devizes:Wiltshire Archaeological and Natural History Society GRUNDY, G.B., 1939, The Ancient Woodland of Wiltshire. WANHM, XXXLVIII, 530-598 HARVEY, R.B., 1998, Shaftesbury Abbey’s 12th century Rentals for Bradford on Avon. WANHM, 91, 76-89 JONES, M., 2004, ‘Cities and Urban Life’, in M. Todd (ed), A Companion to Roman Britain, 162-192. Oxford: Blackwell KELLY S.E., (ed), 1996, Charters of Shaftesbury Abbey, Anglo-Saxon Charters V. Oxford:Oxford University Press/British Academy KELLY, S.E., (ed), 2005, Charters of Malmesbury Abbey, Anglo-Saxon Charters 11. Oxford:Oxford University Press/British Academy KIRBY, D.E, 2000, The Earliest English Kings. London: Routledge LAVELLE, R., 2010 .Alfred’s Wars: Sources and Interpretations of Anglo-Saxon Warfare in the Viking Age. Woodbridge: Boydell Press LAYCOCK, S., 2008, Britannia - the Failed State: Tribal Conflict and the End of Roman Britain. Stroud: The History Press LAYCOCK, S., 2009, Buckles, Belts and Borders: How Soldiers’ Fashions Reflect Political Turmoil in Late Roman Britain. Current Archaeology 20, no.6, 12-19 LEINS, I., 2008, ‘What can be inferred from the regional stylistic diversity of Iron Age coinage?’, in D. Garrow, C. Gosden, J. D. Hill (eds), Rethinking Celtic Art. Oxford: Oxbow Books MYRES, J.N.L., 1986, The English Settlements. Oxford: Oxford University Press RAHTZ, R, 2003, ‘The Dobunnic Area in Post Roman times’, in M. Ecclestone, N. Holbrook, A. Smith (eds), The Land of the Dobunni: a Series of Papers Relating to the Transformation of the Pagan, Pre-Roman, Tribal Lands into Christian, Anglo-Saxon Gloucestershire and Somerset, 24-31. King’s Lynn: Heritage Marketing & Publications Ltd REYNOLDS, A., 2006, ‘The Early Medieval Period’, in N. Holbrook, and J. Jurica (eds), Twenty five years of Archaeology in Gloucestershire: A Review of New Discoveries and New Thinking in Gloucestershire, South Gloucestershire and Bristol 1979-2004, Bristol and Gloucestershire Archaeology Report No. 3, 133-160. Kemble: Cotswold Archaeology. REYNOLDS A. and LANGLANDS, A., 2006, ‘Social Identities on the Macro Scale: a Maximum View of Wansdyke’, in W. Davies, G. Halsall and A. Reynolds (eds), People and Space in the Middle Ages 300-1300, Studies in the Early Middle Ages 28, 13-44. Turnhout: Brepols RIVET, A.L.F. and SMITH, C., 1979, The Place-names of Roman Britain. London: Batsford SHERLEY-PRICE, L., (trans), 1968, Bede: A History of the English Church and People. Harmondsworth: Penguin. THORN, C. and F., 1979, Domesday Book: 6, Wiltshire. Chichester: Phillimore TODD, M., 1999, Roman Britain. Oxford: Blackwell WEBSTER, G., 2003, The Roman Invasion of Britain. London: Routledge WHITELOCK, D. (ed), 1979 , English Historical Documents, Volume I, c.500-1042. London: Eyre Methuen WICKHAM, C., 2005, Framing the Early Middle Ages. Oxford: Oxford University Press WILLIAMS, A., 1999, Kingship and Government in Pre- Conquest England, c.500-1066. Basingstoke: Macmillan Press WILLIAMSON, T., 2008, Sutton Hoo and its Landscape: The Context of Monuments. Macclesfield: Windgather Press YORKE, B., 1990, Kings and Kingdoms of Early Anglo-Saxon England. London: B.A Seaby Ltd YORKE, B., 1995, Wessex in the Early Middle Ages. Leicester: Leicester University Press Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 105-116 A Perambulation of the Hundred and Parish of Westbury by Alison J. Maddock The Perambulation was published by Colt Hoare in 1830 and given the date 1575. There is no indication of its provenance, and while it represents a valuable source of information on boundary features, place-names, land ownership etc, it contains many anomalies and misleading additions. Internal evidence suggests it actually dates from some forty years later. This article reports on research to determine the nature of the apparently lost source document and to dissect the text to reveal ‘original’ information, which has then been used for fieldwork around the boundary. Three extracts are given to demonstrate both the pitfalls of the published text and the sort of information revealed about the hundred and its boundary ( coterminous with the parish) in the early 17th century. Introduction The volume of Sir Richard Colt Hoare’s History of Modern Wiltshire covering Westbury Hundred contains a highly problematic piece entitled ‘A Perambulation of the Hundred and Parish of Westbury, temp. Eliz. 1575’ (Hoare 1830, 54-57). Why problematic? Because despite being widely quoted in 20th-century works of historical reference its source is apparently unknown, its anomalies seem never to have been questioned, its date is readily contradicted by internal evidence and, worse still, disentangling the ‘original’ text from the interpolations of its transcriber with certainty is all but impossible in the absence of the primary document. Moreover, the version of the Perambulation still extant in the handwriting of this transcriber (Richard Harris of Dilton Marsh) indicates that several questionable changes were introduced when it was typeset for publication. The present study attempts to determine whether the original survives in any archive (unsuccessfully so far), and to elucidate the document’s transition to printed form, drawing attention to the pitfalls of taking Colt Hoare’s offering entirely at face value. Evidence for a date of around 1615 is adduced. The research project also encompassed fieldwork along the whole route of the Perambulation, where it follows (and once or twice tellingly diverges from) modern boundaries, unravelling some highly obscure passages, seeking out surviving boundary features, and tracing its correspondence with Anglo- Saxon charter boundaries. Written up, the project’s full results are too lengthy for a paper in this journal, but copies have been deposited at the Wiltshire and Swindon History Centre and WANHS Library for reference purposes. This abbreviated report covers the analysis of the text itself, with some passages selected as case studies to illustrate the process of interpretation through desk research and fieldwork. It should be emphasised at the outset that no suggestion is being made that the Perambulation is other than genuine in its origins. Its peculiarities of wording are many, but as such support rather than undermine authenticity. ‘The Pigeon-hole’, Emms Lane, Bratton, Westbury, Wiltshire BA13 4SA 106 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE [Hawlridi. {evtvood Souse <* Q Redlands- Siorridat XgMtr: i Bre/mii Hilton Sheet Fig.l Map of the hundred from Colt Hoare’s Modern Wiltshire with (inset) outline location on county map The Hundred in context The ancient parish of Westbury, encompassing also Bratton, Dilton, Heywood and part of Chapmanslade, was coterminous with the hundred of that name (Figure 1). Already a separate hundred by the time of Domesday, there are suggestions that Westbury could have been split off at some earlier date from Whorwellsdown Hundred (Anderson 1939, 151). This hundred’s territory curls round Westbury on three sides from north-west to south-east and its conjectural meeting place is markedly peripheral, but would have been central in a united hundred.1 However, charters from the second half of the 10th century show that the intervening boundary from North Bradley to Edington was already significant by then, arguing that any such separation would have had to occur early in the development of Anglo- Saxon administrative hundreds from royal estates (Birch 1893, 364-5 and 495-6). In analysing the text of the Perambulation, certain features require reference to the above- mentioned charters, as will be seen in Example 3 below. They are King Edgar’s descriptive charter of Steeple Ashton and North Bradley AD 964, surviving only as a copy made temp. Edward I (Birch no. 1127; S 727), and his grant of Edington to Romsey Abbey in AD 968 (Birch no. 1215; S 765). These alone coincide with part of the Westbury boundary, leaving much of it without any such early evidence. Two other sources coincident with the parish/hundred boundary are available for comparison with the Perambulation’s wording: a partial perambulation of Selwood Forest along the Somerset/Wiltshire border c. 1632-4, which arose from a forest rights dispute concerning Sir Thomas Thynne (Longleat 10061), and one of Corsley parish from 1636 which has survived by being copied into the Church Book in 1754 (Longleat Corsley 547). The former agrees with the Perambulation; for the latter see Example 1. There are, regrettably, no pre- turnpike maps or estate plans covering the parish and hundred in any detail. A PERAMBULATION OF THE HUNDRED AND PARISH OF WESTBURY 107 Seeking the original text Dating evidence emerges best when the original document’s progress through several hands has been explored and the accretions of the transcribing process are removed, so the question of actual date is left until last. First come the thorny questions of the nature of the original document, where Colt Hoare found it, and why it has apparently vanished. Colt Hoare relied on Dilton Marsh clothier Richard Harris to supply him with much of the material for his section on Westbury Hundred in Modern Wiltshire. Letters to him from Harris and from Henry Wansey, who was responsible for the Warminster Hundred section, suggest that Colt Hoare may have hoped Wansey would cover Westbury too. However, Wansey preferred to enlist Harris, on grounds that the local man had already compiled much material, including some from the public records then held in the Tower of London. Harris agreed to make his information available and assist in any way he could, and although he seems to have seen himself as something of a junior partner to Wansey, as indicated in two surviving letters to Stourhead (WSA 383/907, 6 October and 10 December 1818), his name duly appears on the title page of the Westbury section. Wansey had written to Stourhead on 21 September reporting that: ... in pursuance of our plan I have been over to Mr Harris of Diltons Marsh ... I showed him the perambulation of Westbury hundred the greatest part of which he can make out & when the weather will suit we are with the heads of the neighbouring hundreds to make the perambulation. (WSA 383/907 and Figure 2) By ‘making out’ he probably refers to the difficulties of the old script and the item’s state of preservation, rather than implying it had to be translated from Latin. Archaic forms such as ‘springeth’ and ‘wast’ (for ‘waste’) would be unlikely to occur if this was a translation by Harris. The implication is that this was a loose or at least portable document that was in Wansey’s hands first but may have been passed to him by Colt Hoare. Clearly it was available locally and not a copy of something then residing in some official archive. Nor is there any sign that it has since found its way into any national repository: among those trawled have been the National Archives, British Library and Society of Antiquaries of London. Since it was more likely to have come from ■/£A+^,,,zSci 2/ §> > s^C < ZJ© — *%, j&9-c -A ‘O'sSy- Z_. /<.. — ^ Fig. 2 Facsimile of letter in WSA collections from Henry Wansey to Colt Hoare mentioning the Perambulation the parish chest, the manorial or estate papers of a local landowner, the diocesan archives or the papers of an antiquarian collector, the obvious locations of Wiltshire and Swindon Archives (Chippenham), the WANHS Library, Stourhead and Longleat are among those tried and ruled out along with other county record offices. Nothing has been found in the Stourhead papers at Chippenham or Devizes to explain how Colt Hoare or Wansey first came by the document. The notebook with sketches kept by Henry Wansey when compiling the Warminster volume narrowly escaped destruction and is held by the WANHS Library (DZSWS MSS 3); it does not mention the Perambulation document, though it contains an obscure index entry citing a ‘Westbury book’ that could be interpreted as referring to another notebook, now lost. A remote possibility could be that the missing document was contained in a book of collected Wiltshire manuscripts lent to Colt Hoare by Mrs Crewe nee Hungerford around that time (referred to in an unsigned cover note dated 1818 and specified as not relating to the Hungerfords, WSA 383/907). Neither the Crewe estate, whose library has long since been dispersed, nor the Stourhead library can identify this; if Colt Hoare held on to it then it might have been sold at the time of the Heirlooms Sale in 1883. 108 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE An alternative scenario is that the document was housed with the Westbury parish records. This would be the most usual place for a parish perambulation; the fact that it was also effectively a hundred perambulation might have been coincidental. Many of Westbury’s parish records do survive from the relevant period, so an explanation for loss from them is required. This could be the fire at the offices of local solicitor Pinniger, Finch in January 1935, when many important local government papers were destroyed. Some records of parish government inherited by the Urban District Council in the late 19th century could have been among these. However, the newspaper report of the fire states that the ‘old parish records’ were being stored elsewhere at the time and were saved. In 1889 Canon J. E. Jackson mentioned the existence of ‘an old Perambulation deed, taken three hundred years ago, which describes the parish boundaries with a very curious minuteness’, but we cannot deduce with any certainty whether he was referring to Colt Hoare’s printed version or the original (Jackson 1891, 35). A number of the minor place names from the Perambulation are quoted in The Place-Names of Wiltshire (Gover et al. 1939), most with the source given as ‘1575 Hoare’ but some with the source ‘1575 For ’ which according to that volume should translate as ‘Forest Proceedings (Chancery, Exchequer, King’s Remembrancer)’. Investigation of this apparent clue to the document’s origin has led nowhere and the conclusion must be that the For notation is an error. Harris’s notes, including the transcription of the Perambulation in his own hand, have survived and found their way into the Wiltshire and Swindon Archives (WSA 540/320). Viewing his transcript helps us understand many of the oddities of the printed version (Figure 3). It seems reasonable to assume that Harris’s handwritten text is nearer to the original than the published wording, but it may not be the exact text from which the printer worked. There would be editorial decisions imposed by Colt Hoare (e.g. how to deal with Harris’s marginal notes), and these might have included actual corrections agreed between them, which would account for some of the variations between the two versions. Unless the printer made his own editorial decisions as he set the type, there must have been a second fair copy marked up with instructions, not the one that survives in the WSA, so it is possible that discrepancies might have been inadvertently introduced during this copying. However, some of the changes that appear in the printed text are so obviously errors that the argument tends to favour relying on the handwritten version as being nearest to the ‘truth’ of the original. For example, Colt Hoare’s (or shall we say the printer’s) ‘Dewers Turnings’, describing dog-legs as the boundary follows old field systems on the downs and suggesting a proper name, appear simply to be ‘Deivers’ (i.e. divers) in Harris. The printer’s ‘the mill close to the river’ is ‘Mill Close on the River’ in Harris, making sense in the context of Bratton. Widow Whitaker has become William Whitaker in print, numerous other minor words are at variance, and the transformation of Sedland to Sealland Cross appears to have set a substantial red herring loose in the antiquarian pond.2 However, there are one or two places where one would tend to favour the printed version as making better sense, such as where Harris’s ‘A Round between . . .’ is printed as ‘a bound between. . .’ Changes to capitalisation are frequent in the printed version, as are changes in punctuation - not always to the benefit of clarity. There is another significant analytical complication consequent on not having the original to examine. Harris added explanatory notes and a few thumbnail sketches of locations in his margin, most of which Colt Hoare presents as footnotes (excluding the sketches), but in addition to these the text contains much matter in parentheses. In many cases this is clearly anachronistic, such as references to the turnpike or to some feature being ‘the same to this day’ or ‘now called . . .’. Clearly these must be presumed to be Harris contributions, leading to the suspicion that some or all of the other parenthetical interpolations are his too. Many appear to be alternative readings either where a word was difficult to make out or with the intention of helpfully supplying a substitute where the word is unusual. Sometimes Colt Hoare has simply left them out, as when Harris’s ‘De’Anseys Knap (or Napp)’ is printed as ‘d’Anseys Knap’, or has opted for what appears to be Harris’s own alternative, as when ‘Some time (or in ancient times)’ is printed as ‘in ancient times’, or has left the alternative in but altered the spelling, as when ‘the Brink (or Brough) of the Hill’ has become ‘the brink (or brow) of the hill’. Also puzzling are the instances - perhaps including the last of these examples - where the alternative in parenthesis looks more archaic than the word first given, suggesting that it should be taken as the original word, e.g. ‘highway (or hayway)’. As an additional hazard, Harris’s own handwriting is sometimes uncertain. One or two obscurities in the transcription are A PERAMBULATION OF THE HUNDRED AND PARISH OF WESTBURY 109 omitted by Colt Hoare, but others remain. In one case a particularly tortuous passage can make sense geographically only by assuming that a transcription error has been made, together with the wrong placement of an end bracket which should have enclosed a lengthy interpolation by Harris (see Example 2). Any analysis of the Perambulation text is characterised by such uncertainties. Dating the text Having explained how identification of the original wording has been attempted, the question of date can be addressed. Starting with the obvious clue lying in a reference to the King’s waste (not the Queen’s), and examining the personnel mentioned, it is easily shown that the text as it stands cannot date from 1575. Among the people named (see below), Ley, Phipps, Lamb and Bromwich all suggest a date of around 1610-1616 or even later. Did Harris misread 1575 as 1615? Possible, but not particularly plausible: surely Wansey or Colt Hoare would have spotted this. More acceptable as a hypothesis might be the updating of an Elizabethan parochial perambulation in the second decade of the 17th century, but retaining its original heading.3 Harris does not in fact write ‘temp. Eliz.’ but ‘in the Reign of Queen Elizabeth’: these might be his own words, or they might be copied from the original. He gives no indication of source. To add a further level in its provenance, the document he used could in theory have been a copy of the 17th century version written out at a still later date. There are slight indications that one or two supposed transcription errors may have been made by an earlier writer than Harris. Among the dating evidence are these references: • Waste/woods belonging to ‘Sir James Ley/Leigh Knt’ (Harris gives both spellings). Ley acquired most of his property in Westbury from the late 1590s onwards, completing his clean sweep of the Westbury manors around 1621. He was knighted in 1603, created baronet in 1619, baron in 1624 and Earl of Marlborough in 1626, dying in 1629. From 1624 on, therefore, his title would be Lord Ley and, if the writer was well- informed, the ‘Knt’ would be incorrect after 1619. The exact date of his acquisition of the waste and woods in question, the latter possibly from Henry and Nicholas Phipps (Crittall 1965, 155), unfortunately cannot be pinpointed. • Land of ‘William Bromwick’ in Bratton. He succeeded his father Sefton Bromwich as a minor in 1607/8 (Crittall 1965, 161). • Woods of ‘Simon Slopers of Warminster’, near the Somerset border. Sloper purchased the woods in question from Lord Audley in 1606 (Hoare 1830, 39). • Land of Nicholas Phipps, near Heywood. This is most probably the Nicholas who with his brother Henry was a manorial lord in the Westbury area from 1585 and died in 1615, though it could be his younger son of that name to whom he bequeathed some woodland in Heywood (TNA C 142/355/33 and PCC will at PROB 11/127, 1616). • The J ohn Lamb mentioned, another landowner on the Somerset border, is probably the one who died in 1615 (TNA C 142/355/93); his son was also called John but sold most or all of his Westbury holdings in 1615/16 (Crittall 1965, 157). • Widow Whitaker of Bratton Lower Mill succeeded to the tenancy after the death of her husband Nashe in 1610 (Morrison 1993). Roger Newborough (of Berkley in Somerset) and William Whitaker (of Fairwood), two other landowners to feature in the Perambulation, also shared their first names with other relatives, offering less secure dating evidence in consequence. From analysis of the landholders and James Ley’s titles we see that the date of the Perambulation as it stands is unlikely to be earlier than 1610 and is probably not later than 1619, or 1624 at the latest, with indications that the end date of this range might be brought back to 1615. Westbury Hundred had become split among many owners until Ley gradually purchased most of the manors, and he obtained the grant of the hundred by royal patent in 1610. This event might well have been the spur for a hundredal (as opposed to a parish) perambulation, though no such document is associated with the grant held at the National Archives (TNA C 66/1847/14). Ley’s papers relating to his offices of state have survived, and his manuscript collection later found its way to the British Museum (now in the British Library), but, other than a few sets of accounts for the hundred and a narrative of his family history (WSA 967/14 and 366/1), no other relevant Ley papers from the era in question have been located. From the parish viewpoint, 1610-1615 is also an interesting date because it coincides with Westbury’s 110 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE glebe terrier of 1614, a date that falls into a sparse period for such records in Wiltshire (Hobbs 2003, 455). Although no specific instruction from church authorities to perambulate at that date has been found in diocesan records (nor, for that matter, from the justices in Quarter Sessions), it is possible that the one required by officialdom might have sparked off the idea that it was time for the other to be updated and set down in writing for local use. Against this is the absence of church-related instructions within the text, such as are found in the Corsley perambulation of 1636. In that parish, regular stops for gospel readings were laid down. Their absence in the Westbury Perambulation could argue against a parish origin and for a hundred-based one, but might simply reflect prevailing doctrine at the earlier date, and with so much uncertainty surrounding the text it is also possible that Harris removed any such ecclesiastical diversions. One further point is that there is no mention of a detached enclave of Warminster Hundred at Hisomley near Dilton, within Westbury Hundred, a fact that might be expected to appear in a hundredal but not a parish description. Though its origins remain obscure, it is perhaps more likely that the impetus to perambulate was ecclesiastical rather than secular. The Perambulation As with most, though not all, perambulations, the direction of travel is clockwise. The start and end point is the round barrow called Rowborough south of Bratton village at NGR ST 928499 ( ruyeburwe in the charter of AD 968), now within the Salisbury Plain Training Area (SPTA). The boundary here is with Edington parish, and the route trends south and west across the Plain, bounding next with a projection of Bishopstrow parish (called ‘the Tongue’ in the latter’s enclosure award, though not in the Perambulation) and then successively with the parishes of Warminster and Upton Scudamore, where it leaves the plain for the lusher country around Dilton. At Chapmanslade the adjoining parish is Corsley as far as the Somerset border. The county line is followed northwards until reaching the watercourse that forms the boundary with North Bradley parish. Watercourses are followed north- east to Horse Bridge near Dursley (ST 861541), from where the route crosses to the woods north of Heywood House and runs up to the Edington- West Fig. 3 Archivist Steve Hobbs compares Harris’s text with Colt Hoare’s volume at the Wiltshire & Swindon History Centre Ashton road at ST 884549. The return to Bratton is made via a point south-east of Dunge called lechemere in the charters of AD 964 and 968, where the ancient parishes of Westbury, Edington and Steeple Ashton met, and then via the brook called Stradbrook or Milbourne, which is followed upstream to the springs from which it originates. Climbing the hill behind these, the boundary runs back south to its starting point. Most of the equivalent present-day boundaries give every appearance of coinciding closely with the Perambulation, even when it is difficult to follow due to minuteness of detail, though some minor rulings made under local enclosure acts etc may have fixed slight variations in the line.4 There is one significant deviation which does not seem to have been previously noticed, and another that is relatively recent and easily followed on 19th- and early 20th-century maps (see Example 1 ). In addition, the instructions around the area of the Bishopstrow ‘Tongue’, as they describe the crossing and re-crossing of trackways which have since been altered by army activity, are too complicated to make a certain match. (The Tongue might represent a subsequent arrangement allocating to Bishopstrow some downland that once fell into Westbury Hundred). There are minor discrepancies where short stretches of stream have subsequently been straightened, the boundary following the line of the old course. At the site of Bratton Upper Mill, where there is a classic example of a boundary still following the course of a vanished section of stream rather than the diverted watercourse leading to the mill, and near Brook Hall (ST 848535), where the instruction is ‘leaving the new runing river’, the A PERAMBULATION OF THE HUNDRED AND PARISH OF WESTBURY 111 Fig. 4 Stagnant former watercourse marking the boundary near Brook Hall Farm Perambulation boundary was already following an old line, changes to the watercourses having taken place at an earlier date (Figure 4). The full research paper explores these more fully, and also brings to light a striking number of instances of old (often lost) trackways following, or being followed by, the boundary. There are also stretches with banks and ditches, and one sharp change of direction within the SPTA has recently proved to respect a prehistoric linear feature. Any perambulation of the Westbury sort will normally expect to encounter markers such as boundary stones and pollard trees, especially ‘mere oaks’. Specific features mentioned in the Westbury Perambulation are pits and ‘balls’ or mounds of earth. These seem to have been too small to have survived to the present day, but some of the other boundary markers can be identified and matched to the description, though most are now lost. A few minor place-names with a boundary connotation also occur along the way, including ones with Row and Man(n) elements, although none of the more specific ‘mere’ (OE (ge)maere) type have been found. When reading the Perambulation in Modern Wiltshire or Harris’s transcript, it should be remembered that many of the instructions may progress only a very short distance from one marker to another while others occasionally represent a progression of well over a kilometre. Changes of stated compass direction relating to the former type can be misleading as a result. Indeed, points of the compass in the Perambulation are usually very approximately employed. There follow three extracts as case studies for the content and interpretation of the Perambulation. The text is quoted as it occurs in the handwritten transcript by Richard Harris, including all parenthetical interpolations but not his marginal notes. For brevity there has been some omission of matter where the recitation of minor route markers is irrelevant to the case. Example 1 This exemplifies two or three typical problems encountered in Harris’s transcription and also brings to light an apparently forgotten territorial anomaly between Westbury and Corsley parishes. ... following Chalcott Lane to (Chippenslad) Chapmanslade. (Passing by or) Turning down 112 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 5 Ditched embankment hidden in woodland at Long Hedge near Dilton Court, called ‘Chalcot Ditch ’ in the Perambulation to Biss Ford Lane (by the Side of the Road to Chapmanslade from Chalcott Hollow) and South Croft Stile (now Called Turners Rack close) then Along the lane to South Croft Ditch and Along that Ditch to a Stile (near Chapmanslade) . . . and from that Stile to the Footpath there and the Ground of John Watts’s ... unto Huntley Lane, and so up part of the Said Lane, Athwart the Lane to Rymers Ground and House and Orchard . . . and so up unto the Street of Chapmanslade. All the parenthetical matter can be discarded as being Harris commentary, except that the Chippenslad/Chapmanslade alternative looks rather like one of the reversed type mentioned above, i.e., one would suppose the former version to be the older. The extract begins at ST 847488, where the boundary reaches the A3098 (‘Chalcot Lane’, the Westbury-Frome road), having run along an interesting bank-and-ditch feature near Dilton Court (Figure 5). The lane was altered when it was turnpiked in the 18th century, but according to Harris the new road followed much the same route. At first reading the Perambulation simply follows the road to Chapmanslade, but in fact the boundary used to deviate from the present road line some 600m further along, as shown on the Westbury tithe map and subsequent Ordnance Survey maps. Since the above extract gives us no instructions to leave the lane before reaching Chapmanslade, unless some text is missing here, the conclusion must be that in the 17th century the lane did not follow the same course as the subsequent turnpike. On the 19th- century maps the boundary turns briefly north-west up the lane towards Short Street, then turns west to curve round by a still-existing fence line, cutting off a wedge-shaped group of closes then in Upton Scudamore parish before rejoining the main road near Row Farm. The Perambulation text seems to be evidence that this was the old line of the lane. Today what looks like an old pollard mere oak can be seen along this line. The boundary here was extinguished in 1934, with the creation of the civil parish of Chapmanslade. Harris of course knew about this loop in the boundary line, so he was forced to interpret some of the next portion of text - ‘turning down to Biss A PERAMBULATION OF THE HUNDRED AND PARISH OF WESTBURY 113 Ford Lane’ - as describing it, and in one of his asides (appearing as a footnote in Modem Wiltshire ) erroneously says that this was the way towards Short Street. The excursion down Bissford Lane, in fact the lane leading south-east from Southcroft at the eastern end of Chapmanslade village, is a much more interesting facet of the Perambulation. Until the 1934 reorganisation, the parish boundary ran down the main street of Chapmanslade, so that at the time of the Perambulation dwellings on the north side were in Westbury and those on the south were in Corsley parish - apart from the curious exception which the quoted extract now covers. In a significant difference from any of the modern boundaries the 17th-century one clearly includes a triangular enclave of closes here, around which the boundary runs before returning back up via Huntenhull (Huntley) Lane and a side alley to reach the village street at almost the point from which it deviated. This fascinating diversion must have been a controversial assertion of territory by the parish of Westbury. Surprisingly, the Southcroft triangle is still shown as part of Westbury on its enclosure map (1808).5 The Corsley enclosure map (part of Warminster, 1783) includes it firmly within its own parish bounds, however, and so far as can be made out the 1636 perambulation of Corsley does too (not all of the plots and their owners can be located with certainty). The explanation probably lies in the historical background to Huntenhull manor, once considered to reside administratively with Westbury, leaving a lingering tradition of the enclave belonging to the hundred (if not the parish) of Westbury. Example 2 This is an even worse entanglement, in which Harris misleads and was himself apparently misled by faulty text, creating the suspicion that an ‘original’ copying error had already been introduced to the document from which he was working. He has added long and, as analysis reveals, partially mistaken glosses on the locations, and Colt Hoare’s printer compounds the difficulty by misplacing a closing bracket (the one indicated by the asterisk in the extract) making it appear that the Perambulation has nearly reached Bratton, whereas it is still making its bounds with North Bradley parish south-west of Yarnbrook. As will be shown, the reader should mentally discard the matter in brackets, some of which was in fact omitted in print, while noting that the alternatives ‘strait or streight’ seem - conversely - to be lacking the usual brackets. To Kings Bridge (the same to this day) on to Rittles Lane (Rittles Lane Now called Kittles Lane; is an Old Road that passes between Haywood House Estate in the Hundred of Westbury, and Roud Ashton, or West Ashton woods in the Hundred of Whorwells Down - and so On to below Bratton*) from a Ball on the way to . . . the Meer Oak All which way doth divide The Kings Wast of Steple Ashton (vid.West Ashton in the parish of Steple Ashton) and the Wast belonging to Sir James Ley (afterward Created Earl of Marlborough) and so by a strait or streight line to a Meer Oak . . . then along the way to a Ball Athwart the Highway, Between Heywood and North Bradley Called the Mill Way (before the present New Turnpike road was made) to a Ball by the Ditch of Pikes Copse Corner along by the same Ditch by Another Ball, and so Upwards by the Way to a Young Meer Oak and On to Ryttles Lane, Kings Bridge is Horse Bridge near Dursley as mentioned above. Nearby, after a short distance, the boundary leaves the watercourses it has been following since it left the Somerset border. There seems at first no way to match the ensuing instructions to the current boundary line, having to allow for Steeple Ashton coming into the picture well in advance of the crossing of the Millway (identifiable as the A350), and taking in the confusing double appearance of Kittles Lane; (Harris - or his predecessor? - almost certainly misread K as R, a common confusion in handwriting of the period, and the printer’s transformation of the word to ‘Bittles’ only adds to the muddle). A way through the puzzle has been found, but first the text needs to be dissected to reveal the snares in more detail. The first suspect clause is the instruction to pass ‘on to Rittles Lane’ after leaving the stream. Harris’s parenthetical description of Kittles Lane is more or less correct - it is indeed a lost lane running from the woods north of Heywood House up to the West Ashton-Edington road and in the 17th century was a cause of much dispute between the hundreds of Westbury and Whorwellsdown. However, its appearance is premature. The words ‘On to Ryttles Lane’ are repeated at the end of the above extract, and here they are in the right place, apart from the faulty substitution of the initial letter. A copying error must have crept in, and the variation in spelling suggests that this may pre-date Harris’s work. The lane onto which we should have passed at the beginning of the extract is the one now called Dursley Road. However you read it, passing on to Kittles Lane here is geographically impossible (discounting the unlikely coincidence of a separate 114 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 5 Remains of the ‘Young Mere Oak’ in the woods north of Heywood House, now on the Woodland Trust’s database of ancient trees lane actually called Rittles), but some sort of marker is required to move perambulators on from the bridge to the next set of markers (the series of balls identifiable as leading across the commons belonging to Yarnbrook and Hawkeridge). An error in naming the lane at this point is therefore postulated. The explanation of the reference to Steeple Ashton at a stage where the neighbouring territory in fact belonged to North Bradley, regarded as a separate parish since at least the 15th century (Rogers 1965, 218) lies in an anachronistic survival. Evidence that the area of ‘waste’ in question was still regarded as part of the king’s manor of Steeple Ashton is to be found in a 1617 survey of local coppices (BL AddMS 6027). Harris has thus unnecessarily brought the tithing of West Ashton into the picture at this point in his explanatory gloss. After disentangling this section, the boundary can be followed across the present A350 at ST 871542, to run via the Piked (triangular) Copse miswritten as Pikes and now called Picket Wood and along to the old Kittles Lane. The second of the mere oaks mentioned in the extract (the young one) is shown on the enclosure maps for both Westbury (1808) and West Ashton (1818) and its remains can still be seen as a decaying hulk surrounded by secondary woodland (Figure 6). Example 3 Where they coincide, the Perambulation and the Anglo-Saxon charter boundaries seem to match well, but some disquiet may arise from intrusion of OE names into the former south of the village of Bratton. The following two extracts contain forms almost identical to those appearing in King Edgar’s Edington charter: Ewelmen (great springs) and Padecanstan (Padeca’s stone). Can they really represent survival in folk memory from the Saxon period? A PERAMBULATION OF THE HUNDRED AND PARISH OF WESTBURY 115 (a) ... up [the river] to the Meeting of the Two Springs Anciently Called the EWELM (or Melm) (b) . . . And then to Another Ancient Meer Stone in the Side of the Hill towards Padcanstone and from that Stone diretly in a Strait-line to a Pitt in the edge of that Hill . . . and so . . . Between Eddington field, and Brettons field to a Stone called Patten Stone, anciently Padcanstone Only before these two archaic forms does the word ‘anciently’ appear anywhere in the Perambulation, and to this researcher its use hints at later addition. The capitalisation in (a) is Harris’s (he also underlines the word) and his addition of the parenthetical ‘Melm’ may reflect his familiarity with a stream of that name in his native Dilton (Melme Water). In (b), the insertion (if such it is) of ‘towards Padcanstone’ interrupts the sense of the text, since ‘from that stone’ has to refer to the ‘ancient meer stone’ for the instructions to work.6 When the named stone appears in its correct location it is as Patten Stone, reflecting known contemporary usage. The source of the charter is the Edington Cartulary, which would have been available to Harris or his contacts after being purchased by the British Museum as part of the Lansdowne MSS in 1807 (Stevenson 1987, xxxiii). Intriguingly, it is thought to have passed through Sir James Ley’s hands, so might also have been consulted in his day. If Harris was aware of it, there is no indication of this in Modern Wiltshire. Colt Hoare does indeed add a section of new information from the Cartulary (entirely concerned with local genealogy), which he says was obtained by a correspondent after the rest of the Westbury section had gone to press. Nevertheless, a suspicion lingers that these words could have been inserted into the original Perambulation by someone with access to a translation of the charter text. Conclusion These examples are just three of many that could be used to illustrate the difficulties of interpretation. For such a widely quoted record, the published Perambulation is revealed as a conspicuously unreliable source. Details quoted from it, or arguments based on it, by authorities such as Grundy (Grundy 1920) and the Victoria County History need to be treated with caution. On the other hand, the many intriguing features found on the ground and sifted from the references that it contains offer much scope for further insights and investigation by local historians and archaeologists. The reader is again referred to the full research report to pursue these. As with all boundary studies, the research exemplifies the need for an approach that combines the use of documentary sources and maps with a programme of fieldwork to examine actual features on the ground. Finally, if the publication of this paper leads to the discovery of the original document, the writer will welcome the testing of her conclusions about it. Endnotes 1. The locality once called Whorwellsdown is generally accepted as having been the rising ground stretching eastwards from Cresswelldown Farm in West Ashton (ST 904543) to beyond Raydown and Housecroft farms in Edington (ST 911545). The Cresswelldown location suggested for that hundred’s meeting place would be typically remote from the main settlements but close to the junction of three parishes (Westbury, Edington and Steeple Ashton) and thus near the boundary between the two hundreds (Rogers 1965, 197). This could be evidence for a former combined hundred, but there is no evidence that the meeting place was necessarily at this end of the down. 2. The location is a crossroads on the down south of Bratton. Not far away was Sedcomb, the first element perhaps from OE (ge)set (animal fold), so Harris’s word Sedland is a plausible rendering of the original text, but the printed version Sealland has become received wisdom in Bratton local history, with Sealland Cross appearing as an antiquity on 20th-century OS maps. A note on this ‘lost landmark’ appeared in WANHM 54 (1952, 365-6) suggesting an etymology derived from sealh land (willow land). 3. Testing the possibility that there was indeed a predecessor version in 1575, the only secular event found around then that might conceivably have prompted a perambulation of the hundred was a Webbe/Chocke family settlement in 1574 (Crittall 1965, 150), in which proprietorship of part of the hundred features. 4. One example is a field corner near Fairwood Farm, Dilton, dealt with in Standerwick enclosure legislation in 1818 (Somerset Heritage Centre Q\RDe/4). 5. It is also shown on Greenwood’s 1820 Wiltshire map; he presumably used the enclosure map as his source. For a more detailed exploration of the historical background to this anomaly see the full research report. 6. The misleading introduction of Padcanstone too early on the route led Crawford astray in his assigning of its location on his annotated 6-inch OS map of the area (Crawford c. 1930). 116 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE References ANDERSON, O. S., 1939, English Hundred-Names: the south-western Counties. Lunds Universitets Arsskrift. N.F. Avd.l Bd 35. Nr 5. BIRCH, W. DE GRAY (ed.), 1893, Cartularium Saxonicum. vol.3. London: Charles J. Clark CRITTALL, E., 1965, ‘The Hundred of Westbury’, in The Victoria History of the Counties of England. A History of Wiltshire , Volume 8, 136-192. London: Oxford University Press GOVER, J. E. B., MAWER, A. and STENTON, F. M., 1939, The Place-Names of Wiltshire: Cambridge: Cambridge University Press GRUNDY, G. B. 1920. Saxon Land Charters of Wiltshire, Second Series . Archaeological Journal 72, 8-126 HOARE, R. C., 1830, The History of Modern Wiltshire , Volume 3. London: J. Nichols HOBBS, S. (ed.), 2003, Wiltshire Glebe Terriers 1588-1827. Wiltshire Record Society 56. Trowbridge JACKSON, J. E. 1891. Westbury under the Plain (paper read in 1889). WANHM 25, 33-57 MORRISON, J. 1993. Bratton Woollen Mills Part One. Bratton History Association Journal 3, 14 ROGERS, K. H., 1965, ‘The Hundred of Whorwellsdown’, in The Victoria History of the Counties of England. A History of Wiltshire, Volume 8, 193-263. London: Oxford University Press SMITH, S. 1952. A lost landmark (unattributed Note). WANHM 54, 365-6 STEVENSON, J. H., 1987, The Edington Cartulary , Trowbridge: Wiltshire Record Society 42 Primary sources Wiltshire and Swindon Archives: WSA EA23, Corsley and Warminster enclosure map, 1783 WSA EA76, Westbury enclosure maps, 1808 WSA EA107, West Ashton enclosure map, 1818 WSA EA115, Bishopstrow enclosure map, 1811 WSA T/A Westbury, tithe map and award, 1842 WSA 366/1, Manuscript notes of the Ley family, 17th cent. WSA 383/907, R. C. Hoare’s correspondence from fellow antiquaries etc WSA 540/320, R. Harris’s draft notes on Westbury history for Hoare’s Modern Wilts, c. 1820 WSA 967/14, Accounts for lands in Westbury Hundred, James Greenhill for Sir James Ley, 1612-17 WANHS Library, Devizes: DZSWS MSS 3, H. Wansey’s research notebook for the Hundred of Warminster O. G. S. Crawford’s annotated 1901 edition of 6- inch to 1 mile OS Map XLV NW, c.1930 Longleat House Archive: 10061, Surveys, perambulations, inquisitions, correspondence etc relating to lawsuit involving Sir Thomas Thynne’s attempt to enclose part of Selwood Forest, c.1632 Corsley 547, Abstract of the Church Book, 1754 Somerset Heritage Centre: Q\RDe/4, Berkley enclosure award, 1818 British Library (BL): AddMS 6027, ‘Abstract of divers manors landes and tenements latelye graunted unto Prince Charles [by his father King James]’, Office of Surveyor General, 1617, fos. 130-133 The National Archives (TNA): C 66/1847/14 Grant of Westbury Hundred to James Ley, 3 June 1610 C 142/355/33, Inquisition post mortem, Nicholas Phipps, 1616 C 142/355/93, Inquisition post mortem, John Lambe, 1615 PROB 11/127, Will of Nicholas Phipps, 1616 Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 117-144 Early to middle Anglo-Saxon settlement, a lost medieval church rediscovered and an early post- medieval cemetery in Wilton by Rob De’Athe with contributions by Phil Andrews, John Chandler ; Jessica M. Grimm , Jacqueline /. McKinley , Lorraine Mepham and Ruth Pelting , and illustrations by S.E. James Excavations in 2007-8 revealed the first evidence for Early to Middle Anglo-Saxon settlement in Wilton, in the form of a large sunken-feature building of probable 7th-8th century date, possibly used for the manufacture of fired clay loomweights. The site lay just inside the western defences of the Late Saxon burh and saw further occupation in the Late Saxon period and after, with properties fronting onto West Street bounded by ditches at the rear. Part of the wall of a substantial medieval building, identified as the ‘lost’ church of St Andrew, Ditchampton, was exposed in the parcel of land to the rear of the tenements. Associated with it was a small but tightly packed inhumation cemetery, many of the graves probably post-dating the church’s demolition in the second half of 16th century, after which the churchyard continued to be used as a burial ground into the 1 7th century, so covering a period with few securely dated human bone assemblages. Introduction Despite Wilton’s importance in the Late Anglo- Saxon and medieval periods, first as a royal seat of the Anglo-Saxon kings of Wessex, then as one of King Alfred’s defensive burhs , before developing into the county’s principal urban centre prior to the establishment of the new city of Salisbury, the town has been subject to relatively limited archaeological excavation. However, a series of investigations between 2006 and 2008 on the south side of West Street, culminating in excavations at 41-43 West Street, have added significant new information about the town’s Early to Middle Anglo-Saxon origins and the site of one of its medieval/post-medieval churches. The excavation reported on here was undertaken by Wessex Archaeology between November 2007 and January 2008 on land at the former Wilton Autos site (NGR 409420 131370) in advance of residential redevelopment. The site lies immediately north of land at 35 West Street, which had been subject to archaeological works relating to an earlier phase of development, the results of which are integrated here where appropriate (Figure 1). The site is located c. 0.25km north-west of Wilton town centre, on the narrow peninsula between the Nadder and Wylye river systems (Figure 1). It comprises a c. 0.09ha sub -rectangular plot bounded by West Street to the north-east and Netherwell Lane to the north-west. To the south- east lies a house, garden and outbuildings, with new housing development to the south-west, beyond Wessex Archaeology, Portway House, Old Sarum Park, Salisbury SP4 6EB 118 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 132000 Kingswa House' 22 The Hollows 'Trskine*' Barracks Burden®® rwrq St John’s Hospital WILTON Ditchampton kingsbury’ «iooouidJ Wilton House | Pallid iarr - fridge South Street mim (1995) • Washer n ?,G range j right. All rights reserved. Ordnance Survey Licence number 1 000281 9Ct* © Crown copyright St John’s Hospital Approximate extent of Late Saxon settlement i \\ Excavation area Watching Brief area r ; Evaluation trench T ! Test pit Previous excavation 50 m Ordnance Survey © Crown Copyright 2010. All rights reserved. Licence number 100020449 Fig. 1 Site location and extent, showing excavated areas and related investigations EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 119 which are the channels of the River Nadder. The site is relatively flat and situated at c. 55m above Ordnance Datum (aOD). The geology comprises a spur of Valley Gravel overlying Cretaceous Upper Chalk (Geological Survey of Great Britain 1976 - Salisbury, Sheet 298). Historical and archaeological background Archaeological evidence from Wilton has recently been reviewed as part of the Wiltshire Extensive Urban Survey (McMahon 2004; see also James 1962). The Saxon settlement possibly originated in the 5th or 6th century. By the mid-9th century, documentary references to the occasional presence of the kings of Wessex emphasise its importance as the administrative centre of a royal estate. By the later 9th century, when Winchester emerged as the major town in the kingdom, the importance of Wilton as a royal seat had waned, although it is listed in the early 10th century document known as the Burghal Hidage as part of a chain of defensive burhs created by King Alfred in the late 9th century (Hill 1969). It was also the principal mint for the county in this period, and a nunnery, founded in c. AD930, grew into one of the most important religious houses in the country. The Domesday survey (1086) records eighteen burgesses under seven manors in Wilton, and two other manors held five houses there. The suburb of Ditchampton to the north-west, held by the Bishop of Bayeux, had a small agricultural assessment but contained four mills. By the end of the 11th century there were eight parish churches, with an additional four in the suburbs, indicating a considerable population within what was then the principal urban centre in the county. Wilton declined in importance in the 13th and 14th centuries due largely to the establishment of the planned city of New Sarum (Salisbury) in 1219. Despite the importance of the Late Anglo-Saxon town, archaeological evidence is scarce. However, evaluations on the north side of Netherwell Lane (south of St John’s Hospital) in the 1970s (Anon. 1971; Anon. 1972), and by Wessex Archaeology in 1993, followed by excavation and a watching brief in 1996 and 1997 revealed a well-preserved sequence through the ditch and bank of the burh defences (Andrews et al. 2000). Phases of archaeological work Three phases of archaeological work were undertaken on land at 35 West Street (Figure 1). A test-pit evaluation in 2006 revealed well-preserved medieval remains, with possible structures and occupation remains in test pits 1 and 4, including evidence of metalworking, within the backyard areas of one or more properties fronting West Street (Wessex Archaeology 2006). An excavation south of test pit 4 in January 2007 revealed no features, suggesting that this area was part of the unoccupied flood plain (Wessex Archaeology 2007a). A watching brief in May 2007, however, covering the area of test-pits 1 and 2, recorded part of a substantial foundation for a robbed stone wall and a grave during the excavation of a trench (see Figure 3). Further human remains in a lime-filled pit to the east may be from other disturbed graves reburied at some later date. Previous work at 41-43 West Street comprised two evaluation trenches (10m and 20m long) in October 2007 (Figure 1), which revealed medieval and later features, including ditches, pits and post-holes (Wessex Archaeology 2007b). As a result, following Wiltshire County Council’s recommendation, two areas were excavated covering most of the development footprint - Area 1 (26.5m x 13.5m) in the south-west, and Area 2 (12m x 15.5m) near the West Street frontage (Wessex Archaeology 2008a); an area in between was subject to a watching brief in February 2008 (Wessex Archaeology 2008b). The methodology for the excavation is set out in detail within the Written Scheme of Investigation (Wessex Archaeology 2007c). Full reports on the results of the documentary research, and the finds and environmental analyses, are held in the project archive (Wessex Archaeology project code 60517). Excavation results Early to Middle Anglo-Saxon: Ith-Sth century (Figure 2) The earliest feature was a shallow ditch (678), up to lm wide, running east-west across the southern corner of Area 1. Although no dating evidence was recovered, it was cut by a large Sunken-Featured Building (SFB) (703) and several later features, and the nature of the small charred plant assemblage 120 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 2 Plan and section of SFB 703 EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 121 recovered from it (see below) suggests an Early or Middle Anglo-Saxon date. The SFB comprised a sub-rectangular hollow, aligned east-west, measuring an estimated 7.5m long by 6m wide; its southern side lay outside the excavation area and its eastern end had been heavily disturbed by later features. It was c. 0.4m deep with moderately steep concave sides and a flat base. It contained two hearths and several pits, post-holes and stake-holes. One of the post-holes (690), which was positioned centrally against the west end, was 0.5m in diameter and 0.45m deep (from the top of SFB cut), and there was a similar post-hole (564) near the centre of the east end. Several smaller post- holes (409, 584, 652, 654 and 672) and a group of stake-holes are likely to be contemporary, although of uncertain function. The largest feature was pit 622, measuring 0.9m in diameter and 0.5m deep, while feature 568, measuring lm by 0.5m and 0.5m deep, may have also been a pit, although it could have held a post associated with, or replacing, that in post-hole 564. There were patches of ‘greasy’ orange/black clay (550) up to c. 0.1m thick on the base of the hollow. This layer, probably an occupation deposit, was rich in charcoal and contained a large number of ceramic loomweights, along with animal bone and burnt flint, as well as a bone pin beater and an iron knife. The charcoal probably came from the two hearths (580 and 599), comprising irregular patches of burnt clay on the base of the hollow, in its northern half. These may have been used to bake the loomweights, further examples of which, some incompletely baked, came from the hearths. None of the groups of loomweights displayed any obvious spatial pattern that might indicate the presence of a loom in the SFB. These basal layers were overlain by a largely sterile deposit, c. 0.2-0. 3m thick, of mid-dark brown compact silty loam (553), possibly representing post-abandonment backfilling of the hollow. The hollow was overlain by a cemetery soil (184, see below) derived from the digging and intercutting of post-medieval graves and their disturbance of earlier (Saxon and medieval) contexts. Late Anglo-Saxon-medieval: 10th-12th century (Figure 3) Three ditches (700, 701 and 702) in the south- western part of Area 1 may reflect the rear boundaries of properties fronting West Street. All were between lm and 2m wide and up to 0.5m deep. Ditch 700, which ran roughly north from a shallow terminal, Plate 1 Robbed remains of wall 235, the north-east corner of St Andrew’s church ( scale 2m) was cut by ditch 702, which was aligned north-west to south-east. To their west, the southern terminal of ditch 701, which was aligned north-north-west to south-south-east, was truncated by modern disturbance (see Figure 4), so that its relationship to pits 186 and 411 (below) was unclear. Pits of this period, all of which contained pottery and animal bone, were found in both areas of the site, but dominated in Area 1 where they appear to have been used for the disposal of general domestic waste. The 11 pits in Area 1 (156, 178, 186, 286, 289, 291, 302, 314, 317, 381 and 411), all but three of which intercut, ranged from 1.4m to 2.2m in diameter, and were generally larger than the 12 pits of this period in Area 2 (278, 324, 342, 358, 360, 369, 392, 424, 428, 432, 441 and 469), all but two of which intercut. Medieval: 13th-14th century (Figure 3) St A ndrew ’s church A substantial but heavily robbed wall (235), perpendicular to Netherwell Lane, ran along the south-west edge of Area 1 (to the rear of the earlier ditches), continuing beyond the limit of excavation to the south-east and turning to the south-west at its north-west end (Plate 1). Its foundation trench, which was at least lm wide and c. 0.4m deep, had a 122 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Church wall 235 SITE OF . • ST ANDREW’S \ \ CHURCH Late Saxon/early medieval (1 Oth-1 2th century) feature 13th/14th century feature Medieval feature Grave 105 Undated feature 0 10 20 m ffl Fig. 3 Plan of Late Saxon/early medieval and 13th-14th century features compacted layer of flint nodules at its base, overlain by wall footings of large, sub -rounded flint nodules with occasional lumps of greenstone and sandstone, all bonded by mortar. Surviving elements of the wall included part of an in situ ashlar block at the base of the north-east corner. Inside the corner was a layer of compact silty clay (511), possibly the remnant of a floor surface (or the make-up for a floor), or material that had accumulated after the wall had been robbed. On the outside, a cobbled surface (165) of flint nodules extended north-west beyond the edge of the excavation. EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 123 A lm length of robbed wall (107), c. 1.75m wide, had previously been recorded on approximately the same line as wall 235 c. 15m to its south-east (Figure 3); this was observed in the base of a trench during the watching brief at 35 West Street (although it was not observed in test pit 1 of the 2006 evaluation in the same location). The wall was angled slightly more towards the south than wall 235, perhaps the start of a curve, although this could not be demonstrated with certainty. It too had been heavily robbed with only its footing surviving. Also recorded during the watching brief, c. 5m south-west of wall 107, was an east-west aligned grave, probably medieval, containing an adult male laid supine (105); there were no grave goods or coffin fittings. Documentary evidence (see below) suggests that the wall formed part of St Andrew’s church, and although the foundation date of the church is unknown, the remains indicate a medieval date. In 1564 St Andrew’s was amalgamated with St Mary’s, Wilton, but it appears to have been demolished in the late 16th century, although the former churchyard continued to be used for burial perhaps until c. 1630 (see below), after which its location became largely forgotten. Other features There were four post-holes (354, 355, 356 and 384) in the northern corner of Area 2, three on the edge of the excavation, the most substantial of them (354) containing packing in the form of several large flint nodules. These probably formed part of the rear of buildings fronting West Street. Four other post- holes in Area 2 were less substantial and showed no spatial patterning. Twelve medieval pits were identified in Area 2 (265, 362, 367, 457, 465, 474, 478, 493, 496, 498, 501 and 704), several of them intercutting; four (474, 496, 498 and 501) were heavily truncated by other pits and are not shown on Figure 3. They were 0.8m to 1.8m in diameter and were up to 0.7m deep. A further four pits (150, 158,417 and 688) lay within Area 1 and, as in the earlier period, were generally larger than those on Area 2. In both areas the pits contained domestic waste and are interpreted as rubbish pits. A c. 0.7m diameter well (299) near the centre of Area 2 was built of regularly coursed, finely jointed chalk blocks, medieval brick and some flint nodules. Its western side was cut by a modern pit (not shown), and its depth was not established due to the high water table. The uppermost fill contained medieval pottery and animal bone, and a large heavily worn millstone. Plate 2 Grave (1 12) cutting church wall foundations (scale = lm) Post-medieval ( 1500-1800) (Figure 4) St Andrew’s church is likely to have remained standing for at least the first half of the 16th century with the churchyard used for burial into the 17th century. A large number of graves were revealed in Area 1, in a c. 6m wide area in its southern corner, which documentary and cartographic studies have shown falls within the north-east corner of the churchyard, although no churchyard boundary was identified. The remains of 45 articulated skeletons and much disarticulated human bone representing a minimum of 29 individuals were recovered. Some of the graves cut through a cemetery soil (184), particularly the upper levels, where cuts were fairly clear but shallow. This soil contained mixed material, including Anglo-Saxon loomweights, medieval domestic waste and disarticulated human remains, all probably derived from Anglo-Saxon and medieval contexts by intercutting post-medieval graves. Later intercutting made earlier cuts difficult to distinguish and undisturbed burials were rare. Several graves cut the robbed remains of wall 235 (Plate 2). Residual pottery from some of the graves suggests a medieval to early post-medieval date 124 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE |; , | Post-medieval Modern disturbance [ } Undated feature 0 10 20 m V SUE OF \ ST ANDREW’S CHURCH CHURCHYARD BOUNDARY Fig. 4 Plan showing area of graves and other post-medieval features range for the burials, while the absence of coffin fittings indicates a pre-18th century date. Overall, the majority of the burials appear to date to the first two or three decades of the 17th century. A number of pits in Area I were also of post- medieval date. Pit 167 cut through wall 235 but contained no material to date it more precisely. Pits 222 and 288 were isolated, while others (192, 194 and 196, and 306, 310 and 314) intercut. In Area 2, pit 436 contained late medieval/early post-medieval pottery, while nearby pits 297 and 455 produced post- medieval material. A lm diameter well (202), cutting cobbled surface 165 (see above) in the north-west corner of Area 1, was roughly lined with brick, stone, chalk and flint. Its upper fill contained modern brick, glass, animal bone and residual medieval pottery. EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 125 Documentary analysis by John Chandler The history of the site is bound up with the settlement of Ditchampton (see Figure 1), regarded as one of the ‘suburbs’ of Wilton, but which possibly began life as a community separate from, and equivalent to, pre-urban Wilton and other hamlets or villages in the neighbourhood. Ditchampton is recorded in an Anglo-Saxon charter of AD1045 (www.anglo-saxons.net, charter S1010). Since, as will become apparent, the excavation site adjoined the religious focus of Ditchampton, it is probable that the creation or extension of Wilton as a Burghal Hidage stronghold (laid out along West Street in the late 9th century) impinged upon this early community, and the Anglo-Saxon fortification across the neck of the peninsula between the Nadder and Wylye river systems had the effect of divorcing its church (thus brought within the town) from its territory, and later parish, which lay outside (for a recent topographical analysis see Chandler 2007; Haslam 1984, 122-8). If, on topographical grounds, it is safe to assume that the site falls on part of a rectangular tenement, created during the 9th century urban development, running back from West Street alongside Netherwell Lane, then it is probable that a dwelling occupied the street frontage with backlands, enclosures and paddocks behind. It is likely that it is either the 9th century tenement partly covered by the site, or one of its neighbours, that is referred to in an early 13th century lease by Bradenstoke Abbey (near Lyneham, north Wiltshire), where it is described as ‘a messuage in Wilton near the West Gate’. Bradenstoke was an Augustinian daughter house of Cirencester, which had been endowed with land in Wilton in about ADI 190. The tenement in question had once been held by Laurence the butcher, of Wilton, but was now being leased to Wulricus Tahium of Wilton and his heirs (London 1979, 29 (no.5), 103 (no. 312)). It may also be one of four Wilton tenements, the subject of a feet of fine in AD 1227 (London 1979, 157 (no. 532)). The property presumably remained in Fig. 5 Detail from the 1860 Wilton enclosure map (WSA EA 179), covering the area of the site and showing the parcel of land (13a) formerly occupied by St Andrew’s church. The accompanying award refers to “All that garden near to the said timber yard in Wilton called St Andrews Litten numbered 13a” 126 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Bradenstoke’s possession until the priory’s dissolution in 1539, after which (according to a 1565 survey) it was held by John Byldinge, rector of St Andrew’s church, Ditchampton. It was then described as ‘a parcel of land lying next to the rectory house of St Andrew’s and in the present occupation of Hugh Tinker Esq, lying on the north side of the rectory and extending as far as the common street, and formerly in the possession of the prior of Bradenstoke’ (Straton 1909, vol 1, 182). A marginal note in the manuscript of this survey (WSA 2057/S3) reads modo Doctor Babington (‘now Doctor Babington’). This must refer to Gervase Babington (c. 1549-1610), household chaplain to the Herbert family of Wilton House, who was appointed rector of Ditchampton and St Mary’s Wilton in 1585, and seems to have retained the livings even after his appointment as bishop of Llandaff in 1591 (< Oxford Dictionary of National Biography). The parish church of Ditchampton, referred to in documents from 1291 onwards, was not appropriated by Wilton Abbey, and its advowson passed with the lords of Ditchampton manor until being sold to Sir William Herbert in 1547. The church was united with St Mary’s, Wilton in 1564, and presumably became disused thereafter and disappeared (James 1962, 31), although the parcel of land shown on the 1860 Wilton enclosure map (Figure 5; WSA EA 179, no. 13a) is described as St Andrews Litten (i.e. churchyard). There is a brief hiatus in the Wilton House archives between Babington’s departure in the 1590s and the next estate survey, of 1631-2. It seems plausible, however, to identify the site (or part of it at least) with a 1629 lease on three lives to William Abyn, of ‘a little garden plot in Wilton near St Andrew’s church, between the lands heretofore of Ralph Symes deceased and now of Thomas Antram on the east, and those of St Andrew’s on the south- west, except all timber trees and trees likely to be timber’ (Kerridge 1953, 84 (284)). The site from this time onwards seems to have been largely open, as garden, orchard, hopyard, and later timber yard, and what buildings subsequently occupied parts of it were connected with trade rather than dwellings. Finds Pottery, by Lorraine Mepham A total of 1206 pottery sherds (21, 264g) of Early/ Middle Anglo-Saxon to post-medieval date was recovered. This report focuses on the Saxon and medieval assemblage (1082 sherds; 16, 819g); details of the post-medieval wares (124 sherds; 4445g) can be found in the project archive. The medieval ceramic sequence for the area is reasonably well established, from excavated assemblages from the 13th/early 14th century Laverstock kilns outside Salisbury (Musty et al. 1969), and from sites at Old Sarum and within Salisbury (e.g. Mepham 2000a; 2005; Musty et al. 2001). The precise chronology and potential sources of supply for the ‘Saxo-Norman’ period (i.e. the 10th to 12th century) are not fully understood, although small assemblages have already been published from Wilton (Andrews et al. 2000), Old Sarum (e.g. Stone and Charlton 1935) and Amesbury (Powell et al. 2009). Analysis of the pottery assemblage follows the standard Wessex Archaeology recording system (Morris 1994), resulting in the definition of 15 fabric types (of which three are variants of a single type) (Table 1). All data are held on the project database (Access), which forms part of the project archive. Early I Mid- Saxon Early/Middle Anglo-Saxon pottery comprised a mere seven undiagnostic sherds (129g) in organic- tempered and sandy fabrics. Organic-tempered fabrics have a known currency from the 5th to the 8th centuries. Given their provenance here - three sherds from SFB 703, together with loomweights of Mid- Saxon type (and four from a post-medieval soil layer) - a date range in the 7th or 8th century seems likely. Late Saxon to medieval There is no evidence of continuity between the Early/Mid- Saxon and the Late Saxon/early medieval assemblage. The latter has a start date probably no earlier than the 10th century, and the bulk appears to fall in the 10th to 12th century range, declining into the 13th century with only a few late medieval sherds. Seven ‘ware groups’ have been identified. Laverstock-type coarsewares The most common wares within this group are similar, in terms of the dominant inclusions (abundant rounded quartz grains, coloured pink, grey or clear), to the products of the 13th/early 14th century kilns at Laverstock (Musty et al. 1969). Three subdivisions, based on the size of quartz inclusions, defined for sites in the Salisbury area (e.g. Mepham 2000a), have been used here (E422a, b and c, E422a EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 127 being the coarsest variant), reflecting a general trend through time from coarse to fine. Scratch-marking, apparrently a post-Conquest phenomenon in the region, is frequently seen on the coarser variants; on the finer variants it is shallower and appears more like brushing. Vessel forms at West Street comprise jars (66 examples on the basis of rims), dishes (seven), tripod pitchers (five) and lamps (three). The jar forms match those found within mid-late 12th century pits at the Laverstock kilns (Musty et al. 1969, fig. 7), but also occur in cesspits at Old Sarum dated to the late 11th or early 12th century (Stone and Charlton 1935, fig. 4). Jars with short, stubby, squared rims, sometimes grooved (Musty etal. 1969, fig. 10, 36-8) are infrequent here, and are not seen in the coarser fabric variant E422a; these jars do not seem to appear before the 13th century. There are five shallow, convex dishes; one late 1 1 th/12th century example is known from Old Sarum (Stone and Charlton 1935, fig. 5, 24), although in general this form seems to be 13th century or later (Musty et al. 1969, fig. 12, 56-7). Two distinctive ‘West Country’ dishes, with acute base angles (ibid., fig. 11, 41-4), also have a date range spanning the 12th to 13th century, while a 13th century date is most likely for the single convex bowl with thickened rim, possibly a curfew. All the dishes and the bowl are in the medium-grained fabric variant E422b. Pedestal lamps (Figure 6, 1), of which there are two rims and one almost complete profile here, all in fabric E422b, are known in both late 11 th/early 12th century and 13th century contexts (Stone and Charlton 1935, fig. 5, 23; Musty et al. 1969, fig. 24). Tripod pitchers are the only form which does not have a currency beyond the 12th century. Several stratified (late 1 1 th/early 12th century) examples are known from Old Sarum (Stone and Charlton 1935, fig. 5, nos. 17-9, 22, 25, 29, 31), but none was found in the mid- to late 12th-century pits at Laverstock. A minimum of five vessels has been recorded from the West Street site on the basis of rim/spout sherds, to which can be added one tripod base, and decorated body sherds from seven different vessels; these occurred in fabrics E422a and E422b. Only one tubular spout survived (Figure 6, 2), but other rims have been recognised by the presence of decoration and/or glaze. Decorative techniques used include rouletting and combing (linear or curvilinear) on pitcher necks, but also on bodies, often within bands created by the application of vertical strips (Figure 6, 3-6). All vessels are at least partially covered by a thin, patchy, olive-green to green-brown glaze. These Table 1. Pottery totals by fabric type Fabric code Description No. Weight (g) Q400 Early/Mid- Saxon sandy ware 2 16 V400 Early/Mid- Saxon organic-tempered fabric 3 100 V401 Early/Mid- Saxon sandy fabric with some organics 2 13 E400 Cheddar fabric E 33 634 E403 Michelmersh-type ware 4 85 E420 Laverstock-type fineware I 9 52 E421 Laverstock-type fineware II 155 1945 E422a Laverstock-type coarseware (coarsest variant) 176 3244 E422b Laverstock-type coarseware (medium variant) 493 8127 E422c Laverstock-type coarseware (finest variant) 77 839 E428 West Wiltshire-type coarsewares 38 369 E526 North French monochrome 1 4 C400 Coarse, calcareous fabric 26 935 F400 Coarse flint-tempered fabric 61 434 F401 Sparsely flint-tempered fabric 2 22 Total 1082 16,819 Key: (C = calcareous; F = flint-tempered; Q = sandy; V = organic-tempered) tripod pitchers were initially recognised as a regional type of ‘South-east Wiltshire pitchers’ (Vince 1981), and examples have been found widely over Wiltshire and neighbouring counties. They do not seem to have been a particularly common vessel form, and could be considered as fulfilling the function of higher quality ‘tablewares’; they have been found, for example, at ‘higher status’ sites such as Trowbridge Castle, Wiltshire, Sherborne Old Castle, Dorset, and Carisbrooke Castle, Isle of Wight (Mepham 1993, fig. 38, 31-2; in prep.; 2000b, 104). They also occurred, however, at the village site of Gomeldon, near Salisbury (Musty and Algar 1986, fig. 19, nos. 85-7). Their precise source remains unknown, although the likelihood is that it lies somewhere in the Salisbury area. Laverstock-type finewares These can be fairly confidently attributed to the 13th/early 14th century Laverstock kilns. Sherds are invariably glazed and probably derive from jugs - four jug rims (two with bridge spouts) and two strap handles were found, alongside a number of decorated sherds, all typical of the known range (Musty et al. 1969, figs. 13-18). One sherd from a base of small diameter could be from a bottle (ibid., fig. 22, 179-80). Cheddar-type wares A small quantity of wheel-thrown wares of Cheddar 128 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE type were identified; these are all of Cheddar fabric E, hard-fired and containing a mix of quartz sand, limestone and flint/chert. Diagnostic forms comprise three jars, all with short, everted, slightly thickened rims (Figure 6, 7) similar to examples found at South Street (Andrews et al. 2000, fig. 6, 1). The type was found in late 10th and 11th contexts at the Anglo- Saxon palace at Cheddar (Rahtz 1979, 309-18). Michelmersh-type ware Only a few sherds of wheelthrown Michelmersh- type ware (E403) were present, in contrast to the large group found at South Street (Andrews et al. 2000, fabrics Q400, Q404). Since the publication of the latter site, a kiln site with an archaeomagnetic date of late 10th to 11th century has been located at Michelmersh, which was producing jars comparable to those found at Wilton (Mepham and Brown 2007). One of the sherds found at the West Street site is from the rim of a spouted pitcher of classic Michelmersh type (ibid., figs. 10-11), finger-impressed with an applied, finger-impressed neck cordon (Figure 6, 8). Calcareous wares The calcareous wares are in a single fabric type (C400), containing frequent chalk/limestone. Fabric C400 as identified at South Street is comparable (Andrews et al. 2000). Diagnostic forms, as at South Street, are confined to rounded jars, with everted plain or slightly thickened rims (Figure 6, 9). One jar has a deeply finger-impressed rim. Flint-tempered wares These are made up largely of sherds in fabric F400 from a single context and which may represent a single vessel (there are no diagnostic sherds). Fabric F400 is comparable to flint-tempered wares seen at South Street (Andrews et al. 2000, fabric F401). West Wiltshire-type wares These wares are distinctive by their high mica content. The type encompasses a range of coarsewares, in this instance containing quartz, flint, chalk or sandstone, or combinations of these. Comparable wares were found in quantity in Warminster, where they appeared to have a lengthy currency from at least the 11th century through the medieval period (Smith 1997, 20-4). Vessel forms are likewise difficult to date closely; those seen at West Street are confined to jars, which are typical of the range. Chronology While there is a significant group of material that can be dated as 10th to 12th century, separating this group from any later material is not a straightforward matter, given the currency of some of the wares. A fairly crude division has been made on the basis of the presence or absence of obvious 13th century or later wares, which in this case comprise the Laverstock-type finewares and the post-medieval wares. This dating has informed the overall site chronology, enabling a broad division into early medieval (10th to 12th century) and medieval contexts (13th/ 14th century). Comparison with the assemblage from South Street throws up some interesting contrasts, and could help in tying down the date range of the West Street assemblage. Figure 7 shows the breakdown of ware groups within the Late Anglo-Saxon/earlier medieval assemblages from both sites (the caveat as to the presence of later material amongst some of these ware groups applies to both sites). The proportions of the ware groups differ widely between the two sites, with only the flint-tempered and West Wiltshire types present in broadly comparable proportions. The discrepancy is probably chronologically related. At South Street there are much higher proportions of wares that have a date range of 10th to 1 1th century (Michelmersh and Cheddar types), and the amount of Laverstock-type coarsewares is correspondingly lower; the picture is reversed at West Street, where the presence of tripod pitchers (very rare at South Street) suggests a date range focusing on the late 11th to early 12th century. The 10th/ 11th century wares may, therefore, be largely residual on this site - there are only three features which produced these wares only (pits 278, 343 and 441), and of these only pit 441 contained pottery that had the appearance of an in situ deposit, sherds probably of a single jar in fabric C400 (Figure 6, 9). This assemblage thus augments the small but growing number of Late Saxon/early medieval assemblages from Wiltshire, and usefully highlights a group of material, probably largely of late 1 1th or early 12th century date, which is not otherwise well represented in the other published assemblages. List of illustrated vessels (Figure 6) 1. Pedestal lamp; Laverstock-type coarseware (E422b). PRN [Pottery Record Number] 397, Object No. 45, context 405, pit 411. 2. Tubular spout from tripod pitcher; Laverstock- type coarseware (E422a). PRN 3, context 132, EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 129 ditch 700 (cut 130). 3. Rim from tripod pitcher; combed and applied decoration; Laverstock-type coarseware (E422b). PRN 321, context 316, pit 314. 4. Rim from tripod pitcher; rouletted decoration; Laverstock-type coarseware (E422b). PRN 344, context 359, pit 358. 5. Handle from tripod pitcher; rouletted decoration; Laverstock-type coarseware (E422b). PRN 37, context 157, pit 156. 6. Decorated body sherd from tripod pitcher; combed and applied decoration; Laverstock-type coarseware (E422b). PRN 413, context 426, pit 424. 7. Jar rim; Cheddar fabric E (E400). PRN 250, context 288, pit 286. 8. Rim from decorated spouted pitcher; Michelmersh-type ware (E403). PRN 394, context 405, pit 411. 9. Jar rim; calcareous ware (C400). PRN 330, context 343, pit 441. Fired clay, by Lorraine Mepham A large group of Anglo-Saxon ceramic loomweights (197 fragments from a maximum of 71 weights) was recovered. Most (maximum of 63 weights) 130 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 7 Pottery: pie charts showing proportions of wares from West Street and South Street came from the SFB ‘occupation layer’ (550), with one group in the north-west corner, and others concentrated around hearths; it is possible that they were manufactured in the SFB. Five weights were found in the ‘cemetery soil’ (184) overlying the infilled hollow. Anglo-Saxon loomweights are conventionally divided into annular, intermediate and bun-shaped forms on the basis of the methods of manufacture and profile, reflecting a gradual chronological development from Early to Late Anglo-Saxon (Hurst 1959, 23-4). With the exception of one bun-shaped weight, all are of intermediate form, with several made from pierced discs with flattened surfaces (Plate 3). A Middle Anglo-Saxon date can therefore be suggested. Diameters range from 1 10- 180mm (30 measurable examples); a third of these exceed the ‘standard’ size of 100- 140mm, based on evidence from Mucking and London (Hamerow 1993, fig. 44; Cowie and Blackmore 2008, 196). As only five loomweights are complete, there is limited information on the original weights - these range from 385-897g, although all but the heaviest are under 600g. The normal range for the Early Anglo-Saxon period is 150-550g, with weights exceeding 500g more common in the Middle Anglo-Saxon period (Cowie and Blackmore 2008, 196). The fabric of the weights is consistent with the use of local, probably alluvial clay sources, some containing low levels of sand and rare flint gravel, and occasional organic matter. Many appear to have been under-fired, with soft, friable fabrics, but may still have been functional. Loomweights are generally rare, usually occurring in quantity only in deposits in situ within SFBs. A group of up to 31 weights from an SFB at Harmondsworth, west London, appears to represent a single loom (Farwell et al. 1999, plate 6), as does a group of over 60 weights at Upton, Northamptonshire (Jackson et al. 1969). Two looms are suggested for the group of at least 75 weights from Old Erringham, West Sussex (Holden 1976, 309, plates III-IV). Around 140 loomweights, found in a single SFB at Mucking, may have been stored there (Hamerow 1993, 17, 66-8). On this basis, the West Street weights could represent two looms; a bone pinbeater provides further evidence for weaving. Other finds, by Lorraine Mepham, with comment on the Slag by Phil Andrews The assemblage of ceramic building material is dominated by fragments of flat (peg) roof tile, dated as medieval by their coarse, poorly wedged fabrics and often pale-firing, recorded in quantity from medieval contexts in Salisbury (e.g. Loader 2000). One potential source is known from documentary sources in Alderbury (Hare 1991). Some are glazed on the lower part of the upper surface. In addition, there were four possible glazed ridge tiles and one EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 131 Plate 3 Selection of loomweights from the SFB possible glazed louver or finial, all in pale-firing fabrics comparable to the 13th century Laverstock kiln wares; the kilns are known to have produced roof furniture, although not flat roof tiles (Musty et al. 1969, figs, 25-6). Post-medieval materials include a few more regularly formed flat tiles in finer-grained, orange-red fabrics, three wall/floor tiles, and a few pieces of brick all apparently unfrogged and therefore probably predating c. 1800. Most of the stone assemblage comprises medieval building materials, including slate, shale, limestone and micaceous sandstone roof tiles, and other limestone and greensand fragments used for walling. The use of shale for roofing tiles is interesting, but is known elsewhere in Wessex in the medieval period, for example Shaftesbury, Dorset (Cox and Mills 1991, 173) and at Old Sarum Castle (Wessex Archaeology 2000). In addition, one large, very worn but complete millstone (580mm in diameter) was recovered from the top of medieval well 299. It is in a fine micaceous greensand (glauconitic sandstone) - Shaftesbury Greensand - which outcrops to the west of Salisbury. Two residual copper alloy Roman objects were recovered - a coin dating to the Flavian period (AD 69-96), and an abraded fragment of a bow brooch, probably of Hod Hill type. Other copper alloy objects include a pin, a lace tag, and two strip fittings, mostly from medieval contexts. The lead comprises two window came fragments, and one piece of wire. Much of the ironwork is heavily corroded, and consists largely of nails and other structural items, but other identifiable objects include a horseshoe, at least two knives, and four other possible tools. Over 7kg of ironworking slag was recovered, mostly in small quantities (<100g), but with six contexts containing 100-500g, and just three more than this: medieval pit 474 (683g), post-medieval well 202 (c. 1.7kg), and post-medieval pit 297 (c.2.6kg). Some of the debris comprises clinker or fuel ash slag, not necessarily derived from ironworking, but the rest probably indicates smithing. One piece weighing 725g from pit 297 appears to be a fragment of smithing hearth bottom. A fragment of window glass, virtually opaque through oxidation, has red painted decoration and is likely to come from an ecclesiastical or other high status building, probably St Andrew’s church; another has a grozed edge. The vessel glass is all likely to be post-medieval in date. 132 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Worked bone objects comprise part of a Saxon pinbeater (‘occupation layer’ 550 within SFB 703), a tuning peg (context 159), and a sawn and partly worked length of horn core from ‘cemetery soil’ 184. Ten clay pipe stem fragments were recovered, one stamped with the mark of W. Sayer (c. 1720-50: Atkinson 1970, 187). Spearheads from 22 The Hollows, by Phil Andrews The two Anglo-Saxon spearheads from 22 The Hollows (Figure 1) are also published here because of their intrinsic interest and likely significance in terms of settlement of this period in Wilton. They were recovered during small-scale building work in the late 1990s and were temporarily loaned to Wessex Archaeology for recording. The smaller of the two (Figure 8, 1) has been attributed to S wanton’s type Cl, found in relatively large numbers, but with a date range that appears not to extend beyond the late 6th century AD (Swanton 1973, 49-51). The larger spearhead (Figure 8, 2) is of Swanton’s type C2, the commonest of the leaf-shaped blades, which can be early but continues into the 7th century and beyond (Swanton 1973, 53-55). It is not certain that either of these spearheads came from a grave, particularly as no human bone was reported, but this seems the most likely context and perhaps any grave cut(s) went unseen in the narrow confines of the foundation trench from which they were recovered. Human bone, by Jacqueline /. McKinley Human bone from 126 contexts was subject to analysis, most of it from the south of Area 1 where all or parts of 45 in situ burial remains were recovered from a c. lm depth of cemetery soil, which also contained most of the redeposited bone (total 82 contexts). One deposit of disarticulated bone was from a modern feature in the evaluation trench in Area 1 (Wessex Archaeology 2007b), and four deposits (one articulated bone, three disarticulated) from the watching brief at 35 West Street to the south of the site. The discovery of so large a number of burial remains was unexpected. Limited funding was available for analysis and publication which resulted 100 mm Fig. 8 Anglo-Saxon spearheads from 22 The Hollows, Wilton (scale 1:2) EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 133 in a necessary adjustment to normal procedures; full analysis was undertaken on the articulated remains with limitations on certain areas of data collection; the disarticulated bone was subject to minimum number counts (MNI) according with methods outlined by McKinley (2004). Taphonomy Some of the graves cut through the cemetery soil (184), particularly those in the upper levels where they were fairly clear, but had survived to only shallow depths. The intercutting of graves and the mixing of remains increased with depth through 184 and cuts became more difficult to distinguish. Later features had also truncated some of the burial deposits. Graves extending beyond the limits of the excavation were not pursued as this would only have exposed more remains not under threat from the development. The level of disturbance is demonstrated by the high MNI within the disarticulated bone assemblage (see below) and the percentage of skeletal recovery within the articulated bone assemblage. Within the latter the majority of individuals (54.5%) were represented by less than 50% of the skeleton and only 16 (36.4%) were represented by more than 85%. The bone is generally in good condition (graded 0-1; McKinley 2004) with only moderate fragmentation and slight warping to some skulls. Puncture marks to a right rib from burial 559 (upper level) are indicative of canid gnawing and suggest that some of the burials were originally made fairly shallow and, therefore, that burial in the cemetery was perhaps not overseen as well as it could have been. Demographic data A MNI of 74 was identified from Area 1, 45 from the articulated remains (Table 2) and 29 from the disarticulated bone (Table 3). The assemblage includes a high proportion of immature individuals (51.3%), with an age range from foetal to subadult and a substantial peak between 0.5-5 years (i.e. infants; 20.8% of population) (Table 4). The adult age range extends from c. 18-20 years (burial 276) potentially to c. 55 years (burials 562 and 670). A similar proportion of adult females and males was identified within the disarticulated assemblages, with a slightly higher number of males in the articulated assemblage. The highest proportion of adults of both sexes fall in the 25-45 year range ( c . 54% adults; 41.2% females, 50% males) but a slightly larger proportion of females (41.7%) lay in the c. 35-55 year range, though the potential significance of this is limited given the overlap between the ranges. The proportion of immature individuals is unusually high for an archaeological cemetery population. Other post-medieval cemeteries (1550 - mid- 19th century) subject to archaeological analysis have been found to comprise 9-30% immature individuals with the exception one mid- 19th century London cemetery with 70% (McKinley 2008, table 17). Although the rate from Wilton may appear high, it is very close to the c. 50% of deaths of those below the age of 20 years shown in the London Bills of Mortality for the early 1 8th century (Roberts and Cox 2003, 303-4, table 6.5). Various arguments have been extended to explain the frequently observed dearth of young individuals from such cemeteries. The peripheral location of some graves and the shallow depth of others (with other possible taphonomic factors) are probably largely responsible for figures not always representing a realistic reflection of the mortality rate (Daniell 1997, 127-8; Scott 1999, 26- 7; Roberts and Cox 2003, 316-7; McKinley 2008, 61-63). The immature assemblage includes the rare recovery of remains from a very young foetus (approximately 19 weeks gestation), albeit as a single skeletal element. It may derive from the disturbed burial of a pregnant woman, from a very prematurely born baby, or from a naturally or deliberately aborted foetus. If prematurely born, its presence within the general cemetery soil suggests that it was afforded burial within the ‘main’ part of the cemetery, and that it had been possible to christen it before it died, since those dying un-christened were often excluded from cemeteries or relegated to a specific area (Daniell 1997, 127-8). Alternatively, it could be the remains of a clandestine burial, either of a naturally aborted foetus where there had been no opportunity to christen or possibly of a deliberately aborted foetus (Riddle 1997, 126-166; Scott 1999, 36-7; Roberts and Cox 2003, 316-8). The disarticulated bone assemblage contains a substantially higher proportion of immature individuals (65.5%) than did the articulated remains (42.2%). This may, in part, reflect a higher number of immature individuals buried in the lower (i.e. earlier) stratigraphic levels ( c . 46% of interments) than in the upper ones (c. 27%). Since the lower levels included the graves of adults this part of the cemetery had clearly not been designated just for the burial of children. However, all but one of the seven juveniles (5-12 yr.) within the articulated assemblage, for example, were recovered from 134 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 2. Summary of analysis of articulated human remains Context Grave Quantification Age/sex Pathology 105 104 c. 55% Adult c. 30-35 yr male Amtl; caries; abscesses; hypoplasia; calculus; Sch - 2T, 1L; op - T9-10 bsm, right hip; pitting - right sterno-clavicular; plastic changes - left ilium; poliomyelitis? (left) 106 104R c. 40% a.u.l. Adult c. 40-50 yr male Displaced prox. tibia/fibula joint 171 169 c. 30% s.a.u. Juvenile/ subadult c. 12-13 yr ??male Caries; hypoplasia; calculus; cribra orbitalia ; mv - wormian bone 173 172 c. 98% Juvenile c. 6-7 yr Calculus; cribra orbitalia ; periosteal new bone - tibiae; mv - ossicle at lambda, wormian bone 233 232 c. 97% Adult c. 35-45 yr Pfemale Amtl; caries; abscesses; calculus; endocranial new bone; osteoarthritis - 4 costo-vertebral; op - 2T & LI -5 bsm, left knee, both hips; pitting - IT & 2L ap, right acromio-clavicular, right hip; enth - iliac crest, calcanea, patellae; exostoses - left ulna 237 240R c. 35% Adult c. 30-35 yr male Amtl; caries; abscess; hypoplasia; calculus 241 240 c. 98% Juvenile c. 5-6 y. Caries; cribra orbitalia ; Pinfection - right ankle joint; mv - ossicle at lambda, wormian bones 248 247 c. 90% Adult c. 35-45 yr male Amtl; caries; abscesses; hypoplasia; calculus; periosteal new bone - maxilla; osteochondritis dessicans; Sch - T6-12; op - IT tp, left hip, 2 distal IP joints (hand), 3 costo-vertebral; pitting - right hip, left acromio- clavicular; enth - patellae, calcanea; mv - metopic suture, ossicle at lambda, wormian bones, SI lumbarised 251 250 c. 80% Neonate c. 1-2 weeks - 253 - c. 25% s.a.u. Infant c. 1. 5-2.5 yr - 258 257 c. 42% a.u.l. Adult c. 35-45 yr Periosteal new bone - right tibia; op - left knee, left 2nd T-MtT, right hip; enth - patellae, calcanea 270 269 c. 99% Adult c. 35-45 yr Pmale Amtl; caries; abscesses; calculus; osteoarthritis - 3 costo-vertebral, L5, right hip; periosteal new bone - manubrium , ribs, right pubic crest, left clavicle, scapula, left 5th metatarsal; fracture - radius; osteomylitis - femoral, tibiae & fibulae shafts (Psyphilis); Sch - T8-10; ddd - 3C, L5-S1; op - 1C & all T-L bsm, knees, left ankle & tarsals, left wrist, both shoulders, right elbow, both hips, 2 left distal IP joints (hand), 2 right carpals; pitting acromio- & sterno-clavicular joints, left knee, 3 costo- vertebral; enth - ulna, calcanea; mv - os acromialie 276 275 c. 99% Adult c. 18-20 yr male Caries; abscesses; calculus; Sch - T12; op - distal IP joint (hand); hyper- porosity at bregma; periosteal new bone - left tibia; surface defects - C aps, right dorsal calcaneum; mv - M3s absent, cartilaginous coalition left 3rd C-MtC 327 419 c. 47% a.u.l. Juvenile c. 5-7 yr PPmale - 340 339 c. 55% Neonate c. 2-3 weeks PPfemale 372 371 c. 85% Adult c. 30-35 yr female Amtl; caries; abscesses; calculus; hypoplasia; osteoarthritis - 12th costo- vertebral; calcified cartilage; periosteal new bone - tibiae, fibulae, left calcaneum; juxta-articular erosions - prox. humerus; op - 2C, T4-12, L3-S1 bsm, knees, ankles, 2 left tarsals, right distal IP joint (foot), left shoulder & elbow; pitting - left sterno- & acromio-clavicular; enth - distal humerus, patellae; exostoses - left calcaneum, right 1st MtT; surface defects - anterior patellae, right inferior- dorsal; mv - M3s absent, cervical rib, 4 lumbar vertebrae 5L/Slsame 407 ?=607 406 c. 30% 1. Adult >25 yr male Fracture (with ankylosis) - left fibula; surface defects - anterior patellae, dorsal calcanea; juxta-articular erosions - left 1st MtT, Left 1st MtT-P; op - knees, left Mt-P joint; enth - calcanea 453 452 c. 30% a.u.l. Neonate c. 3 months Periosteal new bone - fibulae 508 509 c. 45% a.u.l. Infant/ juvenile c. 4-6 yr 512 513 c. 28% s.a.u. Adult c. 18-23 yr female Amtl; caries; abscess; hypoplasia; calculus; periodontal disease; periosteal new bone - mandible, maxilla; osteoarthritis - T4-5; op - Cl-2; mv - ossicle at lambda, wormian bones EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 135 Table 2 continued. Summary of analysis of articulated human remains Context Grave Quantification Age/sex Pathology 523 524 c. 99% Adult c. Amtl; caries; abscesses; hypoplasia; calculus; periodontal disease; 30-35 yr male calcified cartilage; ivory osteoma; osteoarthritis - T6-8, L2-3, cost- vertebral; ddd - C3, T9-11; Sch - T4, T6-12; surface defects - left glenoid, left 4th MtT; juxta-articular erosions - MtT; op - T5-12 & SI bsm, ankles, left Mt-P joint, right shoulder, hip & knee; pitting - right hip, sterno- & acromio-clavicular; exostoses - right tibia; enth - calcanea; mv - os acromialie, spina bifida occulta 530 529 c. 97% Foetal/ neonate 40 weeks/birth 532 533 c. 98% Adult c. Amtl; caries; hypoplasia; calculus; Sch - T4-12, L4-5; op - L5-S1 bsm, 20-25 yr male hips, left knee, costo-vertebral; pitting - costo-vertebral; enth - left navicular; mv - L5 semi-sacralised, cartilaginous coalition left 3rd C-MtC 535 536 c. 30% Juvenile c. 10-12 yr Hypoplasia; calculus; cribra orbitalia; mv - wormian bone 538 539 c. 45% a.u.l. Infant c. 3-5 - - 548 c. 48% yr Juvenile c. 7-8 yr ??male Calculus; mv - metopic suture - 549 c. 12% Infant c. 4-5 Caries 558 559 c. 98% yr Adult c. Amtl; caries; abscesses; hypoplasia; calculus; cribra orbitalia ; osteoarthritis 35-45 yr male - T12, left costo- vertebral; ?healed cut - left rib; eburnation - right humerus lesser tubercle; ddd - C3-5, T12, 1L; op - hips, right ankle, shoulders, wrists, right elbow; pitting - right knee, costo- & acromio- clavicular; exostoses - right tibia; enth - right distal tibia, calcanea Amtl; caries; abscesses; calculus; calcified cartilage; surface erosion - left rib shaft; Pfracture - right fibula; periosteal new bone - right distal radius & ulna, tibiae, fibulae; osteoarthritis - 10th costo-vertebral, LI & 4; ddd - C3-7, L5-S1; Sch - T8; pitting - temporo-mandibular, left knee, hips, right MtT-P joint, right MtC-P joint; juxta-articular erosions - hands (Psome rheumatoid); new bone - prox. femora; op - knees, MtT-P joint, shoulders, elbows, hips, prox. IP joint (hand); enth - distal humerus, left prox. ulna, patellae, calcanea; exostoses - left prox. fibula; mv - metopic suture Amtl; abscess; hypoplasia; calculus; fracture - nasal bone; periosteal new bone - visceral rib; osteoarthritis - C4-5, Tl-2, right occipital condyle; ddd - C3-5, C7; op - left shoulder & costo-vertebral; pitting left costo- vertebral; mv - metopic suture Mv - spina bifida occulta Amtl; caries; abscesses; apical cysts; calculus; sinusitis (Psecondary) - left antrum; periosteal new bone - left maxilla; calcified cartilage; Sch - L3-4; op - Cl-2, T7-12 bsm, right costo-vertebral; mv - os acromialie (bi-) Fracture (with ankylosis) - left leg; osteoarthritis - 12th costo-vertebral, 35-45 yr male left wrist; periosteal new bone - left tibia & fibula; Sch - 2T, 2L; op - T9- S1 bsm, knees, left ankle & tarsals, hips, right prox. IP joint (foot), left shoulder & elbow, hands, left costo-vertebral; pitting - hips, right costo- vertebral; enth - right carpal, patellae, left fibula, calcanea, right 1st prox. phalanx (foot); surface erosions - MtT; new bone - prox. femur 602 c. 30% s.a.u. Adult c. Amtl; caries; abscesses; calculus; cribra orbitalia ; soft tissue tumour - C3; 35-45 yr osteoarthritis - 1st costo-vertebral, T3-4; op - Tl-4 bsm, left shoulder; female pitting - temporo-mandibular, left acromio-clavicular; enth - C2; mv - M3s absent 606 607 ?=407 c. 8% a.u. Adult c. Osteoarthritis - 1 l-12th costo-vertebral; op - 2Tbsm, right elbow, right 25-45 yr male MtC-P joint 616 617 c. 30% 1. Adult c. Op - left knee; enth - patellae, left tibia, calcanea; surface defect - 1st 20-35 yr male MtT-P 619 620 c. 99% Adult c. 40- Amtl; caries; abscesses; hypoplasia; calculus; calcified cartilage; hydatid 45yrfemale cyst; osteoarthritis - T4-5; Sch - T5-1 1; ddd - T7; op - hips, knees, right MtT-P, left carpal, right shoulder & elbow, right prox. IP (hand), costo-vertebral; pitting - left hip, costo-vertebral; enth - patellae, calcanea; surface defects - calcanea posterior, 1st prox. phalanges (foot); mv - metopic suture, M3s absent, occipital bunning, ossicle at lambda, wormian bones 561 562 573 581 582 588 589 591 592 c. 98% c. 25% s.a.u. c. 35% a.l. c. 50% s.a.u. c. 60% a.u.l. Adult c. 40-55 yr female Adult c. 35-50 yr female Juvenile/ subadult c. 11-14 yr Adult c. 30-40 yr female Adult c. 136 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 2 continued. Summary of analysis of articulated human remains Context Grave Quantification Age/sex Pathology 625 626 c. 98% Adult c. 40-50 yr Pfemale Amtl; caries; impaction - max. left canine; calculus; sinusitis (bi-); ivory osteoma; endosteal new bone; new bone - metatarsal articular surfaces; osteoarthritis - right costo-vertebral; op - Cl, T2-1 1, Ll-Sl bsm, knees, hips, right shoulder & elbow, left MtT-P, right sterno-clavicular, costo- vertebral; pitting - right acromio-clavicular, T4-5 ap; enth - iliac crest, right olecranon, prox. femora, calcanea, metatarsals; juxta-articular erosions - metacarpal; mv - metopic suture, wormian bones 641 642 c. 99% Adult c. 45-55 yr female Amtl; caries; abscess; calculus; osteoarthritis - 11th costo-vertebral; periosteal new bone - tibiae, fibulae; Sch - T6-12, L2; ddd - C3-6, 4T, L2; op - Cl -2, 4T & SI bsm, knees, hips, wrists, left shoulder, left MtC-P, right hand, right costo-vertebral; pitting - 1st costo-vertebral, C2 ap, right temporo-mandibular; costo- & sterno-clavicular joints, costo-vertebral; enth - distal humerus, patellae, tibiae, calcanea, right navicular & cuneiforms; mv - occipital bunning 646 647 c. 12% a.u. Adult c. 35-55 yr Pfemale Osteoarthritis - right costo-vertebral; ddd - T4-5; new bone/fusion - T4- 12; pitting - 1st costo-vertebral, right sterno-clavicular; op - left costo- vertebral, 2C, T3-L4 bsm 649 650 c. 19% 1. Anfant c. 3-4 yr - 660 661 c. 80% Auvenile c. 5.5-7 yr Caries 660 664 c. 50% Auvenile c. 8-10 yr Hypoplasia; calculus; surface defect - left talus 666 667 c. 12% 1. Adult >25 yr PPmale Op -right knee, left tarsals; surface defect - distal tibia; enth - right distal tibia & fibula, calcanea; remodelling - left calcaneum-talus 669 670 c. 23% 1. Adult c. Osteoarthritis - L4-5, hips; op - 2T & SI bsm, right knee, left elbow & 45-55 yr male wrist, prox. IP (hand); pitting - prox. IP (hand); juxta-articular erosions - hands; Perosive arthropathy - finger phalanges; enth - prox. femur, patellae 674 675 c. 25% a.u.l. Subadult c. 14-16 yr Pfemale KEY: R = disarticulated & redeposited but single individual; s.a.u.l. - skull; axial skeleton; upper limb; lower limb (only where all four skeletal areas not represented); amtl - ante mortem tooth loss; Sch - Schmorl’s node; ddd - degenerative disc disease; op - osteophytes; enth - enthesophytes; mv - morphological variation; C - cervical; T - thoracic; L - lumbar; S - sacral; bsm - body surface margins; MtT - metatarsal; MtC - metacarpal; IP - interphalangeal; tp - transverse process; ap - articular process; prox. - proximal Table 3. MNI identified from disarticulated and redeposited human remains Age categories Foetus c.19 weeks Neonate Infant Juvenile 0-3 mth 1-5 yr 5-12 yr Adult >18 yr Total MIN 1 3 10 5 10 (5 f, 5 m) 29 graves in the lower levels. Their concentration in these early stages of the area’s use for burial, since it clearly represents only part of a larger cemetery, suggests a higher child mortality at this time. This may be linked to one of the numerous epidemics experienced nationally in these years and/or other disease to which children are susceptible (Walker 1981; Roberts and Cox 2003, table 6.18, 335-7). The particularly high proportion of infants (0.5-5 year; 39%) amongst the immature individuals probably reflects the particular susceptibilities within this age group due to the weaned infant’s immune system not Table 4. Summary of age/sex of individuals identified from in situ human remains Un- Female Male Total sexed ?? ? total }} ? total Immature Foetus/neonate 40weeks/birth 1 1 Neonate 0-3 mth 3 3 Infant 1.5-5 yr 4 4 Infant/juvenile 4-6 yr 1 1 Juvenile 5-8 yr 4 - - - - 1 1 5 Juvenile 8-12 yr 2 2 Juvenile/subadult 11-14 yr 1 - - - 1 - 1 2 Subadult 14-16 yr Adult — - 1 1 — — - 1 c. 18-25 yr - - - 1 - - 2 3 c. 20-35 yr - - - - - 1 1 c. 25-45 yr 1 - 1 5 - 1 8 14 c. 35-55 yr - - 2 5 - - 2 7 > 25 yr - - - - 1 - 1 1 Totals 17 - - 12 - - 16 45 EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 137 reaching maturity until they are around six years old (Lewis, 2007 6.4). The age of the immature individuals suggested by the length of various long bones was consistantly less than that suggested by the stage of tooth development. The discrepancy increases with the age of the individual, there being little or no difference with the neonates, up to 6 months in the young infants and up to 2 years in those over 3-4 years of age. A similar discrepancy has been observed in other assemblages of this date and the observations may reflect limitations within the methodology and/or common retardation in growth associated with poor nutrition and childhood illnesses (Molleson and Cox 1993, 150; McKinley 2008, 60). Skeletal indices and traits Sexual dimorphism amongst the adults was most marked, after the pelvic traits, by body size and general robusticity. The skull traits were generally not marked, for example, none of the males have strong nuchal crests and all the mastoid process are fairly large. This dimorphism is emphasised by the stature estimates. It was possible to estimate the stature for all the females and most (15) of the males from the articulated bone assemblage. The female range of 1.44- 1.66m {c. 4’ 7”- 5’ 5”) has only the slightest overlap with that for the males of 1.63 - 1.76m (c. 5’ 4” - 5’ 9”), the averages diverging by 0.14m (female 1.57m, c. 5’ PA”; male 1.71m, c. 5’ 7V4”). The male mean is the same as the average for the period given by Roberts and Cox (2003, table 6.7) whilst the female mean falls just slightly short of the average of 1.60m. Both sets of data have their closest parallels with those from the London cemeteries in the sample ( ibid .). Cranial index was calculated for six females and four males, giving a range of 73.3 - 80.8 (dolicho- mesocranial) with a mean of 77.06 (mesocranial). Both sexes covered the full range with very close averages (76.9 female and 77.3 males). The broad variation suggests that even within this small corner of the cemetery there was no great homogeneity within the group as a whole. Non-metric morphological variations, which may indicate population diversity / homogeneity, were not routinely recorded though some are noted in Table 2. There is a high proportion of individuals with an ossicle at the lambda (five; TPR 24%) and with a metopic (unfused frontal) suture (six; TPR 28.6%). Two adult males share the rare trait of a non-metric cartilaginous coalition in the left third carpo-metacarpal joint (hand). The unusual occurrence of the latter suggests a familial relationship between there two males; though they were buried at either end of different rows within the cemetery group. The relatively high prevalence rates for the other conditions, both of which are shared by two individuals (620 and 248), at least suggests a broad degree of homogeneity between some of those buried within the part of the cemetery. Five of those with metopic sutures were recovered from the southern margins of the site within a c. 5m x 2m area, suggesting a possible ‘family plot’. Pathology Some pathological changes were observed on all except seven (immature) individuals from the in situ burial remains. A summary of the observed lesions and the bones affected is presented in Table 2. The rate of ante mortem tooth loss is higher than the c. 23% given by Roberts and Cox (2003, table 6.15) for the period but considerably lower than that observed at other early post-medieval sites (McKinley 2008, 72; Table 5). Most of the ante mortem tooth loss was linked to dental caries, the rates for which are again much higher than the average of 11.2% recorded for the period, though there were some similar rates within the sample, both London cemeteries (Roberts and Cox 2003, table 6.13). The dental abscess rates are similarly high by comparison with rates from elsewhere (average 2.2% ibid.). These figures are generally suggestive of a poor quality, non-self-cleaning diet high in carbohydrates, exacerbated by poor dental hygiene and possibly sugar consumption; though availability of the latter would generally be limited to the wealthy in the period apparently covered by the cemetery (Whittaker 1993, 52; Molleson and Cox 1993, 47; Hillson 1986, 298). The juvenile stress indicators, cribra orbitalia (pitting in the orbital roof reflective of anaemia; Roberts and Manchester 1995, 166-9) and dental hypoplasia (developmental defects in the tooth Table 5. Summary of permanent erupted dentitions by sex from in situ human remains No. denti- tions Teeth Socket positions Ante mortem tooth loss Caries Abscesses Male 8 169 245 42 30 24 (17.1%) (17.7%) (9.8%) Female 10 151 302 124 37 15 (41.0%) (24.5%) (5.0%) Total 18 320 547 166 67 39 (30.3%) (20.9%) (7.1%) 138 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Table 6. Human bone: summary of number and rates of spinal lesions from in situ remains, excluding infant/juvenile Total no. vertebrae Osteoarthritis SchmorTs nodes Degenerative disc disease Lone osteophytes Lone pitting Male 204 46% total 17 (8.3%) 38 (18.6%) 15(7.3%) 40 bsm (19.6%) 16 (7.8%) Female 224 50.6% total 16 (7.1%) 18 (8.0%) 22 (9.8%) 33 bsm (14.7%) 28 ap (12.5%) 33 (14.7%) Total 443 33 (7.4% ) 56 (12.6%) 37 (8.3%) 73 bsm (16.5%) 28 ap (6.3%) 49(11.1%) KEY: bsm - body surface margins; ap - articular process enamel reflective of illness or nutritional stress; Hillson 1979), although relatively common were only slight-moderate in severity. The overall rate of 20% for cribra orbitalia was largely reflective of the condition amongst the immature individuals (TPR 57%), both being above rates recorded elsewhere outside London (McKinley 2008, 90). Conversely, the 20% rate for hypoplasia is lower than seen within other post-medieval assemblages (ibid., 74). The figures suggest that although parts of the diet of immature individuals may have been compromised they were relatively well nourished and that those who survived into adulthood had not experienced repeated bouts of illness/malnutrition. Five adults (crude prevalence rate (CPR) 19%) had fractures. The fibulae were most commonly affected (three cases), two males showing associated ankylosis indicative of major impact on the bone. One female with a broken nose may have been the victim of domestic violence. The rates fall central to those from contemporaneous sites (McKinley 2008, 79). Infections included two cases of sinusitis (indicative of poorly ventilated housing or work- places), possible poliomyelitis and osteomylitis, the latter possibly linked to syphilis. New bone on the visceral surface of a rib is indicative of a chronic chest infection. Parasitic infection is also indicated by the presence of a hydatid cyst (indicative of tapeworm infestation). Lesions in one cervical vertebra demonstrate at least one individual suffering from a soft tissue tumour. Other lesions are generally degenerative in nature, indicative of osteoarthritis and possibly rheumatoid arthritis, and physical stress indications of age-related wear-and-tear (Table 6). Osteoarthritis in the extra-spinal joints was observed most frequently in the costo-vertebral joints (58 (30.8%) female joints and 32 (22.1%) male). No other joints were affected in the females but 15.8% (three; 8.3% of total) male hip joints had mild-moderate lesions, and one wrist joint (6.25%; 3.8% total). On a very crude level, comparative data from other post-medieval sites suggests those buried at Wilton experienced a physically less stressful lifestyle than those within the large industrial towns, but slightly more so than the trades-folk of some of the smaller towns (Brickley 2006, tables 99 and 101, 107-9; McKinley 2008, 79-85). Animal bone, by Jessica M. Grimm The preservation of animal bone was good, and 5222 bones (approximately 81% of the overall assemblage), including 2529 identified specimens, were selected for analysis, the result of which are summarised here (Table 7). Using Number of Identified Specimens (NISP), cattle bones were most common in the Early-Middle- Saxon period, followed by sheep/goat and a small proportion of pig. Sheep/goat were more numerous than cattle in the 10th-12th centuries, but the proportions were similar in the 13th- 14th centuries. The Minimum Number of Individuals (MNI) shows that sheep/goat were the most numerous animals kept in all periods, followed by cattle and pig, although beef probably remained more important in the diet than mutton or pork. Domestic fowl were also kept during all periods. Only small numbers of wild mammal species are represented, including red deer, hare, rabbit, field vole and mouse, and since only deer antlers, possibly collected, were found, hunting for venison cannot be proved. However, fish remains, including cod, eel, flatfish, mackerel and herring/sprat were recovered in moderate quantities. As most of the skeletal elements of the cattle, sheep/goat and pig are present, it appears that they were processed on or near the site, and numerous butchery marks were noted, with cleavers being the dominant tool for cattle, and knives for sheep/goat and pig. The cattle and sheep/goat assemblages comprise more kitchen waste (meat-rich parts such EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 139 Table 7. Characteristics of the animal bone assemblages by period Period No. frags Preservation index NISP Average weight (g)* Gnawed % Loose teeth % Burnt % Early-Mid- Saxon 1195 3.9 756 18.0 3.0 2.9 5.7 10th-12th centuries 3180 3.8 1286 14.1 1.7 2.7 5.0 13th-14th centuries 847 3.8 487 10.8 1.7 5.0 1.8 Key: *hand collected only; Preservation index = (preservation stage x number of bones / total number of bones in the assemblage) as upper limbs) than butchery offal (heads and feet), but all pig elements are evenly represented, perhaps indicating that they were prepared whole (i.e. roasted) rather than as cuts. The fairly equal proportions of Early-Middle Anglo-Saxon cattle killed each year after the age of two indicates the production of both meat and secondary products like milk, and their use for traction, but they appear to have been butchered older in the 10th-12th centuries, suggesting meat production was less important. In the Early-Middle Anglo-Saxon period a large proportion of the sheep/ goat were killed at an optimum age for meat, the youngest being slaughtered at 6-10 months, almost half being killed in their second year, and none living beyond 5 years. By the 10th-12th centuries, however, they are evenly distributed over the age classes, and by the 13th-14th centuries mainly older animals were present. Although the pig bone assemblage was small, it seems that in the Early-Middle, Late Anglo- Saxon and medieval periods more than half the pigs were killed before they reached two years, the optimum age for slaughter, although the presence of animals older than 3.5 years might indicate breeding on the site; there may have been only immature animals in the 13th— 14th centuries. Charred plant remains and charcoal, by Ruth Pelting Fifty-nine bulk samples were taken from features of different phases and processed for the recovery of charred, waterlogged and mineralised plant remains, and charcoal. Forty-one of the samples (approximately half from the SFB and most of the remainder from medieval pits) were selected for detailed assessment. Identifications follow the nomenclature of Stace (1997). The results of the assessment are summarised here (a full report with tabulated results is held in the archive). Early-Middle Anglo-Saxon Two samples from pre-SFB ditch 678 contained material, including grain of free-threshing wheat and barley, more indicative of Saxon or medieval activity, rather than prehistoric or Romano-British. The SFB was sampled in detail in an attempt to plot variation in occupational activities. Three samples from ‘occupation layer’ 550, and two from pits (584 and 622) produced broadly similar material, each containing between cereal grains in a density of 4.2 to 25 grains per litre dominated by free- threshing wheat and barley, with occasional weeds and hazelnut shell fragments. Occasional grains of oats were also noted and pulses were present in one floor sample. The floor samples tended to produce greater quantities of charcoal than the pit samples, with a relatively high incidence of oak. The post- hole samples were comparable to the floor deposits, although an unusual find was a single possible spelt grain, probably residual. Most of the samples from the hearths produced only occasional grain (although two from hearth 599 produced nearer 100 grains) and surprisingly small quantities of charcoal. The cereal grain was dominated by free-threshing wheat with some barley, with oats and rye present in only one sample. A single pulse and a fragment of hazelnut shell may represent cooking waste, as may two seeds of raspberry/blackberry, two elderberry seeds and occasional fish and mammal bone fragments. Late Anglo -Saxon I medieval The samples from pits 156, 289, 342 and 392 produced grains of free-threshing wheat and barley in varying proportions and occasional grains of oats and rye, with some chaff, along with weeds of arable or ruderal habitats. Large quantities of grain were recovered from a fill of pit 156. The basal fill of pit 156 produced a large waterlogged flot dominated by degraded organic matter, and a range of wild species that are typical of disturbed habitats and likely to have been growing on the site, with seeds of buttercup, sedges and common 140 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE spikerush suggestive of damp grassy conditions. Of more interest were several seeds of probable hemp, widely grown in medieval England as a fibre plant and frequently found on settlement sites (Greig 1995). The lower fill of pit 289 was rich in cereal grains but also contained a much greater number of weed seeds than any other sample on the site, as well as some chaff. The weeds included numerous seeds of stinking mayweed, a troublesome arable weed which forms seed clusters. The chaff and weeds are likely to represent impurities in largely processed grain, the higher number perhaps indicative of a grain deposit which had been less thoroughly cleaned. There was no evidence for cesspit type deposits, suggesting that these pits were backfilled with general domestic refuse. Post-medieval The primary fill of post-medieval pit 194 contained occasional waterlogged material, though preservation was poor and the only species identified were numerous seeds of raspberry/blackberry and a single seed of fig, the latter hinting at a relatively sophisticated diet with the consumption of imported fruit. Discussion The range of cultivated species is typical for a Saxon and medieval site. Only one variety of free-threshing wheat was identified, the hexaploid Triticum aestivum (bread wheat) type, and this is likely to have been the major cereal crop. Barley may also have been a food crop as there is no other evidence for fodder in the SFB, and rye and oats were rare. This pattern has been noted on a number of Saxon sites in southern Britain, such as in Southampton (Monk 1978; Hunter 2005) and the surrounding region (Carruthers 1991; Green 1991; 1994). The evidence for non-cereal crops is very limited, but the presence of hemp seeds may indicate that hemp retting for textile manufacture was taking place in the rivers and streams. While small quantities of indeterminate mineralised remains were noted in the pits, this was insufficient to indicate that the pits were used for the disposal of sewage. The deposits at Wilton are generally typical of semi-rural sites of mixed economy and lack the mineral rich cesspit-type deposits of contemporary urban centres. Discussion, by Rob De’Athe and Phil Andrews Despite a few scraps of worked flint, a Flavian coin and a possible Hod Hill brooch, no further evidence for prehistoric and early Romano-British settlement has emerged from the recent excavations, although earlier work at St John’s Hospital immediately to the north revealed traces of Romano-British agricultural activity sealed beneath the bank of the Late Anglo- Saxon defences (Andrews et al. 2000, 200). The archaeological and documentary findings reported here have, however, added significantly to our understanding of the origins and early development of the town of Wilton. It seems clear that historic settlement at 41-43 West Street began with the Early/Middle Anglo-Saxon SFB. By the late Saxon to medieval period there may have been a tenement running back from the West Street frontage to the church of St Andrew’s, Ditchampton, and at some stage in the middle ages there was probably a dwelling fronting the street, and perhaps a rectory at the back of, or beyond, the site. The church was demolished following the Dissolution, but the former churchyard continued as a burial ground into the 17th century, and the site subsequently remained largely open with any subsequent buildings relating to trade rather than dwellings. The evidence for Early/Middle Anglo-Saxon settlement is of particular significance as this had previously been indicated only from written sources. The presence of a large SFB probably indicates more extensive occupation in the valley bottom as such structures are rarely found in isolation, and are often associated with post-built halls. While none of the finds from the SFB is closely datable, a 7th-8th century date is suggested on the basis of the pottery and loomweights recovered. Possible cemeteries in the area, perhaps associated with the early settlement, may be indicated by the finds of an Anglo-Saxon hanging bowl in the grounds of Wilton House in 1860 (Wiltshire County Council 2004, 11), and the two 6th-7th century Anglo-Saxon spearheads recovered from higher ground at 22 The Hollows c. 0.5km to the north-west (Figure 1). Such a juxtaposition of settlement and cemetery in this period is not uncommon, the settlement strung out along the valley bottom close to the river, with an associated cemetery at the Hollows in a more prominent location overlooking and visible from the settlement. EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 141 The SFB is unusually large - c. 7.5m long by at least 6m wide. A SFB at Upton in Surrey measured 9.1m by 5.5m and is dated to 6th-7th century, while another, at Old Swindon, measured 8.5m by 4.5m and is dated to the 7th-8th century (Tipper 2006, 66). The Wilton SFB appears to confirm the general trend, noted by Tipper (2006), towards larger SFBs from the 7th century onwards, but none of those in the three largest excavated groups in Britain- Mucking, Essex; West Heslerton, North Yorkshire; and West Stow, Suffolk - is as large (ibid., 64-65). Despite its size, however, the SFB may have been a simple, two- post type, although the details of its construction are uncertain. The presence of two hearths indicates that the floor was on the base of the hollow, rather than raised on planks above it. The paucity of pottery, both in the occupation layer and the upper, dumped layer, is not unusual (ibid.), and the structure may not have been a domestic dwelling, but a large ancillary building, possibly related to weaving. The duration of the Early/Middle Anglo-Saxon settlement in the vicinity is unknown, but it is reasonable to assume that there was continuity of occupation in the general area. Chandler has suggested that North Street and South Street follow an early route between crossings of the Rivers Wylye and Nadder, which became the focus of settlement in the Middle to Late Anglo-Saxon period, and that the sub-rectangular area to the east of the route, bounded by Russell Street to the north-east and Kingsbury Square to the south, was the site of a royal compound (Chandler 2007, 56-7). This ‘kingsbury’, which may have contained a ‘palace’ or large timber hall, a church and other accommodation, is likely to have formed the focus of the Late Anglo-Saxon settlement. Certainly, by the early 9th century, Wilton had become an important place within the West Saxon kingdom, and in the late 9th century became a key site in the chain of King Alfred’s fortified burhs. West Street may already have been a broad thoroughfare in the Late Anglo-Saxon period, perhaps accommodating an early market before this was established later in the former ‘kingsbury’ area (Chandler 2007, 59-60). Fronting onto West Street, the site would have lain just inside the gateway through the burh’s north-western defences, which ran between the Rivers Nadder and Wylye. The location of St John’s Hospital, although a medieval foundation, is reminiscent of the Late Anglo- Saxon arrangement of ‘gate churches’ at Wareham, Cricklade and elsewhere; there was also a church over the east gate at Old Sarum. Netherwell Lane (possibly ‘the lane beneath the wall’) may have developed as an intra-mural lane around this time, but there was no evidence from 35 West Street that the defensive circuit continued along the south-west side of Wilton. It is possible that the excavation exposed parts of two Late Anglo-Saxon burgage plots, each up to 55m long and perhaps c. 8.5m wide. However, their West Street frontages lay beyond the limit of excavation and no associated pre-Conquest structural remains were identified, although a number of rubbish pits and rear boundary ditches span the 10th-12th centuries. Pit digging in the backyards of the two properties continued throughout the 13th and 14th centuries when there was possibly a dwelling on the street frontage in one property, and there may have been a rectory house associated with St Andrew’s church near the rear of the site, presumably north- west of the church. The range of finds recovered from pits of all periods is limited, comprising almost exclusively domestic refuse, although a small quantity of slag, mainly from medieval and later features, indicates iron smithing, and crucible fragments from one of the test-pits at 35 West Street suggest copper alloy working in the vicinity. There is also circumstantial evidence for the retting of flax in nearby streams. The variety of plant and faunal remains seem to reflect increasing prosperity of the town in the 10th-12th centuries, with the inhabitants partly growing their own produce but also probably procuring animal products from the market. The finding of structural remains and graves at the rear of the site (and at 35 West Street) represents the rediscovery of St Andrew’s church, Ditchampton, one of the ‘lost’ churches of Wilton, previously thought to lie further to the north-east. It may have been no more than 20m long and occupied a relatively small property lying perpendicular to Netherwell Lane, to the rear of the West Street tenements and on the edge of the Nadder floodplain. The date of its foundation is unknown, but a slight misalignment between wall 235, recorded in the excavation, and the short section of wall observed 15m to the south-west in 2007, hints that the wall might have been starting to curve to the south, perhaps forming an apsidal south-east end, raising the possibility of an Anglo-Saxon date, although a Norman date is equally likely. A Late Anglo-Saxon/ earlier medieval pit (411) which underlay part of the wall supports an 11th or 12th century date - at least for this wall. If an early church did stand on this site, it is likely that the Late Anglo-Saxon defences isolated it from its parish and settlement 142 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE of Ditchampton, which developed (like others in the area) as a suburb of Wilton. Following the church’s demolition, probably in the second half of the 16th century, the churchyard continued to be used for burial, possibly extending beyond its original boundaries, perhaps until c. 1630 when it became a garden. The absence of coffin fixtures and fittings suggests that burial had ended before the 18th century. The human remains are of particular interest as there are few such assemblages securely dated to the 16th or 17th centuries (Church of England and English Heritage 2005, 47). Those buried in the St Andrew’s burial ground appear to lie, therefore, on the cusp between rural and urban life, and probably include tradespeople, artisans and rural labourers, neither particularly prosperous, nor overly poor or deprived. Acknowledgements Wessex Archaeology would like to acknowledge Archstone Lifestyle Homes Ltd, in particular Alan Giles and Mick Parret, for commissioning and funding this work. Thanks are also due to Helena Cave-Penny, formerly Archaeological Officer for Wiltshire County Council, for her assistance and advice during the fieldwork. The archaeological fieldwork at 41-43 West Street was directed by Rob De’Athe and managed by Caroline Budd. The test-pits at 35 West Street were undertaken by Caroline Budd, with the subsequent excavation and watching brief directed by Dave Godden and managed by Brigitte Buss. Hannah Spieler assisted with the recording of the human bone, and Chloe Hunnisett analysed the burial sequence. Jorn Schuster and Kenneth Penn have kindly commented on the two spearheads, the significance of which was first recognised by John Lewis (Wessex Archaeology) when shown them by his neighbour, and we are grateful to Mr Whatley for allowing them to be brought to the offices of Wessex Archaeology for recording. Documentary research was conducted in the Wiltshire and Swindon History Centre, Chippenham, and John Chandler is indebted to the Centre’s staff for assistance in this undertaking. We would also like to acknowledge the Earl of Pembroke for permitting reproduction of several historic maps held in his archive. We are grateful to Chris Roussell for making available his unpublished research on the ancient churches and religious establishments of Wilton. Finally, we would like to thank Andrew Powell for his editorial work, and John Chandler for the benefit of his knowledge of Wilton and for suggesting a number of improvements to an earlier draft of this report. The project archive (Wessex Archaeology project code 60517) will be deposited at Wiltshire Heritage Museum. References ANDREWS, E, MEPHAM, L. and SEAGER SMITH, R. 2000. Excavations in Wilton, 1995-6: St John’s Hospital and South Street. WANHM , 181-204 ANGLO-SAXON CHARTERS, www.anglo-saxons.net ANON. 1971. WANHM 66, 191 ANON. 1972. WANHM 67, 175-6 ATKINSON, D. R. 1970. Clay tobacco pipes and pipemakers of Salisbury, Wiltshire. WANHM 65, 177-89 BRICKLEY, M., 2006, ‘The People: Physical Anthropology’, in M. Brickley and S. Buteux, St. Martin’s Uncovered: investigations in the Churchyard of St. Martin’ s-in-the-Bull- Ring, Birmingham, 2001, 90-151. Oxford: Oxbow CARRUTHERS, W., 1991, ‘The plant remains’ [The Abbots Worthy Settlement], in E J. Fasham and R. J. B. Whinney (eds). Archaeology and the M3, 61 -IS. Winchester: Trust for Wessex Archaeology CHANDLER, J. 2007. Deconstructing Wilton. Sarum Chronicle 7, 56-63 CHURCH OF ENGLAND and ENGLISH HERITAGE, 2005, Guidance for Best Practice for Treatment of Human Remains Excavated from Christian Burial Grounds in England. Swindon: English Heritage COWIE, R. and BLACKMORE, L, 2008, Early and Middle Saxon Rural Settlement in the London Region. London: Museum of London Archaeology Service Monograph 41 COX, E W. and MILLS, J. M., 1991, ‘Kimmeridge shale’ in E W. Cox and C. M. Hearne, Redeemed from the Heath: the archaeology of the Wytch Farm Oilfield (1987- 90), 170-5. Dorchester: Dorset Natural History and Archaeology Society Monograph 9 DANIELL, C., 1997 Death and Burial in Medieval England 1066-1550. London: Routledge FARWELL, D. E., ANDREWS, E and BROOK, R., 1999, Prehistoric, Roman and Early Saxon Settlement at Prospect Park, London Borough of Hillingdon. Salisbury: Wessex Archaeology GREEN, F. J., 1991, ‘Landscape archaeology in Hampshire: the Saxon plant remains’, in J. M. Renfrew (ed.) New Light on Early Farming: recent developments in palaeoethnobotany, 363-77 . Edinburgh: Edinburgh University Eress GREEN, F. J., 1994, ‘Cereals and plant food: a reassessment of the Saxon economic evidence from Wessex’, in J. Rackham (ed.), Environment and Economy in Anglo EARLY TO MIDDLE ANGLO-SAXON SETTLEMENT, WILTON 143 Saxon England, 83-8. London: CBA Research Report 89 GREIG, J. 1995. Archaeobotanical and historical records compared: a new look at the taphonomy of edible and other useful plants from the 1 1th to the 18th centuries AD. Circaea 12(2), 21 1-47 HAMEROW, H., 1993, Excavations at Mucking. Volume 2: the Anglo-Saxon settlement. London: English Heritage Archaeological Report 21 HARE, J. N. 1991. The growth of the roof tile industry in later medieval Wessex. Medieval Archaeology 35, 86-103 HASLAM, J. (ed.), 1984, Anglo-Saxon Towns in Southern England. Chichester: Phillimore HILLSON, S., 1986, Teeth. Cambridge Manuals in Archaeology, Cambridge: Cambridge University Press HOLDEN, E. W. 1976. Excavations at Old Erringham, Shoreham, West Sussex. Part I: a Saxon weaving hut. Sussex Archaeological Collections 114, 306-21 HUNTER, K. L., 2005, ‘Charred plant remains’, in V. Birbeck with R. J. C. Smith, P Andrews and N. Stoodley, The Origins of Mid-Saxon Southampton: excavations at the Friends Provident St Mary’s Stadium 1998- 2000. Salisbury: Wessex Archaeology HURST, J. G., 1959, ‘Middle Saxon pottery’, in G. C. Dunning, J. G. Hurst, J. N. L. Myres and F. Tischler, Anglo-Saxon pottery: a symposium. Medieval Archaeology 3, 13-31 JACKSON, D. A., HARDING, D. W. and MYRES, J. N. L. 1969. The Iron Age and Anglo-Saxon site at Upton, Northants . Antiquaries Journal 49, 202-21 JAMES, M. K., 1962, ‘Wilton’, Victoria History of the Counties of England: Wiltshire 6, 1-37. London: Oxford University Press KERRIDGE, E., 1953, Surveys of the Manors of Philip, Earl of Pembroke and Montgomery, 1631-2, Devizes: Wiltshire Record Society Volume 9 LEWIS, M. E., 2007, Children in Bioarchaeology: the analysis of non-adult skeletal remains in biological and forensic anthropology. Cambridge, Cambridge University Press LOADER, E., 2000, ‘The ceramic building materials’, in M. Rawlings, Excavations at Ivy Street and Brown Street, Salisbury, 1994, WANHM 93, 37-8 LONDON, V. C. M., 1979, The Cartulary of Bradenstoke Priory, Wilts Record Society 35 McKINLEY, J. I., 2004, ‘Compiling a skeletal inventory: disarticulated and co-mingled remains’, in M. Brickley, and J. I. McKinley (eds), Guidelines to the Standards for Recording Human Remains, 13- lb. Reading: British Association for Biological Anthropology and Osteoarchaeology and Institute for Field Archaeology McKINLEY, J. I., 2008, The 18th Century Baptist Chapel and Burial Ground at West Butts Street, Poole. Salisbury: Wessex Archaeology McMAHON, P, 2004, The Archaeology of Wiltshire’s Towns. An Extensive Urban Survey: Wilton. Trowbridge: Wiltshire County Council MEPHAM, L., 1993, ‘Pottery’, in A. H. Graham and S. M. Davies, Excavations in Trowbridge, Wiltshire 1977 and 1986-1988, 101-114. Salisbury: Wessex Archaeology Report 2 MEPHAM, L., 2000a, ‘Pottery’, in M. Rawlings, Excavations at Ivy Street and Brown Street, Salisbury, 1994, WANHM 93, 29-37 MEPHAM, L., 2000b, ‘Pottery’, in C. J. Young, Excavations at Carisbrooke Castle, Isle of Wight, 1921-1996, 98-131. Salisbury: Wessex Archaeology Report 18 MEPHAM, L., 2005, ‘Pottery’, in B. Barber, The development of Trinity Chequer: excavations at the Anchor Brewery site, Gigant Street, Salisbury, WANHM 98, 183-8 MEPHAM, L. and BROWN, L. 2007. The Broughton to Timsbury Pipeline, part 1: A Late Saxon pottery kiln and the production centre at Michelmersh, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 62, 35-68 MOLLESON, T. and COX, M., 1993, The Spitalfields Project. Volume 2: the anthropology. The Middling Sort. York: CBA Research Report 86 MONK, M. A., 1978, The Plant Economy and Agriculture of the Anglo-Saxons in Southern Britain with particular reference to the Mart Settlements at Southampton and Winchester. Unpublished University of Southampton MPhil thesis MORRIS, E., 1994, The Analysis of Pottery. Salisbury: Wessex Archaeology Guideline 4 MUSTY, J. and ALGAR, D., 1986, ‘Pottery’ in J. Musty and D. Algar, Excavations at the deserted medieval village of Gomeldon, near Salisbury, WANHM 80, 154-66 MUSTY, J., ALGAR, D. J. and EWENCE, P F. 1969. The medieval pottery kilns at Laverstock, near Salisbury, Wiltshire. Archaeologia 102, 83-150 MUSTY, J., with ALGAR, D., GERRARD, C. and HADLEY, J., 2001, ‘Pottery, tile and brick’ in P. Saunders (ed.), Salisbury and South Wiltshire Museum Medieval Catalogue Part 3 , 132-212. Salisbury, Salisbury & South Wiltshire Museum OXFORD DICTIONARY OF NATIONAL BIOGRAPHY, www.oxforddnb.com POWELL, A. B., CHANDLER, J., GODDEN, D., MEPHAM, L., STEVENS, C and KNIGHT, S. 2009. Evidence for Late Saxon and medieval occupation near Salisbury Street, Amesbury. WANHM 102, 188-201 RAHTZ, P, 1979, The Saxon and Medieval Palaces at Cheddar. Oxford: British Archaeological Reports 65 RIDDLE, J. M., 1997, Eve’s Herbs: a history of contraception and abortion in the West, London: Harvard University Press ROBERTS, C. and MANCHESTER, K., 1997, The Archaeology of Disease. Stroud: Sutton ROBERTS, C. and COX, M., 2003, Health and Disease in Britain from Prehistory to the Present Day. Stroud: Sutton SCOTT, E., 1999, The Archaeology of Infancy and Infant 144 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Death. Oxford: British Archaeological Reports International Series 819 SMITH, R. W., 1997, Excavations at Emwell Street, Warminster: the Early Economy and Environment of a Wiltshire Market Town. Salisbury: Wessex Archaeology STACE, C., 1997, New Flora of the British Isles (2nd edition). Cambridge: Cambridge University Press STONE, J. F. S. and CHARLTON, J. 1935. Trial excavations in the east suburb of Old Sarum. Antiquaries Journal 15, 174-92 STRATON, C. R. (ed.), 1909 , Surveys of the Lands of William, First Earl of Pembroke. Roxburghe Club (2 volumes) SWANTON. M. J., 1973, The Spearheads of the Anglo- Saxon Settlements. London: The Royal Archaeological Institute TIPPER. J., 2006, The Grubenhaus in Anglo-Saxon England: an analysis and interpretation of the evidence from a most distinctive building type. Yedingham: Landscape Research Centre Archaeological Monograph Series VINCE, A. G., 1981, ‘The medieval pottery industry in southern England: 10th to 13th centuries’, in H. Howard, and E. Morris (eds), Production and Distribution: a ceramic viewpoint , 309-22. Oxford: British Archaeological Reports International Series 120 WALKER, J. B. 1981. The children in the cemetery: child mortality and public health in Lyme Regis 1856 to 1979. Proceedings of the Dorset Natural History and Archaeological Society 103, 5-12 WESSEX ARCHAEOLOGY, 2000, The Bakehouse, Old Sarum Castle, Salisbury, Wiltshire, unpublished client report 47666.02 WESSEX ARCHAEOLOGY, 2006, 35 West Street, Wilton, Wiltshire: archaeological evaluation, unpublished client report 60510.02 WESSEX ARCHAEOLOGY, 2007a, 35 West Street, Wilton, Wiltshire: archaeological excavation report, unpublished client report 60511.02 WESSEX ARCHAEOLOGY, 2007b, Wilton Autos, Wilton, Wiltshire: archaeological evaluation report, unpublished client report 60513.03 WESSEX ARCHAEOLOGY, 2007c, Wilton Autos, Wilton, Wiltshire. Written Scheme of Investigation: method statement for an archaeological excavation, unpublished client report 60515.01 WESSEX ARCHAEOLOGY, 2008a, Wilton Autos, 41-43 West Street, Wilton, Wiltshire: post-excavation assessment and proposals for analysis and publication, unpublished client report 60517.01 WESSEX ARCHAEOLOGY, 2008b, Wilton Autos, Wilton, Wiltshire: archaeological watching brief report, unpublished client report 60516.03 WHITTAKER, D., 1993, ‘Oral health’, in Molleson and Cox 1993, 49-65 WSA 2057/S3: Wilton estate survey, 1565-6, in C.R. Straton (ed.) 1909, but with manuscript additions not in published edition, Wiltshire & Swindon Archives, Chippenham WSA EA 179: Wilton enclosure award, 1860, Wiltshire & Swindon Archives, Chippenham Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 145-156 The Romanesque doorway at Little Langford by Rita Wood The doorway has the only surviving sculpture from the village church of Little Langford of the first half of the 12th century. The carving in the tympanum and lintel is still clear, but the capitals are worn, broken and coated by lichens, making their subjects uncertain, though it is possible to offer a reconstruction of their original appearance, and hence also of their content. The teaching displayed on the tympanum, lintel and capitals is linked to their relative position on the doorway. Similarly, the content of other sculpture discussed is related to its place in the building. The church at Little Langford Little Langford is a small settlement about nine miles north-west of Salisbury, in the green Wylye valley. Served now only by back roads and a footpath, the church may yet be glimpsed from the Salisbury to Warminster and Bristol railway which passes a field’s width to the south: a tiny isolated building with chequered walls and a round-headed doorway. This is now the only significant remnant of the twelfth- century building (Figure 1). The railway followed the course of the turnpike road and the whole village, including the church, was rebuilt about 1864 ( VCH Wilts, 15, 179-80, 182-3). In Domesday Book, the two manors in Little Langford belonged to Wilton Abbey and Glastonbury Abbey; both were held by Edward of Salisbury, then later by his son Walter, d. 1 147, and his descendants {VCH Wilts , 15, 179-80). The presence of billet moulding and the use of a number of motifs to compile an individual and coherent programme suggest a date for the doorway in the second quarter of the 12th century. The font looks 13th-century, so perhaps in our period the villagers went for baptism to Steeple Langford where there is a good square Romanesque font on five pillars but, even if it lacked its own rights of baptism, Little Langford church was given an elaborate doorway with a teaching programme extending over tympanum, lintel and capitals. The tympanum and lintel have been discussed in this journal (Powell 1909); the tympanum, lintel and capitals are illustrated by Charles Keyser (1927, pi. 147); the photograph (WAM 35 (1908), opp. p. 387), looks to be the same one as used by Keyser, but unfortunately none of these sources supply details of the decayed capitals. The tympanum and lintel The stone used is a greyish white limestone, now crusted with lichens of very similar colour and texture. The carving in the tympanum and lintel is flat, sharp-edged and largely on two levels, but the capitals are cylindrical and their carving is more three-dimensional due partly to interlacements among the motifs. The teaching represented by the sculpture on the tympanum and lintel has already been discussed at length by the author in other journals, and the particulars for these areas will be only briefly repeated here. The imagery of the boarhunt, very much in the form seen in the lintel (Figure 2), is widespread 25 St Andrewgate, York, YOl 7BR 146 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 1 Little Langford (Wilts): the capitals, lintel and tympanum of the south doorway. All figures by the author unless noted in English Romanesque sculpture (Figure 3). The boarhunt can be interpreted as the struggle of Christian believers against the Devil; this is an allegory which has only been noted in sculpture at English village churches (Wood 2004, 89-94). The inspiration for linking the boar with the devil comes from Psalm 80:8-13 (RV; Vulgate, psalm 79). The repetition of these motifs over a wide area and without documented connections denies the possibility of a local interpretation of the kind put forward by Powell (1909). The Old Testament psalmist pictured Israel as the Lord’s vineyard: medieval exegetes commonly interpreted Israel and the vineyard as the Church, which had replaced the inattentive Jews as the people of God (Matt. 21:33-43). The psalmist laments that the fruitful vineyard is ravaged by ‘the boar out of the wood’: for medieval interpreters the passage pictures the threat to believers from the destructive activity of the Devil within the Church. In the lintel (Figure 2iii), the huntsman represents the parish priest or, in a wider sense, he may personify the institutional Church; he blows his horn (or preaches) to direct the dogs, the parishioners, in their effort to control the boar, the Devil. In an actual boarhunt, dogs followed the boar, barking, tiring it, and bringing it to bay, but only at their peril did individual dogs try to bring it down themselves or approach near it - dogs were too lightweight to slow its progress, and their teeth could not hurt it. The boar was only defeated by united opposition and by a greater force than dogs possess. Four examples in Figure 2 show dogs literally broken by going too close to the boar, and there are two examples of a dog being trampled: it is not clear what will happen to the third dog on the Little Langford lintel, but it is in a dangerous position. The message of all the boarhunt carvings is that the corporate activity of the Church, such as its organised liturgies, traditions and experience, save the faithful from spiritual dangers, whereas the disobedient who trust their own strength perish. THE ROMANESQUE DOORWAY AT LITTLE LANGFORD 147 (i) Brayton, Yorks (v) St Marychurch, Devon 33^ (vi) Tutbury, Staffs Fig. 2 The seven examples in England illustrating the boarhunt. Diagrams are intended to make comparisons easy and cannot record all details that might be seen on-site. They are prepared from photographs and observation The example at St Marychurch (Figure 2v) depicts the conclusion of the allegory, it shows the master of the hunt arriving on horseback to finish off the boar (literally, to butcher it with a cleaver); this arrival must symbolise the Second Coming of Christ, with Judgement to follow. The tympanum above the boarhunt (Figures 1 and 2iii) shows a scene in heaven, which is indicated by the block of star pattern on the left (Wood 2001, 9, 12-14; Wood 2004, 94-96). There is a tree with three birds on the right, while in the centre under an archway a bishop in a rich chasuble and a stole is giving a blessing. The scene follows on from what is shown in the lintel, that is, the tympanum symbolises the resurrection of believers into the after-life: the bishop under the archway is St Peter at the gate of heaven; the three birds perched in the tree are believers in paradise. A similar tree is carved on the tympanum at Stoke-sub-Hamdon (Somerset) where it is one of several motifs appropriate to heaven. A similar transformation from earthly to heavenly life occurs within the boarhunt scheme on the chancel arch at Liverton. Figure 2iv shows one of a pair of capitals with, on its left, the huntsman with two obedient dogs. Opposite this capital is the one shown in Figure 4, with a crane between a lion and an animal which is a combination of bird and lion: 148 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 4 Liverton (Yorks): chancel arch capital these three creatures can be interpreted as a priest and two parishioners in heaven (Wood 2006b, 125). A third dog at Liverton (Figure 2iv, right) is trampled by the boar: it looks as though the dog on the right at Little Langford is too close to the tusks of the boar and is similarly fated, in which case the three birds in the tree on the tympanum would symbolise the two dogs that chase the boar (obedient parishioners) together with the huntsman (their priest). There can be little doubt that the lintel at Little Langford depicts the earthly battle of believers with the Devil, and that the tympanum shows its happy outcome in heaven. The capitals of the doorway The chief concern of this present paper is to clarify the subjects on the capitals of the doorway (Figures 5 and 11). As interpreted above, the imagery of the tympanum and lintel uses a simple vertical zoning with heavenly things shown above the earthly, an arrangement which is very common in the medieval period, and which suggests that the capitals, like the lintel, probably concern man’s earthly life and have a moral message. Other doorways with an elaborated tympanum and lintel sometimes have only foliage on the capitals, or the capitals may not have survived, so it would be interesting to be able to analyse this extended and original scheme. Unfortunately, after making a careful examination Fig. 5 Little Langford: right capital of doorway of the sculpture, it was found that, due to erosion, breakage and the coating of lichen, it was not clear from the remains precisely what imagery the left capital had contained. This uncertainty made it doubtful that a reliable interpretation could be offered, however, because the other units of the doorway, including the right capital, are intelligible, and because comparisons have been found elsewhere for individual details of the left capital, it was decided to attempt an interpretation. Right capital: trousers The right capital has sufficient legible sculpture remaining to show that it contained a symmetrical composition comprising the figure of a seated man on the angle; his raised hands grip two serpents just behind their heads, while their tails are wound round his legs and hold them to the sides (Figures 5 and 6). The sculptor was not used to the degree of three-dimensional representation necessary for this motif, and for this reason the body of the snake, from some angles, is not quite continuous, but the legs and feet are logically correct, and they make clear what was intended. The man’s face is not well preserved apart from indications of the eyes; there are the remains of his beard on the right (south) face of the capital, so he is probably a layman. He is wearing not the usual skirted tunic but a one-piece garment that covers him from chest to thigh and has sleeves. Since the same surface treatment is used on the chasuble worn by the bishop on the tympanum, the diagonal reeding probably represents a woven fabric, one with a twill weave. It is perhaps underwear that THE ROMANESQUE DOORWAY AT LITTLE LANGFORD 149 Fig. 6 Little Langford: right capital, reconstruction over Figure 5 the man is wearing, for it can still be seen that the sleeves were ribbed in rings around the arm, and in illuminations such sleeves sometimes emerge from below overgarments (for example, Zarnecki 1972, fig. 79). The legs of the man’s costume resemble the loose trousers known as braies, an outer garment before the Conquest but one which was later shortened and worn as underclothing below the tunic. A manuscript example illustrated by Willett and Phillis Cunnington (1969, fig. 8e, 31) shows ‘Job in sickness’ sitting on his ash-tip (Job 2: 7-10) and reduced to wearing only loose calf-length trousers, presumably an undergarment. The manuscript (BL MS Roy. C. VI, fol. 6) is currently given an early twelfth-century date by the British Library, which makes the wearing of braies as an undergarment a little earlier than supposed by the Cunningtons, and brings the garment within the period likely for the carving. The stylisation of fabric by parallel reeding is widespread, and common in sculpture of the Herefordshire School. No doubt different fabrics were being represented: chain-mail; pleated, knitted, quilted and woven fabrics; and stylised drapery have all been suggested. A reeded one-piece garment similar to that worn on the capital at Little Langford is carved at Leominster Priory (Herefordshire; Thurlby 1999, 135-40), where a capital of the west doorway shows a pair of agricultural workers in Fig. 7 Leominster priory ( Herefs ): capital outside the west doorway a vineyard (Figure 7). Their clothing is unlike the ‘knightly trousers’ worn by most men in Herefordshire School carvings (Thurlby 1999, figs. 102, 162). The men on this capital are not engaged in a knightly activity; they hold pruning hooks and, with one foot on the horizontally-trained woody branch and the left hand pulling the soft spiralling shoot taut, they are about to cut rampant summer growth hard back to where it is to bud and flower in spring. Despite the presence of the large leaves, which are only there to emphasize that the shoot is too vigorous, this is a task undertaken in a vineyard in winter. The realistic subject is curious, placed as it is among standard Romanesque imagery of symmetrical birds and serpents in paradisal foliage, and opposite what inevitably becomes the chief focus of interest, the big, alert, lions. There are rich star, beading and cable patterns, and foliage, everywhere - all these motifs are proper to picture heaven and express its abundance and beauty, apart from the two men working in the vineyard. It is to be expected at the main entrance of a monastic church that the attractions of heaven should be displayed, but not this down-to-earth scene of men labouring. The west doorways of monastic churches are usually full of patterns and foliage, but there is a lurking impish face inserted among the bases at the foot of the doorway from cloister to nave at Nun 150 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Monkton Priory, Yorkshire - a caution for careless nuns. Perhaps the labourers too are some sort of warning to those who seek to enter the church thoughtlessly. At Leominster, one of the agricultural workers makes eye contact across the entrance with one of the large lions, and, approaching from the west, the visitor encounters the gaze of one worker and the other large lion. We are surely being challenged to see the human workman as ourself, and the lion as Christ. Is the parable of the labourers in the vineyard (Matthew 20:1-16) being evoked, at least to the extent of thinking of the reward due at the end of the day? A further curiosity, the men’s clothing, if it is underwear, is out of place in a winter scene, it too would have drawn the viewers’ attention, further stimulated their intelligent questioning, and alerted them to the presence of a moral lesson in that particular carving. Unusually, the doorway at Leominster Priory has more carving on two pairs of capitals inside and, if we can trust the entrance as we see it to have been designed as a whole (and all its carvings are certainly contemporary), then the inner capitals can suggest an answer. Changes in the architecture of the nave and west tower at Leominster priory affected the upper parts of the walls and were probably due to a change in the method of vaulting (Clapham 1934, 111-12). The discrepancies in the abaci throughout the west doorway noticed by later writers (Thurlby 1999, 135-140; Pawsey 2001, 40, n. 45) might be due to those changes in the upper architecture, or perhaps to errors in the order of placing the various stones comprising the imposts. In either case, the sculptural scheme at the level of the capitals is unlikely to have been affected. The doorway appears to have been designed with sculpture at this level inside and out from the beginning. The first pair of capitals inside the west front includes vines bearing fruit and only small leaves. In Figure 8, the branches emerge from the mouth of a man on the angle of the capital: he is a ‘green man’, that is, a man resurrected in paradise (Wood 2000b, 8-13). On the opposite capital, fruiting branches emerge from a bestial mask, symbolising life out of death. The sequence experienced on entering the church through the west doorway could therefore illustrate the teaching that pruning or self-control exercised in earthly life (outside the church) is followed by fruition or reward in heaven (in the putative paradise of the nave); the likening of earthly life to winter and the resurrection to springtime was a comparison often made. In a similar vein, Augustine of Hippo discussed the passage on the cleansing or Fig. 8 Leominster priory (Herefs): capital inside the west doorway pruning of the vine (John 15: 1-3) in Contra Faustum (XXII. 20, 28), saying: ‘The great husbandman of the vine [God] uses his pruning-hook differently in the fruitful and in the unfruitful branches; yet he spares neither good nor bad, pruning one and cutting off the other. . . A man, therefore, who acts in obedience to the faith. . . restrains all mortal affections, and keeps them within the natural limit’. In the next section it will be suggested that the capital at Little Langford also represents self-control, but by a different metaphor, not by the cutting away of excessive passions, but by their reining-in. Right capital: snakes The symmetrical composition of a man between two snakes is relatively common on the continent but seldom found in England. However, at Rock, Worcestershire, another site associated with the Herefordshire School, a naked squatting man is accompanied by two snakes that weave in and out of his body and have their mouths to his ears (Figure 9); this carving faces into the nave and would have been intended for the laity to see. A similar composition nearby presents the same topic in a different, remarkably realistic, manner (Figure 10); this carving faces into the chancel and was therefore intended for a clerical audience (Wood 2005, 70-71); Eileen Hamer has suggested that Rock THE ROMANESQUE DOORWAY AT LITTLE LANGFORD 151 Fig. 9 Rock (Worcs): capital on nave side of chancel arch ! may have been a minster church, that is, one with a community of secular priests (Hamer 1992). The capital facing into the chancel shows a man wearing a combination undergarment comparable to that worn at Little Langford and Leominster, though here it is closer-fitting, more tailored (Figures 5, 6 and 7); he is gripping his legs at the knee like the man on the nave side of the chancel arch (Figure 9). He is flanked by two men’s heads which are turned as if speaking to him: it can be inferred that the two men are authorities on the spiritual life encouraging him to exert self-control; the two counsellors might be, for example, St Paul and St Augustine of Hippo. One important Pauline passage in the New Testament on controlling selfish desire is Galatians 5:16-24; this passage is one of the fullest statements of the matter (Barclay 1962, 121-7). Self-control was essential to any successful monastic community, and in our period the restraint of wanton aggression, for example, was enjoined on the laity and is sometimes represented in sculpture. The use of direct realistic imagery for the admonition of the clergy is also found at Kirkburn (Yorkshire), in a corbel on the chancel: a frontal man in a long gown appears to be standing, his legs widely planted, but he is astride a horse whose two tiny front legs survive; the man was probably pulling the reins, but his lower arms and the horse’s head and neck, which would have simulated his penis, are broken off (Wood, 2003, pi. 14). At Barton-le- Street, Yorkshire, a layman controls his ‘horse’ in a similar way (Wood, 2009, 209, fig. 6). In parallel with this Fig. 10 Rock: capital on chancel side of chancel arch interpretation, the snakes on the nave capital at Rock (Figure 9) can be understood as remembered spiritual advice coming from within the man himself, encouraging him in his effort to restrain his lower nature. It is suggested that the two hands gripping and pulling on the men’s legs at Rock, or the coiling of the two snakes at Little Langford, and elsewhere the pulling of a man’s forked beard, were all intended to picture the reining-in of galloping desire. The imagery is widespread - an example of c. 1200 in Cologne is much more sophisticated than the example at Little Langford, but the formation of the image is the same: the snakes pull the ‘reins’ of his forked beard and the man pulls the snakes. At both Rock and Cologne, the man’s biceps are exaggerated to show the great effort that is necessary to attain self-control. (Carlsson 1976, 90-95; Wood 2006a, 137; Wood 2011, 150-152, fig. 16). It is evidently sexual restraint that is implied in the nave capital at Rock (Figure 9), but other impulses of the flesh are just as destructive of the spiritual life, and to be controlled by a believer. As listed in Galatians 5:19-21, these sins include enmity, strife, anger, ambition and jealousy. The loose trousers mentioned above as worn by Job sitting on his midden were worn in an actual, literal, situation, but the sculptural examples just discussed suggest that undergarments might have signalled a symbolic mode of thought, that they indicate a concern with some unseen, personal or interior matter. The symbolism of underwear need not refer only to sexual passions, but to all self-centred desires 152 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Figs. 11a (left) and lib (right) Little Langford: two views of the left capital, south face and east face that conflict with the life of the spirit. The snakes speaking into the man’s ears (Figures 5, 6 and 9) might, for example, represent his memories of church teaching, the voice of his conscience or of his eternal soul. St Paul wrote of the continuous struggle going on within a man between God’s Spirit and the desires of the lower nature (Galatians 5:16-18) or, put more simply, between the flesh and the spirit, or the body and the soul (Barclay 1962, 9-16). Augustine preached about it (Hill 1992, sermons 156.9, 163, 163A); medieval writers composed meditative dialogues between the soul and the body (for example, De Arrha Animae of Hugh of St Victor, d. 1141). That the layman should recognise the existence of this continuous debate or struggle going on within himself was fundamental to moral teaching, and sculptural imagery was therefore contrived to express it, as in the present examples (Figures 5 and 9) in which the physical human body in a symbolic mode is intimately combined with subtle, elusive, spiritual thought represented by the snakes. The perhaps unexpected employment of snakes as a positive image - sometimes conscience, sometimes soul, always whispering of God - may have its origin in the Moralia in Job of Gregory the Great, where he explores the nature of inspiration (Wood, 2011, 152-3). The pruning of the vine, as at Leominster, seems to be a unique example of that particular imagery. A common motif embodying a similar Pauline passage is the equal pair of wrestlers: these could depict the inner struggle of a man with himself, as described in Romans 7:14-23. For the right capital at Little Langford (Figures 5 and 6), it can therefore be construed that spiritual advice or hopes of heaven - memories of the preaching symbolised by the hunting-horn in the lintel perhaps - are enabling the man to restrain his worldly desires. Left capital: a standing man This capital (Figures 11a and lib) probably had a cable-patterned ring, as on the right side of the doorway. The lichens can sometimes give a misleading impression which is hard to dispel, for example, on the angle, in a triangular shape, lichens combine with the worn stone to mimic an upside down face. A definition of the subject cannot be made with complete confidence but, overall, the remains on the angle suggest a standing man. The upper blob must be his head as part of an eye remains on the THE ROMANESQUE DOORWAY AT LITTLE LANGFORD 153 Fig. 12 Lessay abbey ( Manche ): capital in south transept. From Bayle 1992, fig. 637. Reproduced with permission. right (east face of the capital); the shape of the head was like that on the right capital, broad in the skull and narrow below. The man’s body is represented by the triangular skirt of his tunic, an elementary form used in some of the earliest or least sophisticated sculpture; the two slim prongs below the triangle would be his legs, which are often insignificant in sculpture of this period (for example, Wood 2006, fig. 4). The curious ears on the top of the man’s head are undeniably original, but have no parallel in English Romanesque sculpture known to the author. Sculpture in Normandy was particularly influential in this region of England, and there is an example of a man with similar long ears in Normandy, on a voluted capital in the south transept of Lessay abbey (Manche: Figure 12). While the ears are not so long there as they are at Little Langford, they are impossibly narrow for human ears and they are higher on the head than ears would normally be. It is suggested that, in both cases, the long narrow ears raised on top of the head are intended to make the viewer think of the nervous hare, which is sometimes a figure for the defenceless believer who should run for the shelter offered by the Church. This use of the hare occurs at Fishlake and Adel, two churches in Yorkshire, around 1150 and is derived from the Physiologus (Wood 2000a, 31). The remains of the head, body and legs of the standing man on the angle are augmented by the Fig. 13 Bernay abbey (Eure): panel at entrance to chapel in south transept. From Bayle 1992, fig. 167. Reproduced with permission. remains of the lower arm and the tips of the fingers of the hand on the east face of the capital (Figures 1 lb and 17). The hand might have gripped the body of this snake but more probably only reached out to touch it, while on the left (south face of the capital, Figure 1 la), the hand certainly did not reach so far as the snake. There is little room on either side for the large hands that are carved on the right capital. The man’s arms, open symmetrically to left and right, would have shown him in something like the posture of the crucified Christ; this stance occurs elsewhere (Wood 2006, 177, 179, fig. 7). The man is imitating his Lord, to use the passage in Galatians again: ‘those who belong to Christ Jesus have crucified the lower nature with its passions and desires’ (5:24). Left capital: more snakes These are not so perfectly symmetrical as the snakes on the right capital, but the differences between them are probably not significant except in so far as they indicate different sources for the representation of a snake. These snakes do not wind themselves round the man to restrain him, but both hold themselves upright beside him. For comparison, Figure 13 shows 154 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 14 Long Wittenham (Berks): capital at chancel arch, left face. Photo by Ron Baxter; image courtesy of the Corpus of Romanesque Sculpture in Britain and Ireland (www.crsbi. ac.uk) a carving inside Bernay Abbey (Eure, Normandy) at the entrance to a chapel in the south transept; here a man stands between two symmetrical snakes which hold themselves upright in a similar manner. The carving is dated to before 1130 (Grodecki 1951, 7- 67). This man is smiling, chubby and naked - more than just standing between the snakes, he gives the impression of dancing with them: this carving is likely to represent a man in heaven. The man on the capital in Wiltshire, however, is clothed, and - like the Crucified - immovable; his expression is likely to have been solemn, determined. It is suggested that the left capital at Little Langford represented a believer who hears or remembers spiritual advice, which will enable him to persevere in the faith to the end of his earthly life. The two capitals of the plain chancel arch at Long Wittenham (Berks) are all that remain of Romanesque sculpture at the church, but there are several features in common with the capitals at Little Langford, for example, imposts have a slightly hollow chamfer and the capitals have cable pattern rings. The north capital has a snake on either face (Figures 14 and 15); Ron Baxter suggests the south capital ‘may originally have been carved with a two-tiered design of interlacing foliage’, so it is not of immediate interest. The church was built under the patronage of Walter Giffard, earl of Buckingham, Fig. 15 Long Wittenham: capital at chancel arch, right face. Photo by Ron Baxter; image courtesy of the Corpus of Romanesque Sculpture in Britain and Ireland (www.crsbi. ac.uk) c. 1120 (Baxter 2008a); in the twelfth century both Little Langford and Long Wittenham were in the diocese of Salisbury. One of the snakes is of very similar form to the one on the left at Little Langford (Figure 14, compare Figures 11a and 17), while the winged snake, or wyvern, on the other face (Figure 15) is of a more sophisticated kind. The motif between them on the angle has been broken off, but a pointed ear and an eyebrow survive on the right. The tiny church at Padworth, also in Berkshire, retains its plan of c. 1130 with rectangular nave and stilted apse; the patrons in the 12th century are not known ( VCH Berks, III, 413-17; Baxter 2008b). The chancel arch capitals here were carved by a more competent artist (Figure 16), one in the mainstream of Norman proficiency, the lion’s head resembling work in the eastern parts of Romsey Abbey (Hants) of c. 1120-50, and in Normandy c. 1100-20 (Bayle 1979, 126-9, figs. 174, 175, 190; Bayle 1992, 264). On the angle of the capital at Padworth, the lion’s mouth emits beaded streamers which become the tails of two wyverns: this lion’s head is probably the kind of motif which was formerly on the angle of the chancel arch capital at Long Wittenham (Figure 15). A lion can represent the believer transformed and in heaven, just as do the birds in the tympanum at Little Langford (Figure 1) or the lion, crane and bird-lion on the chancel arch at Liverton (Figure 4). THE ROMANESQUE DOORWAY AT LITTLE LANGFORD 155 Fig. 16 Padworth (Berks): capital at chancel arch. Photo by Ron Baxter; image courtesy of the Corpus of Romanesque Sculpture in Britain and Ireland (www.crsbi.ac.uk) Yet, if the lion on the angle at Padworth and formerly at Long Wittenham represents a man transformed and in heaven, then the wyverns and snakes flanking that lion are certainly more than spiritual counsellors supporting a believer in his earthly life: they perhaps represent the soul inhabiting the new, resurrected body. That there are two snakes could be due to the desirability of having spiritual or heavenly things symmetrical, that is, perfect, or to show the superabundance of life in heaven, as with the lavish exuberance of paradisal foliage elsewhere in sculpture. Life in heaven is a subject suitable for capitals of a chancel arch, which is both within the church (the paradise) and close to the altar (the place where God resides). It is this setting which accounts for the delight of the little man in Bernay Abbey, as well as for the transformation of old physical bodies into new spiritual ones, shown by lions and birds, or combinations of them, as on the left side of the chancel arch at Liverton. On the other hand, outside a church, as at Leominster and Little Langford, it is appropriate for the lessons on the capitals of the doorway to be about living faithfully in the world. The two capitals at Little Langford, as reconstructed in Figures 6 and 17, illustrate teaching supporting the earthly life of an individual believer, while the lintel symbolises the corporate life of the Church, and the tympanum pictures the heavenly reward awaiting the faithful. Fig. 1 7 Little Langford: left capital, reconstruction over a photograph by Michael Tisdall Acknowledgements I am very grateful to Professor Maylis Bayle for giving permission for the use of her pictures of the carvings at Lessay and Bernay abbeys in Figures 12 and 13. It was useful to have been able to discuss the clothing worn in various carvings with Sarah Thursfield, though it was not possible for us to reach any definite conclusions about the actuality of what is being worn in most of them. Professor Malcolm Thurlby kindly let me see the introduction to his forthcoming book on Romanesque Gloucestershire, as general statements there were relevant to an assessment of the doorway at Leominster. References BARCLAY, W., 1962, Flesh and Spirit: an examination of Galatians 5, 19-23. London: SCM BAXTER, R., 2008a, Long Wittenham, Berks. The Corpus of Romanesque Sculpture in Britain and Ireland , www. crsbi.ac.uk BAXTER, R., 2008b, Padworth, Berks. The Corpus of 156 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Romanesque Sculpture in Britain and Ireland , www. crsbi.ac.uk BAYLE, M., 1979, La Trinite de Caen. Geneva: Droz BAYLE, M., 1992, Les Origines et les Developpements de la Sculpture Romane en Normandie. Caen: Art-de-Basse- Normandie CARLS SON, F., 1976, The Iconology of Tectonics in Romanesque Art. Hassleholm: AM Tryck CLAPHAM, A., 1934, Royal Commission on Historic Monuments, Herefordshire , III CUNNINGTON, C. W. and CUNNINGTON, P 2nd edn, 1969 , Handbook of English Medieval Costume. London: Faber and Faber GRODECKI, L. 1951. Les debuts de la sculpture romane en Normandie. Bulletin Monumental , 7-67 HAMER, E., 1992, Patronage and Iconography in Romanesque England: the Herefordshire School in context. Unpublished University of Chicago PhD thesis HILL, E., trans., 1992, The Works of St Augustine: a translation for the 21st century , vol III/5. Brooklyn: New City Press KEYSER, C., 2nd ed., 1927, A List of Norman Tympana and Lintels with figure or symbolical sculpture still or till recently existing in the churches of Great Britain. London: E Stock KNIGHT, J. K., and WOOD, R. 2006. St Gwynllyw’s Cathedral, Newport: the Romanesque archway. Archaeol. Cambrensis 155, 163-185 PAWSEY, H., 2001, ‘The Legend of St Etfrid, priest of Leominster’, in A. Malpas, J. Butler, et al., (eds), The Early Church in Herefordshire, ed. Leominster: Leominster Study Group POWELL, J. U. 1909. The Norman tympanum of Little Langford church. WANHM 26, 207-212 THURLBY,M., 1999, The Herefordshire School of Romanesque Sculpture. Almeley: Logaston Press VCH, 1995, The Victoria History of the Counties of England: A History of Wiltshire, volume 15. Woodbridge: Boy dell and Brewer VCH, 1906, The Victoria History of the Counties of England: A History of Berkshire, volume 3. London: Dawsons of Pall Mall for the Institute of Historial Recsearch WOOD, R. 2000a. The Romanesque doorway at Fishlake. Yorkshire Archaeological Journal 72, 17-39 WOOD, R. 2000b. Before the Green Man. Medieval Life 14, 8-13 WOOD, R. 2001. Geometric Patterns in English Romanesque Sculpture. Journal of the British Archaeological Association 154, 1-39 WOOD, R. 2003. The Augustinians and the Romanesque sculpture at Kirkburn church. East Yorkshire Historian 4, 3-59 WOOD, R. 2004. The Romanesque font at St Marychurch, Torquay. Proceedings of the Devon Archaeological Society 62, 79-98 WOOD, R. 2005. The Romanesque tympanum at Fownhope, Herefordshire and the functioning of the Herefordshire School of Romanesque Sculpture. Transactions of the Woolhope Naturalists Field Club 53, 51-76 WOOD, R. 2006a. The Romanesque Church at Melbourne. Derbyshire Archaeological Journal 126, 127-168 WOOD, R. 2006b. The Romanesque chancel arch at Liverton, North Riding. Yorkshire Archaeological Journal 78, 111-143 WOOD, R. 2009. The Occupatio of St Odo and the porch sculpture at Malmesbury Abbey. WANHM 102, 202-210 WOOD, R. 2011. The Romanesque doorways at Dinton and Leckhampstead. Records of Buckinghamshire 51, 139-168 ZARNECKI, G., 1972, The Monastic Achievement. London: Thames and Hudson Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 157-167 The miracle of the horseshoe: a 15th century wall painting at Highworth Church, Wiltshire by Ellie Pridgeon1 and Roger Rosewell2 with photography by Jo Hutchings and Sam Hutchings This paper considers some important yet neglected evidence for late medieval wall paintings in the parish church of St Michael at Highworth and discusses their wider significance. Introduction In 2010, three wall painting tracings were re- discovered during refurbishment at the Wiltshire Heritage Museum. Although known since 1936 when they were reported as ‘lately come to light in Devizes Museum rolled up among some maps’, the tracings received little attention at that time (Cunnington 1936, 420). Recent research suggests that the tracings are of national importance for wall painting studies for several reasons, not least because they provide evidence of the types of mural formerly to be found in local parish churches such as St Michael at Highworth (Wiltshire), a mainly fifteenth-century building on a Norman site (Figure 1). The tracings portray images of St Eligius (or Eloy), a French-born saint with an international following, an angel, and Post-Reformation text lettering, and were drawn in 1861 when the paintings were uncovered (and subsequently destroyed) during the restoration of the church (1850s and 1860s) (Hinton et al. 2010, 2). According to an inscription within the St Eligius reproduction, the painting occupied ‘the pier between Nave and North Transept of Highworth Church’. The tracing measures 1.18m x 2.33m and depicts the saint as a mitred bishop standing in a blacksmith’s workshop performing one of the most well known miracle stories of late medieval Christianity: the shoeing of a restless horse by cutting off one of its legs, fixing the shoe, and then miraculously restoring the limb to the no doubt bewildered animal. The painting dates from the 15th century (Figure 2). This article summarises the life of St Eligius and his cult, considers evidence of medieval devotion in Highworth church and concludes with a discussion of the purpose and function of wall paintings in that process. St Eligius St Eligius (Eloi, Loy, Loye) (c. AD580-660) was born at Chaptelet, Haute-Vienne, France. As a boy he was apprenticed to a goldsmith and later became Master of the Mint to the Frankish kings, Clothar II (584-629) and Dagobert I (629 -39) (Farmer 1982, 130). An account of the treasures displayed on the high altar of the royal abbey of St Denis (near Paris) after its refurbishment in the 1140s, referred to the ‘wonderful cross of St Eloy’, although it is not clear Department of History of Art and Film, University of Leicester; 2 Editor, Vidimus : Online Stained Glass Magazine 158 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 1 The parish church of St Michael, Highworth if the cross had been made by the saint or merely belonged to him (Panofsky 1948, 63). According to the vita , a hagiographical life of the saint written shortly after his death by his friend and contemporary St Ouen, Archbishop of Rouen, Eligius was given just enough gold by Clothar to make a sella regis , or king’s throne (sometimes misinterpreted as saddles) but with amazing skill and ingenuity, Eloi produced two. In the late Middle Ages this feat saw him adopted as the patron saint of goldsmiths (Forsyth 1946, 143-4). The story is depicted in an initial in the Carmelite Missal of c. 1393, a manuscript associated with the London Whitefriars. It shows the bishop presenting a golden saddle to the king (British Library Add. Mss. 29704- OS, fol. 164r). During his time as a goldsmith, Eligius became Fig. 2 Tracing of St Eligius wall painting a priest and thanks to his royal connections he was consecrated Bishop of Noyon-Tournai in AD641. He subsequently founded monasteries at Noyen, Paris and Solignac. He also had a special relationship with the poor: ‘S. Loye loved well poor people. ..the poor people also loved him, that where he went they followed him, and they that would speak with him must ask and enquire of the poor people where he was’ (Ellis 1900, 262). Neither the vita , nor his entry in the Legenda Sanctorum (the Legend of the Saints), better known as The Golden Legend, compiled around 1260 by Jacobus de Voragine, make any mention of the miracle depicted at Highworth (perhaps unsurprisingly), or the saint’s direct involvement with the craft of blacksmithing. Despite the lack of textual sources - the first known written account of the Highworth A 15TH CENTURY WALL PAINTING AT HIGH WORTH CHURCH, WILTSHIRE 159 Fig. 3 Miracle of St Eligius, Freiburg Minster. © Corpus Vitrearum Medii Aevi Deutschland/ Freiburg i. Br. story dates only from the 15 th century, depictions of the leg severing and restoring miracle survive from the mid- 13th century in France, and in Germany and England from the first half of the 14th century. The evolution and dissemination of this miracle story are thus of great interest to historians both as a case study in the transmission of images and the receptivity of medieval audiences to new cults and saints. The Highworth story seems to have evolved from several traditions. First, a story from the vita which claims that after Eligius’ death a horse he had owned was stolen from its rightful owner by a greedy bishop. The aggrieved victim, a local abbot, prayed to Eligius to intercede on his behalf. Thereafter, despite having a previously gentle disposition, the horse suddenly developed ‘a sickness of the feet’, and became, Tike a wild beast roaring and kicking’, when anyone attempted to cure it. Exasperated by its uncontrollable nature the bishop gave the horse away to a matron he loved, but its new owner had no better luck. When she mounted the horse it began to ‘rage’, and threw her to the ground. After further unsuccessful efforts to ‘cure’ the horse, the bishop returned it to the abbot from whom it had been unjustly taken. Within a few days the horse was healed and had regained its former gentleness (Krusch 1902, 726-7). Next, a variation which admonished vanity and told how the saint had hung a sign outside his workshop announcing himself, ‘St Eloi : master of all masters’, until a traveller came and asked him to shoe his horse. St Eloi tried, but failed, since the horse was extremely restless. The traveller then took the horse’s leg, detached it, shoed it and re-attached the leg. St Eloi understood that the stranger was a mighty man and, indeed, it was Christ himself: St Eloi thus learned humility. Finally, by the early- 13th century the story had changed again with St Eloi cast as the man who shoed the horse (Branner 1967, 55-73). This change may be associated with the promotion of the saint’s cult by the Abbey of Noyon which was rebuilt between 1217 and 1240. The abbey owned relics of the saint, which attracted pilgrims. An early image of the miracle survives in a window at Angers Cathedral dated to 1225-35, which also shows other episodes from the saint’s life, including the story of the two thrones/saddles. Thereafter examples of the miracle appeared in glass at Le Mans Cathedral, c. 1235-40 and can be seen in a window at Freiburg Minster (Germany) dated to 1325-30 (Figure 3). One of the earliest depictions of the miracle in English art appears in the margins of the Luttrell Psalter made for Sir Geoffrey Luttrell (1276-1345) of Irnham (Lincolnshire), probably sometime between 1325 and 1335. Folio 52 shows the mitred bishop holding a pair of tongs in his right hand and the severed leg of the horse in his left. Further evidence of the saint’s inroads into popular culture can be gleaned from Geoffrey Chaucer’s Prologue to the Canterbury Tales written in Middle English towards the end of the 14th century, where the author’s introduction to one of his characters, the Prioress, says that ‘Hir gretteste ooth was but by Seinte Loy’ (Cawley 1990, 14). Two strands of devotion appear to have developed as the saint’s cult grew: people seeking individual protection for their horses and his adoption as the patron saint of goldsmiths, blacksmiths and farriers (specialists in equine hoof care, including the trimming and balancing of horses’ hooves and the placing of shoes on their hooves). The earliest surviving representation of the Highworth miracle in English monumental art is thought to be a late 14th century wall painting on the south wall at the parish church of St Botolph at Slapton, Northamptonshire. Although missing most of its surface paint the scene shows, left to right, a tall man with a bishop’s crozier standing by a furnace and holding the leg of a horse, a smaller figure working the furnace, and finally the three-legged horse. A further refinement is a gabled wooden frame, suggesting that the scene is either taking place 160 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE k /\/\/ TOT jC X yv \/y\/ X XXI V7xy^/I A A. 3* >< >< ±4 i m y V x Fig. 4 Miracle of St Eligius, Church of St Botolph, Slapton. © YM Pictures within a building, perhaps a blacksmith’s forge, or that the horse is being supported by a building or something similar (Figure 4). Three other wall paintings survive, all 15th century. At the parish church of the Holy Trinity at Wensley, North Yorkshire, a fragmentary painting shows the horse and a blacksmith at work (Figure 5). At the parish church of St Lawrence at Broughton (now part of Milton Keynes), Buckinghamshire, the central part of the scene is obscured by a later funerary monument. Notwithstanding this loss, some of the peripheral details of the drama remain visible. These include blacksmith’s products and tools: padlocks, keys and horseshoes can all be seen. Different types of hammers are shown, including a farrier’s claw hammer for extracting nails, similar to that illustrated in the Highworth painting. The object directly below the monument plaque may be a slack tub or quench bucket. Metal rods or tools are visible in the trough. Finally, the front part of a horse can be seen to the right of the monument while faded traces of an arm and hand, possibly operating a lever to inflate bellows, creeps around the left hand side of the plaque (Figure 6). The fourth example is at the tiny church of All Saints at Shorthampton, Oxfordshire. Although now fragmentary, it shows the horse supported by a frame while a smith stands behind him. Together with the lost Highworth example, all five images are notable for showing the saint in a narrative context. Examples of both the narrative iconography and St Eligius as a single figure also appeared on 14th and 15th century on rood screen panels, in painted and stained window glass and on alabaster panels which formed altar retables in the pre-Reformation church. It is also likely that he was depicted as a free standing carved statue and he may have appeared on painted clothes and in other media. Rood screens with painted images of the saint as a single figure with attributes survive at Great Yeldham, Essex, where he is shown holding a claw A 15TH CENTURY WALL PAINTING AT HIGHWORTH CHURCH, WILTSHIRE 161 Fig. 5 Miracle of St Eligius, Church of the Holy Trinity, Wensley. © YM Pictures hammer, Potter Heigham, Norfolk, where he also holds a claw hammer and is unusually included in a Doctors and Evangelists set, probably because of the requirements of a donor, and, until it was stolen in the 1980s, and at Hempstead, Norfolk where the saint is shown as a bishop with a hammer in one hand and holding a horse’s leg in the other (Figure 7). A now lost window showing the scene of the saint shoeing the horse has been recorded at West Rasen, Lincolnshire, while at Digby, Lincolnshire, a figure of the saint once shared a window with SS Theobald and Nicholas (Hebgin Barnes 1996, 345, 80). A window depicting a bishop with a hammer previously existed at South Walsham Norfolk (Nichols 2002, 190). Six alabaster tables or panels have been recorded with fragments at St Andrew, Toft, Cambridgeshire, and St Andrew, Freckenham, Suffolk, still displayed in the churches (Cheetham 2003, 39). A well- preserved panel, now in the Victoria and Albert Museum, shows the saint, wearing a bishop’s mitre, holding the severed horse’s leg on top of the anvil with his left hand and while hammering a shoe Fig. 6 Miracle of St Eligius, Broughton. © YM Pictures onto the hoof with his right. The haft of another hammer is slotted into a holder beneath the anvil and two more hammers lie on the ground in front of the horse. Below the anvil, at the front of the panel, there is a trough of water containing what appear to be metal rods. A damaged figure stands behind the saint at the top left of the panel and appears to be pulling on a lever with his left hand to work the bellows at the back of the smithy. To the right of the saint are the hearth and chimney of the forge. Under a roof supported by wooden posts and beams a man stands behind a horse; he looks down at it, his hands held apart in surprise. He wears a flat cap and a tunic. The horse, its left front leg removed and held by the saint, is tied to a post by a rope round its head; it looks out of the panel. On its back is a saddle with a stirrup attached. Above, four horse shoes hang on nails on the roof beams. The panel is dated to 1440-60 (Figure 8). As many of these same features can be seen in the Highworth tracing and the four remaining wall paintings of this scene, it is likely that artists in different media were familiar with the iconography 162 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 7 St Eligius, rood screen, church of St Nicholas, Potter Heigham, Norfolk. © YM Pictures of the image and the details which made it instantly recognisable. Certainly, there seems little doubt about the popularity of the saint in the late middle ages. The wall paintings are found from Yorkshire to Wiltshire, with window glass found in Lincolnshire and East Anglia: at the parish church of St Paul, Bedford, his image is carved as a roof boss. A survey of pre-Reformation churches in West Kent discovered parishioners giving money for candles before images of ‘Saynt Loye’ at more than twenty churches including Bromley, Milton by Gravesend, Sutton-at-Hone, Dartford and Chalk. The will of an early 16th century parishioner of the parish church of St Mary at Eaton Bray, Bedfordshire, bequeathed a small sum for lights associated with St Eliguis in 1521. In both counties his association with masculine occupations is confirmed by the fact that no female testators mention bequests to his image and/or light (Marks 2004, 105-7). Fig. 8 St Eligius shoes a horse. Photo © Victoria and Albert Museum, London On a different scale a number of collective fraternities or guilds dedicated to the saint were formed in the 1 5th century. In 1424 a primarily social and religious Fraternity of St Eligius was formed by London blacksmiths, farriers and loriners (makers of bits, bridles, spurs, stirrups, saddle trees and the minor metal items of a horse’s harness) which chose its masters and wardens at a feast on the Sunday following St Eligius’ day (1 December) (London Metropolitan Archives, GL Ms 5535). In Ipswich, Suffolk, goldsmiths, locksmiths and bladesmiths gathered under the banner of St Eliguis during the town’s annual Corpus Christi procession (Rubin 1991, 262). There was a guild of St Eliguis among tinworkers at Chagford, Devon (Whiting, 1989, 50, 72, 110). The farriers of Derby were responsible for the maintenance of a light before his image at the main parish church of All Saints (Cox 1913, 164). A 15TH CENTURY WALL PAINTING AT HIGH WORTH CHURCH, WILTSHIRE 163 Medieval wall painting: context and outline In the medieval period, church interiors were adorned with a complex array of wall paintings depicting saints, the Life of Christ, Old and New Testament scenes, the Last Judgement, the Seven Deadly Sins, and decorative patterns (Gill 2002; Rosewell 2008). Yet only a small proportion of these images are visible today. Survival has been hampered not only by over-painting with whitewash and texts by 16th and 17th century religious reformers, but also by the continual remodelling of churches from the medieval period to the present day (Aston 1988, 258-9). Fortunately, the whitewashing process of concealment is frequently reversible, and has helped to preserve paintings in many instances. Many were rediscovered behind peeling paintwork during 18th, 19th and 20th century church renovations, while new examples continue to be revealed, for example the Doom painting at St Lawrence, Ipswich - now whitewashed over again for preservation purposes. A principal obstacle to working with wall paintings is that it is complicated to reconstruct exactly how a particular building might have appeared in the medieval period. Most murals that do survive are fragmentary and many more have been subject to inaccurate restoration, or even pigment changes. At Inglesham, Wiltshire, for instance, flaking paintwork reveals centuries of layering, including post- Reformation script painted over a late medieval St Christopher (north wall). Even when churches have a more complete and comprehensive ‘scheme’ of wall paintings, these edifices have frequently lost most of their glass, sculpture and other imagery, making it impossible to ascertain how the various media would have related to and interacted with one another (Gill 2006, 17- 32). This is the case at Lydiard Tregoze, Wiltshire, where fragments of medieval glass indicate that the surviving murals (Martyrdom of Becket, St Christopher, Rood cross with heads, Weighing of Souls, Christ as a Man of Sorrows) were once part of a holistic and interactive design (Rosewell 2008, 299). Despite numerous drawbacks, there are many advantages to working with wall paintings and there are many more surviving murals than extant images in other media. Furthermore, wall paintings cannot easily be repositioned within a church building (unlike glass or sculpture), and are therefore are generally located in their original position. It was common for wall paintings revealed during Victorian restoration to be recorded in watercolour or sketch form before destruction or conservation. Watercolour reproductions exist for various lost murals in Wiltshire, including St Christopher and St Michael at Ditteridge, and St Christopher at Broade Chalke (Trethowan n.d. 9- 10; Whaite 1929, 41, 44). These are all held in their respective churches (and a second Broade Chalke reproduction is at the Society of Antiquities) (Society of Antiquities, BP 49). Richard Kemm also produced a watercolour reproduction of the lost St Christopher painting at Idmiston, executed in 1866 (Wiltshire Heritage Museum, DZSWS. 1982.2 199). A drawing in the church records the lost St Christopher mural at Wilsford, and a pen and ink sketch (included in a letter) confirms the existence of a long-vanished St Christopher image at Durrington (Keyser 1883, 91; Private Collection 1855). Late medieval devotion at Highworth Although no documentation survives regarding the patronage or function of the wall painting of St Eligius at Highworth, it is possible that it was commissioned by a local farrier or by parishioners with a keen interest in the protection of their horses. It is clear from examining sources from other counties that images might be financed through individual sponsorship. At Benenden, Kent, for instance, John Sharpe left £2 12s 4d in his will ‘for painting new of the porch and of St Christopher’ (Hussey 1907, 16). Yet imagery was also funded through communal collections. In the accounts for the Chapel of the Holy Rood in Bodmin (known as the Berry Tower), churchwardens Thomas Phylype and Robert Kyrekeby record gifts received towards the painting of St Christopher: In ‘It[e]m rec[evyd] of the gyftte of dyv[er]s good men to the payntyng of Seynte Cris[tof]or, xvjd’ (1512-14) (Cornwall Record Office, B/Bod 314, membrane 13). It is possible to create a picture of late medieval devotion at Highworth by examining primary sources relating to the church. Highworth is mentioned in the Patent Rolls (the records of unsealed royal grants of official positions, land or commissions) in 1453, when a licence is granted to for the establishment of a chantry in the church: ‘William Ingram, Canon of the cathedral church of St. Mary, Salisbury and prebandary of the prebend 164 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE of Heyworth, Co. Wilts, to found a chantry of one chaplain to celebrate divine service daily at the altar of St. Nicholas in the church of St. Michael, Hayworth, for the good estate of the king and queen Margaret and William and for their souls after death and the souls of Henry V, queen Katherine and the parents of the benefactors of William; the chaplain to be called the chaplain of the Chantry of William Ingram, Heyworth’ (PRO 1910, 66). The chaplain is mentioned again in the 1460/1461 in the Chancery Inquisitions Quod Damnum (The National Archives, Cl 43/452/27). These investigations were undertaken after applications to the Crown (e.g. for licences to alienate property) in order to ensure protection of royal interests. John Abbot of Stanley Abbey, Bremhill, applied to the Crown to grant rent from his manors to John Salve, Chaplain of the altar of St Nicholas in the church of St Michael, Highworth. It is likely that the chantry was situated in the chapel to the north of the chancel, where a piscina denotes the existence of an altar. Early 16th century wills indicate the presence of lights (or candles) at altars and/or before images in Highworth church. Lights were most commonly lit during Mass, although the major images may have had candles burning before them in perpetuity (Marks 2004, 162). It was common practice to leave money or gifts in testamentary bequests to pay for the purchase of wax. This would acquit the donor of any spiritual debts, remind living parishioners of the donor’s devotion to encourage their prayers (intercession), and ultimately help to speed the soul’s path through Purgatory (Kamerick 2002, 98). The will of Richard Rondell (dated 1510) points to the presence of an All Hallows light and a principal (or high altar) light at Highworth: Tt[em] to the high aulter for tithes forgot[t]en iiijs iiijd. It[em] to Alhalowes light xijd and ij q[ua]rters of malte It[em] to the iiij. principall liyht[es] xvjd’ (The National Archives, PROB 11/16). Rondell also left money to ‘the peyntyng of the Roode lofte vis viijd’ (The National Archives, PROB 1 1/16). The will of Walter Arden (1540) mentions ornaments at the high altar: Tt[e]m to the reparatons of the ornaments of the high Aulter of highworth vjs viijd. It[e]m to the behalf of the said churche viij busshels of malte’ (The National Archives, PROB 11/28). Image function in the medieval church: academic context Didactic Function Research has traditionally focused on how medieval images were used for didactic purposes (Male 1961; Duggan 1989, 227-51). The extent to which they functioned in this manner within the church milieu has been exaggerated by researchers in the past, many of whom traditionally portrayed visual depictions (wall paintings in particular) as ‘The Biblia Pauperum, or Poor Man’s Bible’ (Clive Rouse 1996, 13; Male 1961). Such opinions rely too heavily on the letters of Gregory the Great (AD599 and AD600), written in response to the iconoclastic activities of Bishop Serenus of Marseilles: For a picture is provided in churches for the reason that those who are illiterate may at least read by looking at the walls what they cannot read in books Norberg 1982, 768; Martyn 2004, 674, 745). There are a number of reasons why this source should not be taken at face value. First, it is implausible to assume that attitudes towards imagery usage did not change over the 900-year period up to the Reformation (Aston 1988, 24-5). By the 13th century, the prescriptions laid out by Gregory the Great had long been supplanted by attitudes and conditions in which images were more legitimate and powerful (Hamburger 1989, 163). Secondly, as Chazelle points out, there is no reason to believe that the Pope’s words were reflective of the more general attitude towards artwork in the 6th and 7th centuries (Chazelle 1995, 139). Chazelle also maintains that Gregory was writing to Bishop Serenus to chastise him for his destruction of images, and that the emphasis on the value of artistic depictions to the illiterate may have been a passing comment intended as a strategy to change his mind, with little thought about the very specific and complex ways in which images functioned (Chazelle 1995, 139, 148). In theory, any image could be used as an instructional tool if a preacher or priest (or family member) were to explain its significance to a congregation or individual (or if the latter were literate enough to grasp the meaning from an explanation in a book). Surviving sermon literature suggests that priests may not have employed images in a didactic manner to the extent that some scholars have assumed, although it is still possible to find occasional references to visual symbolism in texts (Nilsen 2003, 326, 331). Three A 15TH CENTURY WALL PAINTING AT HIGHWORTH CHURCH, WILTSHIRE 165 sermons of Thomas Brinton, Bishop of Rochester (d. 1389) mention the Wheel of Fortune, likening the turning circle to the troubles and luck of life (Gill 2002, 160). There is a possibility that he had in mind the mid- 13th century wall painting that survives on the choir wall in Rochester Cathedral (Park 1987, fig. 95) Yet we do not know whether the sermon was ever delivered in the Cathedral, and indeed if it were, whether the preacher made specific reference to the painting. Nilsen has also astutely pointed out that it may not have been feasible for priests to use paintings as an effective pedagogic tool during preaching (Nilsen 2003, 326, 331). The conditions of the medieval church were such that the nave would have been crowded and badly lit, and visual depictions may have been too far away or high up for the preacher to refer to or for congregation to view successfully (Nilsen 2003, 326, 331). Mnemonic Function All images are mnemonic in that they serve to remind viewers of their desire for salvation, and ultimately God and hence remind supplicants of meaning, significance and required actions (Yates 1966). Church imagery served to remind the viewer of what he or she already knew, and to recall what was in the mind (and ultimately God himself). Yet a series of narrative paintings depicting a story in a number of scenes, or more explicit and diagrammatic images such as the Seven Deadly Sins, would make little sense to a viewer who did not know who the characters were or what the symbols represented. The mnemonic function of imagery is discussed by Reginald Pecock in his Repressor Over Much Blaming of the Clergy (1449- c.1455), a text defending clergy against perceived unjust aspersions by Lollards (Babington 1860). Pecock deals directly with the relationship between viewers and artwork in churches, and as a clergyman, it is likely that he was aware of how visual depictions functioned in the parish environment. He claims that images remind viewers of ‘God, and of hise benefetis, and of his holi lijf and passioun, and of Seintis and of her holi conuersacioun’ (as well as angels in heaven and moral governance on earth) (Babington 1860, xxii, 172). For Pecock, they were ‘seable rememoratijf” or ‘rememoratijf visible signes’, and: In iholding bi si3t of i3e upon manye dyuerse stories or agis...a man schal in a litil while be embrid now upon the passioun of Seint Laurence, and now upon the passioun of Seint teuen, now anoon aftir vpon the ssioun of Petir, and so forth of anyechaungis (Babington 1860, xxii, 166, 167, 213). Other Functions It is clear that different images types functioned in diverse ways. For instance, the Crucifix was intended to create a very direct, personal and interactive relationship with the crucified Christ in appreciation of the suffering he endured (Swanson 1998, 14). Other images such as the Doom, the Seven Deadly Sins and the Seven Corporeal Works of Mercy, were more specifically moralistic than emotional in purpose, prescribing, describing and reminding viewers of suitable and unsuitable behaviour, and prompting spiritual development (Gill Website). Images of saints on the other hand, growing in number from the 13th century onwards, acquired and performed numerous and varied functions (Marks 2004, 87). Frequently located in the nave or ‘lay space’, their principal function was to act as an approachable intercessor between Man and God, a visual and physical link or channel in the chain between heaven and earth. Through this process, sins could be forgiven and ultimately salvation gained after death (Kessler 2006, 430). Saints were also regarded as figures to be imitated and emulated, and to offer help in time of hardship and need. Julian of Norwich (c.1342-c.1416) likened them to kind friends who sympathised and understood: All the helpe that we haue of speciall sainctes and of all the blessed companie of heauen, the dere worthie loue and the holie endles frinshipe that we haue of them, it is of his godnes (Colledge 1978, 302) Certain saints also provided protection against illness, harm and disease. St Barbara and St Catherine were invoked in childbirth and against unprepared death, and several wall painting inscriptions suggest that seeing the figure of St Christopher would save the viewer from hurt, illness or sudden death that day (Granger Ryan 1993, 370). At Woodeaton, Oxfordshire, for instance, the Anglo-Norman speech scroll inscription in the mid- 14th century mural there reads: ‘Ki cest image verra le jur de male mort ne murra’ (he who sees this image shall not die an ill death this day) (Pridgeon 2008, 94; Tristram 1955, 115). The avoidance of sudden death was particularly significant as it was necessary to be shriven of sin (shrift) and to receive a final communion (housel) if one wanted to limit time spent in purgatory and avoid the pains of Hell (Duffy 2005, 120). 166 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Acknowledgements The authors would like to thanks Dr Miriam Gill for discussions about St Eligius, Dr Phillip Lindley and Dr Cathy Oakes for their interest in wall painting, Lucy Wrapson for conversations about St Eligius roodscreens, Samantha Harper and the Highworth Historical Society for their local knowledge, and Dr Paul Dryburgh for checking the transcriptions. We are also grateful to Dr Jo Mattingly for drawing our attention to the reference to St Christopher in the Berry Tower material, Julia Park (Conservator) for providing information about the Ipswich Doom, and Heather Ault and Lisa Brown (Wiltshire Heritage Museum) for facilitating access to the Highworth tracing. Primary sources British Library. Add. Mss. 29704-05, fol. 164r. Carmelite Missal. Cornwall Record Office. B/Bod 314, membrane 13. Berry Tower Accounts of Thomas Phylype and Robert Kyrkeby Receivers, 1512-14. London Metropolitan Archives. GL Ms 5535. Worshipful Company of Blacksmiths PRO, 1910, Calendar of the Patent Rolls preserved in the Public Record Office , Henry VI, Vol. 6 (1452-1461), London: HMSO Private Collection. Letter from J. W. Hugall to Reverend Richard Webb, 29 October 1855. Society of Antiquaries. BP 49. Fresco by E. Godwin, 1850 (Reproduction of St Christopher wall painting at Broade Chalke, Wiltshire). The National Archives. C143/452/27. Chancery : Inquisitions ad Quod Damnum, Henry III to Richard III. The National Archives. PROB 11/16. Will of Richard Rondell of Highworth, Wiltshire, 1510. The National Archives. ROB 1 1/28. Will of Walter Arden of Highworth, Wiltshire, 1540. Wiltshire Heritage Museum, Devizes. DZSWS. 1982.2199. Kemm Drawings. References ASTON, M., 1988 , England’s iconoclasts: laws against images, Vol. 1. Oxford: Oxford University Press BABINGTON, C., (ed.), 1860, The repressor over much blaming of the clergy by Reginald Pecock. London: Longman, Green, Longman and Roberts BRANNER, R. 1967. Le rouleau de Saint-Eloi. Information d’Histoire de TArt 12, 55-73 BRUNO, K., (ed.), 1902, Vita sancti Eligii episcopi Noviomagensis . Hanovre: Hahn CAWLEY, A. C., (rev. ed.), 1990, Chaucer: Canterbury tales. London: Everyman CHAZELLE, C. M., 1995, ‘Memory, instruction, worship: Gregory’s influence on early medieval doctrines of the artistic image’, in J. C. Cavadini (ed.), Gregory the Great: A Symposium, 181-215. Notre Dame: University of Notre Dame Press CHEETHAM, F., 2003 , Alabaster images of medieval England. Woodbridge: Boydell Press CLIVE ROUSE, E., 1996, Medieval wall paintings. Princes Risborough: Shire COLLEDGE, E. and WALSH, J., (eds), 1978, A book of showings to the anchoress Julian of Norwich. Toronto: Pontifical Institute COX, C., 1913, Churchwarden accounts: from the fourteenth century to the close of the seventeenth century. London: Methuen CUNNINGTON, M. E. 1936. Wall paintings formerly in Highworth Church. WANHM 47, 420-422 DUFFY, E., 2005, The stripping of the altars: traditional religion in England, 1400-1580. New Haven and London: Yale University Press DUGGAN, L. G., 1989. Was art really the ‘book of the illiterate?. Word and Image 5:3, 227-251 ELLIS, F. S., (ed.), 1900, The golden legend or lives of the saints: as Englished by William Caxton. London: Temple Classics FARMER, D. H., 1982, The Oxford dictionary of saints. Oxford: Oxford University Press FORSYTH, W. H. 1946. Saint Eloi and King Clothar. The Metropolitan Museum of Art Bulletin (New Series) 4:5, 143-4 GILL, M., 2002, Late medieval wall painting in England: context and content, C.1330-C.1530. Unpublished University of London PhD Thesis (Courtauld Institute) GILL, M., 2002, ‘Preaching and image: sermons and wall paintings in later medieval England’, in C. Muessig (ed.), Preacher, sermon and audience in the middle ages, 155-80. Leiden: Brill GILL, M., 2006, “‘Reading images”: church murals and collaboration between media in medieval England’, in S. Bigiliazzi and S. Wood, Collaboration in the arts from the middle ages to the present, 17-32. Aldershot: Ashgate Publishing Limited GILL, M., Seven Deadly Sins and Seven Corporal Works of Mercy Website: www.le.ac.uk/arthistory/seedcorn/ contents.html (Accessed July 2011) GRANGER RYAN, W., (ed. and transl.), 1993, The golden legend: readings on the saints. Princeton: Princeton University Press HEBGIN BARNES, E, 1996, The medieval stained glass of the county of Lincolnshire, CVMA (GB), Summary Catalogue 3, Oxford: Oxford University Press HAMBURGER, J. 1989. The visual and the visionary: A 15TH CENTURY WALL PAINTING AT HIGHWORTH CHURCH, WILTSHIRE 167 the image in late medieval monastic devotions. Viator 20, 161-82 HINTON, A., HINTON, E AND WITHE Y, L., 2010, St Michael and All Angels Church, Highworth: a brief history and guide, Publisher/Place unspecified HUSSEY, A., (ed.), 1907, Testamenta Cantiana: A series of extracts from fifteenth and sixteenth century wills relating to church building (East Kent). London: Hughes and Clarke KAMERICK, K., 2002, Popular piety and art in the late middle ages: image worship and idolatry in England, 1350-1500, New York: Palgrave KESSLER, H. L., 2006, ‘Turning a blind eye: medieval art and the dynamics of contemplation’, in J. F. Hamburger and A. Bouche (eds), In the mind’s eye: art and theological argument in the middle ages, 413-439. Princeton: Princeton University Press KEYSER, C. E., 1883, List of buildings having mural decorations. London: Eyre and Spottiswode MALE, E., 1961, The Gothic image: religious art in France of the thirteenth century. London: Fontana MARKS, R., 2004, Image and devotion in late medieval England. Stroud: Sutton Publishing MARTYN, J. R. C., (ed.), 2004, The letters of Gregory the Great 2. Toronto: Pontifical Institute NICHOLS, A. E., 2002, The early art of Norfolk. Kalamazoo: Medieval Institute Publications NILSEN, A., 2003, ‘Man and picture: on the function of wall paintings in medieval churches’, in A. Bolvig and P Lindley (eds), History and images: towards a new iconology, 323-40. Turnhout: Brepols NORBERG, D., (ed.), 1982, 5. Gregorii Magni Registrum Epistularum 1. Turnout: Brepols PANOFSKY, E., 1948, Abbot Suger on the Abbey Church of St-Denis and its art treasures, (Second Edition by Gerda Panofsky-Soergel, 1979). Princeton: Princeton University Press PARK, D., 1987, ‘Wall Painting’, in J. Alexander and E Binski (eds), The age of chivalry: art in Plantagenet England, 1200-1400, 125-130. London: Weidenfeld and Nicolson PHILLIPS, J., 1973, The Reformation of images: destruction of art in England, 1535-1660. Berkeley: University of California Press PRIDGEON, E., 2008, St Christopher wall paintings in English and Welsh churches, C.1250-C.1500. Unpublished University of Leicester PhD Thesis ROSEWELL, R., 2008, Medieval wall paintings in English and Welsh churches. Woodbridge: Boydell Press RUBIN, M., 1991, Corpus Christi: The Eucharist in late medieval culture. Cambridge: Cambridge University Press SWANSON, R. N., 1998, ‘Passion and practice: The social and ecclesiastical implication of Passion devotion in the late middle ages’, in A. A. MacDonald, H. N. B. Ridderbos and R. M. Schluseman (eds), The broken body: Passion devotion in late medieval culture, 1-30. Gronigen: Egbert Forsten TRETHOWAN, M., n.d, All Saints’ Broade Chalke: A visitors’ guide. Salisbury TRISTRAM, E. W., 1955, English wall paintings of the fourteenth century. London: Routledge WHITING, R., 1989, The blind devotion of the people: popular religion and the English Reformation. Cambridge: Cambridge University Press WHAITE, H. C., 1929, St Christopher in English medieval wall painting. London: Benn YATES, F. A., 1966, The art of memory. London: Pimlico Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 168-181 The Savernake Horn by Graham Bathe The Savernake Horn is a remarkably ornate medieval hunting horn incorporating craftsmanship spanning many centuries. A base of carved elephant ivory carries moulded bands of engraved and enamelled silver depicting images of the chase, royal and religious authority, and heraldic beasts. It is claimed as a symbol of office for the hereditary wardens of Savernake Forest, and is arguably the most spectacular of a small corpus of medieval objects symbolising legal tenure through royal patronage. Although the wardenship of Savernake has remained in the same family from 1130 until the current day, with branches incorporating images of hunting horns on their seals since the twelfth century, the Savernake Horn itself is unknown from documentation until its depiction on the Seymour Roll Pedigree of 1604. This, together with the fact that the carrying strap or baldric carries Scottish heraldic designs, raises uncertainties over its provenance and attribution. Description Aspects of the horn’s appearance were given in Britton (1801), Ward (1860), Michael (undated, c. 1900), and Brudenell-Bruce (1949), with more detailed analysis following its acquisition by the British Museum by Cherry (1989; 2005). It has been crafted from an ivory tusk (referred to as an oliphant, from an archaic spelling of elephant). It measures 63cm along its curved outer surface, which has been smoothed into sixteen flat facets. The oliphant conforms with others carved in southern Italy in the twelfth or early thirteenth century (Cherry 2005). A slit in the tusk set back from the bell end, at a point subsequently obscured by ornamental metalwork, shows that it was formerly suspended by a shoulder strap or baldric, fed through this slot and looped out through the bell (visible in Figure 2). This method of carrying is seen on other surviving horns, such as the Boarstall Horn. Ivory was a highly sought-after and elitist commodity in medieval times. On some horns fashioned in Italy, where a school of engraving was established, the ivory itself was carved and ornamented with hunting scenes {ibid.). In the case of the Savernake Horn, its prestige was enhanced sometime in the mid fourteenth century when bands of elaborately ornamented silverwork were moulded around the existing oliphant, each band divided into sixteen panels used for depicting images of hunting and forests (Figure 1). In turn these were embellished with coloured enamel (surviving only partially), which accentuated the lustrous gleam of the metal. With the exception of one panel which depicts a king, where the background is hatched, all others are diaper-patterned and pointed. This is characteristic of basse taille work, in which low- relief engraving holds translucent enamel, allowing light to reflect from the relief, creating an interplay between silver and colour. The artwork has been dated through stylistic correlation with enamels of known provenance, including the Bermondsey Dish of 1325, and a miniature diptych nativity scene of the second quarter of the 14th century (Camber and Cherry, 1977). The largest band of silver, at the bell of the horn, has a flange extending over and protecting the ivory, engraved with pictures of hawks, preening Byeley-in-Densome, Woodgreen, Fordingbridge, SP6 2QU THE SAVERNAKE HORN 169 Fig. 1 The Savernake Horn, carved from a tusk and bearing bands of engraved silver and with wings raised, each striking a different pose (Figure 2). The sixteen framed side panes are remarkably detailed. The focal point of these, facing the person raising the horn to blow it, are panels depicting a bishop, king and forester, each placed below three trifoliate arches (Figure 3). The king, robed and seated, sporting a crown and scepter, and with a sagacious beard and long hair, has his (seemingly gloved) hand raised with open palm as if in greeting. He resembles, to a remarkable degree, surviving images of the three generations of King Edward of the 14th century, especially Edward III. An effigy thought to represent Edward III at the west door of Lincoln cathedral, is virtually identical in features and posture. The king is turned to his right, facing a throned bishop (or possibly abbot), in the panel alongside. The ecclesiastic, who is wearing an episcopal mitre, full length cassock, sleeveless tabard, tippet and hood, is holding a book, and faces the king with similar upturned palm in greeting. To the right of these two figures is a standing huntsman, blowing on a large horn, which appears to be ringed and flanged, and hence quite likely to be a rendering of the Savernake Horn itself. He is wearing a short unbelted tunic (bearing traces of green enamel), a hood with a long tail ( liripipe ) draped behind, pointed shoes, and has a strap over his shoulder. In his left hand he is carrying a long structure, either a sword (it 170 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 2 The hell end of the horn, showing preening hawks on a wide silver flange. A slit in the ivory, through which a carrying strap was once threaded, can he seen centrally, deep within the hell appears to have a wrist guard) or possibly cudgel. All the other panels are framed by simple triple arches above, without tri-foliate divisions, and each has a broadleaved tree within it, drawn to different stylized designs. The panel on the other side of the horn to the King, which would hence face witnesses when it was raised to the lips, has a large sitting lion, with spreading mane and sweeping tail, perhaps conveying regal authority as king of the jungle (Figure 4). Apart from one panel, which bears a prancing unicorn, presumably of heraldic and iconographic significance, all others depict beasts of venery or hunting dogs, positioned around the horn anticlockwise. These include a fox to the right of the lion, male deer with fine antlers, deer without antlers (females, young males or deer which have cast them), and hare. A second band of silver, slightly smaller than the one at the bell, is set a short distance below it. Its enamel is better preserved than the upper ring, and provides a deep-green forest backdrop to each scene. The band bears a silver loop for the carrying strap. Apart from one panel which again shows a unicorn (bearing a massive twisted horn), the panes depict hunting dogs or quarry. The latter include deer, hare and squirrel. In each pane there is again a single stylized tree behind the creature (Figure 5). A third band of silverwork is situated in the middle of the oliphant. This, together with the mouthpiece, is a later addition to the horn, and dates from the 16th or 17th centuries, although it may be a replacement for 14th century originals (Cherry 2005). One panel bears a further loop for attachment of the carrying strap, whilst the other fifteen each THE SAVERNAKE HORN 171 Above: Fig. 3 The largest band of silver, showing those panels facing anyone using the horn, comprising a king greeting a seated bishop, huntsman, dogs and deer. Below: Fig. 4 The largest band of silver, showing panels facing away from the user, comprising deer, lion, fox? and dogs 172 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 5 The central hand of silver, with panels showing dogs, deer and a unicorn bear a small picture. These include a huntsman blowing a horn (in a similar posture to the one on the first ring, although differently dressed), a man on a bucking horse, hunting dogs, deer and hare, with single trees behind. The baldric Chapes (hook plates) were used to attach the baldric or strap to loops on the second and third metal bands of the horn. One of the chapes, depicting a resting deer with fine antlers, appears contemporary with the 14th century silverwork of the horn itself (Figure 6). However all other parts of the baldric, including the leather and metalwork, are either replacements or essentially unrelated items which have been incorporated. The baldric comprises two elements. There is an H-shaped piece of leather, the lower arms of which bear the chapes attaching to the horn (Figure 7). The two sides of the H are connected by a hinged clasp, engraved and embellished with enamel, showing a moth at the top (well drawn, with wide segmented body and appropriate antennae), long-billed birds resembling cranes in lozenges each side, and lion couchant at the base. A total of seven pieces of metalwork, comprising the second chape, the hinged sides of the clasp, buckles at the top of the H, and central medallions, are all decorated with the same shield and coat of arms. This H-piece in turn attaches by the buckles to a shoulder strap, which has six more medallions of the shield. The arms on the baldric are those of the Earls of Moray, comprising three cushions lozenge-ways within a Fig. 6 The single surviving original chape of the baldric, showing resting deer, with the added arms of the Earl of Moray above double tressure (MacDonald, 1904). There is fine tracery around the shield. Ward (1860), drawing on Laing’s Scottish Seals (1850), considered them likely to be those of Thomas Randolph, first Earl of Moray (d 1332), kinsman (probably nephew) and combatant supporter of Robert the Bruce, who was also a commander at Bannockburn, and Scottish Regent after the death of Bruce. Historical references The first written and pictorial accounts of the Horn appear on documents dated within three years of each other. It is depicted on the great illuminated Seymour Roll Pedigree which is dated 1604 (Figure 8). It is out of keeping with other subjects featured on the roll, being much larger (36cm x THE SAVERNAKE HORN 173 Fig. 7 The leather H-piece of the baldric, with central hinged clasp, and medallions showing the Moray shield 35cm), shown without title or identification, and unexplained concerning any relevance it might have to Seymour honours and genealogy. Its position off- centre, disrupting the balance of the roll, raises the possibility that it could be a later addition (Bathe & Douglas, 2012, this issue). However, Hertford was eager to display his noble and royal ancestry and may have sought to emphasize or exaggerate any prestige deriving from objects such as the horn. The image on the Roll Pedigree shows certain differences from the horn surviving today. The bell end shows the etchings of preening hawks on the interior, rather than around a large flange. The largest band of silver shows four human figures etched in panels alongside each other, instead of three, apparently confusing elements from bands 1 and 2. Other panels are not shown in the correct order. In addition the central clasp of the baldric is depicted disproportioned and twice its correct size. The imagery of the lion and other creatures varies slightly from the existing 174 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 8 The first pictorial representation of the horn, on the Seymour Roll Pedigree, dated 1604, where it is shown without title or explanation, and with a baldric different from the one surviving today (WSA 1300/376) clasp, whilst the moth has been depicted as a bird with wings raised. All of these differences are likely to be draughtsman’s errors. Other elements of the baldric however may reflect genuine differences from the item present 400 years later. Both chapes are shown depicting resting deer (as if original), whereas the surviving baldric has the left chape replaced by arms identified as those of the First Earl of Moray. The clasp connecting the two sides of the H-piece is depicted with wide serpentine links instead of today’s butt-hinges. Both the H-piece, and the shoulder strap attached, are depicted in green fabric with elaborate gold designs, unlike the plain leather of today. The strap is shown with 16 medallions of the Moray shield, whereas the current strap has six. The inference is that this image shows an earlier version of the baldric, yet nevertheless not original nor contemporary with the horn, and not commensurate with its symbolism. The horn is first described in the 1607 Latin edition of William Camden’s major topographical work, Britannia, and repeated in the English edition THE SAVERNAKE HORN 175 Fig. 9 The first technical representation of the horn and baldric, in an engraving by J. Basire from Archaeologia in 1 786 of 1610. He states ‘Wolfhall . . . was the house of the noble familie of Seimor, now Earle of Hertford, or of Saint Maur to whom by marriage accrewed a great inheritance of the Esturmies in this tract, who bare argent three demy-lions geules: and from the time ofKingHenrie the second were by right of inheritance, the bailiffes and guardians of the Forrest of Savernac lying hard by, which is of great name for plentie of good game, and for a kind of Feme there, that yeeldeth a most pleasing savour. In remembrance whereof their hunters home of a mightie bignesse and tipt with silver, the Earle of Hertford keepeth unto this day as a monument of his progenitours’. The fact that Camden failed to mention it in his earlier 1586 edition suggests that he may have had his attention drawn to it through the Roll Pedigree being compiled for the Earl of Hertford and completed in 1604. Camden was Clarenceux Herald of the College of Arms, and would have known those compiling the pedigree, and might have had a professional role himself. Camden, however, did not specify that the horn was a symbol of tenure. The first technical drawing of the horn was a detailed engraving by Basire showing its arrangement, carved silverwork and ornamentation in 1786 (Figure 9). It depicts two additional medallions on the H-piece, presumably subsequently lost or removed, but otherwise shows the horn as it appears today, suggesting that replacement of the baldric occurred between 1604 and 1786. The context of the Horn as a medieval symbol of tenure The Savernake Horn is anecdotally claimed to confer royal authority on the hereditary wardens of Savernake (initially the Sturmy or Esturmy family, and then through marriage, the Seymours, Bruces and Brudenell-Bruces). However Ward (1860) reported an alternative belief in a more recent origin, relating to the transfer of Savernake to the Seymours in the mid 16th century. Through Letters Patent of July 1547 the King gifted to his Uncle Edward, Duke of Somerset, Savernake and Braydon Forests, Vastern Park, the castle, borough and manor of Marlborough, the manors of Chelworth, Wootton Basset, Longbridge Deverill and East Monkton, the hundred of Selkley, and lands in Dorset, Warwickshire and Worcestershire in reversion after the death of Queen Katharine Parr, with Glastonbury Abbey, Syon Park, Covent Garden, and many other properties. However, the horn is not specified in the conveyance (Longleat House, Seymour Papers, Box XVIII). The horn is arguably the most spectacular of a series of medieval objects which have come to symbolise legal tenure of office or land, perhaps commemorating events when such privileges were 176 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE conveyed or granted. For example the Boarstall Horn was held by the hereditary wardens of Bernwood and Shotover. It was portrayed in a mid 15th century cartulary showing the King handing a horn and shield to a huntsman with one hand, whilst pointing to the land conveyed with the other. An inquisition of 1266 stated that the land had been held from before the Conquest through possession of a horn. In a legal dispute of 1361, it was claimed that Ranulf, Earl of Chester had conveyed the bailiwick of the Forest of Wirral, using a horn as evidence of title to that office (Cherry 1989). Similarly it was asserted that a horn was used to convey the wardenship of the Forest of Delamere, near Chester, to Ralph of Kingsley during the reign of Henry I. The Horn of Ulph, an eleventh century ivory oliphant probably carved in Italy and bearing images of animals including a unicorn, represented the tenure of land conveyed to York Minster before the Conquest. Its silver trimmings were stripped at the Reformation {ibid.). Romantic tales have become associated with horns of tenure. It has been claimed that the Boarstall Horn commemorates a grant by Edward the Confessor who gifted land to a man who had rid him of wild boar. And Wordsworth’s poem ‘ The Horn of Egremonf , building on an existing Lake District tradition, describes a ‘Horn of the inheritance ’ which, in this instance ‘ none could sound . . . save he who came as rightful heir'. If the image on the Savernake Horn portrays Edward III, this would not be the only prestigious hunting horn with which he was associated. He had granted to Sir John Foxle a bugle horn embellished with gold, which Sir John passed to Richard II through his will of 1378 (Cherry 2005). Sir John Foxle was keeper of the King’s Forests south of the Trent, and presided over inquisitions concerning the Wiltshire Forests. For example he was at Marlborough for this purpose in the first week after Lent in the 48th year of Edward III (PRO E32-318). It is virtually certain that the wardens of Savernake would have been aware of this prestigious item, irrespective of whether it was displayed as a symbol of office at such Inquisitions. Whilst it remains conjectural whether the silverwork of the Savernake Horn was commissioned in emulation of the Foxle ceremonial horn, perhaps from the same workshops, some link between the two is possible. The imagery of the Horn and its potential link with Savernake Some of the animals depicted cannot be identified with certainty. The stylised antlers of most of the engraved deer are not unequivocally palmate, and could be interpreted as either fallow or red. There are no clearly identifiable roe deer. Certain dogs, perhaps especially greyhounds, have collars and are hence unambiguous. However some canines could conceivably be wolf or fox. There are no wild boar, and no obvious badger, marten or wild cat. The squirrel seems incongruous amongst creatures of the chase. However there are no animals depicted indicative of an origin distant from Savernake. The depiction of the King and Bishop, facing and greeting each other, suggests that the horn is commemorating an agreement. References to interactions between religious foundations, Forest officials and the king abound in the medieval annals of Savernake. There are more records surviving from the 13th than 14th centuries. The earliest account relates to 1208 when Hugh Neville, Constable of Marlborough and warden of Savernake was commanded to hang from the nearest oak anyone harming or defaming religious men (WSA 1300/6574). Church leaders holding land in the Forest included the Prior of Easton (Priory Wood near Bitham Pond), Abbot of Hyde (Withy Copse), Prior of Mottisfont (near Leigh Hill), Prior of St Margaret’s (near Postern Gate), Abbess of Wilton (Rainscombe, Overton Heath), and Prior of Okeburn (County Gate where Hampshire, Wiltshire and Berkshire met). Grants were frequently made to such institutions. The Abbess of Wilton was allowed to take 100 fallen oaks in 1246, six score of oak rafters in 1231, and a further 40 oaks in 1251, and then another 60 in 1299 ‘to rebuild therewith certain houses in the abbey that were lately burnt by mischance ’ (Close Rolls). Timber could also be sent further afield. In 1232 a minor order of brothers were allowed to take joists for building their church at Salisbury without cheminage (toll for carriage) or other interference (Close Rolls; Patent Rolls). Oaks for making thirty rafters were granted to Friars constructing a chapel at Oxford, again without impediment or cheminage , and the brothers of Bello Loco (Bedfordshire) were granted 20 oaks for their church (Close Rolls). Savernake timber was also used for constructing THE SAVERNAKE HORN 177 the cathedral and city of New Sarum (Salisbury). In 1269 Henry Esturmy was mandated to allow ten oaks to be taken for the tower of Salisbury from the ‘ wood ofBedwyn within Savemake Forest ’ (presumably the Brails), and two years later Nicholas Longespee (later Bishop of Salisbury) was allowed ten timber oaks from Bedwyn Brail, as gift of the King (Close Rolls). The cannons of the Gilbertine Priory of St Margaret’s benefited especially from the King’s patronage. From 1222 to 1234 he granted them 30 rafters to repair their premises, 30 standing oaks to be taken where ‘this would cause least disturbance ’ for works at their church, timber for their belfry, logs for a forge, 50 tie-beams ( copulas ), and oaks to make roofing shingles. In 1234 ‘ to further the salvation of our souV the King granted additional land to the Priory for the betterments of their house and enlargement of the cemetery ’ (Close Rolls). In 1224 they were granted daily permission for one man to go on foot gathering sticks, and in 1232 this was extended to the collection of dry and dead wood for their hearth by use of a single horse. The brothers and sisters of the Priory and Hospital of St John the Baptist, in Marlborough also benefited. They were granted 20 rafters for extending and repairing their property in 1223, and 10 tree trunks for timber in 1236 (Close Rolls), with a right to have for ever one man going daily for dry and dead wood collecting as much as he could by hand without any iron tool or axe, and to carry the same to the hospital on his back for their hearth (WSA 1300/71; Patent Rolls). Religious houses were also often granted rights of common. In 1250 when, for the well being of his soul and the souls of his ancestors, Geoffrey Esturmy granted to the Hospital of Easton and the ‘ brothers serving God there , 50 acres of his wood in Savernac, viz in Halegodesfolegdene , without interference by the foresters' he also gave them full common of pasture for all kinds of beasts in Savernac'' (WSA 1300/1). In 1270 Henry III had granted the Priory of St Margaret’s at Marlborough the right to depasture sixteen oxen and four cows in the Forest of Savernake, excepting the lawns, without having to pay the usual fine (known as exchapium ) if the cattle accidentally strayed, but instead driving them back into the Forest pasture for ever ’ (Fry 1908; Close Rolls). In 1344 Queen Philippa granted to the Canons of St. Margaret’s, in lieu of a customary portion of ale called toltestre (as granted in a charter of King Henry), common of pasture for 100 demesne swine, quit of pannage, together with sties in the Queen’s Forest of Savernake at all time of the year except the fence month, and except in her lawns there for ever ’ (Patent Rolls). Religious leaders were frequently granted concessions to hunt, or received gift of venison. The Bishop of Salisbury received six deer for the Feast of the Nativity in 1224. The Bishop of Durham was sent five or six deer on various occasions from 1228 to 1235, including for Pentecost and Quadragesima. The Bishop of London was given four deer for a Consecration held in 1244 (Close Rolls). Disputes also arose with religious institutions, sometimes requiring the king’s intervention. In 1232 the warden seized property of the Abbess of Wilton, following transgressions and failure to expeditate dogs (maiming to prevent hunting). The Abbess made representations, and the king pardoned her men, and ordered the lands, woods, beasts and chattels to be returned without delay. In 1236 Forest officials captured and imprisoned men from the Priory of Saint Margaret’s who were gathering thorns in Savernake, for hedging their monastery. Again the King ordered their release (Close Rolls). However, all such grants and episodes relate to the sort of assistance or squabbles that might arise between any neighbours. Whilst there were long standing disputes between bishops and the monarchy, as ecclesiastical lands were afforested, and clergy were imprisoned or fined for trespasses against the vert and venison, any agreements recorded for Savernake were trifling in scale. However certain concessions from which the church could benefit may be relevant. The Carta de Foresta of 1217, reissued several times in the following century, permitted every bishop travelling through the forest to take one or two beasts by view of the foresters, or to blow a horn to give notice if they were not present. The church could also benefit from tithes. In the 29th year of the reign of Edward III, the Dean and Chapter of Sarum made a claim in which they exhibited a charter from the time of King Henry granting them tithes of the Forests of Wiltshire (WSA1300/75). This became a matter of major significance for the wardens, since it enabled them to resist demands for tithes from local churches as vast tracts of land were converted to agriculture for half a millennium. However, there is no evidence that the horn was ever cited in evidence at the many legal disputes concerning tithes. Hence, whilst these issues show that the clergy and the monarchy frequently interacted concerning the Forests, it is uncertain whether any agreements were of sufficient gravitas to deserve artistic commemoration, nor be recognised in the fourteenth century of particular significance to the hereditary wardens of Savernake 178 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE who held or acquired the horn. What is clear however, is that the silverwork of the horn was crafted during a period when the Forest system was undergoing substantial upheaval, something promised for more than a century. In 1330, whilst Edward III was still in his minority, a concession to the nobility was granted as new perambulations were ordered, contracting Forest bounds to those of 176 years before, when Henry II had been crowned. At the Salisbury Forest Eyre of 1330 nearly ninety per-cent of Savernake was removed from Forest Law (WSA 1300/65, PRO E32- 220, PRO El 46/2/34). This was a catastrophic blow to the Forest officials who had been able to exploit their royal authority for private gain. Other Forests were similarly treated, or threatened with removal from Forest Law altogether. Extensive disafforestation was also resented by Edward III when he came of age, and he struggled to re-gain his authority over the lands. The wardens of Savernake, who had secured great personal advantage by their office, endeavoured to minimise any losses. The wardenship was a service associated with the manor of Burbage. It conferred privileges across extensive areas, which the wardens sought to retain, irrespective of any changes in Forest bounds. These included rights to pasture, timber, hunting and the administration of justice. For centuries the wardens would continue to hold Courts which presided over disafforested lands as if any revision of the bounds need not affect them. A symbol of office, implying conferment of royal and ecclesiastical authority, would have been a great asset during such parlous times, and enable the hereditary wardens to maintain social standing throughout lands ostensibly removed from their jurisdiction. Whilst there is no evidence that this is why it was commissioned, it is clear that the horn was fashioned during a period of collapse of the Forests in general, massively so at Savernake, of concern to both the Wardens and King, and that both were seeking to retain their economic and political advantage. There is a tradition that a ceremonial blast on the horn was issued by a huntsman in livery, as a salute to visiting monarchs. It may have been sounded when King James visited in 1603, at the same time that the Seymour genealogy was in preparation (Brudenell-Bruce, 1962). It was reported that, after an otherwise pleasing visit to Savernake, George III reprimanded the Marquess of Ailesbury for failing to 'blow your tenure horn’, stating that 4 When a King of England comes here , and that horn is not blown, your property becomes forfeited to the crown!’ . The Marquess Fig. 10 The 29th Hereditary Warden of Savernake, Chandos Brudenell-Bruce, 7th Marquess of Ailesbury, with the horn at Sturmy House, Durley, in 1961. Photo: Marie Dixon was forgiven when he reposted 'Well sire, I couldn’t get anybody to get any sound out of it’ (Britton, 1801). Chandos Brudenell-Bruce, the 29th hereditary warden of Savernake, who called it the 'Esturmy Horn’, claimed it was last so used to greet King George VI in the year 1940 (Brudenell-Bruce, 1949). He is shown displaying the horn at Sturmy House, Durley, in Figure 10. Linking the Scottish heraldy with Savernake The attachment of a Scottish baldric with an English hunting horn poses a dilemma. There is a little circumstantial evidence linking the horn with the Scottish emblems it embodies. The Bruce family who inherited Tottenham through the female line from the Seymours, traced their ancestry from Robert the Bruce, uncle (or possibly otherwise related) to Thomas Randolph. A genealogical THE SAVERNAKE HORN 179 Fig. 11 Seals of the Sturmy/Esturmy family: left, Geoffrey Sturmy of Glamorgan, cl 160 (WSA 1300/6574); middle, Henry Esturmy, from conveyance of Burbage, 1321 (WSA 1300/16); right, armorial seal of Henry Esturmy of Wolfhall, 1338-1355 (WSA 1300-6574) history of the family emphasizes the connection with the Scottish king by depicting a spider on its cover (WSA 1300/2708), and the Seymour Roll Pedigree includes a narrative concerning Robert the Bruce (WSA 1300/376). Further tangential evidence relates to two swords formerly held at Tottenham. One, bearing the arms of Scotland, was reputed to have belonged to Robert the Bruce. The other, with a hilt inlaid with silver, bears the heraldic devices of the Black Douglas. The blade is dated 1314, with a poem engraved in couplets on each side: This is the sword that once was worn / By the Black Douglas at Bannockburn / By Bannockburn I served the Bruce / Whereof the English made little use (Ward 1860). What is certain is that the heraldic medallions were already emblazoned on the baldric attached to the horn in 1604. The Earl of Hertford (then owner of Savernake as well as hereditary warden) could have replaced the item with his own devices, but chose not to, suggesting that the shields were seen as integral. Further, when the fabric of the baldric was replaced with a leather strap by 1786, the medallions were transferred, again indicating their importance to the family. Provenance of the Savernake Horn No link between the horn and Savernake in medieval times has been proven. The horn’s history, from the 12th century carved ivory, until 1604 when it appears on the Seymour Roll Pedigree, is unknown. Evidence connecting it with the Seymour and Esturmy family is circumstantial. It is known that the symbol of a hunting horn was closely associated with the hereditary wardens. The seal of Sir Henry Sturmy featured a hunting horn in a conveyance of Burbage, Durley and Timbridge in 1321, and a horn of a different design was displayed on his armorial seal used between 1338 and 1355 (Figure 11). One branch of the Sturmy family used a hunting horn on their seals even earlier. Geoffrey Sturmy, Lord of Glamorgan, had a seal showing a man blowing a horn in about 1160. He had settled in Wales, creating an estate out of the waste ‘ where no man had ploughed before’. He also built a chapel at the eponymous Sturmieston (Davies 1991). The ruins of a motte and bailey known as Stormy Castle, which was bequeathed by the family to Margam Abbey, mark the location of his property. The names Stormy Farm and Stormy Down have persisted into the 21st century. Whilst a genealogical link between the Glamorgan and Savernake Sturmys is unproven, it is quite possible. There may also be a connection with the fact that the tenure of Savernake was held on obligation of providing military service in Wales. For example, Henry de Stormy who died in 1295 held Burbage and Durley by the serjeanty of finding one armed squire in the King’s army in Wales, and keeping the bailiwick of the King’s Forest of Savernake (WSA 1300/6574). Furthermore, a Geoffrey Esturmy was unable to participate at 180 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE one Forest Plea (probably in 1245) because he was fighting in Wales, where he died (WSA 1300/6574; PRO E146/2/33). A horn was clearly part of the symbolic equipment of the Chief Warden, who would appoint rangers and foresters. From 1295 till 1305 Henry Sturmy held Burbage and other land by the service of guarding Savernake, and also by providing one man armed with a hauberk, for military duty within the seas. By virtue of this guardianship he claimed rights and appurtenances, including the right to the equipage - namely saddle, bridle, sword, and horn of the foresters of fee, in the event of their death or departure (WSA 1300/6574). These privileges of equitatura or mounture were re-iterated at a Plea of the Eyre of 1490 (WSA 1300/153A). Whilst hunting horns have held symbolic significance for the Sturmy family since at least the 1100s, there is no positive evidence that the Forest has any specific link with the ornately crafted item now bearing its name. A family with an interest in such iconography might have sought to acquire or emulate impressive items seen elsewhere, such as the prestigious horn granted by Edward III to Sir John de Foxle. The horn’s emergence from historical oblivion as late as 1604, and the nature of its depiction on the Seymour Roll Pedigree, militate against any ancient associations. Other images on the Roll Pedigree are portrayed in chronologically relevant locations, whereas the horn is shown near the base. The Pedigree cites evidence to demonstrate the ancestry, honours and inheritance of the Seymour family. However the horn is simply portrayed un- named and without any claim of significance. It is incongruous that it would not be prayed-in-aid, had it been important, like every other evidence cited. The incompatible Scottish baldric, with its purported affiliations with the Earl of Moray, can be plausibly related to the hereditary wardens’ ancestry through Robert the Bruce, but the issue again remains inconclusive. Tantalizing uncertainties concerning any ancient link between Savernake Forest and the Horn persist. It remains possible that it was fashioned for Savernake in the 1300s, at a time of upheaval to the Forest system, possibly embellishing an earlier family heirloom. However, such an attribution stands, not through direct evidence, but because no satisfactory alternative can be found. Acknowledgements Images of the Savernake Horn in Figures 1-7 are copyright of the author, reproduced for non- commercial purposes by courtesy of the Trustees of the British Museum. Images of the Seymour Roll Pedigree (WSA 1300/376) in Figure 8, and Esturmy seals (WSA 1300/16 & 1300/6574) in Figure 11, are reproduced courtesy of Wiltshire and Swindon Archives. The photograph of Chandos Brudenell- Bruce in Figure 10 appears by courtesy of Marie Dixon. I would like to thank Steve Hobbs and Joan Davies for many contributions, and Andrew Douglas for investigating the heraldry and significance of the horn on the Seymour Roll Pedigree. References Primary Sources National Archives ( Public Records Office, PRO) E146/2/34, Perambulation of Savernake, with vills and groves disafforested E32-220, Perambulation 1330-31 based on bounds in the time of Henry III, with vills and groves disafforested E32/3 1 8 : Inquisitions concerning the state of the Wilshire Forests, 35-50 Edward III, including inquisitions on Savernake held before John de Foxle, Keeper of the Forest south of the Trent, at Marlborough, 44 Edw III, 48 Edw III, and 50 Edw III. E146/2/33: Pleas of the Forest held at Wilton, 1245 Wiltshire and Swindon Archives (WSA) 1300/1: Gift by Geoffrey Esturmy to hospital of Easton there, of 50 acres of his wood in Savernake, in Halegodesfolegdene, with rights of common in Savernake. 1300/16: Conveyance by Henry Esturmy of his manor of Burbage, hamlet of Durleigh pasture of Tymerigge, and bailiwick of Savernake 1300/65, Metes and Bounds of Savernake forest 14th century, divided into sections for towns, woods and groves outside forest; 16th - early 17th century copy 1300/71 : 17th century copy of letters patent, to friars of St John’s hospital, Marlborough, giving them leave to gather fuel in Savernake. 1300/75: Searches in the Exchequer chamber relating to Savernake Forest 1300/153A: Plea of the Eyre held at Amesbury. Claim by John Seymour, esq., because of his tenure of the manor of Burbage to the chief Wardenship of Savernake 1300/376: Seymour Roll Pedigree, dated 1604 1300/2708: Genealogy of the Family of Bruce, printed THE SAVERNAKE HORN 181 pedigrees, notes on family history, coats of arms, illustrations and portraits. 1300/6574: Notes, abstracts and copies of material relating to the families of the Earls of Ailesbury compiled from the family and public records. Longleat House Letters Patent 23 July, 1 Edward VI (1547), conferring lands on Duke of Somerset, Seymour Papers, Box XVIII. Printed Calendars Calendar of Close Rolls preserved in the Public Records Office, 2006. Ontario: Tanner Ritchie & University of St Andrews Patent Rolls preserved in the Public Records Office, 2004. Ontario: Tanner Ritchie & University of St Andrews Secondary sources BATHE, G. and DOUGLAS, A. 2012. Forging Alliances: The Life of Edward Seymour, Earl of Hertford, and his commissioning of the Great Illuminated Pedigree of the Seymours, and Monumental Tombs in Wiltshire and Westminster. WANHM 105 (this issue), 182-218 BASIRE, J. 1786. The Savernake Horn (engraving). Archaeologia 3, 24-29 BRITTON, J., 1801, Beauties of Wiltshire displayed in statistical, historical and descriptive sketches. London: Vernon and Hood, J. Wheble, J. Britton BRUDENELL-BRUCE, C., (Earl of Cardigan), 1949, The Wardens of Savernake Forest. London: Routledge and Keegan Paul BRUDENELL-BRUCE, C., (Marquess of Ailesbury), 1962, A History of Savernake Forest. Devizes: Privately published CAMBER, R. and Cherry, J., 1977, The Savernake Horn, in R. Camber (ed.), The British Museum Yearbook 2, Collections and Collectors. London: British Museum CAMDEN, W., 1607, Britannia, siveflorentissimorum regnorum Angliae, Scotiae, Hiberniae, et insularum adiacentium ex intima antiquitate chorographica description. London: George Bishop and John Norton CHERRY, J. 1989. Symbolism and Survival: Medieval Horns of Tenure. Antiquaries Journal 69:1, 111-118 CHERRY, J., 2005, ‘The Savernake Horn An Oliphant Adorned with Metal’, in R. Bork, S. Montgomery, C. Neuman de Vegvar, E. Shortell and S. Walton, S., 2005, De Re Metallica: the Uses of Metal in the Middle Ages. Aldershot: AVISTA Studies in the History of Medieval Technology, Science and Art, volume 4 DAVIES, R. R., 1991, The Age of Conquest: Wales, 1063-1415. Oxford: University Press FRY, E. A., 1908, Abstracts of Inquisitiones Post Mortem, Henry III, Edward I and Edward II, 1242-1326. London: British Record Society MICHAEL, W., nd. (c. 1900), Savernake Forest. Westbury: W. Michael LAING, H., 1850, Descriptive catalogue of impressions from ancient Scottish seals. Edinburgh: Edmonston and Douglas MacDONALD, W. R., 1904, Scottish Armorial Seals. Edinburgh: William Green and Sons WARD, J. 1860. The History of Great Bedwyn. WANHM 6, 261-316 Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 182-218 Forging alliances: the Life of Edward Seymour, Earl of Hertford, and his commissioning of the Great Illuminated Roll Pedigree of the Seymours and monumental tombs in Wiltshire and Westminster by Graham Bathe1 and Andrew Douglas2 Edward Seymour, ; Earl of Hertford (1539-1621), held extensive lands in Wiltshire, at Savernake Forest ( where he was hereditary warden), Amesbury, Bedwyn, Easton, and Wolfhall. His family had achieved substantial advancement through marriage and royal connections: his aunt, Jane Seymour married Henry VIII; his uncle Thomas married Henry’s widow, Queen Catherine (Parr); whilst his father, as Protector and guardian of the child king Edward VI, held virtually all the powers of a monarch. Hertford’s expectations of inheriting massive wealth and titles were all-but obliterated by his father’s execution. Hertford dedicated much of his life to exhaustive and often fruitless attempts to recover his own position and advance that of his sons. The great roll pedigree of the Seymours, an illuminated parchment manuscript 6.83m long, and a series of sumptuous monuments in Westminster, Bedwyn and Salisbury, were designed to trumpet Hertford’s noble connections and broadcast a selective and sanitised view of history. These, together with an entertainment devised for Queen Elizabeth at Elvetham, include some of the most opulent examples of their kind ever produced, and testify to the grandiloquence of a man often embittered by unfulfilled ambitions, striving to assert his eminence. A Brief Biography of the Earl of Hertford The commissioning of the Seymour Roll, and of the various monuments to Hertford’s family, needs to be seen in the context of his long and often troubled life, including the legal judgments against him, and his own lawsuits spanning decades. There is no thoroughly researched biography which focuses on Hertford, although he features in many, often romanticised, accounts concerning Catherine and her sister ‘Queen’ Jane Grey. There are few images of him known; the effigy from his tomb in Salisbury cathedral is depicted in Figure 1. Hertford’s life was characterised by failed aspirations, and lengthy attempts to recover losses of lands and status. The key stages are shown on the timeline in Figure 2. These can be grouped as follows: The rise and fall of Somerset Hertford’s father (also Edward Seymour) was created Duke of Somerset at the coronation of Edward VI, and became Protector of the Kingdom and Dominions during the minority of the king. There is little doubt that Somerset manipulated the boy King Edward VI, who was induced to sign Letters Patent granting lands ‘ to my dearest Uncle Edward’ (WSA 1300/166). In 1547, he was able to secure the Royal Forest of Savernake, across which his family had been hereditary wardens since at least 1130. Somerset held the right to impark and enclose Savernake or other property as he willed. He was granted the Byeley-in-Densome, Woodgreen, Fordingbridge, SP6 2QU 2 48 Woodcote Road, Caversham Heights, Reading RG4 7BB FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 183 Fig. 1 Edward Seymour, Earl of Hertford, from his effigy in Salisbury Cathedral extraordinary authority to prevent any escheator, sheriff, bailiff or other officer or servant of the Crown entering upon the property or doing anything against his liberties and privileges (WSA 1300/166). By the fourth year of Edward VFs reign, in the vicinity of Savernake, Somerset had consolidated with holding with the Collingbourne manors of Kingston, Dormer, and Sunton, plus Hippingscombe, Pewsey, Shalbourne Eastcout and Westcourt, Burbage Darrell, Easton Warren, Manton, Elcot Mill, Castle Mill, Barton, together with the castle, borough and manor of Marlborough (WSA 1300/167). Somerset was deposed in 1549, but restored to the Council the following year, before being arrested again, tried and executed in 1552. He was convicted of felony rather than treason. However shortly after his execution, through the action of his enemies, an Act was drawn up ‘ touching and concerning the limitation of the possession and hereditaments of Edward, late Duke of Somerset’ (WSA 1300/174). The Act acknowledged that the King and others had suffered loss, and those whose property he had obtained 'by extort, power or contrary to justice or equity ’ were to be recompensed. However the Act required his children to be maintained, which needed their rightful inheritance to be identified. In considering how to distinguish between those properties legitimately acquired, and those sequestered unlawfully, Parliament chose the 32nd year of Henry VIII (1540), when Seymour had petitioned Henry VIII concerning his estate. Anything acquired after this date was deemed to have been extorted, and forfeited to the King (WSA 1300/174; CSP Vol. 1 1547-1580). As a consequence Somerset’s son aged 13, lost the major part of his anticipated inheritance, his prospect of marrying royalty (his father had contemplated a match with Lady Jane Grey), and both the dukedom and earldom, in addition to a parent. Adjustments concerning Savernake and the Wiltshire Estates Because Savernake was acquired after 1540, the critical date in the Act of Limitation, Somerset’s son (later: Earl of Hertford) was dispossessed of it. Thus it might have ended. However the Act required his son to be conferred with estates to the values of those held in 1540, and some of the lands seized had already been transferred and were no longer available. In accordance with the Act, Somerset’s son was entitled to recompense for any properties lost. In January 1553 the Lord Treasurer re-assigned to the young Edward the Forest of Savernake, and many of the neighbouring manors and lands held before, including Collingbourne, Hippingscombe, Pewsey, Marlborough, Shalbourne, Burbage, Bedwyn Easton and Grafton (WSA 1300/163). However, The Brails and Southgrove, which had already been conveyed by the Crown, were lost. Hertford and Grey’s putative marriage Shortly after Elizabeth’s accession in 1559, Seymour was restored in blood, and created Earl of Hertford anew (2nd creation). Having made a partial recovery of his position, Hertford hazarded everything through a relationship with Catherine Grey (the sister of executed ‘Queen’ Jane Grey) who, under the will of Henry VIII, was arguably lawful successor to the crown. In August 1561, at a time when Hertford was in France, Catherine found that she could no longer hide her pregnancy, and confessed that the father was Hertford, whom she had married secretly without royal consent. A furious Elizabeth wrote to her Ambassador in France concerning Hertford: 'Ye shall in our name command him to make his speedy repair hither to our realm ’ (BL Hardwicke, Add.35830; fl 83). Cecil explained to the Ambassador 'Yesterday, 11 August, the Lady Catherine Grey discovered herself to be with child and allegeth it gotten by the Earl of Hertford, 184 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Edward Seymour, Earl of Hertford 1539- 1621 SOMERSET EXECUTED. Hertford disinherited from lands and titles. MARRIAGE. Hertford m. Catherine Grey. Both seized when Catherine pregnant. Commission declares union unlawful. IMPRISONMENT. Hertford confined 10 years ATTEMPT TO RECOVER LANDS. Hertford commences lengthy legal case to recover lands unlawfully taken when he was child. LEGITIMACY. Hertford’s sons repeatedly file against bastardisation and disinheritance. REMARRIAGE. Hertford m. Frances Howard MONUMENTS. Hertford erects opulent monuments to mother (1587), grandfather (1590), sister (cl590), 2nd wife (1598) IMPRISONMENT. Hertford in Tower and fined £5000 for claiming his marriage lawful ELIZABETH DIES. James becomes King. Beauchamp’s support withers. RELINQUISHMENT. Hertford accepts lawsuit is hopeless after 47 years, and purchases reversion of lands his sons would have inherited if legitimate. King agrees for Earldom to be created anew on his death. DEATH. Hertford dies. Massive monument of unknown date later erected in Salisbury. 1547 1550 1560 1570 1580 1590 1600 1610 1620 1621 EDWARD VI CROWNED. Hertford’s father created Duke of Somerset Beauchamp and Thomas born in Tower, bastards disinherited from titles and both parents’ lands £15,000 FINE. Hertford repeatedly tries to reduce fines. Land distrained. JOHN SEYMOUR PEDIGREE, by Dethick. BEAUCHAMP’S BETROTHAL. Hertford tries to thwart son’s liaison with commoner. Discovering Beauchamp already wed, Hertford seeks to nullify marriage and imprisons son. ELVETHAM. Hertford arranges lavish entertainment for Queen at Elvetham. TURMOIL ABOUT SUCCESSION. Several claimants. In 1595 Lieutenant of Tower stockpiles weapons in Beauchamp’s cause. SEYMOUR PEDIGREE, completed 1604 MARRIAGE mooted between Arbella Stuart and Hertford’s eldest grandson SECRET MARRIAGE of Arbella Stuart & Hertford’s second grandson, William. BEAUCHAMP DIES. Tiny plaque, Bedwyn. Fig. 2 Timeline showing main stages in Hertford’s life FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 185 to whom also she sayeth that she was married before Christmas last at Cannon Row, in the presence only of the Lady Jane [Hertford’s sister], now deceased, and by a minister to her unknown. This matter is very noisome in diverse ways. The Queen’s majesty is much offended. The Lady is carried by Sir William Saintlow to the Tower’ (BL Hardwicke, Add.35830; fl85). ‘ Nobody can be found privy to the marriage. The Queen and others suspect some greater drift, but I can find none’ (WSA 3790/1/3; (BL Hardwicke, Add.35830; fl 85). At his return and incarceration Hertford complained at machinations to prevent him speaking with the Queen (BL 64, f 1 93). The following month Catherine gave birth to a son within the Tower. In the chapel of St Peter ad Vincula , where his grandfathers Somerset and Suffolk had been interred after execution, he was christened Edward Seymour, thereafter styled Lord Beauchamp. The birth of a potential male contender for the throne raised massive constitutional issues. In January 1562 the Archbishop of Canterbury, the Bishop of London and others were commissioned to investigate the circumstances of the intercourse between Hertford and Grey (CSP Vol. 1 1547-1580; BL Add. 63543, £27). Detailed and intimate enquiries were made concerning their relationship, ranging from where they lay in bed, to what she wore on her head. The interrogatories were specified for each party, who were challenged separately with leading questions (implying a pre-determined outcome) about 'the pretended solemnization’ and 'the pretended marriage.’ The interrogators noticed certain similarities in phraseology, (eg both repeatedly described the beard of the priest as 'auburn’) as if they ‘grownd themselves upon a form of words’ but noticed inconsistencies, and found certain things implausible. They thought it unlikely that Hertford’s sister Lady Jane, aged just 19, who was 'a woman of singular calling, should go abroad out of the house [alone] to fetch a priest ’ or that such a priest should be readily located, and conveniently available and willing to act. They also thought it unlikely that a jointure which Hertford purported to sign giving Catherine £1,000 per year, which (if located) would have provided documentary evidence, should have been casually entrusted by Catherine to a maid-servant unfamiliar with their relationship, who then happened to lose it (NA SP 46/10; BL 33749 ff.92; BL Cotton Faustina A XI; BL Add. 33749). The fact that the sole witness, Lady Jane, had died, and that the priest remained unknown, weakened any case. In May 1562 it was formally declared that no marriage had taken place, and their union was unlawful (NA SP 46/10). This pronouncement doubtless provided a satisfactory solution for the State, since any children were theoretically ineligible for succession (although the Queen would have been conscious that her own illegitimacy had not prevented her assumption of the throne). At least twice in May 1562, presumably through bribing his gaolers, Hertford gained access to Catherine’s cell (Longleat House Portland Papers, Vol. 1, ff.92-3). She again fell pregnant, and gave birth to a second son, Thomas, in February 1563. Further retribution followed. Hertford was fined £15,000 in the Star Chamber, comprising £5,000 for deflowering the Lady Catherine Grey, £5,000 for breaking out of his cell and visiting hers, and £5,000 for re-iterating the said vicious act, together with imprisonment during the Queen’s pleasure (BL Sloane 1786 f.25; BL Cotton Titus C VII f.llb). Most contemporary and subsequent commentators have accepted that Hertford and Grey were indeed married, as they claimed, based solely on the testimony of the couple themselves. When counterclaims were produced stating that their sentence was unjust, such as that produced by John Hales who argued that Catherine was the rightful heir to the crown, it strengthened the Queen’s resolve to prevent her or Hertford’s restoration to favour (CSP Vol. 6 Addendum 1547-1565; BL Cotton Titus C VII, f.llb). The monarchy never acknowledged that a marriage had taken place, and Hertford, Catherine and both their sons suffered from the decision all their lives. As bastards, Beauchamp and Thomas were prevented from inheriting titles or entailed lands, like their father before them. Imprisonment Hertford and Catherine were transferred from the Tower following an outbreak of plague in 1563, and were thereafter held distantly from each other in the custody of a sequence of nobles (CSP Vol. 1 1547-1580). Catherine looked after their younger son, Thomas, whilst Beauchamp was in the care of Hertford’s mother, the Duchess of Somerset. In 1567 Catherine was removed to Cockfield Hall, Suffolk, where she was to be denied access to all strangers (CSP Vol. 1 1547-1580). She fell ill and died in January the following year. The Queen paid the expenses of embalming, employing Heralds, and the costs of mourners at a funeral befitting a person of royal blood, when Catherine was interred at Yoxford in February 1568 (CSP Vol. 1 1547-1580). Hertford was to remain under confinement for almost 10 years, during which time he endlessly pleaded to 186 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE be restored to the Queen’s favour, until his eventual release in July 1571 (Longleat House Seymour Papers Vol. V, ff.99-100d). During his confinement, Wolfhall became ruinous beyond restoration. Hertford’s Fine The sum of £15,000 imposed on him in the Star Chamber was crippling, and its redemption pre- occupied Hertford for over 16 years. During his incarceration Hertford’s officers were repeatedly cautioned. The Queen’s collector stated 7 marvel not a little ’ that Hertford’s receiver had not appeared before him to receive orders for transferring rents, arrears, fines and woodsales. The collector told Hertford’s tenants to submit sums, with the warning 4 Pray do not fail to pay, otherwise you will he troubled by distress at the sheriffs hands, for which I should be sorry for your own sakes,yet may not leave it undone for satisfaction of the Queen ’ (Longleat House Seymour Papers Vol. V ff.5-6). Hertford begged that the annual sum collected 4 be not so large that I and mine may not have living ’ (Longleat House Seymour Papers Vol. V f. 12). The following year the Queen’s collector complained that Hertford’s auditor ('a man of no experience ’) had again denied payment, and had said he 4 trusted that there would be no more demands for that money for the Queen. ’ Hertford was given three days to remedy the situation with a stark notice 4 what hurt might follow to you upon denial I know not, and therefore give you warning ’ (Longleat House Seymour Papers Vol. V f. 13). After Catherine’s death, the fine was reduced to £10,000 and the Queen agreed to accept £700 per year till paid (CSP Vol. 1 1547-1580). The fine was remitted to £5,000 on Hertford’s own release after 10 years. When he begged again, the Queen refused to consider anything until more was paid, whereupon he submitted £1187, and she released a further £1000, leaving £2813. Hertford’s private notes reveal his delusion that the fine 4 was set but for terror ’ (Longleat House Seymour Papers Vol. XI ff.42-46d). In 1572, Hertford again beseeched to be released the rest, claiming 'All men have tasted of your goodness but F (Longleat Seymour Papers V ff.l l-12d). After five more years nothing further had been paid, and Hertford had various drafts prepared in hope of an agreement remitting some or all of the £2,813 outstanding (Longleat House Seymour Papers Vol. V ff.l 4-41). The Queen agreed to remit £1000 so that Hertford 4 would trouble her no more for the rest’ (Longleat House Seymour Papers Vol. XI ff.42-46d). However, Hertford still defaulted, and in 1579, sixteen years after the fine was first imposed, his manors were distrained to ensure payment of the outstanding sum (Longleat House Seymour Papers Vol. V ff.47-68d). Hertford’s attempt to recover lands Through the Act limiting his father’s estate, Hertford lost most of his inheritance when he was a minor. All lands acquired by Somerset after the 32nd year of Henry VIII were seized by the Crown and redistributed. Hertford calculated that, at that time his father had lands valued at £2500 per year, rising to £7,500 by the date of the Act of Limitation in the 5 th year of Edward VI. The lands involved were so extensive there was ample scope for error and unjustified seizure, in addition to what Hertford perceived as an injustice. He estimated that Wiltshire profits erroneously taken from him as a child included the herbage, pannage and custody of Savernake, fees and windfalls there, herbage and pannage of the Brails and Southgrove, the rectory and woods at Amesbury, Grafton manor, Longleat, and Grovely. There was further revenue he had lost from lands in Somerset, Dorset, Lincolnshire, Buckinghamshire, North Wales, Chester and elsewhere. Hertford was particularly concerned about land worth £500 per year, which he perceived was legitimately his inheritance, and had been falsely taken, and transferred to Lord Wentworth (CSP Vol. 7 addenda 1566-1579). Whilst in confinement Hertford had little opportunity to remedy the situation. When his release was imminent in 1571, Sir John Thynne advised him that 'as your enemies seek all ways against you, consider your lands and lose no advantage the law will allow. If it should come to extremity appeal to the Queen’s majesty’ (Longleat House Seymour Papers Vol. V ff.99-100d). Elizabeth was sympathetic (Longleat House Seymour Papers V ff.ll-12d). However, Wentworth vigorously resisted Hertford’s claim and countered that Hertford had more land than he should have as a consequence of deliberate concealment (Longleat House Seymour Papers V f.69-70d). The case dragged on, and in 1580 Hertford complained that 'My suit to her Majesty is now nine years old ’ (Longleat House Seymour Papers Vol. V ff.69-70d). Hertford had explained that he could not 'for love or money’ provide categorical proof that his rightful inheritance had been embezzled, 'being then an infant ,’ whilst his father’s stewards who had best knowledge of the affairs, including John Thynne and Francis Newdigate, had been in prison and excluded from the discussion. Hertford appealed 'to the Queen’s princely liberality,’ whilst she apparently expressed her dislike of Wentworth, whom she said FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 187 abused her (Longleat House Seymour Papers Vol. V ff.4-4d, 69-70d, 73-78). Articles of agreement were drawn up whereby Wentworth would release certain manors in exchange for payments, but two years later the case was still unresolved. At that time the Queen remained sympathetic and said to Hertford that ‘ God knowethyou had wrong ’ (Longleat House Seymour Papers Vol. V ff.88-88d, 118-9). Hertford had also claimed to the Queen that it had been unjust to lose his father’s lands which his father had obtained during the reign of Henry VIII (Longleat House Seymour Papers Vol. V ff.73-78). His notes show he acknowledged that his father was ‘his own carver and might take what he liked when he was Protector, [but] why then should I not enjoy that of which he was no carver viz from 32-38 Henry VIII, when [Henry] allowed nothing except [to those giving] faithful service ’? (Longleat House Seymour Papers Vol. V ff.2-2d). Hertford’s son Beauchamp , his claim on the throne, and his betrothal Beauchamp was a universal disappointment to his father. If Hertford envisaged that his son would admire him, just as he esteemed his own father, his illusions were shattered, and the disrespect was mutual. Beauchamp was a prime candidate for the throne - the lawful heir according to Henry VIII’s will if his legitimacy could be proven (or the issue laid aside). Beauchamp’s refusal to adopt his father’s zeal for advancement resulted in failure to pursue the greatest prize all. Beauchamp made poor progress with learning, and at the age of 20 apologised for having wasted his education (Longleat House Seymour Papers Vol. V ff. 102-3, 110-113, 186-7). During Beauchamp’s early adulthood, Hertford reflected in his private notes that he was ‘ ashamed for my cost and his master’s travail in teaching 3 He commented ‘he pretends conscience, but in truth he does nothing but eat, drink, sleep and play’ (Longleat House Seymour Papers Vol. V ff. 126-7). Hertford’s fury exploded when Beauchamp sought to make his own choice of marriage partner. In common with Tudor conventions, Hertford held extreme views on the importance of furthering royal and noble blood-lines. His invective had even been directed at Sir John Thynne, his father’s most faithful follower, and his own close ally whilst in prison. When Thynne, who had achieved certain eminence (albeit without ennoblement) and sufficient wealth to build Longleat, casually mooted the prospect of a potential marriage between his daughter and a Seymour cousin, Hertford reacted with offensive contempt. Thynne, complaining that his kindnesses had been badly rewarded, warned ‘If you treat your friends as you have me, you will lack them when you need them most ’ (Longleat House Seymour Papers Vol. V ff.89-90d). Certainly, if Beauchamp had followed his father’s example, he would have matched himself to strengthen his (already substantial) claim on the throne. However, to Hertford’s disgust, late in 1581 he became emotionally involved with Honora, daughter of Richard Rogers of Bryanston, a man engaged with pirates and smugglers along the Dorset coast. Hertford already had family connections with the Rogers family through the marriage of his sister Mary, and had previously supported Rogers when he was in trouble for piracy, even helping to secure him a knighthood (Longleat House Seymour Papers Vol. V ff90-101). Rogers was a landowner able to establish discrete harbours, workshops and inns for refitting and victualling ships, in exchange for plundered goods and contraband. He controlled much of the Dorset trade, including around Lulworth, and was publicly shamed following investigations of his activities, although he escaped imprisonment, unlike his brother (Lloyd 1967). Hertford’s writings make it clear that he saw Honora not only as ill-bred, but as a lewd and contemptible slattern. In his notes he called her Onus Blouse - a play on words meaning a burdensome trull or fat, red-faced wench (Oxford English Dictionary). Descriptions of her by others, which Hertford recorded in his copious notes, included a baggage and a whore (Longleat House Seymour Papers Vol. V ff. 110-11 3). In a journal he kept, describing his attempts to thwart their liaison, he applied a similar epithet for his son, substituting ‘His Grace Lord Beauchamp ’ with the words ‘Graceless Bf writing both nicknames as the title on his book cover. Hertford’s anxiety for Beauchamp to attain advancement by conjoining with noble bloodlines, was superseded by the nightmarish notion of Honora becoming Queen. He forbade Beauchamp from seeing her, saying that he must not take such ‘an unwholesome woman ’ as a wife, and instead must marry as he [Hertford] chose (Longleat House Seymour Papers Vol. V ff.97-101). Hertford took strenuous steps to prevent Honora and Beauchamp communicating (Longleat House Seymour Papers Vol. V ff.90-101). Hertford’s personal confidante, Frances Howard, who later became his second wife, wrote to Beauchamp imploring him ‘do not stain your blood with base affection ’, or permit such a lowly mind possess such 188 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE an honourable person, instead making 'his choice like unto himself (Longleat House Seymour Papers Vol. V ff.118-9). Beauchamp claimed that those who talked of his marriage were ‘ greatly overshot ,’ and that he would be pleased to marry according to his father’s wishes. However, on the same day he wrote to Honora calling her mistress, and expressing his devotion. The couple connived to deceive Hertford, preparing a counterfeit ring which they could surrender as evidence that their relationship was over, whilst plotting to marry secretly (Longleat House Seymour Papers Vol. V ff.90-101; 118-9). Hertford sent his servants to Richard Rogers’s house at Bryanston to retrieve any tokens and love letters. Honora claimed that such things were burnt or lost (Longleat House Seymour Papers Vol. V ff. 110- 113). Meanwhile a draft letter was prepared for Beauchamp to sign, saying he had resolved ‘ never to think of so unfit a bargain ’ (Longleat House Seymour Papers Vol. V fl 15). Hertford however was convinced that his son was a liar, and might even place himself in danger. He had Beauchamp forcibly seized and confined (CSP Vol. 2 1581-1590). Beauchamp wrote to Secretary Walsingham apologizing if there had been any offence to the Queen. He initially begged to recover Hertford’s favour, praying to God to turn a ‘ a father’s anger into fatherly remorse,’ but later asked to be removed from his father’s custody, to whom he was 'an eyesore’ (Longleat House Seymour Papers Vol. V ff. 152- 154). In March 1584, Hertford prepared long and detailed interrogatories to investigate his own son’s relationship, mirroring those to which he had been subjected in the Tower, 23 years earlier (Longleat House Seymour Papers Vol. V f. 127). Beauchamp’s testimony also revealed similarities with his father’s history, insofar as he professed he had secretly wed, setting his hand and seal to (as Hertford called it) 'a pretended marriage ’ around Christmastime. Although Beauchamp’s first child, a son also baptized Edward, was born in June 1585, Hertford refused to accept Honora as his daughter-in-law. Beauchamp had been removed from his father’s house-arrest, and Hertford begged the Queen for his son to be returned and confined. However Beauchamp claimed that he would rather kill himself than be in his father’s presence, something which the Queen said she 'knew was true’ (Longleat House Seymour Papers Vol. V ff. 168-9). Hertford then beseeched the Archbishop of Canterbury to get the marriage nullified. Taking account of best lawyers and divines, the Archbishop concluded the marriage was good and could not be dissolved. Hertford refused to accept it, and remonstrated further, but was duly admonished by the Archbishop, who rebuked 'Do not be so carried away that there shall remain no hope of reconciliation. Christian charity requireth forgiveness, God’s law requireth it, and both the law of nature and the written law requireth it in you’ (BL Add. 32092 ff.25 lb-252). Hertford had sought to get his son’s marriage invalidated with the same gusto that he attempted to get his own authorized, taking opposite stances to achieve the same end - personal advancement by linking eminent bloodlines. However there are indications of greater forebodings. Hertford believed that Beauchamp’s relationship placed his son in mortal peril. He said that Beauchamp’s promise 'not to do anything without my leave was obtained [only] when he was in danger of his life’ (Longleat House Seymour Papers Vol. V ff. 1 26). Frances Howard wrote to console Hertford, then her fiance, saying 'You could not have a better occasion to loose yourself from danger’ (Longleat House Seymour Papers Vol. V ff. 168-9). It is unclear whether such danger might have come from those eager to pursue Beauchamp’s case for the crown, from the piratical activities of the Rogers family, or from a Papist cause in the prelude to the Spanish Armada. Hertford’s private journal reveals his antagonism towards his 'madd disobedient sonne’ and his repulsion for 'Onus Blouse ,’ (Longleat House, Seymour Papers Vol. 5, ff90-101, fl26). Following such extremes of language, it is doubtful whether full reconciliation could ever be achievable within the Seymour household. The only known image of Beauchamp, shown in effigy, is reproduced in Figure 3. Hertford's reintegration into society Upon his release from prison in 1571 Hertford had been quickly accepted back into the Court. The following year he claimed that through his attendance he was 'the worse off by £1600 or more. I wish for some occasion to try my dutiful service in some good adventure, which cannot be expressed in a progress’ (Longleat House Seymour Papers Vol. V ff. 1 l-12d). Hertford used the Court to ingratiate himself, advance the causes of seeking to recover his lands or reduce his fines, and seek royal favours. At New Year 1580 the Queen saluted him, and showed that she wore his gift. He promised to obtain for her a fine jewel he had seen (Longleat House Seymour Papers Vol. V ff.92-93d). It is unclear to what extent the Queen genuinely respected or trusted Hertford. She is reported to have used many persuasions initially against Frances Howard marrying him, and said FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 189 Fig. 3 The only known image of Edward Seymour, Lord Beauchamp, son of Hertford, from his father’s monument at Salisbury how little he would care for her (Longleat House Seymour Papers Vol. V fif. 164-5). She also did not doubt Hertford’s ill-treatment of his son Beauchamp. However, Hertford fully recovered his standing at Court, and in 1591, with his new Countess, presented to Elizabeth one of the most lavish entertainments of royalty ever bestowed, when she visited his house at Elvetham as part of the royal progress. Hertford met the Queen with a retinue of hundreds of followers in co-ordinated clothing, and with poets and musicians. Hertford had employed 300 men to prepare for the visit, enlarging his house, adding a great hall for celebrations, decorated with arras and the floor strewn with sweet herbs. A massive lake in the shape of a crescent moon was constructed to stage events, with islands bearing a fort and canon, a ship-island 100ft long bearing three trees set as masts (Anon 1591). In the water was a pinnace with masts, sails and anchors, twelve boats with flags and streamers, and various nymphs and sea-gods (Figures 4 and 5). The Queen’s entertainment at Elvetham reflects the grandiose ostentations of its host, and the lengths he could go to (as he did with the Seymour Roll and monumental construction) to reach the pinnacle of sumptuous magnificence. Fig. 4 The great crescentric pond at Elvetham, arranged for the Queen’s entertainment, 21 September 1591 190 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 5 The abandoned great pond at Elvetham, 2011, with one of the constructed islands Attempts to prove legitimacy It was Beauchamp and his brother Thomas who suffered most from the commission’s pronouncement that their parents’ marriage was unlawful. Whilst the sentence remained in force Beauchamp could not acquire the Earldom. Similarly he could neither inherit his mother Catherine’s estate (which had been seized by the crown, in the absence of a legitimate heir), or in due course his father’s extensive entailed lands. However, legitimacy was indissoluble from a claim on the throne. Any submission by Hertford, on whom the original sentence was conferred, ran the risk of incurring royal retribution, and it was generally his sons who pursued legal attempts to remove their bastardization, spanning 30 years. Beauchamp submitted appeals to the Queen annually for ten years beginning in 1577 when he was just 16 (NA C 270/34/6). Thomas also frequently commissioned his own notarial instruments affirming that he was the legitimate son of Hertford and Catherine, beginning from at least 1583, and annually again from 1589-1592 (CSP Vol. 2 1581- 1590). The brothers’ suits were not just prepared to advance their own interests, but to counter appeals being made by Lord Monteagle. He was the son of Mary Brandon, the half-sister of Catherine Grey, and in 1581 claimed livery of lands descending from Catherine because she died without lawful issue. Beauchamp countered that ‘ she was lawfully married to Edward, Earl of Hertford, by whom she had issue, the petitioner, :’ Descendents of Mary Brandon pursued legal attempts to inherit lands formerly belonging to the Grey sisters from thel560s to 1594 (Cecil MSS v 13 Addenda 1296-1597). From 1592 until 1595 Hertford stood alongside his sons in appealing against the illegality of his marriage (CSP Vol. 4 1595-1597). In 1595 this overstepped the bounds of royal tolerance. He, and possibly Beauchamp too, were arrested and held in the Tower, for seeking a new commission from the Archbishop of Canterbury and the Bishop of London to re-examine and determine the circumstances of his marriage (WSA 3790/1/3; BL Add. 63543 f.41). Frances, Countess of Hertford (his second wife) was overwrought with anxiety that he might be executed. However, Elizabeth wrote assuring her that Hertford’s actions demonstrated his ‘ lewd and proud contempt ,’ but were not capital offences (CSP Vol. 4 1595-1597). She ordered that the Archbishop’s documentation, and all papers relating to any appeals against the pretended marriage of Hertford and Grey should be seized and secured (Cecil MSS v5 1694- 5). Hertford was allowed to return to his own home under house arrest, and a fine of £5,000 imposed with demands for immediate payment. Hertford begged the Queen to be allowed time to sell lands to raise The sudden payment of so great a sum ’ (BL Add. 63543 f.40). The severity of the punishment brought to an end all attempts to pursue the claim of legitimacy during the Queen’s lifetime. Turmoil about the succession , and the death of Elizabeth Elizabeth’s refusal to name a successor generated massive speculation and factionalism, especially in her bouts of illness, and during the years of her final demise, when the issue remained critical and unresolved. As Thomas Wilson, a member of Cecil’s staff reflected c The crown is not like to fall to the ground for want of heads that claim to wear it, but upon whose head it will fall is by many doubted ,’ (Fisher 1936). Wilson prepared for Cecil a list evaluating 12 candidates, and placing James of Scotland, Arbella Stuart, Beauchamp and his brother Thomas as contenders one to four. Because intense factionalism pervaded the governance of England, Beauchamp’s cause could be promoted by those with much to lose if other candidates were successful, irrespective of his own action or proclivities. Elizabeth’s approach kept alive misplaced aspirations, whilst religious and foreign interests supported their own favourites. In 1594 a book entitled A Conference About the Next Succession to the Crown of England, which had been written by Robert Parsons under the pseudonym R Doleman, was published in Antwerp and began to percolate into Elizabethan society. It scathingly examined all candidates, and whilst claiming the Seymours were popular, it confirmed that Beauchamp’s chances FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 191 were diminished through his low-born wife. The book demonstrated that bloodline did not dictate succession, but rather determined a shortlist from which a suitable candidate could be chosen, based on honour, achievement and marriage. The following year, whilst Elizabeth was seriously ill, Sir Michael Blount, lieutenant of the Tower of London, who supported the Beauchamp cause, begun conspiring with others to provision the Tower so that it could be held in the event of the Queen’s death. Blount believed that there were several hundred gentlemen in England who shared his views on the succession and that he would be able ‘to arm four or five thousand with all manner of munitions and furniture ’ (Bruce 1984). Blount’s clumsy insurrection, at about the time Hertford was briefly incarcerated in the same Tower for seeking to get his marriage validated, may have further undermined Hertford’s standing with the Queen. However it appears that Elizabeth was scheduled to visit Hertford in 1601 and 1602 as part of her royal progress, until both were cancelled due to smallpox and foul weather (CSP Vol. 6 1601- 1603). Very little is heard of Beauchamp or his brother during the remainder of Elizabeth’s reign. However in 1596, it was reported that Sir John Smyth urged soldiers, weary from foreign wars, to mutiny and follow Beauchamp and Thomas, apparently without their complicity (CSP Vol. 4 1595-1597; Strype 1824). Then in 1598, in the wake of the minor Armadas of the previous two years, reports of piracy and Papist activity in Cornwall and Ireland, implied that Beauchamp might have been involved or sympathetic to the foreign cause (CSP Vol. 5 1598-1601). Just before Elizabeth’s death Arbella Stuart (the second candidate for the throne in Wilson’s list) sent a manservant to Hertford concerning a possible marriage between her and Beauchamp’s eldest son (Cecil MSS vl2 1602-3). As Gristwood (2003) observes, it seemed that a marriage between a Seymour ( any Seymour) and Arbella would combine two claims to the throne, neither of which, alone, was certain to topple that of James. Hertford was furious at the affrontery, and reported the incident to Robert Cecil, whereby Arbella’s plot was exposed. On her deathbed the Counsel came to the Queen to know her pleasure who should succeed, to which she replied ‘My seat hath been the seat of Kings. And I will have no rascal to succeed me. And who should succeed me but a King.' When the Secretary asked for clarification, she named her ‘ cousin of Scotland .’ An account by Elizabeth Southwell, a maid of honour present, states that the queen gave this proclamation when Beauchamp’s name was specifically raised as a possible successor (Loomis 1996). Elizabeth’s deathbed rejection of Beauchamp and her nomination of James had no effect in law, though it carried considerable influence. It was the Act of Succession in 1604 which removed all doubt - James was the King. According to reports of the diarist and barrister John Manningham, after Elizabeth’s death Beauchamp may have raised 10,000 men in support of his claim for the throne, and there were assertions that he held Portsmouth, had the support of the French, and that the Papists were likely to rise with him (Bruce 1868). The Lords warned Beauchamp he risked proclamation of treason. However the stories sounded erroneous or even risible. Manningham scoffed, ‘He rose in the morning, and means to go to bed at night ’ (Bruce, 1868). It was also reported to the Venetians that Beauchamp was in the west raising foot and horse with the intention of proclaiming himself king, acting at the instigation of France. However they said that the younger Seymour (Beauchamp) was yielding to the elder (Hertford) who, crippled though he was, swore he would have himself carried to London and there sign the proclamation of James’s accession and pledge his son’s hand to the same (Gristwood, 2003). Bruce (1868) claimed that other contenders for the throne were overawed and silenced when they found that the supporters of the James had stolen a march upon them, so that any agitation subsided almost instantly, and ‘the disturbed sea rocked itself to rest. ’ Relinquishment: Hertford ends 47 years of attempts to prove his marriage, and purchases those lands his son would have inherited, if legitimate With Elizabeth dead, and James formally declared as king, any attempts by Hertford to have his marriage to Catherine Grey validated, and his children legitimised, could no longer be construed as treasonous. Following the King’s coronation, Hertford wasted no time is seeking a new commission, to enable his surviving and elder son, Beauchamp, to inherit their mother’s and father’s lands, and his father’s titles. However, Beauchamp and Lord Monteagle contended for the same land and became locked into torturous legal arguments about precedence. An impediment was raised to pursuing an appeal through the courts, and there followed 192 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE fruitless debate about whether another commission could be granted to examine the impediment without intermeddling in such a cause (BL Add. 63543 ff.47- 80, BL Cotton Vitellius CXVI v2 f.5 16). Following legal stalemate, Hertford approached the King directly and in 1605 it was agreed that a commission should re-examine Beauchamp’s entitlement (WSA 3790/1/3; BL Add. 38170, ff.39, 52, 60, 68, BL78). It is sometimes claimed in secondary sources that Hertford’s marriage was recognised in common law at about this time and that the priest who had solemnised the ceremony came forward, some 45 years after the event e.g. Ellis (1827), St Maur (1902) and Locke (1911). Ellis cites Brydges’ edition of Collin’s Peerage for this, which has other proven errors, and this may be the elemental source for this romantic but unsubstantiated notion. Instead incessant legal arguments prevailed, and terms for a new commission could not be agreed. In 1607 Lord Monteagle, as descendent of Charles Brandon by his first marriage, petitioned for lands once belonging to Catherine Grey (descendent of Charles Brandon by his second wife, Mary Queen of France) claiming that Catherine died without issue. Beauchamp counter-petitioned against Monteagle, claiming his case should not be heard until the validity of his parents’ marriage has been re-examined, following the King’s agreement to such a commission (BL, Cecil Papers 1607-8). In 1608 the King consented to the issue being resolved by the Privy Council rather than through the law courts. The Council ignored issues relating to Hertford’s marriage, which left Beauchamp illegitimate, and according to Hertford’s notes ‘ wholly endeavoured to advance the King’s profits. ’ They recommended that Beauchamp should be able to purchase one third of the reversion of his mother’s lands, on payment of 60 years’ value, and purchase the reversion of his father’s lands on payment of 25 years’ value, for such land entailed on heirs male (BL Cecil Papers Petitions 2368). Hertford initially refused, saying the scheme ‘impeaches both ...my conscience and honour’ (presumably because it effectively acknowledged that Catherine gave birth out of wedlock). Hertford claimed that his son should have all of Catherine’s lands ‘were his right admitted’ and Hertford’s lands by inheritance, all without payment. However, later Hertford reconsidered, and said he would lay aside justice, his conscience and honour, and consent to a final conclusion of the case. However he entreated the King to provide more than one third of Catherine’s lands, and to accept just 15 years’ value for the reversion of his own lands (ibid.). After 47 years of legal wrangling a tripartite agreement was drawn up, involving the King (1), Hertford, Beauchamp and Lady Honora (2), and Lord Monteagle (3) whereby lands of Charles Brandon, late Duke of Suffolk, were to be divided, with the Earl and Beauchamp purchasing the reversion of manors and lands belonging to Catherine, and Beauchamp to be created Earl of Hertford (anew) on the decease of current Earl (CSP Vol. 8 1603-1610 & SP 38/9/28/2). j The acknowledgement of Beauchamp’s illegitimacy extinguished prospect that he could resurrect a claim on the throne, and impugned Catherine’s honour. However, it brought to a close nearly half a century of uncertainty. As it happened, Beauchamp predeceased his father in 1612, so was never able to assume the earldom. Beauchamp’s eldest son Edward also died before his grandfather, and it was his second son William who inherited the land and titles when Hertford died in 1621. j The Great Illuminated Roll Pedigree of the Seymours An introduction to the Roll The Seymour Roll is a massive manuscript bearing the date 1604, commissioned by the Earl of Hertford and depicting his ancestry (Figure 6). It includes the lineage of some 750 people, with paintings of 35 individuals, embellished with extracts from documentary sources, facsimile seals, classical quotations, floral decorations, the Seymour heraldic achievement and the first known depiction of the Savernake Horn (Figure 7). The roll pedigree makes only parenthetic reference to Catherine Grey, despite her centrality to Hertford’s ongoing lineage, whilst showing his two later wives. This is in contrast to the Hertford monument in Salisbury cathedral, which depicts Catherine Grey in effigy, but ignores Hertford’s later marriages. It is one of the three large and flamboyant roll pedigrees which Wagner (1952) describes as ‘the most imposing specimens of all.’ These three, comprising the Seymour, Shirley and Vere ; rolls, all date from the Elizabethan and early Stuart age, during a period in which nobility commissioned many genealogies and illuminated pedigrees. The Seymour Roll remained in private hands as property of the hereditary wardens of the Forest of Savernake from its completion in 1604 until it came FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 193 Fig. 6 The Seymour Roll Pedigree. About 40% of a parchment ‘so large that it cannot be fully unrolled in any ordinary room’ (WSA 1300/376) into the public domain in 2009 (WSA 1300/376). Upon the death of John, 4th Duke of Somerset, the last of the Seymour and Somerset wardens, in 1675, it presumably passed to his niece Elizabeth, who was heiress at law, together with the Forest itself. The roll was thus separated from those Seymour papers, including partial pedigrees, which now form part of the Longleat archives. The following year Elizabeth married Thomas Bruce, Earl of Ailesbury, who became the next hereditary warden. From this union the hereditary wardenship has continued through the Brudenell-Bruce family, who bear the titles of Earl of Cardigan, Marquess of Ailesbury, and Viscount Savernake. Perhaps as a consequence of it being privately held, coupled with the fact that it is, as Brudenell- Bruce (1949) observes, ‘ so large that it cannot be fully unrolled in any ordinary room ’ there is scant reference to the Seymour roll in literature. His own Wardens of Savernake Forest includes many ancestral charts, and he was clearly influenced by the content, although he makes few direct references, and introduces earlier generations not shown. Several works on the history of the family do not cite the roll at all. Reed (2008, 2010) in his two meticulously researched papers on the early origins of the Seymour family, appears not to draw from the roll’s contents. Similarly Patterson (1999) in his massive tome Seymour Dawn makes no reference to it. Two early 20th century Seymour family histories are also silent on the roll, yet these may have been indirectly influenced by it through the works of Dugdale and later heralds who drew on the College of Arm’s records in their publications on the peerage. Other than passing mention in Jackson (1875) and Ward (1860) no further references to the roll have been found. Since coming into the public domain, the roll has been described in detail by Douglas (2010, 2011). The parchment The roll measures 6.83m x 1.88m (22’5” x 6’2”), and is attached to wooden poles at each end, the upper of which bears a wire for hanging. It is constructed from 26 parchment membranes, glued together without stitching. These are arranged in nine bands, the top one comprising just two membranes, 194 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE ■- * •* ' ^ StfmmaGknfauk;icvmNohii issim y Honoratissim^lei IMPRIMIS h ^*1 V& U S Ifal i ft ' "i r'~t ■n 1 1 Ij * * 9 * 9 9 f f f f ? ? o -* V* 4 V * ? * •Ghw- • - 5 - * » A fe ' -Y'l f - i l i ® r- , 1 in ; _• - SP, * V * V ? ¥#* f w ; , J J:f ». rr> »'r * * fj_ »■ » y *.* » * *.» ,eW ; : % ffl® * j* * — • - “ " *§ . ' •' J: © 5- H $ © © » 9 © . : » w tops • ■ ■ „ - • ■ • c'g& 9 9 5 *■ r t r. r g 99 y 9 _? • * » * /fr\ ' H3i»^ ' ' # ® ■ ‘ .: - . -. m f ~ % m i w 9 > ^ * * 9 9Wt » * •>•1 ■ '? • If * ■WJ/S"- ■ ■ ... 9 9; P V 99 9 9 V ($>$ b0= v 9 99 9 9 > 9 . Cr O' ?»* % v->S • ••• * t*" _ f f f . J\\ \ m ■■ * y *9 • . . _ " ■ ' .... ; • -A » *»» v ' - •?*»■ ', » « %’ f « f » ,»■ 5 # >! A- *5> 28 9 * -, v~r 9 9 9 9 9V * « ■ © «■ f . “ « ^ ^ 'p ^ ,< , , , ?'•- m * ip f -.'! ■*

\ , Com.’v ( '0 Y • .vtttfof . 6 1 'i.'rt «■ bi'.v ijitt'C :» | fftftolr’ iMcftfi,1 II VVfWf J*VIK«n . t,;/ SfAnttA • ,y (raw Vv«- ^fmoafcrt 7^ ■j^w£»wfc A*’" 5«a# £>wrfi$ ! ^fn'sf- ta«£ fffAm X^W’* fy$s ' -atv^ ’i>ia~£>»vfiY . Us&oJWn, « ' 'xwrt,' \ fU'-vr^K Fig. 15 Extract depicting 3 generations, and 11 of the 750 individuals on the roll. Hertford is central. His roundel reads Edward Seymour, Earl of Hertford and Baron Beauchamp, first married Catherine, daughter of Henry Grey Duke of Suffolk. His later wives have their own roundels. His parents, the Duke and Duchess of Somerset (and his father’s former wife), are on the top row, and his sons Beauchamp and Thomas and their wives on the bottom row. ( Vertical lines foreshortened from original) (WSA 1300/376) 202 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Sources of Evidence No. Deeds and charters 21 Inquisitions Post Mortem 16 State Records, rolls, fines etc 9 Wills 2 Other 1 Fig. 16 Sources of evidence cited on the roll mortem and other documents relating to members of the Sturmy/Esturmy family, and eleven provide similar records relating to Seymours. One piece of evidence cited is in French, and comprises a receipt from Peter Delamere confirming he received 100 sous from Sir William Seymour in 1370/71. Authentification of evidence Of the documents cited in the 49 evidence boxes, 20 were identified as surviving in modern archives. Others may be extant, but have not yet been located. Of these, 15 (75%) were examined and found to have been faithfully, albeit selectively, represented on the roll. There were no examples of misrepresentation. Examples of Text Boxes The centrepiece of the genealogy is a script describing the origins of the Seymours. It states Amongst the nobles and men of war flourished with glory who served under William Duke of Normandy, sumamed the Conqueror, in the celebrated year of 1066 when having invaded England with his conquering standard he routed King Harold in close battle, [are] reckoned those who were sumamed St. Maur .’ It claims the family name derived from a feudal possession in Normandy, bearing a church dedicated to that saint. It mentions family members ‘of noted fame ’ who settled in England. It also describes how, when Henry I seized land from the Welsh, the family settled at Penhow in Wales, a church was dedicated to Saint Maur there, and a place called Seymour Park survives. Despite the inference, any link with the Normandy family has yet to be proven. One box describes a charter of Henry Duke of Normandy and Acquitaine, involving a transfer of his manors of Stapleford, Bossingham and Figheldean, previously held by his mother the Empress Maud, to Henry Hussey. A note held at Wiltshire and Swindon History Centre claims that the original charter, from which this is copied, has been lost, and this provides the only surviving evidence. One evidence box transcribes a charter of 1199/1200 through which King John confirms his grant to Henry Sturmy of the land and bailiwick of the forest of Savernake such as Geoffrey Sturmy his father had at his death, for the accustomed services. Fig. 1 7 Original charter of King John, granting the hereditary wardenship of Savernake to Geoffrey Sturmy, as accurately reproduced on the roll (WSA 1300/12) This transcription is a faithful rendering of the charter which still survives (Figure 17). One box describes an undated deed through which Henry Sturmy, son of Geoffrey Sturmy, granted to his son Henry, at his marriage with Margaret daughter of Lord Hubert Hussey, the Manor of Wolfhall. A further evidence box relates to an inquisition of 1381/2 on the death of Henry Sturmy who held from the King in chief the manor of Burbage for the service of providing one Knight in the King’s army, armed with hauberk, iron helmet and a covered horse, and also held two messuages and virgates in Wyke, one in association with the West Bailey, bailiwick of the forest of Savernake for the service of 12 shillings [at] Marlborough Castle, together with the manor of Stapleford and advowson of the church, and manor of Figheldean. It was declared that his brother William Sturmy, 25, was Henry’s heir. The original from which this is derived appears to be C 136/19/1 in the National Archive. The Duke and Duchess of Somerset These play a pivotal position on the roll, as parents of Hertford. FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 203 ^twalDunfja Fig. 18 Portraits of Duke and Duchess of Somerset, centrally placed on the Seymour roll (WSA 1300/376) Portraits The Duke and Duchess have the only portraits not positioned at the head of their family, and instead inserted at their chronological position (Figure 18). They may also be attempts at true likenesses. The portrait of the Duke closely resembles a painting by Hans Holbein, and engravings derived from this. The Duchess appears more stylised although has some (less conclusive) resemblance to a portrait by Antonio Moro. Descriptions The text for the Duke focuses on his illustrious achievements. It does not mention his execution, whilst suggesting he lost honours because of the malice of his enemies. The full text reads: ‘ The most illustrious Prince Edward Seymour ; Duke of Somerset, because of his virtues, and relationship with the king ( namely the most powerful King Henry had married Lady Jane his sister) his position was raised to the highest public office, and total authority for the army. As a young man, together with Charles Brandon Duke of Suffolk he captured Mount Desiderius or Montdidier in Picardy, honoured with a knighthood, then Viscount Beauchamp, and afterwards was augmented with the honour of Earl of Hertford, and admitted to the most noble order of the garter; when he was sent to Scotland with an army, where he routed the enemy at Leith and captured the fortified city of Edinburgh, and destroyed others. When the most serene Edward, his nephew by his sister took the sceptre, he was created by him Duke of Somerset and Baron Seymour, his governor and protector of the kingdom, commander of the army and held the rank of general, treasurer and Earl Marshal of England, and appointed Governor and Captain of the Isles of Guemesy and Jersey. First with pious endeavours he restored the true religion of god to England then he defeated the Scots most decisively in fair battle near Musselburgh, slaughtering fifteen hundred from the captured nobles, barely a hundred English were lost. Truely here a Prince so great and excellent, whose virtue known at home and at war, by the secret guile of his rivals by whose ambition he seemed to lose the way, and with Fig. 19 Full Hertford Achievement (WSA 1300/376) the greatest lamentation all his honour was taken away for the most trifling of crimes. ’ Large illustrations Beneath the genealogy, sit two major visual elements: Hertford's Full Heraldic Achievement This is presented centrally at the end of the roll, filling the whole of the vertical dimension of the ninth parchment band (Figure 19). The arms are shown with fourteen quarterings, as given in the Table of Figure 20. The Savernake Horn This is the largest representational drawing on the roll, measuring 36 x 35cm. It lies immediately above the heraldic achievement, but slightly right of centre. Its presence disrupts the pictorial balance of the roll. Unlike other sources of evidence presented FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 205 Quarterly of 14: 1 & 14. Royal augmentation granted by Henry VIII in honour of Jane Seymour. 2. Seymour from Somerset, his father. 3. Beauchamp de Hache from Cecilia, great (x6) grandmother. 4. Vivonne, from Cecilia, great (x9) grandmother via Beauchamp. 5. Marshal from Sybil, great (xl 1) grandmother via Vivonne. 6. Clare from Isabella, great (xl2) grandmother via Marshal. 7. Leinster from Eva MacMorrough, great (xl3) grandmother via Clare. 8. Gifford from Rosia, great (xl6) grandmother via Clare. 9. Delamare from Joan Delamare, great (x6) grandmother via Margery de Brockbury, great (x5) grandmother whose father was not an armiger. 10. Sturmy from Maud, great (x4) grandmother. 11. Hussey from Margaret, great (x8) grandmother via Sturmy. 12. William from Isabella, great (x3) grandmother. 13. Coker from Elizabeth, great (x2) grandmother. Fig. 20 Quarterings of the Hertford Achievement in the roll, there is no text or explanation alongside describing any significance attached to the horn. Its depiction may be an afterthought, not incorporated into the original plan. This drawing is the first ever representation of the horn, often claimed as the ancient symbol of tenure of the hereditary wardens of Savernake, but never mentioned hitherto in the extensive family archives. The horn is described more fully, with an image from the roll, in Bathe (2012, this issue). Individuals included on the roll Of the 750 individuals included in the pedigree, 282 are direct line ancestors of Hertford, extending over 19 generations. It contains some 39 monarchs, and at least 206 individuals who have entries in the Oxford Dictionary of National Biography. Multiple collateral lines are extended to the date of the roll. For example, REFERENCE SOURCES Individuals Stirnet.com 660 thePeerage.com 537 Dugdale’s Baronage 365 Burke’s Peerage 337 Burke’s Extinct Peerage 189 The Complete Peerage 176 Cardigan’s Wardens of Savernake 55 Fig. 21 Correlation of individuals recorded in the roll with genealogical reference sources Henry Hastings is the product of eight generations of father, mother and male heir without any record of siblings. Through the use of collateral lines the Seymour roll shows kinship between Hertford and other nobles, and indeed with the King (his 6th cousin, twice removed). The roll portrays at least 57 individuals alive at its completion date, including seven knights, seven barons, nine earls, two princes and the King. Authentification of individuals There is little certainty in any genealogy of the period covered in the Seymour Roll. The royal and noble lines are better documented and researched than other less noble lines, but none of the standard genealogical reference works provides a single, reliable and complete picture of Hertford’s ancestors from which the Seymour Roll might be validated. Comparison of the number of individuals recorded on the roll with genealogical reference works shows a varying degree of corroboration, as shown in the table, Figure 21. Analysis of the direct line ancestors by looking at the degree of corroboration by individual provides a better, yet still flawed insight into the accuracy of the roll, highlighting that some of the claimed links may be weaker than others. Only exhaustive research using original documentation will reveal the true accuracy of the roll. Some relevant and detailed research has been already conducted independently by Reed (2008) who investigated the evidence of the earliest origins of this Seymour line. He concludes that William St. Maur is the earliest provable ancestor. William is the first of the Seymour line included on the roll. The subsequent generations researched by Reed bear some semblance to the roll, although Reed argues for the existence of a generation that is not shown in other Seymour pedigrees and is not on the roll. However, he highlights a weakness in the Seymour lineage that is accepted in other works. He notes that 206 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE f Margaret ? daughter of Sr Henry i Wentworth knight and of i Anne daughter j "\,of Sir John Save^F Sir John Seymor Jane daughter of Sir Nicholas Poyngs knight Henry the eyght King of England France and Ireland Dorathe Seymor marie d to St Clement Smyth f Sir 1 Henry Seymor] | knight j John Seymor one of the genrilbnen i Pensioners j William Seymor Robert Seymor Jane Queue of England Thomas Seymor Lord of Sudky Knight of the Most k Noble Order of j the < iarter Elizabeth daughter of Sir John Seymor knight KATHERIN Queue of England the lass wiffe to fvnige Henry the eighi Gregory >rd Cromwell Edward the sixt King of England France and Ireland r Edoard Duk of Somerset! Lord Protector of England and Knigh' Wof the Garter jf Janne daughter of John Seymor esquire Elizabeth daughter of John Seymor esquire / Anne \ \ daughter of Sir Mtchell I Stanhope , Thomas Seymor sonne and r Lady N Katherin Seymor youngest daughter Lord Edoard Seymor Lady Margret Seymor Mary Seymor Andrew sonne and k heire to Sir Richard J Rogers knight J? ' Lady Anne . ? Seymor ■ Countesse of Warwick { and after maried to < l^nr Edward UmptW f Edward \ [Seymor Erie of] I Hartford J Lam Lord This Pedigre to the right Honorable Edward Erie of Hartford with the collateral descent unto John Seymor esquire one of the gentillmen Pensioners to the Queenes most excellent Majestie. Collected and exemplified according to the records in the office of Arms, the 20 July An Dom 1576 William Dethick York Herauld of Arms Fig. 22 Lower pan of a pedigree prepared by Dethick, York Herald, in 1576. Redrawn from the original at Longleat House in agreement with the Marquess of Bath ( Longleat Seymour Papers vol. 6) Dugdale, a highly respected herald, in his Baronage of England of 1675/6 commences the origins of the Seymour line with Roger Seymour, who married a wealthy widow and heiress Cecily Turberville (nee de Beauchamp) cl 347-8. However, Dugdale would have been aware of earlier ancestors included by Vincent, FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 207 Hancasfc ©eurj> tljc VII Ijab tiuo sons aitii tiuo daughters: bn €li?a - betlj lunge Cblnarb tljp fourth’s baugljter 3 Slrtljur prince of Scales to bo bleb tultljout issue O.c- Jlenrie the VIII of luljome bestcnbeb < Jfflargaret tlje elbest baugljter UjIjo marpeb ttuice itlarie Seconb baugljter marrteb tbitce Cbtuarb tlje H>ixt Jtlaric luljo inarpeb •pijtllip $Unge of £s>payite Cloabetlj «. irst to James IV lung of ss>cotlanb of tofjome slje fjab issue ^erotib Urtljtbalbc ©uglas Carl iUnguis] -of loljome bestenbeb JftrSt tlje iLtng of Jfrante bp Inborn She Ijab issue H>econblp Charles Pranbon Bukc of Suffolk bp iuljom She Ijab issue j" u James tlje V father to fflargaret luljo marrteb the Carl of Jpmox bp taihome She Ijab issue ■» JftrSt jfraunces luljo marrpeb Jjenrp <£rape f Jjatljerin <®rap toljo inarrpeb Buke of Suffolk of s tlje Carl of partforb anb , tuljom issueb •-» fjab issue Seconb Cllenor tuljo ffiiargaret tuljo marrpeb tlje marrteb Carl of Cumb|earle of BerbP anb Ijab issue -erlanb of tnhom tssticlr-' ftlarp marrpeb 1st the Bolphin of Jfrante anb after penrp Porb Barlep anb Ijab issue Che Jorb Barlp before ^ meutioueb anb CljarleS tuljo marrpeb Clijabetlj Cabenbislj anb Ijab issue o fames tlje ^>txt Ha grbella ^Cbtoarb Mentor Porb Peaucham anb Cbotnas ^>epmor Jparbenanbo Porb Strange anb pretljren Fig. 23 Contenders for the throne in late Elizabethan times. Redrawn from the original at Longleat House in agreement with the Marquess of Bath ( Longleat Seymour Papers vol. 6) a herald and protege of Camden from the early part of the seventeenth century, in his Baronage. Vincent is reported to have relied upon and partially quoted a deed which provides evidence of a transfer of land in Woundy to Roger from his brother John. Reed argues that this document, which has not been found in modern archives, may have been falsified. The Seymour Roll shows both brothers and their descent from William St. Maur. Furthermore, it provides a facsimile of the deed in question, in greater detail than reported by Vincent. Vincent also includes four sons of Roger and Cecily, whom Reed also disputes. They too appear on the roll. Whether Vincent was influenced by the roll (or both relied upon common evidence), and whether this is based on a falsified record is unknown. Despite this potential falsification the roll does not make outrageous claims. Some lines are more readily proven than others. However, the central message is that Hertford had a noble ancestry over many generations, and was well connected to contemporaneous nobility. Related and possible precursor genealogies of the Seymours A series of eleven worked pedigrees survive amongst the Seymour papers (Longleat House Seymour Papers Volume VI, ff 9-19). Many of these are prepared in rough. All except one are undated, although the inclusion of certain individuals suggests that most were prepared in the latter years of the 16th century. It is likely that those preparing the illuminated roll pedigree would have been aware of these drafts and the source material from which they were drawn. They include parchments showing pre-conquest genealogy of the Saxon and Danish Kings, Scottish royal ancestry, and the descendents of William I and Edward III. Many genealogies show Hertford’s son, Edward Lord Beauchamp (b.1561), and Arbella Stuart (b.1575) but none show their marriages of 1582 and 1610 respectively. Hertford long-denied the validity of his son’s marriage, so its omission is not conclusive evidence that the pedigrees pre- date 1582. Whilst many of the genealogies comprise rough workings, one professionally executed roll shows the pedigree of the Seymours for eight generations down to the Earl of Hertford, depicting arms in colour (Longleat House Seymour Papers, Box XVIII) (Figure 22). Both of the Duke of Somerset’s marriages are shown. This pedigree also shows one of Hertford’s brothers, a younger Edward (Somerset christened three of his children Edward), with the phrase childe obit 1574.’ Hertford complained of a ‘ villain ’ priest called Braddock who had refused to bury his brother Edward (Longleat House Seymour Papers, Vol. XI, ff.97-101), and this statement and the phrase in the pedigree raise questions about the manner of the two deaths. This pedigree appears to have been drawn up specifically to show the collateral descent of John, 208 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE a gentleman-pensioner to Queen Elizabeth, and an illegitimate brother of the Protector and Queen Jane. It also shows his marriage to Jane, daughter of Sir Nicholas Poyntz, and issue. John’s arms are crossed with a bar sinister, and his crest of the Seymour phoenix carries a similar bar sinister at the foot of the roll. The finished pedigree is signed by William Dethick, York Herald, and dated 20 July 1576. This pedigree shows the involvement of the College of Arms in researching pedigrees for the family at that period. Hertford’s own marriage and children are not shown. Of particular interest is a genealogy showing the four key contenders for the throne, as the descendents of Henry VII, in the latter 16th century Longleat House Seymour Papers Volume VI, ff 9-19). They include Lord Beauchamp and his brother Thomas (d. 1619), Fardinando Lord Strange (d. 1594), Lady Arbella Stuart, and James (described as VI, not I) (Figure 23). This document provides unequivocal evidence that Beauchamp’s legal entitlement to the throne was under active consideration by the family or those close to them, and must have persisted as background awareness during the lengthy formulation of the Seymour roll. Monuments erected during the lifetime of Edward Seymour, Earl of Hertford From 1588, and extending through the period when the Seymour roll was in preparation, Hertford set about constructing a series of opulent tombs eulogising his family. The Duchess of Somerset’s monument in Westminster Abbey On the site of the altar in St Nicholas’s chapel Hertford erected a sumptuous monument, 7.3m (24 ft) high, to his mother (Figure 24). Her effigy dressed in the red robes of a Duchess lies on a sarcophagus decorated with lion’s masks and clawed feet, under an arched canopy, with ornamented obelisks and columns. A Seymour phoenix sits each side of her heraldic arms and the Stanhope crest. A black panel with gold lettering records her attainments in Latin with English translations. She is described almost solely by her relationship to other nobles, and Hertford seized the opportunity to promote the achievements of his father, whose headless corpse Fig. 24 The monument erected by Hertford to his mother Anne, Duchess of Somerset, in Westminster Abbey had been unceremonially interred in the chapel of St Peter ad Vincula , Tower of London. It states ‘Here lyeth intombed the noble Dutchess of Somerset, Anne, Spouse unto that most illustrious Prince Edward, Duke of Somerset, Earl of Hertford, Viscount Beauchamp, and Baron Seymour, Companion of the most famous Knightly Order of the Garter, Uncle to King Edward the Sixth, Governor of his Royal Person and most worthy Protector of all his Realms, Dominions, and Subjects ; Lieutenant General of all his Armies, Treasurer and Earl-Marshal of England, Governour and Captain of the Isles of Guernsey and Jersey. Under whose prosperous Conduct, glorious Victory hath been so often and so fortunately obtained at Edinborough, Leith, and Muscleborough Field .’ She is described as ‘A Princesse descended of noble lignage’ and daughter of ‘the worthy knight Sir Edward Stanhope, by Elizabeth, daughter of Sir Fulk Bourchier, Lord Fitzwaren,from whom our modem Earls of Bath are descended ’ . . . ‘grandson of King Edward the Third, and Queen Eleanor, :’ The monument lists her nine children comprising her sons Edward Earl of Hertford, Henry, and a younger Edward, and daughters, Anne, Countess of Warwick, Margaret, Jane, Mary, Katharine, and Elizabeth. It fails to mention her husband’s execution, or her own second marriage to a commoner - her husband’s steward, Francis Newdigate. Publicly advertising his role, FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 209 Fig. 25 Monument to Hertford’s grandfather, Sir John Seymour, in St Mary’s church, Great Bedwyn the monument states ‘ The Earl of Hertford, Edward, her eldest Son, in this doleful Duty careful and diligent, doth consecrate this Monument to his dear Parent: Not for her Honour, wherewith (living) she did abound, and now departed, flourisheth; but for the dutiful Love he beareth her, and for his last Testification thereof ’ (A smaller monument to the Duchess in the church of Chelsea, where she held the manor, describes her simply as ‘ some time Wife to Edward Duke of Somerset, Protector,’ and is unlikely to have been commissioned by Hertford). Sir John Seymour’s monument in Bedwyn In 1590 Hertford built a grand chest- tomb bearing coats of arms at Great Bedwyn church, for his grandfather, Sir John Seymour, ofWolfhall (Figure 25). He had died in 1536, and was first buried in the church of the Trinitarian Friars of Easton. This had fallen into ruin following the reformation (in which, ironically, Hertford’s father had been such a champion), and Hertford had the coffin and effigy removed to Bedwyn. A lengthy inscription fails to mention any of Sir John’s ancestors and focuses on descendants. Sir John had six sons and four daughters. Prominence is given to ‘Edward, Duke of Somerset, Earl of Hertford, Viscount Beauchamp and Baron Seymour, Uncle to King Edward the Sixth, Governor of his Royal Person, Protector of all his Dominions and Subjects, Lord Treasurer and Earl Marshall of England [who married] Anne Daughter of Sir Edward Stanhope, Knight, by Elizabeth his Wife’ [thereafter described in identical words to his mother’s monument:] Daughter of Sir Foulke Burgchier, Lord Fitzwaryn,from whom the modem Earls of Bath are descended.’ It continues with: ‘ Thomas , Lord Seymour of Sudeley, High Admiral of England, who married Katherine, Queen of England, and Widow to King Henry the Eighth .’ The monument fails to mention the execution of the brothers Somerset and Sudeley, or Somerset’s first marriage. The inscription also records their sister Jane, Queen of England, Wife to King Henry the Eighth, and Mother to King Edward the Sixth.’ Again Hertford’s role in constructing the tomb is proclaimed, and his personal expenses in doing this pointed out: ‘This Knight departed this Life at LX years of age, the XXI day of December, Anno 1536, and was first buried at Eston Priory Church, amongst divers of his Ancestors, both Seymours and Sturmys; Howbeit that Church being ruined, and thereby all their Monuments either wholly spoiled, or very much defaced during the Minority of Edward, Earl of Hertford, Son to the said Duke, the said Earl after ; as well for the dutiful Love he beareth to his said Grandfather, as for the better Continuance of his Memory, did cause his Body to be removed, and here to be intombed at his own Cost and Charge, the last day of September, Anno 1590, in the XXXII year of the most happy Reign of our gracious Sovereign Lady, Queen Elizabeth .’ The Plaque to Hertford’s sister Lady Jane Seymour at Westminster At about the time that Hertford commissioned these two tombs, he had a plaque erected to Jane, his sister, who had been buried in the floor of St Edmund’s chapel, Westminster Abbey. Despite being a smaller, wall-mounted structure, it is extra- ordinarily ornate, in black and gold, with floral patterns and Tudor roses, and with phoenixes each side of the crest, bearing the Seymour arms (Figure 26). From stylistic grounds it can be dated to about 1590, some 40 years after Jane’s death (Peter Sherlock, pers com). In gold lettering on black, she is described as the ‘Noble Jane Seymour, daughter to the renowned Prince, Edward Duke of Somerset, Earl of Hertford, Viscount Beauchamp and Baron Seymour, and to the right Noble Lady Anne, Duchess of Somerset his wife, who departed this life in her virginity at the age of 19 years, the 19th March 1560’ [new style 1561]. Jane was similarly described as dying ‘in her virginity ’ on 210 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 26 Monument erected by Hertford to his sister Jane, in Westminster Abbey the Seymour roll pedigree. Again Hertford draws attention to his own character, stating that ‘ Edward Earl of Hertford and Baron Beauchamp, her dear bother, hath caused this monument to be made .’ Jane had been Hertford’s confidante, and according to his and Lady Catherine Grey’s testimony, had been sole witness to their wedding. Her untimely death three months later deprived them of being able to prove the lawfulness of their marriage, and the legitimacy of their sons. The Monument to Hertford’s second wife at Westminster Abbey The last opulent monument built by Hertford during his lifetime was to his second wife, Frances, Countess of Hertford, who died in 1598, with her ambitions to bear him ‘another pretty boy ’ unfulfilled (Longleat House Seymour Papers V, ff. 168-9). It was constructed at the former altar site of St Benedict, in his chapel in Westminster Abbey (Figure 27). It emulates the monument to his mother, with an effigy in the red robes of a peeress, this time below a double arched canopy, framed with ornate columns, obelisks, figurines and coats of arms. She is described as the ‘Dear Spouse to the noble Edward Earl of Hertford, and Baron Beauchamp, Son of the renowned Prince ■ 'rasriFfCATri US C&KBfVU. CiCS^EGRAT Fig. 27 Monument to Frances, Countess of Hertford, Westminster Abbey FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 211 Fig. 28 Small brass memorial to Beauchamp on the wall of St Mary’s, Great Bedwyn Edward Duke of Somerset, Earl of Hertford, Viscount Beauchamp, and Baron Seymour. A Lady descended of right noble Lineage, being Daughter of the noble Lord William Baron Howard of Effingham, Companion of the most famous Order of the Garter, High Admiral to Queen Mary, and Lord Chamberlain of the Household, and Lord Privy Seal to Queen Elizabeth. Son was he to the right noble Prince Thomas Duke of Norfolk, and Earl of Surrey, Earl Marshal of England. ’ She died 'at 44 Years of Age, the 24th Day of May, Anno 1598, in the 40th Year of the most happy Reign of our most gracious Sovereign Lady Queen Elizabeth .’ Hertford emphasises his role, stating that he, as 'her loving Lord and Husband, much lamenting her Death, in Testification of his great Love towards her, and of his careful Diligence in this doleful Duty, doth consecrate this Monument.'' Beauchamp’s brass plaque at Bedwyn One further monument created during Hertford’s lifetime, but not necessarily commissioned by him, is out of all keeping with the others (Figure 28). It is a tiny and pathetic brass plaque, now mounted on the north wall of Great Bedwyn church, but according to Aubrey formerly on 'a pitiful grafted freestone grave.’ It reads ' BELLOC AMPUS ERAM GRATA GENETRICESEMERUS, TRES HABUI NATOS, EST QUIBUS UNA SOROR’ (7 was Beauchamp; my father a Seymour, my mother a Grey, I had three sons, who have one sister ). This is the sole memorial of Hertford’s son and heir, who, according to the will of Henry VIII, should have been King on the death of Elizabeth I. The diminutive and humble nature of this monument may reflect the fact that Beauchamp, who died in 1612, had been a relentless failure in his father’s eyes. It is also possible that Beauchamp’s plaque was erected by one of his own children. However this is unlikely to have been William, later second Duke of Somerset, because he was in exile, following his unauthorized marriage to Arbella Stuart, cousin to King James. The Hertford monument in Salisbury Cathedral Despite the flamboyance of its progenitors, the family tomb which eclipses all others in ostentation is the memorial to Hertford himself and to Lady Catherine Grey in Salisbury Cathedral (Figure 29). This architectural colossus represents the pinnacle of ‘skyscraper’ monument construction during Elizabethan and Stuart times, and is one of the largest tombs ever seen in England. It is located on the former site of St Stephen’s altar, where rising an estimated 11m (36ft), it wholly occludes the central light of a triple stained glass window. It faces the southern aisle of the nave and choir, and is visible along the whole length of the cathedral (some 130m), with its largest inscriptions, as Peter Sherlock (2008) observes, shouting of Hertford’s and Grey’s certain salvation from the rooftops to all who enter. The monument is a dazzling mix of naturally varied marble, ornamented and gilded stonework, finely honed effigies, figurines of angels and scribes, heraldic devices and lengthy inscriptions. The recumbent life-sized effigies of Hertford (Figure 1) and Grey, lie in prayer on a sarcophagus decorated with lions faces and claws, under a coffered arch. Catherine’s effigy is elevated above Hertford’s, reputedly representing her higher status. Effigies of their two sons - Beauchamp (Figure 3) and Thomas - also praying, kneel dutifully either side of their parents (each facing north), although in reality both had pre-deceased Hertford. There are ornate columns and obelisks depicting weaponry (including firearms), presumably a reference to Hertford’s 212 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Fig. 29 The Hertford Monument is Salisbury Cathedral, with the recumbent effigies of Hertford and Grey flanked by their two sons military role as Lord Lieutenant of Somerset and Wiltshire in 1601. Above the recumbent figures, is a heraldic family tree (suitably portrayed with branches and leaves), depicting seven coats of arms on each side, conjoining as Seymour-Grey at the top. On the Seymour side these are shown as (from the base) Beauchamp of Hache, Esturmy, MacWilliams, Coker, Darrell, Wentworth, and Stanhope. On Catherine’s side they are de Strange, Astley, Ferrers, Woodville, Harrington/Bournville, Wootton and Brandon. Between these is the full coat of Seymour comprising 14 quarterings impaling Grey who has 16 quarterings. Above the arch to the left is the Seymour Augmentation granted to the family by King Henry VIII 'of his mere grace ’ in commemoration of his marriage with Jane Seymour. This also features on a shield worn by the kneeling figure of Beauchamp. At the top is the Seymour crest of a phoenix, rising in flames from a crown, with the Seymour shield (again with the Seymour Augmentation, the Seymour wings, Esturmy demi-lions, with the arms of Beauchamp, MacWilliams and Coker). A unicorn and bull serve as supporters. Two great black panels at the top, either side of a further phoenix, bear quotations in massive lettering, and resemble the appearance of the Roll Pedigree in layout. These include an excerpt from The Phoenix by the Roman poet Claudian, a poem which describes the pretence of death masking the renewal of existence. The quotation ‘ mutata melior procede figura,’ may be translated as ‘ transformed to a more beautiful state,’ foretells Hertford’s and Catherine’s resurrection whilst making further analogy with the rebirth of the phoenix. Below these are two quotations from the first book of Hebrews: ‘ Who made His angels spirits, and his ministers aflame of fire’ relates again to the phoenix, and 'He hath sent them to minister to those who shall receive the inheritance of salvation’ which promises deliverance. An effusive and bombastic inscription in Latin above the recumbent figures, concentrates on Hertford’s eminence as son of the Protector, whilst mentioning his mother’s line more tangentially. Initially the script follows that on Hertford’s mother’s tomb almost verbatim. 'Sacred to the memory of Edward, Earl of Hertford, Baron Beauchamp, son and heir of the most Illustrious Prince Edward, Duke of Somerset, Earl of Hertford, Viscount Beauchamp and Baron Seymour, Knight Companion of the Most Honourable Order of the Garter, Uncle of King Edward VI, and his Governor, Protector of his Kingdoms, Dominions and subjects, Commander-in-Chief of his Armies, Treasurer in general on his behalf, Earl Marshal of England, Governor and Captain of the Islands of Guernsey and Jersey, by Ann his wife, who was descended from an ancient and noble family. ’ In the case of Catherine (who had died over 50 years previously with no memorial hitherto), the monument proclaims her royal lineage for the first time: 'Also of his most dearly beloved wife, Catherine, daughter of Henry and Frances Gray, the Duke and Duchess of Suffolk [which Duchess was] the daughter and heiress of Charles Brandon, Duke of Suffolk, by Mary, sister of Henry VIII, and Queen of France, endowed as her granddaughter and the great-grand- daughter of Henry VII. Equally incomparable by birth and as a wife. Repeatedly they experienced the vicissitudes of fortune. Here at length they rest together in the same harmony with which they lived. ’ The monument then extols the almost endless FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 213 perfections of both: ‘ She was a woman of exemplary probity and piety, of singular beauty and faith, the best and most renowned, not only of her own but of any age. She died piously and at peace on the 22 [recte: 27] of January, in the year 1563’ [recte: 1568]. (In each case a Roman numeral ‘V’ was omitted). ‘ He was a man most upright, a pattern to the nobility . . . [who] excelled in eloquence, in prudence, in innocence, in gravity, and was distinguished for his virtue and learning as much as for the splendour of his noble birth . . . [and was] a strenuous defender of religion. In his youth he shared in the studies of Prince Edward, son of King Henry. Though of abundant wealth, he was even more richly endowed in mind than in goods ; nor did he ever use his power to oppress his dependents. Full of honour and full of years, he paid the debt of nature in his 83rd year, on the 6th of April, 1621. He had two sons by his heroic wife.’’ Underneath the kneeling figure of Beauchamp the inscription reads: ‘ Edward , the first born, Lord Beauchamp, a man in every respect worthy of his titles and of his noble birth, who, seized by death before his father, left three sons by the Lady Honora of the ancient and renowned family of the Rogers: [namely:] Edward Lord Beauchamp deceased; William, now Earl of Hert- ford; and Sir Francis, Knight, together with one daughter Honora, married Sir Ferdinand Dudley, son of Baron Dudley 3 Below the other kneeling figure is inscribed: ‘ Thomas , the younger son, who married Isabel, daughter of Edward Onley, Esq., and, without issue, died before his father :’ On stylistic grounds the Hertford monument is believed to be the work of London stonecutter William Wright, during a period when London provided tombs for most of the country (White 1999; Sherlock 2008). In 1619, two years before Hertford’s death, Wright designed the similar Shrewsbury Monument in Westminster Abbey. In accordance with a proclamation from the Earl Marshall of England, requiring heraldic claims on monuments to be authenticated, his proposals were submitted to the College of Arms for verification and feature in a manuscript called the Booke of Monuments which shows Wright as the mason (White 1994). The Shrewsbury tomb has specific features in common with the Hertford monument in Salisbury, including recumbent effigies under a canopy, an arch coffered four rows deep and with boldly projecting paterae, decorative necking of columns, a composite order and attic storey, and a chest- sarcophagus with lion masks, clawed lions’ feet and pairs of Ionic columns (Figure 30). Wright also had idiosyncratic lettering, with a curving letter ‘E’ with its middle stroke looped, seen on the Hertford Monument. Fig. 30 The Shrewsbury monument in Westminster Abbey, showing close similarities with the (later) Hertford memorial in Salisbury Commissioning of the Hertford Monument There is no documentary indication of when the monument was erected, or who commissioned it (Fletcher 1927). However the un-evidenced assumption that it was erected on the instruction of Hertford’s older surviving grandson, William Seymour, cannot be sustained. There is no evidence that William ever participated in monumental commissions, nor had tendencies to extreme behaviour like his grandfather. William, erected a modest brass plaque at Collingbourne Ducis in memory of a son, Edward, who died 1631, but he is not known to have commemorated any of the other three sons or two daughters who died in his lifetime, his first wife, nor either of his parents. He himself was also buried at Bedwyn without monument. Further his relationship with Hertford was far from close. Hertford claimed in a letter instructing him to ‘ amend your life ’ that William’s behaviour ‘ would cause any grandfather to hate the memory ’ of his grandson (Longleat House Seymour Papers Vol. VI, ff.7-8). William also had cause to resent the way that Hertford had treated his father. William 214 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE also would have no personal memory to justify the effusive eulogy of his great grandfather, Somerset, or his grandmother, Catherine Grey, who had died 69 and 53 years earlier respectively, decades before he was even born. In contrast, the Hertford monument is in keeping with Hertford’s vainglorious personality, his unfulfilled lifetime ambition to be recognised as Catherine Grey’s legitimate husband (after relentless attempts were thwarted or resulted in punishment during his lifetime), and his desire to demonstrate that his son and heir (who had actually been disbarred from inheriting through illegitimacy) was Hn every respect worthy of his titles and his noble birth. ’ If tombs reflect the attitudes of their commissioners, the Salisbury monument bears Hertford’s signature, and is demonstrably distant from other members of his family. As Sherlock (2008) observes, most monumental construction occurs in erratic pulses when key committed individuals erected several tombs, rather than sequences of generations each commemorating their forebears. Tombs were important to occasional people for specific purposes. Hertford had a history of commissioning. William had ample tragedy and cause, but chose not to. Every part of the Hertford monument reflects his pre-occupations. His father is described as an ‘ Illustrious Prince ,’ a title seen on the tombs of his mother, sister, second wife Frances, and on the Seymour roll pedigree. His father’s achievements are described using identical, lengthy phraseology on his mother’s and on his own tomb. It is also wholly in keeping with his excesses elsewhere, commissioning one of the largest illuminated genealogical rolls ever produced, and one of the most lavish entertainments ever presented to a monarch. Hence it is here suggested that the monument was commissioned by Hertford himself, and the inscriptions drafted during his lifetime, although the tomb was not constructed until after his death. There remains some uncertainty about who is interred at the monument. The cathedral register records that Hertford was buried on May 24 1621. It is popularly assumed that Lady Catherine Grey was disinterred from her grave in Yoxford, Suffolk, where she had been buried on 21 February 1567/8, some 53 years previously, and was re-interred in Salisbury, but the monument itself makes no such claim, and there is no evidence from either Salisbury or Yoxford for this. Similar claims have been made that Lord Beauchamp’s body in interred at Salisbury. However the register shows that he was buried on 21 July 1612 at Great Bedwyn following his ‘ decease at Weeke’ (Wick), and evidence that the body was transferred is lacking. However, reburials were not always recorded in the cathedral register. A manuscript in the College of Arms shows that on 14 December 1722 the two sons of Charles, 6th Duke of Somerset, were disinterred from Westminster Abbey and reburied in the Hertford Monument in Salisbury, where their mother had been buried the day before, yet neither register records this (Eward 2002). Until the early 19th century the tomb carried the inscription: ‘ This chapel is the dormitory of the Dukes of Somerset ’ (Fletcher 1927). In fact only two were buried there. John 4th Duke of Somerset was interred on 6 June 1675, and Charles, 6th Duke, on 14 January 1748. Three further Dukes, William (2nd), William (3rd) and Francis (5th) were buried at Bedwyn, without monument. Discussion The pedigree roll and grand funerary monuments were commissioned when society was adjusting at the end of the Tudor and commencement of the Stuart era, amidst factionalism and uncertainty about how change would impact on the nobility and royal patronage. This was the period when both devices were at their peak, with fierce competition amongst the peerage for social status. The roll and the monuments each emphasise (or ignore) certain aspects of history for political, social or cultural reasons, whilst the Salisbury monument commemorating Hertford himself enables him to right those perceived wrongs which life had meted out, in an unfettered way, without fear of mortal punishment. Genealogy and lineage were especially important to the nobility and gentry at this time, even perhaps to the point of obsession (Stone 1965). Pedigrees were being widely commissioned and a small production industry surrounded the College of Arms. Many such rolls survive, though no full register of them appears in the public domain. It would have been impossible for the designer of the Seymour Roll to have drawn up the links of these 750 people, and their heraldic devices, without access to the library and resources of the College of Arms (Robert Noel, Lancaster Herald, pers. comm.). To some extent the limitation of information on lineage would have dictated inclusions and composition. The likelihood is that Heralds and officers of the College, together with other professionals, were FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 215 collectively involved in preparing it. The pedigree maker would have needed to check drafts with his client and College records, and then submit to professional writers and painters who undertook work for the College and its associated clients directly. The College holds reference to at least two charters granting arms to clients at the College, which have features showing close similarities with the floral border of the Seymour Roll. These include a grant by William Camden, Clarenceux, of 1612, to William Lambart, which has a floral design of periwinkles, daisies, strawberries and daffodils, often with botanically incorrect leaves (resembling the Seymour pedigree). The helm and mantle of the coat of arms is also similar. A grant by Gilbert Dethick to Geliy Merricke of 1583, has similar periwinckles, strawberries and daisies (NA SP9/1/14). Further it is clear that the heralds Dethick and Camden had direct links with the Seymours and Savernake, as the former produced the 1576 pedigree (Figure 22), and Camden (1607) described the Savernake Horn. A pair of handwritten volumes of 1620 by Augustine Vincent, called Presidents [= precedents], with a half title 'The Marshalling of all Estates and degrees at Public Assemblies and Funerals" has illustrations in two hands, including Vincent himself, and shows figures with costume, sword, scroll and posture virtually identical to those given on the Seymour pedigree, although not facial expression or complexion. Vincent, the Rouge Rose Herald, was deputy to William Camden, Clarenceux (Metcalf 1887). The regularity of style suggests that the paintings and figures were effectively incorporated through reference to a copybook. The lineage given in the Seymour pedigrees had direct relevance to contemporary interests, and future prospects, as well as a statement of the past. The peerage at this time included many new families. Stone (1965) suggests that the driver for interest in lineage was social integration, the welding into a homogeneous group a patchwork of those with diverse, and sometimes dubious, origins. New peers maintained tradition by demonstrating their shared roots with the old. As Schneider (quoted in Bouquet, 1996) observes, although racehorses have a pedigree they do not have a genealogy, which 'is a statement of social relations, not biological ones." The Seymour Roll abounds with boastful narrative of the successes of Hertford’s ancestors (whilst avoiding mentioning the execution of at least 14). However, whilst a lengthy pedigree might be a useful weapon in the battle for status (Stone 1965), there could be difficulties in communicating its content. It cannot even be certain that the roll was intended for regular display. Its audience was limited, access controlled and the detail meant that the message was simplified to demonstrating antiquity and noble connections, with an inference that kinship should lead to mutual support in common causes. It is unlikely to have been influential or relevant in seeking to restore the inheritance of Hertford’s sons, and data is lacking on which royalty and nobility were ever permitted to see it. The roll also uses heraldic devices as signs of affiliation, which as Coss (2002) notes 'served as icons of power and prestige". Arms might 'openly speak of authority, presence and even political intent" (Ailes 2002). Therefore the use of arms was not just decorative, but also imparted strong messages, demonstrating the ancient nobility and honour of the family. The inclusion of Catherine Grey, Hertford’s first wife and the mother of his sons, on the Seymour Roll, is subtle. She does not feature in her own roundel and shares Hertford’s, although her arms are impaled with his. In part this reflects the difficulties of dealing with three spouses and is a device used at least once elsewhere on the roll. However, the tangential reference to Catherine, whilst the two later wives have their own roundels, is in direct contrast to the Hertford monument in Salisbury, where Catherine lies next to Hertford in effigy. To include Catherine’s ancestry on the roll would carry huge risk - it would become a statement of Beauchamp’s claim to the throne; a blatant and potentially treasonable act. The roll still potentially provided utility in any such argument as Hertford’s credentials as a noble of royal blood himself are demonstrated. However, Hertford is the subject of the roll and not his sons. The collateral lines provide one further insight into the purpose of the roll. All four of the principal potential claimants to the throne on Elizabeth’s death are included: James I of Scotland, Arbella Stuart, Beauchamp and William Stanley. However, critically, only in the case of James is his ancestry shown (from Henry VII). Arbella is given only as the wife of William Seymour. William Stanley’s grandmother, Eleanor Brandon is shown, and described only as daughter of Charles Brandon and Mary, Queen of France. This confirms that the Seymour roll was not designed to challenge James’s right to succession, although it remained important not to undermine such a prospect. The Seymour Roll and the Seymour monuments in Westminster and Wiltshire are out of all proportion to the true distinction of the man who 216 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE commissioned them. Extravagant in size, ebullient in forging connections and using heraldry, they testify to Hertford’s sense of self-importance, keen to share his lineage and seek aggrandisement following a lifetime of failed aspirations. Their very scale shouts one-upmanship. By commissioning monuments in Westminster, Salisbury and Bedwyn, Hertford could ensure his triumphs were broadcast across the land. It is with his own monument in Salisbury that Hertford was most able to sanitise history. The carvings claim that he and Catherine could ‘ rest together in the same harmony in which they lived,' even though the only time they shared the same roof was when they were in the Tower of London. No court of law or monarch had ever acknowledged the validity of their marriage, and Hertford’s purchase of lands which had reverted to the King, was tantamount to acknowledging that his son had no birthright. According to the monument, Catherine, instead of descending into a depressive stupor, could ‘ die at peace.' Hertford’s disrespectful son Beauchamp, who said he would rather die than be in his father’s presence, is resurrected from his own prior death and dutifully kneels at his father’s side. Beauchamp, instead of being castigated for being wedded to a pirate’s daughter, marries into the ‘ ancient and renowned family of Rogers.' And in direct contrast to the Seymour Roll, where Catherine is mentioned only parenthetically, she is elevated as central to his ongoing lineage, whilst his other wives, including his widow, are ignored altogether! The sheer scale of the monument might explain why it was located in Salisbury, which may have been being better able to furnish privileged floor space and accommodate a memorial of this magnitude than Westminster Abbey, where its extravagant claims might also be more closely scrutinised, and where he had created a monument for his second wife, who was ignored on the Salisbury tomb. Hertford had mansions at Wolfhall, Easton, Tottenham and Amesbury in Wiltshire, and at Elvetham and Netley in Hampshire, although he spent a great deal of his last years at Cannon Row in London. It may have been when he lived at Amesbury during the early years of the 17th century that his plans for a monument at Salisbury were formulated and negotiated with the cathedral authorities (although on stylistic grounds, the monument construction was later). Through the monuments, Hertford re-tells history selectively, as it should have been, rather than as it was. He sets his curriculum vitae in stone, subduing alternative perceptions. As sepulchres of lithic name-dropping, Hertford selectively embellishes or sidesteps incidents, honing his image for posterity, whilst drawing on heraldic and genealogical links to bequeath his descendents honour and advancement. Hertford was partially successful in beguiling those who read his selective version of events. Hallam (1854) saw Hertford’s epitaph as ‘ an affecting testimony to the purity of an attachment rendered sacred by misfortune’. Visitors to Salisbury remain enticed by romantic thoughts of Hertford as an innocent victim of love, rather than a man relentlessly scrambling for advancement, whose ambitions could be realised only in death. Conclusions The Seymour Roll and the monuments erected to Hertford’s family are elaborate creations which complement each other. In addition to being visually spectacular, the roll is a meticulously researched, albeit selective, statement of Hertford’s genealogy, containing the relationships of 750 individuals. It was prepared drawing on resources at the College of Arms, probably involving officers, associates and artists linked with the College during its research and preparation. It shows stylistic similarities with certain other contemporary rolls. It accurately reproduces evidence derived from charters, deeds, inquisitions, wills and other sources to support its claims. Its primary purpose was to demonstrate Hertford’s family eminence, and his kinship with other eminent individuals, at least 57 of whom were alive at the time of its completion. It is discerning when emphasising alliances, and ignores ignoble ends. The content of the roll was adjusted to enable important links, however distant, to be displayed (e.g. with King James, Hertford’s 6th cousin, twice removed). The roll has been subject to restoration, and two phases of limited updating. Some 96% of individuals featuring on the roll today were included at its completion in 1604. Approximately 28 individuals were added in the latter 17th century, and a further seven individuals included when the text was re-inked during restoration work undertaken in 1952. The erection of Seymour monuments in Bedwyn, Salisbury and Westminster Abbey enabled Hertford’s self-perceived eminence to be broadcast widely. Whereas the roll is largely a simple statement of links, Hertford used monumental masonry to trumpet his family’s honours and exaggerate their FORGING ALLIANCES: THE LIFE OF EDWARD SEYMOUR, EARL OF HERTFORD 217 nobility, extolling their virtues for posterity. In particular, the Hertford monument in Salisbury, one of the largest tombs ever erected, which is here suggested as being commissioned by Hertford himself, enabled him to salve his own conscience whilst correcting the wrongs that life had meted out. He could achieve what he had failed to gain in his lifetime, falsely validating his marriage to Catherine Grey and the noble birthright of his (legally illegitimate) children. This monument, by showing his and Catherine’s effigy, whilst ignoring his second wife and his widow, contrast with the roll pedigree, where Catherine is mentioned only parenthetically. Hertford’s extravagances were pursued during the Tudor-Stuart transition, at a time of general uncertainty about position and royal patronage, and testify to the ostentation and excesses of a man desperate for social position. Acknowledgements Research on the Seymour Roll by Andrew Douglas formed part of a thesis in fulfilment of an MSc in Advanced Professional Studies (Genealogy) at the University of Strathclyde, Glasgow. Photographs of the Seymour Roll Pedigree (WSA 1300/376) in Figures 6, 7, 9-10, 14-15, 18 and 19, and of the King John Charter (WSA 1300/12) in Figure 17 are reproduced courtesy of Wiltshire and Swindon Archives. Quotations from the Seymour Papers at Longleat House, and the images in Figures 22 and 23 are included by agreement of the Marquess of Bath. We are grateful for professional advice from Steve Hobbs, Archivist at Wiltshire and Swindon History Centre; Robert Noel, Lancaster Herald of the College of Arms; Peter Sherlock, Dean of Theology at Melbourne; and from Susan Eward and the Rev Bill Elliott from Salisbury Cathedral. We would also like to thank Jack Oliver, Joan Davies, Joanna Ramsay and Nathaniel Bathe. References Primary sources Wiltshire and Swindon Archives (WSA) 1300/163: Assignment by William Marquess of Winchester, Lord Treasurer of England 1300/166: Inspeximus of letters patent, Edward IV, conferring dukedom of Somerset and lands. 1300/167: Inquisition post mortem of Edward, Duke of Somerset. 1300/174: Act limiting the possessions and hereditaments of Edward, late Duke of Somerset. 1300/376: The Seymour Roll Pedigree of 1604. 3790/1/3: Interrogation Hertford & Catherine Grey, Lord Bruce MS Book volume VI Longleat House Portland Papers, Vol. 1, Papers of Robert Harley, and Margaret Harley, Duchess of Portland Seymour Papers, Vol. V, Correspondence and papers of Earl of Hertford from 1559 to 1611 Seymour Papers, Vol. VI, Genealogies cl 592- 1595 and Papers of William Seymour. Seymour Papers, Vol. XI, Correspondence and memoranda of Earl of Hertford, 1559 to 1612. Seymour Papers, Box XVIII, Letters Patent 1547-1660 and Pedigree of 1576. British Library Manuscripts (BL) Additional Manuscript 32092, Letter Bishop Whitgift to Earl of Hertford, 4 November 1585 Additional Manuscript 33749, Interrogations of Hertford and Grey in Tower. Additional Manuscript 35830, Hardwicke Papers, documents of the Earls of Hardwicke. Additional Manuscript 38170, Drafts for Commission Beauchamp’s claim to inheritance. Additional Manuscript 63543, Papers relating to Earl of Hertford and Catherine Grey ff27-82. Cecil Papers, 1607, Petitions 2368, The Earl of Hertford’s Petition to the King 1607. Cecil Papers, 1607-8, A Breviate of the Earl of Hertford’s Cause, with petitions by the Lords Mounteagle and Beauchamp. Cotton Faustina A XI, Tranactus causa Lady Catherine Grey and Earl of Hertford, Cotton Titus C VII, fl lb. Sentence in the Star Chamber against Hertford. Cotton Vitellius C XVI v2 f.5 16, Papers re marriage of Earl of Hertford and Catherine Grey Sloane 1786 f25, Account of the Earl of Hertford’s fine in the Star Chamber, 5 Eliz. 1562. National Archives C 136/19/1 : Inquisition post mortem Henry Esturmy, 1381/2 C 270/34/6: Notarial instrument of appeals by Beauchamp against bastardization 1577-1587. SP 9/1/14: Grant of Arms to Geliy Mericke of Hasguard, by Sir Gilbert Dethick, Garter SP 38-9-28-2: Articles of Agreement King James, Hertford, Beauchamp and Mounteagle SP 46/10: Commission into marriage of Earl of Hertford and Lady Katherine Grey, 1562. College of Arms Vincent, A., 1620, Presidents, The Marshalling of all 218 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE Estates and degrees at Public Assemblies and Funerals. (Manuscript volume). Calendars State Papers: Domestic Series of the Reigns of Edward VI, Mary, Elizabeth Ontario: Tanner Ritchie and University of St Andrews Vol. 1 1547-1580 Vol. 2 1581-1590 Vol. 4 1595-1597 Vol. 5 1598-1601 Vol. 6 1601-1603 Vol. 6 Addenda 1547-1565 Vol. 7 Addenda 1566-1579 Vol. 8 1603-1610 Cecil Manuscripts: Manuscripts of the Marquis of Salisbury at Hatfield House Historic Manuscripts Commission. Ontario: Tanner Ritchie and University of St Andrews. Part XIII, Addenda 1296-1597 Part V, 1694-5 Part XII, 1602-3 Secondary sources AILES, A., 2002, ‘Heraldry in medieval England: symbols of politics and propaganda’, in P Coss and M. Keen (eds), Heraldry, pageantry, and social display in medieval England. Woodbridge: The Boy dell Press ANON., 1591, The Honorable Entertainement given to the Queene’s Majestie, in Progresse, at Elvetham in Hampshire, by the Right Honourable the Earle of Hertford, London. BOUQUET, M. 1996. Family trees and their affinities: The visual imperative of the genealogical diagram. The Journal of the Royal Anthropological Institute 2:1. (63 note 4). BRUCE, J. (ed.), 1868, Diary of John Manningham 1602- 1603, London: Camden Society BRUDENELL-BRUCE, C., (Earl of Cardigan), 1949, The Wardens of Savemake Forest. London: Routledge and Kegan Paul Camden, W., 1607, Britannia, sive florentissimorum regnorum Angliae, Scotiae, Hiberniae, et insularum adiacentium ex intima antiquitate chorographica descriptio. London: George Bishop and John Norton Coss, P, 2002, ‘Knighthood, heraldry and social exclusion in Edwardian England’, P Coss and M. Keen (eds), Heraldry, pageantry, and social display in medieval England, 39-68. Woodbridge: The Boydell Press DOLEMAN, D. (pseudonym for Parsons, R.), 1594,^4 Conference About the Next Succession to the Crown of England. Antwerp. DOUGLAS, A., 2010, The Seymour pedigree. Unpublished University of Strathclyde postgraduate diploma etude DOUGLAS, A., 201 1, Reliable genealogical records, or objects of self-aggrandisement and power politics? A critical review of the accuracy and purpose of two ‘great’ roll pedigrees from the early seventeenth century. Unpublished University of Strathclyde MSc dissertation ELLIS, H., 1827, Original Letters illustrative of English History , 2nd Series. London: Harding and Lepard EWARD, S. 2001. From the Archives. Salisbury Cathedral News FISHER, F. J., 1936, The State of England 1600 by Sir Thomas Wilson, in Camden Miscellany. London: Camden Society FLETCHER, J., 1927, The Hertford Monument in Salisbury Cathedral. Devizes: George Simpson GRISTWOOD, S., 2003, Arbella, England’s Lost Queen. London: Bantam Press HALLAM, H., 1854, Constitutional History of England, Vol. 1,214. London: John Murray JACKSON, J. E. 1875. Wulfhall and the Seymours. WANHM 15, 140-207 LLOYD, R., 1967, Dorset Elizabethans. London: John Murray LOCKE, A. A., 1911, The Seymour Family History and Romance. London: Constable LOOMIS, C. 1996. Elizabeth Southwell’s manuscript account of the death of Queen Elizabeth. English Literary Renaissance, 26:3, 482-509 METCALF, W. C. (ed.), 1887, The visitations of Northamptonshire made in 1564 and 1618-19. London: Harleian Society NEW ADVENT, 2009b, New Advent Bible: Wisdom 4. PATTERSON, R. F., 1999, Seymour Dawn, The Origins and Medieval History of the Seymour Family. Aberystwyth: Anduma Publishing REED, P. C. 2008. Descent of the St Maur family of Co Monmouth and Seymour family of Hatch, Co. Somerset. Foundations 2:6, 390-442 REED, P C. 2010. The Seymour family of Hatch, Somerset and de la Mere family of Little Hereford. Foundations 3:3,217-266 SHERLOCK, P, 2008, Monuments and Memory in Early Modem England. Melbourne: Ashgate University of Virginia, 2011, Wisdom of Solomon, from the holy Bible, King James version (Apocrypha). St MAUR, R. H., 1902, Annals of the Seymours. London: Kegan Paul Trech and Co. STONE, L., 1965, The crisis of the aristocracy: 1558-1641. Oxford: Clarendon Press STRYPE, J., 1633, A Survey of the Cities of London and Westminster. London: A. Churchill, et al. WAGNER, A. R., 1952, The records and collections of the College of Arms. London: Burkes Peerage WARD, J. 1860. Great Bedwyn, WANHM 6, 261-316 WHITE, A. 1994. The Booke of Monuments Reconsidered: Maximilian Colt and William Wright. Church Monuments 9, 62-67 WHITE, A. 1999. A Biographical Dictionary of London Tomb Sculptors cl560-cl660. Walpole Society 61, 1-162 YONGE, C. D., 1891, The orations of Marcus Tullius Cicero. London: George Bell and Sons Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 219-224 John Buckler (1770-1851): the topographical artist in Wiltshire by Robert Moody Towards the end of his life, John Buckler, the topographical artist, reckoned that he had drawn over 13,000 sketches of cathedrals, churches and houses during the course of his career with most of these being situated in six counties. The first of the counties to be extensively visited by him was Wiltshire and the purpose of this paper is to give some account of his work in the county that resulted in 690 of his drawings being collected in 10 volumes and now being considered one of the great treasures of the Society's art collection in Devizes. John Buckler was born in 1770 at Calbourne in the Isle of Wight. At the age of 15, he began his long association with Magdalen College, Oxford by commencing employment as clerk to the Steward there, R. B. Fisher, and eventually becoming Bailiff and Collector of London Rents for the college. At the same time he became articled to a surveyor and builder in Southwark and began preparing plans and elevations for speculative development on property belonging to the College. It appears that under the patronage of Dr Routh,1 President of Magdalen, he published in 1797 two engravings of the college, followed two years later by similar views of Lincoln Minster, and thereafter continued to publish many aquatint engravings of cathedrals, churches and private mansions. He also began practising as an architect and accepting commissions to execute drawings of ecclesiastical and other buildings over a wide area of the country. Buckler’s connection with Wiltshire seems to have commenced at much the same time as the publication of his engravings in 1797, as in the Society’s art collection is a watercolour of the south front of Stourhead House painted prior to its remodelling which took place between 1796 and 1800.2 The painting is unsigned but in the view of Richard Hatchwell ‘the title beneath the watercolour is unquestionably in Buckler’s hand. This must be one of the first paintings that Buckler did for Colt Hoare, before he had developed the well-known style in which the majority of his watercolours are done’.3 It seems, therefore, that when a young man, Buckler was already in touch with Colt Hoare,4 who was to continue to employ him as an artist for many years. In about 1801 Buckler was appointed Bailiff and Collector of Rents for Magdalen College and thereafter appears to have had ample time over the following decades to demonstrate his considerable talent in executing meticulous watercolour drawings of topographical subjects. Shortly before 1803, Colt Hoare commissioned Buckler to execute watercolour drawings of all the churches in Wiltshire, including the fonts, monumental brasses and architectural elements of some of them, and of a number of the great mansions of the county and some other buildings, particularly in Salisbury. These drawings, prepared from preliminary sketches,5 together with a small number that Buckler must already have executed and bringing the total to no less than 690, were eventually mounted into 10 volumes the first bearing the dedication: Fair Orchard, South Widcombe, East Harptree, Bristol BS40 6BL 220 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE To Sir Richard Colt Hoare, Baronet, of Stourhead, this Collection of Drawings of the Ecclesiastical Antiquities of Wiltshire made under his munificent patronage, is with respect and gratitude most humbly inscribed by his much obliged and faithful devoted servant, John Buckler. These volumes formed one of the many treasures of the Library at Stourhead until sold in 1883 and were purchased by the Society in 191 8.6 Buckler must have realised that the task that he had agreed to undertake would take more than a few years to complete. In 1803 drawings were completed after visits to fifteen towns and villages in the county including Caine, Devizes and Marlborough. In the spring of the following year Colt Hoare wrote to Buckler: I have received the case of drawings safe - like them very well; and do not object to the prices you have fixed for each of them - if you will undertake to proceed on the same terms I will give you further directions as on [?] & extensive a plan as necessary of the whole County... it will now be necessary to proceed with some degree of regularity so as to finish each parish as you go... if you agree about the continuation of the survey of Wilts, perhaps you might be able to come down to me in Easter week to arrange a future plan. I like the etching of Sarum Cathedral very well - its design point of view so it will be necessary for me to have a view of each front to complete my set. I hope it will not be inconvenient for you to wait for payment of these drawings till Lady Day or Easter by which time you may depend on the money.7 Soon after this Colt Hoare wrote again to him from Stourhead: The drawings arrived safe & I will keep the Romsey collection but to render it complete you must make two outside views of the church... nothing to me appears so interesting as these views of our fine old ecclesiastical buildings and I hope we shall both of us live to form a numerous collection of them; but as Wiltshire is my most material object I must first complete that series - as I cannot find cash for ah. I think it will not be necessary to give you the trouble of a journey to this place as I can settle my plans by letter - 1 should like to take in ah the County (if not too much) between this place & Salisbury - and as many of the churches are very insignificant and some of laborious architecture, I think you will be able to complete them - I will send you a list of the churches as well as a remittance for the remainder of my debt to you before I leave the County for Wales.8 And so it was that in 1804 he complied with Colt Hoare’s request by concentrating on the south of the county, with drawings prepared after visits to at least 56 villages situated entirely to the south of Salisbury Plain. In 1805 he continued to execute drawings of churches and other buildings in the southern part of the county with over 60 places attracting his attention. In the following year he once again added over 60 villages to his list of places visited, now mostly in the north of the county, and in 1807 a further 37, yet again to the north of Salisbury Plain. So far as is known, Buckler did not keep any diary or other record of his work between 1804 and 1807 but his addition of a date to most of his drawings provides what must be a reasonably accurate picture of his travels in the county. In August 1807 he wrote from Salisbury to William Cunnington9 of Heytesbury: Your kind favor has been conveyed to me at this place; which I leave tomorrow for London having completed my intended excursion for the Summer in this part of the country. . . I extremely regretted I had not it in my power to pay my respects in person to you at Heytesbury this summer. I hope next year to be more fortunate. It is with great pleasure I hear that our most excellent and good friend Sir Richard Hoare is making progress in his great undertaking, the history of ancient Wiltshire. I was in hopes I should have had the happiness of meeting him sometime this summer in the neighbourhood of Avebury, but some circumstance or another has prevented him visiting that part of the country for the present. How much was I delighted in viewing the noble Church of Edington which I had not visited till this summer. I consider it by far the finest church I have yet drawn. Steeple Ashton is of a different character, but a fine thing of its kind. I have the satisfaction to inform you I have now completed the survey of above two-thirds of the country and I hope please God to have finished the remainder in about three years...10 Whilst in Avebury in 1 807 he had drawn two views of the church there. He had also executed two of the church at Edington that he had so much admired, and the font and the piscina there. As he had anticipated, by 1811 he eventually completed the enormous task of visiting and drawing views of virtually every church and major building in the county. In 1808, some 40 places had been visited and JOHN BUCKLER (1770-1851): THE TOPOGRAPHICAL ARTIST IN WILTSHIRE 221 in 1809 a further 44, including Malmesbury, where he made fourteen drawings of the Abbey and other buildings there. In 1810 the remaining 32 places in the north of the county were finally visited with just the chapel in Boyton church, Longford Castle and the interior of Stockton House being dated 1811. In calculating the number of places visited between 1803 and 1811 (mentioned above), Salisbury has not been included. Buckler drew no less than 62 images of the Cathedral, Close and city between 1803 and 1812, most of which are to be found in volume 9 of the collection. Some 18 of them were executed in 1805 and so it is clear that he spent some time in the city in that year but made numerous other visits while concentrating on other parts of the county. For instance, in July 1812 he wrote a long letter to Colt Hoare reporting on his visit to various places in Derbyshire, Staffordshire and Warwickshire, concluding: It is my intention to come to Salisbury for a day or two on Monday next and if opportunity offers I propose to take a sketch of Pyt House, but I believe I shall give up Wimbourne Minster this summer... I was very happy to hear from Lewis that you received the two drawings and that you approve of them.11 It is doubtless that this sketch of Pythouse was engraved by George Hollis12 and appears in volume IV, part I of Colt Hoare’s The Modern History of South Wiltshire (1829). As well as providing an invaluable picture of the appearance of many churches before their alteration by later restorers, Buckler’s depictions of a remarkable number of chancel windows totally or partially blocked up supports the proposition that lay improprietors were still neglecting their responsibility for chancel repairs at the beginning of the 19th century. Following completion of Buckler’s work, Colt Hoare arranged for 690 of the pictures to be bound in 10 volumes, after which he listed them all in a small notebook with the first page of it reading: This Collection was made by Mr John Buckler from a personal survey of all the Churches and Mansions within the County of Wilts. The views are arranged in Deaneries, not in Hundreds. The whole series is contained in 10 folio Volumes RCH13 By no means all of Buckler’s images of Wiltshire houses and churches executed between 1803 and 1811 were bound in these volumes. For instance, in the Society’s art collection are two very fine watercolours of Wilton House painted in 180414 and, uniform with these, a painting of the Holbein porch at Wilton. Also in the possession of the Society is a pen on tracing paper drawing of the Holbein porch15 that is almost certainly the preparatory sketch for the later watercolour. In about 1958 Richard Hatchwell published A Catalogue of Books and Drawings relating to Wiltshire, and Books by authors connected with the County that he offered for sale. Included in the catalogue was a collection of fifteen watercolours by Buckler, six of which were dated between 1803 and 1810. It is clear that Buckler was working in Wiltshire again in 1817. Four views of both the interior and exterior of Wardour Castle bearing this date were included in the collection offered for sale by Hatchwell. Further, a sketch of the great drawing room in the Bishop’s Palace in Salisbury, dated 1817, is reproduced in Salisbury: The Houses of the Close with a photograph of the room16 allowing the reader to compare Buckler’s sketch with the room’s present appearance. And in the William Salt Library in Stafford are preserved three watercolours of the church, farm house and monastic remains at Ansty all dated 1817. In 1812, at the request of Colt Hoare, he had produced a design for the front of the inn at Stourton, now known as The Spread Eagle, and was later to design many buildings although none, so far as is known, in Wiltshire. Further, his drawings of very many cathedrals were highly praised by Colt Hoare who, in his catalogue of his library at Stourhead published in 1815, wrote that: . . .to the later [i.e. Buckler] much praise and merit are due, for having by his own means and exertions nearly completed a magnificent and well executed series of the English and Welsh Cathedrals, upon so large a scale, as to comprehend a satisfactory detail of their minute and delicate architecture.17 In 1822 the first part of Colt Hoare’s monumental Modern History of South Wilts was published and in that volume appeared six engravings of Buckler’s drawings with a further four in the next part published in 1824. It is certain that Colt Hoare was instrumental in procuring the many engravings of drawings not only by Buckler but also by Philip Crocker18 and a number of others. 222 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE In May 1822 Buckler rendered to Colt Hoare an account amounting to £67 Os. 6 d. for work done since September 1821, £50 of which Colt Hoare duly paid with the balance being carried forward to the next account delivered by Buckler in March 1823. In the first of these accounts the following charges appeared for drawings included in the first two parts of the Modem History of South Wilts : 1821 £ s d Oct 31 Drawing of Longleat House 4..4..0 Interior of Hall 5..5..0 Effigy in Boyton Church 3..3..0 In the second of the accounts the following charges were included: 1822 £ s d Dec. 7th Moiety of Expenses to Stourhead 5.. 0.. 0 Norman Doorway at Little Langford 1..11..6 Arches doorway Stapleford Church 1 .. 1 1 ..6 Arches doorway & font Durnford Church 1..1 ..0 3 Doorways Winterbourne Church 1..1 ..0 Cross at Stourhead 1..11..6 Fonthill House 1805 2.. 2.. 0 24th Boyton and Stockton houses 2. .2.. 0 Room at Stockton 2..2.. 019 In January 1824 Buckler looked to Colt Hoare for payment of £58 11s .6 d. and at the end of that year rendered an account for a further £46 17s.0d. with a total of £149 7s. 6 d. claimed for work done during 1825 and 1826. Almost all these charges were in respect of drawings of churches and houses within Wiltshire.20 Although in the preface to the second volume of his Modern History of South Wilts (1825), Colt Hoare acknowledged his thanks to “Mr. J. Buckler who made the greater part of the Drawings, and to Mr. George Hollis, who engraved them”, throughout all four volumes published between 1822 and 1844 Buckler and Philip Crocker contributed in almost equal number to the illustrations. However, Colt Hoare’s high regard for Buckler’s talents is demonstrated by Buckler’s name being included in the list of mostly eminent people to whom he presented complimentary copies of the volumes of his Modern History covering the hundreds of Branch and Dole (1825) and Dunworth (1829).21 Some years later Buckler’s grandson noted that it was Colt Hoare’s commission that ‘decided his brains for antiquarian pursuits’ and that later in life he averred his productivity to his ‘uniformly temperate and regular method of living’.22 In 1823 Colt Hoare published his Hungerfordiana: or Memoir of the Family of Hungerford. In this volume are engravings of two drawings by Buckler, one of the Gateway at Down Ampney and the other of Farleigh Castle and Chapel - both being close to the borders of Wiltshire in Gloucestershire and Somerset respectively. In 1810 Buckler was elected a Fellow of the Society of Antiquaries and so would have been well acquainted with John Gage,23 who, in 1829, became director of the Society. On 24 March 1823 Gage and Buckler, in company with ‘Mr Parker’,24 set out from London at the commencement of a tour, their chief intention being to visit Colt Hoare at Stourhead. On the following day the travellers arrived in Bath having on the way visited Avebury, Silbury Hill and Lacock as well as Corsham to view Paul Methuen’s pictures.25 Gage recorded in his journal26 of the tour that, having viewed Longleat House, they: ...got to Stourhead for dinner and found Sir Richard Hoare impatient to see us. We took a slight survey of this elegant mansion and in the evening were occupied with the drawings of places in Wiltshire... On the following day: The Library, the picture gallery and the drawings fixed our attention for some hours and a walk in the grounds filled up the measures until dinner after which drawings and mss were our final resources... On the 28th the visitors made an excursion into Somerset and then two days were spent at Stourhead where the grounds, described by Gage as ‘quite a paradise’, were examined as well as the hot houses with Colt Hoare’s geraniums being especially admired by Gage who, with Buckler, then walked to Alfred’s Tower. On the 31st the travellers left Stourhead and spent the following two nights at Wardour where they ‘surveyed the different apartments in this fine house, the ancient castle and the grounds. . .’. On 2 April they arrived in Salisbury after visiting Wilton House and were too late for the afternoon service in the cathedral. Buckler had, of course, spent many hours in the past sketching and then drawing numerous views of the cathedral. Gage lamented that all the painted glass had been removed from the windows of the Chapter House and recorded that ‘Mr Buckler remembered two of the windows being full of painted glass’. At some stage on their tour, the travellers must have been JOHN BUCKLER (1770-1851): THE TOPOGRAPHICAL ARTIST IN WILTSHIRE 223 joined by Anthony Salvin27 as on the following day Gage recorded : Parted with Mr Parker and Mr Salvin and proceeded with Mr Buckler to Winchester on our return to town. . .Before quitting Salisbury we took another look at the cathedral. The Audley Chapel in the choir is very elegant. How could Wyatt have destroyed the projecting chapels leading from the Lady Chapel?... After describing in some detail his visit to Winchester Cathedral with Buckler, Gage concluded his journal by declaring ‘thus ended an agreeable tour’.28 Not every drawing made by Buckler at Colt Hoare’s request was of an ancient church or of a great house in Wiltshire. In the Society’s collection is a drawing of Bremhill rectory given by Colt Hoare to the wife of the poet William Lisle Bowles in 181 8.29 Another Buckler work belonging to the Society is a watercolour view of the back of the house in Salisbury Close known as ‘Aula Le Stage’, painted in 180530 and, coincidentally, to become Bowles’s house when he was appointed a residentiary Canon of Salisbury Cathedral in 1828. In Richard Hatchwell’s opinion this was an unusual view for Buckler to draw and may have been commissioned by the historian and antiquary William Coxe, Bowles’s predecessor in the house.31 In December 1825, the badly constructed tower of Fonthill Abbey, some 225 feet high, suddenly collapsed. Soon after this Colt Hoare wrote: I was so anxious to see it in its dilapidated state, that I sent an artist32 to take a view of it the week after its fall, and have since had the drawing engraved. Mr Buckler had previously made two large drawings of the Abbey in its perfect state, which, in picturesque effect, are far inferior to it in its ruined state; for the tower, from its excessive height, was out of all proportion.33 It was not only Colt Hoare who commissioned works from Buckler. George Watson Taylor34 arranged for Buckler to execute five drawings of Edington Priory and the remains of the priory house there, all inscribed “To Sir Richard Colt Hoare, Baronet of Stourhead from George Watson Taylor 15 July 1826”. These drawings, with another by Buckler of the monument on the south side of the chancel in the Priory, were purchased by the Society in 1954 and are now bound in one volume.35 Although Buckler continued to produce a phenomenal number of sketches and drawings after 1826 until his death in 1851, none appear to be of buildings in Wiltshire. Thus his greatest legacy so far as the county is concerned is the Society’s magnificent collection of his drawings executed between 1803 and 1811. They form a unique record of many Wiltshire buildings since destroyed or altered. To Buckler is owed a substantial debt. Endnotes and references 1 Martin Joseph Routh (1755-1854), was President of Magdalen College, Oxford between 1791 and 1854. 2 Accession number 1990.590. 3 Richard Hatchwell, Art in Wiltshire from the Wiltshire Archaeological and Natural History Society Collection (Devizes, 2005), p. 112. 4 Sir Richard Colt Hoare (1758-1838). 5 British Library Add. MSS. 36356-97 42 volumes of original sketches by Buckler and his sons. 6 For an account of the purchase of the volumes by the Society, see: The Buckler Collection of Wiltshire Drawings. WANHM 40 (1918), p. 148. 7 Colt Hoare to Buckler, n.d: Bodleian Library, University of Oxford, Buckler and Hollis collection, MS.Eng.lett.a.l, p. 105. 8 Same to same, 4 Mar.1804: Bodleian Library, University of Oxford, Buckler and Hollis collection, MS.Eng.lett.a.l, p. 106. 9 Wiliam Cunnington (1754-1810), antiquary. 10 Buckler to Cunnington, 14 August 1807: WANHS Library, Devizes [hereafter WANHS Lib.], MS 2598. 11 Same to Colt Hoare, 29 July 1812: Wiltshire and Swindon History Centre [hereafter WSHC], 383/907. Pythouse near Tisbury was the recently built seat of John Benett (1773-1852), later one of the members of Parliament for the county. 1 2 George Hollis (1793-1 842), engraver. 13 WANHS Lib., MS.739. 14 Accession numbers 1982.7630 and 1982.7632. 15 Accession number 1982.1862. 16 Susan Leiper, (ed.), Salisbury: The Houses of the Close (London, 1993), pp. 64-5. 1 7 Sir Richard Colt Hoare, A Catalogue of Books relating to the History and Topography of England, Wales, Scotland, Ireland compiled from his Library at Stourhead, in Wiltshire (London, 1815), p. 49. 1 8 Philip Crocker (1780-1 840), surveyor and draughtsman to archaeological excavations. 19 WANHS Lib.: MS.725. 20 Ibid. 21 WANHS Lib. Colt Hoare’s Account Book of the printing expenses and lists of persons presented with complimentary copies. Folio account book bound in vellum. 22 C. A. Buckler, Buckleriana: notices of the family of Buckler 224 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE (1886), p. 51. 23 John Gage, later John Gage Rokewode (1786-1842), antiquary. 24 Probably Thomas Lister Parker (1779-1858), antiquary. 25 Paul Methuen, later 1st Baron Methuen (1779-1849). 26 WSHC, 2606/1, Ms Journal of a tour by John Gage Esq. 27 Anthony Salvin ( 1799- 1881), architect. 28 Op. cit., WSHC, 2606/1. 29 Accession number 1982.8130. 30 Accession number 1991.503. 31 Richard Hatch well, op. cit., p.126. 32 I.e. John Buckler. 33 Sir Richard Colt Hoare, The Modern History of South Wiltshire (London, 1829), vol. IV, pt. I, p. 27. 34 George Watson Taylor (d. 1 84 1 ), owner of the Erlestoke estate. 35 Accession numbers 1982.977-82. Wiltshire Archaeological & Natural History Magazine, vol. 105 (2012), pp. 225-230 The Heytesbury Collection: visitors to William Cunnington’s Moss House by Lorna Haycock Using the extensive collection of letters to and from William Cunnington (1 754-1810) and various printed sources, it has been possible to analyse the status and interests of the visitors to the Moss House and to set the museum in the context of contemporary collections and the upsurge of interest in archaeology, botany and geology at the turn of the 18th century during the Romantic Revival. Among the many treasures in WANHS library is a slim bound volume of manuscript notes entitled ‘The Heytesbury Collection’ describing the contents of the barrows excavated by William Cunnington under the patronage of Sir Richard Colt Hoare. At the back of this volume is a synopsis of the 425 antiquities in the Moss House and a list in the hand of Cunnington’s grandson William Cunnington III of the visitors to the museum in the garden at Heytesbury (Figure 1). The book also contains the famous pencil sketch by J. P. Wyndham of Cunnington and his daughter Elizabeth ‘bringing home the Stonehenge Urn’ (Figure 2) From 1802 onwards, visitors ranged from antiquaries, clerics, poets and local worthies to county society and aristocracy. The discovery of the ruins of Herculaneum in 1715 had kindled interest in archaeology and ‘viewing the antiquities’ had become a recognised pastime for gentlemen, particularly since the issue of the journal Archaeologia in 1770, and the publication in The Gentleman’s Magazine of the Herculaneum excavations.1 The site would have been known to the more affluent among Cunnington’s visitors, such as the Earl of Pembroke, the Marquis of Tavistock, and Sir Rchard Colt Hoare, while on the Grand Tour: ft/Mift ef a ft'ff'.uut feemft (y foft £t /tt'ityhit , in /ie/r, /if- ftriftrH itiift S/'tY/n f‘; h<‘*t Tc b/tjj 3t n ftri ft>t , n. 7 itfltitintf ft/rt/ n Jif/H i cefty of *7 'tttadi uftt/ift fic ft'iMiifcd ft .nttiftPii j Stud \ ftu.i ~i/t( fuetn'A j a. W yc- 7 / / 5K-C- fc 9/ } c ■■'Wtvuf/i, i /jez/ft ■Av '(i/s/; ^ / fCr-tr* /ssp /, f/tUi+pt, w f/eyK'-urPi Mey . /ffiz &o-Cs\ eft fy'M 9ft*/* /hi.a.fe • / J/*uJt** , %a \ dt&y 9/- Jt*. fftr'fitwjihxdjy Me*. hr* AW j&Trf hr' Je***fU , r/Jl, faftv sfurticn^- 1 " Y/t Multuei < AAs ftetdi. gc<< - ‘Lift-*' /i i /> JW' — fry' ■ ^»4 . A. A'i. r /< tf eg ftfrf — A/r '/ Z. edd'/reA. K/Jn'J/aeu ftp ‘iAt./-. i /, ,, f .'• * C./A -• M J ftr . . __ A/,- A;// . AL-en; /Ait'- td . AA-J1 . ftr J-rAu, c/3n^r<-~ , ra. f - nH JL/.C ef f Ad. MfM Jt lhawt , . fttnA . A i t : y /. / Jy . -£*/. &rA^/