— WILTSHIRE STUDIES

The Wiltshire Archaeological and Natural History Magazine

Volume: 24. 2007s 3

4 hen

a a |

Sau

ja MLE”

ot yen is wf

The Wiltshire Archaeological and Natural History Magazine Volume 94

2001

Published by

The Wiltshire Archaeological and Natural History Society 41 Long Street,

Devizes, Wilts. SN10 1NS

Telephone 01380 727369

Fax 01380 722150

e.mail wanhs@wiltshireheritage.org.uk

Founded 1853

Company No. 3885649

Registered with Charity Commission No. 1080096 VAT No. 140 2791 91

Frontispiece. Stained glass medallion and trefoils formerly in Salisbury Cathedral Chapter House. See Fig. 15, p. | 28

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE VOLUME 94 (2001)

ISSN 0262 6608

© Wiltshire Archaeological and Natural History Society and authors 2001

Hon. Editors: Joshua Pollard, BA, PhD, and John Chandler BA, PhD. Hon. Local History Editor: James Thomas, BA, PhD, FRHistS.

Hon. Natural History Editor: Michael Darby, PhD, FRES

Hon. Reviews Editor: Michael Marshman, ALA.

Editorial Assistant: Lorna Haycock, BA, Dip.ELH, Cert Ed.

We acknowledge with thanks a publication grant from Berkeley Homes (Hampshire) Ltd., towards the cost of publishing “The Excavation of a Saxon Settlement at Cadley Road, Collingbourne Ducis, Wiltshire’, by Jo Pine.

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’ (this volume) 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.

Typeset in Plantin by John Chandler

and produced for the Society by

Avonset, 11 Kelso Place, Upper Bristol Road, Bath BA1 3AU Printed in Great Britain

i THE NATURAL | HISTORY MUSEUM

15 FEB 2001

PURCHASED GENERAL LIBRARY |

{

Contents

Excavating the Sanctuary: new investigations on Overton Hill, Avebury, by Mike Pitts, with 1 contributions by Joshua Pollard and analyses by Jacqueline I McKinley and Amanda Rouse

Savernake Forest oaks, by Jack Oliver and Joan Davies 24 Murder at Brookside Cottage: a dark deed in North Wiltshire, by Kay Taylor 47 The Ruin of a great Wiltshire Estate: Wardour and the Eighth Lord Arundell, by 56 Barry Williamson

The Excavation of medieval and post-medieval features at 3 Kingsbury Square, Wilton, 68

by Kate Taylor, with contributions by Sheila Hamilton-Dyer and Jane Timby Cotton wadding: Marlborough College and the path to war, by Brian Edwards 75 The Excavation of a Saxon settlement at Cadley Road, Collingbourne Ducis, Wiltshire, 88

by Jo Pine, with contributions by Mathew Gleave, Sheila Hamilton-Dyer, Steve Ford, John Letts, Nicola Powell, David Richards, Chris Salter, Jane Timby and David Williams

The Thirteenth-century stained glass of the Salisbury Cathedral Chapter House, by Sarah Brown 118 The Urbanity of Marlborough: a Wiltshire town in the seventeenth century, by LL. Williams 139 Archaeological investigations at the Roman villa, Netheravon, 1996, by Mick Rawlings 148 The Coleoptera (beetles) of a Salisbury reedbed including twenty-two species new to Wiltshire, by 154 Michael Darby

The Excavation of a cremation cemetery of the Bronze Age and a flint cairn at Easton Down, 161

Allington, Wiltshire, 1983-1995, by David J. Ride

From tiny seeds ..., by Antoinnette Rawlings A James Bennett of Salisbury (1797-1859): jeweller and newspaper proprietor, by Robert Moody 182 The Marlborough Mount revisited, by David Field, Graham Brown and Andrew Crockett 195 An Antiquarian visit to the Rollright Stones, by Frances Peters 205 Sheldon Manor: medieval settlement and land-use in the clayland region of north-west Wiltshire, 209

by Graham Brown

Notes and Shorter Contributions 218

The Glow-worm (Lampyris noctiluca, (Linnaeus, 1758)) in Wiltshire: records and questions, by Michael 218 Darby Neolithic activity in Ducks Meadow, Marlborough, by Emma Harrison, with contributions by Jane Timby 219

and Graeme Walker

The Devizes Millennium Horse, by Mark E. P. Hows and Peter T: Greed 223 The Rev Edwin Meyrick and his family: a brief investigation of a forgotten Wiltshire clergyman, by Tony 226 Pratt

Recently discovered Tegula Mammata from North Wiltshire, by Bob Clarke 228 Fonthill House, by Michael Darby 230 Ludgershall Castle: two addenda: The Ludgershall gem set seal reconsidered, by John Cherry; Iron prick 234

spur, no. 3, by Blanche Ellis

Excavation on Roundway Down, by Sarah Semple and Howard Williams 236 Swanborough Tump, by Sarah Semple and Alex Langlands 239 Excavation and Fieldwork in Wiltshire 1999 243 Reviews 257 Peter Sherlock (ed.). Monumental Inscriptions of Wiltshire: an edition, in facsimile, of Monumental 257

Inscriptions in the County of Wilton by Sir Thomas Phillipps, 1822, by Steven Hobbs Martin Green. A Landscape Revealed: 10,000 Years on a Chalkland Farm, by Philip Aslett. 258

R. H. Thompson and M. J. Dickinson. Sylloge of Coins of the British Isles 49: The Norweb Collection 258 Tokens of the British Isles, 1575 - 1750, VI, Wiltshire to Yorkshire, Ireland to Wales, by Paul Robinson

A Millennium Mixture, by Michael Marshman 260 Obituary 265 John Musty 2605

Index, by Philip Aslett 269

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.

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 Chief Executive.

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. Whilst 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 1NS.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 on disk, Word or Rich

Text Format files are preferred.

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 Téne 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 following that of the Magazine. If possible, all original artwork should not exceed A3 before reduction. 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.

Offprints: Ten offprints of each article will be given free (to be shared between joint authors). Offprints are not given for notes and reviews.

WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY SOCIETY

COUNCIL (to 19 May 2000)

VICE-PRESIDENTS

R.G. Hurn

N. Davey, OBE, PhD, DSc, FSA N.J.M. Anderson

E.K. Annable, BA, MA, FSA H.F. Seymour, BA

Elected Members

D.N. Shelton, BA, BSc., Dip.Ed., FRGS (1993)

Lt. Col. C. Chamberlain (1994) M. Darby, PhD, FRES (1995)

B. Coupe, LDS, RCS, MSc. (Dent.) (1996)

D. Lovibond, BA (1996)

M.J. C. Smith, BA, FICE (1996)

C. A. Shell, MA, MMet, PhD (1997) R. Sneyd (1997)

Mrs G. Swanton, BA, Dip.Ad.Ed. (1997)

D.J. Williams, MA (1997) D. Field (1998) T. Schadla-Hall (1998)

Maj. Gen. G. M. G. Swindells CB (1998)

E. Stanford ARBS (1998)

Mrs P. Sneyd, PhD, BSc., CBiol, MIBiol (1999)

FOUNDATION TRUSTEES R.G. Hurn

H.F. Seymour, BA

DrT.K. Maurice, OBE

J.F. Phillips, BSc.

Ex-officio Members

M. Corney, BA (Hons)

Chairman

Col. D. Part, OBE (Military), TD, DL (London) (1997)

Deputy Chairman P. Taverner, MA

Chairman, Archaeology Committee

D.N. Shelton, BA, BSc, Dip.Ed., FRGS Chairman, Buildings & Monuments Committee C.J. Perraton, MA, CBiol, MIBiol.

D.J. Williams, MA A.G. Lansdown

Nominated Members

Mrs L. Bennett, Mrs P. Rugg

T. Price I. R. P. Hopkins Mrs J. Brunt

Chairman, Natural History Committee

Chairman, Programme Committee Hon. Treasurer

Members, Wiltshire County Council

Member, Devizes Town Council

Member, Kennet District Council Member, Swindon Borough Council

P.R. Saunders, BA, FSA, FMA, FRSA Director, Salisbury & South Wiltshire Museum

BOARD OFTRUSTEES (from 19 May 2000)

Chairman Col. D. C. Part, OBE, TD, DL

Deputy Chairman P. Taverner, MA

Elected Trustees

Lt. Col. C Chamberlain

C J Perraton, MA, CBiol, MIBiol. Mrs D Robertson

D Roseaman

CA Shell, MA, MMet, PhD

A Snow

E Stanford, ARBS

Mrs G. Swanton, BA, Dip.Ad.Ed. W E Verity

D.J. Williams, MA

OFFICERS

Chief Executive

Curator

Deputy Curator

Assistant Curator (Natural Sciences) Sandell Librarian

Education Officer

Nominated Trustees

Mrs L Bennett Mrs P Rugg

A Mills

Mrs J Triggs

P.R. Saunders, BA, FSA, FMA, FRSA

In attendance: Ms A. Cutforth

In attendance: Ms A. Cutforth

County Museums Officer, Wiltshire County Council

Member, Wiltshire County Council Member, Wiltshire County Council

Member, Devizes Town Council

Member, Kennet District Council Director, Salisbury & South Wiltshire Museum

County Museums Officer, Wiltshire County Council

G. Chancellor, BSc, PhD, AMA, MIMegt P.H. Robinson, PhD, FSA, AMA

Mrs A.J. Rawlings, BA (Hons), MA, AMA A.S. Tucker, BSc, AMA

Mrs L. Haycock, BA (Hons), Dip ELH, Cert.Ed.

Ms J. Harvest, Cert.Ed, DipSpLD

Wiltshire Archaeological & Natural History Magazine, vol. 94 (2001), pp. 1-23

Excavating the Sanctuary: New Investigations on Overton Hill, Avebury by Mike Pitts!

with contributions by Joshua Pollard’ and analyses by Jacqueline I McKinley’and Amanda Rouse?

The Sanctuary is an important but poorly understood part of the complex of ceremonial structures in the Avebury World Heritage Site. Located and excavated in 1930 by Maud Cunnington, it consists of seven concentric pit circles, one of which held small megaliths, a second megaliths and posts and the others just posts. The excavation was published promptly and many of the finds survive, but no archive was known until the realisation in 1999 that W. E. V. Young, excavation foreman, had kept a detailed record of the dig in his diary. A small re- excavation was conducted to reconcile apparent discrepancies between the diary and the published report. Although all features had been previously examined, the new dig resulted in both important new finds (a large group of lithics, including nine arrowheads, part of a copper alloy awl and in situ amphibian bones) and new insights into

the nature of the Sanctuary’s construction.

_ Further developments on this study will be posted at www.hengeworld.co.uk

INTRODUCTION

In 1930 Maud and Ben Cunnington located and substantially excavated the former stone circles first identified by John Aubrey in 1648, and named ‘the Sanctuary’ by William Stukeley in 1743. These rings stood on a low but locally prominent chalk brow, Overton Hill or Millfield, at the southern end of the two parallel rows of megaliths known as the West Kennet Avenue, which connects the Sanctuary to the stone circles at Avebury (SU 118680, Figure 1; Cunnington, M. E. 1931; Piggott 1985; Ucko et al 1991).

The excavation found not only the two stone circles, but also six concentric rings of post holes (one in the same circuit as the inner stone circle). The post holes and the artefacts they contained allied the site to Woodhenge, discovered in 1925 and excavated by the same team in 1926 and 1927, and through

Woodhenge to nearby Stonehenge (Cunnington, M. E. 1929). From the moment the Sanctuary was uncovered, archaeologists have debated what stood in the post holes. The focus of this discussion has been on whether or not the posts supported a roof, and to what extent different rings of posts (and stones) were standing at the same time.

In 1999, at the suggestion of Ros Cleal, Curator of the Alexander Keiller Museum in Avebury, I consulted manuscript diaries written by William Young and now held in the library of the Wiltshire Archaeological & Natural History Society. Although not credited in the excavation report, Young, a skilled experienced digger, was foreman at the Sanctuary. His extensive diaries begin in 1930, a few months before the dig at the Sanctuary (and thus after Woodhenge, where he also worked). They contain an extremely detailed record of the excavation that has never been referred to before. As well as information on aspects not covered in the published report, the

1 125 High St, Marlborough, Wilts SN8 1LU; 2 Dept. of Humanities and Science (Archaeology), University of Wales College Newport, PO Box 179, Newport NP6 1YG; 3 Wessex Archaeology, Old Sarum Park, Salisbury SP4 6EB; 4 Cardiff School of Engineering, Cardiff University, PO Box 917, Cardiff CF2 1XH

bo

Windmill Hill

ca

Silbury Hill

West Kennet —

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

West Kennet Enclosures XN

Long Barrow

Figure 1. The Sanctuary is located at the southern end of the West Kennet Avenue. Large dots are barrows, small ones natural

sarsens (based on Smith 1965)

diaries reveal several inconsistencies. Despite careful analysis, it proved impossible to match some of the post holes in the diaries with post holes in Cunnington’s report (different numbering schemes are used). In particular, Young and Cunnington appeared to be describing a different number of rings near the centre of the site.

It was decided that only excavation would resolve these conflicts and thus allow full use to be made of both sources. Accordingly, in view of the great academic and public interest in the Sanctuary, permission was sought for a small excavation within the area excavated and described by both Cunnington and Young. This was granted by the Department for

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 3

Culture, Media and Sport (the site is now a Guardianship monument within the Avebury World Heritage Site) in July 1999, and excavation took place between 23 August and 14 September. The site was restored to its former state of cylindrical concrete post- markers rising from grass turf.

The aim of this article is to describe the new excavation, to summarise the previously unseen record of Young’s diaries, and essay a reconciliation of all sources. Lengthy tables combining all available information on excavated features can be found at www.hengeworld.co.uk; printed copies have been given to the museums at Devizes and Avebury. Coincidentally, in 1999 I found the skeleton of the adolescent buried against one of the megaliths, excavated in 1930 and believed destroyed in 1941, at the Natural History Museum. This is also described.

Surviving artefacts from the 1930 dig are in the museum of the Wiltshire Archaeological & Natural History Society in Devizes. There appears to be no archive, apart from Young’s diaries and a few associated letters. Retained animal remains and the

human skeleton are at the Natural History Museum. Antler, which was not sent to London in 1930, appears to have been discarded. There are some draft diary pages, previously unidentified, and some negatives of Young’s photos in Avebury museum, where all finds and records from the 1999 dig are held.

THE 1930 EXCAVATION

On 26 April 1930 Maud Cunnington wrote to Young to say that the owner and farmer of the land she and her husband had identified as the likely site of the Sanctuary (Jack Osmond), had granted them permission to excavate. She did not know how much work there would be — some stone holes, perhaps a ditch, perhaps nothing — but she hoped he would be able to help. On 21 May, Young received a card from Cunnington saying they had not yet found anything (they began on Tuesday 20 May; the field was planted with young sugar beet).On the Friday, the long 2m

Figure 2. The Sanctuary stone and post ring (C) in 1930, with Robert Cunnington, holding notebook, standing in an excavated stone pit. Photo W. E. V. Young, © Alexander Keiller Museum

4 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

wide trench exposed two stone pits, and Cunnington wired Young in the evening asking him to start work Monday morning ‘for a week or a fortnight’.

Young’s diaries (about which more will be said later) are our best source on how the project was managed. Daily entries describe features he excavated himself, but other pits receive scant mention. When he arrived to dig, several holes of the outer stone circle and of the post ring B (the ‘Fence ring’) had already been emptied. He left on 16 June, to meet a prior digging job at Ham Hill, Somerset, when one pit of ring F (the ‘7 Foot ring’) and much of ring G (the ‘6 Foot ring’) were still unexcavated. His photos of the dig (the same six prints are mounted in the two main diaries) are the only ones known (Figures 2, 3).

The ‘excavation party’ consisted of Ben, Maud and Colonel Robert Cunnington (an army-trained surveyor), Maud’s brother (E. L. Pegge), Mr C. W. Pugh and four or five paid men, including Young, and E. Hambridge and F. White, who, like Young, had worked at Windmill Hill in 1928 and 1929. Even by today’s standards this was an experienced team.

The diaries record a few archaeological visitors: J. P. Williams-Freeman (president Hampshire Field Club), R. C. C. Clay, Harold St. George Gray (curator Taunton Museum), Robert Newall, W. J. Hemp (Secretary Royal Commission on Ancient Monuments for Wales and Monmouthshire), Charles Drew (Curator Dorchester County Museum) and George Engleheart. The Wiltshire Archaeological & Natural History Society visited in force on the day of the discovery of the burial beside stone C12 (3 June). Hambridge had found some human skull fragments, but Maud Cunnington, recorded her nephew Robert, disliked ‘publicity’, and withheld the news from all but a ‘single lingerer’ — the sort of detail not present in Young’s journals (Cunnington, R. H. 1954). It is possible that notes or photos from some of these visits may yet be found.

Cunnington chose the dig as the subject of her talk at the A. G. M. of the Wiltshire Archaeological & Natural History Society in Trowbridge (Times 1 August; Nature 2 August 1930). She described the six rings of post holes, and suggested these were succeeded by the two stone circles when the Avenue and Avebury rings were built. The Cunningtons had bought the site, and planned to mark pits by concrete pillars, as they had done at Woodhenge. Her full report was printed the next year, with contributions from J. W. Jackson (animal remains), Arthur Keith (the human skeleton) and A. S. Kennard and B. B. Woodward (non-marine molluscs), followed by a note based on T. Woodhead’s examination of the charcoals

(Cunnington M. E. 1931; Cunnington and Woodhead 1931). Of the subsequent studies of this site, Smith (1965, 244-7) and Pollard (1992) looked at the finds in Devizes museum, but others refer only to the original published report.

THE ADOLESCENT BURIAL BESIDE STONE C12

Maud Cunnington wrote to Arthur Keith at the Royal College of Surgeons, London on 18 August 1930, asking if he would look at the skeleton they had found buried beside stone hole C12. His analysis was written on 14 October and incorporated verbatim in the excavation report (mss. in R. C. S. library; Cunnington 1931, 330).

Cunnington delivered several human remains from her excavations to the college, where she was held in some respect. M. L. Tildesley, Curator of the

Figure 3. The adolescent burial beside stone hole C12, from a photo published as a postcard by W. E. V. Young (given to Pitts by I. Smith)

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 5

Stone Hore

Seace “in: To ter.

Figure 4. Sketch of burial pit beside stone hole C12 from Young’s Diary, 4 June 1930. The black circle below the knees is the Beaker pot (not to original scale)

Department of Human Osteology, wrote of her that her ‘archaeological accuracy and distinction need no words of praise from me’ (Annual Report of the Museum, R. C. S. 1932, 28). The skeleton was accessioned into the museum’s collection.

The college was severely damaged in a bombing raid in 1941. It seems to have been assumed by archaeologists at the time that all their specimens were destroyed, a belief that became part of archaeological folklore. Many skeletons did survive, however, and unknown to archaeologists were removed to the country for the duration of the war, and ultimately transferred to the Natural History Museum. The Sanctuary adolescent’s post cranial remains arrived in 1951, and the head in 1955. They were listed as coming from a long barrow on Overton Hill, accession PA SK 48 (R. C. S. 4.0372). There is no doubt, however, that these bones are those from the Sanctuary. The background to this and other skeletal discoveries is described elsewhere (Pitts 1999; 2000b).

Keith said the skeleton was a male of about 14 years. Burl suggested it was a girl, because the head was pointing south. ‘It was a Beaker custom in Wessex to bury women with their heads to the south, men

with their heads to the north’ (Burl 1979, 198). While there may be a tendency for bodies to be laid out this way, it is not universal. McKinley, who kindly examined the skeleton at the Natural History Museum, is not prepared to identify the sex. The mandible is masculine, but the sciatic notch is wider than 90°, suggesting a female hip. The individual was still young, and McKinley feels that attribution of sex should be left open (Pitts 2000b, 130-1).

Young describes ‘slight traces of charring on upper side of larger limb bones’ and McKinley confirms this impression but feels that it is dark coloured staining, not burning. Cunnington did not comment.

THE 1999 EXCAVATION

The main aim of the excavation was to expose post holes described in both Cunnington’s published report and in Young’s diaries, so that a reconciliation between the two might be achieved. There were other subsidiary goals:

1. It seemed likely that objects that would now be retained had been left in the ground; careful excavation, including the use of sieves where appropriate, would assess this factor.

2. It was thought possible that in situ fill could still exist in some postholes: the 1930 records make it clear that distinguishing artificial fill was often difficult. If found, there was the chance that such fill could contain material suitable for radiocarbon dating. When curator of Avebury museum, I had looked in vain at existing collections for appropriate material (Pitts and Whittle 1992; see Radiocarbon Dates below).

3. Excavation would allow study of the spatial relationship between the concrete markers and the actual post holes.

A 5 x 5m square was sited to take in post holes E4 (thought before excavation to be the one described by Young 1n particular detail) and D5 (Figures 5, 6, 12). The latter is the only ‘single’ post hole in the ring of ‘double’ holes of the D or ‘Bank Holiday ring’. It is apparent from Young’s diary that this was the first pit in this ring to be excavated. It is possible that it too was a double pit, but not recognised at the time. Excavation conditions were difficult, and the team had no experience of double pits (there was none at Woodhenge). They worked by backfilling as they went, so may not have been able to re-examine the pit once it was realised that others were oval and not round. Thus there was a chance that a complete post pipe and associated remains lay undisturbed in this pit.

6 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

ie

50m

Figure 5. The Sanctuary enclosure and concrete markers surveyed by Dave Field for National Monuments Record/English Heritage in 1999. Dashes are rectangular blocks (stone holes), crosses circular (post holes). The square is the 1999 excavation trench; the small shape 1n the north east corner a recumbent sarsen

In the event, we found excavation as difficult as our predecessors, and sieving and the quantity of artefacts made progress slower than anticipated. It was decided to leave D6, and what soon became clear may be a uniquely valuable feature at D5, for future study.

(Fence ring (B) may be worth re-investigation for a similar reason. 32 pits are described as 30cm diameter (Cunnington) or 60cm (Young), only two (B33 and B34) being larger (Cunnington: 60cm diameter; Young: 90cm) and with post cores. Much of this ring had been excavated before Young’s arrival on site on May 27. But B33 and B34 were completed on June 6. Young comments (Diary 6/6/30) of B33 that ‘previously the core only of this hole had been taken out, and the “packing” left in situ’.)

Pollard assisted in direction and with provision of most of the tools and volunteer labour, which came from the University of Wales, Newport, assembled at Beckhampton for the larger concurrent project

(Gillings et a1 1999). Jennifer Garofalini assisted with the excavation while she was preparing her thesis on the use of VRML in archaeology (Garofalini 2000).

The modern soil was excavated in three spits over the whole trench to a depth of about 50cm, on to a firm loam or broken chalk surface assumed to be the base of the 1930 excavation. In the north west corner of the trench, chalk was reached at a slightly greater depth, matching Young’s description of how they deepened a 3 x 3m area in the centre in an unsuccessful search for small holes (Young 1930a, 12 June). No previously unrecognised pits or stake holes were identified (Figure 7).

Soil was sieved through 5mm mesh. Several modern coins were found, particularly around concrete marker F3, and a collection of pieces of quartz and polished stone had been buried near marker F4. Most of the coins are from 1995 or more recently (10), all but two of the nine others dating from the previous ten years. Apart from one coin, these

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY i

THE SANCTUARY 1930/1999

© Fencering ©

Stone-and-post ring

ae ©

O ae Ce "ears Es e

CD 10-foot ring () CQ 2

2 O Q O 7-foot ring O pee ole O : O @ O : \\ © € O © O © oO O (@) O 5 Oo 2) re) Oo fe) O re) © O OB O eo) 0 5 10 metres MEC/MWP.

(SE eee ee eee eee

Figure 6. The six inner rings of the Sanctuary, with the 1999 trench superimposed. Within the trench, the pits as excavated are drawn in thick lines, as plotted by Cunnington in dashed lines. The post positions are taken from Cunnington’s Plate I (where they are shown ‘as measured at the bottom’), in which the only stone pit marked is H3. Both post and stone pits appear in Cunnington’s smaller scale Plate I (with no reference to how the pits were planned). There are significant differences in the detail of post pit shapes in Cunnington’s two plans (compare particularly ring E: in Plate I they are all drawn as ‘double’ pits, but in Plate II several are small and circular). The registration of colours in Plate I (red stone pits, black post holes) in copies available to me is not perfect, so it seems unwise to place too much weight on the precise shapes and positions of stone pits in Plate I. They have been drawn schematically here. The trench was located by scaling it to Cunnington’s plan, and finding the best visual fit between the eight post holes we both excavated.

8 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

finds were in the top spit, and clearly constitute offerings by modern visitors. The range of coins is interesting: as well as low denomination change of UK currency, there were two Belgian coins, one Spanish and one Russian.

The only coin found at greater depth was a US 1 cent piece, dated 1930 (near marker F5). Perhaps this was buried by someone on the 1930 team, in the same way that the Cunningtons’ ancestor William Cunnington left initialled and dated lead plaques in his excavations. Colt Hoare had suggested the plaques in 1804 as a cheaper alternative to ‘Mr Bolton’s new copper coinage’(Cunnington, R. H. 1975, 57, 63).

Eight post pits were totally excavated. Significant quantities of prehistoric artefacts were recovered, including over 1000 of flint (of which 70% were in the soil) and a handful of small prehistoric sherds (all in soil). Also found were 301 pieces of sarsen (87% in soil), 41 bone (75%), 78 burnt flint (73%) a Roman sherd and a fragment of prehistoric copper alloy awl. The flints (described below by Pollard) include nine arrowheads (Cunnington reported three from the entire 1930 dig) and a further 68 retouched pieces.

THE SANCTUARY

i,

As Pollard indicates, the assemblage is of considerable significance, not just locally, but as a rare example of systematically recovered material from a late Neolithic timber circle. Pollard also describes the sherds and the awl. Tabulated provenance data for all finds are in the archive.

The locations of the post holes match well with Cunnington’s published survey, albeit this is sketchily drawn (see Figure 6). The concrete markers, however, can only be taken as approximate representations of post pipes (Figure 7). We do not know precisely where these pipes were in relation to pit sides, but it is apparent that G2 cannot be on the site of the pipe, and G3, F4 and F5 are probably not. Perhaps an error of + 20cm between the centres of post pipes and concrete pillars should be allowed. F3 is a special case. It had been broken off and recemented twice, each time moving to the west until it was about 50cm from its post hole.

Comparison of ring diameters as mapped by the N. M. R. (concrete posts: Figure 5) and Cunnington (post holes) suggests typical divergences of no more than 30cm. While this is enough to discourage the

Lil C y; | < ar Q E3 oy es £3 y I ae 3@ ( @ ! ap ald! ° \ ‘ @ \ | i4 | ae : re [a : 4 ~ (aN Ss EO | ‘A AS ane ( es =. ! = ke a : \ (e:- | We > . aS ~ eo) ee : 4 x % 7 3 (yee ve te ey | @ She PELL Ey ae | “@ Ea a aN QA “ Shaw \ A A G vs | J NY AS a + \\ \ | \ eS \?> ean \ NY | s ig = = x \ ES in WW?) ls a \ <) | | Sas aN ‘\ y | t — Ne \4 Woo ) ‘ a Z 3 : -_—— + 3 metres MWwP.

Figure 7. Left: Plan of fully explored part of 1999 trench showing features and lines of profiles in Figure 9. Right: 1930

concrete markers

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY

a 2 metr THE SANCTUARY 0 eves

010 TPN

62 63 G3 Se ee a a Sn ae in aa ees ieee (eT SE N section line eo FS E5 €3

SE

SS

£4 in plan

chalk ndge at level B

pit edge atlevelA 7 re

0 05s 1 metre MWP

Figure 8. Top row: profiles of pits in D and G rings after Cunnington (1931). Below: pits excavated in 1999

10 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

search for ancient high precision units of measurement, it is not perhaps as bad as might have been expected from comments made at the time by Keiller. He wrote to A. D. Passmore 21 October 1936 of R. H. Cunnington’s equipment: ‘one 30 foot tape —muuch stretched, and never hitherto corrected since purchased a good many years ago — a broomstick “notched to feet”; and two slips of wood nailed to the top of the aforementioned broomstick to form some sort of sight vane. To this, of course, must be added Mrs Cunnington’s umbrella’ (Alexander Keiller Museum ref 78510174). Keiller’s love of his own surveying equipment had as much to do with technology as it did with relevant precision.

The pits were excavated by layer (1930 backfill) without sectioning, except for E3 which lay under the trench edge. This profile showed the pit to have been refilled from the east (Figure 8).

All post holes were wider near the base than higher up, as described by both Cunnington (M. E., 1931) and Young (1930). The profiles were quite irregular (Figure 8). It was not possible to say if any of this undercutting was caused by natural collapse. The major difference between the holes exposed in the 1999 excavation, and the impression given in the 1930 report was the way in which some of the pits clearly cut through each other. The original plans, which show post holes ‘measured at the bottom’, do not reveal this, and Cunnington makes no comment in the text. This important new observation will be considered further below. The practice of using the bases to map holes also had the effect of visually under-estimating pit size.

The three ‘double’ post holes (E3-5) all had indications of having been dug slightly deeper at the outer, eastern end. In E3 and E5, the deepening was bordered by a slight curving step on the bottom (Figures 7-8). The evidence in E4 was more dramatic. When first re-excavated, this pit had a substantial step of firm, clean chalk on the western, inner side that had every appearance of being undisturbed rock (Figure 13). It was left to weather, and when carefully explored around the western edge of the pit the step was revealed to be hard packed chalk rubble and powder (034). This was removed down to another very firm surface. This also was later revealed to be the top of clean chalk fill, mixed with pale orange silt (035). The final pit base in convincing chalk rock clearly demonstrated the effect of two roughly circular pits being dug beside and partly intercepting each other (Figure 8).

When it was realised that substantial in situ fill had been left in E4, the other pits were carefully re- examined. A small amount of chalk fill was found in the bottom of F5 (Figure 8) and F3. No objects were

found in any of this chalk. During excavation, a few small pockets of apparently original fill were found in crevices in the edges of pits E3, E4 and G2, some of which contained amphibian bones (see below).

LITHICS by Joshua Pollard

A total of 1058 pieces of identifiable worked flint and 60 pieces of burnt flint were recovered during the 1999 excavation (Table 1). The bulk of the material derived from surface contexts [001], [002] and [003], although sizeable collections of flint also came from some of the feature fills (eg E4 and E5). None of this material was found within its original context; though the Cunnington practice of returning spoil to the feature from which it was dug, or an adjacent one, means that some spatial integrity in regard to the original distribution of artefacts might be retained. Mixing of material from topsoil and feature fill contexts has, however, taken place. It is evident from the variable condition of the material from surface contexts, which ranges from rolled to very fresh, that the flint here includes pieces from both post hole fills and the pre-excavation ploughsoil. Because of the derived context from which the material came, fully contextual analysis was considered inappropriate. Contextual difficulties aside, the assemblage is still of enormous significance, not just in regional terms, but as one of only a handful comprising systematically recovered material from a late Neolithic timber circle.

Raw Materials

The thin but fresh cortex on many pieces indicates a local chalk source for the flint. It is of variable quality, a number of pieces incorporating thermal fractures and internal flaws. Additionally, there are two flakes with a thin, worn and stained cortex that most likely derive from gravel deposits. There is no conclusive evidence of high quality mined flint being utilised, though it is not impossible that this was employed in the manufacture of some of the more elaborate artefacts such as arrowheads.

Composition, Technology and Implement Forms (Figure 9)

Several debitage forms and implement types are present. Given the relatively low percentage (8.1%)

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 11

Table 1: Worked flint from the 1999 excavations

Context Preparation Rejuvenation Unretouched Chips flakes flakes flakes [001] 16 13 65 36 [002] 42 27. 348 Wi [003] Z 1 2 [009] L 3 1 [037] 1 = 8 2 D7 1 1 3 1 ES - 1 8 2 E4 9 8 54 4 E5 9 12 87 11 E3/F3 1 2 8 1 F4 1 - 4 1 G2 ~ 2 11 1 G3 6 3 27 3 Total 86 69 625 173

Cores Flaked pieces Shatter frags. Retouched Total 2 - - 11 143 4 4 12 27 D0) - - 1 4

- - - 1 2

- ~ 3 12 - - - - 6

- - - 2 13 = 2 1 8 86 3 ) 3 9 137 1 - - 3 16 - - - - 6

1 - 2 7, - - 1 1 41 11 9 17 68 1058

of preparation flakes (defined as having 60% or more cortex on their dorsal surfaces), many of which are small anyway, it is probable that nodules were brought on to site in a partially prepared or tested state. Otherwise, all stages of core reduction are present. Chips (taken as pieces less than 15mm in maximum dimension) make up 16.4% of the total; a reasonably high figure that must in part reflect the extent of sieving. Amongst the material from surface levels [001] and [002], which was more systematically sieved than the feature fills, the percentage rises to 20.5. Their presence is important, insomuch as it indicates in situ knapping (Newcomer and Karlin 1987). Implements comprise 6.4% of the total, a not unduly high figure.

The majority of flakes are broad, hard-hammer struck and without prepared platforms. Rejuvenation takes the form of trimming areas of step-fracturing, or removing ridged/keeled surfaces. Most of the cores are systematically worked down and consist exclusively of single or multi-platform examples from which flakes have been struck (there are no blade cores). In this respect the bulk of the debitage is typical of late Neolithic industries in the region, and across southern England (Smith 1965, Holgate 1988). Flakes with facetted platforms are present, though in very small numbers (in the order of 1% of the flake assemblage). They do however hint at another mode of working that was taking place during the later Neolithic. Levallois-style working was used in the production of pre-formed blanks for chisel and oblique arrowheads; and similar flakes were also used to manufacture a knife from E3/E4 and a denticulate

from G3. This involved working discoidal cores with carefully prepared platforms, the flakes from which are broad (tending to oval), thin and exhibit dorsal flake scars intersecting from several directions. Apart from these few implements there are few signs of this kind of working taking place on the site. There is one unmodified Levallois flake from [002], but no characteristic cores. Chisel arrowheads and the few other implements worked on Levallois flakes were most likely being brought on to the site as finished implements.

Additionally, there are a small number of soft- hammer struck blades/bladelets, mostly from surface contexts [001] and [002]. These make up around 5% or just under of the debitage assemblage. A percentage at least of these are likely to be of 4th millennium BC date, and relate to one or more short-lived episodes of pre-monument activity. Other diagnostic 4th millennium BC forms such as blade cores, leaf arrowheads and fragments of polished axe, are absent, though the narrow flakes with fine, regular “edge- trimming’ (Smith’s ‘Class A’ utilised flakes: 1965, 92) are likely to be earlier Neolithic. Blade/narrow flake production need not, however, be confined to the 4th millennium BC. Soft hammer struck narrow flakes characterised the broadly contemporary assemblage from the West Kennet palisaded enclosures (Whittle 1997, 92), and were utilised in the production of microdenticulates at the Sanctuary.

The range of recovered implement forms is quite restricted (Table 2). Scrapers dominate, followed in numerical frequency by miscellaneous retouched

12 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

| a °

5 ee eee CM

Figure 9. Worked flint and copper alloy from the 1999 excavations. (1-6) chisel arrowheads; (7) oblique arrowhead; (8) triangular arrowhead; (9) barbed-and-tanged arrowhead; (10) denticulated flake; (11) microdenticulate; (12) edge-trimmed flake; (13) knife; (i4) notched flake; (15-16) scrapers; (17) copper alloy awl (drawn by J. Pollard)

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 13

Table 2: Retouched and utilized pieces

Context Scrapers Arrowheads

[001] 6 2 [002] 6 4 1 3 [003] 1 : [009] 1

[037] 1

D7 f 2 E3 1 1 E4 3 : 2 E5 2 1 E3/F3 : 8 Z F4 z E x s G2 c 1 = - G3 5 = 2 =

Total PDs 9 6 6

Notched Knives

Micro- Den- Edge- Misc. Total

denticulates ticulates trimmed retouched

- - - D, LT

2 1 4 6 Bia - - - 1

- - - - 1

- - - 2 5)

- - - - 2,

1 - - 1 8

- - - 3 9

1 - 1 - 8}

- - - B

- 1 - - 1

4 2 by 15 68

flakes, arrowheads, knives, and modified flakes (notched, edge-trimmed, denticulated and so forth). This might be characterised as a ‘light’ tool kit, with heavy-duty objects such as axes being conspicuous by their absence. In itself this is interesting given the scale of wood-working that most probably took place on the site. The proportion of arrowheads is unusually high. The scraper to arrowhead ratio is 2.3:1, compared with 5:1 from the later Neolithic West Kennet Avenue occupation site (Smith 1965), and 9:1 from the West Kennet enclosures (Whittle 1997). The situation is, however, reflected at other timber circles in Wessex, such as DurringtonWalls (Wainwright and Longworth 1971) and Woodhenge (Cunnington 1929), where oblique forms are far more prevalent.

Of the nine arrowheads, six can be classified as chisel, one as oblique, one as barbed-and-tanged, and one as triangular. This replicates quite nicely the range of arrowhead types discovered during the Cunnington excavations. Three of the chisel arrowheads are intact, as 1s the triangular example, though hinge-terminating bending fractures on the tip and tang of the barbed- and-tanged suggest breakage through impact. With its square barbs this is probably of Green’s Conygar type (Green 1980). Worked through extensive bifacial retouch, the triangular point is more problematic. Whilst it may be an unfinished barbed-and-tanged, similar examples are known in some numbers from the upper fills of OD II on Windmill Hill, in association with Early Bronze Age lithics, and from the West Kennet Avenue occupation site (Smith 1965). The

form is unusual and concentrations of these are known from Wessex, the Breckland and Yorkshire Wolds (Green 1980).

Characteristic Beaker/Early Bronze Age implements are significantly well represented. In addition to the barbed-and-tanged arrowhead from E5, there is a knife with semi-invasive retouch from {002], and scrapers with similar working from [002], ([037], E4 and E5. Classic ‘thumbnail’ scrapers are, however, absent. The scrapers are quite heterogeneous, and include several examples with styles of working more typical of 2nd rather than 3rd millennium BC industries. In addition to ‘typical’ side and end scrapers, some quite worn, there are straight- end (3 examples), denticulate (2) and composite (1) forms. The latter include examples on thick and irregular flakes and thermally fractured pieces. Amongst the straight-end scrapers there is one with inverse retouch. Additionally, there is one concave scraper from [003].

The Lithic Assemblage from the 1930 Excavations

The recovery of 1058 pieces of worked flint from the very limited re-excavation of the monument stands in sharp contrast to the 79 pieces of lithic material that survive from the 1930 excavations (Table 3). (This material is currently held in Devizes Museum, and further details are given in the original report and in Pollard 1992.)

14 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Table 3: The lithic assemblage from the 1930 excavations

Flakes Cores Scrapers Arrowheads Misc. retouched Knife Sickle Total Surface - - 13 2) 1 1 1 18 Post and stone holes 44 1 9 4 3 - - 61 Total 44 1 22 6 4 1 1 79

It is clear that the recovery strategy adopted during the original excavation was selective, particularly with regard to the retention of debitage. Of the 44 flakes from the feature fills, 41 of these are from a knapping cluster around C15 (Cunnington 1931, 321). The excavators were a little more systematic in the recovery of implements, though they were either failing to recognise miscellaneous flake tools such as denticulates and edge-trimmed flakes, or were discarding them in favour of more elaborate retouched forms such as scrapers and arrowheads.

Comments

The 1930 assemblage is misleading in scale, and to some extent in representation. The 1999 re-excavation has therefore facilitated a more realistic characterisation of the Sanctuary lithic assemblage. Perhaps the most striking feature 1s the sheer quantity of lithic material from the site, especially given its recovery from such a limited area. The total lithic assemblage from the monument must be in the region of several thousand pieces.’To put this into perspective, the total from this one small area is almost identical to the assemblage from the more extensive work on the West Kennet palisaded enclosures (Whittle 1997, 90-93), and exceeds in terms of density per square metre concentrations within the lithic scatter on the southern slopes of Windmill Hill (Whittle et al. 2000). Of course such high densities need not be uniform across the monument, and if the spatial distribution of material from the Cunnington excavations can be taken as representative, the 1999 excavations were located in a part of the site particularly rich in artefactual material (Pollard 1992).

The assemblage includes diagnostic pieces of 4th to early 2nd millennium BC date, spanning pre- monument episodes of activity and the construction and use of the post and stone circles. However, the bulk of the material can be assigned to the later Neolithic and earlier Bronze Age, and therefore relates specifically to the monument. In situ working may account for at least some of the lithic material,

though other objects were perhaps being brought into the monument specifically for deposition (such as the chisel arrowheads). There may be complex processes of production, movement and discard taking place within the structure, and it could be misleading to read the range of material at face- value, as a record of functional activities focussed upon the site. The working, transformation and (quite formalised) deposition of lithic materials are attested, and within such a ‘charged’ arena as the monument may have taken on a greater significance than was normally afforded within the context of routine production and use. If we think of timber circles like the Sanctuary as places where the categorisation of people, things and relations was reaffirmed, lithic tools and the techniques of working stone could have been openly employed here as tokens of social roles and responsibilities (Edmonds 1995).

POTTERY by Joshua Pollard

There are 14 fragments of prehistoric pottery, in grog-, sand- and flint-tempered fabrics, both reduced and oxidised. All the pieces are small, most being crumbs, the absence of more or larger fragments indicating fairly systematic recovery of pottery by the original excavators. There is one feature sherd, a simple bevelled rim in a flint-tempered fabric from [002]; most probably Ebbsfleet ware. This has faint traces of diagonal cord? impressions on the bevel and a diagonal whipped cord ‘maggot’ on the interior? surface.

COPPER ALLOY AWL by Joshua Pollard

From E4 is a damaged and badly corroded copper alloy awl, 20mm in surviving length and 4mm in

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 15

maximum thickness (Figure 9.17). The surviving end is pointed and squared in cross-section. The broken end has an oval cross-section and perhaps terminated in a squared end. Similar Early Bronze Age awls are common grave finds, though are also known from monumental contexts at Windmill Hill (Hamilton 1999) and Stonehenge (Lawson 1995).

ANIMAL REMAINS

analysis by Amanda Rouse

Amanda Rouse very kindly examined the small bones found in pockets of compact chalky fill, thought to be Neolithic, about half way down the sides of pits E3, E4 and G2. The largest group, context 031, was from E3. About a dozen small bones or fragments are described by Rouse as ‘definitely a toad. The bones could well be from the same individual, on the basis of size and there being no repetition of a particular element’. Bones include a humerus, 9th vertebra, scapula, maxilla, frontoparietal, tibia-fibula and radio-ulna. About ten small pieces that ‘look amphibian’ were found together elsewhere in E3. ‘The only bones that were not too fragmented to be identifiable are certainly toad (femur and humerus). They are from a smaller animal than that represented by 031’. Also from E3 were individual finds of a frog humerus and a toad tibia-fibula or frog/toad radio-ulna.

A single toad tibia-fibula was found in G2, and a possible frog (not toad) femur in E4.

Several toad remains were found in post holes in 1930. Wilfrid Jackson identified the ‘remains of at least eight individuals ... represented by numerous bones’. Cunnington wrote that ‘quite a number of these bones were found in a very fragmentary state, only a selection of these being sent to Dr. Jackson’ (Cunnington, M. E. 1931, 331). Neither writer says which holes contained these toad bones.

In his Diary, Young refers to ‘frog bones’ from pits C19, C23, C29 and E3. Where context detail is given, these bones were found in the chalk, not the _ post pipe (all pits had ‘distinct cores’): ‘in rubble filling’ (C23), ‘in rammed filling’ (C29) and ‘in packing’ (E3) (Young 1930a, 28-30/5/30; 13/6/30). The bones found in 1999 were also in chalk. This would seem to deny Cunnington’s suggestion that the presence of toad bones ‘may be due to the animals creeping into the decaying trunks for shelter’ (1930, 331).

RADIOCARBON DATES

Whether the animals were deliberately buried when posts were erected, or whether they tunnelled into the freshly dug or even partially filled pits, the amphibians represented by articulated skeletons would seem to be ideal candidates for radiocarbon dating. Mammal bones do survive from the 1930 dig (albeit mostly dipped in organic preservative), but there are no articulations: the wide date range of artefacts from the site means that scattered bones need not be contemporary with the timber or stone rings. Unfortunately, however, in 1999 not enough toad bones were recovered for dating purposes. Further excavation may eventually allow such a date to be obtained.

At present, then, the only chance for absolute dating of the Sanctuary lies in the newly re-located human skeleton. At the time of writing, samples are being prepared for analysis by the Oxford Radiocarbon Accelerator Unit. Results should be available to be reported in Notes of the next volume of Wiltshire Archaeological and Natural History Magazine.

RECONCILING THE ARCHIVES

In 1930, Young’s Sanctuary journal was nearly not written, and 25 years later most of the diaries, an extraordinarily valuable resource on Wessex archaeology between the wars, were nearly lost. Towards the end of his life, when the diaries’ future was under discussion, Young wrote a statement about them and his relationship with Keiller and the Ministry of Works (document in Devizes Museum, probably typed by D. G. King). He refers to a letter to him of 19 November 1955. ‘Did you know’, wrote Stuart Piggott, ‘that your Diaries got put into the Sotheby’s sale by mistake? I wrote off in haste and got them withdrawn, but A. K[eiller] was then already on his deathbed’.

At least three sets were created more or less concurrently. Two of these, Diary with Archaeological Notes (Young 1930a) and Leaves from my Journal (Young 1930b) are now in Devizes museum library. A third series, which I refer to here as the Draft Diary (Young 1930c), is represented by a handful of loose pages in Avebury museum. A variety of internal evidences suggest that Draft, written in pencil with

16 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

numerous crossings out, was the original on which the formal Diary was based, and this in turn was shortened to produce Leaves. Nonetheless, all versions contain unique details, occasionally contradictory. As well as the sequence proposed above, it may be that all three drew on the same daily notes that were only later turned into a narrative journal.

Murray (1999, 58-9) suggests that the diaries were kept at Alexander Keiller’s instigation, but I have not found any definite evidence for this (contra Pitts 2000a). In 1930 Young lent Keiller ‘the second copy of this years diary’ (Diary 17.12.1930: perhaps a reference to a volume of Leaves?), but a year later he ‘sent him my original diary for the period April 23rd -July 10th 1930, for him to keep’ (Diary 13.10.1931: these dates do not match any known diary). By then Keiller was posting Young monthly ‘allowance’ cheques to help him through the winter, and it may be that some arrangement had been made over the diaries, which in the earlier years are substantially archaeological in content. In 1936, when Young began full-time work with Keiller, Keiller had the Diary series case-bound and boxed for retention by the Morven Institute of Archaeological Research in Avebury.

Keiller was frustrated at being beaten to the Sanctuary by the Cunningtons, and put considerable pressure on Young to spy for him on proceedings. The detail of the Sanctuary journal may be the result of this. Both Diary and Leaves appear to begin in 1930, and the surviving pages of Draft also cover part of the Sanctuary excavation (26 May to 16 June 1930). The formality of the style, and the tardiness with which Keiller received copies (eg on 1.11.1933, Young wrote to Keiller enclosing his diary for 1933 up to June 30- ‘the other one will follow in due course’) suggest that they were written up some time after the events described, albeit with detailed notes (which would constitute a fourth ‘diary’). This is important in considering his description of the Sanctuary excavation.

In spring 1930, Young was planning his season of archaeological fieldwork. He derived great pleasure from archaeology, but he also relied on the work for a significant part of his income. H. St. George Gray wrote to him on 17 April, inviting him to Ham Hill, Somerset, in June. On 26 April, Maud Cunnington told him she had permission to dig the Sanctuary, but did not know when they would start or how much work there would be. Keiller had been planning to employ him at Meon Hill, Hants, but cabled him on 7 May to say that the field was under thick crops and could not be dug (it had been discovered by O. G. S. Crawford from the air, and was excavated in 1932-3,

with Young as foreman, by Dorothy Liddell). Young then wrote to Cunnington, pleading for information on whether or not there would be work for him at the Sanctuary, but received another non-committal reply. As we have seen, the dig eventually began on 20 May, Young was wired to join on the 23 and started work on the 26. It seems probable that had Meon Hill not been under a ‘thick crop’, he would instead have worked there with Keiller, and his record of the Sanctuary would not exist.

The diaries contain considerably more detail about features excavated at the Sanctuary than hitherto available (to single out one item, he refers to a previously unknown ‘small round jet bead with very small perforation’ from C27).There can be little doubt that at least most of this information was recorded at the time of excavation. The relevant parts of Diary (Young 1930a) appear to have been written out before, and of Leaves (Young 1930b) after the publication of Cunnington’s report (M. E., 1931). For example, measurements for the lump of sarsen in hole C21 in the June 16 entry of Leaves, not given in Diary, match precisely those in the published report. But Leaves does not contain substantial alterations so much as abbreviations, and some apparently significant differences remain between these manuscripts and the ‘official’ record, particularly concerning rings E and F.

The correct identification of post holes is important because of the amount of information in Young’s diary not present in Cunnington’s report. Pollard demonstrated (1992) that animal remains and artefacts are far from randomly distributed around the site, and proposed that this reflects what people were doing there. Unfortunately, matching all pits with complete confidence does not seem to be possible.

Before re-excavation, all available information was tabulated. Pits in each of the sources were matched following the excavation sequence by ring. It appeared that Young described two post pipes in the ‘double’ E pits, contrasting with Cunnington’s insistence that they had only one (eg E3). Apart from a brief reference to them in Leaves on 16 June, the F pits seemed to be absent.

The new excavation resolved this dilemma. The F and E pits were seen to intersect, showing what Young described as ‘recesses’ at the inner end of the E pits to be the separate post holes of the F ring. This interpretation leaves only the contradiction that Young described post pipes in the F holes, while Cunnington denied their presence.

The tables were changed to reflect this, and these final versions appear at www.hengeworld.co.uk and in the prints at Avebury and Devizes museums.

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 17

Secrion Traoucn Post Hove

Scare Gin. To IFT.

THE Two CORES ARE IN LINE WITH THE CENTRAL Post HOLE.

rete ie fe e

. VL’ ~~) io) > NN

Hoole a

OS 5S9Sa9s5.

CiR CLES.

Soss

5

g

BaEGRe:

Ramen CHALK CORE

Figure 10. Young’s sketch of a post hole in ring E. The section is his drawing from Diary, June 13 1930. The plan is traced from

Draft of the same date (not to original scale)

Because they are inextricably linked, Young’s descriptions of E and F holes remain in the E list, while parts of these are also in the F list.

Alone in all his surviving records, Young’s post hole 2, Circle 5 is treated to a schematic section drawing (Figure 10). In Draft there is a sketch of the profile and plan, and in Diary a more formal version of the profile only. Nothing appears in Leaves.

With the hindsight of the 1999 dig, it is possible to interpret this as showing an E hole to the west (left), the larger F hole on the east and the circular top of a G hole to the south (in plan). The question is, which of these pits are they?

Before the new excavation the large pit was thought to be E4, the others remaining unidentified. Excavation revealed that E4, as shown by Cunnington

18 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

in her plan (M. E., 1931, Plate ID, had a step, absent in Young’s description of his Hole 2 (Figure 13). E3 is the only other pit in that ring that has a configuration with the adjacent F and G pits similar to Young’s plan, and although in Cunnington’s plan it too has a step, no such step was revealed on excavation. Both Young (Hole 2) and Cunnington (E3) refer to a circular hollow on the base of the pit, which they interpreted as a cut in the chalk to take the post (Young gives dimensions of 25cm across and 5cm deep: Figure 10). Like the step (the hollow was not visible in 1999), this may have been formed in hard packed fill. It seems likely, then, that Young’s sketch represents E3, F3 and G2.

The clearest correspondence is with ring C, where excavation occurred in unbroken sequence from pit to pit, and several specific indicators occur (for example, the presence in both sources of a human mandible in one pit or a large block of sarsen in another). The only certain identification in ring D is the first pit dug, D5. It seems likely that excavation progressed in an anti-clockwise manner, but several of the pits remain unmatched. The reference to half sectioning of Hole 10 inYoung’s diaries suggests D10 (cf Cunnington, M. E. 1931, Plate III.4); while the dates of excavation suggest D2.

The distinctive decorated rim sherd illustrated by Cunnington as Plate VII.1 is drawn by Young in the diaries. This would indicate that Young’s Hole 2, Circle 6 was probably G3, supported by his comment that “There was only about six inches of wall on either side, separating [Hole 2] from the two oblong holes’. But he describes Hole 2, by this interpretation mistakenly, as lying between what other evidence suggests to be pits E2 and E3, while G3 is between E4 and E5.

According to Robert Cunnington, describing work at Woodhenge and the Sanctuary, spoil from an excavation was shovelled immediately into the nearest space that had been previously dug out. ‘It meant’, he wrote, ‘that the site looked untidy, and indeed meaningless except to a trained eye’ (Cunnington, R. H. 1954, 229).

It is easy to imagine how pits could be confused at the Sanctuary, which was dug at great speed, by either Cunnington or Young, without extremely careful record keeping. Young himself got in a muddle transcribing his own records of ring C. In Leaves he seems to have jumped text, confusing several post holes. The jet bead appears in Leaves in stone hole C30; Draft, however, agrees with Diary in placing it in C27 - albeit using yet another pit numbering system —which seems the safe attribution. On the other hand,

Cunnington refers to pot sherds from pits D13 and D15 (Cunnington, M. E.1931, Plate VIII and table p 325): the D ring has only 12 pits. With only eight post holes to excavate in 1999, with nearly as much time as the Cunningtons had to dig the entire site, keeping them separate in the mind was nonetheless a demanding task.

None of this totally invalidates the important and detailed record. But it does indicate that neither Cunnington nor Young should be treated as ‘the’ correct version, and that both sources need to be thought of as occasionally slightly out of focus.

WHAT DID THE SANCTUARY LOOK LIKE?

Cunnington’s report contains nothing on her impressions of the Sanctuary structure, other than her conclusion that it was probably not roofed (M. E., 1931, 309). She was more forthcoming in her public lecture. The stone circles had been put up to replace the earlier wooden rings, which were erected for ‘ceremonial purposes’. ‘It is not necessary to picture these timbers as merely bare posts’, she said. “They could have been coloured and adorned in many ways, perhaps even carved into various forms’ (“The Avebury “serpent”: another “Woodhenge” in Wiltshire’. Times 1 August 1930).

Meanwhile Robert Cunnington placed a large roof over the whole site (Cunnington, R. H. 1931). It was Piggott, however, who swung opinion, and the majority of later reconstructions of all timber henges have been strongly influenced by his vision (Piggott 1940; cf Musson 1971; Burl 1976, 318-20, on the Sanctuary, which 1s different from Burl 1979, 124-5 and 193-6; Castleden 1993; English Heritage 1995; Lees 1999; and recent television films using virtual reality models).

Piggott himself acknowledged the inspiration of working with Gerhard Bersu at Little Woodbury. The extent of this influence has recently been highlighted with the publication of some of Bersu’s sketches, not least of an iron age house on the Isle of Man (Evans 1998, Figure 4). Like most commentators, Piggott seems to have followed gut feeling as much as evidence. ‘Great dim rafted halls of magic and ritual’ appealed more to his imagination than the ‘forests of naked posts in which we have all so long and so dismally wandered’ (1940, 221-2).

Three lines of evidence have been used to support Sanctuary reconstructions. The apparent replacement

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 19

of posts in the ‘double’ post holes suggested an extended sequence of phases, supported by the wide range of Neolithic ceramics (detailed by Smith 1965, 244-7). On the other hand the logic of the plan as a unitary design pointed to a much shorter history, perhaps implying contemporaneity of all posts. Although little has yet appeared in print, it is probably fair to say that majority feeling amongst specialists now tends towards a one or two phase structure of unroofed posts and stones (Pollard 1992; Gibson 1998).

Thanks to Young’s diaries and the 1999 excavation, we now have two important new contributions to this debate. Reconciling the apparent contradictions between these sources results in a very different understanding of what was happening on Overton Hill 4000 years ago. It is this that is perhaps the major contribution of the new discoveries.

By extreme good fortune, Young was at Piggott’s lecture to the Royal Archaeological Institute on 7

February 1940 (Young 1940; Pitts 2000a; 2000b). After Piggott had impressed the audience with his drawings and speech, Young was asked by the chair to comment. He offered no more than a few pleasantries, but in his diary expressed his anger. ‘One point in particular’, he wrote, ‘I totally disagreed with. This was [Piggott’s] maintaining that the double post holes at the Sanctuary were the result of re-placements carried out during the gradual growth of the hut, thus flatly contradicting the evidence which came to light when those post holes were excavated’. They had thought of this at the time, he wrote, and cut long sections. These ‘proved that the two posts in each particular case had been erected at one and the same time’.

The 1930 diaries make it clear they did indeed cut several sections, although few were apparently recorded, and the possibility that ‘double’ post holes were intercutting features was considered and rejected. Maud Cunnington thought all posts contemporary.

Figure 11. Interpretation of E4 as a succession of backfilled and re-excavated pits

20 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Newall’s copy of the Wiltshire Archaeological and Natural History Magazine in which her report appeared is in Avebury Museum. He wrote in the margin, quoting a letter she sent in 1931. Someone had suggested that each oval pit held two posts supporting lintels, like Stonehenge trilithons. ‘It seems quite a good idea’, she wrote. ‘I can’t think why none of us hit on it before?’

In light of earlier comments, I believe respect is due the original excavators’ opinions of what they dug up. The question then is what do we make of the apparent intersection of several post holes in the small 1999 trench, and what only very special pleading would dismiss as evidence for two pits in the ‘double pit’? E4, implying that posts were not all standing at the same time?

A modern excavator would have no hesitation is describing the E4 hole recorded in 1999 as the result of one roughly circular pit cutting through another (Figure 8). If we consider the profile more closely, however, it is apparent that more than one recut is present (Figure 11). The compressed top of layer [034] looks like the base of a pit, as does the top of [035] and again the chalk rock below it. The vertical face of these layers to the east is clearly the effect of a pit being dug through them. This left a sharp angle at the edge of [034], as would be expected if that layer was buried under other fill, removed in the 1930 excavation. However, this is not the case with [035], which very clearly sloped gently towards the deeper end of the pit (Figure 11). This could indicate that the pit bottoming on [035] had a very uneven base, but it could also be a relict of yet another pit dug down through older refilled pits. It is impossible now to be dogmatic about this, but the suggestion is that E4 is the cumulative effect of five pits, dug, refilled and redug (Figure 11). In fact, if we accept the evidence of this, then we are recognising the possibility that many more pits than this were once excavated within this small, rectangular patch of ground.

Pit F4, contrastingly described by Young as but a ‘recess’ within pit E4 and by Cunnington as a completely discrete post hole, was dug through the backfilled pits at the inner, western end of E4, or dug away by these excavations: unfortunately we cannot now tell which. Almost all multi-phase reconstructions, perhaps unconsciously following a law of progress that starts with a roofed hut and ends with a temple, place these F pits at the start of the sequence.

It is interesting that there were only two post pipes in this three dimensional palimpsest (if we believe Young’s diaries), and significant that one of these

Figure 12. At completion of excavation in 1999, the four deep holes are (left to right, N trench side top left): E3 (under baulk) and G2; F4 and E4; G3 (round pit); F5 and E5. Other shallower hollows are archaeological excavation (1930 and 1999) of ‘natural’. Concrete posts for D5 and D6 (double) behind. © Mike Pitts

should be in what seems to be, even with the very limited evidence we have, the latest of all the holes dug (excepting the possibility of the pit with the other pipe, F4): the deep, eastern end of E4.

All the evidence can be accommodated if we imagine a succession of pits being dug in the chalk, presumably for tall oak posts. The timbers were held in place by packing chalk hard around them. Excavating D5, Young wrote (1930a, 9 June): “The chalk packing round the core was so firmly rammed that until we reached a depth of 3 ft 6 in [1.1m] it was mistaken for the undisturbed side’. They made the same error excavating E4, leaving a step of packing that we, too, in 1999 initially believed to be chalk rock.

That the Neolithic workers were repeatedly digging through redeposited chalk rather than rock would be one reason to pack filling particularly hard. The presence of only one post pipe in E4 means that previous posts had been removed. That experienced field archaeologists missed that they were digging through recut features could reflect both this very hard packing of clean chalk, and also that posts were not left long in the ground. There was no time for timber to rot, for soil to fall into the pit and contaminate the chalk, for packed chalk rubble to become cemented by percolating rain water. The purity of the fill might even suggest that the site had been completely stripped down to a chalk surface.

So we have large posts being erected, taken down and replaced, perhaps over a brief period. This is hardly an activity to be envisaged beneath a heavy thatched roof, and supports Cunnington’s initial idea that the posts stood free or lintelled. Yet at the same

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY Zl

time, this process is controlled. There remains the fact of the site plan, a regular, patterned arrangement with corridors, circles, posts of recurrent sizes in pits of recurrent depths, which shone through the transience of individual timbers and acts of construction, outlasting those last posts allowed to decay, undisturbed, in the ground, and, after four millennia, even the very megaliths (cf Pollard 1992).

Piggott made a well known and telling comparison between the Sanctuary and a building of Christian worship. “To interpret the Sanctuary’, he said, ‘as a one-period plan may in fact be analogous to considering a parish church of to-day, embodying the alterations and enlargements of successive generations of the pious, as representing an original Norman structure on the site’(Piggott 1940, 196).

With the evidence before us, we can see that this image may be strongly misleading. The Sanctuary no longer appears to us as a monument, a contrived pile that could be extended but not moved. We see rather a process, an intellectual, mythical construct whose existence and meaning lay in the repetition of ceremony, physical effort and pattern (Pitts 2000a). The appearance of the site may have varied significantly from one year to the next, as posts came and went, or stayed and weathered, so that there never were discrete ‘phases’ we could hold up in reconstructions.

It appears on current evidence that only two of the seven rings at the Sanctuary may have consisted of repeatedly renewed posts, the E ring and the D ring. Perhaps we are looking at two categories of circle. On the one hand are the outer rings: the outer stone circle (Young’s photo of ring A under excavation suggests that if posts stood between the stones, the

Figure 13. Post hole E4 after excavation of all fill replaced by Cunnington in 1930. The trowel sits on the step later removed as undisturbed neolithic packing. © Mike Pitts

holes would not have been found; at that point they were looking only for two stone circles), the small and numerous posts of the Fence ring (B) and the serried ring of alternating posts and stones (C). These might be analogous to the immovable ditches and banks seen on some similar sites. On the other hand, enclosed by the first group, are the rings of moving posts, the D and E pits, the F pits which were dug into holes that had formerly held posts and the large G pits squeezed into the gaps (the lower belling of these pits might be accounted for if they had to be dug between already standing posts; as Young says in his diaries, if they had continued down with vertical sides, they would have been too narrow to accommodate a person and very difficult to excavate).

With its emphasis on people and movement rather than memorial and structure, this interpretation is quite different from the normal antiquarian conception. But it resonates well with Gillings and Pollard’s recent description of an Avebury megalith as a living entity, and Avebury itself as ‘not a structure for the ancestors but .. . a carefully choreographed gathering of them’ (Gillings and Pollard 1999, 184; their italics).

FINAL DISCUSSION

What implications does this have for other later Neolithic timber circles? The evidence has been usefully summarised by Gibson (1994). The Sanctuary appears to be unique in its two rings of ‘double’ post holes, and the two stone circles are also distinctive. It could then be argued that the scene here described, of rapid post replacement at a location whose sanctity was of greater permanence than the structures it supported, is also peculiar. Perhaps, though, it is better to think of every location as unique in its own way. The range in size of timber henges is very great (diameters less than 5m, to c 100m at Stanton Drew); the number of rings varies from one to nine; some are circular, some oval, some irregular in plan (at North Mains, a circle and an oval both stood inside a ring ditch, but not, apparently, at the same time); some have standing stones, in a variety of arrangements, while most have none; some have enclosing ditch and bank, some do not; some are on hill tops, some in valleys. And the relationship of these rings to adjacent structures also varies considerably (the Sanctuary, for instance, is at the end of a stone avenue; Sarn-y-bryn-caled is close to a cursus; Durrington Walls rings are inside a ‘superhenge’ earthwork but otherwise unenclosed; and so on).

22 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The influence of Woodhenge, the first such site to be described, should not be under-estimated. The plan gives an air of unitary substance, and combined with Piggott’s interpretation of a roofed building, has spurred archaeologists to look for permanent and covered constructions. At Durrington Walls, Wainwright consciously sought another Woodhenge (Pitts 2000b, 57-8). Unconsciously, many of us have done the same with what is actually a highly varied tradition held together more by the presence of Grooved Ware than any precise architectural pattern.

This is exemplified particularly well at Balfarg, Scotland. Gibson reproduces the post holes of Mercer’s six concentric ‘rings’, omitting the many other excavated features (1994, figure 37). Yet a statistical study commissioned by Mercer himself allowed for only one clear ring (circle A, of large post holes) in a mass of other pits that could equally result from the sort of activity proposed here for the Sanctuary (cf Pitts 2000b, footnote 121).

Perhaps the message is that nothing should be taken for granted. Further excavations of these sites should be scrutinised for indications of post replacements (bearing in mind that rapid post turnover could leave little trace) and duration of activity. The way in which even Woodhenge was excavated might caution us against over-confidence in the conventional interpretation of this site (cf Pitts 2000b, 44-5). There are also many pits at the Sanctuary, and many opportunities for further re- excavation. The dance continues.

ACKNOWLEDGEMENTS

Amongst the 20 who assisted at the 1999 dig, I should particularly like to thank Jennifer Garofalini and Dave Wheatley (Southampton University), John Joyce and Chris Parfitt (University of Wales, Newport), Melanie Pomeroy (English Heritage), and Gill Swanton, and of course Josh Pollard, whose advice, help and encouragement have been invaluable. The skeleton would not have been located without the assistance of Claire Jackson (Royal College of Surgeons), Rob Kruszynski (Natural History Museum) and Karen Walker (Wessex Archaeology). Others whose help is much appreciated include Alex Bayliss and Amanda Chadburn (English Heritage), Ros Cleal (National Trust), Pamela Colman (Wiltshire Archaeological & Natural History Society), Dave Field (NMR, English Heritage), Chris Gingell (National Trust), Jackie McKinley (Wessex Archaeology), Mike Parker Pearson (Sheffield University), Amanda Rouse (Cardiff University), Colin Shell (Cambridge University) and Isobel Smith.

References

BURL, H. A. W. 1976. The Stone Circles of the British Isles. London: Yale University Press.

BURL, H. A. W. 1979. Prehistoric Avebury. London: Yale University Press.

CASTLEDEN, R. 1993. The Making of Stonehenge. London: Routledge.

CUNNINGTON, M. E. 1929. Woodhenge. Devizes: George Simpson.

CUNNINGTON, M. E. 1931.The ‘Sanctuary’ on Overton Hill, near Avebury. WANHM 45, 300-35.

CUNNINGTON, R. H. 1931.The ‘Sanctuary’ on Overton Hill. Was it roofed? WANHM 45, 486-8.

CUNNINGTON, R. H. 1954. The Cunningtons of Wiltshire. WANHM 55, 211-36.

CUNNINGTON, R. H. 1975. From Antiquary to Archaeologist: a Biography of William Cunnington 1754-1810. Aylesbury: Shire.

CUNNINGTON, M. E. and WOODHEAD, T. W. 1931. Report on charcoals from “The Sanctuary” on Overton Hill. WANHM 45, 484-5.

EDMONDS, M. 1995. Stone Tools and Society: Working Stone in Neolithic and Bronze Age Britain. London: Batsford.

ENGLISH HERITAGE 1995. Stonehenge and Neighbouring Monuments. London: English Heritage.

EVANS, C. 1998. Constructing houses and building context: Bersu’s Manx round-house campaign. Proceedings of the Prehistoric Society 64, 183-201.

GAROFALINIL, J. L. 2000. The Construction and Analysis of Three Dimensional Models ... [using VRML and data from the Sanctuary]. University of Southampton Department of Archaeology M. Sc. Thesis.

GIBSON, A. 1994. Excavations at the Sarn-y-bryn-caled cursus complex, Welshpool, Powys, and the timber circles of Great Britain and Ireland. Proceedings of the Prehistoric Society 60, 143-223.

GILLINGS, M. and POLLARD, J. 1999. Non-portable stone artefacts and contexts of meaning: the tale of Grey Wether (www.museum.ncl.ac.uk/Avebury/stone4.htm). World Archaeology 31, 179-93.

GILLINGS, M., POLLARD, J. and WHEATLEY, D. 1999. Longstones Field, Beckhampton: An Interim Report On The 1999 Excavations. Privately printed report.

GREEN, H. S. 1980. The Flint Arrowheads of the British Isles. Oxford: British Archaeological Reports 75.

HAMILTON, M. 1999. Bronze? In Whittle, A., Pollard, J. and Grigson, C. The Harmony of Symbols: The Windmill Hill Causewayed Enclosure, Wiltshire (Oxford: Oxbow Books), 343.

HOLGATE, R. 1988. Neolithic Settlement of the Thames Basin. Oxford: British Archaeological Reports 194. LAWSON, A. 1995.The copper alloy bead and awl. In Cleal, R. M. J., Walker, K. E. and Montague, R. (eds), Stonehenge in its Landscape: Twentieth-century

Excavations (London: English Heritage), 430.

EXCAVATING THE SANCTUARY: NEW INVESTIGATIONS ON OVERTON HILL, AVEBURY 23

LEES, D. 1999. The ‘Sanctuary’, Avebury. An architectural re-assessment. WANHM 92, 1-6.

MURRAY, L. J. 1999. A Zest for Life: the Story of Alexander Keiller. Swindon: Morven Books.

MUSSON, C. R. 1971. A study of possible building forms at Durrington Walls, Woodhenge and the Sanctuary. In Wainwright, G. J. and Longworth, I. H. Durrington Walls: Excavations 1966-1968 (London: Society of Antiquaries),363-77.

NEWCOMER, M. H. and KARLIN, C. 1987. Flint chips from Pincevent. Jn Sieveking, G de G and Newcomer, M. H. (eds), The Human Uses Of Flint And Chert: Papers From The Fourth International Flint Symposium (Cambridge:Cambridge University Press), 33-6.

PIGGOTT, S. 1940. Timber circles: a re-examination. Archaeological Journal 96, 193-222.

PIGGOTT, S. 1985. William Stukeley: An Eighteenth-Century Antiquary (2nd ed). London: Thames and Hudson. PITTS, M.W. 1999. The stuff of archaeology. Past 32, 1-2. PITTS, M. W. 2000a. Return to the Sanctuary. British

Archaeology 51, 15-19.

PITTS, M.W. 2000b. Hengeworld. London: Century.

PITTS, M.W. 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: .are-examination. Proceedings of the Prehistoric Society 58, 213-226.

SMITH, I. F. 1965. Windmill Hill and Avebury: Excavations by Alexander Keiller, 1925-1939. Oxford: Clarendon Press.

UCKO, P. J.. HUNTER, M., CLARK, A. J. and DAVID, A. 1991. Avebury Reconsidered; from the 1660s to the 1990s. London: Unwin Hyman. ;

WAINWRIGHT, G. J. and LONGWORTH, I. H. 1971. Durrington Walls: Excavations 1966-68. London: Society of Antiquaries of London.

WHITTLE, A. 1997. Sacred Mound, Holy Rings: Silbury Hill and the West Kennet Palisaded 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. WANHM 93, 131-180.

YOUNG, W. E. V. 1930a. Diary with Archaeological Notes 1. Manuscript Diary, Library of the Wiltshire Archaeological & Natural History Society (Devizes).

YOUNG, W. E. V. 1930b. Leaves from my Journal I. Manuscript Diary, Library of the Wiltshire Archaeological & Natural History Society (Devizes).

YOUNG, W. E. V. 1930c. Loose pencilled diary notes 26 May - 16 June 1930. Alexander Keiller Museum ref 20000590 (Avebury).

YOUNG, W. E. V. 1931. Diary with Archaeological Notes 3. Manuscript Diary, Library of the Wiltshire Archaeological & Natural History Society (Devizes).

YOUNG, W. E. V. 1940. Diary with Archaeological Notes, January-July 1940. Manuscript Diary, Library of the Wiltshire Archaeological & Natural History Society (Devizes).

24 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Big Bellied Oak, October 1999

Wiltshire Archaeological & Natural History Magazine, vol. 94 (2001), pp. 24-46

Savernake Forest Oaks by Jack Oliver' and Joan Davies?

Climate and soil conditions affect the current distribution of the two native oaks common in England. The Sessile oak prefers a humid site with high rainfall and fairly good drainage. It is the dominant species in the oak woods on the high ground 1n the north and west of the country. The Pedunculate (English) oak requires a more fertile soil and 1s very common in southern and eastern parts. In Savernake Forest we expected to find all or most of the oak trees to be the Pedunculate oak. Instead, following detailed examination of the leaves, leaf and acorn stalks, our survey of the oaks in Savernake Forest found that the two native oak species were both common and regenerating naturally. The Pedunculate and Sessile oaks occurred in roughly equal quantities, whether as veterans or as the much more numerous saplings. Neither species was as common as the hybrid between them Quercus x rosacea. Commonest of all were intermediate trees of all ages which appeared to be introgressed, hybrids back-crossed with either parent. These sometimes produced as many or more acorns as the ancestral parent species, although 1999 was not a very good acorn year for Savernake oaks in general. These three (or five depending on definition) native oak taxa were widespread, making Savernake a mixed oak forest.

‘The original endemic mutant Savernake cluster oak (Q. robur var cristata) still survives. Five of its progeny are in the arboretum area but none of these small trees produced acorns in 1998 or 1999. Of the three non- native oak species noted, we found more than 20 specimens of only one, the American red oak (Q. rubra).

The named and unnamed ancient oaks of great girth and age had been managed at times by coppicing, pollarding or both, or left as maidens or specimen trees, such diversity being a fine attribute of this great forest. In the 1999 survey, over 170 large oaks were measured. The oak with the largest girth found in the forest was the Big Belly Oak. This is a hybrid, Quercus x rosacea, with a girth of nearly 1 Imetres and a coppiced-ring circumference of 14metres; it is very probably the oldest oak in the Forest. Parts of the coppiced base could be a thousand years old. Of the 13 great oaks with girths above 7metres, only the True Braydon Oak retains its full height and spread. The tallest oak in the forest is probably a Turkey oak (Q. cerris) with a girth of 5.3metres. The forest 1s huge and the work 1s continuing. In 2000 a further 260 oaks have been measured and labelled.

INTRODUCTION

British people have always liked oaks. The oak is a tree of imposing stature and great longevity. It is regarded as an emblem of hospitality and strength. The tree, its leaves and fruit, the acorn, are featured in many ancient myths and folk traditions. Ancient tribes worshipped in sacred oak groves and traditionally couples were married under an oak tree. In the legends of many countries the acorn was said to be man’s first food. The products from the oak tree have been very useful to man. For centuries the

hard durable timber of the oak has been used in buildings; it was renowned for building our naval Men-of-War fighting ships, with the ‘crooks and knees’ from the English oak being especially prized. The smaller wood provided firewood and charcoal, extracts from the bark were used for tanning and in the autumn acorns in the woods and forests provided pannage for pigs. Today the oak is the paragon of the modern emphasis on the values of biodiversity. More species of insect are associated with oak than any other tree or indeed any other plant, in Britain.

1. High View, Rhyls Lane, Lockeridge, Marlborough SN8 4ED 2. Ballard’s Piece, Forest Hill, Marlborough SN8 3HN

26 ‘THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The oak has featured strongly in the history of Britain and there are many tales about notable old oak trees, their size and how long they are thought to have lived. Dryden portrayed the life of the oak tree as follows:

The monarch oak, the patriarch of trees,

Shoots rising up, and spreads by slow degree;

Three centuries he grows, and three he stays

Supreme in state, and three more in decay.

The typical growth of an oak includes a period of quite rapid early growth for around 80-150 years, followed by a gradual slowing down. After about 250 — 350 years decline sets in, branches die and diameter growth slows right down. Trees surviving beyond 400 years are often hollow. Old trees gradually replace most of their heavy branches with a shorter head and are described as being ‘stag headed.’ However, it is known that if the tree is coppiced and/or pollarded it can live for 900 years or more. When a tree is coppiced its life cycle starts again and when it is pollarded the rate of growth slows down until a full head is redeveloped.

The largest oak with the greatest girth in the British Isles is the English Oak tree growing at Bowthorpe, Lincolnshire. In 1997 it had a girth of 12.75metres. Savernake Forest is considered to be one of the top five European Forests for veteran trees. Ancient oaks with large girths are to be found there, and although there have been articles and books on the history of Savernake Forest, we have not unearthed much detail on the natural history of the Savernake oaks themselves. This article is intended to give some detail of these forest oaks at the end of the millennium so that comparisons can be made again in 100 or 1000 years hence.

SHORT OUTLINE HISTORY OF THE FOREST TREES

The name Savernake is of Saxon origin. There are references to Safernoc in 934, Savernac in 1156 and Savernak in 1275. The Concise Oxford Dictionary of Place Names says the name is probably derived from a river name identical with Severn. Other people have suggested ‘sweet fern’ or ‘gravel’ and even a ‘hare’, but if the end of the word is considered, all the three old forms oc, ac and ak, are old names for an oak. The place where the oak trees grow? The Old English ac meaning oak has come down to us in ‘acorn, the fruit of the oak tree.

The 934 reference is to the Saxon King Athelstan who, in a Charter, refers to certain crofts lying ‘alongside the woodland which is called Safernoc’. Although Savernake Forest does not appear in the Domesday Book, it is known that the area became a Royal Forest soon after the Norman Conquest and Richard Esturmy, who owned land at Burbage, was appointed the first warden. He is the ancestor of the present Marquess of Ailesbury, whose son the Earl of Cardigan, is the owner and the 31st hereditary Warden of Savernake Forest.

In medieval times a forest was an area of land, usually belonging to the king, set apart for hunting wild beasts and game and having special laws and officers of its own: it was a district subject to Forest Law. As a Royal Forest, Savernake consisted of small woodlands and coppices dotted about with areas of rough grass, heath, fern and scrub. There was sufficient woodland to protect the fallow deer during the winter months and records tell us that timber from the Forest was used to repair Marlborough Castle and to build and maintain two new mills on the river Kennet. The Forest fluctuated in size reaching a maximum in the twelfth century, when it consisted of five bailiwicks. The one known as La Verme, (home farm), was always the bailiwick of the Warden himself. This was in the triangle between Marlborough, Hungerford and Burbage. Disafforestation occurred in 1330; the Forest reverted back to its original size and the surrounding area was no longer subject to Forest Law. The post of hereditary warden stayed in the Esturmy family until 1427 when it passed through the female line to the Seymours.

In 1547, Edward Seymour became Duke of Somerset, Protector of the Realm and the absolute owner of the Forest. Savernake is unique in retaining the name ‘Forest’ and not being called a ‘Chase’ when it ceased to be a Royal Forest. Before this time the wardens had managed the woodland that happened to exist. The Protector’s account books show that he planted some trees; one plantation was at ‘the Great Dych’ which was probably near Tottenham. Deer Parks were made in 1598, by fencing 2,000 acres of land north of Martinsell, between the Pewsey and Burbage roads to form the Great Park, with a further 1,300 acres, separately paled at Brimslade. The residue, the area known today as Savernake Forest remained un-paled. An early sketch map, drawn in about 1600 shows the residue of Savernake Forest as an open area in the centre, with Isbury Copse, Bullstoke, Burch Wood, The King’s Wood and Morlee Wood drawn as small separate woods around the edge. The deer parks had a very short life; by 1675 the fences

SAVERNAKE FOREST OAKS

had fallen into disrepair. In 1718 the Great Park was developed on modern farming lines with compact, 200 acre individual farms and the deer were encouraged back to the residual forest.

In the early 18th century the Forest was in a bad and unprofitable state: it was described as being lightly wooded with few trees, deer were destroying the coppices and the old oaks were becoming more picturesque and worthless. Improvements started about 1745, when Charles Bruce, Earl of Ailesbury, planted a clump of beech trees on Three Oaken Hill, (this name appears in medieval records) near to where the column now stands, and beech along the southern part of Grand Avenue. It was the next Warden, Thomas Brudenell Bruce, who, during the second half of the 18th century, laid down the foundation of the Forest as we know it, with tree-lined rides meeting at Eight Walks. In addition to the beech avenues, oaks with some sweet chestnuts, ash, elm and occasional firs were planted in areas of open ground. He also undertook the destruction of heath and furze. There were five nurseries at Crabtree, Bagden, Braden Hook, Park and Birch Copse supplying a succession of young trees. Many trees were not planted out until they were seven or even fifteen years old. To give these young trees protection against deer, a number of enclosures in the form of banks topped with fencing were constructed. In addition Tottenham Park was reconstituted as a deer park to help to reduce the number of deer roaming wild in the forest. It was also about this time that ‘borderers’ lost their grazing rights in the forest. This period of tree planting continued until 1814. The next warden did not plant trees, he built Tottenham House. His son turned Tottenham Park and the Forest into ‘one great whole’ by constructing, in 1870, a 16 mile deer fence of wood paling on top of a bank around almost the whole of the central block of woodland; only Birch Coppice and Savernake Wood were excluded. Systematic forestry with trees being cut and trees being planted in rotation had never been practised at Savernake. Thus in the mid-nineteenth century, an agent reported that most of the oaks were nearly a hundred years old, with the old oaks in Savernake being four hundred years of age and useless for timber.

The deer kept the undergrowth down, the trees were well spaced out, and although amongst trees, one would not have felt hemmed in. Richard Jefferies described the beeches along Grand Avenue as looking like a continuous Gothic arch, as in an aisle of a forest abbey. The Forest was beautiful, but within a forest deer park it is difficult for young trees to survive. There was a further planting period from 1894 to 1911,

27

mainly in the outlying woods. In the Forest ‘flowering chestnut’ (presumably horse chestnut trees), copper beeches and groups of rowan were planted and protected from deer with wire netting. In the twentieth century there was a change in the national attitude. Hereditary titles and honours were allowed to continue but the free inheritance of wealth was not. In 1939, the 28th Hereditary Warden leased the sylviculture rights of the Forest to the Forestry Commission. One clause in the lease said that reasonable endeavour should be made to regenerate the old Forest by natural means. To aid natural regeneration and to allow trees in new plantations to develop, the deer were driven into Tottenham Park. During the Second World War the Forest became a giant ammunition dump, guarded by American soldiers. The Forestry Commission continued its work with Land Army girls being employed in Thornhill Nursery to sow local acorns and tend the young trees. After the war plantations of oaks from this nursery were planted in the western part of the forest and conifers in the eastern part.

Most of the present extent of Savernake Forest, 905 hectares, was notified as an S.S.S.I. in 1971 and 1988, largely on the grounds of exceptional biological diversity. The outline description and reason for notification included the following summary of the geology and dominant trees:

The forest lies on a plateau of Upper Chalk, covered by

Clay-with-Flints and dissected by dry valleys. The Chalk

is exposed along the valley sides; however small deposits

of Valley Gravel, Reading Beds and Bagshot Sands, together with the clay, give rise to particularly acid conditions in places.

Relicts of the ancient wood pasture remain, represented by the distinctive open-crowned specimens of sessile and pedunculate oak, which are scattered about the site. Eighteenth and early nineteenth century plantations of beech and oak have now assumed a semi- natural structure but the Forest consists mainly of twentieth century beech or oak plantation. These stands are supplemented by naturally regenerated silver birch, ash, downy birch, rowan and willows. Wych elm, field maple, holly and midland hawthorn aiso occur and hazel is locally frequent in the former coppice areas. Hawthorn is abundant and, like blackthorn, forms stands of scrub in the open spaces where it is a valuable nectar resource for deadwood insects.

The preceding description implies natural regeneration of about 15 tree species, but this count should now be raised to 20 or more, with perhaps a further 5 or more fertile tree hybrids back-crossing and reproducing naturally. Savernake Forest is known for its tall beautiful beeches, many (or even most) of

28 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

which are characterized by very narrow-angling of the main branches. This seems to be a genetic propensity, as these beeches grow like this even when not crowded; furthermore the Forest has some normally spreading beeches, even though many of these are densely shaded and surrounded. Both types of beech grow taller than the two native oak species in the Forest (and elsewhere), and cast a deeper shade. This also applies to some of the rapidly heightening recently planted conifers, especially the Douglas firs.

This sketch of the history of Savernake Forest trees is simplified, but illustrates our contention that systematic destruction of the native trees has never been absolute over contiguous wide areas or continuously sustained over centuries. Patchy survival and recovery of deciduous summer forests in general and the persistence of occasional ancient oaks in particular has been the general pattern.

GENETIC ASPECTS

Even for specialists, oaks are known to be particularly difficult subjects for genetics and cytology. Tannin concentrations interfere with cell preparations, and the oak chromosomes are very small and homogeneous. Nearly all species of oak are diploid with 24 chromosomes.

High degrees of sterility are reported when crosses are attempted between different species of oak, including the two native species. However the botanical literature contains references to the abundance of hybrids in natural oak populations. Hybrids of the indigenous British oaks (Q. robur x Q. petraea) have been described for oak populations in Scotland, Ireland, Wales, NW, SW, SE and central England. Natural populations of Q. roburx Q. petraea hybrids have been reported also from France to Yugoslavia (Wigston in Morris & Perring 1974, see also Minihan & Rushton 1984).The most recent New Flora of the British Isles (Stace 1997) states that the hybrid Q. petraea x Q. robur ( = Q. x rosacea)‘. .. occurs throughout the British Isles in areas where one or both parents occur, occasionally being commoner than either ... and is fertile’ The other main flora of the British Isles (Clapham et al 1989) comments on the frequency of Q.x rosacea where the parents grow together.

Wigston (1974) explains the paradox as follows: ... the interspecific fertility barrier (between pure Q. petraea and Q. robur) may be relatively unimportant if the hybrids are fully fertile with either parent. It is also possible that reduced fertility will

‘

be little barrier to introgression if hybrids and backcrosses have any competitive advantage over parental genotypes.’ He then discussed the different statistical analyses based on multiple samples relating mainly to 7 or so leaf characters to classify oak woods as 4 types: dominated by Q. robur, by Q. petraea, by intermediates, or mixed woods with intermediates. The discriminatory characters separating Q. robur from Q. petraea can be independently variable and most probably under polygenic influences, although the presence or absence of pubescence on leaf undersides could be mainly under the control of a single allele. Even so, when present the quantity and distribution of pubescence may be polygenically governed.

In short a pure Q. petraea and a pure Q. robur might be able to produce an FI hybrid only occasionally, and the hybrid might not be a good acorn producer by selfing or crossing with another F1 hybrid - few or no F2 progeny. However repeated back crossing (perhaps with first one and then another parent over the generations) may produce trees with intermediate balanced characters, or trees inclined towards one or the other ancestral parent species (introgression). These recombinant types by natural selection might become more successful acorn producers, during some years or in certain localities, than either of the pure ancestral parent species.

AIMS AND METHOD OF STUDY

The initial idea was simply to measure the girths of the largest trees in Savernake Forest as part of the Wiltshire Botanical Society (W.B.S.) project of recording the largest and veteran trees in the county around the start of the new millennium, in order to provide records for the future. Girths unless stated otherwise were measured at 1 /% metres (5ft) above ground level, as closely as possible following the rules suggested by the Tree Register of the British Isles (TROBI founded in 1988).

Subsequently it seemed appropriate to look into the histories of any named and/or large oaks, and to record interesting features such as past coppicing or pollarding, shape, size, hollowness and botanical variation. For these largest oaks details were collected per tree on girths, descriptions, 6-figure grid references, description of sites, data for precise botanical designations, and where possible any historical records and associated stories.

SAVERNAKE FOREST OAKS

29

Savernake Forest

Picnic site

1 Amity Oak

2 Duke's Vaunt

3 King of Limbs

4 Slingsby Oak

5 Original Queen Oak 6 Big Belly Oak

7 New Spiral Oak

8 True Braydon Oak 9 Cathedral Oak

10 Surveyed Oak

Eight Walks is near the centre of Savernake Forest. In order to make the project manageable, the survey was carried out in eight sectors (octants) of the Forest, demarcated by each of the 8 straight roads, drives or walks radiating from the centre. Twelve O’clock Drive (12-OD) runs due north to south. Octants I — IV are east of 12-OD and octants V — VIII west of 12-OD.

IDENTIFICATION

There are two native oak species and numerous intermediates between them, as well as some introduced oak species. Twenty eight samples were sent to the expert referee for the Botanical Society of the British Isles to authenticate such oak taxa not previously recorded for Savernake Forest, and to provide the authoritative guidelines for the correct taxonomic designations of the 150 largest oaks in the Forest and saplings around them (see Acknowledgements).

Birch Copse

The main features scrutinized for each of the big oaks were those that discriminate best (Stace 1997, Rich & Jermy 1998) between the two native oak species: petiole (leaf-stalk) lengths, peduncle (acorn stalk) lengths and pubescence (where available), leaf bases (two features) and pubescence or not on the underside of the leaves (examined by binocular microscope at x30 magnification for reasons given subsequently). Some attention was also given to the number of leaf lobes and the depth and regularity of the leaf-lobing. Five or more measurements of petiole length were taken from the leaves of each tree to give the extremes and median per tree. Based on the preceding, the taxonomic designation was decided for each tree. The BSBI referee had emphasized that in looking at the oak hybrids and intermediates, one should take into account as many of the preceding features as possible rather than leaning too heavily on microscopy alone or petiole length alone, although these two components tend to be the most satisfyingly definite to measure, being less susceptible to judgment than the others. On any one tree, for instance,

30 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

pubescence and petiole length were much more constant than leaf shapes, leaf bases and peduncle lengths for Savernake Forest native oaks.

Finally note was made of other unusual variations not mentioned in the floras, or not defined as characters known to discriminate between the two native oak species.

MICROSCOPY

Examination of the underside of the leaves using a x8 or x15 magnification standard botanical lens was not adequate for Q. x rosacea. All leaves were checked, and where there was pubescence, hair lengths measured, using a binocular microscope with a x30 magnification and an eyepiece graticule. Tiny webs from minute Arachnida could mimic leaf pubescence.

Pure Q. robur leaves and the acorn stalks were almost hairless, apart from a few simple appressed hairs apparent on the leaf under surfaces of a few trees. Pure Q. petraea leaves had numerous small stellate hairs (type A below), widespread and consistently spaced on the under surface of the leaf laminas, but not overlapping; also concentrations of larger stellate and other hair types (type B-E below) alongside leaf veins, but especially in the vein angles, on the under surfaces.

A Stellate hairs with approximately 4-6 slender rays neatly appressed to the laminar surface with a spread of just under 0.2mm.

B Bottlebrush or brush hairs, initially projecting vertically but spreading out to a greater or lesser extent away from the surface.

C Simple hairs, vertical, appressed or angled alongside the veins, uncommon in vein angles.

D Intermediate between B and C, two or more vertical branches from a single base.

E Untidy stellate hairs, not neatly appressed as in type A preceding, and obviously longer.

The longest of the hairs of type B, C, D, and E were 0.4mm, (exceptionally 0.8 mm). Types B, D and E could be dense and tangled in vein angles and alongside larger veins, commoner than type C and always bigger than type A. Where the short acorn peduncles of Q. petraea were available, any of types B, D and E were common and sometimes closely massed and tangled, and types A and C were occasional.

Q. x rosacea trees with general features evenly balanced between the two ancestral parents, and intermediate trees inclined either towards Q. robur

or Q. petraea all had widely varied leaf pubescence. Many of these trees had much smaller sublaminar stellate hairs than type A above. These could have fewer rays than the 4-6, with a spread of less than 0.1mm, making them hard to see with a standard botanical lens especially as such rays were very slender and closely appressed, and often irregularly and inconsistently scattered, dependent on the leaves chosen. Likewise the hairs alongside veins and in the vein angles might be as large and as densely tangled as types B to E above or smaller, sparser and more erratically organized. Types C and D could be more common than B and E on the leaves of some hybrid trees. Hairs of types B, D, and E might only be found in leaf vein angles, or be fairly common on some leaves, but far less so on others from the very same tree. Acorn peduncles varied in length and pubescence even on the same tree, but usually with some or many B, C and D type hairs.

In summary, intermediacy in the hybrid native oaks could be indicated on acorn peduncles or under the leaf surface by smaller or fewer hairs of rather less complex types, or in different relative proportions, often with erratic or inconsistent scatter; or they could be similar to the Q. petraea pubescence, but with other features (see next section) typical of Q. robur (see also Oliver 2000).

PETIOLES, PEDUNCLES, LEAVES AND ACORNS OF NATIVE OAKS

Specimens were taken from 159 of the largest oaks in Savernake Forest, usually one or more branchlets of average size and appearance for the tree in question, carrying between 5 and 40 leaves. The petioles (leaf stalks) were measured to find the minimum and maximum lengths for that tree, giving the spread (or range) from which the median value was calculated. There was found to be variation in both the petiole lengths and spread for different oaks. Some oaks (all Q. robur) had most leaves with no petioles, and at the other extreme there was a Q. petraea whose petioles reached 30mm. On 103 trees the spread was 5mm or less and for 2 trees as great as 20mm.

The graph shows the variation in median lengths and the large number of trees that are above the 2- 3mm range for Q. robur but below the 13-25mm range for Q. petraea (Stace 1997). In the case of Savernake Forest oaks this was as high as 107 out of

SAVERNAKE FOREST OAKS

Petiole Lengths

Number of Trees

We S85 SAAS S17 A921. 23:.25 | Median Length

the 159 trees sampled. Stace gives the exceptional range of the Q. robur petioles as 0 — 7mm. Even taking these extremes, 64 trees still had petiole lengths which fell between but belonged to neither ancestral parent.

In other respects than pubescence and petiole lengths, characteristics of the pure species were also less common than characteristics of trees with hybrid features. Q. robur has cordate and auriculate leaf bases, with usually 8, 6 or fewer deep irregular lobes to the leaf outline. Q. petraea has cuneate (sloping) leaf bases without auricles, and usually 10, 12 or more lobes to the leaf outlines, regularly but shallowly incised. Q. robur has glabrous (hairless) peduncles (acorn stalks) 2-9cms long, whereas Q. petraea has peduncles 0-2(4)cms long with clustered hairs. There were 3 main types of intermediacy for the majority of native oaks in Savernake Forest:-

1 Intermediate, indeterminate features: petiole lengths fitting neither parent species, leaf bases semi- auriculate, subcordate or cuneate, leaf outlines between the 2 parent species; peduncles 2cms long with a few hairs.

2 Discrepant features: e.g. leaf lobing and pubescence like Q. petraea, but leaf base typical of Q. robur. There were 2 trees with short petioles and short peduncles, and 2 with long petioles and long peduncles, which combinations are quite wrong for the ancestral species.

3 Variable features: many trees had leaves which varied so much in lobing and in their leaf bases or range of petiole length that some leaves could be found on a single tree which could fit Q. robur, others fitting Q. petraea whilst still others matched neither satisfactorily. In particular, a large proportion of the trees on which acorns were found had peduncles which on any one tree could be as short as 1mm or as long as 6cms. Many of these infructescences in 1999 had some or most acorns aborted or semi-aborted;

=

31

such trees contained bunched, intermediate and semi-elongated infructescence stalks, with only a few bearing healthy mature acorns (at best usually only one per infructescence) by September.

This paper mainly concerns the oaks of greatest girth, but all concomitant observations would indicate that the saplings and younger trees were just as variable and intermediate as the ancient oaks — if anything more so in respect of their leaves.

The preceding paragraphs concentrate on features known to discriminate between Q. robur and Q. petraea, but some other variations were noted. Some of these might have been environmentally induced, but more probably indicate genetic propensities. 2 trees noted had leaves of only 4-6cm long over the entire tree, and 6 were noted for huge leaves, 16-20cms on most of the boughs, whether in light or shade. Acorn sizes and shapes for both native species are given in a standard Flora as ‘1.5-4cm, ellipsoid or oblong...... cups 1.5-2cm diameter’ (CTM1989), but several trees were found in the angle between the A4 road and Grand Avenue with sub-spherical acorns. 2 trees were found with very slender, cylindrical (but mature and healthy) acorns, with acorn-cups consistently less than lcm in diameter.

Savernake Forest native oaks therefore are immensely variable in their fruits and foliage. The majority fall between Q. robur and Q. petraea in accepted characters which discriminate between these two species.

THE OAK TAXA IN SAVERNAKE FOREST

The ensuing proportions are reasonably closely in line with the specifications made by the BSBI Oak Referee on the initial 28 specimens submitted to him. The submission of those might have been influenced by some inadvertent selection bias, whereas the percentages below relate to all the 159 native oaks of greatest girth that had been found and identified in 1999 in Savernake Forest. The two ‘probable’ categories are simply a recognition that different botanists have different opinions as to the boundaries between Q. x rosacea and the two parent species (Rich & Jermy 1998).

Quercus robur L. Pedunculate or English oak (QR) 21% of the total. There were also sprinklings of pure QR saplings, but

32 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

not necessarily to 21%. The impression was that QR was perhaps less common in the deeper parts of the forest and more common in fringe and picnic areas, but it was fairly common in all 8 octants.

Q. robur (pedunculate) var cristata (Henry, 1917).

The Savernake cluster oak. Leaves densely

clustered, asymmetric, oblique and twisted; acorns only one per peduncle, with flattened apex and a depression containing the styles. Acorns fertile, a proportion (or most?) of which come true, possibly by parthenogenesis (see also Grose 1957). There are 6 small trees to be seen in the Arboretum area (in Octant VII). No acorns, seedlings or saplings seen in 1998 or 1999 on these, or on the ancestral mutant tree (tag no. 08866) at SU 216653 (near the south edge of Column Ride in Octant V).

(Probable) Q. x rosacea Bechst.

Semi-intermediate or introgressed native oak, robur features dominant (QSR). 13% of the total, but possibly commoner in young trees. Throughout the Forest, all 8 octants.

Q. x rosacea.

Hybrid native oak, fully intermediate (QS). At 27% the commonest of any taxon in Savernake Forest. 38% of the largest oaks of all (the 39 oaks greater than 595cms in girth) were QS. Young QS trees were common.

(Probable) Q. x rosacea. Semi-intermediate or introgressed native oak, Q. petraea features dominant (QSP). 17.5% of the total, with young QSP trees fairly common in all octants.

Q. petraea (Matt.)Liebl. Sessile or Durmast oak (QP). Also 17.5% of the total, with young trees fairly common in all octants.

Combining some of these percentages, it can be seen that approximately 1/3 of the Savernake Forest oak trees incline towards or are Q. robur; approx. 1/3 incline towards or are pure Q. petraea; and nearly 1/3 are pure Q.x rosacea with balanced characters. Alternatively all intermediate and probably introgressed types of Q.x rosacea, including the back-crossed trees which incline towards one or other ancestral species account for

more trees than both the pure species together, approaching 2/3 of all Savernake native oaks. There were no markedly exclusive concentrations of any one of these taxa 1n certain parts of the Forest. They were well spread around, generally ‘all mixed in together’.

Q. rubra L.(Q. borealis F.Michx.)

American red oak. The 4th most common oak taxon in the Forest, nearly all (or all?) from previous plantings. None were over 200cm circumference, but some were much smaller, only 30cm in girth or less. One seedling was noted near the A346, where some large red oaks had been cut back to their stumps and were sending up new shoots; but no acorns were found in 1999. Noted in the following octants:- IV &V East and west of Twelve O’clock Drive, VI South of Great Lodge Drive, VII Around & between White Road intersects with Long Harry and Grey Road, and in greater numbers flanking parts of Grey Road. Also a grove of about 6 trees near the main camping and picnic site: one tree was seen well north of Cadley Church, 5yds east of the A346 road, but more can be seen every late autumn when they are conspicuous either side of the A346.

Q. coccinea Minchh.

American scarlet oak. Labelled as such, a small grove in the arboretum (Octant VID), but very similar to the Q.rubra preceding. No acorns in 1999.

Q. ilex L.

Holm or evergreen Oak. Three low branching trees on the grassy triangle at the road junction near to Sicily Cottages (Octant III). One over 2 metres in girth but no acorns in 1999,

Q. cerris L.

Turkey oak. A majestic tree towering above the forest canopy, with a girth of 5.30 metres, grows 15 metres west of Twelve O’clock Drive in Octant V at SU 225 665. This is probably the tallest oak, and possibly the tallest tree in Savernake Forest. A naturally seeded sapling was also found in the cluster oak grove in the arboretum area in Octant VII, but much broken and damaged either by deer, or in order to clear the area around the cluster oaks. No acorns were found in 1999.

SAVERNAKE FOREST OAKS

NATIVE OAKS IN SAVERNAKE FOREST

Some of the ancient oaks appear, from the appearance of the basal stool, to have been coppiced in the distant past. A few of these have a basal circumference of 8 metres or more, comparable with the ancient Savernake chestnut coppice rings, and probably dating from the same eras. A greater proportion of the oaks had once been pollarded, from 5—12 feet. Rackham (1974) states that pollards are characteristic of non- woodland sites and wood boundaries, but occasionally pollarded oaks had been so closely spaced that the illusion of woodland is now given. Systematic pollarding was practised from the 15th century or earlier. It prolongs the life of the oak so managed and most of the famous ancient oaks have started as non- woodland specimens, often pollards.

Savernake Forest was much more open in the past, and chestnuts, beeches and oaks seem to have been coppiced or pollarded at times in previous centuries. However the majority of the ancient oaks in Savernake Forest are gnarled and nodular hulks, hollowed, mis- shapen and stag-horned, and show evidence of phases of management followed by long periods of neglect. Some oaks appear to have been coppiced, allowed to grow, then subject to pollarding cycles, then left for major branching to develop from the pollarded head. Others, probably boundary, hedgerow or embankment oaks have been once cut or broken at intermediate levels, 4 or 5 feet, or had once been hacked around on one side, but not on the other.

Other factors have shaped the Savernake oaks and influenced longevity. Some have been naturally ‘pollarded’ by gales, or just lost large limbs, or been shifted to 45° slant but have survived. Glazed frost and snow have snapped horizontal branches. Crown die back and decay of the standing tree following sudden isolation and exposure has made some oaks vulnerable to other invading pathogenic and saprophytic fungi (see Murray 1974). The very numerous and lofty great beeches in Savernake Forest cast deep shade that overshadows and kills many of the oak branches. Beeches, growing to 46m (Stace 1997) are taller trees than the oaks and although oaks have a longer life than beeches, some Savernake pollard and standard oaks have died in the deep shade before the beeches become senescent. In Savernake Forest a number of ancient oaks in the Postwives Walk and the Birch Copse areas appear to have been heavily shaded in the past, especially in the Duke’s Vaunt locality. There are some huge black ancient oak trunks, often split with broken

33

segments pointing up within the enveloping conifer canopy, with dead regenerated branches which might have survived had they not been shaded out. It is not certain how many of the massive dead hulks were wholly or partly the consequence of the living oaks having been overtopped by beeches, chestnuts or tall conifers. Douglas firs, for instance can grow to 65m high in Britain. The policy now is to thin conifers around veteran oaks, but not too suddenly because of the danger of light and exposure shock to a previously enveloped tree. After 2 or 3 years an open glade is created around the veteran oak.

Despite the preceding factors there are in addition to oak plantations abundant oak saplings of all sizes springing up in many parts of Savernake Forest. No systematic counts were made of the types of naturally regenerating young oaks, but the taxa QR, QSR, QS, QSP and QP all were common, with the strong impression that hybrids and introgressed trees accounted for 2/3 or more of these young oaks.

In 1999 there were 46 oaks found with girths greater than 600cms (at or near 150cms above ground level). 5 of these (11%) were pure Q. robur, 7 (15%) were pure Q. petraea. The remaining 34 (74%) are ancient hybrids or partial intermediates, probably introgressed, (See Table 1).

A wrong impression can be given if excessive emphasis is placed on the gnarled, contorted, often stunted ancient pollards of great girth alone. Savernake oaks vary immensely. Some have smooth trunks, others are heavily burred. There are many huge open grown spreading oaks (as in Holt Pound Avenue), but even more semi-open grown lofty trees (as in the main picnic and barbecue areas). In the dark depths and around Postwives Walk (Octant V) there are a few splendid tall oaks with tall straight trunks, with beautiful green mossy fluting ascending from up to 2.5 metres from the ground level. Many of the fine tall oaks have girths of between 350 and 550cms. The most impressive and the tallest oaks do not necessarily, or even usually, have the greatest girths.

SOME GREAT AND NAMED SAVERNAKE NATIVE OAKS FOUND IN 1999

Big Belly Oak

Grid reference SU213 657 in OctantVI, attached plate code 06924. Q. x rosacea. Bechst. (i.e. a hybrid). Girth 10.8metres.

34 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Van on the A346 road passing behind the Big Belly Oak

Besides being marked on Ordnance Survey Maps as the Big Belly Oak it has been called, according to W.Maurice Adams in his Sylvan Savernake of 1900 ‘Decanter’ or ‘Big-bellied Oak’ because of its shape. It is the very conspicuous oak on the eastern side of the main A346 road from Marlborough to Burbage, south of Cadley. It actually bulges into the A346 road and lorries have knocked bits off its side. It is nodular, burred and well named, seemingly the result of ancient coppicing as well as pollarding. At a girth of 10.8m it probably has the greatest circumference of any Wiltshire oak. There is another oak comparable for size in Spye Park. Looked at from the road the Big Belly Oak does not give the impression of immense size, but viewed from inside the forest it is broad enough to hide a passing lorry. The combinations of different levels of the A346, undulant ground on the east side, the old pyramidal base, bulges, burrs and large nodules, all conspire together to make measurements unreliable. It is possible to make the girth (at 150cms from the ground) anything from 10.7 to 11.3m. The basal measurement is at least a remarkable 14m around the original coppice stool, itself irregular in shape and height, even with a substantial segment removed on the west, the side of the A346. Taking all this together the Big-belly may be the oldest oak in Savernake Forest, parts of its base being older than the Forest yews and ancient sweet chestnuts.

Legend records that if you dance naked twelve times anticlockwise around the tree at midnight, the devil then appears. We have noi validated this. Brambles to the east and dangerously close semi- continuous heavy traffic to the west of the trunk together seem to have discouraged any test dancing in recent years.

In Elizabethan times and earlier, the road going south from Marlborough was not along the line of the A346 road, but further to the west. Just to the east of the big Belly Oak there are the remains of the old Roman road from Old Sarum to Mildenhall, (Cunetio). With religious houses at Easton, Lilbourne and Ramsbury, this track could have remained in use during the middle ages with the Big-Bellied, pollarded oak tree used as a landmark. On the sketch map of 1600 showing the perambulation of 1301, the Braden Way is probably along the same line as this Roman road. It was the boundary between the Forest and a narrow tongue of land coming northwards into this part of the Forest owned by William of Lilbon and the Prior of Easton. In this case the Big Belly Oak could be an old boundary marker. Just to the west of the oak, on the other side of the road, there is a very narrow band of woodland, which is part of Savernake Forest. The forest pale of 1870 ran along the western side of this narrow strip and according to the Andrews’ and Dury’s map of 1773, and-on Greenwood’s map of 1820 this was the line of the boundary between the hundreds of Selkley and Kinwardstone.

King of Limbs

Grid reference SU242 660 in Octant III, attached plate code 09246. Q. x rosacea. (QS). Girth 10.30 metres

Another well-named oak tree, with huge spreading limbs. It has a hollow trunk that was once burnt out inside but it has made a subsequent recovery. From its size this must be a very old oak. It is in an area marked as Birch Copse on modern maps. As this tree has such wide spreading branches it must have been growing in an open area and in the past it could have been used as a marker. It is just inside the Savernake

King of Limbs

SAVERNAKE FOREST OAKS

parish boundary, in a small jutting out triangle. At the apex of this triangle lies the boundary between Burbage and Little Bedwyn Parishes. At one time the old track along the column ridge, known as Three Oaken Hill, continued north-eastwards to meet the A4 road at Knowle farmhouse. This track would have passed close by the King of Limbs.

The undergrowth in the Forest is variable and in summer this oak can be difficult to access. In 1875 Richard Jefferies in this description of Marlborough (Savernake) Forest said ‘More bracken. What a strong, tall fern! It is like a miniature tree. So thick the cover, a thousand archers might be hid in it easily.’ In August 1999 the bracken canopy surrounding this tree varied from 1.5 to 2.4m high. A giant bracken frond (Pteridium aquilinum (L.) Kuhn) of 4.78m (15ft 10ins) was found nearby, growing up through young Douglas firs. This is a record frond or leaf size for Wiltshire, probably for England and possibly for Europe.

Cathedral Oak or Monarch

Grid ref. SU205 680 in Octant VII, attached plate code 09500. Quercus robur L. (QR). Girth of 9.95 metres

This ancient pollard is on the boundary of Furze Coppice, next to a well-used path. The 1999 notice claimed this to be the oldest Savernake Oak, from 996AD. We believe that both Cathedral and Monarch are new names, but the notice has given rise to an even newer name -the ‘Millennial Oak’. It is very difficult to estimate the age of an old tree and in fact this tree may not be the oldest Savernake oak as there is no evidence of a basal coppice ring. It has a very big head (it has not become stag headed) and only the third greatest girth of the Savernake oaks. As the tree is in good health and only slightly hollow, its girth should continue to expand at a reasonable rate, so it has the potential to become a champion in the future. The Cathedral Oak is not on a parish boundary but on a boundary associated with the management of the Forest. Furze Coppice Lodge was the home of a ‘quarter keeper’, whose main occupation was to protect the deer. They had allowances of firewood, a few acres of land for their own sheep, cows and horses, fees from deer skins and selling fern. Yearly they had to mark the horses and domestic animals belonging to the inhabitants of Marlborough who had signed an Agistment Deed with the Forest Warden. In 1705 it cost eighteen pence for each mare and gelding to be marked.

35

Cathedral Oak

Duke’s Vaunt

Grid reference SU238 665 in Octant III, attached plate code 08990. Intermediate or introgressed Native Oak with Q. robur features dominant (QSR) Girth 8.9metres (but previously the girth was probably more; note the separate halves below).

Sketches of this tree were made in 1802 and 1893. The sketch of 1802 shows a majestic old tree in an open position with wide spreading branches and a tall hollow trunk, with openings down one side. On the top left-hand side of the base opening there is a large bulge in the trunk. In 1893 the tree is without its top part, with the same opening in the base and bulge on the side. There is a report, in the Gentleman’s Magazine of 1802, when its girth was given as 30ft (9m), of twenty boys having been shut up inside the hollow in about 1762. At that time the tree had a door hung on it and the oldest men said that the oak had been in the same decayed state for as long as they could remember. Thus in 1700, three hundred years ago, it was in a state of decay. Maybe the Duke’s Vaunt could be 900 or more years old and started life during the reign of a Saxon King. It is situated just within Savernake parish, a few yards from the

36 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

boundary with Little Bedwyn, which is also the boundary between Selkley and Kinwardstone hundreds. Four great oaks are recorded on the Forest Estate Map of 1786, with the Duke’s Vaunt named as the Duke’s Font. It is known from the perambulation made in 1301 that, the then Warden, Henry Sturmy, owned land at South Lease and Knoll on the eastern side of this part of the Forest. If it is assumed that this land passed down to the Warden, the Protector, the Duke of Somerset, what is more natural than, when he became the owner of the Forest in 1547, the magnificent old oak, still in its prime and within the Forest, just on the western side of his own land, should be named the Duke’s Vaunt (or Font).

In 1900 it was claimed still to have the largest trunk of any tree in the forest, although owing to an accident in 1880 it narrowly escaped destruction by fire and its sides had somewhat fallen in. In about 1950, some of the branches were supported by chains tied to a near- by beech. The beech has gone, the oak remains: it is in two halves with hanging loops of chain. Because of surrounding conifers, planted in 1979, the oak tree is now in heavy shade. A start has been made to let in more light by felling a few of the close conifers. In time more of the surrounding Douglas firs will be felled. Bill Ayers, the Forester, says that it is important to let the light in gradually so the ancient tree is not suddenly subjected to bright sunlight or strong winds.

Yin

Duke’s Vaunt

Unnamed

Grid ref. SU212 667 in Octant VII, 70 m. N. of Great Lodge Drive, tag No 08977. Quercus x rosacea. (QS). Girth 7.6 metres at 120cms, measured at this level on account of the low divergence level (200cms) of the two massive limbs. Coppice girth of 9.5metres.

Possibly this was once an ancient coppiced oak as well as a likely old pollard. This oak is situated close to the highest point on Great Lodge Drive at 559ft above sea level. This Drive, which goes from Great Lodge Farm to Eight Walks, first appeared on Greenwood’s map of 1820. From the Andrews and Dury map (1773) a track from St Michael’s Farm going northwards, into the forest, passed this tree. This track descended into the valley to join the one to Braydon Hook which came from Great Lodge Farmhouse. Part of this farmhouse dates back to Elizabethan times. Old pollarded oak trees are often found on or very close to an ancient boundary. This does not appear to be so in this case.

Unnamed Grid reference SU224 659 in Octant V. Quercus x rosacea. (QS). Girth 7.4metres.

The trunk is more flattened than cylindrical, and like the previous one it has two huge divergent limbs at 220cms. It is probably an ancient pollard. This unnamed ancient oak is very near to the site where the King Oak once stood. The Earl of Cardigan in his very interesting book “The Wardens of Savernake Forest, published in 1949, describes a visit he made, as a boy, with his grandfather see the King Oak — “The King Oak was standing with other giants (includes this unnamed tree) in a curious glade in the Forest — a glade where the bracken, so profuse elsewhere, would never grow. At this time the King Oak was a very decayed veteran that would not last much longer. A picture post card of 1920 shows the dead trunk surrounded by a low wooden fence. During the war, the remains of this famous tree were taken away as souvenirs by American soldiers. The unnamed tree lives on.

‘Roman Road Oak’

Grid reference SU 217 677 in Octant VIII, attached plate code 09209. Quercus robur (QR). Girth of only 5.7metres at 150cms, but the coppice base at 30cms is 8.8metres in circumference.

This oak is situated at the intersection of the Roman Road with Red Vein Bottom. It is the southern one of a pair of oaks. The size of the coppiced base probably

SAVERNAKE FOREST OAKS

indicates great antiquity. This is one of a number of ancient coppiced or/and pollarded oaks situated on this old Roman road to Old Sarum. None have been found on the other Roman road through the forest.

Surveyed Oak

Marked on Ordnance Survey maps, SU 217 676 in Octant VIII, tag No 08947. Q. x rosacea (QS). Girth of 7.4metres

This oak is situated a few metres east of the Roman Road, in a dense thicket of wych elm, beech, hawthorn and blackthorn. It is another probable ancient pollard, again with two huge limbs diverging at 220cms. The Ordnance Survey has surveyed this oak as it defines the boundary between Savernake and Mildenhall parishes. It is in an area without any forest banks.

‘Slingsby Oak’

Grid ref. SU 224 654. in OctantV, 120m W. of Twelve O’Clock Drive, tag 08830. Intermediate or intro- gressed Native Oak, Q. petraea features dominant (QSP).Girth 7.1metres.

In the summer this oak is in deep shade, but it is conspicuous because of the large green metal deer- watch. It is possibly an old pollard, with two huge boughs from 300cms that are very narrowly angled. In September 1999 there were spherical galls, as large as 0.5cm diameter, on all the leaf buds and all the (aborted) flower buds, and no acorns forming.

True Braydon Oak

SU 216 670 in OctantVII, attached plate code 08958. Intermediate or introgressed Native Oak, Q. robur features dominant (QSR) Girth 7.1metres.

A magnificent oak, of great height as well as considerable girth. It is bigger, better and more beautiful than the other contender as the ‘Braydon Oak’. A massive fractured-off fallen bough itself has a girth of 370cms. The Braydon Oak is marked at this grid reference on a pre-war large-scale map. It is close to the Roman road to Old Sarum and to Braydon Hook Lodge. There were two keepers of importance in the Forest, one at Bagden (now Savernake Lodge) the other at Braydon Hook Lodge. Braydon Hook is named in the perambulation of 1301. In a 17th-century sketch map there is a track called the Braden Way.

3H

Unnamed

SU 212 657 in OctantVI, attached plate code 06920. Intermediate or introgressed Native Oak, Q. petraea features dominant (QSP). Girth 7metres

This unnamed oak is in the narrow band of the forest on the west of the A346 road. The main trunk is slanting at 40°, with many big fallen boughs, but the tree still is generally healthy. This oak probably existed before the main planting period of 1750. It is at the edge of the forest close to the boundary between the Selkley and Kinwardstone hundreds. It is strange that today both the eastern and western boundaries of Savernake Forest coincide with the old hundred boundary, the forest filling the southern bulge in the Selkley hundred.

The Original Queen Oak

Grid ref. SU 224 657 in Octant V, attached plate code 08809.Q. petraea (QP), on the west side of Postwives Walk. Girth of 6.5 metres.

This oak tree is in the position marked as the Queen Oak on present Ordnance Survey maps and the 1887, 6 inch to 1 mile, County Series Map. Lucy’s Marlborough Official Guide of 1922 states ‘The

Original Queen Oak

38 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Queen Oak, is still a handsome tree and fenced like her consort.’ The King (now gone) and Queen Oaks were at opposite ends of a glade. There are no remains of the fence and this old tree can no longer be called handsome. It is hollow, with no visible signs of pollarding and at about 450cms from the ground a fractured jagged stump is directed obliquely skywards like a pointing hand. A large part of the bole is dead and there is missing section (of about 280cms) consequent upon an avulsed bough now on the ground. This fallen bough is itself 400cms in girth. The irregular shape of the hollow trunk, with its avulsed edges makes itis hard to make an accurate measurement of the girth, which is about 6.5 metres.

A local Forestry Authority report of 1998 says that the Queen Oak with a girth, in 1997, of 6.3m, (probably the girth was calculated from a measurement of the diameter) may have been planted to celebrate the marriage of Jane Seymour to Henry VIII in 1536. Jane Seymour was the daughter of the Warden, Sir John Seymour. This planting date was made using a formula of John White’s that has since been found to overestimate the age of oak trees. On the other hand if an allowance is made for the missing section of the trunk the girth would probably be more than 9 metres and as such the Queen Oak could date back to Tudor times. The Queen Oak is not marked with the King Oak on the Estate Map of 1786 and it is probable that a graceful tree was thus named ‘Queen Oak’ in Victorian times.

‘Spider’ and Unnamed Oak

Grid reference SU 224 657 in Octant V, plate codes 08808 & 08810. Q. petraea (QP) & QSP. East of Postwives Walk in the region of the Queen Oak. Girths 6.45 & 6.6 metres

These are just two of the many large ancient oaks in this wild part of the forest with disappearing paths. Probably both are ancient pollards. The Spider Oak has a squat, black, hollow, broken trunk with central ascending boughs, and eight spreading low branches arching up slightly before descending towards the ground. The second divides at about 250cm into two huge boughs spiralling up to the canopy. The winding paths in this part of the forest point towards an ancient origin. This could be the location of the medieval ‘King’s Wood’

False Braydon Oak

Grid reference SU217 668 in Octant VII, attached plate code 08968.Quercus robur (QR). Girth of 6.2metres.

This oak tree, north of Great Lodge Drive, has a zig- zag trunk with many dead branches. Recently (and almost certainly wrongly) known as the ‘Braydon Oak’, possibly because the True Braydon Oak is much less easily accessible, surrounded as it is by hollies, brambles and hawthorns. By July 2000 this tree seemed to be succumbing to oak die-back disease.

‘New’ & ‘Old’ Spiral Oak

Grid references SU215 665 & 213 664, Octant VI, tag 08870 & 08871. QS and QR. Girth 5.95 and 6.15metres

Both Spiral Oaks have spiralled trunks, very similar to the trunks of sweet chestnuts. The New Spiral Oak is very well named. It is a hybrid oak just NE. of the SE. corner of the western Thornhill Nursery enclosure, with very strongly and strikingly spiralled attached boughs and similarly strongly spiralled fallen branches.

The Old Spiral Oak, the one shown on an old map, is a pedunculate oak SW. of the preceding, a squat old pollard with 2 huge spreading low limbs from 10ft. These limbs are still somewhat spiralled, but with jagged and stag-horn diebacks. The tree has an irregular coppice base measuring 830cms at Ift. One of these trees is adjacent to the line of the Roman road to Old Sarum.

The New Queen Oak

Grid reference SU223 655 in OctantV, attached plate code 08814. QSP. Girth 4.7metres

A ‘Queen Oak’ should be a handsome tree. The current staff at the Forest Enterprise Office, Postern Hill, asked Valentine Cope, a retired forest foreman who had worked in the forest for very many years, to confirm which tree is the Queen Oak. He identified the tree that he knew as the Queen Oak as the tall tree above with the long straight trunk. This oak has a very long straight fissure (caused by lightning?) on the eastern aspect, with markedly raised healing lips on either side. It is in a cleared area 30 or so metres west of Post Wives Walk and south of the position of the Queen Oak marked on O/S maps. We have called this handsome tree the ‘New Queen Oak.’ With a girth of 4.7 metres it is not large enough to date from 1536.

Maybe this ‘New Queen Oak’ will continue to flourish and be the oak tree used to commemorate Jane Seymour and Savernake in Tudor Times. Jane Seymour, daughter of a Hereditary Warden, was the mother of Edward VI, and the Queen who was buried next to Henry VIII. The remains of the King Oak were

SAVERNAKE FOREST OAKS

The New Queen Oak

removed 55 years ago and the Duke’s Vaunt, the pride of the Duke of Somerset who was the Protector of the Realm and the brother of Jane Seymour, will not survive much longer. See ‘Subsequent Additions’ heading for another tree associated with the Tudors.

Amity Oak Grid reference SU232 676 in Octant I, attached plate code 01920. Quercus petraea. Girth of 4.35metres.

This oak is drawn as a large tree and named ‘Emity Oak’ on a map of 1608 dealing with a Preshute tithe dispute in the north Savernake area. It is marked on a boundary line in the same position as the Amity Oak on modern Ordnance Survey maps. On the Savernake Forest Estate Map of 1786 this tree is labelled as the ‘Emety Oak’. The difference in names of this tree and Duke’s Vaunt (Font) could have been due to the draughtsmen not being able to understand the local dialect. Another local name of ‘Ashlade’ has remained. In 1603 the oak was on the eastern edge of Ashlade Coppice, in 1786 in Ashlade and now it is on the eastern side of Ashlade Firs.

At some time in the past the old oak has been replaced with a new one. Despite the small girth in comparison with many other Savernake oaks it is

39

marked on the old Ordnance Survey map of 1880 and new maps alike because of its position and it retains the name ‘Amity Oak.’ It is the most slender of the named oaks, and one of the tallest with a long straight unbranched trunk then two very narrowly angled vertically ascending boughs, one of which now appears to be dead. It is situated at the junction of Sawpit and Amity Drive on the side of the old forest pale bank of 1870. It marks the intersection of the boundaries of the three civil parishes of Little Bedwyn, Mildenhall and Savernake. Before 1935, when the parishes of North and South Savernake were combined to form Savernake parish, it was at the junction of the parishes of Little Bedwyn, Mildenhall and South Savernake with Brimslade and Cadley. Normally civil parishes are based on the older church boundaries. Savernake parish is unusual in that it does not have an ancient parish church. It was not until 1854 that Christ Church, Cadley, was built by the Marchioness of Ailesbury. From the size and shape of the tree, the Amity Oak is about 250 years old. Maybe there was a ceremonial planting by the hereditary Warden of Savernake Forest, with representatives from Mildenhall, Little Bedwyn and possibly Preshute parish churches to define the edge of the forest and to encourage goodwill between all parties.

40 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

FUTURE GIANTS AND EXISTING GREAT GIRTH VETERANS

Many of the impressive oaks with vertical straight trunks, minimal or no die-back, few broken branches, and good height or symmetrical spreads had girths only in the 3-5 m bands. Good spreading oaks were noted in some of the open fields around Eight Walks, and Crabtree Common, and the great grassy avenue at Holt Pound (Octant IIT) where there is now also the greatest-girth beech in the Forest (nearly 7m) at the Warren Farm end. The biggest concentration of fine semi-open grown oaks is to be found in the public amenity areas at the periphery of Octant VI, around the grassy car parks, barbecue and picnic sites. West of the Braydon Oaks there is a quiet and less accessible semi-open oak grove with big spreading trees. Around the site of the long dead King Oak and around Postwives Walk (Octant V) amongst ancient veterans in this wild part of the forest, there are a few huge younger naturally regenerated oaks of good height, often with mossy fluted bases which are heading for great girths 300 years hence.

No great-girth oaks were found in Octants I & II, but gnarled veterans, often pollarded or sometimes coppiced in the distant past (occasionally both) were scattered in a variety of habitats over the rest of the Forest. Table 1 shows that all the native oak taxa occur in all three important bands. There are both species, their hybrid, and apparently introgressed intermediates all found in appropriate proportions in the 4-5, 5-6 or 6-7m girth ranges.

Some oaks of varied girths at 150cm had obviously enlarged bases, even if these were very irregular, or partly concealed by burrs, fallen branches, brambles or hollies. Some of these seemed to be ancient coppice stools, although others could have been stumps cut back to keep paths clear in past years, especially when lopsided. These ancient coppice bases were less common on oaks than on sweet chestnut trees, but the oaks of any size could have such enlarged bases. ‘Table 2 shows that 7 of the 33 great-girth or named oaks had probably been used for decades or centuries before the currently measured main trunk was allowed to develop. Table 2 also illustrates the taxonomic variety and scatter in the Forest of the named and greatest- girth oaks. They include broken and hollow, pollarded and squat, nodular and burred, stag-horn, as well as solid and lofty trees.

EPIPHYTES

There are 4 main categories of vascular plant epiphyte on the Savernake oaks, not altogether typical (even allowing for occasionals) of the lists reported from other oakwoods (compare Morris & Perring 1974, Rose 1974):

1 Ferns

Five species were noted. Polypody, intermediate polypody, broad buckler-fern, bracken and the male- fern. The first three produced spores in situ, and intermediate polypody was probably the commonest vascular epiphyte to colonize old oaks in Savernake Forest, and certainly the plant with the most dense

Table 1. Girths of the Largest Savernake Oaks at 150cms above ground, according to Taxonomic

Designations, (1999)

GIRTH cms OR QSR Qs QSP 350-499 19 10 15 10 500-599 10 7 17 12 600-699 5 2 8 5 700-799 1 3 2 800-899 1

900-999 1

1000-1099 2

TOTAL 35 21 45 29

% of Total 21% 13% 27% 17.5%

QP Q TOTAL unclassified 17 5 715 5 2 54 7 27 6 1 1 2 29 7 166

17.5% 4%

SAVERNAKE FOREST OAKS

4]

Table 2. List of 33 of the largest girth and/or named Savernake Oaks found in 1999. Girth in cms measured at 150cms from the ground. Bracketed numbers show basal or ancient coppice-ring circumference in cms measured at lft. Petiole lengths

shown in mm.

Name or Situation Tag No. Oct Map ref. Taxa Petiole Girth Coppice SU lengths,mm cms _ girth min max Amity Oak 01920 I 232 676 QP 10 20 435 S of Birch Copse 08982 Ill 237 659 QP 20 30 685 Duke’s Vaunt 08990 Il 238 665 @SR- 5 10 890 King of Limbs 09246 III 242 660 Qs i 14 1030 E of 12 O’clock Drive 08769 IV 225 659 OSE 10 20 615 Ayers Oak 08772 IV 226 659 QP 5 20 610 New Queen Oak 08814 Vv 223 655 QSP 7 15 470 Spider Oak 08808 Vv 224 657 QP 10 16 645 Turkey Oak 08828 Vv 225 655 QC 530 Original Queen Oak 08809 Vv 224 657 QP 10 25 650 N.E. of Queen Oak 08810 Vv 225 657 QSP 4 12 660 Slingsby Oak 08830 Vv 224 654 @SPS 25 25 710 W of 12 O’clock Drive Vv 224 659 QS 5 20 740 Field boundary near 8 walks 08758 Vv 223 663 Qs 9 12 645 W of 12 O’clock Drive 08777 Vv 224 659 QS 10 30 670 (900) New Spiral Oak 08870 VI 215 665 Qs 8 16 595 Old Spiral Oak 08871 VI 213 664 QR 3 5 615 Big Belly Oak 06924 VI 213 657 QS 5 15 1080 (1400) W of A346 Rd 06920 VI 212 657 @SP, 3 115 700 S of Gt Lodge Drive 08043 VI 215 667 QS 5 10 610 (1200) Cathedral Oak 09500 VII 205 680 QR 2 3 995 NW edge of White Rd 08191 VII 206 678 Qs 4 9 615 N of White Rd, E of A346 08565 VII 202 674 QR 4 if 620 (800) White Rd\Grey Rd intersect 08168 VII 203 675 @SP 6 18 675 (770) False Braydon Oak 08968 VII 217 668 QR 1 2 620 True Braydon Oak 08958 VII 216 670 QSR_ 2 5 710 Near Braydon Oaks 08965 VII 217 670 QSR 4 6 605 (840) S.W. of Grey Rd 08724 VII 204 674 @SP 2 fl 655 (810) N of Gt Lodge Drive 08977 VII 212 667 QS 5 7 760 (950) N of Gt Lodge Drive 08883 VII 214 670 QP 10 25 655 A4\Grand Av angle 08941 VIII =. 211 682 Qs 6 11 670 Roman Rd Oak 09209 VIII = =217 677 QR 0 2 570 (880) Surveyed Oak 08947 VIII 217 676 Qs 9 13 740

and extensive coverage on a number of individual trees. Polypody also spreads by rhizomes along the branches of oaks, and can tolerate deep shading in the summer months.

2. Trees, Woody plants, Scramblers

and Climbers The following tree seedlings and saplings were noted as epiphytes, in order of commonness: elder, ash, sycamore, rowan, hawthorn, beech, Norway maple, hazel, oak and holly. One elder had flowered and fruited as an epiphyte.

Bramble was the most common flowering plant epiphyte, and 3 bramble plants had in 1999 produced

blackberries from their perches in the crotches or fractured parts of old oaks, usually ancient pollards. Raspberry plants, sometimes with raspberries, were seen on 11 oaks, seedlings and young plants of honeysuckle on 4, and young epiphytic plants of dog- rose, redcurrant, gooseberry and ivy (seedling) on at least one oak each.

3. Herbaceous Plants

The special case of the Big-Belly Oak aside, there were few weedy species noted. The commonest herbaceous epiphyte was wood-sorrel, which (unlike bramble and raspberry preceding) was almost entirely confined to lower parts of the trunk, most obvious in

42 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

A Fern Epiphyte on an Oak Tree

May. Next, found even in the dark parts of the forest, followed (surprisingly) bifid hemp-nettle, then common hemp-nettle, the hemp-nettles often seen in May as seedlings at various heights. The stinging- nettle was the next most common epiphyte, but seldom seen in flower. Cleavers, herb-robert and short-fruited willowherb were each noted on 4 or more oaks. Finally there were at least 10 other herbaceous species found as occasional on oak trunks or (rarely) higher up.

4. Grasses

Cocksfoot and rough meadow-grass were occasional as epiphytes; annual meadow-grass and wood false brome also occurred.

Of the total of 43 vascular plant species recorded on oak trunks and branches by May 2000, only 11 were seen to have produced spores, fruits or seeds in situ. Most individuals were young plants doomed to die without reproducing.

In January 2000 the Big-belly Oak at the side of the A346 road, carried 2 fern species (polypody and broad buckler-fern), 10 herbaceous species (wood avens, 5 plants; wood dock, 1 plant: herb-robert, 11 plants; stinging nettle, 4; short-fruited willowherb, 5; garlic mustard, 2; ground-ivy, 2; ivy-leaved speedwell, 3; cleavers, 24 mainly seedlings; dandelion, 8 and bramble, 6 plants). Only the 2 fern species, one herb- robert, 2 willowherbs and one dandelion and one

bramble had produced spores or seeds as epiphytes. There were also 3 grass species growing on the Big- belly Oak (cocksfoot, 3 plants; rough meadow-grass, 4; annual meadow-grass, 1;). This veteran oak has a bizarrely shaped conical, almost pyramidal, trunk and it protrudes into a busy main road, is not in shade and is subject to much air turbulence which carries road-verge seeds into its trunk crevices.

Most of the other old oaks were much more sheltered, shaded, out of range of roadside weed seeds, and in much quieter less turbulent situations, and vascular epiphytes were only seen on a few of these. However there nearly always were non-vascular epiphytes on their trunks and branches, bryophytes (the mosses and liverworts), lichens and algae. Savernake Forest has a list of at least 156 lichen taxa, including a number of rare species, with most epiphytic lichens largely confined to old oaks. Algal and moss species are largely responsible for the green trunks of most beeches and oaks in Savernake Forest, several moss species in particular giving beautiful rich green effects when the bases of the oaks are fluted. 104 bryophyte taxa are recorded for the Savernake Forest SSSI, a number of which are epiphytic: 2 mosses and 2 liverwort species found are known to be associated almost exclusively with old trees in ancient woodlands (for detail, see Stern 1996 and SSSI 1981).

OAK DIE-BACK DISEASE

This condition has been known since the 1920s (Gibbs & Greig, 1997). There has, however been recent concern that a more serious form of oak die- back has been progressing in parts of Britain and Continental Europe over the past few years. Complex interactions are almost certainly involved (Gibbs, 1999). Oaks stressed by a great variety of adversities, for instance drought, can lose the use of one or more branches which are ‘sealed off’ and die. The tree subsequently puts out newer shoots and recovers, even by forming a new and more contracted canopy than hitherto. The dead branches are then left protruding, the staghorn oak condition. In the recent more fulminant outbreaks, single or small irregular groups of oaks develop yellowing of the foliage for no good observable reason (e.g. no drought, mildew, no insect defoliations or aphid invasions in the previous year). Next year if the tree has only a 15% coverage of the foliage, some of it yellowing, death will probably ensue even if some new shoots have been put out.

SAVERNAKE FOREST OAKS

The condition could be in three stages: general debilitating factors, secondary specific pathogenic attacks, and thirdly root pathogens. The usually non- lethal root pathogens (e.g. Phytophthora) seem to increase in virulence (Gibbs, 1999).

There have always been staghorn oaks in Savernake Forest. Some - or most - recover to renew their health and vigour. The Old Spiral Oak (QR) in 1999 looked in such a poor state that we hardly thought it worth measuring. However by June 2000 it had renewed its foliage in a good recovery. By contrast, the False Braydon Oak (QR) had only a 10% coverage of (mostly) yellowing foliage in July 2000, and looks to be on its way out owing to the onset of the new wave of oak die-back disease. Until recently, Savernake Forest had few trees affected, but earlier this year, we were shown, by the Earl of Cardigan, some newly diseased large oaks in the Deer Park behind Tottenham House and in the Savernake Lodge areas. In July 2000, as well as the involvement of the False Braydon Oak, there were individual trees and small irregular groups showing signs of the Oak Die-back Disease noted along-side Grey Road and by Twelve O’clock Drive south of Eight Walks. Some young plantation oaks, all pure Q. robur, were affected; a rather ominous sign. So far nearly all affected oaks were noted to be Q. robur; or the hybrid with the Q. robur features predominant. Q. petraea seems so far to be much more resistant, and this also applies to the hybrid where the Q. petraea features predominate. If this observation is sustained as generally accurate, this becomes an additional reason for planting Savernake acorns in any new plantations, rather than bringing in Q. robur acorns or seedlings from elsewhere, as seems to have happened in the past. The naturally regenerating young oaks, with substantial proportions of Q. petraea and Q. rosacea (Q. petraea features very much in evidence), are so far not affected by oak die-back disease.

SUBSEQUENT ADDITIONS NOTED IN 2000

One of us ((EO) combed the Forest again in the spring, summer and autumn of 2000. At least 175 more oaks were found in the 350 - 499cm girth band (Table 1), and 65 more in the 500-599 band. There are 18 more to add to the 600 — 699 range, and 3 more to the 700-799cm range. Coppice girths at lft ranged from 610 — 1100cms for fourteen of these additional oaks. For the heartland of the forest the

43

taxonomic proportions are comparable to those shown on Table 1.

On Long Harry Drive, near to the intersection with Church Walk at (SU208 675) was re-found the oak marked on the 1786 Estate map, once designated “Long Harry’ and best now called ‘Big Harry’ (or ‘Old Harry’). This extraordinary tree has a girth of 700cm at 5ft, but springs from the side of a gigantic hemispheroid coppice base 5-6ft high, indicating a pre-Tudor origin (at least). This make plausible the supposed link with hunting or courting forays by King Henry VII. The oak is probably a hybrid, but inclines towards Q. robur. A second veteran, nicknamed ‘the Hornet Oak’ was occupied by a hornets’ nest with its main exit at face level and the dead heartwood being hollowed by these insects. One of the fine named oaks on the periphery of the forest is in the grounds of St Katharine’s School. It is the majestic ‘St Katharine Oak, with a girth of 650cm and is pure Q. robur.

DISCUSSION AND CONCLUSIONS

Grose in ‘The Flora of Wiltshire, 1957’ stated that no woodlands in Wiltshire are dominated by the Durmast Oak, Quercus petraea. The tree does occur in Savernake Forest but is scarce. The dominant oak is Quercus robur. In the same book there is a report from John Wildash that earlier notices of Quercus petraea for the Forest were found to be Quercus petraea x robur. Thus nearly 50 years ago there was a query about the taxa of Savernake Forest oaks. Although Grose listed a group of old trees near Thornhill Pond in the Forest as Q. x rosacea, this like other possible hybrid records of oaks was forgotten. The name Q. x rosacea was resurrected officially by botanical taxonomists in the last 11 years.

The recent British Floras (Rich & Jermy 1998, Stace 1997) consider that Q. x rosacea, in parts of Britain at least could be common. Stace refers to Q. x rosacea being ‘.....occasionally commoner than either parent’. The slightly older flora (CTM 1989) suggests Q. x rosacea to be “.....rather frequent where the parents occur together ..... most common in Scotland & N. England; and in N. Ireland (Antrim). Variably fertile due to introgressive hybridization.’ Rushton, writing in Rich & Jermy (1998) encapsulates the problems thus ‘Not only are Q. petraea & Q. robur variable in themselves, but have also widely introgressed producing variable, fertile hybrids. This results in inconsistent identification of both parents

44 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

and their hybrid Q. x rosacea by different botanists, as the limits of the species are a matter of opinion.’ Note the five words from ‘introgressed’ to the end of the first sentence, implying sustained hybridization between Q. x rosacea itself and its parents (back- crossing), and the same process continuing with successive generations.

At the other end of the scale, Aas (1991) in Germany and Boratynski et al (1997) in Poland give amore restricted picture, with only 3% of intermediate oaks in the Gdansk State Forests being identified as Q.x rosacea. In between Belous (1997) in the Ukraine, Brookes & Wigston (1997) in England, French & Murphy (1994) in Cornwall, Minihan & Rushton (1984) in N. Ireland, Mihailescu & Ciobanu (1990) and Stanescu & Sofletea (1992) in Romania, and Wigston (1974) on British oaks all tend to leave the question of commonness of Q. x rosacea open. The results from our survey carried out in 1999 give Q. x rosacea as the dominant natural tree for Savernake Forest. It is certainly not Q. robur, unless one’s attention is confined to the plantation oaks derived from imported acorns.

Rich & Jermy (1998) illustrate in full detail the ranges of leaf variation between Q. petraea, Q. x rosacea and Q. robur; and Stanescu & Sofletea (1992) have eight drawings of Q. x rosacea leaves from the Bejan Forest in Romania. These illustrations of Q. x rosacea leaves are remarkably similar to the leaves from the majority of naturally regenerated oaks in the Savernake Forest. The most important of the nine characters made for identification were those using microscopy. Leaf pubescence features of the three oak taxa are detailed by Rich & Jermy. These again match our findings, but we have also added more types of hair, and drawn attention in particular to the barely visible half-size sublaminar stellate hairs with a spread of less than 0.1mm found on many Q. x rosacea Savernake oaks, compared with the 0.2mm spread of Q. petraea sublaminar hairs. Hairs on the undersides of the leaves which are found alongside the veins and in the vein angles are more diverse in type and larger than the sublaminar stellate hairs for both Q. petraea and Q. x rosacea Savernake oaks (see microscopy section and Oliver 2000).

A crude guide for identification is petiole lengths. All Q. robur Savernake oaks had most petiole lengths of less than 7mm, and 86% had all of their petioles between 0 —7mm. For Q. petraea 70% of the trees had their median petiole lengths in the 13 — 25mm range, but the spread was very variable. As for Q. x rosacea there was 84% with median petiole lengths between 7-13mm.

From the preceding discussion, it is clear that Savernake trees here designated QS are unequivocally Q. x rosacea. Those semi-intermediates QSP and QSR are probably introgressed back-crossed hybrids, although some botanists say this could only be proved with DNA analyses, and they could simply represent variability within the three taxa. Whatever the taxonomic boundaries, all three main native Savernake Oak taxa show large ranges of different genetic variations, not necessarily only related to hybridization and back-crossing.

The endemic Savernake Cluster Oak (Henry 1917) is an extreme variant of Q. robur, for instance. Although fertile, it is hard to see how the Cluster Oak could be favoured by natural selection; but it is very probable that many of the other native oak variants in the Forest, especially introgressed hybrids, represent a rich stock of oak genetic diversity well adapted to cope with the different geological, climatic and edaphic conditions found over the thousands of years in the Savernake area. There could be strains useful for trials elsewhere — for instance, Q. x rosacea is one of the hardwood taxa suggested for farm timber resources in New Zealand ( Ledgard & Giller 1998) and is starting to be tried in regeneration experiments in the Scottish uplands (Humphrey & Swaine 1997). Q. x rosacea 1s being considered in the Zlatna forest- decline trials in those parts of Romania suffering pollution by sulphur dioxide and heavy metals, including zinc and cadmium, (Mihailescu & Ciobanu 1990). Genetically, Savernake Forest could be classified as one of the great mixed forests as categorized by Wigston(1974). Whatever the octant or underlying geology, whatever the conditions in Savernake Forest, wherever the native oaks are regenerating naturally they are producing this great range of phenotypes, seen in the youngest saplings to the most ancient veterans.

Of the four non-native oak species, Q. rubra, the American red oak would seem to have the potential for spread by seed (Stace 1997) in future years, but few acorns have been produced to date.

Savernake Forest remains a patchwork quilt of varied habitats with old and new mixed plantations, plantations of conifer or oak, and mixed deciduous forest with remote undisturbed areas deeply shaded in summer by either beeches, oaks, sweet chestnuts or common (suckering) limes. There are naturally regenerated oak thickets, groves and glades, numerous ancient banks and boundaries, and open fields with small clumps of old oaks or beeches. Fine trees surround the amenity areas. In addition to the well- known beeches along Grand Avenue there are other

SAVERNAKE FOREST OAKS

grassy trackside or roadside great limes, horse chestnuts, sweet chestnuts, beeches or oaks. Ancient oaks, uncommon or rare in most of the rest of Europe, have been preserved successfully in all the preceding habitats, but some others have not survived beech shading and a few have succumbed to Douglas firs or other conifers, especially in the Birch Copse area.

It would seem that some of the histories and ages attached to certain veteran oaks have been exaggerated, embroidered or perhaps even fabricated by the Victorians or others. Although girths increase with age, measurements are subject to wide margins of error on irregular, burred, nodular, partly rotted or split trunks, and age estimates simply based on girths are full of pitfalls. None of the ancient oaks are likely to go back as far as the early Saxon Kings. Giant, veteran oaks are normally found in parkland, and Savernake Forest is very unusual in having large old oaks in a forest. It will be interesting to see what happens to girth records for the biggest oak trees, as existing records list Q. robur or Q. petraea, and many of our giants are clearly the hybrid between these two species. Frequently veteran oaks are found on old boundaries. Amity Oak and the Duke’s Vaunt are on parish and hundred boundaries, and a few of the large oaks are on the line of the Roman road to Old Sarum, which might have been a boundary line. The others are scattered about, mainly in the western part of the Forest and not on any known boundary.

Some of the great existing veteran oaks need protection from shading, but possibly even more important is the marking of significant trees. The essential practical measure is metal labelling, or notices in situ. Repeatedly we found that historic trees such as the Great Beech of Savernake, the Braydon Oak, the Spiral Oak and others have been lost, or their names misapplied to the wrong trees. With overgrown or lost paths and erratic and inconsistent and vague map records, even the well known veterans such as Duke’s Vaunt, King of Limbs, Queen Oak and Amity Oak could be hard to find, or (as in the case of the last two) hard to name the correct tree with certainty. The Cathedral Oak had a helpful and interesting notice, and many of the big trees flanking the roads, avenues, picnic, camping and barbecue areas already have galvanized iron, zinc or lead plates with number codes, presumably so that any safety-precaution tree surgery can be quickly and efficiently implemented. In 2000, thanks to help and permission from Bill Ayers of Forest Enterprise, we have extended this

45

system to the important trees in the deeper parts of the forest. The tagging has included the oaks from Tables 1 & 2 and those enumerated under the ‘Subsequent Additions noted in 2000’ subheading. Coded tags have also been attached to 22 other tree species of exceptional size, significance or longevity. A detailed tree-by-tree report is in preparation for the Wiltshire Botanical Society, which will relate the plate codes to map references, site or other details, special features, and girths measured in 1999 or 2000.

We would encourage preservation of the rich genetic diversity of the native Savernake Oaks by the use of Savernake acorns in all future plantations. The recent, patchy outbreak of Oak Die-back Disease in Q. robur emphasises the importance of genetic diversity. There should not be a puritanical absolutist conservationist policy with fanatical elimination of the limes, hollies, yews, Norway maples, sycamores, red oaks or non-native conifers. The patchwork quilt variety in this forest 1s one of its main glories. However the collected acorns should be labelled as Q. x rosacea, Q. robur, Q. petraea, or as semi-intermediates, or as from named trees, so that progress of the progeny could be studied more specifically in future decades and centuries. A relatively uncomplicated research project would be to record acorn production by Q. x rosacea oaks in comparison with acorn production by pure Q. robur and Q. petraea, and acorn production by veteran trees.

Oaks, like yews, marry biology to history and myth. The Savernake oaks can be linked with English social and landscape history. The genetic mixing of the two oak species native to our islands and Europe may be as great or greater than in any other of the European oak forests and with many more great girth veteran trees. We hope this paper provides a firmer baseline for future botanical and historical studies than has been possible in the past.

ACKNOWLEDGEMENTS

Our thanks are due to Allen Coombes (the BSBI Oak Referee) for authentication of specimens, to Brian Rushton of the University of Ulster (Coleraine) for answering queries on some genetic aspects and to David Rose of the Disease & Diagnostic Advisory Service for information and a guided explanation on oak die-back disease. Locally we appreciate help from the Earl of Cardigan, Hereditary Warden of Savernake Forest, from Patrick Cashman of English Nature, and Bill Ayers and Andy Glover of Forest Enterprise.

46 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

References

AAS, G., 1991, ‘Crossing experiments with Pedunculate & Sessile Oak’, Allgemeine Forst und Jagdzeitung 162(8) 141-145. In German, Eng. summary.

ADAMS, W. M., 1904, Sylvan Savernake & its Story.

AILESBURY, Marquess of, 1962, A History of Savernake Forest. Woodward, Devizes.

BELOUS, V.I., 1972, ‘Hybrid populations of Quercus robur & Q. petraea in the Ukraine’, Lesovedenie, 6 37-46. In Russian, Eng. summary.

BORATYNSKI, A., et al, 1997, ‘Pedunculate, Sessile and Hybrid (Quercus x rosacea) Oaks in selected seed stands in the R.D.S.F. Gdansk area’, Sylwan. 141(5), 41-49. In Polish, Eng. summary.

BRENTNALL, H. C., 1941, ‘The Metes & Bounds of Savernake Forest’. WA&NHM, 49, 391-434.

BROOKES, P. C. & WIGSTON, D. L., 1979, ‘Variation in the morphological & chemical characteristics of acorns from populations of Quercus petraea, Q. robur and their hybrids’, Watsonia, 12(4), 315-324.

CARDIGAN, Earl of, 1949, The Wardens of Savernake Forest. Routledge. London.

CLAPHAM, A. R., TUTIN, T. G. and MOORE, D. M., 1989, Flora of the British Isles. Cambridge University Press, Cambridge.

FOREST ENTERPRISE, 1998, Named Trees of Savernake Forest. (local)

FRENCH, C. N., & MURPHY,R. J., 1994, Checklist of the Flowering Plants & Ferns of Cornwall and the Isles of Scilly. Cornish Biological Records Unit, Univ. of Exeter.

GIBBS, J. N., and GREIG, B. J. W., 1997, Biotic and Abiotic Factors affecting the Dying Back of Pedunculate Oak, Quercus robur L.

GIBBS, J. N., 1999, Information Note, Forest Research, “Dieback of Pendunculate Oak’, Forestry Commission.

GROSE, D., 1957 (reprinted 1979), The Flora of Wiltshire, Wilts Archaeological & Natural History Society. E.P. Publishing Ltd. Wakefield.

HADFIELD, M., 1974, ‘The Oak and its Legends’, 123- 129 from Morris & Perring (below).

HENRY, A., “The Cluster Oak of Savernake Forest’, The Gardeners’ Chronicle, 27.1.1917.

HUMPHREY, J. W. & SWAINE, M. D., 1997, ‘Factors affecting the natural regeneration of Quercus in Scottish

oakwoods’, Journal of Applied Ecology, 34(3), 577- 584.

JEFFERIES, R., The Hills & the Vale — Marlborough Forest. Oxford University Press.

LEDGARD, N., & GILLER, M., 1998, ‘Deciduous hardwood species: early silvicultural options for growing timber on farms’, New Zealand Forestry 42(4), 16-21.

L.U.B.G.M., 1922, Lucy’s Official Borough Guide to Marlborough.

MIHAILESCU, A. & CIOBANU, C., 1990, ‘Industrial pollution of forests and forest soils in the Zlatna region’, Revista Padurilor, 105, 3-4. In Romanian, Eng. summary.

MINIHAN, V. B. & RUSHTON, B. S., 1984, ‘The taxonomic status of oaks (Quercus spp) in Breen Wood, Co. Antrim, Northern Ireland’, Watsonia, 15, 27-32.

MORRIS, M. G., & PERRING, F. H., (eds.), 1974, The Briush Oak. Published for the BSBI by E.W. Classey Ltd, Faringdon.

MURRAY, J. S., 1974, ‘The Fungal Pathogens of Oak’, 235- 249, from Morris & Perring (above).

OLIVER, J. E., 2000, Quercus x rosacea in Savernake Forest. BSBI News, 84, 31 — 34.

RACKHAM, O., 1974, ‘The Oak in Historic Times’, 62- 79 from Morris & Perring (above).

RICH, T. C. G., & JERMY, A. C., 1998, 1998 plant Crib, BSBI London.

ROSE, F., 1974, ‘The Epiphytes of Oak’, 250-273 from Morris & Perring (above).

S.S.S.1., 1981, Notification of Savernake Forest under Sec. 28 of the Wildlife & Countryside Act 1981.

STACE,C., 1997, New Flora of the British Isles, Cambridge University Press, Cambridge.

STANESCU, V., & SOFLETEA, N., 1992, ‘Oak Hybrids in the Forest Bejan (Deva)’, Revista Padurilor, 107(2), 2-3. In Romanian, Eng. summary.

STERN, R. C., 1996, Savernake Forest S.S.S.I. Management Plan, Second Revision.

TROBI, (Tree Register of the British Isles) Founded 1988. Database of over 125,000 notable trees in Britain and Ireland. Wootton, Beds

WIGSTON, D. L., 1974, ‘Cytology & Genetics of Oaks’, 27-43 from Morris & Perring (above).

WILKINSON, M., 1982, Compilation of Field Surveys and Reports on Savernake Forest, 22.11.1982.

W.R.O.S.F.E.M., 1786, Wiltshire and Swindon Record Office, Savernake Forest Estate Map of 1786.

Wiltshire Archaeological & Natural History Magazine, vol. 94 (2001), pp. 4

7-55

Murder at Brookside Cottage: a Dark Deed in

North Wiltshire by Kay Taylor

This paper is a study of the oral and written accounts relating to the trial and conviction of gypsy Edward Buckland for the brutal murder of widow Judith Pearce on the Seagry—Sutton Benger border in 1820. It also questions the quality of the judicial system in late Georgian England.

A gruesome murder that occurred on the border of the small north Wiltshire village of Sutton Benger on 13 November 1820 had such an impact on the lives of the villagers that it was still a topic of conversation among their descendants more than one hundred and fifty years later. On moving to the village in 1978 the author gained the impression that the killing had happened within living memory of those recalling the event. It was only on being shown the victim’s grave in the parish churchyard that it became apparent that the accounts were not those of eyewitnesses, but had been handed down through the generations. This telescoping of time-scales has been observed, by historians and anthropologists, both in the informal tales and folklore of early modern England, and in the African system by which village ‘remembrancers’ transmitted a stable ‘official’ local tradition to succeeding generations.' It was, therefore, interesting to find the same process occurring in the twentieth century, in the heart of rural England, with the local population keen to initiate interested incomers into the secrets of an exceptional episode from their collective past. An ‘eye-witness’ account of the murder was preserved, until recently, by the late Dennis Selwood of Seagry. It transpired that the tale he told had been handed down to him when he was a boy, by his great-grandfather, who had, in turn, been given the details by his father. This ancestor had been a boy in Sutton Benger at the time of the murder, and apparently retained a vivid memory of the events of that fateful night and the subsequent trial.’ Recollections of some of the events are also to be

found in the pages of the nineteenth-century diarist and incumbent of Langley Burrell, Francis Kilvert but, as with Dennis Selwood’s version, these rely on the memories of elderly people many years after the event.’

The Selwood family tradition preserved the memory of the murder from the days of those that experienced the drama through five generations, with this and other oral versions outlasting a locally held written account. A newspaper cutting, reporting on the trial at the 1821 Salisbury Lent Assizes, had been circulated around Sutton Benger in the earlier part of the twentieth century. This cutting, which has long since disappeared, played a vital part in keeping alive memories of the murder, albeit in a distorted form, as the readers then recounted what they had read to others, the story changing slightly with each retelling. Sutton farmer John Lea used to hire out the cutting to interested villagers for one penny. The money raised contributed towards the upkeep of the memorial stone to the victim, which had been erected by public subscription in 1821.’ In his classic work on English local history W. G. Hoskins warned that, while oral evidence should not be dismissed altogether as a source of local history, it had to be subjected to rigorous checks.’ In this case the oral accounts of the murder given by the villagers differed from the contemporary written reports in some key areas. The main inaccuracies occurred, not so much in the details of the murder itself, as in the addition of an extra crime (burglary) and in the ultimate punishment of the alleged offender. The most common, but

5 Lee Crescent, Sutton Benger, Chippenham SN15 4SE, and University of the West of England

48 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

erroneous, version of the story was that a gypsy had brutally murdered an elderly widow and was apprehended the morning after, still in possession of goods stolen from her cottage. A group of villagers marched him off to the Bell Inn where he was tried by a local ‘court. Having been found guilty he was hanged on a hastily erected gallows on the neighbouring Bell Piece, which at that time was a vacant plot of land. Bell Piece retained a reputation for being haunted until a housing estate was built there in 1981.° Another popular misconception was that the gypsy was the last person to be publicly hanged ini Devizes market place. This myth can be discounted for two reasons: public executions at Devizes continued into the 1860s, and the gypsy was actually hanged at the county gaol at Fisherton Anger, near Salisbury.’

The murder obviously caused a great stir in Sutton Benger and the surrounding villages as the most exciting occurrence in living memory. It also caused considerable friction with the neighbouring village of Seagry, whose inhabitants claimed that the murder ‘belonged’ to them. Brookside Cottage, where the murder was committed, was situated on the Seagry side of the parish boundary. However, the victim’s relatives lived in Sutton Benger, and the Sutton folk were directly involved in the apprehension and conviction of the alleged perpetrator, so the villagers there felt that their claim was the greater. Although the victim was buried in the Sutton churchyard, the Rev Anketell still considered the murder to be a valid part of Seagry’s past, more than sixty years later, including it in his 1886 Collections for a History of Seagry.* Tronically, the boundary between the two parishes was moved with the coming of the M4 motorway in the early 1970s, and Brookside Cottage is now in Sutton Benger. D. R. Woolf observed that almost every rural community contained or abutted on a field, hill, river or ruin which it associated with a saint or local hero or with a memorable event’ and Brookside Cottage fulfilled this role for both villages. Kalvert noted, in 1875, that this lone house between Sutton and Seagry was known as Murder Cottage.!” The nickname continued to be used by local schoolchildren from both villages up to World War II, and one childhood pastime was to dare those brave enough to go up to the cottage, which they believed to be haunted.'' In fact the attractive stone cottage currently occupying the site was not the scene of the murder, having been built c.1850.The original cottage was a simple thatched wattle and daub building with a downstairs kitchen and buttery, and an upstairs sleeping area reached via a ladder. The cottage’s thin

lath wall construction had enabled the intruder to gain access and the occupants to effect their escape. After the murder this cottage lay derelict for nearly thirty years, as no-one wanted to live there, and it was eventually demolished.

The gypsy, Ted Buckland, was arrested the morning after the murder and taken before the local magistrate, Mr Coleman, at Langley Fitzurze (Kington Langley). From there he was taken to the county gaol at Fisherton Anger to await trial at the following Lent assizes. However his trial by the media started much sooner. Less than three weeks after his arrest, and some three months before the trial proper, the following poem appeared in the Bath Herald.'” The anonymous poet introduced the work thus:

Lines (in imitation of Crabbe) on Judith Pearce, who was most inhumanly murdered by a gipsey, at Seagry, in this neighbourhood, 1n the month of November last. A lonely cottage stands beside the way,

A white thatched cot, with honey-suckles gay;

There JUDITH PEARCE, a widow lived alone,

By a rough quarry of blue-coloured stone;

Where lurked a wretch, of Egypt’s wandering race,

A wretch forlorn, without a mark of grace,

Whom ruffians left, for such a rogue was he,

That even the vilest shunned his company;

Dark was his face but darker still his mind

To pity, and to every tender feeling blind.

He had no friends, nor knew the joys of home,

But muttering, through the dews of night would roam, Brooding on fancied wrongs, with secret pride,

On words, or looks, or benefits denied.

Round his gaunt side a rope for girdle swung,

From which a light, short-handled hatchet hung;

A tattered garment did the village fright,

A coat by day, a blanket all the night,

Which round his neck a butcher’s skewer confin’d, Fit fastening such a filthy dress to bind.

JUDITH had often a kind warning given,

How far his ways were from the ways of Heaven;

And once too, JUDITH (which would kindle strife

In greater persons) asked him — ‘Where’s your wife?’ Once fire denied — a common courtesy;

Yet there seemed danger in his quick black eye;

And so there was, for as she lay in bed,

At night the thatch was blazing o’er her head,

And EDWARD BUCKLAND, so the villain call,

Was met in haste, close to the village wall;

And if as on some villainy he mused,

The evening salutation he refused;

Suspected, taken, he escapes at last,

And all supposed the danger now was past —

When JUDITH’S brother, in the dead of night, Heard his grand-niece who shook with cold and fright, Tell how she ‘scaped the murderer’s hand by flight; “Wake! Wake! She’s murdered!’ was the frightful cry;

MURDER AT BROOKSIDE COTTAGE

High Street

To Chippenham

To Langley

Burrell

Map of Sutton Benger showing position of Brookside Cottage

‘I heard the blow! I almost saw her die.’

They found her lying on the garden mould,

Mangled with dreadful wounds, quite dead and cold, A sight to shock the weak, and almost scare the bold.

It would have been difficult to prevent potential jury members from reading such a blatantly biased account of Buckland’s character and assumption of his guilt, the poet’s verdict pre-empting the eventual outcome of the court case. Whether or not Buckland actually

49

Brookside Cottage

All Saints Church

To Christian Malford

Sutton Lane

committed either the earlier arson or the murder is now, however, a matter of historical speculation. Nevertheless, the publication of such a defamatory account in a local newspaper, prior to the trial, must raise questions over the fairness of his trial and subsequent conviction.

The judicial records for the Easter 1821 Quarter Sessions are unfortunately missing from the court rolls held at the Wiltshire and Swindon Record Office, so

50 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

much of the following account has been taken from contemporary newspaper reports and a pamphlet issued in the wake of the execution.'* Journalists from the Devizes and Wiltshire Gazette and the Salisbury and Winchester Journal were present in court throughout the trial, on 15 March, and reported the proceedings at great length. The Salisbury journalist recorded that the presiding Justice, Mr Holroyd, had been ‘sorry to observe that the calendar was uncommonly heavy, both in regard to the number of prisoners and the nature of the crimes laid to their charge’. Of the ninety prisoners presented at the Lent Assizes four were indicted on murder charges, and there were sixteen others to be tried for capital offences, such as highway robbery, forgery and horse and sheep stealing.

As indicated in the above poem, the gypsy at the centre of the Pearce murder, Ted Buckland, alias Buckley, was a well-known, although not particularly well-liked, local character, who had lived in the area around Sutton Benger, Seagry and Christian Malford for over twenty years. At the end of the eighteenth century he had been one of the band of gypsies that regularly camped in tents opposite the gate to Sand Furlong in Sutton Lane,'? which was then a green lane and not a metalled road. He had apparently acquired such an evil reputation that his own tribe ‘kicked him out.’? After leaving the gypsies he lived a solitary existence, sleeping rough, doing odd jobs, and begging from house to house, mainly for salt, tinder and needles. He seldom travelled far, and occasionally earned money by selling matches. His distinctive appearance enhanced his reputation as an evil little man, with his swarthy complexion, black hair, bushy grey whiskers and long beard. In 1820 he was 66 years old, only five feet three inches tall, and described as being darker than normal for a gypsy. He was a common sight wandering around the area wrapped in a dirty old blanket which, as the anonymous poet observed, he tied round his waist with string and pinned across his chest with a meat skewer.'° In summer he would go about barefoot and sleep, wrapped in his blanket, in the quarries to the north of the village,'’ off the Seagry Road. In winter he sometimes managed to persuade Mr R. Hull to let him sleep in his barn, and Sutton farmer John Russ, who had known him for more than twenty years, occasionally took pity on him and gave him straw to lie on.

Old Ted was the obvious suspect when the murder was discovered as his relationship with the pious victim had been tempestuous in the past. Judith (Judy) Pearce was a God-fearing 58-year-old widow, who

lived in an isolated cottage by the brook at the foot of Seagry Hill. She worked at Church Farm in Seagry, and out of Christian charity was known to share a crust with Ted Buckland when he came begging. This was, however, usually at the cost of a lecture on mending his evil ways. In the spring of 1820 Ted had demanded more than food, wanting to warm himself by Mrs Pearce’s fire, but his menacing manner alarmed her so much that she slammed the door in his face. She barred and bolted herself in until he went away. That same night the thatched roof of her cottage was set on fire, but was quickly brought under control by some passers-by, with the help of a heavy rainstorm. Buckland was apprehended for the crime but he somehow got away and was not seen in the area for about six months. On his return many believed (as they asserted with the benefit of hindsight during the trial) that he would want his revenge on Judy Pearce for having accused him of the arson attack.'* By mid-November Buckland was back in North Wiltshire, and begging for clothes from a Mrs Ann Flower. As he had called on that charitable lady on a Sunday 12 November she agreed to let him have some of her husband’s old clothes if he returned the following day. On the Monday she let him have an old great-coat and breeches, and these clothes were to play an important part in Buckland’s trial and conviction.

At 8 o’clock on the morning of Tuesday 14 November Buckland was found at the side of Sutton Lane, cooking his breakfast over a fire. He was promptly arrested by the Sutton constable Richard Ellery, with the help of a group of villagers. These people had been roused about three o’clock in the morning by the excitement of the night’s events and were keen to see the culprit quickly brought to book. Ignoring the oddity of anyone who had just committed a particularly brutal murder staying in the vicinity and calmly cooking breakfast, they marched Buckland off to see Mr Coleman, the magistrate, at Langley Fitzurze.'!° There are a number of ways in which Buckland’s behaviour that morning could be interpreted. It is conceivable that he was as evil as everyone believed, that he did not care about what he had done, and his presence was a bluff in the hope that the murderer would be assumed to have been long gone. However, it is equally possible that he really was not very bright and did not have the wit to leave the area after committing his heinous crime. The third possibility is that he was actually innocent of the murder. Constable Ellery took possession of the great- coat, the state of which was crucial to the prosecution case. Mrs Flower told the magistrate that there had

MURDER AT BROOKSIDE COTTAGE

been no blood or dirt on the clothes when she had given them to him the day before. Buckland countered that the blood on his breeches was from a horse and his coat had got muddy when he knelt down to get a drink of water.

The only witness in the case was young Elizabeth Cottle who had been persuaded to move in with her grandmother after the cottage fire in the spring. The twelve-year-old girl and Mrs Pearce were asleep in the room above the buttery when they were woken by a noise. Having dressed and lit a candle they went down to the buttery. There were sounds of an intruder in the kitchen so they barred the door, but the intruder managed to force it ajar, by chopping at it with an axe. Mrs Pearce was injured in the struggle to force the door shut and block it with furniture. Once it had gone quiet she managed to make a hole in the thin lath wall of the buttery so that she and Elizabeth could escape to the village for help. As they were crossing the garden their assailant struck Mrs Pearce, knocking her to the ground, and then grabbed the child. She struggled free and as she fled she heard the man strike her grandmother for a third time. Elizabeth testified in court that she had to run about half a mile to her great-uncle’s house in Sutton Benger, and while she was telling her sorry tale the clock struck three. In answer to the prosecutor Mr Merewether, Elizabeth identified Buckland as the man who had struck her grandmother and grabbed her, although she admitted that she had not been able to see him clearly.

Many villagers contrived to take part in the trial, as witnesses for the prosecution, while Buckland conducted his own defence. Elizabeth’s uncles William and Thomas, her aunt, and neighbours Daniel Powell and John Price had all returned to the cottage where they found Judy Pearce dead in the garden. When they checked the cottage they found that nothing was missing. Rev Christopher Lipscomb, the vicar of Sutton, was praised for his ‘great pains to bring this - villain to justice.’*° He had had a model of the cottage and grounds made by the village carpenter, which he used to re-enact the crime for the benefit of the jury. Daniel Powell produced the door-post from the buttery and Thomas Ferris, landlord of the Bell Inn, had found a hatchet that had been discarded in the brook, more than three months after the murder, the blade of which it was demonstrated fitted exactly the cuts in the door post. William Greenwood of Christian Malford claimed that it was the same hatchet Buckland had tried to exchange with him for a hatful of potatoes. A modern observer would probably doubt the certainty of his assertion as his identification was based mainly on the lettering on the handle of the

51

hatchet, although he admitted he was unable to read. Greenwood reinforced the public perception of the gypsy as the despicable rogue portrayed in the anonymous poem published in the Bath Herald, by saying that he did not go ahead with the exchange as he was not convinced that Buckland had come by the hatchet honestly. The medical evidence supplied by Dr Joseph Hayward confirmed that Judy Pearce had died as a result of four wounds to the head, made with a blunt instrument. He conceded, under the prosecutor’s examination, that it was possible that the corner of a hatchet could have made the wounds. Finally Constable Ellery gave evidence that the dirt found on Buckland’s great-coat was deemed to match that from the wattle and daub cottage walls. Buckland’s conduct of his defence appears to have been consistent with his reputation of being an illiterate and simple person. Throughout the trial he had been prone to outbursts, such as calling Elizabeth Cottle ‘a damned liar, and William Greenwood a ‘lying rascally whelp, but he repeatedly refused to cross-examine the prosecution witnesses. At one point during the proceedings he began to behave so wildly that he almost convinced the judge he was insane. However, as the trial wore on the judge came to the conclusion that this was a ruse on the part of the defendant to gain the jury’s sympathy. When invited to present his case for the defence Buckland merely gave a low bow and said: ‘It is my wish to oblige your lordships as far as I can. I'll swear upon ten thousand books that I never killed the woman. May I be damned to ever-lasting if I had anything to do with the hatchet.’ In his summing up the judge reviewed the evidence presented and admitted that there was not much of a case without the hatchet. As had already been demonstrated the doctor had only admitted, after careful questioning by Mr Merewether, that it was possible that the corner of a hatchet could have produced the wounds that caused Judy Pearce’s death. It had not been conclusively proved that the hatchet found in the brook months after the murder had ever actually belonged to Buckland. Evidence of his alleged ownership had been provided by William Greenwood, based on the lettering on the handle of the hatchet — which he was unable to read, ‘not being a scholar.’ The anonymous poem, published the previous December, had already informed the public that Buckland’s normal mode of dress included a hatchet hanging from his rope girdle. This is difficult to square with the notion that the gypsy had come by the tool dishonestly and then tried to exchange it for food. Despite these flaws in the evidence it only took the jury one minute to return a guilty verdict, after which

THE WILTSHIRE PRC MAE Ore Cee AND NATURAL HISTORY MAGAZINE Account of ihe TRIALS, Sic 0b ane : AN ASHER and EDWARD BUCKLAND,

“were executed in the Court of the County Gaol of Wile) in Fisherton-Anger, Satutday, March 17, 1821;, - | °°"

4

2 ER Oe Ba ge” Sag I

a , NTA eyed

= Tike Assuaes Cr the Coonty cf Wilts, holden et 5 Py city of New Sarura, Seturceyv, March 30, 1821,

JOHN ASHEN, eged 74 yesre,

as indicted for the wilful Murder of Patrick ATKcy at the Bull Ina, Warminster, on Suaday, Augast #7, 1820. Ut appeared in evidence that Asher travelled the Country _esa@ vender of quack medicines, and that the cgpsc of the horrid deed for which be ra justly suffered, was ijereusy. Asher, bad for some time, cohebited with eo worten Cwil-

ton, nemed Sarzh Lewis; eth this woman, the Deceared,

who also travelled the country with ele cloths, bedaine’

sequainted, and frequently met her at diferent huuses of svsort for people of their ea.ployrwent. > oN. . On Sunday afternaon the old man wa: erinkn. &ith cthess at the Bull inn, Warmiinter, end went out into the back yard, aod sut one beach there, between two other men; soon efter which the young Womgn ceme out 0 the game place, ard eat down upnae heap of siraw, oppusite vo wich goon afer came out Patrick MKey, wa placed Liseoctf by ber side, and dcpan to pluy pith her, tating her -sgoved by the neck, upon which the wld gaan je 8 sege mse ‘from between the two men, ard pulling 6d © haife,\yavw Bis theec stays with such feree, thas eccording & ihe * Surgeon's report, euck of them wus mortal. :

., M'Key languished till between eleven aud twelve clock > fhe.‘next dav, and teen expire’, leaving atife and teo children at Newry in the North ef Irevand. - an '

1 ~ °

FOR MURDER. .° 4s

pihers, €0 tbey- must*xct b>

Be j Rt EDWARD BUCKLAND aged 66 scan, ~~

was also indicted for the- wilfal Murder’ of Sudith Peicce, at Seagry. >. < 2 hytirg

Bocktand was-a-mad of snost notorinas character, guian by several names, and living inno house. We cumedn the night & the residence of Judith Peirce, with u reseturiya as it anpears, of taking ewer ber bile In revenge for ber bucing accased bim some time beture uf ectting the thatch of the house on bre. Thre villain Grst beat th one side pf the house, near a small window, andeniered. Peirce ted only @ youns woman ah ber. Oo beuring the wrmse Aboy got ep, sod«inking a hicht, came down, and reticed te @ eourD where they fastened down the wooden catch to keep (mm from them. He heaving e hatchet, cut th: door so far gme:. &s to make @ cutat the womun wn the keed. They Bec

“{ from the house to e garden gate,trying ty escaze, but be

foliowing, gave We poor :nao-anutler ue on the beac,

‘Jend afecr thats ghird.cut ~whicu breegb: her wievdung tc

the ready =

.” Great préise is Cue tc the Her. C Uicscomb, vcaege -q Sutton Benger,. who took great: peins to beng this site “{ to justice, and by prodscimg « nest mode! ofebe howe-and getden, by which the Jury hed @ vciy cleat idea of che

business. Also by proceniy the very Latchet wh w dict: be bad killed the woman, 23 welj rs nea tle (rage cf this being the very fostrument with wrich ile macde? wos eficcted. ch ea (lias . The Jadge,”.an: passing che awful sentenor tf Peata,- which he did in the most cficering meurer, gaye ithe! Pri- soncrs to anderstand, Che arybey shewed my merey To yet 16 be ahem -cav drone. m ~~ a”

“nm?

i ‘ bet ie

him. ayes

The unfortonate Malelse ‘as, after theirenpJetezsticn, bebaved very penitent,. Jricing va fer edyer-wib the, Chaplain, ombing every ctoucment in Y power tie the freat and heincuus sias they bad committed, end Creting mercy Geum: tact Judge before-whom ibey were sborily :o. appew ne ee ae eae

. Abost non, after frevioce- peoyer, the Maklectors at-- cended the Drop prepered for their executiciy and on} which ‘appeared to public wiew but a short: tine ke- fore the Executioner discharged, bis ‘duty py lausching them fate eternity. © Jbey metibeir ignominturs end with aech fortitude and comprure; beieg perfectly resigned to

Asher was poro at Narket Laviygton, and bes left.2 wile and threfchildres, iw 3

Buckland was torn at Crea: Bedwin, 00d dss alto Icfte wife and-theee children. eee 2 :

Anonymous pamphlet describing the trials and double execution, 1821

MURDER AT BROOKSIDE COTTAGE

the judge donned his black cap and pronounced the death sentence. Buckland was led away, still protesting his innocence. The newspaper reported that his last words to the court were, ‘ I hope gentlemen you won’t take away my life; upon my soul, I did not kill the old woman.’

Edward Buckland was hanged in the courtyard of the county gaol of Wiltshire in Fisherton Anger on Saturday 17 March 1821 in a double execution with another convicted murderer. John Asher, a 74 year old itinerant vendor of quack medicines, was also tried at the Salisbury Lent Assizes and convicted of murdering Patrick McKey in a jealous rage, at the Bull Inn, inWarminster. A double scaffold was erected at the prison and the pair were dispatched simultaneously. The journalist covering the event reported that Buckland looked around for members of his gypsy tribe and asked, from the scaffold, if any of his people were there. Just before the drop fell he announced to the crowd, “They are going to murder me’, and then asked the hangman if he could, ‘hang me up alittle then let me down again’. The journalist’s view of the execution was very different from that of the anonymous author of the trial pamphlet. He was at pains to stress the remorse and repentance of the convicted men as a moral lesson for his readers, rather than admit the possibility that one of the men might have been unjustly convicted, and was still protesting his innocence to the bitter end. The pamphlet contained the following account of the hanging:

The unfortunate Malefactors, after their condemnation, behaved very penitent, joining in fervent prayer with the Chaplain, making every atonement in their power for the heinous sins they had committed, and craving mercy from that Judge before whom they were shortly to appear.

About noon, after previous prayer, the Malefactors ascended the Drop prepared for their execution, and on which they appeared to public view but a short time before the Executioner discharged his duty by launching them into eternity. They met their ignominious end with much fortitude and composure being perfectly resigned to their fate.*!

The pamphleteer concluded his account with the information that each man had left a wife and three children.

The day after the execution the Sutton vicar, the Rev Christopher Lipscomb, preached in the parish church of All Saints a lengthy sermon on the evil murder. The sermon was published in pamphlet form, and copies sold for a shilling.*? The whole episode of the murder, trial, execution and sermon had such an effect on the villagers that a collection was made for a memorial stone to be placed on Judy’s grave in Sutton

53

@ Sermon,

PREACHED

IN THE PARISH CHURCH

oP

SUTTON-BENGER,

On SUNDAY, March the 18th, 1821, BEING

THE DAY: AFTER THE EXECUTION

oF

EDWARD BUCKLAND,

FOR THE

MURDER

or

JUDITH PEARCE.

BY THE

Rev. CHRISTOPHER LIPSCOMB,

FELLOW OF NEW COLLEGE, OXFORD, AND VICAR OF SUTTON- BENGER, WILTs.

a CHIPPENHAM: PRINTED AND SOLD BY J. Mo COOMBS, High-Dtreet,

DOREY DO

PRICE ONE SHILLING.

Cover of Rev Christopher Liscomb’s 1821 published sermon relating to the murder.

Benger churchyard. The Salisbury and Winchester Journal reported that £7 had already been received, and that, ‘the overplus [is] to be given to the family of the deceased.’ The memorial has unfortunately suffered the ravages of time, despite being recut many years ago, and the lower portion of the inscription is no longer readable. It was fully transcribed in the early 1980s, reading as follows: This stone was erected by public subscription in memory of Judith Pearce of the Parish of Seagry Widow Aged 58 years. Her blameless life of Diligence and Honesty was terminated by an Act of the most malicious Barbarity. On the night of the 13 of November 1820 She was cruelly murdered by Edward Buckland a Gipsey whose crime was providentially brought to light and he was executed 17 March 1821

54 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The area around Brookside Cottage has attracted many ghostly sightings since the execution, the restless spirit being that of Edward Buckland, wrapped in his distinctive blanket. Those claiming to have met the ghost include long-standing residents, such as Dennis Selwood’s mother Ada in 1901, and newcomers to the area, as recently as the 1970s.”’ The apparition is of a small man wrapped in a dirty old cloazk, and has been described by people who had apparently never heard of the gypsy or how he used to dress. He appears at the point where the footpath from Christian Malford meets the Seagry Road, not far from Brookside Cottage. For those that believe in the restlessness of the spirits of those wronged in life, the wandering of Ted’s ghost raises the question: was he really innocent as he always claimed? Was he hanged for a murder he did not commit? Or is his restlessness because he has been, ‘damned to everlasting torment,’ by his own words, for using the hatchet? However, there are more down to earth reasons for challenging the outcome of this case.

It certainly seems that Buckland was a victim of the prevailing prejudices of his time, suffering from

Tombstone of Judith Pearce, Sutton Benger churchyard

the widely-held assumption that all gypsies were untrustworthy, despicable rogues. The ‘respectable’ majority all relished their parts in bringing to book this evil little man. It is not hard to feel some sympathy for the gypsy over the way his case was handled, even though the circumstantial evidence would seem to indicate that Buckland was correctly convicted of the crime for which he paid the ultimate price. At best uneducated and at worst mentally deficient, he was left to conduct his own defence with no help or advice from learned counsel. Newspaper reporting in the run-up to the trial was highly prejudicial to the defendant. The medical evidence, and that relating to ownership of the possible murder weapon, was highly suspect, and the only witness was a frightened twelve-year-old girl, startled from her sleep and stumbling around in the dark. It is still possible, today, to appreciate the impenetrable quality of the blackness of a November night as the village street lights, which do not yet extend as far as Brookside Cottage at the foot of Seagry Hill, are turned off around midnight. The truth will never be known, but the question remains: was Edward Buckland the victim of a gross miscarriage of justice, or was he justly convicted despite the short-comings of his trial?

References

Primary Sources

WSRO 2859/1: ‘Account of the Trial &c of John Asher and Edward Buckland who were executed in the Court of the County Gaol of Wilts in Fisherton-Anger, Saturday March 17, 1821, (Easton, Printer, Sarum, 1821).

Newspapers and Sermon

Bath Herald, 2 December 1820.

Devizes and Wiltshire Gazette, 22 March 1821.

Salisbury and Winchester Journal, 26 March 1821.

Rev Christopher Lipscomb, ‘A Sermon Preached in the Parish Church of Sutton Benger, 18th March 1821, Wiltshire Sermons, vol 2, (Wiltshire Archaeological and Natural History Society Library, Devizes).

Books

H K Anketell, Collections for a History of Seagry, (Devizes, 1886).

W G Hoskins, Local History in England (Longman, 1972).

Kilvert’s Diary 1870-1879 (Bracken Books, London, 1992).

Journals and Articles

W K Griffiths, ‘Edward Buckland,’ Kilvert Society Newsletter.

Alessandro Portelli, “The Peculiarities of Oral History,’ History Workshop, no. 12, Autumn 1981, pp 96-107.

MURDERAT BROOKSIDE COTTAGE

Raphael Samuel, ‘Local History and Oral History,’ History Workshop, no.1, Spring 1976, pp 191-208.

Elizabeth Tonkin, ‘Investigating Oral Tradition, Journal of African History, 27 (1986), pp 203-213.

D R Woolf, ‘The ‘Common Voice’: History Folklore