Full text of "The Wiltshire Archaeological and Natural History Magazine"

See other formats

ayia

THE
WILTSHIRE ARCHAEOLOGICAL
AND NATURAL HISTORY MAGAZINE

VOLUME 89
1996

This volume of the Magazine is dedicated, with affection and gratitude, to the memory of
Margaret Guido, FSA, Vice-President of the Society, who died on 8 September 1994.

Published by
Wiltshire Archaeological and Natural History Society
41 Long Street
Devizes SN10 INS
Telephone (01380) 727369

Registered Charity Commission No. 309534 V.A.T. No. 140 2791 91

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE
VOLUME 89 (1996)

ISSN 0262 6608
© Wiltshire Archaeological and Natural History Society and authors 1996

Editor: Kate Fielden, BA, D.Phil.
Hon. Natural History Editor: Patrick Dillon, B.Ed., Ph.D., C.Biol., MIBiol., FLS
Hon. Local History Editor: James Thomas, BA, Ph.D, FRHist.S
Hon. Reviews Editor: Michael Marshman, ALA
Editorial Assistant: Lorna Haycock, BA, Dip.ELH, Cert.Ed.

Change of Title
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.

We acknowledge with thanks publication grants for this volume from the following bodies: The Gleeson Group
plc, for the paper ‘Prehistoric Sites and a Romano-British Settlement at Butterfield Down, Amesbury’ by Mick
Rawlings and A.P. Fitzpatrick; Esso Petroleum Co. Ltd, for the paper ‘A Romano-British Farmstead and
Associated Burials at Maddington Farm, Shrewton’ by Jacqueline I. McKinley and Michael Heaton; and the
Council for British Archaeology for Graham Brown’s paper ‘West Chisenbury: Settlement and Land-use in a
Chalk Downland Landscape’.

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.

Origination by
Avonset, Midsomer Norton, Bath
Set in Plantin
Printed in Great Britain by
The Looseleaf Company, Melksham, Wiltshire

WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY SOCIETY
COUNCIL AND OFFICERS (31 July 1995)

PRESIDENT
Professor S. Piggott, CBE, FBA, FSA

VICE-PRESIDENTS FOUNDATION TRUSTEES
Professor W.H. Dowdeswell, MA, FIBiol. B.H.C. Sykes, MA
R.G. Hurn R.G. Hurn
N. Davey, OBE, Ph.D., D.Sc., FSA H.F. Seymour, BA

Dr T.K. Maurice, OBE
J.F. Phillips, B.Sc.

Chairman
N. de PE.W.Thomas, MA, FSA (1993)

Deputy Chairman
Professor W.H. Dowdeswell, MA, FIBiol. (1992)

Elected Members

Dr C.A. Shell, MA, MMet. (1990) C.J. Gingell (1992)
M.J. Smith, BA, FICE (1990) Miss S.F. Rooke (1992)
DJ. Williams, MA (1990) N. de VE.W. Thomas, MA, FSA (1992)
F.K. Annable, BA, FSA, FMA (1991) Ms A. Borthwick : (1993)
Mrs S.J. Buxton (1991) L. Luckett, BA (1993)
E. Elliott, B.Sc., F.B.Ps.S. (1991) C.J. Perraton, MA, CBiol., MIBiol. (1993)
J.M.G. Kirkaldy, Ph.D, M.Sc. (Econ.) (1991) D.N. Shelton, BA, B.Sc., Dip.Ed., FRGS (1993)
N.J. Anderson (1992) Lt Col C. Chamberlain (1994)

Ex-officio Members

Mrs G. Swanton, BA, Dip. Ad. Ed. Chairman, Archaeology Committee
Miss K.J. Fielden, D.Phil. Chairman, Buildings & Monuments Committee and Editor
*Mrs P. Slocombe, BA Chairman, Industrial Archaeology Committee
Mrs G. Learner, BA Chairman, Library Committee
Mrs P. Sneyd, Ph.D, B.Sc., CBiol., MIBiol. Chairman, Natural History Committee
*D.J. Williams, MA Chairman, Programme Committee
R.S. Webber Hon. Treasurer (desig.)
D. Rowe, Dip. Ed., Cert. Ed. Hon. Publicity Officer
R.C. Hatchwell, FSA Hon. Keeper of Prints and Drawings

Nominated Members

Mrs D.J. Main, Miss D.J. Matthews and A.H. Goring Members, Wiltshire County Council
A.R. Taylor Member, Devizes Town Council
TR. O’Sullivan Member, Kennet District Council
A.P. Bishop Member, North Wiltshire District Council
T. Carbin Member, West Wiltshire District Council
Mrs P. Slocombe, BA Wiltshire Buildings Record
R.L. Pybus, MA, FLA, FBIM, MILAM, FRSA Director, WCC Library & Museum Service
P.R. Saunders, BA, FSA, FMA Curator, Salisbury & South Wiltshire Museum

*also Elected or Nominated Member

OFFICERS
Secretary P.M. McG. Coston
Curator P.H. Robinson, Ph.D, FSA, AMA
Assistant Curator Miss A.J. Hoog, BA, MA
Assistant Curator (Natural Sciences) A.S. Tucker, B.Sc.
Assistant Curator (Biological Recorder) Miss S. Scott-White, MA
Sandell Librarian Mrs P. Colman, Dip. ELH, FRSA

Assistant Librarian Mrs L. Haycock, BA, Dip. ELH, Cert. Ed.

Contents

Prehistoric Sites and a Romano-British Settlement at Butterfield Down, Amesbury

by MICK RAWLINGS and A.P. FITZPATRICK
with contributions by D.A. ALLEN, MICHAEL J. ALLEN, A BURNETT,

ROSAMUND M.J.CLEAL, M.CORNEY, ANDREW CROCKETT, S.M.DUGGAN, J.EGERTON,

M. FAIRBROTHER, P.A. HARDING, M. HENIG, J.I. MILLARD and SARAH F. WYLES

A Romano-British Farmstead and Associated Burials at Maddington Farm, Shrewton 44

by JACQUELINE I. MCKINLEY and MICHAEL HEATON

with contributions by RACHAEL SEAGER SMITH, DAVID MURDIE, MICHAEL J. ALLEN,

JOHN A. DAVIES, S. HAMILTON-DYER, P. HINTON and R. MONTAGUE

West Chisenbury: Settlement and Land-use in a Chalk Downland Landscape
by GRAHAM BROWN

Royal Itineraries and Medieval Routes
by NORMAN HIDDEN

Wiltshire Deer Parks: An Introductory Survey
by KENNETH WATTS

William Lisle Bowles: The Making of the Bard of Bremhill
by DOREEN SLATTER

A Provisional Checklist of Fossil Insects from the Purbeck Group of Wiltshire
by A.J. ROSS and E.A. JARZEMBOWSKI

The Common Stinging Nettle (Urtica dioica L.) in Wiltshire, 1994
by JACK OLIVER

Notes

An Early Anglo-Saxon Wrist-clasp from the Parish of Baydon by JOHN HINES

The Identification of Alentone/Allentone in Wiltshire Domesday by JASON ST. JOHN NICHOLLE
The Silver Seal-matrix of Geve of Calstone by JOHN CHERRY

A Medieval Heraldic Roundel from Potterne by PAUL ROBINSON

An Addition to the Stourhead Canon by ROBERT J. MAYHEW

Pen Pits and Sir Arthur Bliss by RAYMOND J. SKINNER

Excavation and Fieldwork in Wiltshire 1994

106

116

130

130
132
134
136
138
139

144

Reviews

Michael Aston and Carenza Lewis (editors), The Medieval Landscape of Wessex (JOHN CHANDLER)
Richard Bradley, Roy Entwistle and Frances Raymond, Prehistoric Land Divisions on
Salisbury Plain. The Work of the Wessex Linear Ditches Project (MARK CORNEY)
D.A. Crowley (editor), A History of Wiltshire, VCH Vol. 15: Amesbury Hundred and Branch
and Dole Hundred (K.H. ROGERS)
A.P. Fitzpatrick and E.L. Morris (editors), The Iron Age in Wessex: Recent Work (ROY CANHAM)
J.K. Kirby (editor), The Hungerford Cartulary (STEVEN HOBBS)
Michael Parsons, Farley with Pitton, Vols. 1-11 (MICHAEL MARSHMAN)
F.E. Warneford (editor), Warneford Star Chamber Suits, WRS Vol.48 (STEVEN HOBBS)

Obituaries

Richard Atkinson, CBE
Grace Fairhurst

Leslie Grinsell, OBE
Peggy Guido

Index

154

154
155

156
157
157
159
159

161

161
161
162
163

165

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 also maintains
the Wiltshire Biological Records Centre at the Museum.

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 Secretary.

Contributions for the Magazine should be on subjects related to the archaeology, history or natural history
of Wiltshire. There is no fixed length. Papers, notes and reviews should be typed on one side of a page only, with
good margins and double spacing. The style for footnotes, references and so on should be that found in this
issue. The author-date system is preferred for references and footnotes should be avoided unless essential.
Contributions of article length should be accompanied by a summary of about 100 words. Two copies, one of
which is a top copy, should be sent to the editor at the Museum, 41 Long Street, Devizes, Wiltshire, SN10 1NS.
A further copy should also be retained by the author. The editor and subject editors will be pleased to advise
and discuss with intending contributors at any stage during the preparation of their work. They will also supply
notes, if requested, which may be helpful in explaining house style and in giving advice on the compilation of
references and bibliographies, and the preparation of iilustrations.

Proofs: authors will receive galley proofs only.

Offprints: ten offprints of each article will be given free or shared between joint authors of articles (not notes
or reviews). Further offprints may be ordered from the printer at galley proof stage, when prices will be
indicated.

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.

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 1-43

Prehistoric Sites and a Romano-British Settlement at
Butterfield Down, Amesbury

by MICK RAWLINGS and A.P. FITZPATRICK

with contributions by
D.A. ALLEN, MICHAEL J. ALLEN, A. BURNETT, ROSAMUND M.J. CLEAL,
M. CORNEY, ANDREW CROCKETT, S.M. DUGGAN, J. EGERTON, M. FAIRBROTHER,
P.A. HARDING, M. HENIG, J.I. MILLARD and SARAH F. WYLES

Work in advance of house building allowed the planning and limited excavation of prehistoric and
Romano-British sites. A possible Late Neolithic pit-ring henge was identified, while a ring ditch, a
crouched inhumation and a decorated chalk plaque were considered to be probably of Early Bronze Age
date. A pit containing the major portion of a large Beaker and fragments of two other Beakers was found
adjacent to a large ditch which 1s probably of later Bronze Age date. Early Roman occupation lay largely
outside the areas examined. A village-like later Roman settlement was identified, where a wide range of
environmental and artefactual evidence indicates a mixed farming economy. Parts of an outlying
enclosure of the same date were examined. An early Sth-century gold coin hoard was also found.

INTRODUCTION

The archaeological investigations considered in this
report were undertaken by Wessex Archaeology for
The Gleeson Group plc. Work was conducted in
several stages from February 1990 until 1993, both
prior to and during construction of a new housing
estate on 23ha of open ground on the eastern outskirts
of Amesbury (centred on SU 166414: see Figure 1).
The estate, named Butterfield Down after architect
William Butterfield (1814-1900), who worked on the
parish church of St Mary and St Melor, Amesbury, is
still under construction and there is a possibility that
further archaeological work may take place.

Archaeological investigations began with a sys-
tematic surface artefact collection over the whole site
(Figure 2), and were followed by the excavation of a
series of 21 small ‘keyhole trenches’ (Figure 3) and a
subsequent watching brief. In October 1990, the
findspot of a hoard of eight gold coins just to the
north of Trench 22 was examined (Figure 1). A
geophysical survey covering 3.5ha, designed to
locate concentrations of features in the area to the
north of Trench 22 was carried out in the same
month. A smaller area to the west of Trench 22 was
included in the survey as a means of testing the
feasibility of geophysical survey techniques on this
site; this area was to be stripped and investigated as
Trench 23 in November 1990.

Two areas were subjected to an archaeological
watching brief during topsoil stripping early in 1991
(Figure 3, Trenches 25 and 26). An evaluation com-
prising geophysical survey and trial trenching was
carried out in December 1991 in an area at the south-
western edge of the development site (Figure 4).

In August 1993 a further area along the eastern
edge of the site was examined prior to development
(Figure 5) and a Beaker pit was found.

The methodology and results of each phase of
work are discussed below.

Geology and topography

The site is situated on the western part of the predom-
inantly flat-topped ridge of Cretaceous Upper Chalk
known as Boscombe Down (Figure 1, B). No deposits
of clay-with-flints have yet been recorded and the soils
are typically thin rendzinas, becoming slightly thicker
on the slopes. The land slopes down gently to the
north and west. Most development so far has taken
place on the more level ground at the southern edge of
the field, between 110m and 115m OD.

Archaeological background

Butterfield Down is situated within a rich archae-
ological landscape (Figure 1, B-—C). Stonehenge
is less than 5km to the west; there is a group of
well-preserved Bronze Age barrows 1.3km to the

Swindon
Reading
@

_ ESBERY

. a
Salisbury
Southampton

Road

Key:
at Area of
L...t development

“Holders

SA ff,

Y, Built-up area
V4

aye Trenches

Cropmarks

Location
Ww of Hoard

a Find spot
& S.M.R.no.

® Round barrow
& S.M.R. no.

24 Other S.M.R. features

All contours in metres O.D.

Figure 1. Butterfield Down: Location in relation to Amesbury and previously known sites and monuments (A-B);
distribution of cropmarks, excavated trenches, and location of gold coin hoard (C)

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN

Romano-
British
Pottery

YM Dense scatter

Lilli 6
//// Light scatter

Figure 2. Butterfield Down:

Ceramic
Building
Material

distribution of selected materials recovered in fieldwalking survey

4 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

north-east on Earl’s Farm Down, and a bowl barrow
immediately to the south of the site (Sites and
Monuments Record (SMR) 688) was destroyed in
1951. The existence of further barrows is suggested
by ring ditches recorded as cropmarks to the south-
west (SMR 720 and 773). The SMR records a
possible round barrow within the development area
(SMR 689) noted by O.G.S. Crawford on his private
6-inch map but this monument cannot now be
located either on the ground or in aerial photographs.

A large ditch (SMR 745) which forms one
element of the Earl’s Farm Down network of Bronze
Age linear features (Bradley, Entwistle and Raymond
1994, fig. 22), runs up to the eastern edge of
Butterfield Down and continues across the devel-
opment area. Another ditch (SMR 749), which had
a low bank on each side, lay to the south-east of the
site but is now largely built over; it may also be part
of the Earl’s Farm Down network, possibly a
continuation of SMR 745.

Archaeological work during building at Boscombe
Down airfield, to the east of the present site, revealed
widespread occupation during the late Iron Age
and Romano-British periods (Richardson 1951).
Cropmarks recorded in recent aerial photographs at
Southmill Hill, less than 1km to the south-west of
Butterfield Down, have been interpreted as further
evidence of activity in the later prehistoric period
(McOmish 1989, fig. 5). Aerial photographs taken in
1990 by the RCHM(EB) show several large and diffuse
linear features, possibly trackways, running into the
central area of Butterfield Down. Also visible is a
group of ditched enclosures at the eastern side of the
development area and several shorter lengths of linear
cropmarks (Figure 1, C). The enclosures are close to
the line of the large linear ditch, SMR 745.

The numerous known features and findspots of
the Romano-British period provide evidence of
intensive activity at that date in the area.
Immediately south of the site, construction work in
1951 revealed two middens containing Romano-
British pottery as well as a ditch, pits and a roadway
(SMR 303). Further south, a pot was found in
c. 1842, within which were bronze and silver coins of
the 3rd and 4th centuries AD and some silver finger
rings (SMR 305, and Corney, below). Part of a
Romano-British inhumation cemetery (SMR 306)
was located during railway construction in 1900
north of the site, and further sherds of Roman
pottery have also been found close by (SMR 307).
‘To the north-west of Butterfield Down a bronze coin
of the late 3rd century AD was found in a garden
(SMR 318) and to the south-west a series of late

2nd- to 4th-century AD coins was recovered in 1922
and 1972 (SMR 304).

In addition, a Neolithic stone axe and several
items of metalwork have been found within recent
years at Butterfield Down by local metal-detector
users and identified by Salisbury and South Wilts
Museum. The metalwork includes a possible Late
Iron Age sword hilt pommel, a Roman disc brooch,
a medieval harness pendant and a post-medieval
copper alloy strap end.

Throughout this report the terms early and late
Roman refer to the Ist-2nd and 3rd—5th centuries
AD, respectively.

Fieldwork
by MICK RAWLINGS
METHODOLOGY

Phase 1: Fieldwalking

An extensive fieldwalking survey was carried out
across the whole of the development site in
February 1990. Although there was a low crop
cover, ground conditions were reasonable and
artefact recovery was good. All artefact types were
collected along a series of 25m transects at 25m
spacing, on a grid based on the National Grid, and
a full report was prepared (Wessex Archaeology
1990).

Phase 2: Excavation stage 1

In July 1990 a series of small trenches was opened at
the southern end of the development area (Figure 3,
Trenches 1—21). After machine-stripping, all features
were hand-cleaned and planned and a sample of
feature types was excavated to assess the date and
nature of earlier activities. Most of the trenches were
placed where a concentration of Romano-British
pottery was identified during fieldwalking. Trenches
1-13 were located on the sites of proposed houses,
Trenches 14—19 along the routes of proposed roads,
and Trenches 20 and 21 were positioned to examine
a linear feature recorded on aerial photographs
(Figure 1, C). A watching brief was subsequently
carried out during construction within the overall
area in which Trenches 1—21 were located; features
were identified but not excavated: this area was
recorded as Trench 22.

Phase 3: Excavation stage 2

In November 1990 a second stage of excavation
(Trench 23) was carried out ahead of development
to the west of the previous excavations. The principal

WwW

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN

poqewoo] d1B [Z—-[ SOYOUILT, YOTYM UIYIIM BIE [[eIDAO dU] SI ZZ YOUSLLL

<(saseyd [[e) somnqeay [eorBopoaeyore puke 9Z—] soyouaIL, JO UONedO] :uUMOG PlEeyseung “¢ 91nB1,J

SU 164
Geophysical survey grid
¢ =
Extant building \
_| \S
. 27
- i
Ss —<——— - +
z/ X
° S
cl |e"
£ / vi] . |
2
2 x |
= she : 2) : ae u {
. 33
\3, “
ke ‘i la
NS \ S ~
2 aN ts , s
e SS SS
wv N WG
: N ~S NN $s
mS NN Sa NS
N \ “30 \ 28 x
N x LN Cs) X .
\ YN k Nin \
X S S N N
~~ \ N wy pS
NN SF Le | N N wl iks S
[yh ws * gw
\ ; »
411 BS - NY
Key: SS |
S.J Features located b eee Ree)
\NS ; z oe
‘| geophysical survey ae Sep
ag
. S
Sy
Ferrous disturbance ed
[WJ Archaeological feature stone
: within Trench
[=~] Non-archaeological feature
0 100
=cCj’jxToNr oe rene om
WA

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Figure 4. Butterfield Down: geophysical survey and trial trench evaluation in the south-west corner
of the development area

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 7

objective of this work was to recover an accurate plan
of a substantial area of the site. The whole area
(c.0.54ha) was stripped using a_ mechanical
excavator and the exposed chalk was then hand-
cleaned. All features were planned (Figure 6) and a
sample was excavated. In general the density and
distribution of the features in the trench confirmed
the results of the geophysical survey within this area,
which had indicated a large number of discrete
features but no substantial linear ones. A similar
pattern was recorded by a geophysical survey in the
area to the north of Trench 22.

Phase 4: Watching Brief

An archaeological watching brief was carried out
during pre-construction work to the north of the
Phase 2 excavation area (Figure 3, Trenches 25 and
26). Following machine-stripping of the soil cover,

Areas excavated

Features visible after stripping

all visible features were planned and recorded. No
excavation was undertaken within these trenches.

Phase 5: Evaluation

A geophysical survey was undertaken across the area
of land located immediately to the west of Trench 23
(Figure 1) prior to construction work. A number of
trial trenches were dug to investigate areas suggestive
of high or low archaeological potential and a sample
number of archaeological features was excavated
(Figure 4).

Phase 6: Excavation stage 3

Following machine-stripping of the soil cover over an
area of land located on the eastern edge of the
development area (Figures 1 and 5), all visible
features were planned and recorded, and a sample of
feature types was then excavated.

Figure 5. Butterfield Down: Beaker Pit (2) and prehistoric and Romano-British ditches at the eastern edge of the
development area

8 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

- 411 411

Key:

Trench 23

SES

Figure 6. Butterfield Down: plan of archaeological features in Trench 23 (all phases)

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 9

FIELDWALKING RESULTS

Almost all of the pottery found was Romano-British
and spanned the whole of that period. There was a
clear concentration in the southern part of the
development area (Figure 2, I), within which was an
area of larger sherds.

Three distinct concentrations of brick and tile
were identified (Figure 1, II, A-C). Of these, only B
could be positively dated as Romano-British; C
mainly comprised modern building debris but A
contained few datable pieces.

The main concentration of worked flint was along
the eastern edge of the development area (Figure 2,
III, B, with a small discrete cluster indicated at A).
Most of the flint was heavily plough-damaged making
precise identification difficult; however, the majority
of tools were scrapers. Apart from the scrapers only
one tool, a fabricator, was found. Overall the tech-
nology of the assemblage shows little evidence of
deliberate blade production and this, together with
the restricted range of tool forms, suggests that the
majority of the worked flint can be dated to the later
Bronze Age.

Earl’s Farm Down Linear

Flint

Clay loam

Silty clay

Clay silt

The distribution of burnt flint (Figure 2, IV) is
similar to that of worked flint, with a small cluster in
the centre of the field (A) and a broader concentration
(B) along the eastern edge. In the course of other
surface collection surveys in the vicinity of Amesbury,
there has been a consistent association between higher
levels of surface burnt flint and areas of later Bronze
Age activity (Richards 1990).

A few fragments of quernstone were also recovered
during the surface collection, as well as a Roman coin,
all from within the main scatter of Romano-British
pottery.

The fieldwalking highlighted areas of greater
archaeological potential within the development area. A
definite concentration of worked flint associated with a
scatter of burnt flint suggested activity during the later
Bronze Age along the eastern edge of the field, which
may be associated with a linear ditch and enclosures
adjacent to this area identified on aerial photographs.
The other cropmarks seen in aerial photographs appear
to be associated with an extensive area of Romano-
British settlement, within which a discrete concen-
tration of ceramic building material (Figure 2, I, C)
may derive from one or more substantial buildings.

Figure 7. Butterfield Down: sections through prehistoric ditches at the eastern side of the development area

10 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The excavations and associated
works

PREHISTORIC

Beaker pit and linear ditch

A number of small pits were recorded following
machine-stripping of an area of land at the eastern
edge of the development site (Figure 5), and two of
these were subsequently excavated. Pit 2 was
irregular in both plan and _ profile, generally
measuring 0.8m in diameter and 0.3m in depth. The
single homogeneous fill contained the fragmented
major portion of a large decorated Beaker (Figure
15, 1) and sherds of two other Beakers, along with
several pieces of worked flint. Pit 6 was more circular
in plan and measured 0.75m in diameter and 0.15m
in depth. A few pieces of worked flint were recovered
from the single fill, but no other finds were noted.

Located close to these pits were two nearly
parallel ditches aligned north-east/south-west. The
smaller and more northerly of these (Figure 5, 18)
was 1.9m wide at the surface and 0.5m deep. The
profile is of a shallow, flat-based feature with a
deeper central gully which suggests a recut of the
ditch (Figure 7), though no such event was
discernible within the fill sequence. A single sherd
from a coarse Beaker was found in the basal fill, and
a sherd from a Collared Urn was found in the upper
fill along with several small sherds of Romano-
British pottery. A bronze coin of Romano-British
date was found on the surface of this feature.

Ditch 21, to the south, is almost certainly the
large linear feature recorded as SMR 745, visible as
a cropmark and extant earthwork for over 5.5km
eastwards from Butterfield Down. This has been the
subject of previous investigations (Cleal et al. forth-
coming) and is suggested to be part of a land division
system of probable Late Bronze Age date (Bradley,
Entwistle and Raymond 1994). The present ditch
was 4.15m wide at the surface and 1.45m deep. It
was V-shaped in profile and had a small, vertical-
sided, flat-based slot at the base (Figure 7). A chalky
basal fill extended up the sides of the ditch. The
upper fills were darker and more loamy, and
contained several sherds of pottery, predominantly
of Romano-British date but including one of Late
Iron Age date and one modern. No finds were
recovered from the lower ditch fills.

Trench 23

A ring ditch in the south-west corner of Trench 23 was
probably 20m in diameter, although not all of it lay
within the excavations (Figure 6, 3005). Two sections

were excavated across the ditch revealing that it was
1.5m wide at the top, 0.8m wide at the bottom and
0.6m deep, with a flat base. A substantial quantity of
worked flint was found within the ditch fill sequence:
much of it was primary knapping debris, while
elements of blade production on the cores indicated a
later Neolithic/Early Bronze Age date. Two small
sherds of undiagnostic Roman coarseware were found
in the upper fill of the ditch; no pottery was found in
the primary fill.

Four pits lay adjacent to the ring ditch. To the
south-east a small oval pit (2948) had been cut
through the outer edge of the ditch. It was about 1m
long and 0.35m deep with steep, straight sides and
contained a large assemblage of worked flint charac-
teristic of the later Bronze Age along with three
sherds of Late Bronze Age pottery. Three sherds of
Romano-British pottery are considered to be
intrusive.

A circular pit (2943) just to the north of the ring
ditch was 2m in diameter and 0.6m deep. The upper
fill contained a few pieces of undiagnostic worked
flint and a small quantity of pottery, including one
sherd of later Neolithic Peterborough ware and two
sherds of Middle—Late Bronze Age Deverel-Rimbury
type. Although one of the remaining sherds was
Romano-British, the others were certainly prehistoric
and in a fabric similar to the Peterborough ware
sherd. No finds were recovered from the basal fill.

To the east of this pit was a slightly smaller
circular pit (3004) 1.2m in diameter and 0.6m deep,
containing only a few undiagnostic worked flints in
the upper fills. On the flat base of the pit was the
crouched inhumation of a child aged approximately
12 years (Plate 1) who had suffered from a severe
abnormality of the lumbar vertebrae which might
have caused permanent paralysis. There were no
accompanying grave goods, but in the compacted
shallow fill below the skeleton was a tiny sherd from
an accessory vessel, possibly an Early Bronze Age
“incense cup’ (Figure 16, 1).

Pit 2964 was 1.5m in diameter and 0.45m deep.
A few small fragments of human bone were
recovered from the upper fill but there were no finds
in the lower fills. To the south-east of this pit was a
subrectangular feature (2998), 2.2m long by 0.7m
wide and 0.35m deep, with steep sides and a flat
base. No artefacts were found but the colour and
texture of the fills were similar to those found in the
prehistoric features. Three similar subrectangular
features lay close by and could have been part of a
small, segmented ditch enclosure adjacent to the
ring ditch.

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 11

Plate 1. Butterfield Down: (child) Burial 3004, probably Early Bronze Age, from Trench 23

ROMANO-BRITISH

Trenches 1-22. Features of Roman date were
recorded in every trench but Trench 19 (Figure 8).
The highest density of features was in the central and
western parts of the area (Trenches 2, 3, 4 and 7),
with markedly lower densities in the east (Trenches
10-11) and south-west. A wide range of feature
types was recognised and sixteen individual features
or groups of features were excavated.

The wide linear feature 1302 (Trench 13) was
shown to be only 0.25m deep and to contain a
substantial quantity of late Roman pottery. It
appeared to continue into Trench 12. The feature
correlates with one recognised in air photographs
(Figure 1, C) and its interpretation as a trackway was
confirmed by the excavation. A shallow feature
(1202) excavated in Trench 12, which had a basal
layer of packed burnt flint, may be the remains of an
associated trackway.

A feature recognised in air photographs and
identified in Trenches 20 (2001) and 21 was shown
to be 0.7m deep with a flat base. Although the only
dating evidence was four undiagnostic Roman

sherds, the feature may be either a ditch enclosing
the eastern part of the site or a part of a
contemporary field system next to the settlement.
Other linear features included a short and very
shallow curving gully (Trench 8, 808) and the
terminal of a ditch (Trench 2, 202) of similar size to
2001, which had been recut at least once. The earlier
ditch contained only Roman coarsewares; the recut
included diagnostic late Roman types.

Two pits, 312 (Trench 3) and 404 (Trench 4)
were excavated to a depth of 1.9m and 2m, respec-
tively, before being backfilled. Large quantities of
sheep bones were recovered from pit 404, in
particular mandibles and foot elements. Substantial
quantities of late Roman pottery were recovered from
both pits but pit 404 also contained a single sherd of
Early Bronze Age date as well as a fragment of a clay
spindle whorl and two bone objects.

Two further pits, 304 and 306 (Trench 3), were
c.0.35m deep and contained undiagnostic Roman
coarsewares, although two sherds of the later Bronze
Age were found in 306. Three groups of features
were excavated in Trenches 1, 6 and 15. In each case
the irregular plan of the group was shown to be the

fas/vm

oor os 0

yom uoljoNIJsUOD Hulinp pahieasgo sainjedy |

Pale payenedxa UIYIIM PaAtasqo sain}ee4 pele |

Aay

ae eae
(e)

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

tle

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 13

~
~

~
i
~~
—

fil 1.95m0.D.
RAY

SEJ

Figure 9. Butterfield Down: plan and section of ditch or pit 310 in Trench 3

result of a series of shallow intercutting scoops. The
pottery from the groups in Trenches 1 and 6 was late
Roman but that from the group in Trench 15 was
early Roman.

Feature 310 (Trench 3) was similar in profile
(Figure 9) to pits 312 and 404, although this one was
excavated to the base, a depth of 2.7m. It was not
clear on excavation if this was a ditch terminal or an
elongated pit. The complete corpse of a crow (either
carrion or hooded) had been placed on the chalk
bedrock and covered with a layer of dark soil (370).
No other finds were recovered from this layer but the
two overlying fills (369 and 367) contained large
quantities of late Roman pottery and other finds such
as animal bones and oyster shells. The upper part of
the pit contained a deposit of chalk rubble (358) and
a final fill of brown silty clay loam (357). Finds from
357 included late Roman pottery and a small chalk
plaque with incised decoration (Figure 14 and Plate 4).

A dumbbell-shaped feature (802) in Trench 8
was demonstrated to be a small, oval dryer or oven,

with an ovoid hearth at the south linked by a short
flue to the stokehole at the north. The hearth
contained blocks of burnt chalk and 2nd-century or
later coarsewares.

At the north-west corner of Trench 4 a sub-
rectangular feature, possibly the remains of a cellar
or a sunken-floored building (442), was 3.7m
long and 2.7m wide, with an average depth of
c.0.4m (Figure 10). The undulating base rose
slightly in the north-east corner and along the
eastern side there was a small ledge. Towards the
southern end of the ledge was an oval depression
0.2m deep (444) which, if contemporary, may have
supported a post or been a step. There were two
stakeholes at the north-east corner of the feature.
Large quantities of 3rd- and 4th-century pottery
were recovered.

A number of postholes were found, including an
alignment which represented a fence running across
Trenches 2, 7 and 9. Other fence-lines may exist and
it is possible that the site was divided internally.

14 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

A
. x | ecua

a
111.82m0.D.

112.29m0.D.
TN

Silt loam

Compact chalk rubble Pottery Silty clay loam

Figure 10. Butterfield Down: plan and section of sunken-floored building 442 in Trench 4

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 15

Plate 2. Butterfield Down: corn drier 2267, Trench 22

After the excavation of Trenches 1-21, topsoil
was stripped from the whole area by the developer,
and this was called Trench 22. The use of a tracked
bulldozer made the recognition of archaeological
features difficult. Even so, some larger features were
recorded and planned (Figure 8). Four features
(2215, 2235, 2265, 2275), which were similar in
size, shape and alignment to the sunken-floored
feature, were augered and the depths at their
centres ranged from 0.35m to 0.43m. These
features may also have been cellars or sunken-
floored buildings.

The sole feature (2267) cleaned and accurately
planned within Trench 22 was a very well-defined
rectangular pit 2.7m long and 2.5m wide (Plate 2)
with a smaller circular element attached to one side.
The pit had mortared flint walls which were capped
by flat limestone slabs in places. The evidence from
the corn drier or malting kiln excavated subsequently
in Trench 23 suggests that pit 2267 may also have
been part of a similar structure.

Trench 23. Three sections were excavated across a
group of intersecting ditches in the north-east corner
of this trench (Figure 6). Although some of the
ditches contained late Roman pottery, the great
majority of diagnostic pottery from these features was
early Roman. An isolated shallow ditch (2801) to the
west of this group contained a quantity of pottery with
only one diagnostic sherd of the late Roman period.
However, in the absence of features exclusively of
1st-2nd-century date, the activities which took place
in this area are not clearly understood.

A large rounded feature (2816), 5m in diameter,
was sampled and found to be an irregular hollow
0.4m deep with two fills containing much pottery
along with animal bone, shell, ceramic tiles and
nails. The pottery included a quantity of early
Roman date but the presence of diagnostic fine
wares of the late Roman period suggests that the
earlier material was residual. Similar features at sites
of Romano-British and earlier date have been
considered to be ‘working hollows’.

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Corndrier 3020.

Flint

Limestone

Cream mortar

(Reversed section)

3020

Figure 11. Butterfield Down: plan and section of corn drier 3020 in Trench 23

112.14mO.D.
TAN

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 17

To the west of this feature, two sections were
excavated in and across a ditch (2346) adjacent to
the western edge of the trench. The ditch was 0.5m
deep and the lower fill contained late Roman pottery.
It cut a shallow steep-sided pit (2806) of late Roman
date, and also a smaller, undated, ditch only 0.05m
deep (2345). A deeper pit to the east (2813) which
was excavated to a depth of 1.25m had vertical sides
and was 1.75m in diameter. As with several features
considered above, although early Roman pottery was
recovered, the presence of a substantial amount of
late Roman material suggests a 3rd- or 4th-century
date for the infilling of the pit. A dump deposit in the
fills contained many fragments of cob walling. At the
lowest excavated level, a complete horse skull was
found directly on top of the substantial part of a
sheep skeleton. Although described here as a pit, this
feature could also be interpreted as a shaft or a well.

A rectangular feature (2810) to the east of the pit
was totally excavated and was very regular in form,
some 0.55m deep, with vertical sides and a gently
sloping base. Few finds were recovered from this
feature but it appears to be of late Roman date.

The terminal of a shallow ditch (2940) to the east
was also excavated, but only one, undiagnostic,
sherd was recovered. Another ditch terminal (2847)
to the south-west comprised part of a group of
ephemeral curving ditches enclosing a subcircular
area within which ditch 2346 terminated. Ditch
2847 was 0.2m deep with steep sides and a flat base.
A few sherds of late Roman pottery were found in
the single fill of the ditch, whilst a small metal bird
thought to be a sceptre-head (Figure 13) was found
on its surface. At the eastern terminal of the southern
segment was a small rectangular grave (2845) which
contained the remains of an infant aged approx-
imately six months. Only the cranium and long
bones were present but it is probable that the body
was complete when buried and that soil conditions
have destroyed all but the most resilient elements.
The stratigraphic relationship between the ditch and
the grave was not clear and no diagnostic pottery was
recovered from the grave. A second infant burial lay
in a shallow subrectangular grave (2952) to the
south-east. This burial was even less well-preserved,
with only a few teeth and bone fragments remaining.

Plate 3. Butterfield Down: corn drier 3020, Trench 23

18 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

No diagnostic pottery was recovered from the grave;
however, it is likely that both burials are of Roman
date. A shallow scoop (2942) close to the western
edge of the trench was also undated and its purpose
is unknown.

Further to the south-east was a circular pit
(2989), 1.5m in diameter, which was excavated toa
depth of 1.2m. The pit was filled with a series of
dump deposits of chalk rubble and flint nodules, and
occasionally a mixture of both. Only two sherds of
pottery were recovered, one of which was early
Roman. This pit was adjacent to the intersection of
two straight, shallow, but incomplete gullies which
may represent the corner of a building. To the north-
west, and on a similar alignment to one of the gullies,
was a wall-footing of crushed chalk faced on either
side by a single layer of flint nodules (3007). The
foundation was cut into the fills of a large hollow
(2987) made up of several intercutting scoops and
shallow pits. The wall-footings were only 0.1m high
and survived in two short sections forming a right-
angle, the flint nodules on the outside surviving as
only a single course. This probably represents the
corner of a building of which no other definite trace
was recovered; it probably survived due to the
unconsolidated fills in the hollow (2987) causing
subsidence which had made foundations necessary.
No datable pottery was recovered from the foundation
trench but late Roman sherds were found in the fills
of the hollow, along with some early Roman
material.

In the south-east corner of the trench a
subrectangular feature (3020) 4.85m long and 3.2m
wide was aligned north-west/south-east. Its eastern
side had been cut by a modern service trench. Upon
excavation the feature was found to be a well-
constructed T-shaped subterranean structure (Figure
11 and Plate 3), a typical example of a corn drier or
malting kiln, possibly sited within a rectangular
building.

Access to the stokehole area, approximately one
metre below present ground level, was by a series of
steps cut into the chalk at the western side. The stem
element of the T-shaped flue was 3m long and 0.4m
wide; the axial element was 2.3m long and 0.35m
wide. The fire-pit was indicated by a more heavily
burnt patch of chalk bedrock adjacent to the
stokehole, although lesser amounts of burning were
recorded throughout the flue.

The hot air passed along the flue to heat a
rectangular pit 3.2m long, 3m wide and 1m deep.
Walls of flint nodules set in a compact mortar (3018)
lined the pit and the eastern wall continued beyond

the pit to line the edge of the stokehole: these walls
were internally faced with a single layer of flint
nodules. Several small stakeholes were cut into the
top of the east and west walls.

On each side of the fire-pit a short section of wall
(3016) had been added, perpendicular to the
sidewalls. These later walls were slightly wider and
faced with courses of flint nodules and chalk blocks.
Only two courses were recorded, placed on a ledge of
chalk bedrock and clearly distinct from the walls
lining the rectangular pit. They extended towards the
fire-pit for approximately 0.55m but a pink sandy
stone mortar with a few flint nodules (3015),
concentrated mainly at the base, was bonded onto the
end of them. These walls overhung the fire-pit slightly
and as flat slabs of limestone were recorded in the
upper part (Figure 11) it is likely that these were part
of an arch over the fire-pit at the mouth of the flue.

Beyond the flue arch walls and either side of the
flue, the interior of the structure was excavated to a
depth of 0.3m below the present ground level. At
this depth there was a compact surface of light grey-
brown mortar (3017) within which were occasional
flint nodules. This surface is likely to have been the
base for a raised drying floor of some sort but no
evidence was recovered to indicate the nature or
position of such a floor.

The base of the flue was filled with a dark grey
deposit of ashy material which lay directly on the
chalk bedrock. Along the stem of the “T” this deposit
was 0.15m deep but in the axial element it was slightly
thicker, rising sharply to a depth of 0.3m at each end.
Six samples of this ash were collected at regular
intervals along the flue and all contained carbonised
barley and wheat which had not germinated. Weed
seeds were also present in two samples.

Considerable quantities of mortar and building
materials were recovered from the filling of the flue,
along with large amounts of pottery, mostly of late
Roman date, and animal bone, indicating the demo-
lition or collapse of the superstructure and deliberate
refuse disposal.

Adjacent to the western edge of the corn drier
was a small dumbbell-shaped oven (2322). The
western element of this was a shallow subcircular pit
with a flat base which was cut directly into the chalk
bedrock. As this showed evidence of burning it may
have been the stokehole. It was linked directly to the
eastern element, a more regular circular pit, 1m in
diameter and 0.25m deep. This was lined with a
0.15m thick layer of orange-brown clay within which
some smaller flint nodules were visible. The internal
surface of the lining had been fired to a yellow colour

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 19

with a much harder texture. An ashy deposit was
found in the northern part of the base of the
chamber but it was not examined and no diagnostic
pottery was found.

Several other similar, though less well-preserved,
dumbbell-shaped features were recorded (Figure 6:
2469, 2470, 2617, 2620, 2841, 2748 and 2809). At
the eastern end of 2809 a single plain /ydion brick
had been used as the base of the oven or hearth.

As in the earlier excavations, alignments of
postholes partitioning the site were recorded. Clear
examples can be seen in the south-east part of the
trench, where one alignment runs parallel to, and
south of, trackway 2492, while two other alignments
form a right-angle to the north-east of corn drier
3020.

Other areas: The watching brief carried out during
machine-stripping of two further areas in the
southern part of the development site resulted in the
planning of many more features (Figure 3, Trenches
25 and 26). Although none of these was excavated,
almost all of the pottery observed on the surface of
the features was of Romano-British date. The plan of
the features in these trenches suggests a continuation
of the settlement already investigated in Trenches
1—23, with further trackways, ditches, fence-lines
and, possibly, sunken-floored buildings.

In the south-west area which was subjected to
geophysical survey and trial trench evaluation, the
dense evidence of Romano-British occupation
recorded to the east was not present (see Figure 4).
Instead, a number of linear features were located and
excavated, and it appears likely that this area contains
several small ditched enclosures located adjacent to
the settlement. These are similar in size and form to
the enclosures recorded by aerial photographic
survey in the eastern part of the development area
(Figure 1, C) and partially investigated by excavation
(Figure 5). Ditch 15 here was shown to be 0.75m
wide and 0.35m deep, with moderately sloping sides
and a slightly irregular base. It formed the north-
eastern side of a square or rectangular enclosure,
45m wide. Three sherds of Romano-British pottery
were found in the fills of this ditch. A second ditch
(26) was curvilinear in form and was also of
Romano-British date, but does not appear to be part
of the group of enclosures.

The finds

The following accounts are summaries of fuller
reports in the site archive. Artefacts recovered by

fieldwalking are only mentioned when they are
considered to be of intrinsic interest.

The Roman gold coin hoard
by A. BURNETT

In October 1990 a small hoard of eight gold (and
possibly one silver) coins was discovered by a local
metal-detector user approximately 90m to the north
of Trench 22 (Figure 1, C). A full description of the
coins is published in Burnett 1992.

Two joining sherds of pottery, probably from the
vessel within which the coins were deposited, were
also found (Figure 16, 8) and showed signs of recent
breaks. The pot is a small New Forest colour-
coated, plain, globular beaker (Fulford 1975a, Type
30.12), a type which became more common after
AD 340-50 (Fulford 1975b; 1979).

The Butterfield: Down hoard comprises eight
gold solidii and possibly one silver siliqua. There are
solidii of four emperors, one of Gratian (AD
367-83), two of Valentian II (AD 375-92), two of
Arcadius (AD 383-408) and four of Honorius (AD
393-423). Analyses of similar coins indicate that the
gold content is high (c.98%).

Analysis of the silver siliqua of Arcadius (AD
383-408) indicates that it is of fine silver (90%). It
is rare to find single silver coins in hoards of gold
issues of this period, and it is the only silver coin
known to have been found at Butterfield Down. It
was found at the same time as the gold coins and is
of the same date, although it is more worn. A date of
about AD 405 is suggested by the presence of the
latest coin in the group, making this one of the latest
coin hoards known from Roman Britain.

Following the discovery of the hoard, a 2.5m’
trench (Trench 24) was hand-excavated in the
area of the findspot in order to recover any further
finds or subsoil features within which the pot
might have been placed. Since, however, the finder
was unable to re-identify accurately the exact
findspot, it may be that the trench was dug outside
the area of the hoard. No further gold coins or
other parts of the pot were found, but the edge
of a group of small features was revealed, along
with three stakeholes. None of the features seemed
to have been disturbed so if the hoard vessel
came from within the area excavated it is likely
that it was located at the base of the present
ploughsoil. Although two copper alloy coins were
found within Trench 24, they were not made
available for study.

20 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Other Roman coins

by M. CORNEY

In addition to the eight or nine coins from the hoard
and the bronze issues found at the same time, a
further 925 coins were recovered from the general
area of the Down by metal-detector users. All have
been examined and, where possible, full identification
has been made. Two hundred and thirty-three (25%)
were illegible, the high percentage reflecting the
crude cleaning methods employed.

300AD7

Irregular
Regular

200AD 4

Date

100AD 4

xi Xb XVa XVI

eo Nana Seve

"Vinay vin 4 xp

Issue Period

Figure 12. Butterfield Down: Roman coins by coin period

The numbers of coins by issue period (Reece
1972) are presented in Figure 12 (including those
recovered from controlled excavation). The pattern of
loss shows little evidence for coin use from the 1st—3rd
centuries AD. A dramatic increase occurs from the
middle of the 3rd century AD with a high rate of coin
use and loss continuing into the early 5th century. In
period X (AD 259-275) 59 coins (52% of the period
total) are barbarous issues. The peak occurs in the
period AD 364-378 with a total of 288 coins
representing 40% of the identifiable assemblage. Con-
tinued coin use into the early 5th century AD is well
represented with issues of period XVI (AD 388-402)
representing 9% of the total — a high figure which

conforms to the general pattern of late Roman activity
in the central Wiltshire region (Corney, in preparation).

Other numismatic evidence, perhaps related, for
late Roman activity in the area east of Amesbury
comes from the discovery, in 1842 or 1843, of a
mixed hoard of silver rings and bronze and silver
coins ranging from Postumus (AD 259-268) to
Theodosius II (AD 375-392) from New Covert,
500m south-west of Butterfield Down (SMR 305
above; VCH 1957, 30).

Objects of silver and copper alloy
by S.M. DUGGAN

A badly damaged undecorated silver finger ring with
an average diameter of c.19mm was found by a
metal-detector user. Although likely to be Roman, a
closer date cannot be assigned. Some 53 copper
alloy objects were found, only eight of which were
recovered during excavation, five in fieldwalking and
the remainder by metal-detector users. They are
considered by type below and unless the context is
given, the piece was found by metal-detector users.

Four of the six brooches are early Roman and
include Dolphin, bow, and Lamberton Moor types.
A large trumpet brooch with a lionesque knop is
likely to be of late 2nd- or early 3rd-century date; a
similar example was found at Exeter, Devon
(Holbrook and Bidwell 1991, fig. 102, 27). The final
brooch is T-shaped with a head-loop and a lozenge-
shaped plate inlaid with rhomboidal and clover-leaf
and dot shaped mouldings which may have been
enamelled. This brooch shares several characteristics
with the Flavian Caerleon type which is found in
south-west Britain (Collingwood and Richmond
1969, fig. 103, 28-9; Wedlake 1982, fig. 53, 55) but
the lozenge-shaped plate is unusual and _ is
comparable with flat rhomboidal brooches of 2nd
century AD date (Crummy 1983, fig. 14, 78).

Three finger rings and one key ring were found. A
finger ring from the upper fill of the ‘working hollow’
2816 in Trench 23 has a well-preserved bezel,
containing an amber-coloured glass or enamel setting,
and hoops on either side which include a winged
moulding. Similar examples of 3rd- or 4th-century
AD date have been found at Colchester, Essex
(Crummy 1983, fig. 50, 1777-85) and Cirencester,
Gloucestershire (Viner 1982, M2; fig. 54, 37).

Pieces of four separate armlets were found, all 3rd-
or 4th-century types. A fragment of plain armlet was
found within one of the fills of the corn drier 3020,
while a section of a two-strand plaited wire armlet was
found on the surface of the fill of ditch 2333.

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 21

A single broken probe, or possibly a pin, was
found in ‘working hollow’ 2987 in Trench 23;
comparable examples come from Colchester and are
thought to be of early 2nd-century date (Crummy
1983, fig. 65, 1929-32). Fragments of four belt
plates or buckles have been found, one of which is
possibly Ist century AD (zbid., fig. 151, 4211), but
the others are likely to be post-Roman. A spoon with
a pear-shaped bowl and three linear incisions at the
base of the handle is paralleled (albeit without
incisions) by a find from Colchester (zbid., fig. 73,
2014) dated to the early 2nd century. Part of a spoon
with a round bowl of a type usually dated to the late
Ist and 2nd centuries AD was found in corn drier
3020.

In addition to the figurine identified as a sceptre-
head, discussed below, there are fragments from four
copper alloy figurines. Three of these are single feet:
one a claw with four nails, one an animal foot with
five ‘toes’, and the third badly worn and indistinct.
The fourth piece is a finely modelled bust of a
woman with her hair gathered into a knot at the nape
of her neck. As the base is sharply angled, it is clear
that it was attached to another piece, possibly a stand
of some form.

Finally, a pendant mount from a rare type of post-
medieval sword-belt fitting was also discovered (cf.
Gaimster 1988).

Sceptre head
by M. HENIG

A small copper alloy figure of a bird perched atop an
iron rod (Figure 13) was found in the upper part of
ring ditch 2847 in Trench 23. The bird is 43mm long
from beak to tail and weighs 24g. It has a long curved
beak and a rather small head, flattened at the top
with a suggestion of brows above the eyes. The wings
are folded upon its back, the pinions being indicated
by means of long grooves, and it has a squared-off
tail. The bird is probably intended to be an eagle,
although the bill is more like that of a chough.
Comparison may be made with figurines of
birds from the temple site at Woodeaton (Kirk
1949, 31, nos. 4-5) and from Ramsden, near
Finstock (Henig and Chambers 1984, fig. 1, 1),
both in Oxfordshire. An eagle in a cache of religious
paraphernalia found at Willingham Fen, Cam-
bridgeshire, may also be noted (Rostovtseff 1923,
94, pl. iv, 5). The feet are missing in all these cases
SO it is uncertain whether they were votives given to
deities (referring to the widespread belief in augury
whereby birds as denizens of the air could reveal the

0 10 20 30

Figure 13. Butterfield Down: copper alloy bird on iron
mount from Trench 23

will of the gods in their flight patterns) or were
elements in regalia.

The uppermost element in a priestly head-dress
in the hoard from Felmingham, Norfolk (Gilbert
1978, fig. 5A) supported a bird whose foot alone
remains. However, much more pertinent to the
Butterfield Down eagle is a bird from the same
cache, probably a corvid, standing upon a globe
which was evidently fixed on an iron staff (ibid.,
fig. 4D). This was surely a sceptre head or tipstaff
of the same type as the eagle, designed to be
carried in religious processions. The shafts of the
Butterfield Down and Felmingham Hall sceptres
may have been of bare iron, but it is perhaps
more likely that copper alloy sheeting with an
iron armature gave strength and solidity to the
object.

The only other bird sceptre known from Roman
Britain is an owl from Willingham Fen, Cambridge-
shire (Rostovtseff 1923, 94, pl. iv, 4). Sceptres
topped by birds are best known in the form of
the eagle-sceptres of the Roman emperors (Strong
1976, pl. 127; Kent 1978, pl. 549). A number of
sceptre heads have been recognised in Roman
Britain, mainly representations of deities though
some may portray emperors (Henig and Leahy
1984). It is probable that they would have been used
away from shrines in processions designed
to bless the fields and propitiate the gods who looked
after the community. Consequently, the Butterfield
Down sceptre is not out of place in an otherwise
secular context, and it emphasises the important
part religious ceremonies played in daily life.

22 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Iron objects

by M. FAIRBROTHER

Some 43 iron objects and 143 iron nails were
excavated but none have been radiographed or
conserved. Nine small cleats were found, six small
oval examples (cf Manning 1985, pl. 61, R60-4) and
three longer, narrower, oval ones (ibid., pl. 61,
R54-9), all probably from the soles or heels of boots.
Other items for personal use include a pin with a
domed head, and a stylus found in corn drier 3020
of a type thought to be early Roman in date (zbid.,
fig. 21, Type 1). The five knives were of different
types (ibid., fig. 28, Types 10, lla, 13, 16 and 21)
and although those which could be dated appear in
the mid Ist century AD, the types are long-lived.

A shackle, perhaps for animals, consisting of a bar
forming two-thirds of a circle, with eyes at the ends,
one holding a round loop and the other a long narrow
loop bent into a shallow V-shape was also found (zbid.,
fig. 23, 7). The shackle would have been closed by
passing the narrow loop through the round loop and
securing it with a chain or padlock. The heel portion
(including the hook) of a hipposandal was also found.

There are few tools — only three wedges, one of
which was associated with a fragment of a heavy iron
bar but does not seem to have been attached to it.
Pieces of structural ironwork comprise a joiner’s dog
(tbid., pl. 61, R52), a heavy metal bolt, a possible
loop fitting and a double spiked loop (ibid., pl. 61,
R34-50). The remaining ironwork comprises six
pieces of strip binding, a single horseshoe and an
assortment of pieces of wire, rods or bars.

Metalworking slag
by ANDREW CROCKETT

A total of 28 pieces weighing 1581g was recovered
from the site. The most common type is ferrous and
quite dense, deriving from smithing rather than
smelting. Not enough material was found to identify
firmly any areas used for metalworking but it is
worth noting that slag was found in two ‘working
hollows’: 2973 and 2987 in Trench 23.

The worked flint from pit 2
by P.A. HARDING

A small assemblage of worked flint was recovered
from the feature which contained portions of three
separate Beakers (Figure 5, pit 2). Most of this flint
was found in the upper fills of the pit, associated with
Vessel 1. ‘The assemblage is in mint condition and

the presence of chips suggests a date contemporary
with the Beaker.

No refitting was possible, although distinctive
cortex forms and patterns in the flint indicate that a
minimum of two nodules may be _ represented.
Proportionally, there is more broken material than
unbroken, which suggests that knapping took place
nearby. The chips are uncharacteristic of those
produced during platform preparation which is in
accord with the flakes, only one flake showing
evidence of platform preparation. The chips are better
interpreted as accidental by-products of knapping.

Four flake tools were present. An end scraper
made on a long broken flake (Figure 15, 4) had been
retouched at the distal end by semi-abrupt, direct,
regular flaking to a short convex edge. An end scraper
made on a thin non-cortical flake (Figure 15, 5)
showed semi-abrupt, regular, direct discontinuous
retouch and some inverse retouch, especially towards
the proximal end. An end scraper made on the
proximal end of a thin, broken flake (Figure 15, 6)
showed semi-abrupt, irregular, direct retouch
forming a convex edge, whilst another flake had been
retouched along one edge (Figure 15, 7) with
marginal, inverse flaking to form a straight edge.

There is an insufficient quantity of flint in the
assemblage to allow firm conclusions to be drawn.
However, despite the absence of any diagnostic tool
types, the overall character, typology and condition
of the material suggest that it is contemporary with
the Beaker pottery. The fairly high proportion of
tools and the scarcity of cores may indicate domestic
or ritual activity rather than industrial production.

Portable stone objects

by A.P. FITZPATRICK and J.I. MILLARD

Prehistoric

A rough-out for a Neolithic stone axe was found
by a metal-detector user. It was thin-sectioned by R.V.
Davis on behalf of the Council for British Archaeology
Implement Petrology Committee (ref. 1858/W 1428),
revealing a highly altered medium grained gabbro. The
rock is probably Cornish, and possibly from. the
Falmouth area. This may suggest that the piece is of
earlier rather than later Neolithic date.

A rectangular, decorated chalk object (Figure 14;
Plate 4) was recovered from the upper fill of pit 310
in Trench 3. There is incised decoration on the two
faces and on three of the edges, the fourth being too
worn to establish whether the decoration extended
all over the object. No parallels of Roman date are
known to the writers but the piece is similar to the
two Late Neolithic plaques from the Chalk Plaque

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 23

Pit, Amesbury (Harding 1988; Varndell 1991) anda
date in the Late Neolithic or Early Bronze Age seems
likely, particularly as the decoration on the plaque
appears to echo that on Beakers (Lawson 1993).
Forty-four quernstone fragments from 17
separate querns were recovered. Of these, 16 were
small hand querns, with a diameter of less than
0.5m, in conglomerate, coarse and fine sandstone,
and coarse limestone. The other stone is a Greensand
millstone with a diameter of 0.62m, and is a type
which is usually associated with mechanically
worked mills (Cunliffe 1971, 153, fig. 71). This was
found on the surface of an unexcavated feature in the

north-eastern part of Trench 23, close to a group of
ditches which may be early Roman.

Further quern fragments were recovered from a
range of contexts, including the fills of corn drier
3020. Five limestone mortar fragments were found,
two of which may be of Purbeck Marble. They are
the same size and type as an example from
Colchester (Buckley and Major 1983, fig. 79, 2804);
dished profiles and smoothed interior surfaces are
evident on four of the five fragments, one of which
had an extant lug. Seven whetstone fragments were
found in a variety of features; all had two surviving
surfaces, at least one of which was polished.

| WA /SEJ

Figure 14 (above) and Plate 4.
Butterfield Down: incised
chalk plaque from Trench 3.
Scale 1:2 and actual size

24 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Shale object
by ANDREW CROCKETT

Two-thirds of a solid circular shale object, 27mm in
diameter and apparently lathe worked, was found in a
rubble fill within corn drier 3020. One surface is flat,
the other slightly domed, giving a thickness of 6mm at
the centre, and the piece would have weighed c.6.4g.
Although the piece is broken, it appears to be a gaming
counter. Bone, glass, stone and ceramic gaming pieces
or counters are well known (Crummy 1983, 91-6;
Holbrook and Bidwell 1991, 229, 267, 275, 279) but
the use of shale for such pieces is not paralleled at the
sites referred to or in the large assemblage of shale
objects from Silchester (Lawson 1976).

The glass
by D.A. ALLEN

A total of 22 glass vessel fragments and one glass
bead was found. Almost all are of colourless or
greenish colourless glass, with only two small blue-
green chips and one dark blue fragment which,
together with the absence of blue-green bottle
fragments, is typical of a later Roman collection.

The dark blue fragment was the only piece of
typically Ist century AD tableware present. It was
found in one of the upper fills of corn drier 3020 and
although the precise form cannot be determined, the
strong colour and optic-blown ribbed decoration
were commonly used for jugs, jars and bowls during
the second half of the Ist century AD.

From the Flavian period onwards, colourless
glass replaced bright colours for tablewares and
several fragments of this type were found. One of
these, from the lower fill of pit 404 in Trench 4, is the
rim of a very common later 2nd- and 3rd-century
cup (Isings 1957, 85b; Allen 1988, 293, no. 44).

Fragments from later Roman containers were also
found. One piece, from the surface of an unexcavated
feature 2756 in ‘Trench 23, was probably from a
mould-blown barrel-shaped bottle of a type often
made and signed by Frontinus or Felix (e.g. Harden
et al. 1968, 44, no. 79, from Faversham, Kent) and
quite common during the later 3rd and 4th centuries.

A small green square-sectioned bead was found
on the surface of an unexcavated feature 2699 in
‘Trench 23 and is a common Roman type of 3rd—4th
century date (Guido 1978, fig. 37, 6-7).

The pottery from pit 2
by ROSAMUND MJ. CLEAL

Three Beakers are represented in the assemblage
recovered from pit 2, one by a single sherd, one by

two conjoining sherds, and the other by half to two-
thirds of a vessel. The last is made up largely of
conjoining sherds, and it seems likely that the vessel
was complete when placed in the feature and has
subsequently suffered some destruction. Sherds
were examined at x20 magnification, following
standard Wessex Archaeology procedures.

Vessel 1

This vessel is represented by a large, crushed portion

of a Beaker (Figure 15, 1) and a sherd count is

therefore not relevant. The total weight of the
surviving portion of the vessel is 2949g, and at least
three-quarters of the Beaker is present.

Fabric: hard, with smooth surfaces and a hackly

fracture. It contains the following inclusions, of

which grog is the most obvious:

Grog Sparse to moderate (<15%), ill-sorted,
sub-rounded to sub-angular fragments
of grog. Mostly fine (<lmm) but
some as large as 5mm; in the largest
fragments inclusions of grog and
quartz sand are visible.

Flint Sparse (<3%), well-sorted, angular
fragments (<4mm), not well-prepared.
Shell Occasional (i.e. not present in every

sherd), <3mm.
Quartz Sand Moderate (<15%), well-sorted, rounded

grains (<lmm, most <0.5mm).
Firing: the vessel is patchily oxidised on the exterior
to shades of red and orange; the core is mainly
oxidised as is the interior surface.
Decoration: the decoration is incised and comprises
horizontal zones of filling, including a deep, herring-
bone-filled, running chevron; ladder-patterns; lattice;
and a narrow, filled, running chevron. It is also clear
that a second, narrow, filled, running chevron had
been planned on the upper belly, but this had been
rejected after laying out in outline around at least
some of the circumference. This was then replaced
by a plain zone with a wide zone of lattice below, the
outline of the running chevron not being entirely
erased. The decoration, although complex, is not
executed with particular care, and is not consistent
around the circumference of the vessel.
Condition: the sherds are in good condition, with
little wear on the surfaces. There are traces of a
carbonised residue on the interior surface.

Vessel 2

‘Two sherds comprising one rim sherd (36g) and one
body sherd (24g) of a comb-decorated Beaker (Figure
15, 2). The exterior rim diameter is c.170mm.

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN

TCADA oar

SRI OL
YS GSN
YS Yeh

iT i

(td

ae
\

BIN

th
NSH

VY
mn
|
)

0 100
{+s —— gee “jeer 11 17)

100

jee aE EE 1 7

As
\A s é A Pottery
ili Vix WN
KI h it NY (\: WOAe Flint
is Pa By

Figure 15. Butterfield Down: Beaker pottery and associated flints from pit 2

Fabric: moderately hard with rough, poorly-finished
surfaces and a hackly fracture. It contains the follow-
ing inclusions, of which bone is the most obvious:
Bone Sparse to moderate (5-10%), well-
sorted, angular fragments. Maximum
size <5mm, most <3mm, unevenly
distributed. The pieces are white,
soft and appear to be well-calcined,
indicating that they were calcined in
an oxidising atmosphere: this must
have occurred prior to their being
added to the fabric as the pot is only

Flint

partially oxidised. In particular, some
very well-calcined pieces only show
in broken edges of the vessel within a
matrix of unoxidised body core. The
fragments of bone cannot be differ-
entiated as either animal or human
(J.I. McKinley pers. comm.). The
fragments are unevenly distributed
throughout the fabric.

Rare (<3%), ill-sorted, angular
fragments. Maximum size <10mm,
most <5mm. Not well-calcined.

26 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Pebbles Rare (<1%), ill-sorted, rounded.
Maximum size <7mm. These are
probably flint.

Moderate (c.10%), well-sorted
rounded grains, most <0.5mm.
Sparse (5%), well-sorted, rounded,

Quartz Sand

Iron Oxides/

Glauconite dark grains. All <0.5mm but some
too small to measure at x20
magnification.

Grog There are some small fragments of

commiunuted potsherd (grog) present,
but these are hard to differentiate
from the matrix. Maximum size
c.4mm, rounded, sparse to moderate
(<?10%).
Firing: the vessel is patchily oxidised on the exterior
with one area of clear orange colouring 70mm from
the rim. The remainder of the exterior is pale
orange-brown, indicating only partial oxidation. The
interior is partially oxidised to pale brown, and the
core is black.
Decoration: this comprises rectangular-toothed comb
impressions, but the worn state of the impressions
did not allow the tooth length to be established. Comb
size varies but seems to be about 2 x Imm. It is possible
that a shorter comb was used for infilling than for the
outlines, but this is not certain. There is a horizontal
band of ladder pattern below the rim, separated from
it by a single horizontal line of impressions. A
reserved chevron motif runs around the neck of the
vessel: the background is filled with horizontal lines
of impressions. This is motif 32ii of Clarke’s
Southern British Group 4 (Clarke 1970, 427).
Condition: both the edges and the surfaces of the
sherds are weathered. On the exterior the weathering
is severe enough to obscure some of the comb
impressions.

Vessel 3

This vessel is represented by a single sherd weighing

34¢ (Figure 15, 3).

Fabric: hard with a hackly fracture. The exterior

surface is rusticated and the interior smoothed. It

contains the following inclusions, of which bone is
the most obvious:

Bone Sparse (<5%), well-sorted, angular
fragments <lmm. The fragments are
soft, white, and do not react with
dilute (10%) HCL. They are unevenly
distributed.

Flint Sparse (<5%), well-sorted, angular
fragments <2mm, not well-calcined,
and unevenly distributed.

?Sandstone Occasional (only one fragment
observed) whitish inclusion made up
of sub-angular quartz grains in an
opaque creamy white, non-
calcareous matrix. The fragment is
very friable.

Rare (<3%), small dark grains, some
black, some reddish. Some at least
do not respond to a magnet and may
therefore be glauconite rather than
iron oxides. The grains are fine and
mainly too small to measure at x20
magnification.

Sparse (<c.7%) fragments which are
difficult to distinguish from the clay
matrix and mainly small (<1mm).
The matrix is slightly micaceous,
with rare fine mica, too small to be
measured at x20 magnification.
Quartz Sand__ Rare fine grains, <0.5mm.

The fabric is similar to that of Vessel 2 but shows
greater attention to preparation of the additives, with
the smaller size of the bone fragments being
particularly noticeable.

Firing: the exterior is fired to a pale orange-brown,
the interior to mid-pale brown, and the core to
black. The hint of orange in the exterior surface
colour suggests that the fully oxidised colour of the
clay would be a clear orange. The core of the sherd is
obscured by a tarry deposit along the broken edges.
Decoration: this 1s plastic and comprises fingernail
pinching in vertical ribs, with a zone of horizontal
decoration which almost certainly consists of
horizontal ridges also formed by fingernail
pinching, although this is only preserved in a small
area of the sherd. It is not clear from the sherd
which way up it should be viewed, but the convexity
of the vessel profile in the area covered by vertical
ribs strongly suggests that this is the belly of the
vessel.

Condition: the sherd shows some wear, particularly
on the interior surface where the partially oxidised
surface has partly worn away. The edges are
abraded but show a quite distinct, tarry, carbonised
deposit. This is also present where the sherd is
freshly broken, suggesting that it is present
throughout the wall in this part of the vessel. Likely
explanations for this deposit are either that the clay
was highly carbonaceous and that this part of the
vessel has carbonised during firing or use, or that an
organic material has been absorbed by the vessel
wall during use and subsequently carbonised during
post-firing heating.

Tron Oxides/
Glauconite

Grog

Mica

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 21.

OTHER PREHISTORIC POTTERY FROM NEAR PIT 2

A single, plain, body sherd (3g) in a grog-tempered
fabric was recovered from the basal fill of ditch 18
(Figure 5). The exterior surface and half of the core
were fired orange, the interior surface and the rest of
the core were black. This sherd is almost certainly
from a coarse Beaker.

A single rim sherd (23g) in a soft, grog-tempered
fabric was found in the upper fill of the same feature.
The sherd is undecorated and the rim is internally
bevelled. This is probably from a Collared Urn,
although the base of the collar is not present. A few
sherds of Romano-British and later pottery were also
found in this context.

DISCUSSION

The Beakers are not certainly associated, as there is
some doubt that Vessels 2 and 3 were deposited at
the same time as Vessel 1. On the grounds of the
extremely unusual fabrics of Vessels 2 and 3, these
two Beakers at least are likely to be contemporary. It
seems unlikely that Vessel 1 should represent
separate use of such a small feature, but not
impossible. There is no human bone, either unburnt
or cremated, with the deposit and it must be
assumed that it is not a funerary one.

The occurrence of bone as a tempering material
is extremely unusual, but not unique in the Neolithic
and Bronze Ages. Seven cases of Neolithic and
Beaker date are cited by Smith and Darvill as being
known at the time of their writing (1990, 152), of
which one at least is certainly Beaker (from Lough
Gur, Co. Limerick).

In terms of Clarke’s (1970) classification, Vessel 1
exhibits motifs 4 and 5 (ladder pattern and lattice-
filled band) of the Basic European Motif Group, a
herringbone-filled version of motif 27 (deep, filled,
running chevron) of the Late Northern British Motif
Group, and motif 33 (filled, running chevron) of the
Southern British Motif Group. The use of a
Northern Group motif is interesting, but the vessel
shows no other attributes diagnostic of that group,
and it is difficult to see it as other than a Southern
tradition vessel. Within that tradition it is assignable
to Clarke’s Developed Southern British Group (S2),
in which the filled neck/zoned belly style of
decorative organisation, present on Vessel 1, is
especially characteristic (Clarke 1970, 210). A
feature of the apparently domestic assemblages
which include Beakers of this group is large
rusticated Beakers which display horizontal rows

and vertical columns of finger-pinching (zbzd., 1970,
214), such as occur on Vessel 3.

Vessel 2 may be ascribed to either Clarke’s
Developed Southern (S2) or to his Late Southern
(S3) Beaker Groups; as the lower body is missing it
is not possible to establish whether the belly is filled
or zoned, and this is one of the criteria for separating
these groups.

In terms of Lanting and van der Waals Steps, only
Vessel 1 is certainly assignable, to Step 6 (Lanting
and van der Waals 1972). However, both Clarke’s
classification, and that of Lanting and van der Waals,
can no longer be seen as useful in determining the
likely date of a Beaker, since the British Museum
Beaker dating programme has failed to provide
support for either scheme from well-associated
radiocarbon determinations, and in their stead it
may only be suggested that there is a clear time band
into which Beaker use falls, approximately between
2600 and 1800 cal BC (Kinnes er al. 1991). The
Clarke scheme remains useful in that it indicates
broad patterns of, particularly, decorative similarity,
and of associations between types of Beaker, as in
this case, where it indicates that it is not unusual to
find Beakers similar to Vessels 1 and 2 with
rusticated vessels such as Vessel 3, on apparently
domestic sites.

The other pottery
by J.1. MILLARD

Introduction

The pottery assemblage comprises 6394 sherds
(124,567g), the majority of which are of Romano-
British date, with prehistoric and post-medieval
material also present (Table 1). Analysis followed the
standard Wessex Archaeology analytical recording
systems (Morris 1992a; 1992b), and for this purpose
was divided into two parts. First, a brief scan of the
pottery from Trenches 1—22 and surface collection
from the unexcavated features in Trench 23 was
undertaken. Pottery from excavated features in Trench
23, which has the greatest stratigraphic integrity, and
from Trench 24 (the hand-excavated trench around
the gold coin hoard) was then recorded in detail.

Pottery recorded by scanning

The presence of Roman wares of known type or
source, for example Black Burnished ware, and all
pottery not of Roman date (prehistoric, medieval/
post-medieval) was recorded, but detailed analysis
was not conducted. The remaining Roman pottery
was divided into broad fabric groups on the basis of

28 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

the dominant inclusion type: grey and oxidised
sandy fabrics, grog-tempered fabrics, flint-gritted
fabrics, and fine wares of unknown type. The
number and weight of sherds by fabric group for this
part of the analysis are shown in Table 1. In addition,
18 rim types were identified and recorded on a
presence/absence basis (Table 2).

Pottery recorded in detail

Methodology. The prehistoric and Roman pottery
from features in Trench 23 and from Trench 24 was
divided into five broad fabric groups on the basis of
the dominant inclusion or known source: flint-gritted
fabrics (group F), shell-tempered fabrics (group S),
grog-tempered fabrics (group G), sandy fabrics

Table 1. Quantification of pottery recovered up to 1992 by fabric

Type scanned detailed
recording recording
n0. weight fabric n0 weight
Prehistoric 10 69g Fl 1 22g
F2 2 llg
F3 af 38g
F4 14 28g
Gl 2 7g
Sl 3 14g
Grey wares 1820 20039¢ Q100 1161 18136g
Q101 344 5179g
Q102 123 1354¢
Q103 46 533g
Oxidised wares 297 3801¢g Q104 160 2290g
Q105 29 440g
Q106 14 114g
Q107 3 63g
Q108 5 45g
Grog-tempered 574 20094¢g G100 321 12819¢g
G101 4 80g
Flint-gritted 3 150g
Savernake ware 4 44¢
Black Burnished ware 438 5732g BB1 331 4152¢g
BB1 (var) 4 214g
Oxfordshire 132 1345g CC 51 387g
White ware 8lg
Parchment 3 6g
Mortaria 21 1023g
Oxidised 4 llg
New Forest 172 1815g Parchment 3 77g
Red-slipped 15 129g
Stoneware 82 785g
Greywares 1 23g
Rhenish 1 2g 9 57g
Samian 20 191g 29 384g
Amphorae 11 1609g 35 17500g
Fine wares unknown source 17 07g Q110 8 190g
Q112 1 2g
2 2¢ Q113 8 47g
Q114 3) 25g
Medieval/Post-medieval 9 107g 4 12g
Total 3510 55807¢g 2884 68700g

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 29

(group Q), and fabrics of known type or source
(group E). The first four groups were then sub-
divided into separate fabric types, according to the
range and coarseness of inclusions present (Table 3).
Details of vessel form, surface treatment, and decor-
ation for the Roman pottery were also recorded.

Prehistoric pottery

A small quantity of Late Neolithic and Late Bronze
Age pottery was recovered from Trenches 3, 4, 13, 17
and the surface of Trench 23. Twenty-nine prehis-
toric sherds were recovered from excavated features
in Trench 23. Six fabric types were identified: four
flint-gritted (F1—F4), one grog-tempered (G1), and
one shell-tempered (S1). All of these fabrics were
represented by very small numbers of sherds. In most
cases the lack of diagnostic material precludes close
dating; the majority of it was found redeposited in
later contexts.

Two sherds (fabric F2) bear impressed decoration
and have been identified as Late Neolithic Peter-
borough ware in a fabric similar to that already recov-
ered in the Stonehenge area (Cleal with Raymond
1990, 235). One sherd (G1) may derive from an
accessory vessel (Figure 16, 1), probably an incense
cup of Early Bronze Age date (cf Annable and
Simpson 1964, no. 445; R.M.J. Cleal pers. comm.).
A second sherd in the same fabric is also likely to be
of Early Bronze Age date. The single sherd of fabric
F1 derives from a thick-walled vessel, most probably
an urn of Middle Bronze Age Deverel-Rimbury type,
and is similar to the Early/Middle Bronze Age fabric
types recovered in the Stonehenge area (Cleal with
Raymond 1990, 241). All other handmade sherds are
prehistoric but not datable more closely.

All sherds except the fragment of ?incense cup
were found with later material and must therefore be
considered as residual. The fragment of the ?incense
cup was found beneath the body in burial 3004
(Trench 3). As it is a tiny fragment it may simply
have been incorporated in the fill of the grave but it
is considered as a grave good here.

Roman pottery

Thirty-seven fabrics were identified and quantified
(Table 1) and the coarse and fine ware fabrics are
listed in Table 4. Eighteen coarse (Table 2) and one
fine ware rim types were identified and quantified
and the six most common rim types are illustrated
(Figure 16, 2-7). Imported fine wares are repre-
sented by Rhenish ware and samian. The Rhenish
ware sherds are all from Trier-type beakers, produced
between AD 150-250 (Greene 1978, 18).

No attempt was made to attribute the samian to
production centres but it is likely that the great
majority are Central Gaulish. Four sherds had been
repaired. Form 18/31 platters were the most
common form, with one sherd from a form 18
platter, five from a form 79 platter, and two possibly
from a form 43 mortarium. Apart from form 18,
which can be dated to the mid-late Ist century AD,
all the forms are 2nd century.

Nearly equal amounts of Oxfordshire and New
Forest products were recovered (Table 1); however,
the New Forest material is represented mainly by
stoneware-type colour-coated wares and_ the
Oxfordshire vessels by red-slipped wares. Oxford
wares include oxidised colour-coated wares, white
ware and parchment ware and the forms include the
carinated bowl type C81, ‘dog bowl’ type C94, and
mortaria types C97 and C100 (Young 1977). One
sherd may be from a type C88 bowl.

Table 2. Romano-British Coarse and Fine Ware Rim
Forms excluding New Forest and Oxfordshire products

* = forms identified from amongst scanned material
+ = forms identified from amongst pottery recorded in detail

R100 Rim form undiagnostic
+*RI101 Everted rim jar (3rd—4th C)
+*R102 Everted rim jar (lst-2nd C)

ARAO3 Straight-sided bowl with grooved rim
+*R104 Flanged bowl (late 2nd—3rd C)
+*R105 Dropped flange bowl (3rd—4th C)
+*R106 Dog dish, shallow bowl (2nd—4th C)
+*R107 Storage jar
+*R108 Flagon
+*R109 Carinated bowl/dish (1st—-2nd C)
+*R110 Wide-mouthed jar (1st-2nd C)
+*RI11 Narrow-mouthed jar

ARID Butt beaker (1st-2nd C)

*R113 Shallow bowl with bead rim (2nd—4th C)
+*R114 Flat-rimmed bowl (2nd C)

*R115 Shallow dish/lid (1st—4th C)

*R116 Shouldered bowl (1st—2nd C)

ARITA7 Bead rim jar (1st C)

*R118 Reeded rim bowl (1st—2nd C)

R122 Everted, flattened rim jar with double

or single grooves on upper rim surface
(1st-2nd C)

+ R123 Wide-mouthed, necked jar/bowl] with
out-turned bead rim

+ R124 Shallow bow] with internal bevel

+ R125 Lid-seated rim vessel

+ R126 Deep bow] with clubbed rim

+ RI27 Fish dish; shallow oval dish

+ R128 Bow! with clubbed rim (2nd—4th C)

30 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Table 3. Romano-British fabric group totals from Trenches 23-4 by feature: detailed recording

Feature Context Grog BB1 Quartz Quartz Samian New Oxford- § Amphorae Rhenish
(all) (coarse) (fine) Forest shire
no wt no wt no we no wt no wt no wt no wt no we no wt
2306 2306 3 15 11 82 41 252 I a 12 3 10 5 45
2322. 2312 1 =. 122
2330 2329 3 39
2336 1 13
2338 6 37
2347 2 67 2 15
2348 1 6
2332 2326 2 4 2 13 1 1
2333 2327 12 186 Tet 32 273 1 11 2-29 38
2335 2334 1 6 119 70 1231 1 8 i= 1 8 t 3133
2337 2328 4} 71 #18 171 1 5
2346 2342 6 121 8 115 26 434 4 18 23 i829
2343 4 20 8 64 1 1
2801 2800 1 36 7 58 32 737 2 28 2 13
2806 2807 1 2
2809 2805 2 45
2809 2 20
2810 2803 2 120 4 27 13 89
2811 2811 4 136 2 30 2 56
2813 2812 5 120 11 77 76 1165 1 3 2 14 1 16 3 5
2814 8 539 132 55 1113 3 164 1 31 3... 31 1 228 2 996
2824 19 1397 27 634 199 4139 9 50 2 62 2 31 8 335 8 2304 1 3
2816 2331 17 249 44 514 146 2479 1 5 22 289 16 136
2341 2 26 25 753 73 1295 1 2 14 144 4 58
2821 2324 2 22 8 53 9 44 1 2 8
2349 3 13
2820 2 18 9 99 3 8
2822 2323 1 10 f) 10 +14 68 1 1
2823 2325 3 45 5 18 15 95
2829 2825 7 45
2841 2828 22 170 1523 1 4
2845 2835 1 2. 3 7 4 25
2847 2846 1 28 5 23 2 5
2938 2937 5 43
2943 2827 1 9
2948 2947 2 6 1 3
2955 2844 4 115 10 133 97 1366 1 2 2 18 7 46 2414067
2953 2 29
2954 3 33
2972 2960 1 50 4 29 14 4131
2973 2838 1 28 2 25 26 267 1 } 1 2
2974 15 129 1 2 1 7 1 2
2975 1 18 1 8
2983 2979 1 5
2987 2826 Z 20 19 114 1 3 1 9
2830 2 2 29 343 2 6

2989 2950 1 115
2962 1 2

3010 2971
3020 2300
2301
2305
2307
2309
2315
zal 2316
a2 2317
2319
2802
2843
2956
2976
2316
2815
3024
3025
3026
3027
3028

23:. 668° > 36
33 5
een 1
3
2
57 1101 54
ear oe PT
1-126. 12
103 5802s 11
HOMO7T4. I
20 1237
1.>. 26
1

411 255 3867

SNe}

70 111 1328
4 23 227
13 15 186
13 13 126
548 235 3476
8 98

22 VA 272
88 18 405
105 20 285
2 40

11 241

3) Or +282
59 2554

3 31 423
shel

3.7 S39

12 1 16

*1 Special Find 5031
*2 Special Find 5054

= Oe

OF BF 1 6

1 2 1 14
1 16
1" 33

5 43 4 138
1 66
I. A2

6 45

3 64

Figure 16. Butterfield Down: prehistoric and Romano-British pottery from Trench 23
1. Possible incense cup, Early Bronze Age, fabric G1, pit 3004 2. Everted rim jar, 1st-2nd century AD, fabric E101, rim
R102, ditch 2337 3. Everted rim jar/bowl, 3rd—4th century AD, fabric Q102, rim R122, pit 2816 4. Everted rim jar, 3rd
century AD, fabric Q100, rim R102, kiln/oven 2841 5. Dog dish, 2nd century AD onwards, fabric E101, rim R106, pit
2816 6. Flanged bowl, late 2nd-3rd century AD, fabric E101, rim R104, pit 2816 7. Drop-flanged bowl, 3rd-4th century
AD, fabric E101, rim R105, pit 2816 8. Hoard vessel, New Forest colour-coated globular beaker
9. Face pot, fabric Q110, pit 2813

32 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

‘The New Forest products include colour-coated,
parchment, red-slipped (Fulford 1975a, fabrics 1a,
lc, 2a), and grey wares. Forms include single
examples of jug types 18 and 22, two examples of the
indented beaker type 27, one each of beaker types 30
and 33, two type 63 bowls, an example of a type 106
mortarium and six unidentifiable rims (zb7d.). Four
fine sandy wares of unknown or uncertain origin
were defined (fabrics Q110, Q112-14; Table 4;
detailed descriptions in archive) but these fabrics are
represented by very few sherds (Table 1).

The rim of a narrow-mouthed, flagon-like vessel
with an applied and incised face mask (Figure 16, 9)
occurred in fabric Q110 (rim type R108). The origin
is unknown, but it may be a fairly local product as
undecorated sherds in the fabric were recorded
throughout the site. The decoration does not fall
into the face mask traditions identified by
Braithwaite (1984) and Butterfield Down lies
outside the previously known distribution area of
face mask pots, which was restricted to the north
and east of Britain. The form of the vessel is unusual
for this type of decoration, since it has previously
been recorded in only one instance, near Carlisle,
Cumbria (ibid.).

The single sherd of fabric Q112 derives from a
2nd-century poppyhead beaker with barbotine
decoration of a type produced at the Highgate Wood,
Kent, Upchurch and Oxfordshire centres amongst
others (Tyers 1978). Fabric Q113 may be a product
of the colour-coated ware industry of north Wiltshire
which was in operation c.AD 125-40 (Anderson
1979, 11). This fabric has also been recognised at an
enclosed settlement at Figheldean in the Avon Valley
(Mepham 1993, fabric type Q114).

Thirty-five sherds of Dr. 20 amphorae from
southern Spain, mostly from a single vessel, were
recovered. This is the most common type found on
Romano-British sites; the vessels would have con-
tained olive oil and were produced from the mid Ist to
the end of the 3rd century AD (Peacock and
Williams 1986, 136). The rim has been removed and
the handles had apparently been sawn off. However,
the base also seems to have been sawn in half,
suggesting that the neck and handles were not
removed to reuse the vessel.

Thirteen coarse ware fabrics were recognised;
nine are broadly defined and may include products
from more than one source. None of the fabrics
identified are restricted to early forms, with the
exception of Q103, which is represented by five
identifiable sherds: four everted rim jars of early
form (R102), and one wide-mouthed jar (R110).

Table 4. Romano-British coarse and fine ware fabrics,
excluding New Forest and Oxfordshire products

E101 Black Burnished Ware (BB1)

E102 Black Burnished Ware (BB1 variant)

G100 Wheelthrown, coarse grog-tempered fabric

G101 A finer version of G100

Q100 Wheelthrown, coarse grey wares without
glauconite

Q101 Wheelthrown, coarse grey wares with
probable glauconite

Q102 A finer version of Q100

Q103 A finer version of Q101

Q104 Wheelthrown, coarse oxidised wares without
possible glauconite

Q105 Wheelthrown, coarse oxidised wares with
probable glauconite

Q106 A finer version of Q104

Q107 Wheelthrown, sandy fabric with poorly
sorted quartz grains and displaying irregular
firing conditions

Q108 A finer version of Q105

Q110 Wheelthrown, white-slipped, well-sorted
oxidised sandy fabric with sparse iron oxides

Q112 Wheelthrown, hard, dense, fine unoxidised
fabric with rare quartz grains

Q113 Wheelthrown, oxidised, slightly sandy fabric
with sparse iron oxides, colour coated

Q114 Wheelthrown, hard, dense, fine, oxidised
fabric with rare quartz grains and iron oxides

This may be the result of insufficient data rather than
any genuine trend. Other early vessel forms in coarse
ware fabrics include an everted rim jar (R102) in
BB1, while from the scanned collection there are butt
beakers (R112), shouldered bowls (R116) and bead
rimmed jars (R117), which show Iron Age
influences. Savernake ware, an early Roman fabric,
was amongst the material scanned from Trench 4.
The most commonly represented fabric types
with regard to rim sherds (Q100, Q101 and Q102)
all include examples of vessels dating from the Ist to
the 4th centuries, such as everted rim jars (R101,
R102), flanged (R104) and dropped-flange (R105)
bowls and ‘dog dishes’ (R106). The two grog-
tempered fabrics were frequently found in thick-
walled storage jar forms, although only one datable
rim sherd was present, an early form of everted rim
jar (R102). Black Burnished ware (BB1) in its early
and late forms does show continuity of availability
from one production centre. It is found in everted
rim jars of early (Figure 16, 2) and late (Figure 16,
3) forms, flanged (Figure 16, 6) and dropped-flange

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 33

(Figure 16, 7) bowls, ‘dog dishes’, and one example
of a ‘fish dish’ (late 3rd—late 4th century AD)
(Gillam 1976; Seager Smith and Davies 1993, 252).

Estimated Vessel Equivalents (EVEs) (Orton
1980) were calculated for each form as a means of
assessing statistically their relative numbers in the
assemblage (Table 5). This showed that the two
most common forms were the early and late forms of
everted rim jars (R101, R102) and that there were
roughly equal numbers of early and late vessel forms
(nine and six forms, respectively; Table 2) for which
a definite date range could be given. There were,
however, twice as many later than earlier vessels,
suggesting that the main period of site activity was in
the 3rd to 4th centuries.

Table 5. Estimated Vessel Equivalents from Trenches 23-4,
detailed recording. (EVEs were not available for rim forms
124, 127 and 128 as the sherds were too small to allow
calculation) |

Rim form EVE Rim form EVE
R101 7.45 R111 0.16
R102 igo 2 R113 0.08
R104 1.02 R114 0.05
R105 2.48 R123 0.12
R106 3.40 R124 -

R107 1.16 R125 0.07
R108 0.40 R126 0.30
R109 0.13 R127 -

R110 0.15 R128 -

The only coarse ware fabric type which can be
attributed to a known source is Black Burnished
ware (BB1) from the Wareham/Poole Harbour area of
Dorset (e.g. Williams 1977). Sources for the other
fabrics are uncertain, since grey wares in particular
are difficult to characterise and the assemblage prob-
ably represents the products of several different prod-
uction centres. Anderson (1979) has defined one centre
of grey ware manufacture in north Wiltshire c.42km
to the north of Butterfield Down where kilns are
known at Purton, Whitehill Farm and Toothill Farm.

The presence of what is probably glauconite in
fabrics Q101, Q103, Q105, and Q108 might indicate
a source close to outcrops of Upper Greensand which
occur in north and west Wiltshire. A production
centre at Westbury is suggested by kiln furniture and
wasters found there (Rogers and Roddham 1991).
Other possible sources are the New Forest kilns,
which produced grey wares alongside the fine wares
in the late Roman period, but these wares have not
yet been sufficiently well characterised for identifi-
cation to be possible.

Distribution

Trenches 1-22 and Trench 23 surface collection: certain
fabrics were commonly found; grey wares, oxidised
wares, BB1, grog-tempered fabrics, Oxfordshire and
New Forest products were present in all but Trenches
5 and 9. Only three flint-gritted sherds were recovered,
all from the surface collection in Trench 23.

There were also isolated examples of other fine
wares which occurred in greater quantities in the
excavated features of Trenches 23 and 24. White-
slipped ware was recovered from Trenches 1, 3-4, and
7, samian from Trenches 1, 3-4, 7, 12, and 17, and
amphora sherds were found in Trench 3. All these
wares were also retrieved from the surface of Trench
23, together with one sherd of Rhenish ware and two
of probable north Wiltshire colour-coated ware. Four
sherds of Savernake ware were found in Trench 4 —
the only occurrence of this type on the site.

No distinct early features were identified in the

analysis and although a number of early rim forms
were identified they occurred with later material.
Early everted rim jar forms were recovered from 11
trenches and the surface of Trench 23; examples of a
carinated bowl/dish rim, a wide-mouthed jar, a
shouldered bowl of Ist/2nd-century date, and a Ist-
century butt beaker were recovered from Trench 4.
The shouldered bowl form was also recovered from
Trench 22. A bead-rimmed jar form was identified
in Trench 15 and a Ist/2nd-century reeded rim bowl
form in the assemblage from the surface collection in
Trench 23. With the exception of the concentration
in Trench 4, the early forms were distributed
randomly across the site.
Trenches 23 and 24. Roman pottery was recovered
from most of the excavated features and no
significant clustering was observed. Feature 3020, a
corn drier, showed a concentration of fabric types
Q100, Q101 and Q104, but it is likely that the
pottery was deposited over a period of time after the
structure went out of use.

The variety of early and late Roman types
recovered, together with certain fabric types which
have a limited date range, show that the site was
occupied throughout the Roman period. None of the
locally-made wares seems to have been restricted to
any chronological period, and non-local wares were
present throughout the occupation of the site. The
relatively small amount of early Roman wares, such
as Savernake ware, or easily definable early forms of
BB1, indicates activity at Butterfield Down was at its
peak in the late Roman period when the large
proportion of New Forest and Oxfordshire wares
may reflect the status of the site.

34 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Occupation at Butterfield Down seems to begin
at the time that occupation at nearby Boscombe
Down West declined (Richardson 1951). The latest
forms at Boscombe Down West are very similar to
the early rim forms at Butterfield Down and Trench
4 in particular has a concentration of these,
including shouldered and carinated bowls. The
range of fabrics and forms in the late Roman
assemblage is paralleled nearby at the substantial
settlement west of Durrington Walls where little
early Roman material was found (Swan 1971).

Although there is a greater emphasis on the
late Roman period at Butterfield Down, the range
of fabrics and forms is similar to that found at
Figheldean (Mepham 1993). If the difference is not
due to the different dates of the sites, the proportion
of fine wares at Butterfield Down, 6.7% by weight,
which is comparable to the 7% at the site west of
Durrington Walls (Swan 1971) may suggest that these
sites were of higher status than that at Figheldean.

Ceramic building material

by J. MILLARD

Some 256 pieces of ceramic building material
(14,340g) were recovered. Of these, 120 pieces
(11,934g) are Roman, 134 pieces (2,395g) medieval
or post-medieval and two pieces (11g) are undated.
The post-Roman material consists mainly of small
fragments of brick or roof-tile and is almost entirely
from unstratified contexts. ‘The even distribution of
this material is likely to be the result of agricultural
practices, probably manuring.

The Roman material includes 86 fragments of
brick, six of tile of uncertain form, seven of tegulae,
one imbrex and two of comb-patterned flue tile.
Eighteen other fragments may be dated to this
period on the basis of fabric and form but are of
uncertain type. The material was evenly distributed
across the excavation, the only concentration being a
group of 64 fragments from a single plain /ydion
brick measuring 430mm x 290mm x 35mm which
had been used as the base of a hearth or oven in
Trench 23 (Figure 6, 2809). There was no obvious
use or reuse of ceramic building material in the
mortared structures such as the corn driers.

The fired clay
by ANDREW CROCKETT
A total of 107 pieces (2,155g) of both single-faced

and double-faced cob walling was found. This
material is oxidised throughout and some pieces

have suffered from burning after use. Most of the
cob walling formed part of a dump in pit 2813
(Figure 6), while a further large group was found in
the fill of the excavated corn drier 3020.

Approximately half of a spindle whorl (fabric G101,
35mm diam., <12mm thick, c.25g, central perforation
10mm diam.) was found in pit 404, Trench 4. Roman
spindle whorls are usually made from worked stone,
shale, or broken pottery (Leach 1982, 217), and it is
rare to find examples made from clay.

The building stone
by J.I. MILLARD

Some 137 flat stone slabs, all probably tiles, were
found. Most are of limestone, although there are
fourteen of sandstone. The tiles can be divided into
two broad groups: thin slabs likely to be roofing
material, and slightly thicker pieces more suitable for
flooring. Two of the roof tile fragments have
surviving nail holes and on one of these ferrous
corrosion products are visible adjacent to the hole.
Nine of the fragments are much thicker than the
floor tiles, and their function is unknown. Other
building material consists of two dressed blocks, one
of limestone and the other of Greensand, both
probably architectural fragments. Unworked pieces
of Oolitic Limestone and sandstone may also have
been building materials.

Worked bone and ivory

by J.. MILLARD

One ivory and one bone pin were found in the fills of
corn drier 3020; a second bone pin and a carved
sheep metatarsal were recovered from the fills of pit
404 in Trench 4. The ivory pin is 78mm long, has a
spherical head and the shaft tapers at both ends, a
form paralleled in bone pins thought to date from after
AD 200 (MacGregor 1985, figure 64, nos. 8-10).
Unless it is actually highly polished bone, the pin was
probably imported. The importation of elephant and
fossil (mammoth) ivory to the Mediterranean world
is known in the 4th century (Krzyszkowska 1990)
but it is unknown whether raw materials or only
finished objects reached mainland Britain.

The bone pin from the corn drier is similar but is
only 26mm long. Only a 56mm portion of the shaft
of the third bone pin survives and this tapers at both
ends. The natural grooves on both the anterior and
posterior surfaces of the sheep metatarsal have been
enlarged using a knife.

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 35

Paleo-environmental material
The carbonised plant remains

by MICHAEL J. ALLEN

Six samples of ash were recovered from regular
intervals along the flue of corn drier 3020. They
were processed using standard flotation methods,
and the 500um flots were assessed. All samples
contained carbonised grain (barley, Hordeum sp.,
and wheat, Triticum sp.), none of which showed signs
of germination. Two of the samples contained chaff
elements (e.g. rachis fragments, etc.), and weed
seeds (Polygonum sp. and cf. Bilderdykia convolvulus)
were common in two samples and present in two
others.

These finds suggest the burning of crop waste as
well as cleaned grain, and could indicate disposal
and burning of farm waste in the structure after it
had gone out of use; it is more likely, however, to
reflect the probability that ‘corn driers’ actually had
a variety of uses (van der Veen 1989).

Animal bones
by J. EGERTON

Nearly 4,000 well preserved but severely fragmented
animal bones were recorded from Trench 23,
principally from late Roman contexts. Only securely
dated finds are considered here.

The early Roman assemblage was dominated by
sheep (over 50%) but cattle were also found in
significant numbers with other: domestic animals
present (Table 6). Although the majority of bone was
recovered from pits and ditches, the relatively high
proportion of teeth indicates poor preservation overall
of the assemblage. Over 15% of the assemblage was
weathered but only three gnawed and two butchered
bones were recorded.

Table 6. Animal bones from early Roman contexts
in Trench 23

Teeth Other elements Total
Cattle 2 9 11
Sheep if 12 19
Horse - 3 3
Dog - 1 1
Chicken - 1 1

Total fragments 78

Total identified 35

% identified

Late Roman contexts contained 3771 fragments of
which 1746 (47.7%) of the bone from features was
identifiable (Table 7). The assemblage was heavily
fragmented due to physical breakage and butchery
prior to deposition and also in part because of post-
depositional breakage. The sample produced only
17 complete and mature long bones (sheep 8, horse
5 and cattle 4). The percentage of identified bone
demonstrates a common variation between feature
types and identifiable fragments, with pits offering
better preservation, and proportions of species were
constant across the site.

Table 7. Percentages of species from late Roman features in
Trench 23 excluding contexts with special deposits/large
quantities of bones, e.g. pit 404

Cow 35.9 Chicken 0.6

Sheep 54.8 Red Deer 0.13
Pig 4.1 Hare 0.06
Horse 3] Bird 0.06
Dog tt Amphibian 0.13

The relative proportions of species is unsurprising,
with the proportion of sheep (55%) slightly larger
than cattle (36%) on this chalkland site. Apart from
two special contexts, most of this site reflects small-
scale primary and secondary butchery waste.

Some cattle on the site were used as draught
animals (see below), but the limited ageing data also
suggests some were killed on maturity for meat and
57% of the cattle fragments were from high meat-
bearing bones. Generally cattle produce thirteen
times more meat per animal than sheep, so setting
aside secondary products, they were the most
important animal in the food economy (Grant 1984;
Done 1986; Millett 1990).

Most of the butchery marks (cuts and chops) were
associated with primary butchery, but some (28%)
were associated with disarticulation of the foot ele-
ments. One animal had spavin which is associated
particularly with draught animals resulting in their only
being able to manage light loads; two bones displayed
septic arthritis and one had an exotosis on the foot.

The remains of at least 46 sheep, comprising 55%
of the assemblage, were found. Most bones were
fragmented but only 22 butchery marks were noted
and 40 bones had been gnawed by canines.

A few neonatal deaths probably associated with
lambing were recorded. The majority of bones,
however, indicate a mature age at death (i.e. three
yearst+) and this is supported by the dental data. As
three good fleeces can be obtained by the time an
animal is three-and-a-half years old, this suggests

36 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

‘Table 8. Animal bones from pit 404

Context Phalange 1 Phalange 2
44] 50 62
445 5 41

sheep rearing was primarily for wool and other
products rather than meat.

Pit 404 in Trench 4 contained a large collection
of foot and tooth elements (Table 8). These elements
are very well preserved, mostly from mature animals,
and lack cut marks so it is likely that they are either
waste products of industrial processing, such as
skinning or tanning, or comprise a special animal
deposit. The group is paralleled, for example, by a
group from a Roman well at Oakridge II,
Basingstoke, Hampshire (Maltby 1993) where a
collection of head and foot elements was in an
excellent state. However, at Butterfield Down the
metapodial bones were not found with the tooth
elements so it seems likely that metapodia were used
for bone tools. It may not be a coincidence that a
sheep metatarsal had been worked into a tool.

Only four horses were identified but a large
proportion of the fragments were teeth (33%); this,
together with the rarity of butchery (just two cuts
associated with disarticulation), suggests that horses
were seldom, if ever, eaten at this time. A complete
skull of a 19-year-old horse was found in pit 2813 in
Trench 23. It is not clear if this was a special deposit
for as well as containing seven complete and
associated sheep long bones, it was accompanied by
a seemingly ordinary mix of waste bones.

Only four pigs were recorded, suggesting that
they were not an important part of the food economy
though the limited ageing data does suggest that they
were bred on the site and were killed whilst young.

There are, perhaps, rather few dog bones consid-
ering the canid gnawing of seventeen elements (1.1%
of the assemblage). One very young puppy was found
in a ditch in Trench 3. Two bones of red deer were
found, one an antler from a mature animal which had
been sawn off below the burr. Other animals
represented included hare, chicken and amphibian.

Of special interest is the burial of a crow, whether
carrion or hooded is uncertain, in a pit or ditch
(310) in Trench 3. The preservation of the bones,
which was very good, endorses the suggestion that it
was a special deposit.

Discussion

The animal bones give keen indicators of the
activities on this late Roman settlement in an area
which lacks well-investigated sites. Animal hus-

Phalange 3
44 11 11
4 - - 45

Astragalus Calcaneus Teeth

bandry was clearly of a similar mode to that of other
small rural settlhements of the period. The cattle
bones are intensely butchered and there is evidence
for the use of all parts of both sheep and cattle.

Assessment of the land Mollusca

by SARAH F. WYLES and MICHAEL J. ALLEN

Samples were taken from a subsoil hollow in the centre
of the ground enclosed by the ring ditch and from the
ring ditch itself (2500). The subsoil hollow is undated
but if, as seems likely, it is earlier than the ring ditch,
the presence of species such as Vitrea contracta and
Carychium tridentatum which are found in leaf litter
and tall grass would suggest tall grassland. The
presence of open country species (Vallonia spp. and
Pupilla muscorum) indicates long, ungrazed grassland,
possibly with some localised scrub habitats in the area.

The primary fills of the ring ditch were barren
and there were few finds from the secondary fills.
However, the presence of Vallonia excentrica in the
lower secondary fill and Helicella itala, Pupilla
muscorum and Vallonia excentrica in the upper
secondary fill suggests that after the monument was
built it was surrounded by well-established, short-
turved grazed grassland, which fits well with what is
known of the contemporary landscape in the area.

Columns of contiguous samples were taken from
ditches 15, 18, 21 and 26 in the east of the site and
were dominated by open country species. In the case
of Ditch 21, the presumptively later Bronze Age
Earl’s Farm Down ‘Linear’, the sequence showed no
significant change in the major species composition
with the exception of Vertigo in the upper fills. Some
variation within the Vallonia species was noted,
however, indicating the potential for discerning
changes in the open country environment.

Marine Mollusca
by SARAH F. WYLES

Fragments from at least 158 oysters, which may have
been dredged from natural beds rather than farmed,
were recovered from Roman contexts all across the
site. The numbers are too small for oysters to have
been other than an occasional supplement to the diet
of the inhabitants.

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 37

Discussion

by MICK RAWLINGS, A.P. FITZPATRICK and
ROSAMUND M.J. CLEAL

Prehistoric

Late Neolithic pit ring

One certain and one possible prehistoric monument
were found within the area which was later the site of
a Romano-British settlement. Although only one pit
could be partially excavated (2998), it is probable
that the four subrectangular pits in the south-west of
Trench 23 were related, forming a small ‘pit ring
henge’-type structure c.10m in diameter. Similar
monuments are increasingly well known in southern
England.

At Conygar Hill on the Dorchester, Dorset,
bypass, two similar structures were found, one of
which contained Late Neolithic Grooved ware (Smith
forthcoming) and there are two further examples
from Dorchester-on-Thames, Oxfordshire, although
the ditch segments there were up to 2m deep
(Atkinson et al. 1951, Site II, phase I and Site IV).
Late Neolithic Peterborough ware was found in Site
II at Dorchester-on-Thames (where Site IV is likely
to be contemporary) and was used in funerary
practices. A comparable monument at Barrows Hill,
Radley, Oxfordshire is probably also of Late
Neolithic date and may have been funerary
(Chambers and Halpin 1984, 6-7).

Other small, circular Late Neolithic monuments
are known in the immediate vicinity of Butterfield
Down. The first phase of barrow Amesbury G71 on
Earl’s Farm Down, c.1.3km away, was a small ring
ditch of probable Late Neolithic date (Christie
1967). At Butterfield Down, Peterborough Ware was
found in pit 2943 c.20m to the north-west of the pit
ring henge. The unfinished stone axe provides
further evidence for Neolithic activity on the site, but
perhaps at an earlier date in the period.

Pit 2

Beakers similar to those recovered from pit 2 have
been found on domestic sites elsewhere, but the
context and typology of the Butterfield Down
material may indicate a non-domestic mode of
deposition. Vessel 1 is an extremely large Beaker,
apparently larger (an terms of approximate height
and diameter) than all but one of the vessels
illustrated in Clarke 1970 (the exception being a
rusticated vessel from Great Barton, Suffolk: Clarke
1970, fig. 916).

The volume of Vessel 1 is approximately eight
litres (calculated by division of the internal profile
into conic frustra), which falls within the sort of
volume typical of storage, or possibly food
preparation, rather than for individual eating and
drinking. The volume places it towards the upper
limit of the range for Beakers, which appears to lie
mainly between one and five litres, averaging at three
litres (Thomas 1991, fig. 5.8).

The fact that there is some carbonised residue
adhering to the interior suggests that it had held
organic contents at or before deposition. The
possibility that its use and deposition were not
domestic is apparent, but this hypothesis cannot
easily be tested. If analysis of the organic residues
were to be undertaken, it might shed some light on
the nature of the material held within the Beaker.

Although Butterfield Down Vessel 1 is excep-
tional in terms of size, it is not unique in terms of its
deposition within a small, apparently isolated site.
Although non-funerary Beaker sites are not as
common in Wessex as they are in some other regions
(e.g. eastern England), other vessels have been
found in similar circumstances in the area. At
Barrow Pleck, Rushmore (Cranborne Chase),
sherds representing slightly less than half
an incised Beaker not dissimilar to Vessel 1 were
found in the top of a periglacial feature, where
they had probably been placed in a slight hollow
formed by the slumping of the fill (Cleal 1991,
148, fig. 7.3, Po). A less ambiguously domestic
site with at least one similar Beaker is close to
Badbury Rings, Dorset, where two pits and a
posthole were filled with sherds of probably more
than fifteen vessels (Gingell with Dawson 1987),
one of which was represented by approximately
one quarter to one third part of its total and which
could be classed as a Clarke’s Final Southern Beaker
(S4).

The occurrence of small numbers of Beaker
sherds as scatters or in small isolated features cannot
be considered unusual, either locally or within the
region (e.g. the widespread occurrence of small-
scale Beaker scatters in the Stonehenge area, Cleal
1990, fig. 154). As such, the sherds of Vessels 2 and
3 would not occasion particular comment. The
presence of the large and virtually whole Vessel 1,
however, indicates that this deposit cannot be
regarded in quite the same light. Given this, it is
tempting to speculate that the extremely unusual
nature of the fabrics of Vessels 2 and 3, with their
bone temper, may also reflect a non-domestic or not
wholly domestic function.

38 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Early Bronze Age chalk plaque and ring ditch

Other evidence of later Neolithic or Early Bronze
Age date is provided by the chalk plaque (Figure 14).
Although it was found in a late Roman pit, the object
finds its best parallels with the Late Neolithic chalk
plaques from the nearby Chalk Plaque Pit c.200m to
the east, which were associated with Grooved Ware
(Harding 1988; Lawson 1993). The decoration on
the Butterfield Down piece is closer to that on
Beaker pottery, which may suggest that it is slightly
later in date.

A further prehistoric monument, the ring ditch,
does not appear to have encircled a central grave cut
into the natural chalk. Although graves may have
been dug in the southern part of the ring, it is quite
possible that the ring ditch did not contain any
burials, or that they may have been made in the now
destroyed mound. The lithic assemblage from the
ditch suggests a Late Neolithic-Early Bronze Age
date. Although there are local parallels, such as the
penannular ditch of Winterbourne Stoke Barrow 44
(of a similar size) which are thought to be Late
Neolithic (Green and Rollo-Smith 1984), an Early
Bronze Age date for the Butterfield Down example is
considered more likely.

The crouched inhumation burial 3004 is
probably broadly contemporary with the ring ditch
and may have been a satellite burial. In view of the
predominantly funerary contexts of incense cups it is
likely that the small fragment — possibly of one of
these vessels — found in the grave is a formal grave
good rather an accidental introduction during the
digging of the grave.

The snails from a hollow in the centre of the ring
ditch monument are likely to pre-date it and they
suggest an open, long-grassed environment which
may have contained some scrub. However, by the
time that the fill of the ditches began to stabilise, the
monument lay in well-established, short-turved,
grazed grassland, an environment which is well
documented in other analyses in the area. It is also
noteworthy that the fills of features of prehistoric
date have a different colour from those of Roman
ones which is probably due to an increase in the
quantities of chalk resulting from the reduction of
soil depth by ploughing, probably during the
Middle—later Bronze Age, a trend which is again well
documented in the later prehistory of the chalklands.

The Earl’s Farm Down linear ditch

Sample excavation of the large linear ditch at the east
side of Butterfield Down failed to provide any clear
evidence for the date of this feature. It forms part of a

major component of a network of such ditches known
as the Earl’s Farm Down complex which, along with
other similar networks in the Salisbury Plain area, has
been tentatively assigned an original construction
date within the later Bronze Age (Bradley, Entwistle
and Raymond 1994, 122). They are seen as territorial
divisions which represent the formal organisation or
re-ordering of the landscape. The summary evidence
from the mollusca suggests that the ditch lay in an
open country environment although detailed analysis
might indicate changes within it.

Although restricted by the limited amount of
excavation possible, the evidence from Butterfield
Down is a useful addition to our knowledge of the
later Neolithic and Bronze Age periods in the
Stonehenge area and the variety of later Neolithic/
Early Bronze Age special deposits and funerary and
ritual monuments in Wessex (Barrett, Bradley and
Green 1991, 58-139).

Roman

Glass and pottery from the Ist and 2nd centuries
AD were recovered from a variety of contexts, but
structural evidence for early Roman activity on the
site is restricted to a few features, in particular a
group of right-angled ditches in the north-east of
Trench 23 (Figure 6). It is possible that early Roman
activity at Butterfield Down was mostly outside the
areas so far examined since the brooches discovered
as surface finds would seem to suggest occupation at
this time. The lack of evidence for Iron Age activity
offers a contrast to the continuity through the later
prehistoric and Romano-British periods seen at local
sites such as Chisenbury Warren (Bowen and Fowler
1966, 50-2) and Figheldean (Graham and Newman
1993).

The late Roman settlement at Butterfield Down
covered at least six hectares and appears to have
been unenclosed. The bulk of the pottery is later
Roman and there was a dramatic increase in coin
loss in the settlement at this time.

Our understanding of the layout of the settlement
is limited since excavation was restricted to the sites
of modern houses and roads (Trenches 1-21).
However, the evidence is consistent with the clearer
picture given by the larger area of Trench 23. Here a
single, shallow, right-angled wall-footing and a ring
gully were the only certain traces of the foundations
of buildings; no clearly defined buildings or
residential compounds were identified. Although
there are hints that there may have been a temple on
the site (see below), the character and quantity of

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 39

finds strongly suggest a settlement in which buildings
left little, if any, archaeologically obvious traces.

Timber-framed buildings may have rested on sill
beams as at, for example, the early phase of Skeleton
Green, Hertfordshire (Partridge 1981) or on stone
joist supports, as at the late Roman settlement at
Wanborough, Wiltshire (Anderson and Wacher
1980). The cob walling from Butterfield Down could
have come from such timber-framed buildings. One
sunken-floored feature was excavated in Trench 4
while several similar examples were observed during
the watching brief (Trench 22). These may be the
remnants of sunken-floored structures or cellars
representing the only surviving evidence of
buildings, a situation known for example at the
extra-mural settlement at King Harry Lane, St
Albans, Hertfordshire where large, late Roman
‘cellars’ dated to the 3rd century are the only
structural evidence for buildings (Stead and Rigby
1989, 7-11, fig. 4, 7-8).

Some of the buildings at Butterfield Down may
have been roofed with clay and stone tiles instead of
thatch, and some may have had stone tiled floors.
The two possibly architectural fragments suggest
that more imposing buildings may have stood in the
settlement, perhaps in the area of the clay tile scatter
located during fieldwalking (Figure 2, H, B). The
posthole alignments show that the settlhement was
divided by fences, some of which presumably
enclosed buildings. The hollow ways in Trenches
12-13, 23 and possibly 17, represent substantial
tracks or roads, confirming indications in aerial
photographs of a series of trackways passing through
the settlement.

A range of evidence informs us of the activities of
the inhabitants of Butterfield Down. Plant macrofossils
from corn drier 3020 indicate that barley and wheat
were ‘dried’ while chaff from the same samples
suggests that the crops were also being winnowed
and threshed on site. The number and variety of the
corn driers or kilns identified might suggest either
that cereal processing was undertaken on a small,
perhaps household, basis or that it was a more
important activity. The presence of a millstone,
probably from a mill driven by animals, as well as
numerous querns, points to the latter possibility.

Cattle and sheep appear to have been the most
common farm animals. The combined evidence of
age and butchery marks points to the killing of cattle
on maturity and the primary butchering of high
meat-bearing joints indicates that these could have
provided a principal source of meat. The evidence of
pathology also suggests that some cattle were used as

draught animals and it may have been animals such
as these on which the iron shackle was used.

In contrast, the sheep identified appear to have
been kept to maturity, presumably for their fleeces
and other products. The discovery of a large
assemblage of feet and teeth suggests that some
animals at least were butchered in such a way as to
allow their hides to be kept. Some of their bones
were worked into tools. Pigs were also eaten, as were
hares and chickens. Horses seem to have been
butchered rarely and they are likely to have been kept
primarily for riding, as beasts of burden, and for
traction; the horseshoe, and perhaps the hipposandal
also, would have been worn by these animals.

Very few tools which might be indicative of other
tasks undertaken by the inhabitants were found. The
single spindle whorl testifies to the spinning of wool
and the slag shows that some smithing was
undertaken; the knives could have served a variety of
uses. The discovery of a stylus indicates conditional
literacy, at least, an ability which is likely to have
been quite rare in settlements of this sort (Evans
1987). The glass and pottery demonstrate something
of the range of storage and table vessels used and, as
would be expected on a predominantly later Roman
site, foodstuffs imported in amphorae are rare.

Evidence of the religious beliefs of the inhabitants
is provided by the burial of infants within the
settlement (2845 and 2952, Trench 23), a practice
which is particularly common in late Roman rural
settlements (Struck 1993). The sceptre-head is a
notable discovery and it may not be accidental that it
was discovered on the surface of the shallow ring
gully 2847. Such dating evidence as there is from the
gully suggests a late Roman date, a period in which
domestic buildings were usually rectangular, which
raises the possibility that it may have been a temple.
However, as the infant burial found in one of the
gully terminals would be appropriate to a domestic
setting the question must remain open. In any case,
the distinction between sacred and profane should
not be drawn too rigidly and certain deposits may or
may not derive from religious acts.

The crow, whose skeleton was found in the base
of feature 310 (Trench 3), had clearly been placed
there and covered over deliberately. This may be
paralleled in a 4th-century deposit at Foxholes
Farm, Hertfordshire where a cockerel was placed at
the bottom of a pit and flanked by two coins, the
upper part of the pit being packed with flints
(Patridge 1989, 49, 208-9). In pit/shaft 2813 at
Butterfield Down (Trench 23) a horse’s head had
been placed on top of part of a sheep, and a sherd

40 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

from a face mask pot was found within the overlying
fill. These pots are often associated with religious
activities on settlements (Braithwaite 1984, 124).
The deposit of sheep heads and feet in pit 404 may
represent waste from the processing of hides, but it
could also be a votive deposit (Scott 1991, 117).

It is possible that specialised religious buildings
stood on Butterfield Down. The large number of late
Roman coins and other late Roman metalwork such
as the figurines, spoons and bracelets, could derive
from a temple. The evidence for one or more
buildings with tiled roofs crowning the summit of the
Down, and the fragments of architectural masonry
found in the excavations might also support this
suggestion.

The setting of the site

Excavations at nearby sites allow Butterfield Down
to be placed in its local context. The later prehistoric
and early Roman site of Boscombe Down West
(Richardson 1951) is located c.3km to the south-east
and its occupation appears to have declined as
Butterfield Down developed, the latest pottery forms
at Boscombe Down being very similar to the earliest
here.

Butterfield Down shares many similarities with
the site at Durrington Walls (Wainwright 1971) 3km
to the north-west. This is an extensive, unenclosed
late Roman settlement with a ceramic assemblage
which also indicates some early Roman activity.
There is a lack of clearly identifiable buildings,
though there are a number of small ovens or kilns,
together with a well-constructed corn drier which is
almost identical to the one excavated at Butterfield
Down (Figure 11, Plate 3). The ceramic assemblages
are analogous, with similar ratios of fine to
coarsewares.

‘Turning to the broader range of settlement types
within the region, discussion of Romano-British
rural settlement has historically been linked to early
observations concerning the lack of villas in the
region of Salisbury Plain. This absence, along with
other factors, led Collingwood and Myers (1937,
224) to suggest that the Plain formed part of an
imperial estate, an idea which has enjoyed enduring
popularity, though more recently it has been
suggested that poor soil conditions were responsible
(Esmonde Cleary 1989, 106).

However, whilst some villas are known, the
number of nucleated settlements (Graham and
Newman 1993, 51), together with the evidence of
recent air photographs and surveys of extensive
Romano-British field systems and sites such as

Church Pits, Knook Down East and Knook Down
West (Britannia 23, 1992, 297-9, fig. 20-2) suggest
that soil conditions on the Plain were not a constraint.

Instead, social factors may be one reason for the
paucity of villas. While accepting that in the early
Roman period Salisbury Plain (amongst other
regions) might have been part of an imperial estate,
Hingley has suggested that in the late Roman period
the estate might have been partitioned and sold to
private landowners: thus villas ought not to be too
readily expected, and the wealth expended elsewhere
in building villas might here have been used in
different ways (Hingley 1989, 156-61), for example
in material goods. Hingley’s suggestion, however,
that wealth was invested in goods rather than in
buildings (o0p.cit.) is unconvincing as such objects are
also found at villas. Nonetheless, his distinction
between individual and community is valuable, and
the size of the settlement at Butterfield Down is large
enough to represent a ‘village’. The apparent absence,
so far, of lavish dwellings at Butterfield Down may
indicate the collective ownership of wealth.
However, some settlements were occupied from the
Iron Age and throughout the Roman _ period
(Graham and Newman 1993, 52) and the absence of
villas from the downlands of the Plain could reflect
its distance, both physical and social, from major
towns and the ideas of Romanitas which they
embodied (Scott 1991, 116). Clearly, further and
more detailed work on Romano-British sites within
Salisbury Plain, and their integration within wider
landscape analyses, is necessary.

The Archive

The archive is deposited in Salisbury and South
Wilts Museum, 65 The Close, Salisbury, Wiltshire
SP1 2EN.

Acknowledgements. The Butterfield Down project was commissioned
by Gleeson Homes and the help and support of the company’s
employees, in particular Nigel Hogg, Mark Davies and Clive
Wilding are gratefully acknowledged. The majority of the fieldwork
and much of the post-excavation work was paid for by The Gleeson
Group plc, who also financed the publication of this paper. The
work in 1993 was supported by Wiltshire County Council.
Additional financial assistance towards the post-excavation work
was provided by The Guinness Trust and Lord Moyne.

Much help was given at all stages by Mark Corney and other
officers of the RCHM(E) and Helena Cave-Penney of
Wiltshire County Council. Clare Conybeare, then of Salisbury
and South Wiltshire Museum, helped in liaising with local
metal-detector users, while Nick Griffiths identified some
items of metalwork. Assistance and advice during the post-
excavation period were given by Andrew Burnett, Frances Healy,
Lorraine Mepham, Elaine Morris, and Jessie Williams.

The project has been managed successively by Julian C. Richards,
Peter J. Woodward, Richard Newman and A.P. Fitzpatrick. The

———EE—————

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 4]

fieldwalking was carried out under the control of Martin Trott. The
excavations were directed by Mick Rawlings with additional
supervision by Hugh Beamish, Neil J. Adam, Lawrence Pontin,
Andrew B. Powell, Rachael Seager Smith, and P.A. Harding.
Trench 24 was excavated by Vince Jenkins. Aerial photographs were
organised by Graham Brown, who also assisted with the
excavations. The photographs are by Elaine Wakefield and Mick
Rawlings, and the drawings by Linda Coleman, Julian Cross and
S.E. James.

BIBLIOGRAPHY

ALLEN, D.A., 1988 “Roman Glass from Corbridge’ in M.C.
Bishop and J.M. Dore, Corbridge: Excavations at the
Roman Fort and Town 1947-80, English Heritage
Archaeology Report 8, London, 287—93

ANDERSON, A.S., 1979 The Roman Pottery Industry in North
Wiltshire, Swindon Archaeology Society Report 2,
Swindon

ANDERSON, A.S. and WACHER, J.S., 1980 ‘Excavations at
Wanborough, Wiltshire: an Interim Report’, Britannia
11, 115-26

ANNABLE, F.K. and SIMPSON, M.A., 1964 Guide Catalogue of
the Neolithic and Bronze Age Collections in Devizes
Museum, Devizes

ARTHUR, P., and MARSH, G., (eds.), 1978 Early Fine Wares
im Roman Britain, British Archaeological Report British
Series 57, Oxford

ATKINSON, R.J.C., PIGGOTT, C.M., and SANDARS, N.K.,
1951 Excavations at Dorchester, Oxon, Oxford

BARRETT, J.C., BRADLEY, R.J., and GREEN, M.J., 1991
Landscapes, Monuments and Society: The Prehistory of
Cranborne Chase, Cambridge

BOWEN, H.C. and FOWLER, P.J., 1966 ‘Romano-British
Rural Settlements in Dorset and Wiltshire’, in C.
Thomas (ed.), Rural Settlement in Roman Britain, CBA
Research Report 7, 43-67

BRAITHWAITE, G., 1984 ‘Romano-British Face Pots and
Head Pots’, Britannia 15, 99-133

BRADLEY, R., ENTWISTLE R., and RAYMOND, F., 1994
Prehistoric Land Divisions on Salisbury Plain, English
Heritage Archaeological Report 2, London

BUCKLEY, D.G., and MAJOR, H., 1983 ‘Quernstones’, in
Crummy (1983), 73-6

BURNETT, A., 1992 ‘Boscombe Down, Wiltshire’, in R.
Bland (ed.), The Chalfont Hoard and Other Roman Coin
Hoards, Coin Hoards from Roman Britain 9, London,
359-60

CHAMBERS, R.A., and HALPIN, C.E., 1984 Barrow Hills,
Radley, 1983-84 Excavations: an Interim Report, Oxford

CHRISTIE, P.M., 1967 ‘A Barrow Cemetery of the Second
Millennium BC in Wiltshire, England’, Proceedings of
the Prehistoric Society 33, 336-66

CLARKE, D.L., 1970 Beaker Pottery of Great Britain and
Ireland, Cambridge

CLEAL, R.M.J., 1990 ‘The Ceramic Sequence: Summary
and Discussion’, in Richards 1990, 242-46

CLEAL, R.M.J., 1991 ‘Cranborne Chase — the Earlier
Prehistoric Pottery’, in J. Barrett, R. Bradley, and M.
Hall (eds.), Papers on the Prehistoric Archaeology of
Cranborne Chase, Oxbow Monograph 11, Oxford,
134-200

CLEAL, R.M.J., with RAYMOND, F., 1990 “The Prehistoric
Pottery’, in Richards 1990, 233-46

CLEAL, R.M.J., ALLEN, M.J., and NEWMAN, C., in
preparation ‘An archeological and environmental study
of the Neolithic and later prehistoric landscape of the
Avon Valley between Durrington Walls and Earl’s Farm
Down’

COHEN, H., 1982 Description historique des monnaies frappés
sous Pempire romain VIII, Paris

COLLINGWOOD, R.G., and MYERS, J.N.L., 1937 Roman
Britain and the English Settlements, Oxford

COLLINGWOOD, R.G., and RICHMOND, I.R., 1969 The
Archaeology of Roman Britain, London

CORNEY, M.C., in preparation, “The late Roman settlement
pattern of the Wiltshire chalk’

CRUMMY, N., 1983 The Roman Small Finds from Excavations
im Colchester, 1971-79, Colchester Archaeology Report
2, Colchester

CUNLIFFE, B., 1971 Excavations at Fishbourne (Sussex)
1961-69, Vol II: The Finds, Society of Antiquaries of
London Research Report 26, Leeds

DONE, G., 1986 ‘The Animal Bones from Areas A and B’,
in M. Millett and D. Graham, Excavation on the
Romano-British Small Town at Neatham, Hampshire,
1969-79, Hampshire Field Club and Archaeology
Society Monograph 3, Winchester, 141—47

ESMONDE CLEARY, A.S., 1989 The Ending of Roman Britain,
London

EVANS, J.,; 1987 ‘Graffiti and the Evidence of Literacy and
Pottery Use in Roman Britain’, Archaeological Fournal
144, 191-204

FULFORD, M.G., 1975a New Forest Pottery,
Archaeology Report British Series 17, Oxford

FULFORD, M.G., 1975b “The Pottery’ in B.W. Cunliffe,
Excavations at Porchester Castle, Vol I, Roman, Society of
Antiquaries of London Research Report 32, London,
270-367

FULFORD, M.G., 1979 ‘The Pottery’ in G. Clarke, Pre-
Roman and Roman Winchester, Part II: The Roman
Cemetery at Lankhills, Winchester Studies 3, Oxford,
221-37

GAIMSTER, D.R.M., 1988 “Two Post-Medieval Sword-belt
Fittings from Pyecombe, West Sussex’, Sussex
Archaeological Collections 126, 245-47

GILBERT, H.M., 1978 “The Felmingham Hall Hoard,
Norfolk’, Bulletin of the Board of Celtic Studies 28,
159-87

GILLAM, J.P., 1976 ‘Coarse Fumed Ware in Northern
Britain and Beyond’, Glasgow Archaeology Fournal 4,
57-80

GINGELL, C.J., with DAWSON, R.J., 1987 ‘An Earthwork
near Badbury Rings in Dorset’, Proc. Dorset Natur. Hist.
Archaeol. Soc. 109, 65-78

GRAHAM, A., and NEWMAN, C., 1993 ‘Recent Excavations
of Iron Age and Romano-British Enclosures in the
Avon Valley, Wiltshire’, WAM 86, 8-57

GRANT, A., 1984 ‘Animal Husbandry’ in B. Cunliffe,
Danebury: an Iron Age Hillfort in Hampshire, Vol 2. The
Excavations 1969-78: The Finds, CBA Research Report
52, 496-548 and fiche 16:A2—17:E9

GRAY, H. St GEORGE, 1937 ‘A Hoard of Late Roman Coins
from Shapwick Heath’, Proceedings of the Somerset
Archaeology and Natural History Society 82, 163-70

GREEN, C., and ROLLO-SMITH, S.M., 1984 ‘The
Excavation of Eighteen Round Barrows near Shrewton,
Wiltshire’, Proceedings of the Prehistoric Society 50,
255-318

British

42 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

GREENE, K., 1978 ‘Imported Fine Wares in Britain to AD
250: A Guide to Identification’, in Arthur and Marsh
(1978), 15-30

GUIDO, M., 1978 The Glass Beads of the Prehistoric and
Roman Periods in Britain and Ireland, Society of
Antiquaries of London Research Report 35, Leeds

HARDEN, D.B., PAINTER, K.S., PINDER-WILSON, R.H., and
TAIT, H., 1968 Masterpieces of Glass, London

HARDING, P.A., 1988 “The Chalk Plaque Pit, Amesbury’,
Proceedings of the Prehistoric Society 54, 320-27

HENIG, M., and CHAMBERS, R.A., 1984 “Two Roman
Bronze Birds from Oxfordshire’, Oxoniensia 49, 19-21

HENIG, M., and LEAHY, K.A., 1984 ‘A Bronze Bust from
Luaford Magna, Lincolnshire’, Antiquaries Fournal 64,
387-89

HINGLEY, R., 1989 Rural Settlement in Roman Britain,
London

HOLBROOK, N., and BIDWELL, P.T., 1991 Roman Finds from
Exeter, Exeter Archaeology Report 4, Exeter

ISINGS, C., 1957 Roman Glass from Dated Finds,
Archaeologia Traiectina 2, Groningen

KENT, J.P.C., 1978 Roman Coins, London

KINNES, I.A., GIBSON, A.M., AMBERS, J., BOWMAN, S.,
LEESE, M., and BOAST, R., 1991 ‘Radiocarbon Dating
and British Beakers: The British Museum Programme’,
Scottish Archaeological Review 8, 35-68

KIRK, J.R., 1949 ‘Bronzes from Woodeaton, Oxon’,
Oxoniensia 14, 1-45

KRZYSZOWSKA, O., 1990 ‘Ivory and Related Materials’,
Bulletin of the Institute of Classical Studies, Supplement 59

LANTING, J.W., and VAN DER WAALS, J.D., 1972 ‘British
Beakers as Seen From the Continent’, Heliniwmn 12,
20-46

LAWSON, A.J., 1976 ‘Shale and Jet Objects from Silchester’
Archaeologia 105, 241-75

LAWSON, A.J., 1993 ‘A Late Neolithic Chalk Plaque from
Butterfield Down, Wiltshire’, Antiquaries Journal 73,
183-85

LEACH, P.J., 1982 ‘Stone and Fired Clay Objects’ in PJ.
Leach, Ilchester Vol. I: Excavations 1974-75, Bristol,
Western Archaeological Trust Excavation Monograph
$.. 217223

MACGREGOR, A., 1985 Bone, Antler, Ivory and Horn, London

MALTBY, J.M., 1993 “The Animal Bones from a Romano-
British Well from Oakridge II, Basingstoke’, Proc. Hants
Field Club and Archaeol. Soc. 49, 47-76

MANNING, W.H., 1985 Catalogue of the Romano-British Iron
Tools, Fittings and Weapons in the British Museum, London

MCOMISH, D.S., 1989 ‘Non-hillfort Settlement and _ its
Implications’, in M. Bowden, D. Mackay and P.
Topping (eds.), From Cornwall to Caithness: Some Aspects
of British Field Archaeology, British Archaeological
Report British Series 209, Oxford, 99-110

MEPHAM L., 1993 “The Pottery’ in Graham and Newman
1993, 25-34

MILLETT, M., 1990 The Romanisation of Britain: an Essay in
Archaeological Interpretation, Cambridge

MORRIS, E.L., 1992a Data Levels Guidelines, Wessex
Archaeology Manual 2, Salisbury

MORRIS, E.L., 1992b The Analysis of Pottery, Wessex
Archaeology Manual 4, Salisbury

ORTON, C., 1980 Mathematics in Archaeology, Cambridge

PARTRIDGE, C.R., 1981 Skeleton Green: A Late Iron Age and
Romano-British Site, Britannia Monograph 2, London

PARTRIDGE, C.R., 1989 Foxholes Farm: A Multi-period Site in
Hertfordshire, Hertfordshire Archaeological ‘Trust
Monograph 1, Hertford

PEACOCK, D.P.S., and WILLIAMS, D.F., 1986 Amphorae and
the Roman Economy, an Introductory Guide, London

REECE, R., 1972 ‘A Short Survey of the Roman Coins found
on Fourteen Sites in Britain’, Britannia 3, 269-76

RICHARDS, J.C., 1990 The Stonehenge Environs Project,
English Heritage Archaeology Report 16, London

RICHARDSON, K.M., 1951 “The Excavation of Iron Age
Villages on Boscombe Down West’, WAM 54, 123-68

ROGERS, B., and RODDHAM D., 1991 “The Excavations at
Wellhead, Westbury 1959-66’, WAM 84, 51-60

ROSTOVTSEFF, M., 1923 ‘Commodus-Hercules in Britain’,
Journal of Roman Studies 13, 91-109

SCOTT, E., 1991 ‘Animal and Infant Burials in Romano-
British Villas; a Revitalisation Movement’, in P.
Garwood, D. Jennings, R. Skeates and J. Toms (eds.),
Sacred and Profane, Oxford University Committee for
Archaeology, Monograph 32, Oxford, 115-21

SEAGER SMITH, R.H., and DAVIES, S.M., 1993 ‘Roman
Pottery’ in P.J. Woodward, S.M. Davies and A.H.
Graham, Excavations at Greyhound Yard, Dorchester,
1981-84, Dorset Natural History and Archaeology
Society Monograph 12, Dorchester, 202-89

SMITH, I.F., and DARVILL, T.C., 1990 ‘The Prehistoric
Pottery’, in A. Saville, Hazleton North, Gloucestershire
1979-82: The Excavation of a Neolithic Long Cairn of the
Cotswold-Severn Group, English Heritage Archaeology
Report 13, London, 141—52

SMITH, R.J.C., forthcoming ‘Excavations along the route of
the Dorchester Bypass 1986-88’, Wessex Archaeology
Report, Salisbury

STEAD, I.M., and RIGBY, V., 1989 Verulamium: The King
Harry Lane Site, English Heritage Archaeology Report
12, London

STRONG, D.E., 1976 Roman Art, London

STRUCK, M., 1993 ‘Kinderbestattungen in romano-
britischen Siedlungen — der archaeologische Befund’, in
M. Struck (ed.), Rémerzeitlicher Graber als Quellen zu
Religion, Beviélkerungsstruktur und Sozialgeschichte,
Mainz, Archaologische Schriften Institutes fur Vor- und
Fruhgeschichte der Johannes Gutenberg-Universitat
Mainz 3, Mainz, 313-18

SWAN, V., 1971, “The Coarse Pottery’, in Wainwright,
1971, 100-16

THOMAS, J., 1991 Rethinking the Neolithic, Cambridge

TOLSTOI, J., 1908 Monnaies Byzantines, St Petersburg

TYERS, P., 1978 “The Poppy-head Beakers of Britain and
their Relationship to the Barbotine Decorated Vessels of
the Rhineland and Switzerland’, in Arthur and March
(1978), 61-108

VARNDELL, G., 1991 “The Worked Chalk’, in IH.
Longworth, A. Hearne, G. Varndell and S.P. Needham,
Excavations at Grimes Graves, Norfolk, 1972-76,
Fascicule 3: Shaft X; Bronze Age Flint, Chalk and Metal
Working, London, 105-106

VEEN, M. VAN DER, 1989 ‘Charred Grain Assemblages
from Roman Period Corn-driers in _ Britain’,
Archaeological Journal 146, 302-19

VINER, L., 1982 ‘Copper Alloy Objects’, in A. McWhirr, L.
Viner and C. Wells, Romano-British Cemeteries at
Cirencester, Cirencester Excavations 2, Cirencester, M2:
B8—C4

PREHISTORIC AND ROMANO-BRITISH SITES AT BUTTERFIELD DOWN 43

VCH 1957 Victoria County History of the Counties of England.
A History of Wiltshire, Volume I, part 1, London

WAINWRIGHT, G.J., 1971 ‘The Excavation of Prehistoric
and Romano-British Settlements near Durrington
Walls, Wiltshire, 1970’, WAM 66, 76-128

WEDLAKE, W.J., 1982 The Excavation of the Shrine of Apollo
at Nettleton, Wiltshire, 1956-71, Society of Antiquaries
of London Research Report 40, London

WESSEX ARCHAEOLOGY, 1990 ‘Butterfield Down,
Amesbury. Stage 1 Archaeological Assessment —

Surface Collection’, Salisbury, unpublished client
report

WILLIAMS, D.F., 1977 ‘The Romano-British Black-
Burnished Industry: An Essay on Characterisation
by Heavy Mineral Analysis’, in D.P.S. Peacock (ed.),
Pottery and Early Commerce: Characterisation and
Trade in Roman and Later Ceramics, London,
163-220

YOUNG, C.J., 1977 Oxfordshire Roman Pottery, British
Archaeological Report British Series 43, Oxford

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 44-72

A Romano-British Farmstead and Associated Burials
at Maddington Farm, Shrewton

by JACQUELINE I. McKINLEY and MICHAEL HEATON

with contributions by RACHAEL SEAGER SMITH, DAVID MURDIE, MICHAEL J. ALLEN,
JOHN A. DAVIES, S. HAMILTON-DYER, P. HINTON and R. MONTAGUE

The discovery of two burials during construction of a pipeline booster station by Esso Petroleum Co Ltd
at Maddington Farm, Shrewton, led to full excavation of the site. The site proved to be part of a small
Romano-British farmstead of probable 3rd—4th century date, comprising a small circular posthole

structure with associated hearths and pits adjacent to field boundary ditches that appeared to have also

functioned as the focus for a small inhumation cemetery.

INTRODUCTION

The Site

In 1961, Esso Petroleum Co Ltd laid an oil supply
pipeline from the Fawley Refinery at Southampton
to Avonmouth near Bristol (Figures 1 and 2,
context 1246), a distance of some 80 miles. More
recently, as part of the ongoing programme of
management and maintenance, it became apparent
that a ‘booster station’ was required to maintain oil
pressure across the varied relief crossed by the
pipeline. For safety reasons such installations
require accessible sites away from built-up areas.
Maddington Farm near Shrewton was chosen as the
most suitable site, partly on the grounds that it
impinged on no known archaeological sites. An
environmental impact assessment of the proposed
site and its immediate surroundings (RSK
Environment Ltd 1992) revealed some evidence of
prehistoric activity and it was recommended that a
watching brief should be conducted during
preliminary stages of construction.

The pipeline passes c.400m_ north-east of
Maddington Farm which lies on the B390 2km west
of the village of Shrewton, between Salisbury and
Devizes (Figure 1). The farm buildings are situated
on the north side of the road at the bottom of one of
the numerous dry valleys that dissect the southern
edge of Salisbury Plain, joining the Till valley at
Shrewton. The ground rises to the north, south and
west to the level of the surrounding downs, at
approximately 120m OD. The downs are crossed by
a network of bridleways and footpaths, one of which

follows a south-west—north-east route from the east
side of the main farm buildings, crossing an
east-west bridleway c.140m north of the site. The
area of excavation lay immediately to the west of this
bridleway, c.500m from the farm buildings at SU
0490 4450, on arable land on the south-facing slope
just below the summit of the Chalk spur (Figure 1).
The soils of the area are humic rendzinas and typical
palaeo-argillic brown earths. Clay-with-flints also
occurs locally.

Construction work, in September 1993, com-
prised slight re-routing of the existing oil-carrying
pipe through the booster pump set several metres
into the natural chalk (Figure 2: pipe diversion
trench). To the north of the pump the ground was
terraced to house the station buildings and the area
to the south was stripped and made up with scalpings
to provide hard-standing for the contractors prior to
landscaping. Emptying the 1961 pipe trench of
backfilled material revealed that the pipelaying
operations had cut through a formerly unnoticed
burial (121) outside the construction site. A second
burial (128) was found in the north edge of the 1993
pipe diversion trench. Wessex Archaeology was
instructed to excavate fully the construction site to
the north of the pipe diversion trench.

Archaeological setting

Maddington Farm lies on the southern periphery of
Salisbury Plain, close to the Stonehenge and
Avebury World Heritage Site, and within one of the
richest concentrations of prehistoric monuments

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 45

Contours in metres OD

1 Ze Boe
ee pe TE) | m7

[ ] DEVIZES

SALISBURY. |

409

Figure 1. Maddington Farm, Shrewton: site location plan

and archaeological sites in Europe. Knowledge of
the Romano-British period has increased substan-
tially in recent years as a result of the ongoing
surveys of South Wiltshire and the Salisbury Plain
Training Area (SPTA) by the Royal Commission on
the Historical Monuments of England. Intensive
settlement and land-use in and around the area of
the Plain in the Romano-British period is now
apparent, often demonstrating a continuum of use
from the prehistoric to the late Romano-British
periods (M. Corney pers. comm.). Ten Romano-
British settlements, evident as earthworks, have been
reported to date (WAM 1994) ranging from the 6-ha
settlement at Compton Down to the c.22-ha village
at Charlton Down, both c.10km north-east of the
site. The overall picture shows a range of occupation
from farmsteads to large-scale settlements, including
several villas, and a complex field system extending
over the Plain.

The excavation

The excavation was restricted to the c.45.5 x 40.0m
area of terracing north of the pipe diversion trench.
Although adjacent areas to the south and east,
including the bridleway up to the site, were stripped
of topsoil and prepared for contractors’ compounds
and vehicular access, no subsoil disturbance was
involved and hence no threat was posed here to
archaeological deposits.

There was no space on the site for topsoil storage
and stripping was undertaken by the pipeline
contractors on a piecemeal basis during upgrading of
the vehicular access. Archaeological excavation pro-
ceeded as areas became available, generally about
one third of the site being accessible at a time. This
necessitated more tracking and spoil handling than
would normally be acceptable on archaeological
sites, resulting in some disturbance and compression
of the uppermost archaeological deposits. A team of

46 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

[ 1120 : rae CMe os ae + ee oe 7
1128. 1143 1150 KB a
1191."

1130 &D1084 re ae
1126 - 922° 62° 400 yer? o

) 1189! : ny

1 1065 &
1090 11030 — tah % e
fe) zy ' 2

39 5 1135 NTFS y, 1187 2 177@
1124 1122 : 5° of 1206

t
1
at

| 1002
[ee

Digg, Section (Fig.3) _
N,

Key: :

e@ Postholes

o Unexcavated
postholes

: .. Stakeholes

Figure 2. Maddington Farm, Shrewton: plan of excavated area showing all features

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 47

four archaeologists (assisted on one occasion by the
main contractors) worked for four weeks between
20 September and 14 October. Archaeological
deposits were examined in sample slots, half-sections
or by total excavation in the case of inhumations.
Bulk soil samples were taken from the visually most
promising dated contexts and processed for palaeo-
environmental materials. A single soil column was
extracted from a machine-cut section through a
colluvial deposit for molluscan analysis (Figures 2
and 3).

The archive of materials, site records and
detailed reports is stored at Wessex Archaeology’s
offices at Old Sarum under the archive code W564;
it will be deposited at Salisbury Museum in due
course.

Archaeological deposits
by DAVID MURDIE and JACQUELINE I. MCKINLEY

All archaeological deposits were contained within
features cut into natural chalk. The site having been
ploughed, there were no archaeological deposits
surviving above the natural chalk and few
stratigraphic relationships to support detailed site
phasing.

Structure 1118

In the north-western area of the site, an almost
complete circle of eight postholes at c.2m intervals
formed the outline of a 5.5m-diameter structure
around a single central posthole. A tenth posthole
may have been masked or destroyed by feature
1135 which could account for the apparent gap in
the north-eastern side of the circle. The postholes
were of approximately 0.34m diameter and 0.17m
depth, with vertical sides and flat bases, and filled
with dark brown silty loam. There were nine other
postholes situated immediately north of 1118, with
similar dimensions and fills. Their apparent
concentration in the immediate vicinity of the
structure suggests that they might be associated
with it.

Hearths

There were six hearths, all situated in the north-
western area of the site, three within 3m of the south
edge of structure 1118 (1038, 1034 and 1051), and
three in a small group in the north-west corner of the
site (1090, 1084 and 1128). Of these, 1038, 1034
and 1090 were shallow, cut to a maximum surviving
depth of 0.16m, rectangular to ovate in plan,
1.10-1.50m long by 0.55—0.75m wide, with steep

sides and flattish bases which were discoloured and
scorched. In each case the bottom fill contained
large quantities of charcoal, sealed beneath layers of
silt and silty clays containing varying amounts of
chalk fragments.

Features 1051 and 1084 each consisted of two
contiguous circular or bowl-shaped depressions, the
former 1.50m long with a maximum diameter of
0.56m and maximum depth of 0.19m, and the latter
1.25m long with a maximum width of 0.50m and a
maximum depth of 0.18m. The fills were similar to
those of 1034 and 1038 and evidence of burning was
observed on the bases and sides.

Feature 1128 consisted of a circular depression
connected via shallow linear troughs to two smaller
pits to the west and south. The two pits (1126 and
1130) contained black silty clay primary fills beneath
compact, fire-reddened powdery chalk, whilst 1128
itself contained a single fill of charcoal-flecked
yellowish-brown silty clay and chalk rubble.

All the hearths produced Romano-British pottery
and burnt flint with limestone and quern fragments,
iron nails, fired clay (possibly hearth lining) and flint
flakes.

Pit 1195

Situated in the south-west part of the site, c.lm
north-west of the western terminal of ditch 1157,
this feature displayed a bell-shaped profile, 2m wide
at the surface and 1.7m at the base, with a depth of
1.33m. The primary fills (1199, 1201, 1210, and
1217-1220) comprised thin layers of highly organic
silts lying against the sides of the pit but tipping
predominantly from the eastern edge, with a
combined thickness of 0.3m. They were sealed by a
0.5m-thick dump (1198) of dark greyish-brown silt
that extended the full width of the pit, sealed below
a 0.1m-thick layer of loose chalk rubble (1197). The
uppermost fill (1196) was a yellowish-brown silty
clay with occasional chalk and flint inclusions, cut
on the eastern side by feature 1215. Most of the
finds were derived from fills 1196 and 1210 and
included Romano-British sherds, animal bone,
worked bone pins (SFs 76 and 77), worked flint,
worked stone fragments and ceramic building
material.

Linear features

A series of linear features crossed the site diagonally,
apparently enclosing the main concentration of
features described above and all with at least one
terminal within the site. The earliest (1064) ran
east-west and was cut by one of a series of three

48 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

concentric curvilinear ditches (1155, 1156 and
1157) which curved slightly from the west corner of
the site to the north-east.

Feature 1064, excavated in four segments, was
22m long and aligned east-west with distinct
terminals at each end. It had almost vertical sides
and a flat base, with a maximum width of 0.70m and
maximum depth of 0.53m at the western terminal,
becoming gradually shallower and narrower towards
the eastern end where it was 0.5m wide by 0.3m
deep. The single fill comprised a chalky, dark
yellowish-brown silty clay. The western terminal cut
the eastern edge of animal burial pit 1043 (see
below), and although these features are separated
stratigraphically, their close juxtaposition suggests
they may be roughly contemporaneous. Fragments
of human bone were recovered from an area 1.2m
west of the eastern end of the feature where they
appeared in concentration: it was unclear whether
they represented a disturbed shallow inhumation
burial or redeposited bone. Feature 1064 cut the
upper fill of feature 1022 (see below) and was itself
cut by ditch 1155.

Features 1155, 1156, and 1157 had similar bowl-
shaped profiles, 0.35-0.50m wide by 0.10—0.15m
deep, were co-terminal and appear, stratigraph-
ically, to be contemporaneous. The longest (1155),
excavated in five segments, was at least 35.5m
long, extending beyond the northern edge of the
excavation. It curved across the centre of the site
from west to north-east, cutting the upper fills of
ditch 1064, and was apparently respected by 1157
and 1156. The most northerly of the three (1157)
was 10.5m long, with three phases of a distinct
terminal at its west end, and with a_ possible
continuation (1278) 3m beyond its less well-defined
eastern end. The southernmost of the three, 1156,
excavated in three segments, had distinct terminals
at both ends and appeared to cut 1269, the fills of
the quarry pits/lynchet (see below). All three features
contained single fills of chalky, brown silty clay with
occasional secondary deposits of flint nodules. Of
the eleven defined layers from the excavated
segments, four contained worked flint and Romano-
British pot sherds.

There were twenty-seven stakeholes clustered
around the south-western terminals of ditches 1157,
1156 and 1155, within an area c.3 x 3m. Excavation
of six revealed narrow profiles, 0.07m wide by 0.08m
deep. The absence of similar features from the rest of
the site suggests that these are indeed stakeholes and
that they are associated with the ditches. None
contained finds.

Postholes

In addition to the postholes described in relation to
structure 1118, a further nine postholes were
identified. Five of these occurred within the hearth
group in the north-west corner of the site, three
cutting 1269, the fills of the quarry pits/lynchet (see
below) in the south-east of the site, and one isolated
in the centre of the site. The postholes varied in
shape from oval to square and were 0.26—0.85m in
diameter and 0.13—0.33m deep. All were filled with
brown silty clay and contained small quantities of
artefacts, including fragments of Romano-British
pottery, animal bone, worked flint and an iron nail.
No structural groupings were apparent.

Quarry pits and the lynchet, context 1269

The site was dominated by an extensive linear spread
(1269) of mixed loamy materials describing a broad
amorphous arc running north-east to south-west,
apparently bounded on the west by the linear features
described above. It corresponded to a broad linear
surface depression that developed during initial plant
movement across the site, and was seen to extend
beyond the north edge of the site running across the
natural contours. It was investigated in a series of hand
excavated sondages (Figure 2, various) and a single
machine-excavated trench (Figure 2, 1288; Figure 3),
that collectively revealed a roughly linear concentration
of amorphous intercutting pits in a rather ambiguous
relationship to the grave group described below and a
shallow colluvial deposit partly contained within
them; some, such as 1244, 1221 and possibly 1224,
were visible as discrete features on the surface, while
the majority were defined only after removal of the
overlying deposits in ‘box trench’ sondages.

The sondages revealed amorphous, intercutting
pits typified by 1022. This was trapezoidal in plan,
4.0m long by 3.8m wide with irregular sides and a
generally flat base 0.6m deep, and had been cut
through the west end of grave 1026 (see below). It
was filled with several erratically interleaved layers of
silty clays, dumps of compacted chalk fragments and
lenses of fine, dark humic loam, the uppermost of
which was cut by ditch 1064. Generally, artefacts
were found only in the upper fills of these features,
with the bulk, including fragments of human and
animal bone, pot sherds, flints, fragments of non-
local stone, and one iron nail, recovered from a single
context (1060) that filled features 1059 and 1061.

The machine sondage 1288 (Figure 2, 1288; and
Figure 3) across the centre of 1269 revealed a
shallow, typical calcareous colluvium (1250-1253)
(Allen 1992), sealing the chalk rubble fills of a

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM

ULUNOS UvOsNT[OU! Jo UOTsod Surmoys “ayoudy pue sud Arrenb aul ‘soysirp Iesur] ssoioK UOTDas sUOIMaTYS “Wey UOISUIPpeYy *¢ 2INS1.J

S9°06S ave
feee 968 Goweo-stt

wud

saijouc — -— —= eee eee)

Z T 0

yn? : ~
osnTTOW aowzorert 3846 =

50 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

number of underlying quarry pits (Figure 3, 1266
and 1267). The colluvium appeared to be resting in a
shallow negative lynchet in the hillside formed by the
western edge of the quarry pits, represented here by
feature 1266, and may originally have extended out
over feature 1267 as well. Stabilisation horizons were
clearly visible within it, one of which (1268) was
marked by an artefact-rich layer (1252) containing
fired-clay, animal bone, pottery and tile. The
uppermost layer (1250) was cut by ditches 1155 and
1156. A single column of soil samples was taken
through this combined sequence for mollusc analysis.

Graves

Grave 121 lay outside the excavation area,
approximately 16m north-west of the bridleway on
the southern edge of the 1961 oil pipeline trench
(1246) by which it had been cut; it formed a shallow
scoop cut into natural chalk, c.0.75 x 0.50m,
containing the badly fragmented upper and lower
limb bones from an inhumation.

Grave 128 was excavated in section from the north-
ern face of the 1993 oil pipe diversion trench. A sub-
rectangular vertical-sided and flat-bottomed cut of
uncertain dimensions, it contained a flexed inhum-
ation (124) with iron hobnails along the soles, aligned
east—west with the skull to the west, and was sealed by
lynchet fill (122) equivalent to 1253 (above).

Grave 1002 was situated on the western edge of
the site. A shallow, rectangular cut with vertical sides
and a flat base, 2.40 x 1.20m, and 0.5m deep, it was
cut into natural chalk and contained a badly crushed,
extended, prone inhumation (1005) aligned north—
south with the skull to the north. A skeleton of a
small dog (1019) laid on its left side, lay head to head
below the upper torso. The presence of a coffin was
indicated by layer 1004, a dark rectangular area of
silty clay extending around the inhumation and
defined by 41 coffin nails in two lines at the head and
foot of the grave (SF 30-59). Other artefacts include
a bronze coin, and an As of Faustina II (AD
161-175; SF 60) recovered from the mouth.

Grave 1008 (Figure 4) lay on the western edge of
the site, 0.80m north of grave 1002. A rectangular cut
2.15 x 1.15m, with vertical sides and a flat base 0.70m
deep into natural chalk, it contained an extended,
supine inhumation (1012), aligned north-south with
the skull to the south. A coffin was indicated by a
rectangle of pinkish grey silt (1011) extending to
within 0.18m of the grave edge, defined by 42 iron
coffin nails extending 0.5m from the head and foot
ends of the grave to a height of 0.40m from the base.
Other artefacts comprised five very small copper-alloy

rivets recovered with leather fragments (one 77 sztz)
around the disto-medial sides of the feet (SF 29) and
a single sherd of Romano-British pottery.

Grave 1026 was situated in the north-east area of
the site on the eastern edge of the quarry pits. A
shallow ovate feature, 0.9m x 0.75m, it was cut into
1269 (quarry pit/lynchet fills) but, curiously, was
also cut at its west end by quarry pit 1022 (see
above). It had sloped sides and a flat base, 0.07m
deep, and contained a badly disturbed inhumation
(1037), crouched on the left side, aligned south-west
to north-east, the skull removed by 1022.

Grave 1177 was in the north-eastern corner of
the site, forming a shallow 1.25m x 0.85m ovate cut
into natural chalk with gently sloped sides and a flat
base, 0.12m deep. It contained a disturbed inhuma-
tion (1179), tightly crouched on the left side, aligned
south-east to north-west with the skull at the south-
east. There was charcoal flecking in the grave fill
within the immediate area of the skeleton.

Animal burials

Feature 1043, situated in the central western area of
the site, was cut by the western terminal of ditch
1064. Oval in plan, 1.40 x 1.10m, with vertical sides
and a flat base cut into the natural chalk to a depth of
0.62m, it contained the articulated skeleton of a calf
(1058) laid on its right side with the skull to the west
and feet together. The burial was placed directly on
the chalk base and sealed beneath a single deposit
(1044) of chalk rubble and silty clay.

Feature 1154 was situated 2.50m east of the
eastern terminus of linear feature 1064 and directly
in line with 1064 and feature 1043. Similar in plan
to 1043, 1.30m long x 0.60m wide, though with a
bowl-shaped profile of only 0.1m = depth, it
contained the disturbed fragments of a cattle
skeleton, almost certainly disturbed during earlier
plant movement. The feature was filled with a
chalky, dark brown silty clay (1153).

Feature 1143, situated within the hearth group
in the north-west corner of site, comprised a 1.2m
long by 0.64m wide rectangular cut in the chalk,
with steep sides and a flat base to a depth of 0.37m.
It contained the articulated skeletons of five piglets,
all oriented north-south, sealed beneath a chalky,
dark yellowish-brown silty clay (1144) from which
two small abraded sherds of Romano-British
pottery were recovered.

Pyre debris pits
Three features (1224, 1240 and 1244) all situated at
the southern margin of the site, cutting the lynchet

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM ail

119.53mOD
DN

Area of leather and small
copper alloy rivets

118.83mOD
uN

Depth of coffin nails:
mOD
OO 119.08 -118.99
© 118.98 - 118.89

A 118.78 - 118.69

118.82mOD
nw

fue

O-

118.78mOD
ms

15A ©O7

119.35mOD
a

Figure 4. Maddington Farm, Shrewton: grave 1008

i
|
|
|

52 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

deposits (1269) described above, contained relatively
large quantities of fire-derived materials such as
charcoal and burnt flint, and cremated human bone.
In view of the absence of these materials from the
rest of the site, particularly the cremated bone, they
are assumed to have been deliberate deposits. On the
basis of comparisons with other, as yet unpublished,
cemetery sites at Baldock and East London
(McKinley forthcoming), they are referred to here as
pyre debris pits.

Feature 1224, the largest of the three, 4.0 x 2.60m
wide and 0.60m deep, was sub-rectangular in plan
with a very poorly defined upper edge. Excavation of
a quarter segment of the pit revealed its true
dimensions to be at least twice those apparent on the
surface. It was filled by a single deposit (1225) of very
dark greyish-brown silty clay containing abundant
charcoal, burnt flint, cremated human bone, animal
bone, Romano-British and Iron Age pottery.

Feature 1240, cut away on its south-west side by
the 1961 oil pipeline trench (1246) was sub-
rectangular in plan, 1.80m long north to south by at
least 1m wide, with steep sides and a flat base 0.30m
deep. It was filled by a single deposit (1241) of very
dark brown silty clay containing an abundance of
charcoal flecks, burnt flint and cremated human
bone. The feature was well-defined and cut into the
surface of the lynchet deposits.

Feature 1244, in the centre of the group, was sub-
rectangular in plan, 1.60 x 0.80m, with sloping sides
and a flat base 0.22m deep. It was filled by a single
deposit (1245) of very dark brown chalky, silty clay
containing abundant charcoal flecks, burnt flint and
cremated human bone.

Miscellaneous pits and hollows

The remainder of the excavated features comprised
shallow amorphous pits and hollows of dimensions
varying between 0.6m to 2.25m in length to 0.46m
deep, and of varying orientations. They were concen-
trated principally in the north-west corner of the site
and all were filled with similar brown silty clays
containing few artefacts, and of no apparently obvious
function or association, although the concentration of
features in this area is perhaps significant.

Finds

Roman Coins
Identified by JOHN A. DAVIES

SF 60, context 1005; resting on the palate of
inhumation burial 1005.

Faustinia II, As, AD 161-175. Obv. and rev:
illegible.

SF 74, context 1000: topsoil.

Constantius II, AE2, AD 353-355. Obv: DN
CONSTAN [TIVS PF AVG]; rev: [FEL TEMP
REPARATIO] falling horseman. Mintmark missing
(off-flan).

Metalwork
by R. MONTAGUE

Two items of copper alloy and about 264 of iron
were recovered. The exact number of iron objects
cannot be given as some of the hobnails from grave
128 were fragmentary. The metalwork was recorded
in its unconserved state; detailed records are
available in the archive.

Copper alloy objects

A spatulate sheet object, possibly part of a toilet set,
was found in the fill of feature 1224. Five identical
copper alloy rivets, 2.10x 1.60mm, with one flat
head and the other hemispherical, were recovered
from around the disto-medial foot bones of skeleton
1013 (Figure 4), and may represent some form of
decoration on leather sandals. One rivet remained in
situ, set through a fragment of leather with the
hemispherical head uppermost.

Tron cleats

Three cleats were recovered. Two came from the
cluster of hobnails along the soles of the feet of
skeleton 124 in grave 128. The occurrence of cleats
together with hobnails as a form of both strength-
ening and protection against excess wear of shoe soles
is not unknown: six of the 144 graves at Lankhills,
Winchester, which contained hobnails also contained
cleats (Clarke 1979, 322-325, figures 38 and 39).
The third cleat was recovered from feature 1223.

Tron hobnails

A total of 162 hobnails was recovered, with
approximately 160 complete and _ fragmentary
examples from the soles of the feet of skeleton 124 in
grave 128 (c.80 from each foot). This total is fairly
high. Of the 43 graves from the Romano-British
cemetery at Poundbury, Dorchester, only two
produced a higher total (c.290 and c.325). Peaks in
the frequency distribution were noted at c.10, c.35
and c.50 hobnails per boot (Mills 1993a, 99; table
10). A hobnail was also recovered from the fill of
posthole 1084, and another from the fill of pit 1195.

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 53

Coffin nails

A total of 83 coffin nails of Manning’s (1985, 134)
Type 1 was recovered from graves 1002 (41) and
1008 (42, Figure 4). These were all very similar in
form, with flat round heads (diam. 13.6-21.6mm)
and square-sectioned shanks (length 59.1—78.3mm).
This type of coffin nail is common in Romano-
British cemeteries (Mills 1993b, 114-116). Some
nails were clenched over, indicating a thickness for
the wooden planks of 23.9-56.2mm. Twenty had
observable traces of mineralised wood, variously
located under the head, along the shank or in both
areas.

The coffin nails in both coffined burials were
clustered at the foot and head; none occurred along
the sides in 1002 and those in 1008 spread to a
maximum extent of 0.50m along the sides (Figure
4). Similar distributions were observed at Alington
Avenue, Dorchester (Romano-British graves 268
and 3661; Davies et al., forthcoming), and in many
of the 451 graves at Lankhills (Clarke 1979, figures
47-66).

Other nails

Other nail types include seven flat- and round-
headed nails which were clustered at the waist area
of the dog burial 1019 in grave 1002. These are
much shorter than the coffin nails from grave 1002,
with complete examples averaging 24.8mm in
length, and may represent the remains of some sort
of fitting or harness worn by the dog. Six other nails
were recovered, all from features producing pottery
of Romano-British date.

Tools
An iron awl, measuring 96.5x5.0mm, was
recovered from feature 1193, and a probable knife
handle (cf. Manning 1985, pl. 53, Q5) was found
during topsoil stripping of the central area of the
site.

Flint

Forty-one pieces of worked flint were examined and
recorded by Philip Harding; details are in the archive.
According to his observations the assemblage
principally comprised small and frequently broken
flakes, with a single example of a double-sided
scraper. There were no cores or other primary waste.
The material varies in patination and condition and,
as most of it was recovered from contexts otherwise
dated to the Romano-British period, can be
considered largely residual on this site.

Pottery
by RACHAEL SEAGER SMITH

The pottery assemblage comprises 563 sherds
(4800g), of which the majority are of Romano-
British date; small quantities of prehistoric pottery
are also present.

The pottery has been analysed in accordance
with the standard Wessex Archaeology recording
system for pottery (Morris 1992). It was divided into
four broad fabric groups based on dominant
inclusion types: flint gritted (Group F), grog-
tempered (Group G), sandy (Group Q) and fabrics
of known type or source (Group E). These broad
groups were then further subdivided into 24 fabric
types based on the range and coarseness of
inclusions present. The following terms are used to
describe the quantity of inclusions: occasional = less
than 1%; rare = 1—2%; sparse = 3-7%; moderate
= 10-15%; common = 20-25%; abundant = 30%+.
Each of these fabrics was assigned a unique,
chronologically significant fabric code.

The pottery has been quantified (using both
number and weight of sherds) by fabric type for each
context and details of vessel form, surface treatment,
decoration and manufacturing technique have been
recorded. Pottery fabric totals are given in Table 1,
and Table 2 summarises the vessel forms present in
each fabric type.

Prehistoric pottery

The prehistoric pottery has a potential date range
from the Late Neolithic/Early Bronze Age to the Iron
Age although dating is hampered by the lack of
diagnostic sherds. Three fabric types were identified:

Fl. Hard fabric; sparse to moderate crushed angular flint
<1.5mm, sparse quartz <0.25mm, rare iron oxides
<0.5mm. Unoxidised.

F2. Hard, fine-grained sandy matrix with sparse crushed
angular flint <4mm. Unoxidised.

G1. Soft fabric; rare to sparse grog <lmm, rare quartz
<0.5mm, rare iron oxides, soft white particles, both
<0.25mm. Predominantly oxidised but interior surface,
margin and core may be unoxidised.

Late Neolithic/Early Bronze Age Beaker activity in
the area is indicated by the fabric G1 sherds. Two of
these, from pit 1195, have square-toothed comb
impressed decoration and, while undecorated, the
three joining sherds of this fabric from quarry pit
1022 are probably also from a Beaker vessel. Both
the flint-gritted fabrics are represented by featureless
body sherds from thick-walled vessels only, but
based on the appearance of the sherds, fabric F1 is

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Table 1. Distribution of pottery by feature and fabric, showing number and weight of sherds

| PREHISTORIC ROMANO-BRITISH COARSEWARES

Feat

cont

1001

FINEWARES

G1|F100} Grog | Q Q | Clay | BB1 | NF p | Unprov | NF Oxf

Sam
coar |glauc]| discs fine fine

fine

1001

1013

—

1 5 | 4 5 D
log |35g | 10g 16g Alg

res

1183

1185

OW S192 13
18
1195

| PREHISTORIC | ROMANO-BRITISH COARSEWARES

FINEWARES

NF p|} Unprov | NF Oxf

fine | fine

ce Fl | F2} Glj F100) Grog Q | Clay} BB1
es | discs
.
Pe

Beles etliole bok

are

He i
aa x3]
1161

ance
1184 ee 3
pret ze| ooe| | aa

1186
xe

1199

os
“ES ee |
1195] 1210 44 =

470g
1195} 1198 1
mall 20g
1195] 1196 2
5g
1206| 1207 1
2g
1
Ag
2
3g

1244] 1245
1264) 1248 1

5g
1268] 1251 | 2 Onlik2 1

190g 651g] 70g 19g
1269] - 3 1 2

og 5g 29g

Eze

1276| 1277 15 Pea eas

54g 37g
Total nos. 12| 4 alla 72 252) 938 WP 7 12 52 8 19 5
diotal 15 | 82] 11) 3 452 |1703} 280} 733 | 687] 111 486 86 142 | 9
weight (g)
Key: Grog = fabrics G100 and G101 NF fine = fabrics E161 and 162

Q coar = fabrics Q100, Q102, Q106, Q197, Q110 and Q111 feat = feature
Q glauc = fabrics Q101 and Q103 cont = context

Unprov fine = fabrics Q104, Q105 and Q109

56 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

probably of Middle—Late Bronze Age date while
fabric F2 is probably Iron Age.

Sources for the prehistoric fabrics are uncertain
although all three are likely to be fairly local. Grog
was the predominant inclusion amongst the Beaker
sherds recovered in the area of the Stonehenge
Environs Project (Cleal with Raymond 1990,
237-238). No direct parallels were found between
the flint-gritted fabrics and the material from the
Avon Valley (Mepham 1993) or Butterfield Down,
near Amesbury (Millard 1996) although flint-gritted
wares of similar broad date bands do occur at both
these sites. The three joining body sherds of the sand
and flint-gritted fabric F2 from pyre debris pit 1224
may have been deliberately trimmed to form a
roughly circular counter.

The distribution of the prehistoric pottery by
feature is given in Table 1. The majority of
prehistoric sherds occurred in association with
Romano-British material and must therefore be
considered residual. In three cases, linear feature
1155, the calf burial 1043 and the possible pyre
debris pit 1244, sherds of prehistoric pottery were the
only datable artefacts recovered and may therefore
indicate the date of the feature, but the possibility
that they are redeposited cannot be excluded.

Romano-British pottery
‘Twenty-one fabric types were identified and these
have been divided into coarse- and finewares.

Finewares

Seven fineware fabrics were identified. The only
Continental imports recognised are five sherds (9g)
of samian. These sherds have not been assigned to
any production centre and all are very small and
abraded. Only one vessel form was recognised, a
Dragendorff 33 cup, a type common from the late
1st to late 2nd centuries AD.

The other finewares from known sources comprise
various products of the late Romano-British Oxford
and New Forest production centres. Oxfordshire red
colour-coated wares are the most numerous amongst
this group. Vessel forms include flanged bowls (Young
1977, 160, type C51), hemispherical bowls with bead
rims (zbid., 160, type C55); mortaria are indicated by
the presence of body sherds only. New Forest
finewares comprise both the red-slipped and colour-
coated wares (parchment wares are here considered
to be part of the coarseware assemblage). The only
vessel form recognised was the indented beaker
(Fulford 1975, 50, type 27) but all the sherds of this
fabric are from closed forms.

Three fineware fabrics from unknown sources
were also recognised:

Q104. Hard fabric; common quartz and rare iron oxides
both <lmm. Oxidised, often with thin unoxidised core.
Wheelmade.

Q105. Hard, fine-grained fabric; rare to sparse iron oxides
both <0.25mm. Oxidised. Wheelmade.

Q109. Hard, fine-grained fabric; sparse quartz <0.5mm,
rare to sparse iron oxides <2mm and sparse mica
<0.25mm. Oxidised with unoxidised inner margin and
core. Wheelmade. Exterior surface coated in a thick,
creamy-brown slip.

The source and date of the buff sandy ware (fabric
Q104) are uncertain although it is likely that these
sherds are derived from flagon forms. It is possible that
the very fine-grained oxidised ware is an example of a
minor New Forest product, in which the red-slipped
ware fabric, fired to a higher than normal temperature
although not reaching that typical of the colour-coated
wares, is used to produce open bowl forms more
usually found in the coarser, sandier parchment ware
fabrics. This fabric is not mentioned by Fulford
(1975) but examples do occur amongst the kiln
assemblage from Pitt’s Wood (Swan, in preparation).

All the sherds in the white-slipped red ware fabric
(Q109) derive from a single flagon, the neck and rim
of which is missing. The application of a white slip to
hide an otherwise red firing fabric is a phenomenon
found widely across southern England but the
production centres of these vessels remain largely
unknown. White-slipped red wares are generally
dated from the mid Ist—late 2nd century AD.

A greater range of imported fineware fabrics was
recovered from the sites at Figheldean (Mepham
1993) and Butterfield Down, Amesbury (Millard
1996). This, however, is likely to be a factor of
chronology; at Shrewton the majority of activity
seems to have occurred from the later 2nd century
AD onwards, outside the period of currency of these
finewares. ‘Together, the Oxfordshire and New
Forest finewares form 7% of the assemblage, a
higher figure than that from Durrington (Swan
1971), Figheldean (Mepham 1993) and Butterfield
Down (Millard 1996). However, at all these sites,
the same emphasis on closed forms from the New
Forest and open forms from the Oxfordshire region
is apparent, even though the numerical importance
of sherds from these centres is more variable.

Coarsewares
Fourteen coarseware fabric types were identified,
including two of known source. Four of these fabrics

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM Dil

Table 2. Romano-British coarsewares: vessel form by fabric, giving number of occurrences

VESSEL FORMS

eds | LO

—_—_+

Key to Vessel Form:

1=flat-flange bowl/dish; 2=everted rim jar; 3=dropped-flange bowl/dish;
4=bow!l with inturned rim; 5=attenuated jar; 6=flagon body; 7=lid;
8=clay disc; 9=everted rim with squared end; 10=dog-dish; 11=rolled rim storage jar;

12=flagon rim; 13=necked jar with a hooked rim;

14=narrow necked jar with a rolled rim; 15=small everted rim jar; 16=flanged bowl;
17=high shouldered jar; 18=necked jar with groove at junction of neck and shoulder.

are ‘catch-all’ types and include products from more
than one source. The correlation of fabric types and
vessel forms is shown in Table 2.

E101. Black Burnished ware (BB1); for fabric description
see Williams (1977).

E160. New Forest Parchment ware; for fabric description
see Fulford (1975, 26).

F100. Soft, fine-grained; moderate quartz <0.5mm, rare
angular crushed flint <lmm, rare iron oxides <0.5mm.
Oxidised.

G100. Hard, soapy; moderate poorly-sorted grog <5mm,
rare quartz, mica and iron oxides, all <0.5mm.
Predominantly oxidised. Handmade. Surfaces smoothed or
burnished but with grog frequently protruding, especially
on interior.

G101. Hard; sparse grog up to 5mm, crushed, angular
flint <2mm, rare iron oxides. Oxidised. Handmade.
Surfaces smoothed but larger inclusions protrude.

Q100. Hard; moderate to common quartz <0.5mm, rare
iron oxides <0.5mm. Unoxidised. Wheel- and handmade
examples. A ‘catch-all’ group for all sandy greywares
without probable glauconite.

Q101. Hard; common quartz, sparse to moderate
probable glauconite, rare iron oxides, all <0.5mm.
Unoxidised. Wheel- and handmade examples. Probable
glauconite very visible, often giving speckled appearance. A
‘catch-all’ group for all sandy greywares with probable
glauconite.

Q102. Hard; moderate to common quartz <0.5mm, rare
iron oxides <0.5mm. Oxidised. Wheel- and handmade
examples. A ‘catch-all’ group for all oxidised sandy wares

without probable glauconite; includes intentionally
oxidised fabrics as well as oxidised examples of fabrics that
are more usually unoxidised.

Q103. Hard; common quartz, sparse to moderate
probable glauconite, rare iron oxides, all <0.5mm.
Oxidised. Wheel- and handmade examples. Glauconite(?)
very visible, often giving speckled appearance. A ‘catch-all’
group for all oxidised sandy greywares with probable
glauconite.

Q106. Hard, fine-grained; common quartz <0.25mm, rare
iron oxides <lmm; occasional elongated voids <2mm.
Unoxidised. | Manufacturing technology uncertain.
Characteristically a very dark greyish-brown in colour.
Q107. Hard, fine-grained; moderate quartz <0.25mm,
sparse but very visible red iron oxides <0.5mm, rare mica.
Generally unoxidised but exterior margin of some
examples oxidised. Manufacturing technology uncertain.
Q108. Hard, coarse; common quartz <0.25mm, sparse
angular flint up to 10mm, rare soft, white calcareous
particles, red iron oxides, both <lmm.. Oxidised.
Handmade. Used exclusively for clay discs.

Q110. Hard, very fine-grained; common quartz, rare black
iron oxides, both <0.25mm. Unoxidised. Wheelmade.
Characterised by a very white core with dark grey surfaces.
Q111. Hard; common quartz, rare iron oxides, both
<0.5mm. Unoxidised although some examples have
partially oxidised core. Manufacturing technology
uncertain. Surfaces smoothed or burnished.

The distribution of Romano-British pottery by
feature is given in Table 1. The sandy wares, in
particular the ‘catch-all’ fabric groups (Q100—103),

58 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

are clearly derived from a number of different
sources and probably span a wide date range. Kilns
to the west of Swindon are known to have been
producing sandy greywares from the early 2nd
century AD until the end of the 4th century
(Anderson 1979) while the presence of probable
glauconite in fabrics Q101 and Q103 suggests a
source in the region of the Upper Greensand areas of
west and north Wiltshire. Greyware wasters and kiln
furniture have been found at Westbury, on Upper
Greensand in the west of the county (Rogers and
Roddham 1991, 51). The presence of New Forest
greywares amongst this assemblage is also indicated
by the bow] with the inturned rim (Fulford 1975, type
7), while everted rim jars and narrow-necked jars with
rolled rims are also known amongst the repertoire of
the New Forest greyware potters (zbid., types 30.5 and
31.2, respectively) as well as at a variety of other
centres. The vessel forms have a recognised date from
the mid 2nd century onwards, although the sherds of
fabrics G100 and Q107, which are also found at
Figheldean (Mepham 1993, fabrics G100 and Q113)
and Butterfield Down, Amesbury (Millard 1996,
fabric G100), indicate the presence of earlier
Romano-British material amongst the assemblage.
Fabrics F100, Q106 and Q110 are represented by
body and/or base sherds only.

‘The three Black Burnished ware forms recognised
are the characteristic and most widely distributed
products of this industry during the later 3rd—4th
centuries, although some of the more fragmentary
everted rim jars may be of later 2nd-century date.
Other later Romano-British material amongst the
assemblage includes the small everted rim jar in New
Forest parchment ware, which has a date range of
c.AD 300-380 (Fulford 1975, 74) and the greyware
vessels, probably also from this source and
contemporaneous with it. The clay discs, fragments
of which were found in storage pit 1195 and feature
1268, were made exclusively in the coarse, sandy
oxidised fabric with large flint inclusions (Fabric
Q108). The function of these objects is uncertain.
Various uses, including storage jar lids, cheese-press
lids and components of ovens or other heating
structures have been suggested. Examples of similar
discs occurring in local fabrics are known from
various sites in south Oxfordshire (Miles 1978, fig.
57, 32; Sanders 1979, fig. 28, 124-127; Wessex
Archaeology 1993) but no published examples from
sites closer to Shrewton have been identified during
the preparation of this report.

The paucity of diagnostic sherds recovered
increases the difficulty of dating the Romano-British

assemblage with any precision. The majority of
excavated features contained no pottery or insufficient
diagnostic sherds to be assigned anything more than a
general Romano-British date. No early Romano-
British (1st-2nd century AD) groups were identified
but quarry pit 1191 contained material of 2nd—3rd-
century date, while storage pit 1195 contained sherds
spanning the mid 2nd—4th centuries. A larger number
of features containing later Romano-British (3rd—4th
century) pottery were recognised. These included
quarry pits 1022, 1059 and 1061, hearths 1086 and
1090 and ‘miscellaneous’ features 1029, 1065, 1067,
1227, 1268 and 1276.

The assemblage contains the usual range of
fabrics and forms typical of a southern English small
Romano-British farming community of compara-
tively low status, and is broadly comparable with
those from other sites in the vicinity (Swan 1971;
Mepham 1993; Millard 1996).

Other artefacts
by RACHAEL SEAGER SMITH

Worked stone

Fragments from eight worked stone objects were
recovered, representing seven rotary querns and a
probable roof tile.

Three of the querns are of Greensand, the nearest
outcrops of which occur within 15km of the site on
the Greensand ridge of north Wiltshire, although at
present the best known production centre of
Greensand querns is at Lodsworth, West Sussex
(Peacock 1987). The other quern fragments are of
various sandstones. With the exception of one
sandstone example of a lower stone from feature
1084, all the querns were too fragmentary to deter-
mine which stone was represented. The fragment of
the lower stone has iron staining visible inside the
central pivot hole. The diameters of only two of the
fragments could be measured. One of these, of
Greensand, found in posthole 1103, measured
approximately 0.80m in diameter, while the other, a
fragment from a very well-worn sandstone quern
found in pit 1195, measured approximately 0.40m.

The probable roof tile is represented by three thin,
flat fragments, two of which join, of fossiliferous
limestone. These were found in hearth 1038 and the
patchy, slight reddening of the surface of all three
fragments may indicate their exposure to heat.

Three of the quern fragments (two sandstone,
one Greensand) were found in feature 1084 and a
further fragment came from posthole 1103; the roof
tile fragment was recovered from feature 1038: all
comparatively close to structure 1118.

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 59

Non-local stone

Three fragments of unworked, non-local stone were
found (i.e. not naturally occurring within c. 15—20km
of the site). One small chip of fine-grained limestone
was recovered from quarry pit 1061 while a small
piece of sandstone and a slightly fossiliferous
limestone fragment were found in pit 1195.

Fired clay

A total of 58 fragments (216g) of fired clay was
recovered. Although all were small, shapeless and
featureless, they were recovered from the fillings of
hearths (40 pieces from feature 1126, eight from
feature 1090 and ten from feature 1038): thus it is
possible that they are derived from hearth linings,
oven covers or similar structures, although none of the
fragments are vitrified or show signs of over-exposure
to extreme heat. All the fragments from features 1126
and 1090 are soft, very chalky and off-white in colour,
while those from feature 1038 are harder, deep
reddish-brown in colour and contain a_ larger
proportion of clay, although still with chalk inclusions.

Worked bone

A complete pin (object no. 76) with an elaborately
decorated head (Figure 5) and an incomplete bone
point (object no. 77) were found in the primary
fillings of pit 1195. No direct parallels have been
found for the pin although it is encompassed by
Crummy’s type 6 pins with bead- and reel-shaped
heads which are broadly dated to the 3rd and 4th
centuries AD (Crummy 1983, 24). The pin has an
elaborately decorated head comprising a cylindrical
bead decorated with an incised zig-zag and reel
surmounted by a small conical bead. It is carefully
made and finished, with uniform surface polish on
the plain, tapering, circular cross-sectioned shaft
which is not quite straight (length 130mm). The top
of the pin head also has very high polish. The bone
point is damaged at both ends (surviving length
109mm, width 5-11mm, thickness 4-8mm). It was

| made from a curving bone fragment and has a
| tapering, square cross-section. The surface is well-

polished for c.40mm at the narrowest end, but much

less well finished over the remainder. The function of
| this object is uncertain although it is likely to be of
Romano-British date.

| Ceramic building material

} One piece of ceramic building material was found on
the surface of the unexcavated quarry pits/lynchet
| (1269).
| Romano-British date but it is too fragmentary to

The fragment is almost certainly of

Figure 5. Maddington Farm, Shrewton: bone pin;
actual size

identify the brick or tile type from which it was
derived (cf. Brodribb 1987, 3).

Palaeo-environmental material

Human bone
by JACQUELINE I. McKINLEY

Inhumed human bone was recovered from eight
contexts, comprising five discrete inhumation
burials (124, 1005, 1012, 1037 and 1178), one
probable burial (121), one disturbed/redeposited
burial (1145), and bone redeposited in a quarry pit

60 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

(1020). Small quantities of cremated human bone
were recovered from eight other contexts (Table 4).
Full details are in the archive.

Methods

Cremated bone was analysed following the writer’s
standard procedure (McKinley 1989).

Age was assessed from the stage of tooth
development and eruption (van Beek 1983); the
stage of ossification and epiphyseal fusion (Gray
1977; McMinn and Hutchings 1985); the length of
immature long bones (Bass 1987); the pattern of
degenerative changes in the pubic symphyses and
fourth ribs (Brooks 1955; Iscan er al. 1984 and
1985); tooth wear patterns (Brothwell 1972); and
the general degree of cranial suture fusion and
degenerative changes to the bone.

Age categories:
infant 0-4 years (young 0-2 years;
older 3—4 years)

juvenile 5-12 years (young 5-8 years;
older 9-12 years)
subadult 13-18 years (young 13-15 years;

older 16-18 years)
young adult 19-25 years
mature adult 26—45 years (younger 26—30 years;
older 31—45 years)
older adult 45+ years
Sex of adults was assessed from the sexually
dimorphic traits of the skeleton (Bass 1987). Cranial,
Platymeric and Platycnemic indices were calculated
according to Brothwell (1972) and Bass (1987).
Stature was estimated using Trotter and Gleser’s
regression equations (1952; 1957). Pathological lesions
and morphological variations/non-metric traits were
recorded, and diagnoses suggested where appropriate.
Anatomical terminology used is according to Gray
(1977) and McMinn and Hutchings (1985).

Results

Condition of bone. Bone from the two deep graves, 124
and 1012, was in good condition, though some
fragmentation had occurred as a result of the heavy
weight of chalk rubble grave fill, while some breakage
of 124 had resulted from the ‘salvage’ nature of
recovery. Bone from the other graves was generally in
rather poor condition, especially the spongy bone of
the vertebrae and innominates, and was root-marked.
The bone from these shallow graves was often
fragmentary, partly as a result of heavy plant crossing
the site immediately prior to excavation, but some

fragmentation was of less recent origin and had
probably resulted from agricultural disturbance.
Demography. A minimum of seven individuals was
identified from the inhumed bone: one young
juvenile, one older juvenile, three adult females
(mature, older mature/older and older), and two
adult males (young and older mature). The general
physique of all the individuals tended towards the
gracile. The older mature adult 1005 could be
identified only as possibly male due to contradictory
morphological traits: the general size of the skeleton
and skull morphology indicating a female, whilst the
size of the articular surfaces and innominate
morphology suggested a male.

Unfortunately, the assemblage represents probably
only a small part of the cemetery and further
demographic comment is therefore precluded.
Skeletal indices. Stature estimates were made from
four skeletons: two females, 124 (160.5cm) and
1012 (154.4cm); two males, 1005 (169.0cm) and
1179 (166.5cm).

Cranial indices could be calculated for only two
individuals: both were mesocrany. Platymeric
(anterior—posterior flattening of proximal femur
shaft) and platycnemic (medio-—lateral flattening of
the tibia shaft) indices were calculated for four
individuals: three platymeric (1005, 1012, 1179)
and one eurymeric (124); two platycnemic (1005,
1179) and two eurycnemic (124, 1012). The
numbers involved are too low to apportion any
significant interpretation.

Pathology

A summary list of lesions/pathological conditions
and morphological variations according to bone
groups affected is presented in Table 3.

Dental disease. Full or partial dentitions were
available for six of the seven individuals, and four of
the five adults had some dental disease. One
hundred and thirty-eight teeth were recovered and
125 sockets in five left and right mandibles, and six
left and five right maxilla, were available for
examination. A total of 158 tooth positions were
recorded (including crown crypts).

Of the teeth, 16/158 (10%) were lost ante mortem
from four of the five adults: 9/72 (12.5%) maxillary,
7/68 (10.3%) mandibular, 11/38 (28.9%) female
and 5/57 (8.8%) male. The much greater percentage
of female tooth loss is, at least in part, related to the
occurrence of more females in the older age
categories. Although significant comparison of such
small numbers with other contemporaneous sites is
not easy, the overall percentage of tooth loss is close

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 61

Table 3. Human bone: summary of inhumed bone, showing percentage of skeleton recovered, age and sex of the
individual(s) and pathological lesions/morphological variations

context|% skeletal | age sex pathology
recovery

121 c.8% older juvenile periostitis: r.femur a
124 c.0% older adult female |o.a.: r.p.humerus, r.p.radius, r.costo-clavicular, costo-

vertebral, bi-lateral acetabulae, cervical, lumbar, sacro-
iliac; d.d.d.: cervical; trauma: fracture |.d.fibula, ?atlas
transverse ligament; hyperostosis: 4th & 5th lumbar,
manubrium; ?infection: 5th lumbar & 1st sacral surfaces;
o.p.: r.scapula, r.d.radius, r.d.ulna, |.p.ulna, |.trapezium,
l.1st metacarpal, |.finger phalanges, atlas-axis, thoracic &
lumbar bodies, r.patella; exostoses: 1.1st metacarpal,
l.finger phalanges, iliac crests, lp. & d.tibiae, 1.fibula,
r.d.fibula, patellae, calcanea; pitting: lateral clavicles; d.l1.:
r.& |.1lst p.foot phalanges, r.lst metatarsal; spina bifica
occulta; m.v.: occipital bunning, calcaneal double facets,
small 11th & 12th thoracic rib facets, fusion r.foot 5th
middle & distal phalanges

1005 c.75% older mature adult |??male | calculus; p.d.; hypoplasia; caries; abscesses; 0.a.:
thoracic; trauma: ?fractured l.rib; o.p.: atlas-axis,
l.scapula, middle finger phalanges, r.d.humerus & p.ulna,
r.d.tibia; pitting: |.d.tibia; exostoses: |.calcaneum, |.tibia
& fibula, r.p.femur, iliac crests, r.patella; d.1.: r.p. femur;
hyperostosis: r.femur neck; m.v.: atlas double facet

1012 older/mature older |female |calculus; p.d.; hypoplasia; caries; abscesses;
adult ?sinusitis; calcified tissue; Schmorl’s: lumbar,
thoracic; 0.a.: lumbar; o.p.: thoracic, atlas, axis, sacrum;
m.v.: atlas double facet, accessory neural foramina 6th—7th
cervical, pseudo-facet sacrum & r.ilium

| 1020 1) young
juvenile=1145

2) adult=?1037

1037 mature adult female | calculus; p.d.; ?abscess; periostitis: r.clavicle, m.v.:
Pabsent mandibular M3
1045 1) 4—6yr. calculus; hypoplasia
2) adult
=?1137/1179
1179 c.75% young adult male calculus; p.d.; cribra orbitalia; periostitis: r.ventral

ilium, femoral necks/proximal shafts, r.clavicle; bone
rarefication/vascularity: acetabulae, 1|.femoral neck;
d.l.: l.acetabulum, |.femur head, r.lst p.foot phalanx;
exostoses: calcanea, |.d.finger phalanx; o.p.: atlas-axis;
m.v.: maxillary 12 ‘pegged’, retention ].deciduous canine,
double facets calcanea, r.squatting facet

| Key:
| r., right; 1., left; p., proximal; d., distal; o.p., osteophytes; 0.a., osteoarthritis; p.d., periodontal disease; d.d.d., degenerative
| disc disease; d.l., destructive lesion; m.v., morphological variation; M, molar; I, incisor.

62 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

to that from Baldock Area 15 (12.4%; McKinley
forthcoming) and Cirencester (8.5%; Wells 1982).

‘Tooth loss tends to increase with age and may be
related to one or more factors. Diet and dental
hygiene may influence other dental diseases which
predispose to tooth loss. Tooth loss from excess wear
may be precipitated by periodontal disease (gum
infection), which all the adults showed to some
degree.

Dental caries were noted in dentitions 1005 and
1012. Of the whole assemblage, 12/138 (10%) teeth
had carious lesions: 6/53 (11.3%) female and 8/50
(16% male); 8/72 (11.1%) were maxillary and 6/66
(9.1%) mandibular. The majority of lesions were
in the molars. In half of the affected teeth the
crown had been fully destroyed. Where it was
possible to ascertain the origin of the lesion, all were
cervical with the exception of one small occlusal
lesion. All dentitions showed some degree of
calculus (calcified plaque), which was heavy in 1005,
covering the occlusal surface of the third molars.
Dental abscesses were present in three dentitions:
1005, 1012 and 1037. Of the whole assemblage,
7/126 (5.5%) of sockets had abscess lesions, of
which 4/46 (8.7%) were female and 3/51 (5.9%)
were male (3/53 (5.7%) were maxillary and 4/73
(5.5%) mandibular). The general condition of the
dentitions suggests a relatively poor level of dental
hygierie.

Trauma. There is a well-healed spiral fracture in the
left distal fibula of 124, with associated exostoses at
the distal interosseous ligament attachments of the
fibula and left tibia. The exertion of a violent lateral
force 1s indicated (Adams 1987), resulting in rupture
of the ligaments and fracture of the bone. No lesions
were noted in the talus. One left rib from 1005 has a
well-healed fracture, which may have resulted from a
fall or a blow to the chest.

Infections. Evidence for possible sinusitis was
noted in 1012, where areas of irregular new bone
were seen on the wall of the left antrum. Non-
specific periostitis was observed in three individuals.
Infection of the periosteum (the membrane covering
the bone) may result from direct introduction of
bacteria via a wound or fracture, or spread from foci
elsewhere in the body through the blood stream. The
juvenile skeleton 121 has areas of periosteal new
bone on the distal right femur shaft, but incomplete
skeletal recovery limits diagnosis. Inhumation 1037
has lesions on the superior side of the mesial right
clavicle shaft; no other associated lesions were noted.
The surface proximity of the bone suggests a soft
tissue wound.

Extensive periostitis was observed in 1179, a
young adult male, including areas of the right ventral
ilium, both femoral necks and shafts, and the medio-
dorsal right clavicle shaft. Destructive lesions were
also noted in the left acetabulum and left femur
head, with bone rarefication/vascularity (the right
side bones have not survived). The overall form of
the joints remained intact. In this case, the focus of
infection may have been in the hip joint(s) and
spread to the ilium and proximal femoral shafts. The
adjacent body surfaces of the 5th lumbar and Ist
sacral vertebrae from 124 showed erosive pitting
with slight surface new bone suggestive of infection.
Degenerative disease. Osteoarthritic lesions (osteo-
phytes, eburnation and pitting) were noted in the
joints of three individuals. Eight joints were affected
in 124, thoracic and lumbar joints were affected in
1005 and 1012, respectively: basically the result of
age-related wear-and-tear, whilst other predisposing
factors include previous disease, injury and obesity
(Adams 1987).

Osteophytes (irregular bone forming on the
margins of joint surfaces) and exostoses (irregular
bone forming at tendon and ligament insertions)
may both reflect age-related wear-and-tear, though
they may also be associated with specific diseases or
trauma (see above). Extensive lesions were noted in
124, with osteophytes at eleven joints/joint groups,
and exostoses at nine (including trauma related,
discussed above).

The florid hyperostosis observed on the ventral

surfaces of two lumbar vertebrae in 124, may
indicate the early stages of diffuse idiopathic skeletal
hyperostosis (DISH). Extra-spinal manifestations of
the disease may be represented by the extensive
exostoses noted (Rogers et al. 1987), and hyper-
ostosis in the manubrium.
Deficiency disease. Mild cribra orbitalia (pitting in the
orbital vault), believed to result from a metabolic
disorder associated with childhood anaemia, was
noted in 1179. Slight to mild dental hypoplasia
(developmental defects in the tooth enamel: Hillson
1986) was observed in three dentitions. Spina bifida
occulta, the lesser, and pathologically insignificant,
form of spina bifida (Adams 1985) was present in
the inferior sacrum of 124.

Cremated bone

A summary of the cremated bone is given in Table 4.
The bone recovered may be divided into two groups:
bone, probably already scattered, accidentally rede-
posited within graves and other contexts; and bone
deposited in pyre debris pits 1224, 1240 and 1244.

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 63

Table 4. Summary of cremated human bone

Context total weight fragments

1145 <1.0g
1178 <0.1g vault
1201 1.6g long bone & articular
surface
1217 0.2g long bone (13mm)
1225 6.5g vault (25mm); finger
phalanges;
femur shaft (12mm)
1240 2.2¢g long bone & articular
surface
1245 9.0g vault (18mm); humerus
(12mm)
& distal finger phalanx
<41> O.lg

The quantity of bone in each case is very small, and
it is not believed that any of it represents the remains
of a disturbed cremation burial. There is no evidence
of disturbed graves, nor of any associated artefacts.
The presence of cremated bone in the three
charcoal-rich pits does, however, suggest that
cremations took place within the vicinity. The pits,
with their charcoal, burnt flint and cremated bone,
probably represent dumps of pyre debris such as
have been found at the contemporaneous sites of
Baldock (Burleigh and Stevenson pers. comm.) and
Hooper Street in London (McKinley in prep.).

The bone fragment size is small, as may be
expected, but it was possible to identify some
fragments which were of older subadult/adult size.
The small size of the fragments may be one reason
why they were not collected for burial, perhaps
having been overlooked. Much of the bone appears
incompletely oxidised (blue/black colour).

The dense charcoal deposits within the three pits
imply the remains of more than a single cremation
pyre. In a recent experiment (Marshall and
McKinley in prep.), the writer found that from
900kg of wood used to build a pyre, 3.82kg of
charcoal fragment size >2mm (c. 14 litres) remained.
Over time this would break down to smaller sized
fragments. The implication is that an unknown
number of cremation burials is located within the
vicinity of the site.

| Animal bone

by S. HAMILTON-DYER

Introduction

| The condition of bones varied from excellent, with

fine surface details preserved, to chalky and eroded.
The assemblage is summarised in Table 5 in which

minimum numbers of bones or, where possible,
individual animals rather than total fragment
numbers, are given for reconstructable bone
fragments, the animal burials and the puppy bones
from pit 1195.

Results

Grave 1002. Underneath the prone inhumation
1005, included as part of the burial, was the
complete skeleton of a dog (1019). The body had
been laid on its left side with the head partly under
the human skull. The pelvis lay under the lower
thorax with the hind legs towards the right elbow of
the inhumation. The entire skeleton was recovered
including one of the internal ear bones. Some of the
smallest elements came from sieved samples,
including 23 small fragments of bone, the size and
appearance of which suggest that they were gut
contents. All the bones, including vertebrae, have
fused epiphyses but skull sutures are still visible
implying that the animal was old but not extremely
aged. This individual was male as evidenced by an os
penis.

There are several minor pathological lesions.
Three of the smaller teeth had been lost ante mortem
with the alveolus infilled or in the process of infilling.
The 4th and 5th lumbar vertebrae are almost inter-
locking with the presence of lipping osteophytes.
One of the metacarpal bones shows evidence of
healed fracture with the foreshortening of the bone.
The skull had a small partly healed hole near the
right-hand nasal/maxilla suture, and there is a small
crack in the left frontal with slight porosity of the
surrounding bone.

Measurements of the bones give an estimated
withers height of around 46cm based on the factors
of Harcourt (1974). This ‘medium’ size (around the
size of a modern border collie) is common in Iron
Age and Romano-British material.

Burial pit 1043. This ovate pit contained the skeleton
of a calf (1058). The skeleton is essentially complete,
including all toes and epiphyses. Two small bones
are absent, one fibula and one patella. The skull,
although fragmented, had not been chopped or pole-
axed, and shows small horn cores developing. None
of the bones were fused, apart from the distal
epiphysis of the humerus which had begun to fuse to
the shaft at the time of death. In the jaws, the
deciduous 4th premolar and Ist molar were in wear
with the developing 2nd molar just visible in the
jawbone. The combined tooth and _ epiphyseal
information indicates that the animal was probably
around a year old. The immaturity of the bones

64 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Table 5. Animal bone: species distribution (non-burial contexts)

Feature context horse cattle sheep pig LAR SAR

unid dog small fowl unid eel amph Total

/goat mammal mammal bird

‘Topsoil 1000 2 - - 1 3
Q/L 1269) - - 1 1 1 3
pit 1022 1020.) - - 4 ~ - 1 - - — 5
F 1041 1040) — 1 - - ~ 1
pit 1043 1044. - 1 ~ 1 1 5 - 8
pit 1059 1109 —- 1 ~ - _ 1 2
F 1060 1061 3 - 3 —- 23 14 - 6 - 49
F 1067 1070 3 3
hearth 10841085 — ~ 2 2
F 1086 1087) — 1 — - 2 3
hearth 1090 1091 = = 1 1
ph 1122 1123 1 1
F 1124 1125 1 ~ 1
F 1146 1145 1 1
F 1183 1184 —- - 2 - 1 - - - 3
F 1185 1186 — 1 - - ~ 1
F 1191 1162 - 1 = - - ~ - 1
pit 1206 1207. - 1 2 ~ = 1 4
pit 1221 1222 2 2
F 1226 1223 - 2 ~ - 3 1 6
pit 1227 1228 —- - 1 1
F 1243 1242 - 2 - ~ 1 - - — = - ~ - 3
F 1265 1257) - 1 4 - 3 - - - — - - ~ ~ 8
F 1266 1253 - ~ _ - = 1 - - - - ~ - 1
F 1265 1251 - - 3 — 3
pit 1195 1196 1 2 12 - 2 ‘A 18
pit 1195 1199 — - 5 - - — 59 3(72) - 3 21 2 2 101(170)
pit 1195 1201) - _ 4 1 - 3 14 1 23
pit 1195 1201 1 1
pit 1195 1219 1 — 1 = - = - 2
total 1195 No. 1 2 21 1 2 5 74 3 1 3 27 2 3 145(214)

% 0.7 1.4 14.5 0.7 1.4 3.4 51 2.1 0.7 2.1 18.6 1.4 2.1
total excl No. 3 12 20 be By ee21 15 ih 0 0 0 0 0 116
1195

% 2.6 10.3 17.2 0.9 31.9 18.1 12.9 6 0 0) 0 0 0
Grand total No. 4 14 41 2 29 26 89 10 1 3 27 2 3 261(330)

% 15 5.4 15.7 0.8 14.9 10

Key:

34.1 3.8 0.4 LiL, 10.3°.°-0.8).. lel

LAR=large ungulate (probably mostly cow but may also include horse
SAR=small artiodactyl (probably mostly domestic sheep/goat)

precluded measurement. Examination of the bones
did not reveal any butchery or pathology.

Burial pit 1143. This pit contained the articulated
skeletons of five piglets. The individual animals were
allocated separate context numbers: 1138, 1139,
1140, 1141 and 1142. There was some mixing

between these contexts and some bone was also
recovered from the pit fill, 1144. Photographs and
drawings reveal that the piglets had been placed in
the pit entire but slightly overlapping. The skull of
1139 was recovered almost intact and contained the
remains of small cesspit/compost fly pupae in the

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 65

nasal passages. The piglets were all of similar age and
size, but not identical. All the bones were unfused. In
the jaws, the 1st molar varied from just erupted to
still hidden in the jaw crypt. The deciduous 4th
premolar was just in wear in all cases. The piglets
were certainly not neonatal but probably died or
were killed at less than four months old (Bull and
Payne 1982). The variation in size and ageing may
indicate that they were from different litters,
although litter mates can vary considerably. As with
the calf there is no evidence of butchery, dismem-
berment or cause of death.

Pit 1154. The 76 broken fragments recovered are in
poor condition. They comprise 16 cattle bones from
a right hind leg, from the tibia down, and more than
one set of toes; all epiphyses are fused. A much
fragmented cattle rib is also present. The remaining
23 fragments could not be reconstructed or
identified. The pit had been truncated by machine
stripping and it is unclear whether this bone group
represents a complete or partial burial, or disposal of
unrelated material. Measurement of the broken but
complete metatarsus enabled an estimate of withers
height to be made of 112.3cm (von den Driesch and
Boessneck 1974).

Pyre debris pits. The 44 animal bone fragments from
feature 1224 are a mixture of horse, cattle and sheep.
Although eroded, none appears burnt. There are
remains of two jaws each of horse, cattle and sheep,
from different individuals. Other fragments include:
horse pelvis, cattle ulna, pelvis, ribs and foot bones,
and sheep femur and upper teeth. Two of the cattle
bones had been chopped. Pit 1244 contained one
small burnt fragment of unidentified mammalian
bone.

Pit 1195. The 214 fragments from this pit constitute
the only other large group of bones. Three fragments
were recovered from the bottom fills, layer 1219 and
1210; these include a mouse or vole humerus. Layer
1201 contained a pig 3rd phalanx which may have
been through the digestive system. Bone from layer
1199 includes 72 bones of three neonatal puppies.
This layer also contained many small fragments,
including bird, fish and amphibian (including
common frog, Rana temporaria and common toad,
Bufo bufo) bones which were not recovered elsewhere
on the site. In addition to fowl leg bones, the bird
bones include toes, tracheal rings and ossified
tendons. The two fish bones are of small eels
(Anguilla anguilla). Some of the many small
fragments of mammal bone have the appearance
usually associated with digestion. The only bones

| from the main domestic mammals were a sheep toe

and four sesamoids. Many of the smallest elements
in this layer would probably have been missed
without sieving. The material from the upper fill is
quite different, being a mixture cf sheep bones
including head, foot and limb bones, together with a
horse axis and two cattle humerus fragments.

Other contexts. The remaining 25. contexts
contributed just over 100 fragments in total. Forty-
nine of these are from feature 1060 and include six
bones of a dog, three horse teeth, three sheep/goat
bones and unidentified sheep/pig and cattle/horse-
sized fragments.

Discussion

Apart from the numerous bones in the animal
burials, this group of material is a small assemblage
with only a few fragments from each context. Pit
1195 had the largest group of material, some of
which is probably associated with human cess, the
presence of which may also have aided bone
preservation. Overall, sheep is the most frequent
animal identified, with cattle second. Horse is
present in two contexts. Pig bone, other than the
piglet burials, is confined to a maxilla in the quarry
pit 1269 and the toe in pit 1195. If repeated on a
larger scale this high level of sheep and low numbers
of pig bones would resemble the late Romano-
British assemblages at Owlesbury and Winnall
Down, both in Hampshire (Maltby 1985); however,
both these sites had higher numbers of horse bones.
The fragment numbers in this assemblage are so
small that additional material from the unexcavated
deposits could alter these proportions considerably.
The age of the animals probably indicates that they
were bred at or near the site.

Charred plant macrofossils
by P. HINTON

Charred plant remains were extracted from seven
samples recovered from five of the hearths (1034,
1038, 1049, 1090 and 1084), the animal burial 1143
and pit 1195. The assemblage is summarised in
Table 6.

Cereals

The condition of many of the seeds is poor and the
cereals are mainly very fragmentary grains (Table 6).
The overall morphology of some cereals may be
compared with Triticum dicoccum (emmer wheat)
and T. spelta (spelt wheat), and the presence of these
two wheat species is confirmed by the more readily
identifiable chaff fragments in some _ samples.
Triticum aestivum s.l. (bread wheat) has not been

66 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

‘Table 6. Charred plant macro-fossils

HEARTHS PITS

FEATURE NO. 1034 1038 1049 1090 1084 1143 1195
Sample volume (litres) 6 10 3) 10 5 10 25
Triticum cf dicoccum Schubl. (spelt wheat) grains 1 10 1 1 10

glume bases 1 3
Triticum dicoccum/spelta (emmer or spelt) grains 3 4

glume bases 1 3 1 2

spikelet bases 1 6 1
Triticum sp. (undiff. wheat) 7 16 fr if 11 2 3
Hordeum vulgare L. emend Lam. (hulled barley) 3 10 3 4
cf Hordeum vulgare L. emend Lam. 1 3 1 1

cf Avena sp. (oats) 1
Cerealia indet. — fragmentary grains 35 6.100 ¢.3 c.4 6.50) 6:2 c.3
(indeterminate cereals)

Papaver sp. (poppy) 3
Urtica dioica L. (common nettle) 1
Chenopodium/Atriplex sp. (goosefoot or orache) 1 1
Polygonum aviculare s.1. 1 1
Rumex cf crispus L. (curled dock) 1
Rumex sp. (dock) 1
Viola sp. (violet or pansy) 1
Aphanes arvensis L. (parsley piert) 1
Vicia cf tetrasperma (L.) Schreber (smooth tare) 1 1
Vicia/Lathyrus sp. (vetch or vetchling) 1 3 1 1
Trifolium cf pratense L. (red clover) 2
Lithospermum arvense L. (corn gromwell) 21 frs
Plantago lanceolata L. (ribwort plantain) 1 1
Odontites vernus (Bellardi) Dumort (red bartsia) 1
Galium aparine L. (cleavers) 1 1 2 2
cf Valerianella dentata (L.) Pollich 1
(marrow-fruited corn salad)
Compositae indet. (small-seeded Senecio type) 2
Luzula cf campestris (L.) DC (field wood-rush) 1
Carex nigra/ovalis (common or oval sedge) 1
Lolium perenne L. (perennial rye-grass) 2 12
cf Phleum pratense L. (timothy grass) 2
Graminae indet. (small-seeded grasses) 2
Key: fr = fragment;
frs= <5 fragments
All records are for seeds/fruits unless otherwise indicated
recognised but even in better condition these grains form and trace of radicle depression. None of the
are not always distinguishable from spelt and the outer surface remains.
presence of this free-threshing wheat cannot be ruled The numerous indeterminate cereal fragments
out. have been estimated, by volume, to an approximate
The barley is slightly better preserved and two equivalent number of grains. Larger fragments were
samples (13) and (14) include grains which appear selected and smaller ones estimated from sub-
to be naturally asymmetrical and therefore indicate samples of the larger flots, e.g. sample 14. The badly
the presence of six-row barley. The identification of degraded condition of these grains, and the main
the one Avena (oat) grain is somewhat dubious and fragments, suggests that a considerable amount of

is based solely on its size, more or less cylindrical chaff and other more fragile items may have been

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM

lost, either at the time of charring or after deposition.
It would be unwise therefore to imply a specific stage
of crop processing from the proportions of chaff and
weed seeds: it is only possible to suggest that these
were not deposits of cleaned, fully prepared grain.

Other species
All the other plants represented are open ground
species which could all occur in cultivated soil, but
the grasses, vetches, clovers, wood-rush and sedge
are perhaps more likely to have a grassland origin.
Perennial rye-grass (Lolium perenne) and_ the
probable timothy (Phleum pratense), are character-
istic pasture grasses. The one sedge seed has not
been securely identified but is likely to be oval sedge
(Carex ovalis), which occurs in damp or dry
grassland, more frequently the former. Common
~ sedge (Carex nigra) which is similar in form (but
_ distinguishable when better preserved) prefers boggy
conditions and this context seems unlikely here.
Summary
| The plant remains indicate that cultivated cereals
included the glumed wheats, emmer and spelt, and
barley. The possible oat is more likely to have been a
_ weed. All these could well have been grown in the
vicinity, and the chaff and weed seeds suggest that
threshing and cleaning operations were carried out
| nearby. If the grassy species were not associated with
the crops then hay might be suggested as an
_ alternative source.

Land Mollusca
by MICHAEL J. ALLEN

Methods

A column of eight samples was taken from the
lynchet/colluvium revealed in sondage 1288 (Figure
3) for Mollusca, and processed by standard methods
(Evans 1972). A single spot sample from a possible
turf line (context 1111) was also taken and assessed
for Mollusca. The flots (retained on a 0.5mm mesh
aperture) were scanned under a stereo-binocular
microscope and the semi-quantitive record of each
species identified is presented in Table 7.

Results
Overall shell preservation was good but numbers
were relatively low. In the old land surface (context
1111) this may be due to the weakly calcareous
nature of the turfline. In the lynchet, however, the
low whole shell numbers in the flots are typical of
colluvial deposits where shells are broken during the
deposition of the hillwash, indicated by the large
numbers of broken fragments observed in the
residue fractions (2mm, 1mm and 0.5mm). Shell
numbers are therefore high enough for valid and
statistically viable palaeo-environmental analysis.
Overall, the assemblages almost exclusively
comprised open country species. The only shade-
loving specimen was from the putative turfline and
the only catholic species recorded was Trichia hispida
which is common in hillwash deposits. The Mollusca
from the buried turf line were open country

Table 7. Land snails: semi-quantitive record of species identified

Lynchet

Catholic species

| Context 1253 1281 | 1251 | 1240
44 45 46
| Open country species
| | Pupilla muscorum C (icp ieee a ORGS ER | B A A

Vertigo spp. Cll ia ae

Helicella itala PCr eb a eee

Vallonia spp. B B C eer

Intro. Helicellids Ca ae

Trichia hispida | ome eer
\ Shade-loving species
Discus rotundatus C | | | E
|| Burrowing species
| C B A ASA } A + AER
‘| Comment on totals 10 30 25 15 20 15 30 40 60

iKey: A= >10; B = 5-9; C= <5 items

68 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

specimens typical of very open, dry, possibly bare
soil. A single specimen of Discus rotundatus, a shade-
loving species, was recovered. The deposit was
dominated by the Vallonias (of which both V’ costata
and V excentrica were noted) and Pupilla muscorum.
‘The assemblages were typical of lynchet deposits but,
significantly, species composition changed through
the contiguously sampled column indicating that,
had sufficient dating evidence been recovered,
analysis to detect subtly changing habitats associated
with the use of the local landscape during the
formation of the natural north-south feature might
have been possible. However, as the dating evidence
is no more precise than broadly Romano-British, and
the quantities of residual prehistoric material and the
obvious reworking of the deposits during more than
one phase of activity suggest that the deposits are of
limited environmental value, more detailed analysis
of the assemblage has not been undertaken.

Discussion

The discovery of the previously unknown Romano-
British site at Maddington Farm has added to the
growing corpus of information on human activity
within the area of Salisbury Plain during this period.
It has also served to demonstrate the value of
maintaining archaeological surveillance of linear
construction projects such as pipelines, even when
they pass through areas with no prior evidence of
archaeological activity. However, the relatively small
area of the excavation precludes any wide ranging
discussion of its content or significance.

The farmstead
Evidence of ‘domestic’ activity is confined to the
north-western area of the site, to which the linear
features 1155, 1156 and 1157 appear to form a
boundary. The overall indication is that the area
examined is situated on the margins of occupational
activity, essentially at the edge of a field. At six
metres in diameter, the post-built structure 1118,
though undated, is smaller than most other examples
of ‘native’ or Iron Age roundhouses and appears to
lack any internal features except for the central post
which would be unnecessary on such a small roof
span. It is likely, therefore, that it did not have a truly
domestic function and may have been no more than
a hut. Though undated, it appears to have been
respected, spatially, by the hearths and other features
dotted about it, the dated examples of which fall
within the 3rd to 4th centuries AD.

Charred plant remains were recovered from the
hearths associated with the hut. These included

cereals, predominantly wheat with some barley, in a
not fully prepared state and probably grown locally.
The presence of some chaff and weeds of cultivation
suggests that threshing and cleaning may have been
carried out nearby. Other plant remains are all open
ground species, some of which occur in cultivated
soils but also including a variety of grassland species,
such as vetch, clover, wood-rush and sedge, and
characteristic pasture grasses such as timothy and
perennial rye grass. Some of these may also indicate
the presence of straw on the site, in which case one
possible function for structure 1118 could have been
as a byre (though no samples were taken from
features associated with this structure).

The animal bone assemblage is not large but
includes typical domestic and farm animals
including young individuals. Sheep/goat, cattle and
pig are all present with some horse and several dogs
—one buried with, presumably, its master — and three
new-born puppies.

The purpose of the east-west ditch 1064 is
unclear, as it does not appear to form a boundary. At
its western terminus it cuts the edge of the calf burial
pit 1043, while 2.5m east in a direct line from the
eastern terminus were the truncated remains of
another probable animal burial, containing cattle
bone. The reasons for these two, and the other
animal burials, is unclear; no evidence of butchery
was noted on any of the bones from the burials,
though this does not preclude the possibility that the
animals had been killed rather than died of natural
causes; they may simply represent the most
convenient method of disposal of carcasses at the
edge of a field, or, alternatively, they may be of a less
expedient nature. In any case, the alignment of the
ditch and the two animal burials is surely too close to
be pure coincidence, and for that reason they are
considered, here, to be associated. Ditch 1064 was
cut or replaced by ditches 1155, 1156 and 1157.
These clearly form components of a maintained
boundary with an entrance at the south-west end,
demarcating or enclosing the majority of the
‘domestic’ features in the north-west corner of the
site, and forming the northern or upslope margin of
the quarry pits and lynchet deposit 1269,
components of which they cut.

However, the relationship between the quarry
pits and colluvium 1269, the boundary ditches and,
hence, the ‘farmstead’ area remains unclear. There
are burials sealed by the colluvium (128) and burials
cut into it (1026), and though the ditches 1155,
1156 and 1157 appear to form the northern
boundary of 1269 thereby pre-dating it, all except

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 69

1157 clearly cut into it and 1064, the earliest of the
four, disregards it altogether. The conundrum can,
perhaps, be best explained by the gradual nature of
colluvium deposition, and the particular undis-
turbed topographic and agricultural conditions
necessary for it. The irregular pits, over which it
appears to have settled, are most probably marling
pits and their occurrence is not surprising in view of
the presence of heavy, clay-rich, acidic clay-with-
flints soils nearby. Their concentration here in a
broad linear sweep across the contours of the hill
must indicate marginal land at the edge of
cultivation areas, conditions also predisposed to the
stabilisation of colluvial deposits. The molluscs
sampled from within it are typical of hillwash
deposits. The two processes appear to have run more
or less in tandem here in an area which was, perhaps,
later formally demarcated by the boundary ditches.
The intercutting of the pits themselves and the
graves with the colluvium, can easily be understood
in the light of the time span necessary for this depth
of colluvium to develop.

Pit 1195 may have been dug originally as a
storage pit, of the sort common on chalkland Iron
Age sites, but the series of thin organic silt layers at
its base included small fragments of apparently
digested animal bone, so that it may have been used
from time to time as a convenient latrine. These
layers were sealed by deliberately dumped material,
above which were upper fills containing bones of
three neonatal puppies, other animal bone, pottery,
two bone pins and other material -consistent with the
pit having been used for opportunistic or ritualised
rubbish disposal. The function of large pits is open
to debate but there is little evidence in rural locations
that they were dug for the sole purpose of domestic
rubbish disposal. Unfortunately there is no evidence
for the primary use of this pit.

Although there is evidence of prehistoric activity
on the site in the form of a few sherds of pottery and
flint work, this is considered residual or redeposited,
mostly occurring with Romano-British material.
Dating evidence from the artefacts, particularly the
pottery, indicates a predominantly late Romano-
British (3rd—4th century) date, with some earlier
Roman pottery also present.

The cemetery

The cemetery appears to be that of a normal, small
domestic type, associated with an agricultural com-
munity containing inhumation and, by implication,
cremation burials. Although no cremation burials
were found the presence of cremated human bone in

the probable pyre debris pits would suggest there are
some in the vicinity. Such mixed cemeteries were not
unusual in the Romano-British period. The full
extent of the cemetery is unknown.

The common burial posture in Romano-British
inhumations was supine and extended. Crouched
burials, being more common in rural areas, are seen
as a persistence in native burial practice in the first
two centuries AD (Philpott 1991). Inhumations
1037 and 1179 were crouched in shallow graves, 124
flexed, 1005 extended and prone and 1012 extended
on the left side. One of the crouched burials, 1037,
and probably a second shallow burial, 1046, were
cut by later Romano-British features, which would
suggest they are earlier in date than the deeper
burials which also contained later Romano-British
dating evidence.

Prone burials are relatively rare, and occur more
frequently in the late Romano-British period
(Philpott 1991). Some demonstrate traits implying
execution, others of careless or deliberately disres-
pectful deposition; those with a more ‘formal’
appearance are often placed on the edges of the
cemetery. Inhumation 1005 was apparently buried
with care, implying respect; it was laid in an extended
position within a coffin, a bronze coin in the mouth
and a small dog placed across/under the chest, head
on shoulder. The grave was immediately adjacent to
that of 1012, and did not appear to be separated from
the others. It is not inconceivable that the coffin was
accidentally inverted during deposition, though one
might perhaps expect more movement of the body
within the coffin during turning if that had been the
case. Alternatively, some unknown ritual may be
indicated. The pronation of the dead to stop them
from rising has been suggested as one possibility
(Harman et al. 1981). Unfortunately, the crushing of
this shallow grave prior to stripping of the site might
have destroyed explanatory evidence.

Coins are frequently found in the mouths of
inhumations, usually where no other associations are
present (Philpott 1991). One inhumation from the
rural cemetery at Ilchester (grave 37) was found with
an illegible coin in its mouth and was accompanied
by a dog, as here. The coin of Faustina II from
inhumation 1005 provides an earliest possible date
for the burial of AD 161-175, though, as the coin is
extremely worn, indicating a long period of
circulation, it could be substantially later than this.

Dogs are not uncommon companions in
Romano-British graves (Philpott 1991), where they
are believed to represent pets or possibly status
hunting dogs. The medium-sized, male animal with

70 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

1005 was apparently quite elderly, and is most likely
to have been a personal pet. Similar associations of
dogs with humans in inhumations of Romano-British
date have been found at Baldock, Hertfordshire
(Hamilton-Dyer forthcoming), Lankhills (Clarke
1979, grave 400) and Figheldean (Graham and
Newman 1993, grave 61). Though usually intact, there
are a few instances where dogs have been decapitated
or otherwise mutilated: for example, Alington Avenue,
Dorchester (Maltby forthcoming, graves 3661 and
4389). There are no_ references to _ possible
fittings/harnesses of the type noted from 1019.

Flexed inhumation 124 had c.80 hobnails at its
feet, the footwear obviously being worn by the
deceased at the time of burial. Hobnails are
commonly recovered with burials, especially in rural
sites (Philpott 1991). Other types of footwear do not
appear to have been recorded in burials, though the
presence of leather shoes is considered likely
(Philpott 1991). The presence of a type of leather
sandal or patten, is demonstrated in inhumation
1012 where five small copper-alloy rivets, one 77 situ
in a fragment of leather, were recovered from around
the feet. Blue/green and dark brown staining was
noted on the posterior surfaces of both Ist and 2nd
metatarsals and the left 1st proximal phalanx, which
suggests a copper-alloy riveted strap crossing the
foot. The rivets probably served a decorative rather
than a practical purpose, being confined to the
medial side of the feet (Figure 4). The shoes were
apparently worn at the time of burial.

Inhumation 1012 was coffined, and in a deep
grave cut. Extended on the left side, it appears that,
during deposition, the body slid against the right side
of the coffin (probably because of the way it was
lowered into the deep grave) then slumped forwards,
slumping further still during decomposition. Forty-
two coffin nails were recovered from the head and
foot ends of inhumantion 1012 (Figure 4): the
number appears unusually large for purely construc-
tional purposes and may indicate a decorative use for
some of them.

The variation in burial posture, including
crouched, flexed and extended burials, does not
appear to relate to the age or sex of the individuals.

Summary

Together, the structural, artefactual and _ paleo-
environmental evidence points to marginality; the
excavation appears to have clipped the edge of a
small, Romano-British farmstead, the nucleus of
which may lie closer to the top of the spur, probably
c.140m to the north, adjacent to the east—west

bridleway which is believed to follow the line of the
Roman road between Salisbury and Devizes. The
small round building associated with hearths seems
most likely to have had an agricultural use rather
than a domestic one. The environmental evidence
suggests a typical low-level mixed farming economy
with both arable and pastoral elements.

The full extent of the cemetery is unknown, the
inhumation burials being fairly widespread across
the site, close to edges of the fields. More burials are
likely to exist, and certainly the presence of hitherto
undetected cremation burials is strongly indicated by
the occurrence of cremated human bone within the
pyre debris pits. The period over which the burials
were made is unknown but since at least one of the
graves (128) was sealed by soils associated with the
formation of the lynchet, a fairly lengthy timescale
within the later Romano-British period is indicated,
consistent with the occupation of a small farm over
several generations.

Acknowledgements. The project was funded by Esso Petroleum Co
Ltd and carried out by Wessex Archaeology. Particular thanks are
due to Mr R. Chapman, Esso’s project manager who initiated the
works and whose assistance and co-operation throughout the
excavation aided its progress and was much appreciated. The
writers are grateful to Mark Corney (RCHM(E)) for his advice on
the archaeoiogical background.

The full-time excavation team comprised Keith Cooper and
Jeremy Fry, under the supervision of David Murdie and Jacqueline
I. McKinley, with the occasional assistance of Elaine Wakefield and
Bill Timmins, and several of the other site contractors. The
environmental samples were processed by Sarah Wyles. The project
was managed by Michael Heaton. The drawings were produced by
Julian Cross and Erica Hemmings and the report was initially edited
by Julie Gardiner.

REFERENCES and BIBLIOGRAPHY

ADAMS, J.C., 1985 Outline of Orthopaedics, Edinburgh/
London

— 1987, Outline of Fractures, Edinburgh/London

ALLEN, M.J., 1992 ‘Products of Erosion, and Prehistoric
Land Use of the Wessex Chalk’, in M.G. Bell and J.
Boardman (eds.), Past and Present Soil Erosion:
Archaeological and Geographical Perspectives, Oxford,
37-52

ANDERSON, A.S., 1979 The Roman Pottery Industry in North
Wiltshire, Swindon Archaeological Society Report 2

BASS, W.M., 1987 Human Osteology, Columbia

BEEK, G.C. van, 1983 Dental Morphology: An Illustrated
Guide, Bristol

BRODRIBB, G., 1987 Roman Brick and Tile, London

BROOKS, S.T., 1955 ‘Skeletal age at death: The reliability of
cranial and pubic age indicators’, American Fournal of
Physical Anthropology 13, 567-97

BROTHWELL, D.R., 1971 Digging up Bones, London

BULL, G., and PAYNE, S., 1982 “Tooth eruption and
epiphysial fusion in pigs and wild boar’, in B. Wilson, C.
Grigson and S. Payne (eds.), Ageing and Sexing Animal
bones from Archaeological Sites, British Archaeological
Reports (British Series) 109, Oxford, 55-72

ROMANO-BRITISH FARMSTEAD AND BURIALS, MADDINGTON FARM 71

CLARKE, G., 1979 Pre-Roman and Roman Winchester: Part I

The Roman Cemetery at Lankhills, Winchester Studies 3,
Oxford

CRUMMY, N., 1983 The Roman Small Finds from Excavations
at Colchester 1971-79, Colchester Archaeological
Report 2

CLEAL, R. and RAYMOND, F., 1990 ‘The Prehistoric
Pottery’, in J. Richards, The Stonehenge Environs Project,
English Heritage Archaeological Report 16, London,
233-46

DAVIES, S., BELLAMY, P.S., HEATON, M.J., and WOODWARD,
P., forthcoming: Excavations at Alington Avenue,
Fordington, Dorchester, Dorset Natural History and
Archaeological Society Monograph Series

DRIESCH, A., VON DEN, and BOESSNECK, J., 1974 Kritische
Anmerkungen zur Widerristhéhonberechnung — aus
Ldangenmapfen vor- und friihgeschichthcher Tierknochen,
Saugertierkundliche Mitteilunger 22, 325-48

DRIESCH, A., VON DEN, 1976 A Guide to the Measurement of
Animal Bones from Archaeological Sites, Harvard,
Peabody Museum Bulletin 1

EVANS, J.G., 1972 Land Snails in Archaeology, London

FARWELL, D.E., and MOLLESON, T.1., 1993 Poundbury
Volume 2: The Cemeteries, Dorset Natural History and
Archaeological Society Monograph 11, Dorchester

FULFORD, M.G., 1975 New Forest Pottery, British
Archaeological Reports (British Series) 17, Oxford

GRAHAM, A., and NEWMAN, C., 1993 ‘Recent Excavations
of Iron Age and Romano-British Enclosures in the
Avon Valley, Wiltshire’, WAM 86, 8-57

GRAY, H., 1977 Anatomy, New York

GUILBERT, G., 1982 ‘Post-ring Symmetry in Roundhouses
at Moel y Gaer and Some Other Sites in Prehistoric
Britain’ in PJ. Drury, Structural Reconstruction;
Approaches to the Interpretation of the Excavated Remains
of Buildings, British Archaeological Reports (British
Series) 110, Oxford, 67-86

HAMILTON-DYER, S., forthcoming: “The Animal Bones’, in
G. Burleigh, forthcoming (see J.I. McKinley,
forthcoming)

HARCOURT, R.A., 1974 “The Dog in Prehistoric and Early
Historic Britain’, Fournal of Archaeological Science 1,
151-76

HARMAN, M., MOLLESON, T.I., and PRICE, J.L., 1981
‘Burials, bodies and beheadings in Romano-British and
Anglo-Saxon cemeteries’, Bulletin of the British Museum
Natural History (Geology) 35(3), 145-88

HILLSON, S., 1986 Teeth, Cambridge
Archaeology

ISCAN, M.Y., LOTH, S.R., and WRIGHT, R.K., 1984 ‘Age
estimation from the ribs by phase analysis: white males’,
Fournal of Forensic Sciences 29, 1094-104

— 1985 ‘Age estimation from the ribs by phase analysis:
white females’, Journal of Forensic Sciences 30, 853-63

LEACH, P., 1982 Ilchester Volume I: Excavations 1974-75,
Western Archaeological Trust Monograph 3

MALTBY, J.M., 1985 ‘The Animal Bones’, in P.J. Fasham,
The Prehistoric Settlement at Winnall Down, Winchester,
Hampshire Field Club Monograph 2, 97-112

— 1988 “The Animal Bones from the 1984/85 Excavations
at Alington Avenue, Dorchester, Dorset’, unpublished

_ Ancient Monuments Laboratory Report 182/88

— 1987 ‘The Animal Bones from the Excavations at
Owslebury, Hampshire: an Iron Age and Romano-

manuals in

British Settlement’, unpublished Ancient Monuments
Laboratory Report 6/87
— forthcoming: The Animal Bone, in S. Davies er al.,
forthcoming (A.M. Lab report no. 192/88)
MANNING, W.H., 1985 Catalogue of the Romano-British Iron
Tools, Fittings and Weapons in the British Museum,
London
MARSHALL, A., and MCKINLEY, J.I., in prep. Observations
of an Experimental Pyre Cremation at Guiting Power,
Glos.
MCMINN, R.M.H., and HUTCHINGS, R.T., 1985 A Colour
Atlas of Human Anatomy, London
MCKINLEY, J.I., 1989, ‘Cremations: Expectations,
Methodologies and Realities’, in C. Roberts, F. Lee and
J. Binthiff (eds.), Burial Archaeology: Current Research
Methods and Developments, British Archaeological
Reports (British Series) 211, Oxford 65-76
— forthcoming: “The Human Bone from Baldock Area 15
Cemetery’, in G. Burleigh, Excavations at Baldock: An
Iron Age and Roman Settlement, Vol. 1: Inhumations
and Cremations, Letchworth Museum Service
— in prep. Cremated Bone from Burials and Cremation-
related Contexts, in Barber and Bowsher, East London
Cemeteries, Museum of London, English Heritage
Monograph Series
MEPHAM, L.N., 1993 “The Pottery’
Newman 1993, 25-34
MILES, D., 1978 “The Roman Pottery’, in M. Parrington,
The Excavation of an Iron Age Settlement, Bronze Age
Ring-Ditch and Roman Features at Ashville Trading Estate,
Abingdon (Oxfordshire), 1974-76, London, C.B.A.
Research Report 28, 74-8
MILLARD, J.I., 1996 “The other pottery’ in M.N. Rawlings
and A.P. Fitzpatrick, ‘Prehistoric Sites and a Romano-
British Settlement at Butterfield Down, Amesbury’
WAM 89 (this vol.), pp.27ff.
MILLS, J.M., 1993a ‘Hobnails’, in Farwell and Molleson
1993, 99
— 1993b ‘The Coffin Nails’, in Farwell and Molleson
1993, 114-16
MORRIS, E.L., 1992 Guidelines for the Analysis of Pottery,
Wessex Archaeology Guideline 4, Salisbury

PEACOCK, D.P.S., 1987 ‘Iron Age and Roman quern
production at Lodsworth, West Sussex’, Antquaries
Journal 67, 61-85

PHILPOTT, R., 1991 Burial Practices in Roman Britain; a
Survey of Grave Treatment and Furnishings, British
Archaeological Reports (British Series) 219, Oxford

RCHMEa, in prep. The Earthworks of Salisbury Plain, Vol.
I: South Wiltshire, HMSO Monograph

— b, in prep. The Earthworks of Salisbury Plain, Vol. II:
The Salisbury Plain Training Area, HMSO Monograph

ROGERS, B. and RODDHAM, D., 1991 “The Excavation at
Wellhead, Westbury 1959-66’, WAM 84, 51-60

ROGERS, J., WALDRON, T., DIEPPE, P., and WATT, I., 1987
‘Arthropathies in palaeopathology: the basis of
classification according to most probable cause’,
Journal of Archaeological Science 14, 179-93

RSK ENVIRONMENT LTD, 1992 Avonmouth to Fawley Booster
Station Environmental Review for Esso Petroleum Company
Ltd, Helsby

SANDERS, J., 1979 “The Roman Pottery’, in G. Lambrick
and M. Robinson, Iron Age and Roman Riverside
Settlements at Farmoor, Oxfordshire, CBA Research
Report 32, London, 46-54

in Graham and

72 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

SWAN, V., 1971 ‘The Coarse Pottery’ in G.J. Wainwright,
“The Excavation of Prehistoric and Romano-British
Settlements near Durrington Walls, Wiltshire, 1970’,
WAM 66, 76-128

— in prep. Romano-British Pottery Production in the New
Forest

TROTTER, M. and GLESER, G.C., 1952 ‘Estimation of
stature from long bones of American whites and
negroes’, American Journal of Physical Anthropology
10(4), 463-514

— 1957 ‘A re-evaluation of estimation of stature based on
measurements of stature taken during life and of long
bones after death’, American Journal of Physical
Anthropology 16(1), 79-123

WAM, 1994 ‘Excavation and Fieldwork in Wiltshire 1992:
The Archaeology of Salisbury Plain Training Area;
Prehistoric and Romano-British’, WAM 87, 158-59

WEBB, P.A.O., and SUCHEY, J.M., 1985 ‘Epiphyseal union of
the anterior iliac crest and medial clavicle in a modern
multiracial sample of American males and females’,
American Fournal of Physical Anthropology 68, 57-466

WELLS, C., 1982 “The Human Burials’, in A. McWhirr, L.
Viner and C. Wells, Romano-British Cemeteries at
Cirencester, Cirencester

WILLIAMS, D.F., 1977 “The Romano-British Black
Burnished Industry: An Essay on Characterisation by
Heavy Mineral Analysis’, in D.P.S. Peacock (ed.),
Pottery and Early Commerce: Characterisation and Early
Trade in Roman and Later Ceramics, London, 163-220

WESSEX ARCHAEOLOGY, 1993 ‘Mill Street, Wantage,
Oxfordshire, Archaeological Site Investigations’,
unpublished client report W603, Salisbury

YOUNG, C.J., 1977 The Roman Pottery Industry of the Oxford
Region, British Archaeological Reports (British Series)
43, Oxford

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 73-83

West Chisenbury: Settlement and Land-use in a Chalk
Downland Landscape

by GRAHAM BROWN

West Chisenbury, a detached tithing of the parish of Netheravon until 1885, is typical of many tithings
along the valleys of the Wessex chalk downland. Each had a proportion of the land resources, the
amount dependent on the size of the settlement, with meadow restricted to the valley floor and arable

cultivation extending from the settlement towards the downs. At the extremities of the settlement lay the
pasture and waste. At West Chisenbury well-preserved settlement earthworks representing a cluster of

medieval farmsteads are evident. Since the tithing hes within a military training area it has not suffered

from the agricultural improvements of the past century to the same extent as those elsewhere in the
county and, consequently, the archaeological landscape can still be studied in detail. This paper
attempts to place the medieval settlement of West Chisenbury in its landscape context by combining the

data from field survey and field investigation with the available documentary evidence.

INTRODUCTION

The shrunken settlement at West Chisenbury is
situated within a broad meander on the west bank of
the river Avon at about 92m above OD on Lower
and Middle Chalk and valley alluvium (Figure 1).
The settlement remains are located partly on the
first river terrace above the flood plain, with the
remainder spilling out onto the flood plain. On the
opposite side of the river is the hamlet of East
Chisenbury. This pairing of settlements is charac-
teristic not only of the Avon valley but also of other
medieval settlements on the south Wiltshire
chalkland. Examples include Steeple Langford and
Hanging Langford in the Wylye valley and North
Burcombe and South Burcombe in the Nadder
valley (RCHME (a) in preparation).

Investigation of the site itself is one element of a
research project being undertaken by the author
which aims to assess the nature of medieval and
post—medieval settlement in the Avon valley north of
Salisbury. The area also forms part of a study being
undertaken by the Royal Commission on_ the
Historical Monuments of England (RCHME (b) in
preparation).

The Avon valley was, and still is, an important
route for travel through the chalk massif of
Salisbury Plain. A wealth of Roman and Saxon
material also points to its longevity as a favoured
environment for settkement. A series of finds of
Roman building debris and associated building

foundations suggests that the Avon valley was the
focus of a number of structural complexes (for
example, Annable 1968, 119). Given the pattern
and nature of Roman settlement of Salisbury Plain
it seems reasonable to suggest that these valley
bottom complexes functioned in part as estate
management centres (RCHME (b) in preparation).
There are also finds of Saxon material here
(Annable 1967, 125) and the roadway utilising the
valley floor has a potential Saxon precursor
(Cossons 1959, 254). It is not intended here to
claim a direct line of continuity in settlement from
the Roman period but, rather, that these earlier
estates established a heavily organised and settled
landscape which laid the foundations for subse-
quent exploitation of the valley floor.

HISTORICAL BACKGROUND

At the time of the Domesday survey, West
Chisenbury was held by Nigel the Physician as one
of three estates that belonged to Netheravon church.
It was the smallest of the estates and assessed as
having land for only 1 plough, 3 smallholders with
> plough. With 6 acres of meadow and pasture 4
furlongs long and 2 furlongs wide, it was valued at
£3 (Thorn and Thorn 1979, 56.3). In contrast East
Chisenbury, the second largest of his estates, was
valued at £13. The third estate was some distance
from Netheravon, at Stratton St Margaret, and was
valued at £16 (ibid., 56.1—2).

74 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

DEVIZES
e

Ee SALISBURY ®
\
WEST

CHISENBURY
,

COMPTON

- 2-100 = -

EAST
CHISENBURY
\

\
a2

Figure 1. General Location of West Chisenbury, (RCHME, Crown copyright)

By the 12th century West Chisenbury was sub-
infeudated and from the beginning of the 13th until
the late 14th century it was held by the de la Foyle
family. In 1313, 2 carucates and 6 = acres
of meadow were held by the chaplain, Philip
Dyonys, for a rent of 40s. (Pugh 1939, 84) until
1337 when this land, including the messuage,
reverted to Henry de la Foyle (Elrington 1974, 48).
By the mid 14 century the estate had passed to John
Breamore and in 1361 the demesne messuage

included a dovecote and garden. Additionally there
were 240 acres of arable land worth 3d. an acre, 12
acres of meadow and common pasture for 6
working cattle, 20 oxen and grazing for 400 sheep
(Stokes 1914, 276). In 1428 it was in the
possession of Robert Browning when the estate
formed part of the Netheravon prebend of
Salisbury Cathedral; and in 1649 the holding at
West Chisenbury contained an old decayed barn of
4 bays (Bodington 1919, 301). The estate remained

WEST CHISENBURY: SETTLEMENT AND LAND USE

in the Browning family until 1708 when it was sold
to John Flower.

Throughout much of the medieval and post-
medieval periods a number of grants of land were
made at West Chisenbury to various individuals
including monastic houses, and it was not until the
early 18th century that the estates were finally
combined.

The largest of these estates was one of 5) virgates
that was conveyed to Peter Bacon in 1201 by Roger
de la Foyle, the mesne lord of the capital manor.
Bacon’s descendants retained the estate until 1624
when it was sold to William Rolfe (Fry and Fry
1901, 124-6). A year later it was sold to John
Merewether and in 1648 it was enlarged by 81 acres.
The estate was known as the manor of Chisenbury
or Chisenbury de la Folly and was held in free socage
(free tenure but no military service) from the
Browning family in the early 17th century; it
comprised the capital messuage, 180 acres of arable
land, 10 acres of meadow, and 12 acres of pasture. It
formed part of a much larger estate which included
land in Whiteparish, Upavon, and Rushall (zbid., loc.
cit.; Pugh 1939, 81). In 1720 the land was sold to
John Flower.

In the 13th century there is also mention of a
tenement at West Chisenbury held by John de la
Roches. In 1252 his descendants held a third of a
carucate and 2 virgates at Chisenbury and Coombe
in Enford. In 1354, not long after the Black Death,
the Roche land at West Chisenbury comprised a
messuage, 3 virgates of land, and 2 acres of meadow.
The estate was eventually sold to John Flower in
1723 and merged with the capital manor (Stevenson
1980, 174).

In 1227 a gift by Richard de la Folie of a
messuage and an unspecified amount of pasture to
the Augustinian Canons of Maiden Bradley was
confirmed; they retained the land until the
Dissolution in 1536 (Chettle and Kirby 1956, 296,
301). A small amount of land was also held by the
newly founded Preceptory at Ansty, a minor house
of the Knights Hospitallers (Stevenson 1980, 174).

By the early 18th century West Chisenbury was
worked as a single farm located on the eastern side of
the present A345 road. In 1776 West Chisenbury
was purchased by William Beach who also owned a
substantial estate in Netheravon (Stevenson 1980).
In 1861 Beach sold his Netheravon and West
Chisenbury estates_to Lord Normanton and thirty
six years later West Chisenbury was one of the first
estates on Salisbury Plain purchased by the War
Department. Since this date the estate has been in

government ownership; nevertheless, farming cont-
inues on the lower slopes in much the same area as at
the time of the Domesday survey.

During the 18th century the pattern of land
tenure began to change and new farms were built on
the downs as more land was cultivated, a process
that accelerated after parliamentary enclosure. West
Chisenbury was no exception to this process and a
field barn was built on the downs sometime after
1796. Farming expansion reached its peak in this
region by 1879 and was followed by an agricultural
depression which was to last into the early 20th
century (Perry 1974). During this period rents on
the Beach estate fell by 40 per cent which led to a
reduction in arable cultivation and much of the down-
land reverted to pasture (Stevenson 1980, 177).

A water mill is recorded at the time of the
Domesday survey at East Chisenbury; this mill may
be the same one that was owned by the Proctor of
Ogbourne who, in 1230, had to pay 2d. annually to
Roger de la Foyle for the use of the leat and sluice
which were evidently within the bounds of West
Chisenbury (Chibnall 1951, 57).

A chapel of ease is recorded at West Chisenbury
in 1405, although a chaplain is mentioned in 1313
(see above); by 1535 it was no longer standing
(Jackson 1867, 268). In c.1650 a suggestion was
made that West Chisenbury should be linked to the
neighbouring parish of Enford for ecclesiastical
purposes, since attendance at the parish church at
Netheravon by the hamlet’s residents was so low
(Bodington 1919, 300).

Population estimates are difficult to assess for
West Chisenbury since it was a detached tithing of
Netheravon parish until the late 19th century. Many
of the tax returns and other population indicators
have been obscured by their inclusion within the
parent parish assessment. In the tax list of 1332 East
and West Chisenbury were regarded as one vill and
19 individuals are listed, including Henry de la
Foyle, who was lord of West Chisenbury manor
(Crowley 1989, 65). Since East Chisenbury was the
larger of the two estates at the time of the Domesday
survey, it seems likely that a little over 200 years later
the majority of the 1332 tax payers were from there.
The 1377 poll tax lists 111 payers within Netheravon
parish but again does not separately identify West
Chisenbury (Beresford 1959, 308). Taxation assess-
ments for the 16th century show that Netheravon
was the third most highly rated parish in its hundred.
By the 19th century West Chisenbury was listed
separately and in 1811 the population was 38, rising
slightly to 47 by 1891.

70 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

1000 METRES

ee a ]

in “SEs

/| CHISENBURY ~~

DEVIZES

LAVINGTON
WAY

DOWN NE THERAVON

UPAVON

WEST
CHISENBURY

River Avon

COMPTON

ARABLE MEADOW

Figure 2. West Chisenbury tithing showing expansion of cultivation onto the downs and extension of meadow along the
River Avon. (RCHME, Crown copyright)

ARCHAEOLOGICAL BACKGROUND

In 1926, during the construction of houses 100
metres to the north of the settlement (¢ on Figure 3),
a pagan Saxon burial was discovered (Cunnington
1930, 84). The location of this burial, within 300
metres of the tithing/parish boundary is significant,
given that Bonney (1966, 25-30) has shown that
there is a marked correlation between pagan Saxon
burials and land boundaries. Bonney comments
‘that those boundaries ... were in being as early as
the pagan Saxon period and they imply the existence
of a settled landscape...’. Unfortunately, the bounds
of the Enford Saxon charter which, according to
Grundy (1919, 228), include West Chisenbury,
cannot be ascertained clearly. ‘There is even some
doubt as to whether the charter refers to Enford in
Wiltshire or a similarly named settlement in
Hampshire (Sawyer 1968, 174). In spite of this, the

present day boundary of West Chisenbury may still
reflect a much earlier, possibly Roman, land holding.

THE TITHING LANDSCAPE

The tithing boundary is formed by the river Avon on
the east while to the west it extends in a finger-like
projection onto the open downland of Salisbury
Plain. Much of the remainder of its perimeter is
marked by a bank and elsewhere, as on Chisenbury
Down and Lavington Way Down, its course was
formerly indicated by boundary mounds and a
double linear ditch. In places, the tithing boundary
clearly follows the lynchets of ancient fields. This 1s
particularly evident near the linear ditch where the
boundary divides Compton and West Chisenbury.
The use of ancient fields as boundaries can be seen
elsewhere on Salisbury Plain: for example, the

WEST CHISENBURY: SETTLEMENT AND LAND USE

southern boundary between Bratton and Edington is
stepped as it negotiates a prominent lynchet. By
contrast, in the Till valley the parish boundary of
Berwick St James near the deserted village of
Asserton ignores a series of massive strip lynchets
which it crosses (C. Dunn pers. comm.).

The communication pattern close to the
settlement was dominated by tracks either running
between the settlements on the west bank of the river
Avon, or by those leading out to the arable fields and
pasture. A major track, (a—b on inset to Figure 3) lay
above the first river terrace and was the route
between Netheravon and Upavon. However, this
track became less important when the A345 road
was constructed 200 metres to the east and through
the abandoned settlement earthworks, some time
between 1849 and 1889. Another track led north
close to the river towards Upavon. Branching from
the track (a—b) were three further tracks. The
southerly one is followed by the A345 road. To the
north, the second track survives as a hollow way and
hedge-line that led directly to West Chisenbury
Farm. The third track (c—d on Figure 3 and inset) led
from the settlement onto the downs. Formerly this
track continued east at d, through the medieval
settlement, and on to the flood plain (m and n on
Figure 3). However, it has been diverted at d to
follow the hedge-line in a south-easterly direction
towards West Chisenbury Farm.

Beyond the settlhement a number of tracks
crossed the tithing landscape of which the most
notable are the two in the extreme west. The first
was the drove track between Netheravon and
Devizes which lies beside a linear ditch, perhaps the
Ealdan Dic (Old Ditch) of the Enford Saxon charter
(Grundy 1919, 234). Crossing this track, with a
corresponding break in the linear ditch, was another
track that led from Upavon to Tilshead; in the north
it survives as a deeply incised hollow way cutting
diagonally through an ancient field system. Further
south it lies parallel to ridge and furrow and later
becomes a slight hollow way beside a prominent
field lynchet.

Three types of land-use have been identified
within the tithing: meadow, arable, and pasture and
waste. The Domesday arable was probably located
to the north-west of the settlement in a natural bowl
with relatively steep slopes on three sides. By 1361
cultivation had expanded south onto the higher
ground above the settlement. At 3d. an acre the West
Chisenbury arable was similar in value to that of
other downland tithings: for example, at Market
Lavington in 1293 there were ‘374 poor acres upon

hgh

the hill, and the acre was worth 2d.’ (Fry 1908, 192).
Although there appears to be no evidence for the
type of field system in operation here during the
medieval period, by analogy with other downland
estates in the area, it is likely to have been an open
two field system by the 13th century (Tate 1945,
139). There are no further indicators of arable
expansion at West Chisenbury during the medieval
period. Occasionally, however, there may have been
temporary intakes; this certainly occurred on other
downland estates in the area such as at Shrewton
(Bennett 1887, 35).

By the late 18th century cultivation had spread
westwards onto Chisenbury Down. The area here is
bisected by three deep valleys which were shown as
being uncultivated on a map of 1849 (HRO map,
M20). The steepness of the slopes, and the absence
of strip lynchets would suggest that they have never
been cultivated. It is therefore likely that neither the
western part of Chisenbury Down nor Lavington
Way Down witnessed substantial episodes of
cultivation in the medieval period.

There were three distinct areas of meadow at
West Chisenbury. To the north a narrow band of
meadow, 7 acres in area, extends from the settlement
and terminates at the tithing boundary, whilst to the
south of the track leading to East Chisenbury there is
a much more extensive meadow of 17 acres.
Sandwiched between the two is an area known in
1898 as Saucers Meadow (MoD 1897): this covers 4
acres and may have been named after an eponymous
resident of West Chisenbury mentioned in 1329
(Stokes 1914, 44) and may also be a remnant of the
6 acres referred to in the Domesday survey. The
meadows to the north and east were later converted
to floated water meadows, although the larger
meadow in the south does not appear to have been
adapted for this purpose.

The two main areas of pasture and waste were on
Lavington Way Down (100 acres) and Chisenbury
Down (about 518 acres in 1796). The former was
probably named from the track that formed its
northern boundary. However, by the late 19th
century this name and that of Chisenbury Down had
inexplicably been changed to Slay Down and
Compton Down, respectively. Interestingly, the
incongruous shape of Lavington Way Down,
projecting as it does into the parish of Upavon,
suggests that it may once have been included in that
parish.

A breakdown of the land-use in West Chisenbury
tithing from the Domesday survey until the late 19th
century is given in the Appendix (see page 83).

78 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

- 1794

(la
2
Ly,

West Chisenbury

Moke

100 METRES,

2 >
; , ms
aygrennn cet en TEU

c
ALE
Ate Me

©Crown Copyright

138

oT
ly,

nti,

>
=
$

fo}
2
E
6

COMMISSION

i HISTORICAL

=
Ss ROYAL

inset shows the hamlet at
the time of Enclosure.
(RCHME, Crown

Figure 3. The earthworks
at West Chisenbury. The

WEST CHISENBURY: SETTLEMENT AND LAND USE

FIELD SURVEY OF THE EARTHWORK REMAINS
(Figure 3)

(The italic letters in the following account refer to
the letters on the plan.)

The settlement earthworks extend from the top of
the river terrace eastwards on to the flood plain. In
places the area of settlement on the flood plain is
lower than the present course of the river indicating
that as the settlement expanded the river has been
canalised and therefore restricted to its present
course in the east. In January 1995 after prolonged
heavy rain the river burst its banks and temporarily
reverted to its former meandering course through
the settlement. The creation of later water meadows
in the east has also tended to mask the settlement
remains.

The Street Pattern Within the Settlement

The principal surviving settlement earthworks are
laid out on either side of a former street now
represented by a hollow way. This hollow way is best
preserved at its western end between c and d. East of
d it is traceable as a very broad linear depression
crossed by at least three step-like scarps which may
represent an extension of settlement (see below), or
later agriculture, after this section of the street had
been abandoned. Near m the former street descends
on to the flood plain and survives as a pronounced
hollow way which was later used as a main drain for
the northern water meadow. It soon changes
direction to run south to join a second street (k-/) at
n. East of n it remains in use as a track that leads to a
river crossing to East Chisenbury.

The second street lies to the south and again
leads from the downs to the settlement. At its
western end it survives as a hollow way, c.5m wide,
beside a hedgerow and fence-line, which gradually
fades out near the present A345 road. Formerly this
street continued east at k and / to the former farm-
stead (w) to meet the northern street at 7.

The Settlement

The settlement earthworks appear to fall into three
elements. The first is located at the top of the river
terrace and extends westwards from e— and is largely
confined to the area between the two streets. It also
extends on to the northern side of the northern
street where scarps representing at least four
regularly laid out rectilinear properties are visible,
bounded by an irregular scarp (7-j) which fronted the
former street, and a boundary bank, or headland,
g-h. These properties are associated with step-like
scarps crossing the street suggesting that when this

section of the street was abandoned some
restructuring of the properties may have taken place
with their boundaries being extended to take in the
area formerly occupied by the street. The best
preserved earthwork is an enclosure at p which
measures c.25m’ and appears to be the remains of a
walled structure; it has been bisected by the present
road.

In the area between the streets the earthworks are
poorly defined and have been damaged by later
quarrying and by the present farmhouse. A promi-
nent sub-square platform, measuring 50m’, lies at g.
The earthworks associated with the southern street
(k-l) are now very fragmentary and have been largely
destroyed.

The second element comprises the settlement
remains located east of e-f on the flood plain.
Earthworks representing former rectilinear prop-
erties lie chiefly on the north side of the northern
street between m and n. Some are artificially raised
suggesting that, as today, this area and that to the
north, were prone to flooding. Setthement remains,
now represented by amorphous scarps on the eastern
side of this street, have been largely destroyed by the
construction of the later water meadows. However, a
rectangular hollow, clearly the site of a former
building, is still visible at w.

The third element lies north of and behind the
properties near m, immediately adjacent to the river.
The Inclosure map of 1794 shows a ‘T-shaped feature
in this area (inset to Figure 3). The earthworks
comprise two platforms. The larger of the two (r) is
partly surrounded by a ditch 10m wide and up to
1.3m deep which encloses a trapezoidal area
measuring 35m by 25m; there are no visible internal
details and its southern limit is poorly defined. The
second platform, immediately to the north, is
smaller and consists of a rectilinear raised area about
0.4m in height which, on the north, has been
truncated by the river. To the west, on rising ground,
is a circular feature (s), 8m in diameter and 0.3m
high, enclosed by a slight ditch.

Two houses, each one with an enclosed garden
and outbuildings, are shown on an aerial photograph
taken in 1925 (NMRC SU1352/3). Their sites can
be identified on the earthwork survey: the first is a
rectangular structure situated on the former street
slightly to the west of m, the second is situated on the
raised ground at v with a small bridge over the
hollow way leading towards the farmstead. Map
evidence shows that the houses were abandoned
between 1925 and 1939, probably after the houses at
t were constructed.

80 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The Water Meadows

There are two water meadows at West Chisenbury.
The first is situated to the north of the settlement,
beyond the surveyed area to the north of 2, and
extends as far as the tithing boundary. This meadow
was not surveyed since the earthworks, apart from
the main drain, are very fragmentary. However, the
main drain is an important aid in understanding the
chronology of the settlement earthworks on the flood
plain since it was laid out through the settlement
remains and utilised the hollow way north of n.

‘The second meadow lies within the surveyed area
and is bounded by a stream on the east and the
hollow way (7) to the west. Although this area has
been degraded, probably by later agricultural activity,
enough of the system remains for an interpretation to
be made. The meadow slopes gently in a southerly
direction and is of herring-bone construction; it also
slopes from the east and west towards the centre. The
stream to the east is lower than the main carriage and
therefore did not form part of this meadow. Water
was supplied to the northern part of the meadow
from two sluices, while the southern part was
supplied from another sluice to the north of v. A
channel (y), possibly a former hollow way to the
settlement in this area, extends from the hollow way
m—n towards the river. This channel was probably the
main carrier for the water meadow in the south. The
principle of the water meadow and how it functioned
was similar to those described by Cowan further
south on the river Avon (Cowan 1982).

DISCUSSION

West Chisenbury was a detached tithing of
Netheravon until the late 19th century. It is conceiv-
able that this tithing may once have been part of a
much larger Saxon parochie centred on the royal vill
of Netheravon and included the neighbouring parish
of Enford. It has been noted in Hampshire that one
of the reasons for the break-up of a parochie was the
granting of land to the church, particularly in the
10th century (Hase 1994, 62). A similar process may
have occurred here when, in 934, King Aethelstan
granted to the monks of St Swithun’s priory in
Winchester 30 hides in Enford (Grundy 1919, 228);
this grant would have effectively left West
Chisenbury detached from Netheravon. A_ link
between Enford and Netheravon is evident in 1086
since Netheravon church owned land in East
Chisenbury in the parish of Enford (Thorn and
Thorn 1979, 56.2).

The present arrangement of earthworks and the
Domesday evidence appear to suggest that the
hamlet of West Chisenbury may have developed as a
regular, single row settlkement, bordering the north
side of the northern street. Alternatively, it may have
developed as a settlement between the two streets
that later expanded beyond the northern street. The
sizes of the properties are similar to those that were
excavated at Gomeldon in south Wiltshire (Musty
and Algar 1986) and also equate to the sizes of 13th-
century buildings given by Dyer (1994, 155). This
arrangement of small squat sub-rectangular plots is
the normal form of medieval settlement in the
Wessex chalklands, quite unlike the crofts and tofts
that are seen in some other parts of the country. The
settlement on the flood plain mirrors the layout of
that on the river terrace and presumably represents
later expansion, The area between the two streets,
together with the land to the south, may represent
the demesne messuage; barns and paddocks may
have occupied the lower land immediately to the east
on the flood plain. Within this area, the platform at g
is a likely location for the chapel of ease (see above).
This would imply, therefore, that the manorial
complex lay wholly within the bounds of the street
system, partly on the river terrace with the remainder
spilling onto the flood plain between m and n.

To the north, and separated from the main
settlement complex by an area which is susceptible
to flooding is a small paired moat-like site (r). Moats
in Wiltshire are uncommon, particularly on the
Chalk (Lewis 1994, 184), and the incidence of a
possible one here is therefore of particular signi-
ficance. The Ordnance Survey recorded a moat at
Knighton Farm, 7km south of West Chisenbury on
the river Avon, but it is more likely that this feature is
an angled hollow way. The interpretation of the site
at West Chisenbury is complicated by the use of the
channel to the north of r as the main drain for the
northern water meadow; whether this channel was
cut during the construction of the water meadow, or
whether it is using an existing feature, is unclear.
Moat sites fulfil a variety of functions but were
primarily manorial residences (Le Patourel and
Roberts 1978, 46). Whether the site here was for
habitation or for some other purpose is uncertain;
although it is very small, habitation is not unknown
on such small sites: at Willoughton in Lincolnshire,
for example, a moat 35m x 45m with an associated
fish pond was the site of the grange of an alien priory
(Everson et al. 1991, 218). The site may therefore be
another, or an alternative manorial complex —
possibly even the demesne messuage of John

WEST CHISENBURY: SETTLEMENT AND LAND USE

Brymmore documented in 1361, with the homestead
occupying one platform and a garden the other, and
with the dovecote lying to the west. Alternatively, it
may represent a monastic holding, possibly that of the
Augustinian estate recorded here in 1227. It appears
to lie on the edge of the main settlement complex,
indicating that it is presumably later in date than the
main part of the hamlet. A further possibility is that it
could represent a post-medieval ornamental water
garden feature associated with the river. The
development of ornamental gardens in this area is not
uncommon: across the river at East Chisenbury, for
example, a leat was constructed in the 17th century in
order to create a water garden in the newly emparked
estate. A similar process may have taken place at West
Chisenbury at this time when considerable water
management of the meadows was being undertaken.

During the post-medieval period there were a
number of agricultural improvements, the most
important of which in this area was the floating of the
water meadows. The importance of water meadows
has been described in detail elsewhere (Davis 1811;
Kerridge 1953; Bowie 1987). Although the date of
their establishment along the river Avon in unknown,
it is likely to have been sometime in the 17th century
when they were also in use along the Wylye valley. Two
types of water meadow construction have been
recognised along the river Avon north of Amesbury.
They were either constructed with side carriers that
were set at right angles to the main drain, or they were
herring-bone shaped with perhaps additional
subsidiary carriers. It is unlikely that these two types
represent a chronological development even though
the herring-bone type appears less well developed and
more haphazard; rather, they were constructed to suit
the topography of the particular area. For example, the
northern meadow at West Chisenbury is long and
narrow with the main drain in the centre, and parallel
to the river. This provided enough space for the
construction of the side carriers at right angles to the
ditch. In contrast, herring-bone meadows appear to
have been constructed in less uniform topographies
such as in areas of river meander. An unusual feature
of the water meadow to the east of the settlement is
that it overlies some abandoned properties and also
utilises the hollow way as a main carrier. This indicates
a date, possibly in the 17th century, by which time the
eastern part of the settlement was abandoned.

The stream which formed the eastern boundary
of the water meadow did not form part of the
meadow and appears to be more substantial on the
1794 map than it is today. It is conceivable,
therefore, that this stream may have been a leat for a

water mill, possibly the one recorded in the 11th and
13th centuries at East Chisenbury (see above). The
later site of the East Chisenbury mill was at the river
crossing between the two hamlets.

Beyond the settlement, the tithing landscape
mirrors that of other chalk downland riverine estates
with meadow on the flood plain, arable close to the
settlement and extending on to the downs, and
pasture beyond. This division of land resources has
been dated to at least the later Anglo-Saxon period
(Hooke 1988, 140). However, a notable difference
in this area in comparison with others on the Wessex
chalkland is that the downs have not been cultivated
in modern times and the pattern of ancient and later
fields is still largely extant.

Prehistoric and Romano-British fields covered
much of Lavington Way Down and Chisenbury
Down, although in the areas on the lower slopes near
the settlement that have been regularly cultivated
since the medieval period, little trace of these fields
survives. Ridge and furrow is also evident on parts of
Chisenbury Down, beyond the area of ‘permanent’
arable, and extends west over much of Thornham
Down and Charlton Down. These fields are
arranged in a rather haphazard fashion, varying in
area from 300-500 acres, with 5—7m between ridges.
They have clearly been superimposed on the ancient
fields and do not respect their layout; they also
encroach on _ parish and_ tithing boundaries
(Crawford 1935, 91). Since the ridge and furrow has
not destroyed the ancient fields it is likely that it
represents temporary intakes during periods of stress
on land resources, possibly during the Napoleonic
Wars.

The plan of West Chisenbury shows a large,
roughly rectangular area, probably the manorial
area, with the modern farm set in the middle and
large scarped closes to the south which were clearly
once paddocks. Further traces of paddocks are
evident to the north-west of the farm with
earthworks to the north and north-east, possibly
including the site of the chapel. To the north of this
rectangular area is a regular single row settlement,
containing perhaps four properties. Later settlement
expansion occurred over the flood plain with a
curving hollow way leading to the river-crossing.

West Chisenbury may have originated as a Saxon
settlement on the site of the later manor and lying
directly opposite the ancient river crossing. This
developed into the manorial area with the remainder
of the hamlet to the north, probably during the late
Saxon or 11th—12th centuries, causing the disloca-
tion of the track to the river-crossing and thus the

82 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

development of the curving hollow way or street
from the settlkement to the river crossing. Later,
possibly during the 12th or 13th centuries when land
was granted to various individuals, the settlement
expanded in the only direction possible, over on to
the flood plain, since the manor lay to the south and
the arable fields to the north and west. What effect
the Black Death had at West Chisenbury 1s
unknown; however, it was not deserted at this time
since there was a land transaction in the mid 14th
century. The settlement probably began to shrink in
the mid 16th century and continued into the 17th
century with the demise of the chapel, the manorial
barn recorded as being in decay, and _ the
construction of the water meadows over former
properties. In the early 18th century the settlement
became a single farm with a few isolated cottages.
Although this sequence of development is paralleled
in many other places in England, the water
management on the flood plain is unusual, first by
canalising the river to restrict its course and secondly
by the creation of a water meadow over an area of
former settlement.

Acknowledgements. | would like to express my gratitude to David
McOmish of RCHME, who assisted in the survey and gave me
much encouragement during the preparation of this paper, and to
colleagues at RCHME who commented on the text, particularly
Chris Dunn who edited the text and gave much valuable advice.
The illustrations were drawn by Deborah Cunliffe. The Record
Offices in Wiltshire and Hampshire assisted by making the
documents and maps available. Finally I would like to thank the
Ministry of Defence Land Service and the farmer, Mr Baxter,
without whose co-operation the fieldwork would have been
impossible.

Survey Methods and Archive. The site was surveyed at a scale of
1:1000 using a Wild TC 2000 theodolite and intrigated EDM for
the main control framework; the archaeological detail was recorded
using taped offsets. The archival report and survey diagram (NAR
No. SUI5SW 44) have been deposited in the National Monuments
Record at the National Monuments Record Centre, Kemble Drive,
Swindon, SN2 2GZ and are open to the public for consultation.

REFERENCES

Unpublished Sources

HRO Hampshire Record Office map, MP40

MoD Schedule of sale of West Chisenbury Farm, 1897
(held at the office of the Defence Land Service in
Durrington)

NMRC Aerial Photograph SU1353/3, dated 2.6.25

West Chisenbury Inclosure Map 1794

Published Sources

ANNABLE, F.K., 1967 ‘Excavations and Fieldwork in
Wiltshire, 1966’, WAM 62, 124-131

ANNABLE, F.K., 1968 ‘Accessions to the Museum’, WAM
62, 107-115

BENNETT, THE REVD CANON, 1887 “The Orders of
Shrewton’, WAM 23, 33-39

BERESFORD, M.W., 1959 ‘Poll Tax Payers of 1377’ in E.
Crittall, (ed.), VCH Wilts 4, London, 304-313

BODINGTON, E.J., 1919 “The Church Survey in Wilts,
1649-50’ WAM 40, 297-317

BONNEY, D.J., 1966 ‘Pagan Saxon Burials and Boundaries
in Wiltshire’, WAM 61, 25-30

BOWIE, G.G.S., 1987 ‘Water Meadows in Wessex — A Re-
evaluation for the Period 1640-1850’, Agricultural
History Review 35, 151-158

CHETTLE, H.F., AND KIRBY, J.L., 1956 ‘Priory of Maiden
Bradley’, in E. Crittall and R.B. Pugh (eds.), VCH Wilts
3, London, 295-302

CHIBNALL, M., 1951 Select Documents of the English Lands of
the Abbey of Bec, London

COSSONS, A., 1959 ‘Roads’ in E. Crittall (ed.), VCH Wilts
4, London, 354-271

COWAN, M., 1982 Floated Water Meadows in the Salisbury
Area, South Wiltshire Industrial Archaeological Society
Monograph 9

CRAWFORD, 0.G.S, 1935 ‘Superimposed Cultivation-
Systems’ in ‘Notes and News’, Antiquity 9, 89-91

CROWLEY, D.A., (1989 The Wiltshire Tax List of 1332,
Wiltshire Record Society (WRS) Volume 45, Trowbridge

CUNNINGTON, M.E., 1930 ‘Saxon Burials at West
Chisenbury’, WAM 45, 84

DAVIS, T., 1811 General View of the Agriculture of Wiltshire,
London

DYER, C., 1991 Hanbury: Settlement and Society in a
Woodland Landscape, Leicester University Press

DYER, C., 1994 Everyday Life in Medieval England, London

ELRINGTON, C.R., 1974 Abstract of Feet of Fines Relating to
Wiltshire for the Reign of Edward III, WRS, Vol. 29,
Devizes

EVERSON, P., TAYLOR, C., AND DUNN, C., 1991 Change and
Continuity. Rural Settlement in North-West Lincolnshire,
RCHME, London

FRY, G.S., AND E.A., 1901 Abstracts of Wiltshire Inquisitiones
Post Mortem, Charles I, 1625-1649, London

FRY, E.A., 1908 Abstracts of Wiltshire Inquisitiones Post
Mortem, Henry III, Edward I and Edward II, 1242-1326,
London

GRUNDY, G.B., 1919 “The Saxon Land Charters of
Wiltshire’, Archaeological fournal 76, 143-301

HASE, P.H., 1994 “The Church in the Wessex Heartlands’ in
M. Aston and C. Lewis (eds.), The Medieval Landscape
of Wessex, Oxford, 47-81

HOOKE, D., 1988 ‘Regional Variation in Southern and
Central England in the Anglo-Saxon Period and its
Relationship to Land Units and Settlement’ in D.
Hooke (ed.), Anglo-Saxon Settlements, Oxford, 123-151

JACKSON, THE REV. CANON, 1867 ‘Ancient Chapels in the
County of Wiltshire’, WAM 10, 253-322

KERRIDGE, E., 1953 “The Floating of the Wiltshire Water-
meadows’, WAM 55, 105-118

LE PATOUREL, H.E., AND ROBERTS, B.K., 1978 “The
Significance of Moated Sites’ in A. Aberg (ed.),
Medieval Moated Sites, CBA Res. Rept. 17, 46-55

LEWIS, C., 1994 ‘Patterns and Processes in the Medieval
Settlement of Wiltshire’ in M. Aston and C. Lewis
(eds.), The Medieval Landscape of Wessex, 171-193

MUSTY, J., AND ALGAR, D., 1986 ‘Excavations at the
Deserted Medieval Village of Gomeldon, near
Salisbury’, WAM 80, 127-169

PERRY, P.J., 1974 British Farming in the Great Depression,
1870-1914, Newton Abbot

WEST CHISENBURY: SETTLEMENT AND LAND USE

PUGH, R.B., 1939 Abstracts of Feet of Fines Relating to
Wiltshire for the Reigns of Edward I and Edward II, WRS,
Vo}. I, Devizes

RCHME (a), in preparation: Wiltshire — The Field
Archaeology of a Wessex Landscape. Part 1: South
Wiltshire

RCHME (b), in preparation: Wiltshire — The Field
Archaeology of a Wessex Landscape. Part 2: Salisbury
Plain Training Area

SAWYER, P.H., 1968 Anglo-Saxon Charters: An Annotated
List and Bibliography, London

STEVENSON, J.H., 1980 ‘Netheravon’ in D.A. Crowley
(ed.), VCH Wilts 11, London, 165-181

STOKES, E., 1914 Abstracts of Wiltshire Inquisitiones Post
Mortem Edward II AD 1327-1377, British Record
Society, London

TATE, W.E., 1945 ‘A Handlist of Wiltshire Enclosure Acts
and Awards’, WAM 51, 127-173

THORN, C., AND THORN, F., 1979 Domesday Book.
Wiltshire, London

Appendix

The table below provides an indication of land-use and is an attempt to dissect the workings of the tithing from
the Domesday survey until its acquisition by the Ministry of Defence (MOD) in 1897.

Arable
1086 13 plough
150a.(622 ha)*
1361 240a.(100ha)
1048 the capital
manor was
c.246a.
1796 518a.(216ha)
1898 505a.(210ha)

Pasture/Down Meadow
4 x 2 furlong 6a.

(400 sheep) 12a.(demesne)

447a.(186ha) 5a.(demesne)
469a.(195ha) 26a.

* This presumes the demesne plough cultivated 100 acres and a tenant plough cultivated 50 acres (Dyer 1991, 67, note 34).

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 84-87

Royal Itineraries and Medieval Routes

by NORMAN HIDDEN

Edited records have been published of itineraries made by the monarch and his court in connection with

the royal dispensation of justice during the reigns of Henry I and Edward III. Various factors

determined the royal routes taken and in particular the stopping places chosen. Additions to, or

deletions from, the list of places visited may have occasional significance in indicating a township’s

development or decline; and apparent changes of route may suggest an extension or improvement to

certain roads. The routes covered in this brief examination mainly are those between royal residences in

or near the capital to Wiltshire, with special reference to Salisbury, Marlborough and Hungerford.

One of the earliest ways in which the Plantagenets
dispensed justice throughout their kingdom was by
the monarch himself travelling about the country
with his court (Curia Regis), both taking local
counsel and settling important legal cases in the
region visited. Records of these royal itineraries have
been carefully compiled! and it is possible, from the
reign of Henry I onwards, to ascertain most of the
routes journeyed. They included royal visits to
Wiltshire, particularly to Marlborough and Salisbury,
as well as to other places in the region which lay en
route or within reach. As early as 1100 Henry I made
overnight stops at Salisbury and Marlborough. In
1106 a ‘return’ journey from the west proceeded
from Alveston in Gloucestershire via Marlborough,
Ludgershall, and Salisbury, and thence to the royal
residence at Windsor.*

Overnight stoppage at a particular place could be
dictated by a number of factors. These included
whether any business had to be conducted there, and
whether suitable accommodation existed in the
locality where the monarch might lodge and be
entertained. The first choice of royal stopping places
was obviously one of the king’s own residences; hence
the frequent mention of Ludgershall, Marlborough
and Clarendon Park near Salisbury. A second choice
might be an ecclesiastical house (e.g. Sandelford
Priory near Newbury) or the substantial residence of
a local landowner. Failing this, there were the local
resources offered by a township with its inns, taverns
and other miscellaneous accommodation. A major
factor, however, was the distance which the monarch

1. Itinerary of Henry I (ed.) W. Farrer (Oxford, 1920); Court,
Household and Itinerary of Henry II (ed.) R.W. Eyton (London
1878); Itinerary of King John (ed.) T.D. Hardy (London 1835);
‘Itineraries of Henry III’, Theodore Craib (typescript, P.R.O.
1923); Itmmerary of Edward I (ed.) E.W. Safford (List & Index
Society, vols. 103, 132 and 135 (1925)); Itinerary of Edward II
and his Household (ed.) Elizabeth M. Hallam (List & Index
Society, vol. 211 (1984)).

and his accompanying household might be able to
accomplish in a day’s ride, taking into consideration
both the time of the year and the hours of daylight.
Where the king had finished his business and was
returning homewards on the last stage of his journey,
he could ride well ahead of his commissariat; and
with his time unconsumed by any further business
en route, he might cover a good fifty miles or more,
as Richard I did from Marlborough in 1189. Having
stayed the night of 29 August there, he arrived back
at his residence at Windsor the next evening.* The
frequent number of instances of a day’s journey from
one stopping place at Marlborough to another at
Newbury (or vice versa) shows not only the
importance of these two towns, but also that the
distance between them, somewhat less than twenty
miles, on a more or less flat road was reasonable for a
one-day journey with all the royal baggage involved.

In 1222 Henry HI set out from Woodstock on 28
December, stayed two nights at Oxford, spent New
Year’s Eve at Hungerford, and arrived the following
day at Marlborough.* His decision to stop at
Hungerford rather than travel another eight miles or
so seems likely to have been determined solely by the
logistics of the situation. He had ridden from Oxford
that morning, which was a good thirty miles away,
with a steep ascent over the Downs, and may have
felt, in the short daylight of mid-winter, that it would
be unwise to push on to Marlborough either through
the forest of Savernake or along the narrow road via
Ramsbury, north of the river Kennet, that same
evening.

* Unless otherwise indicated, specific references to royal itinerary
details (marked *) have been derived from the appropriate
volume listed above.

ROYAL ITINERARIES AND MEDIEVAL ROUTES

He could not have found Hungerford a con-
venient stopping place, for there is no record of his
stopping there on any subsequent occasion, even
when travelling homewards from Marlborough via
Newbury (e.g. in 1223, 1225 and 1226), nor on
those occasions when he travelled from Marlborough
to Reading (e.g. 1234, 1235 and 1236) and stayed
the night at Sandelford Priory.* Nevertheless, the
stop at Hungerford is the first mention in any reign of
this town’s location along the royal routes and is good
evidence that by this date the town had come into
substantial being.

In a later journey in 1241 Henry III, returning
homewards, passed through Hungerford again. That
he may have experienced an involuntary and probably
infuriating hold-up there, would seem to be suggested
by a mandate he dictated on his arrival at Sandelford,
expressing his dissatisfaction with the state of the
bridge over the river Kennet, and slapping a fine of 5
marks on the township (villatam) for the inadequacy
of its bridge.* The phrase used (‘pro defecto pontis’)
presumably refers to the bridge’s ill-repair, but may
mean the absence of a bridge at all, entrance to the
town itself having traditionally been across a ford in
the river. The issue is confused, since the ford was
across what was then known as ‘the Bedwyn stream’
and is now known as the river Dun, a tributary of the
Kennet, at a point near its confluence with the major
river. The Kennet would be very much more likely to
have needed a bridge to cross by, and there is evidence
of one at Eddington at this date which crossed to
Charnham Street. The responsibility for its upkeep
lay with the lord of the manor of Hidden-cum-

_ Eddington? and not with the town of Hungerford, for
neither Hidden-cum-Eddington in Berkshire nor
Charnham Street in Wiltshire were incorporated in
_ the town and for centuries fiercely maintained their
| independence.* Was Henry III’s irritation in
connection with the bridge, or lack of it, due to his
knowledge that the lord of both Hidden and
_ Hungerford was none other than his former friend but
now adversary Simon de Montfort?

Later royal itineraries to Wiltshire resulted in
overnight stops at Hungerford without complaint.
' Thus Edward I stopped in the town on no less than
| three occasions: in 1286, 1289 and in 1302.* His
}son Edward II stopped there in 1308.* As his
| journey was merely from Newbury to Marlborough,
) which he would have done comfortably in a day, it

_ would seem that some business was awaiting him at
|

| 2. Calendar of Close Rolls 1237-1241, p. 375.

| 3. S.F. Wigram (ed.), The Cartulary of St. Frideswide’s (Oxford
1895), ii, p. 337.

Hungerford. His next visit was in 1320 en route to
Marlborough, another short day’s journey. Disap-
pointingly, none of these visits by Edward I or
Edward II has been chronicled by the Victoria
County History for Berkshire, which is able to record
only that Edward III passed through the town in
1331 and again early the next year.’

Edward I’s journeys in this area are particularly
interesting because whereas the itineraries of his
father Henry HI had been mainly along the old-
established route from London to the west on which
Hungerford, and more particularly Charnham
Street, lay, Edward I made use of Hungerford also as
a stopping off place on journeys to or from the south.
Thus in 1286 he came from Amesbury to Upavon
and thence to Hungerford.* Some specific business
clearly brought him there, for he then turned west to
Marlborough (which he could have reached direct
from Upavon), staying there three nights before
returning eastward to Reading. Similarly, in 1289 he
journeyed from Reading to Hungerford, and thence
by a direct north-south route to Amesbury and
Clarendon Park near Salisbury.* It would thus seem
that a southern route to Salisbury via Hungerford
had come into regular use by the latter part of the
13th century.

It is noticeable that in the previous century
royal itineraries had shown a more complicated
north-south route, such as via Marlborough and
Ludgershall, to Salisbury, a route which was con-
siderably more circuitous. In 1189 Richard I had
returned to Windsor from Winchester via Salisbury
and Marlborough, and in 1203 King John came back
home from Portsmouth via Burbage, Marlborough
and Newbury.* Evidently, the recognised north—
south route was via Marlborough at this date. Of
course Marlborough was an important centre; it had
royal connections and doubtless there was always
business to be done there, but had there been an
alternative suitable route, it is unlikely that travel-
worn monarchs would not have made use of it
whenever possible, especially on their homeward
journey when they would be unfettered by the need
to conduct local business.

It will be noted that the routeing of royal
itineraries through Hungerford in a north-south
direction occurs at a date by which the town had been
replanned, the little village higgled around the parish
church having been superseded by a long wide street
(present day Bridge Street and High Street) a quarter

4. See, for example, P.R.O. ASSI 2/1, f. 188.
5. VC.H. Berks, vol. iv, p. 185.

86 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

of a mile to the east with spacious burgage sites on
either side. This road pointed straight to Salisbury
and in later years was often referred to as ‘Salisbury
Street’. At the junction where it crossed an older road
leading from Kintbury and going westwards past the
church, the town held its market. Historians have pf Mood bi
been unable to date this development precisely, Fee Soa
cautiously offering an outside time range of between
1170 and 1296.° The evidence of the town’s appear-
ance (or non-appearance) on the royal routes fits well
into this. Henry III’s stop overnight in 1222 confirms
that the town was in existence then, and Edward I’s
use of the north-south route in 1286 suggests that a
new road system had developed which could provide
an impetus for a successful market prior to the
earliest documentary evidence of such a market in
1296.’

King Edward’s journey from Downton (or
possibly Dinton) and Salisbury to Hungerford in late
January 1286, via Amesbury and Upavon, is revealing
in other ways too. The wardrobe waggon train took
four days to make this winter journey, of about 40
miles, struggling over steep hills. The King and his
household went on ahead, reaching Amesbury and
Upavon on 20 January and Hungerford the next day.
Here the royal household divided, the King, the
Queen and their immediate attendants going on to
stay at Marlborough with the Queen Mother (22-24
January). The rest of the household remained at
Hungerford during the King’s visit to Marlborough,
and were joined there by the wardrobe waggon. The
King then returned to London via Denford and
Crookham on the 25th. The location of Crookham
suggests that the route taken homewards was along
the old Kintbury road, south of the river Kennet. In a
further journey in 1289 Edward I travelled from
Newbury to Marlborough via Hampstead Marshall
and Hungerford. There can be no doubt that to do so
he would have used the route to Hungerford via
Kintbury. A map centuries later® refers to the road
from Hungerford via Kintbury as ‘the old and great
Market road’, suggesting a function that had become
eclipsed in the 18th century by the Bath road north of
the river, which in its turn is today overshadowed by
the M4 motorway.

In considering the information provided by the
royal itineraries which concerns routes and roads, the

Greville C. Astill, He ; ene Ce Hungerford to Marlborough section of the London to
6. Greville C. Astill, Historic Towns in Berkshire (Berks Arch. Ctee. . : ‘ “Ws
Publication no, 2 (Reading 1978);p..29. Bristol road (scale in miles). Reproduced from J. Ogilby’s
7. VC.H. Berks, iv, p. 187. Britannia (London 1675) by kind permission of the

8. Berks R.O. D/EB P1: sketch map by Wm. Watts c.1750. Guildhall Library, Corporation of London

ROYAL ITINERARIES AND MEDIEVAL ROUTES

human factor should be borne in mind. Royal visits
must have caused great excitement in a small market
town, with their glamour, bustle and activity; they
must have caused also a good deal of disruption,
some anxiety and probably not a little fear. On all
these journeys, demands were made on _ local
inhabitants by a small advance guard of royal
officials purchasing or requisitioning food and
lodgings for the arriving party and fodder and
stabling for their horses. Some idea of the sweat and
effort which went into making the royal governance
work successfully may be seen in the accounts of the
royal Wardrobe and Household.’ Thus, payment is
made to porter Hicke for one cart drawn by five
horses, which took four days to carry the wardrobe
equipment from Downton in Wiltshire’ to
Hungerford, a journey of under forty miles. This
represents an average speed over the hills of less than
ten miles a day. The whole journey from Exeter to
Hungerford may have been some 130 miles and
porter Hicke was paid for eleven days on the road.
Local purchases on this journey in 1285-6 included

9. B.F. and C.R. Byerley (eds.), Records of the Wardrobe and
Household: 1285-6, (HMSO 1977), p. 15.

10lbs of grain at 4d. a pound, and 1d. for a sack to
contain the purchase; 16lbs of vetch at 1d. a pound;
5lbs of liquorice at 4d. a pound; 1 inkwell and a
supply of ink for the scribe (total cost 4d.); 2d. for
some spice; and another 2d. for some white powder
required by porter Hicke. No one would have been
busier than the royal clerk of the court for he would
be visiting, in advance of the king, towns within a
radius of ten miles, issuing summons to local
officials, empanelling juries, obtaining information
about breach of the assizes and the sale of sub-
quality goods.!°

The records of royal itineraries on these court
journeys are valuable accounts of the administration
of justice coram rege, and also throw occasional
fascinating illuminations on the social history of the
time. In addition, as this article tries to show, they
may provide an extra means of dating the rise or fall
in importance of small towns which lay along their
route, as well as being an indicator of the possible
changes in east-west and north-south routes leading
into Wiltshire and the south west.

10. Ibid., p. xxv.

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 88-98

Wiltshire Deer Parks: An Introductory Survey

by KENNETH WATTS

WG. Hoskins suggested that ‘the reconstruction of medieval deer parks and their boundaries is one of

the many useful tasks awaiting the fieldworker with patience and good local knowledge’.! This paper
provides an introduction to the subject of the early deer parks and attempts to locate all known and

probable examples in Wiltshire. It is written in the hope that local historians will be encouraged to make

use of their local knowledge and contacts to research their local parks and add to our knowledge of this

worthwhile, but neglected, subject.

INTRODUCTION

Although medieval Wiltshire was heavily forested
and contained many deer parks, these early parks
have generally been neglected as a subject for
modern research. Parks were significant elements in
medieval, Tudor and Stuart landscapes and society,
providing sport, an important means of social
distancing, skins, and fresh meat at all seasons at a
time when it was not readily available in winter.

Deer parks normally occupied large areas of
undeveloped land and have consequently tended to
be lost as pressures of increasing population led to
parks being entirely disparked or reduced in favour
of agriculture. Many ‘home parks’ around country
houses were converted into ornamental landscaped
gardens in the 18th century. Evidence for the former
existence of many early deer parks survives in early
document rolls and in place-names. Occasionally
signs may be seen on the ground in old field
boundaries or in eroded linear banks and ditches
surviving from the park pales which surrounded
early parks. The aim of this paper is to establish the
locations of all the Wiltshire deer parks. It must be
emphasised that its subject is the pre-1700 deer
parks, the functional parks created to provide sport
and venison. It is not concerned with later ornamental
landscaped parks, except where they superseded and
occupied sites of earlier deer parks.

The word ‘park’ is used today in a wide variety
of applications, for example in car park and for
places of public recreation. During the 18th century
it was used for the landscaped ‘parkland’ which
surrounded country houses, but in the early
medieval period a park was a fenced area devoted to

1. W.G. Hoskins, The Making of the English Landscape (London
1955), p.76.

the keeping of deer. The word derives from the Old
English pearroc meaning enclosure, the essence of a
park being that, unlike a forest or a chase, it was
physically separated from the surrounding country-
side by a fence or barrier. In the early medieval
period it was usually parco or parcum, and early
references in the Close Rolls include:

1230: parco de Wutton at Wootton Bassett (33);

1242: parco suum de Corsham at Corsham (16);

1253: parcum suum de Knoel at East Knoyle (80);
and

1256: in parco regis de Divisas at Devizes (30).?

During the Tudor and Stuart periods the word
became parke.

A park may be defined as an area of at least 30
acres enclosed to retain deer. /- warren was an
enclosure for small game such as rabbits and hares,
and was inferior to a park. ‘Rights of free warren’ were
rights to take small game, but not deer which were
royal animals. Owners of deer parks were at times
required to breed horses in the national interest and
the term ‘horse park’ was occasionally used. Early
bishops and abbots were often men of action and
enthusiastic hunters, and monasteries and abbeys
frequently had deer parks associated with them. In the
early medieval period deer parks were created and
owned by bishops at Ramsbury (43), Downton (60),
East Knoyle (80), and Potterne (71), and by abbots at
Malmesbury (3), Garsdon (4), and Stanton St
Quintin (10). After the Dissolution the deer parks
attached to former monastic establishments provided
admirable areas for their new secular owners to
convert into landscaped grounds.

2. Numbers in brackets after references to individual parks refer to
the numbering in the Schedule and Distribution Map appended
to this paper.

WILTSHIRE DEER PARKS

ORIGIN OF DEER PARKS

The Romans hunted deer in England and contained
them in enclosures called haye which became the
haga of the Saxon Charters. Isaac Taylor suggested
that hay or haigh was of similar meaning to park,
being ‘a place surrounded by a hedge... usually an
enclosure for purposes of the chase’.? The haga of
the Anglo-Saxons were often palisaded enclosures,
and it is likely that the first true deer parks were
those of Kings Aethelstan and Cnut. Anglo-Saxon
references to deer-falds are also known, fald being Old
English for a staked-off enclosure. Many deer parks
were created by kings, nobles and prelates in the
12th century, and even more in the 13th and 14th
centuries when most great men owned at least one
park or chase. Later lords of the manor aspired to
owning a park, such parks being often created on
waste land away from the manor house at the
extremity of the parish, for by this time there was
growing pressure on productive land from an
increasing population. Deer parks continued to be
made through the Tudor period; although most
enclosure was to create sheep pasture and increase
arable, some was to continue the medieval practice
of emparking to accommodate deer. The Com-
mission appointed in 1548 to enquire into enclosure
was directed to investigate engrossing, enclosing and
emparking, and parks for deer continued to be made
into Stuart times.

FORESTS AND CHASES

A forest was a royal hunting-ground, not necessarily

wooded, which was unfenced and administered by
, royal officials according to a severe code of Forest
| Laws which protected the deer as royal beasts
| belonging to the monarch. The forest was effectively
| an area over which Forest Law was applied in addition
| to common law. Some forests included parks from
_ which deer were released to be hunted. Sometimes a
| subject was licenced to create a park within an area
| subject to Forest Law. The Norman and Angevin
_ kings constantly extended their forests, and this
| practice was a major cause of friction between King
| John and his barons in 1215. The reason for
| extending forests was not for hunting considerations
alone. Mutilation or death, the early punishments for
offences, had gradually been superseded by a system
of fines which provided a lucrative source of revenue.

3. I. Taylor, Words and Places (London 1911 edn.), p. 102. Taylor,
who wrote in 1864, may be regarded as suspect by modern
etymologists, but The Hague in Holland was formerly a hunting

An open area used for hunting deer by a subject,
who was normally a great noble to enjoy such
rights, was called a chase. Examples in Wiltshire
are Aldbourne, Cranborne, and Vernditch Chases,
together with the former Collingbourne Chase. If a
chase was acquired by the Crown it became a forest,
as technically happened at Aldbourne Chase in 1399
when its owner, the son of John of Gaunt, usurped
the throne as Henry IV.

Forests formerly covered a large proportion of
Wiltshire. During medieval times the English
countryside consisted of islands of cultivation within
the general waste, and Speed’s map (1610) shows
many forests then surviving. Both Elizabeth I and
James I leased off great areas of forest, but Charles I
attempted to reclaim them and it was this fact,
together with his attempts to re-afforest large areas
and revive Forest Law, which contributed to the
Civil War. :

Forest Law was imported by the Normans and
enforced at special Forest Courts. It was a code
of regulations which promoted the welfare of deer
as royal beasts to the detriment of the local people.
By the late 12th century, when Forest Law was at
its peak prior to crumbling under Richard I and
John, it is believed to have applied over about one
third of England. In the 13th century it was
administered over nine Wiltshire forests: Selwood,
Melksham, Chippenham, Braydon, Savernake,
Chute, Clarendon, Grovely and Melchet.* Forest
Law created great tension between monarch and
subjects because great areas were being preserved for
deer at a time when more land was needed for
agriculture to feed an increasing population. It was
terminated by the 1640 Act for the Limitation of
Forests which was passed to counter Charles I’s
attempts to revive Forest Law and _ re-afforest
substantial areas.

LICENCING AND SIZE OF PARKS

A subject wishing to create a park in or near a royal
forest required a licence to impark because parks
relied on attracting deer — which despite being wild
were regarded as the king’s beasts — through deer-
leaps or /ypiatts (leap-gates) which allowed deer to
enter but not escape from the park. A proposal to
include a deer-leap was often included in the licence

lodge of the Princes of Orange.
4. Melchet Forest was transferred to Hampshire by boundary
change in 1895.

90 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

for a park. An example is the 1267 licence for deer-
leaps at the parks at Wootton Bassett (33 and 34):

Grant for life to Philip Basset that he may have
a deer-leap at his new park under his town of
Wotton and another at his old park under his
manor of La Fasterne, within the metes of the
forest of Bradene, so that if any deer enter the
parks by the said deer-leaps, they shall remain
to him.’

Areas of deer parks vary considerably. Licences
suggest that they covered a minimum area of 30
acres, such small parks being used for stockholding.
A hunting park needed to be at least 100 acres, and
a large park would contain 500 acres or more.° Some
very large parks such as Clarendon (59) were several
miles around their perimeter. A very small park of 30
acres could be contained within a simple circular
area 440 yards in diameter, a smal! to medium park
within a circle 790 yards diameter, and a large 500-
acre park within a 1750 yards diameter circle. Parks
were of course invariably irregular in shape as
dictated by terrain, but such rationalisation into
circles is a simplification which is useful as a rough
guide to areas to be anticipated.

NUMBERS OF PARKS

Many more parks existed than are shown on the
early county maps. Saxton’s map of Dorset showed
eleven parks, but in their investigation into Dorset
deer parks Professor Cantor and J.D. Wilson
immediately found 44 parks and later increased their
findings to 49 definite and 42 possibles.’ Speed’s
map of Wiltshire has 22 parks, but Professor
Cantor’s gazetteer in The Medieval Deer Parks of
England (1983) lists 42 parks (plus one possible) in
Wiltshire. Research by the present writer has
unearthed evidence for the existence of over 90
Wiltshire parks. The parks shown by Saxton and
Speed (by a ring fence symbol) were those which
survived into the Elizabethan and Jacobean periods,
but there can be no doubt that by the 16th century
very many medieval deer parks had been disparked.
The Schedule and Distribution Map appended to
this paper show 90 parks, some of them multiples.
If it is assumed that each was an average park of
200 acres, the total area of Wiltshire parks amounted

5. Calendar of Patent Rolls, 1266-1272, p. 177.

6. Imperial measure has been used throughout this paper as all
historical references used imperial units.

7. Proceedings of the Dorset Natural History and Archaeological

to 18,000 acres in a county of about 864,000 acres,
representing about 2.1 per cent of the county area.
It has been estimated that at the beginning of the
14th century there were about 3,200 parks in
England covering about 2 per cent of the country,®
an assessment which is consistent with the above
calculation for Wiltshire which was arrived at inde-
pendently. Today deer parks are seldom recognised
and their boundaries are not much in evidence.
Consequently, in Wiltshire only one park pale
(Devizes Old Park: 30) is shown on the 1:50 000
map, and only one more (Ramsbury: 43) at the
larger scale of 1:25 000.

CHARACTERISTICS OF PARKS

The topography required for a deer park was a
mixture of woodland and pasture, the former to
harbour the deer, the latter to provide open areas
over which they could be hunted after being
unharboured with all the traditional rites of venery.
A secondary use of parks was to grow timber for
building, and some park trees were allowed to grow
uncoppiced into large timber trees with clear trunks
and their lower foliage was browsed as high as the
deer could reach. Such trees were typical of
parkland. A further feature was the keeper’s lodge
which was often provided, particularly when the
park was an ‘out-park’ remote from its home
establishment. Lodges were often moated, the moats
being stocked with fish for the table, as were the
ponds and lakes frequently found in deer parks. In
the early 1260s the Constable of Marlborough
Castle was twice instructed to stock the ponds at
Elcombe Park (38) with bream.”

Very early deer parks were no doubt surrounded
by simple high fences to enclose the deer, but at the
time early deer parks were being created a long
tradition existed of delineating boundaries with
banks and ditches, from the Bronze Age through the
Iron Age and Romano-British periods to the Dark
Ages and into early medieval times. There was a
similar tradition of building stockade fences for
various forms of enclosure. These two elements of
linear bank and ditch and stockade fencing soon
came to be combined in the park pales which
enclosed early deer parks (see Figure 1 which
illustrates the development of the park pale). In early

Society, Vols. 83-100 (1961-1978).

8. J.M. Steane, The Archaeology of Medieval England and Wales
(Beckenham 1985), p. 168.

9. Calendar of Close Rolls 1261-1264, pp.7 and 321.

WILTSHIRE DEER PARKS

STAGE ONE

Early deer parks were probably
surrounded by stockade fences of
oak logs 8 or 9 feet (2.4m or 2.7m)
high which prevented deer from
entering or leaving the park. Such
‘fences required huge amounts of
timber.

STAGE TWO

It was soon realised that economy in
fence height could be achieved by
providing a linear ditch inside a
lower fence, probably about 5 feet
(1.5m) high. This had the added
advantage of allowing deer to enter
but not leave the park for its full
perimeter.

STAGE THREE

Later, further economies were made
by abandoning stockade fences and
providing higher fencing of cleft oak
post and rail. Ditches were made
discontinuous to save labour, with
‘pitfalls’ provided at intervals
opposite short lengths of lower
fencing. These ‘deer-leaps’ allowed
deer to enter but not leave the park
at these points.

Development of the Park Pale KW 1994

Pale fence
(stockade) 4

Ground level

Pale fence
(stockade)

i General pale height

Post and rail fence |
lower at deer-leap

——. —_. —____ —__. —.

‘Pitfall’

Park side Y Outside.

Figure 1

92 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

medieval times the stockade consisted of narrowly-
spaced oak trunks or split oak palings with tops
usually pointed to prevent water standing on the end
grain. Later, to economise on timber, the stockade
construction was abandoned and cleft oak post and
rail fencing was substituted in the royal parks. In
north-west Wiltshire, where stone was readily
available, stone walls were often used for enclosing
parks instead of fences. Hedges were also sometimes
used as an alternative to fences or walls for park pales.
Associated with the pale fence in early parks was a
continuous ditch excavated inside the fence which
allowed deer to leap into, but not out of the park. The
reduction of the full height pale fence to a lower fence
with a ditch offered considerable economies in
maintaining the pale. Fences required periodic repair
and occasionally complete renewal, and it was obvi-
ously much less expensive to maintain a lower fence
associated with a practically maintenance free ditch.

A good early 14th-century description of a
hedged and ditched park pale survives from East
Grafton Park (47) where John de Havering was
required to

...4nclose them on that part towards the
foreign lands there lately disafforested with a
great dyke 6 feet high and 7 broad, and with a
hedge, making the crest of the dyke so that the
king’s deer could not get out of the enclosure
but could enter without hindrance.!”

Around the middle of the 14th century continuous
ditches tended to be dispensed with to save labour,
and deer-leaps (also known as /ypzatts or saltory — the
latter from the Latin word meaning to leap) were
introduced. These consisted of short lengths of
lowered fencing with pits called ‘pitfalls’ opposite
and inside them which allowed deer to enter but not
escape from the park. It is possible that such deer-
leaps led to the expression to ‘drop in’ as well as the
word ‘pitfall’. Evidence of deer park ditches on the
ground is now rare. They have generally been
ploughed out and are at best merely slight swellings
in the ground or old field boundaries which have
adopted the line of the former park pale.

The dearth of labour which resulted from the
mortality of the late 14th century caused by the
Black Death may have contributed to the abandon-
ment of laborious continuous ditches and the
introduction of deer-leaps, the pale fences being
generally increased in height to compensate for the
lack of a ditch and to prevent the escape of park deer.

10. Victoria County History, Wiltshire, Vol. 4, p. 424.

As a broad principle it may be assumed that the later
the park the less likely are linear earthworks to be
associated with it.

Royal forests or groups of forests were adminis-
tered by a warden. Often a man of considerable stand-
ing, he was assisted by a number of foresters and
verderers. There were ‘riding’ and walking’ foresters,
the former being superior. The principal forester was
sometimes an hereditary office paid by modest
emoluments but carrying rights such as husbote and
heybote, rights to timber for the repair of buildings
and fences. His authority sometimes enabled him to
procure additional unofficial benefits. The parker
was an official appointed to look after one or more
deer parks. The common names of Forester, Forster
and Parker arose from the occupations of the officials
responsible for administering forests and deer parks.

Three types of deer were protected by Forest Law
and encouraged into parks. Red deer, being the
largest, were regarded as the noblest and most prized,
but fallow deer were the most usual park deer. After
the mid 14th century the smaller roe deer were
neither protected nor encouraged because it was felt
that their presence discouraged red and fallow deer.
Inferior deer were known as ‘rascals’. Fallow deer
fawn in late June or early July and during the ‘fence
month’ from 20 June to 20 July it was illegal to enter
the forest and disturb the deer. Throughout the
medieval period deer prospered under royal
protection, but a decline in deer numbers set in
during the 17th century as a result of more land being
taken into cultivation and was accelerated as a result
of the Civil War. There is evidence that deer parks
suffered during the war, for example at Wardour (85)
where the deer were turned loose during the siege of
the castle, and at Bromham (29) where house and
park were destroyed.

DECLINE OF DEER PARKS

In the Tudor period attitudes to deer parks began to
change. From being sources of sport and meat, parks
now began to be regarded as decorative settings for
the ostentatious houses which were being built, often
from the spoils of the Dissolution of the monasteries.
Some new deer parks were founded during the reign
of Henry VIII, but by the early 17th century parks
were under pressure from an increasing population
and its need for more agricultural land. So great was
James I’s love of hunting that there was a final
flourish as he encouraged the founding of new parks
at, for example, Bowood (21), Savernake Great Park

WILTSHIRE DEER PARKS

(44), and Savernake Brimslade Park (45). When the
Commissioners for the Sale of Crown Lands were
appointed in 1604 they were instructed that no
forest, chase or park was to be sold, and Charles I
continued this policy when raising funds in 1628.

Many deer parks were converted to farms during
the Tudor and early Stuart age, and by the beginning
of the 17th century Forest Courts had been
abandoned in many Wiltshire forests, although
Charles I tried to revive them. The pursuit of deer
gradually gave way to the hunting of foxes and by the
end of the 18th century the deer hunt had become
unusual except in the few areas, all outside Wiltshire,
where it has survived to the present day.

SOURCES

The principal sources used to research the former
deer parks listed at the end of this study are as
follows:

Maps. The initial search was into maps, old and
new, beginning with the maps of the 16th and 17th
centuries when demand arising from widespread re-
allocation of land led to great advances in
surveying and mapmaking. Large scale maps were
produced, to be followed by county maps to a
smaller scale. The county maps of Saxton, Speed
and Morden indicate the parks which had survived
into Tudor and Stuart times by a stockade symbol.
Saxton was the first cartographer to show deer parks
in his county atlas of 1579, and he was followed by
Speed and Morden with their maps of 1610 and
1695, respectively. Speed plotted 22 parks in
Wiltshire, although according to John Aubrey!! he
was aware of 29. On his 1695 map for Camden’s
Britannia Morden also indicated 22 parks, omitting
Speed’s Oaksey Park (1) but adding a park at West
Lavington (72). Other maps consulted were
Andrews and Dury’s 1773 map of Wiltshire, the
relevant sheets of the first edition of the Ordnance
Survey, and many Tithe, Enclosure and estate maps.

Archival sources. The second primary source was
provided by early documents. A single reference —
for example a licence to impark or a royal gift of deer
— is often the only positive evidence for a deer park
ever having existed. Two particular problems arise in
searching the transcripts of the early documents.
Although they are frequently available in printed
form, the early rolls are in Latin. The second

11. J. Aubrey, Natural History of Wiltshire (Newton Abbott 1969
reprint), p. 58.

problem is the inadequacy of the indexing. Many
more references to deer parks are found by simply
scanning the printed transcripts than by relying on
the indexes, probably in the proportion of four to
one. Research into documentary sources was
generally restricted to the printed volumes of
extracts from document rolls, but the Victoria County
History: Wiltshire and the many volumes of this
Magazine occasionally draw attention to early
documentary references to deer parks.

Place-names are vitally important indicators for
the former existence of deer parks. Relevant place-
names were discovered by close scrutiny of old and
new maps, and by detailed reference to the invaluable
The Place-Names of Wiltshire (see note 13 of this
paper). Etymology is a particularly specialised sub-
ject, but the following place-name elements provide
useful indicators for the former existence of deer parks:

Park is an obvious indicator although it is often
applied to landscaped grounds of the 18th century
and is sometimes a fanciful modern name. In my
own experience it is, however, surprising how often a
‘park’ name which has been dismissed as unlikely to
indicate a deer park is ultimately proved by documen-
tary sources to have survived from an authentic park.

Lypiatt in all its variant spellings is an early name
for a leap-gate. It appears in connection with many
deer parks, for example at Corsham (16).

Haye or Hey was an enclosure, often for keeping
deer. Isaac Taylor wrote ‘a haigh or hay 1s a place sur-
rounded by a hedge, and appears to have been an
enclosure for purposes of the chase’.'? Many Hayes
names occur around East Knoyle Park (80) and
elsewhere.

Lodge names frequently survive from parkers’
lodges in deer parks. Examples include Lower Lodge
Farm and Great Lodge Farm at Chippenham (19),
Great Lodge Farm at Savernake Great Park (44),
and ‘Collingbourne Lodge’, the last marked on the
Andrews and Dury 1773 map at Collingbourne (53).

Bower is an old name for a lodge or arbour for
ladies and is sometimes associated with a park.
Bowerhill at Seend Park (27) was shown by Andrews
and Dury as a moated site with the name ‘Bower
Island’.

Lawn derives from /aunde meaning an open glade.
It survives on the site of several former parks, for
example Lawn Farm at both Tisbury (83) and at
Wootton Bassett Vastern Great Park (33).

12. I. Taylor, op. cit., p. 102.

94 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Brail or Broyle is ‘a common forest term for a park
or wood stocked with deer or other beasts of the
chase and enclosed as a rule with a wall or hedge’.'
Several examples occur around Bedwyn in Bedwyn
Brail, Brail Farm, and Wilton Brail.

Stalls may possibly survive from ‘buckstalls’
which were framed nets used to catch deer. The
name appears at Stalls Farm at Longleat (74), and in
several places (Bournelake Stalls, Spackman Stalls,
and Bushy Leaze Stalls) around Hailstone Park (32)
at Cricklade.

Bers or Bars were barriers for enclosing deer.'
The name survives in Bars Bottom near Bedwyn
(50) and ‘Bar Ground’ (which is now Park Grounds
Farm) on Andrews and Dury at Vastern Great Park
at Wootton Bassett (33).

Dog Kennel has been suggested as ‘probably
originally a nickname given to some very small
or insignificant dwelling or hovel’.'° This may
sometimes be the case, but ‘dog’ and ‘kennel’ names
are often found associated with known former deer
park sites where they probably indicate the places
where deerhounds were kept. The many examples
include Kennel Farm (formerly ‘Dog Kennel’) at
Clarendon (59), Dog Kennel at Farleigh (63), and
‘Dog House’ on Andrews and Dury at Savernake
Great Park (44).

Breach is generally taken to mean ‘land newly
taken into cultivation’,!® but its frequent appearance
at known deer park sites suggests that it may
sometimes commemorate deer-leaps which were
effectively breaches in the park pale. The many
examples include Breach Lane at both Wootton
Bassett Vastern Great Park (34) and Southwick Park
(65), and Breach Farm at Compton (90).

Rail or Rayles may be a recollection of the post and
rail fencing of a park pale, as at Rodbourne Rail Farm
(formerly ‘Rayles’) at Malmesbury Cole Park (3).
“The Rayles’ on a 17th-century map of Wardour Park
(85) had become ‘Rails’ on an 18th-century map.

Pale is an obvious indicator word for a deer park,
for example in Park Pale Pond at Wardour (85).

Castles or sites of former castles, are a good starting
point in searching for deer parks. It is no
coincidence that O.G.S. Crawford combined
“Medieval Castle Mounds and Parks’ in Chapter 10
of his Archaeology in the Field (1953), for as he
pointed out they were often related. Many castles
were founded in Norman and Plantagenet times

13. J.E.B. Gover, A. Mawer, and F.M.Stenton, The Place-Names of

Wiltshire (Cambridge 1939), p. 332.
14. H.C. Brentnall records that ‘Bers means a deer enclosure, a
park in fact’: ‘The Metes and Bounds of Savernake Forest’,

when the monarchs and their nobles were addicted
to deer hunting. Consequently a deer park was a
usual concomitant to an early castle and at least one
deer park is to be found associated with most of the
early castles of Wiltshire.

General reading. The early topographers such as
Leland and Aubrey often mentioned deer parks,
but after such parks went out of fashion all but the
most obvious examples such as Clarendon (59) were
ignored by later writers, and knowledge of the
location of deer parks was lost. During the 1950s the
work of pioneering fieldworking local historians such
as O.G.S. Crawford and W.G. Hoskins, together with
an awakening interest in medieval archaeology, led to
the recognition of the important place deer parks had
occupied in medieval society. More recently, Leonard
Cantor and Christopher Taylor have pursued the
subject and drawn attention to the extent to which
deer parks have gone unrecorded.

Fieldwork. O.G.S. Crawford observed that ‘the final
test of all documentary sources is in the field’ and
exhorted all who were interested to ‘Go and walk
along what looks, on the map, like the boundary of a
park, and mark it in’.'” Fieldwork is a most rewarding
activity, but in connection with deer parks there are
particular problems. Most archaeological sites are
recognised and access is usually provided, but deer
parks are generally not acknowledged and by their
nature occupied large areas of land which have
subsequently been taken into cultivation. Access is
therefore often difficult. Public rights of way may
provide restricted access to the sites of most former
parks, but such restriction limits investigation in the
field.

BIBLIOGRAPHY

The only published recent work on Wiltshire deer
parks is by Professor L.M. Cantor whose work
includes 43 parks in Wiltshire. An earlier record is
the 1583 Note of Parks in the County of Wilts in the
State Papers which briefly described the deer parks
which existed at that time. This was extensively
drawn upon by E.P. Shirley whose work (also listed
below) described the parks existing in 1867, many
of which were early examples. In 1892 J.Whitaker
produced a study describing the parks which existed
towards the end of the 19th century. The major
publications are:
WAM 49 (1941), p. 423.
15. Gover, Mawer and Stenton, op.cit., p. 376.

16. Ibid., p. 423.
17. O.G.S. Crawford, Archaeology in the Field (London 1953), p. 199.

i)
Ol

WILTSHIRE DEER PARKS

BERESFORD, M., History on the Ground (Gloucester 1984),
Chapter 7, ‘A Journey through Parks’, pp.187—236 (no
examples from Wiltshire)

CANTOR, L.M., ‘The Medieval Parks of England’, Geography
Vol. 64, Part 2 (1979), pp.71—85, republished as:

CANTOR, L.M., The Medieval Deer Parks of England, A
Gazetteer (1983)

CANTOR, L.M., AND WILSON, J.D., “The Medieval Deer
Parks of Dorset’, Proceedings of the Dorset Natural
History and Archaeological Society, Vols. 83-100
(1961-1978), passim

CRAWFORD, 0O.G.S., Archaeology in the Field (London 1953),
pp.189-197

HOSKINS, W.G., The Making of the English Landscape
(London 1955), pp.73-76

HOSKINS, W.G., Local History in England (London 1959),
pp.114—-116

HOSKINS, W.G., Fieldwork in Local History (London 1967),
pp.51—54

PATTON, J., ‘How Deer Parks Began’, Country Life,
16 September 1971, pp.660-—662

SHIRLEY, E.P., Some Account of English Deer Parks (London
1867), pp.99-103

TAYLOR, C.J., Fieldwork in Medieval Archaeology (London
1974)

WHITAKER, J., The Deer Parks and Paddocks of England
(London 1892), pp.166—172

Schedule of Wiltshire Deer Parks

(Numbers refer to the distribution map, Figure 2)

The following schedule lists all the known and probable early deer parks in Wiltshire. The reference numbers are
common to both the schedule and the distribution map. Most of the parks are authenticated by documentary
references. Several are included on grounds of probability rather than conclusive evidence. The degree of
probability for the inclusion of each park is indicated on both the schedule and the distribution map. The writer
of this paper is preparing descriptions of the individual parks and would be interested to hear from anyone who
has information on any of the parks, particularly those classified C (probable) and D (only possible).

Key:

Number and Name of park / Map reference / Probability: A—definite, B—almost certain, C—probable, D—only
possible / Date of first documentary reference found to date / Notes

ee
FKP OoOoOONaA UN BP WNY

es
COONAN D UW & W WH

wow NY WY WH
BwoNw re ©

. Oaksey / 990926 / A / 1336. Became first Duchy of Lancaster, then Royal park. Disparked 17C.
. Charlton / 970900 / B/ 1580. Late park replacing an earlier one at Stonehill to its east.

. Malmesbury / 941853 / A/ 1235. Several parks owned by the Abbots of Malmesbury.

. Garsdon / 967873 / A/ 16C. Abbot of Malmesbury’s park, acquired by Richard Moody.

. Corston: West Park / 971843 / B/ 1453. Park probably belonged to the Abbot of Malmesbury.

. Somerford: Maunditt’s Park / 956851 / B/ 1451. Park associated with Great Somerford Castle.

. Dauntsey / 009801 / B/ 16C. Formerly Abbot of Malmesbury’s park.

. Clack (Bradenstoke) / 997796 / C / 1538. Park associated with Bradenstoke Abbey.

. Draycot / 935785 / A/ 16C. Created by the Cernes. Acquired by the Long family.

. Stanton St Quintin / 895798 / A/ 1602. Owner Abbot of Cirencester. Disparked 16C.

. Castle Combe / 838775 / A/ 1327. Created by Dunstanvilles beside their castle. Disparked 17C.

. Yatton Keynell / 866779 / C / 1354. Probable short-lived park associated with Castle Combe.

. Kington St Michael / 899773 / C / 17C. Belonged to Abbot of Glastonbury. Disparked 16C.

. Colerne / 836729 / C/ 1311. Mystery park possibly associated with castle mound to its east.

. Hartham / 861721 / D/ none. The name suggests an early park, possibly associated with Corsham.
. Corsham: East / 880710 / A/ 1242. Earl of Cornwall’s park. Became Royal park. Disparked 16C.
. Corsham: West / 856700 / B/ 1300. Second Earl of Cornwall’s park. Probably disparked early.

. Easton / not known / D / 1292. There may be confusion with Crux Easton in Hampshire.

. Chippenham (Pewsham) | 937714 / B/ 1299. Crown park. Pewsham and Chippenham probably same.
. Bremhill / 985743 / C / 1592. Probably illegal park created by Sir Henry Bayntun.

. Bowood / 987701 / A/ 1619. Royal park known as ‘King’s Bowood’. Broken up after Civil War.

. Lacock / 912675? / C / 1260. Tentative. Possible Longespée deer park. Location uncertain.

. Bowden / 940675 / B/ 1583. No early references. May once have been part of Spye Park.

. Spye / 950675 / A/ 1605. Probably an early park acquired by Bayntuns after Civil War.

. Monkton Farleigh / 819656? /B/ 12C. Park associated with Cluniac Priory and de Bohuns.

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

—

®2

a ee
/~> NN ‘S
@ 1
at i @4 J
5@ @a6 @ 35

1@ 2 O13 @ 39 Me
O 40
@ 20
( 140 O15
@ @19
17 wee . @21 43 @ 2®e@
f 2299 @@24 49
) @ 25 @29 44@ O @51
48
Q 26 @31 s5@ ? @50
aa 578 O28 @ 30 46 @@47
63
XR o%4 @ 69 e71
@65 2° " 5%
O70
\ ess an a Te |
@ 68 Q 56 of
@ 73 (
4 @74 @75 O76
J \
ve Oo 77 a
4 @ 78 @ 89
SA @79 80 @82
= os 57 59
Xo @33 @.90 @ @
\ ®
5
Q@ 86 be @ 58
X i 60@ 62@
@
~~ a wv v~,,
s71Gf 10 . seo ie,
Led Miles ws’
Note: numbers refer to those SSS Kms
15 10 5 0

in the preceding Schedule
LEGEND @ Definite

® Almost certain
@ Probable

DISTRIBUTION MAP

SHOWING THE DEER PARKS OF WILTSHIRE

O Only possible

KW 1995

Figure 2

WILTSHIRE DEER PARKS

26. Holt / 856626? / C/ 1316. Existed 1316 when owner was John de Holte. Site uncertain.

27. Melksham / 910610 / B/ 1299. Crown park given to Despensers.

28. Seend / 925610 / A/ 1309. Royal park granted to Despensers, then given to de Bohuns.

29. Bromham / 976659 / A/ 1583. May be early. Belonged later to Bayntuns. Wasted in Civil War.

30. Devizes Old Park / 995609 / A/ 1149. Very early Royal park. Disparked by 1595.

31. Devizes New Park / 008632 / A/ 1157. Second early Devizes park at Roundway.

32. Hailstone (Cricklade) / 083944 / A/ 1236. Early Fitzwarin park. Probably disparked very early.

33. Wootton Bassett : Vastern Great Park / 049822 / A/ 1230. Created by the Bassets. Very large.

34. Wootton Bassett : Little Park / 053803 / A/ 1267. Second Wootton Bassett park.

35. Lydiard Tregoze / 089850/ A/ 1256. Owned by Tregozes, then the St Johns. Landscaped 18C.

36. Swindon / 163833? / D/ 1247. Tentative, although there is a 13C reference to park at Swindon.

37. Burderop / 166802 / B/ 1583. May be early. Existed 1583. Good park pale ditch.

38. Elcombe / 138803 / A/ 1260. Created by the Lovells, forfeited and granted to Comptons.

39. Woodhill / 061769 /B / 1304. Early park owned by de Besylles. Probably disparked early.

40. Aldbourne / not known / D / 1307. Earl of Lincoln had a park here in 1307. Position not known.

41. Snap and Upham / 229771 / C/ 1606. References indicate a park here. In Aldbourne parish.

42. Littlecote / 300700 / A/ 1520. Park of the Darells. Henry VII hunted here.

43. Ramsbury : Old Park / 255710 / A/ 1246. Early park of the Bishops of Ramsbury. Large.
Ramsbury : New Park / 260703 / A/ 14C. An early second park of the Bishops.

44. Savernake : Great Park / 190660 / A/ 1622. Large late park created early 17C.

45. Savernake : Brimslade Park / 209634 / A/ 1625. Late park created early 1600s.

46. Savernake : Suddene Park / 245614/ A/ 1547. Early park at Wolfhall.

47. Savernake : East Grafton / 256610 / A/ 1347. Existed in early 14C.

48. Savernake : Tottenham / 246644 / A/ 1547. Replaced Suddene as principal Savernake Park.

49. Savernake Lodge / 233667 / D/ none. Some evidence of park around Savernake Lodge.

50. Savernake : Bedwyn Parks / various / A / 1231. Several early parks around Bedwyn.

51. Savernake : Chisbury / 273654 / A/ 1260. Created by Matthew de Columbers mid 13C.

52. Conholt / 320546 / D/ none. Possibly early deer park, but uncertain.

53. Collingbourne / 275528? / A/ 1253. Emparked mid 13C by William de Valence.

54. Ludgershall / 263516? / A/ 1216. Royal park beside castle. Granted to Despensers.

55. Everleigh / 215538? / A/ 1234. 13C deer park owned by de Montfort, then Duchy of Lancaster.

56. Coombe / not known / C / 1288. Tentative. There may be confusion with Castle Combe Park.

57. Wilton / 100305 / A/ 1578. Late creation of the Herberts in 1540s. Landscaped in 16C.

58. Faulston / 073423 / A/ 1618. Emparkment by Bayntuns in 1387 may have replaced earlier park.

59. Clarendon / 181302 / A/ 1223. Very large Crown park, largest park in Wiltshire, 7 miles round.

60. Downton / 200227? / A/ 1283. Park of the Bishops of Winchester. Existed 1283. Site uncertain.

61. Loosehanger / 213912 / C/ 1684. Mentioned as park in 1684, but could be earlier.

62. Newton / 242226? / A/ 1253. Imparked 1253 by William de Valence, Earl of Pembroke.

63. Farleigh (Wiltshire Pk) / 809582 / A/ 1431. 1431 reference to the park at Farleigh Castle.
(Farleigh Hungerford was in Somerset, but its park was in Wiltshire, hence its name).

64. Trowbridge / not known / D/ none. Speculative. There should be a park associated with the castle.

65. Southwick / 845558 / A/ 1246. Adam de Grenville enclosed unlicenced park here in 1246.

66. Brook / 851525 /A/1323. Park here in 1323. Lord Willoughby de Broke took his name from Brook.

67. Rood Ashton / 888563 / C / 1248. Uncertain. There is a licence for a park at “Little Aston’ (1248).

68. Westbury / 861509? / A/ 1230. Deer park at Westbury in 1230. Location not known.

69. Keevil / 920584 / A/ 1318. Earl of Arundel owned a deer park here in 1318. Location uncertain.

70. Erlestoke / 965535? / D/ none. Possible early deer park. Landscaped early 18C.

| 71. Potterne/ 010574 / A/ 1353. Park of the Bishops of Salisbury. Existed 1353. Probably very large.

72. West Lavington / 006523 / B/ 1695. Probable late park enclosed by Sir John Danvers in 17C.

| 73. Corsley / 825460 / A/ 1572. Late park created by Sir John Thynne in 1570s. Shown by Saxton 1579.

| 74. Longleat/ 815430 / A/ 1422. ‘Parco de Hornyngesham’ in 1422. Probably earlier.

75. Warminster / 877423? /B/12C. Deed refers to ‘parks’ at Warminster. 1327 Crown park.

| 76. Heytesbury / 932428 / C/ 1320. Lord Badelesmere imparked here 1320. Went through many hands.

Ts
78.
79.
80.
81.
82.
83.
84.
85.
80.
87.
88.
89.
90.

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Sherrington | 9553752 / D/ none. Possible park of the Giffords towards Great Ridge Wood.

Stourton / 967343 / A/ 1427. 1427 Lord Stourton licenced to impark 1000 acres. Landscaped 18C.
Zeals / 795313 / A/ 1246. Geoffrey de Zeals imparked illegally in 13C. Later licenced.

East Knoyle / 877301 / A/ 1253. ‘Parco suum de Knoel’ (1253). Owned by Bishops of Winchester.
Mere / 849298 / A/ 1268. Earl of Cornwall created park late 13C. Records of park breaks 1296.
Fonthill / 9333152 / A/ 1373. Many successive parks here. ‘Le parke’ in 1373.

Tisbury / 927297 / A/ 1376. In 1376 Sir Thomas West was licenced to add to his park here.

West Hatch / 909280? / A/ 1280. In 1280s there was a park here. Location uncertain.

Wardour / 930260 / A/ 1382. Reference 1382 to ‘Parkmede’. Later two parks.

Donhead (Wincombe) / 880241 / C / 1552. 1552 Lord Pembroke probably created park at Wincombe.
Tollard Royal / 945173 / A/ 1227. Park mentioned 1227. 1615 recorded as hedged and ditched.
Rushmore / 956189 / B/17C. Some evidence that Robert Cecil in 17C created park here.

Grovely Lodge / 046340 / C / 1589. Drawing of 1589 shows park fence enclosing deer at Grovely.
Compton (Chamberlayne) / 031299 / C/ 1274. Reputed to be an early deer park but uncertain.

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 99-105

William Lisle Bowles: The Making of the Bard
of Bremhill

by DOREEN SLATTER

Wilham Lisle Bowles was fortunate in attaining a way of life in which he could fully exercise his many
talents. Early success and his attractive personality brought him many friends and made him known to
a wider social circle than other literary Wiltshire clergymen at the time enjoyed. Although his personal

papers have been dispersed, his poems and books contain autobiographical and topical allusions. These,
together with the letters and memotrs of those who knew him, provide material for an understanding of

his career and of the importance of its background setting at Bremhill. This paper attempts to establish
Bowles’ development as a parson poet and to suggest the ways in which his family and friends

contributed to it.

William Lisle Bowles’ appointment as vicar of
Bremhill in 1804 placed him in possession of an
interesting church, standing on high ground, with a
comfortable old house just below it, commanding an
extensive view towards the Marlborough Downs.
Bremhill was a well endowed living, providing the
means and opportunity for him to enlarge the scope
of his activities. He seems to have determined, as a
resident parson, to take his new responsibilities
towards his parishioners more seriously than had
previously been the case. He saw acting as a county
magistrate as a proper extension of his functions
and, from time to time, his voice was heard on larger
issues affecting the Church. But he continued to
write poetry and historical works and took his place
in social gatherings. In spite of extreme absent
mindedness, and probably many absences from
home, he seems to have been popular with his
parishioners. It is hoped, by considering incidental
references to his surroundings, family and friends, to
present a larger view of Bowles in West Country
society. Two contemporary accounts of his garden at
Bremhill Vicarage provide further information about
the scene in which he moved. However, any present
appreciation of Bowles must be limited by the fact
that his papers have been dispersed and many are in
the United States of America.

Bowles came to Bremhill with a considerable
reputation as a poet and with antiquarian and artistic

1. First the curacy of East Knoyle, then the rectories of Chicklade,
and Dumbleton, Glos. See D. Slatter, ‘The Revd William Lisle
Bowles (1762-1850: The Need for a Re-appraisal’, WAM 86
(1993), p. 138. An examination of the parish registers might
reveal how far Bowles was resident.

2. In 1798 Charles Bowles published a translation of the custumal

tastes ready to profit from his new situation. The
background details of his early career are little
known. He prefaced his small book, Scenes and
Shadows of Days Departed ... (1837) with some
autobiographical notes about his early childhood in
which he referred to the influence of his parents
upon his character. However, his remarks included
no references to his education at Winchester and
Oxford, where he came under the influence of Dr
Joseph Warton, headmaster of Winchester College,
and then of Thomas Warton, his brother, Camden
professor of Ancient History at Oxford and poet
laureate, 1785-90. Their poems and teaching must
have stimulated Bowles’ own talent, so that he began
to publish his poems as early as 1789. At the same
time, he was embarking on a career in the church,
following the example of his father, grandfather and
great grandfather.

After being ordained, he lived for some time in a
house in the parish of Donhead St Andrew while
holding preferments elsewhere.'! The influence of
various people in the vicinity of Donhead may have
contributed to the development of his career. His
younger brother, Charles, who had established
himself as a lawyer in Shaftesbury, becoming
Recorder of the town in 1804, was one of these.
Charles Bowles had probably developed an interest
in local history at an early stage as his profession
enabled him to handle ancient documents.* He

of the manor of Gillingham, Dorset, and the confirmation of
the charter of Gillingham by Elizabeth I. In it he is described as
notary public. He went on to write the History of Chalk Hundred
(1833) for Sir Richard Colt Hoare. His obituary notice
appeared in The Gentleman’s Magazine, n.s. vol.8 (1837),
pp.90-1.

100 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

provided information for the second edition of John
Hutchins’ The History and Antiquities of the County
of Dorset (1796-1815), an undertaking supported
by Sir Richard Colt Hoare.’ It may also be relevant
to mention another relative, the Revd Peregrine
Bingham, rector of Edmondsham, Dorset, and later
of Berwick St John, a brother-in-law of Charles and
William. In 1804 Bingham published a Memorr of his
father, the Revd George Bingham, who had been a
collaborator of Hutchins.

It is important that the Bowles brothers became
acquainted with Sir Richard Colt Hoare of
Stourhead, probably in about 1800. Hoare, who had
returned from travels in Italy and Wales, then
decided to devote himself to the investigation of the
early history of his county. His Ancient History of
Wiltshire (1812-21), recorded the excavation of long
barrows and other tumuli, with detailed drawings by
Philip Crocker, a surveyor from the Ordnance
Survey, of the sites and objects discovered.
Unfortunately the work did not give full dates of the
different excavations and seldom mentioned the
names of those present. Hoare relied primarily upon
William Cunnington of Heytesbury,' and different
friends and neighbours, such as the botanist A.B.
Lambert of Boyton,’ who happened to be available.
Exceptionally, in Volume I, there is a description of
the excavation of a tumulus at which William Lisle
Bowles and Richard Fenton, the Welsh poet and
topographer, were present.° A dramatic thunder-
storm broke out during the proceedings, forcing
those who had been watching to take shelter in the
pit that had been dug. On his return home, Bowles
was inspired to write a poem which Hoare, his
artist’s eye similarly affected by the drama of the
scene, included in his published account.

Colt Hoare’s work has earned him recognition as
a distinguished pioneer of field archaeology in this
country. But the legacy of William Stukeley’s ideas
about the Druids continued to affect the appre-
ciation of the pre-Roman past. This was reinforced
by a new interest in Welsh history and culture, which
Hoare shared, and an attempt to connect the tradi-

3. J.Hutchins, The History and Antiquities of the County of Dorset
(1861-64), 3rd edn., vol.3, p.11. The editors of this edition
acknowledged that the editors of the former edition were under
‘particular obligation’ to Charles Bowles.

4. See R.H.Cunnington, From Antiquary to Archaeologist, A
Biography of William Cunnington, 1754-1810, ed. J.Dyer
(Princes Risborough, 1975).

5. Aylmer Bourke Lambert (1761-1842), original Fellow of the
Linnean Society and Vice President, 1796-1842. Volume 2 of
his work on the genus Pinus (1824), was dedicated to Colt
Hoare.

tional Welsh bards with the Druids.’ In his poem,
Bowles supposed that the tumulus was the last
resting place of a ‘white hair’d Druid Bard sublime’.
Richard Fenton had witnessed the excavation and
the Dictionary of National Biography article about
him states that he was a friend of Bowles as well as of
Hoare. Hoare remained a friend of Fenton until he
died in 1821 and a portrait of him still hangs at
Stourhead.

The lore of the Druids surely contributed to
enhance the mystique with which Bowles was
surrounded as a poet, famous since the success of his
Fourteen Sonnets Written at Picturesque Spots on a
Journey in 1789. Colt Hoare wrote to him as ‘My
dear Bard’® and he, Crabbe and Moore were referred
to as the three Wiltshire bards.? When Bowles moved
to Bremhill in 1804, he found no difficulty in
reconciling his role as a famous poet with that of a
country clergyman. He set out to make his vicarage
reflect his personality and his idea of the place of the
parish clergyman in society. He therefore decided on
certain alterations to the appearance of his house
and started to create a garden on Shenstonian
principles which would convey a moral message. He
justified his inclusion of a description of the garden
in his book The Parochial History of Bremhill (1828)!°
by saying ‘One of my objects besides miscellaneous
information on parochial objects, was, in the present
age of clerical obloquy, to exhibit the clergyman and
his abode in their proper moral position in English
Society’.

Very little is known about the construction of the
vicarage garden but it is evident that Bowles must
have begun the work soon after his appointment. By
1810 he was in a position to give information to the
third Lord Lansdowne who was restoring the house
and gardens at Bowood. Writing to Bowles on 4
September, Lord Lansdowne acknowledged his
advice and added that he remembered Josiah Lane,
previously employed by Bowles, as ‘an excellent
executive workman’ but one needing to be
supervised.'! Presumably Lane was the builder of
rock work and what Bowles called ‘a kind of cave

6. Sir R.C. Hoare, The Ancient History of Wiltshire (1812-21), repr.
EP Publishing Ltd., 1975, vol. 1, pp.238-41.
7. By Edward Williams (1747-1826), a Welsh bard. See ‘Druids’ in
British Heritage, eds. A. Isaacs and J.Monk (Cambridge, 1986).
8. Sotheby & Co., Sale Catalogue, 24 March 1936, p.54, no. 143.
9. E.g. when they attended a dinner at the opening of the Bath
Institute, Jan. 1825: WAM 34 (1905-06), p.230.
10. The Parochial History of Bremhill (London, 1828), (hereafter
Bremhill), p.xvi.
11. G. Greever, A Wiltshire Parson and his Friends (London, 1926),
p. 99.

_—_ ae

WILLIAM LISLE BOWLES

with a dripping rill which falls into the water
below’,!”? the subject of a poem by Bowles in 1808.
After he had taken up residence at Bowood in 1812,
Lord Lansdowne began the practice of taking some
of his house guests to see Bowles in his garden, ‘one
of the prettiest spots in the county’.!? By 1814, the
attractions of the place had become sufficiently well
known to be recognised by an article in The
Gentleman’s Magazine, probably written by Bowles’
friend, Archdeacon Robert Nares.!* This claimed it
as an abode of genius, comparing it with The
Leasowes and Hagley. The Biographical Dictionary of
Living Authors (1816) also referred to the garden.!°
The description by Nares and Bowles’ own later
account were written as perambulations of the site
though starting from different sides of the vicarage.
Changes or additions to the garden had probably
been made during these fourteen-year interval
between the two and these perhaps account for some
difficulty in reconciling the two versions. Further-
more, the number of features mentioned does not
coincide.

A letter to Bowles from Hannah More, probably
written in 1809 when she was living at Barley Wood,
Somerset, touches on one feature of the garden, an
urn commemorating Bowles’ brother, Henry, who
had died prematurely in 1804.!° It appears that
Hannah More had commissioned a similar um to the
memory of a person not named in the letter, but
presumably her friend Bishop Beilby Porteus of
London, who died in 1809. She wrote to Bowles
asking his advice about details of the dedication, her
words suggesting something more formal than the
inscription eventually used on the pedestal. Bowles
was asked to make arrangements with Mr King, the
statuary, probably a member of the firm established
in Bath and London,” to execute her order; possibly
the same person had supplied the urn in Bowles’
garden. Mounted on a pedestal, it is illustrated on
p.224 of Henry Thompson’s The Life of Hannah

12. Bremhill, p.256.

13. Greever, op. cit., p.99.

14. The Gentleman’s Magazine, vol. 84 (1814), pt.2, pp.203-04.

15. J.Watkins and F.Schoberl, A Biographical Dictionary of Living
Authors of Great Britain and Ireland (London, 1816), pp. 35 and
416.

16. The Gentleman’s Magazine, n.s. vol.4 (1835), pt.2, p.246.

17. R. Gunnis, Dictionary of British Sculptors, 1660-1851 (London,
1968), pp.228-29. See: ‘King, Charles’ and ‘King, Thomas &
Sons’.

18. G. Grigson, Places of the Mind (London, 1949), p. 12. The Latin
inscription is included in the account of Henry Bowles by W.
Munk (ed.), The Roll of the Royal College of Physicians of London
(London, 1878), vol. II, p.445. There is a memorial tablet to
Henry and his wife, Penelope, on the wall of the north aisle of
Winchester Cathedral. I am grateful to Dr P.Robinson of

101

More... (1838). It was of a very simple form, without
handles, but with a lid. The monument stood in a
group of trees on a high point in the Barley Wood
garden. In contrast, Geoffrey Grigson, who saw
Bowles’ urn at Bremhill in about 1948, said it was
shaded by dark trees, just beyond the rill and the
water.'® He quoted a Latin inscription in memory of
Henry Bowles, apparently on the urn itself, which
the description implies was standing on the ground.
According to the account in The Gentleman’s
Magazine, the urn stood ‘on a gentle ascent’ above ‘a
pleasing cascade’.

A friend who made a definite contribution to the
garden was Samuel Rogers, the poet. Rogers
entertained Bowles in London and they also met as
guests at Bowood, but it is unclear how they came to
know each other. According to the writers of
Reminiscences and Table Talk of Samuel Rogers (1903),
who give no dates to their anecdotes, Rogers found
Bowles excessively timid.'!? He himself, however,
could be a somewhat forbidding person. Maria
Edgeworth wrote after breakfasting with him in 1830
that he was ‘not more yellow than ever nor more
satirical, for both are impossible’.2° Nevertheless,
this contrasting pair shared an interest in garden
design. For example, Rogers had considered buying
Pope’s villa at Twickenham when it came on the
market.?! In 1817 Bowles invited him to make a
contribution to the Bremhill garden by sending him
white sticks with which to mark spots selected by
Rogers for planting trees.” Bowles’ description of
the garden in his Parochial History of Bremhill
mentions a large ‘Indian shell’ given by Rogers and
marked by an inscription by his nephew, Peregrine
Bingham the younger.”*

It is significant that both Bowles and Rogers knew
Sir Uvedale Price of Foxley, Herefordshire, though
again it is not clear how this came about. Writing to
Rogers in 1824, Price said that he had known
Bowles for some considerable time. He went on ‘I

Compton, near Winchester, for this information.

19. Reminiscences and Table Talk of Samuel Rogers . . . Collected from
the Original Memoirs of Dyce and Sharpe, ed. G.H. Powell
(London, 1903), pp.200-01.

20. E.Inglis-Jones, The Great Maria (London, 1959), p.230.

21. Table Talk of Samuel Rogers, p. 13.

22. P.W. Clayden, Rogers and his Contemporaries (London, 1889),
vol. 1, pp.250—51. In the same letter, Bowles refers to his boat
‘with flag, gardener and pony’.

23. Bremhill, pp.251-52. The inscription was said to be by the
author of The Pains of Memory (1811), but the text goes on to
make it plain that the author was the younger Peregrine
Bingham (who later became a lawyer), not his father, the
clergyman, to whom the poem is generally attributed.

24. 6 Oct. 1824. Clayden, op. cit., vol. 1, pp. 387-88.

102 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Bremhill Parsonage, from an engraving by R. Sands, after W. Bartlett, c.1820, reproduced in Bowles’
Parochial History of Bremhill ... (1828). (Photograph by Derek Parker)

should have been very glad to have met him, and
heard him perform his water-music and do the
honours of his water-party’, possibly upon a barge
on the canal below the house. Price was the author of
the long and detailed Essay on the Picturesque as
Compared with the Sublime and the Beautiful... (1794),
a work which must certainly have affected Bowles’
taste. Writing to J.B.Nichols, Bowles called it ‘the
sweetest book in the English language on the
picturesque and beautiful’.*” Rogers and Price knew
Sir George Beaumont, named by Bowles as one of
the visitors to his garden.

The chief inspirations of the Bremhill garden must
have been the surrounding landscape and a re-
awakening of his boyhood interest in garden planting.
As he surveyed the prospect from his study window,
Bowles would have been reminded of different ages
of English history. In the valley immediately below
were the remains of Stanley Abbey, recalling the
development of monasticism in the medieval church.
The borough of Calne with its fine Perpendicular
church, enlarged by prosperous clothiers, and
Bowood Park not far away, the centre of enlightened

25.1 am indebted to Mr R. Hatchwell for this quotation from a
letter now in the USA.
26. Scenes and Shadows of Days Departed. . . (London, 1837), p.xix.

patronage, lay in the foreground to the south. In the
distance were the Downs, with famous prehistoric
monuments beyond, evidence of an ancient past then
hardly understood. Bowles put on record that he
owed his feeling for music to his mother and his
appreciation of landscape and the beauty of nature to
his father whom he had watched, as a boy, planting
trees and shrubs at Uphill Rectory, Somerset, and at
Barton Hill House, Shaftesbury.”° Bowles seems to
have been particularly attached to Uphill, which he
returned to visit in later life. Although the grand
gardens at Stourhead and Fonthill must have inter-
ested him, Uphill was on the scale of his own grounds.

Bowles clearly wished the vicarage garden at
Bremhill to be judged as part of a larger scene,
in conjunction with the vicarage, church and
churchyard. By the time he had published his
Parochial History of Bremhill, he had made certain
embellishments to the vicarage ‘in consonance with
ideas of picturesque propriety’.*’ He was referring to
the power of architecture to affect the imagination by
means of association of ideas, a subject much
discussed by writers of the time.** Bowles explained

27. Bremhill, p.249.
28. E.g. by Sir Joshua Reynolds in his Thirteenth Discourse first
delivered in 1786 to students of the Royal Academy.

WILLIAM LISLE BOWLES

in his History that by ‘parapetting the whole [the
vicarage and buildings] with a simple gothic
ornamental railing, such as appears on the church at
Stourhead, unity has been given to the exterior, and
the long low roofs have put on an ecclesiastical
appearance’.”? Both Bowles and Colt Hoare may
have had in mind the exterior of Lacock Abbey,
where the Gothic Revival work of Sanderson Miller
had adapted and enlarged a medieval ecclesiastical
building. The effect was not lost upon Maria
Edgeworth when she was taken to Bremhill by the
third Lord Lansdowne in September 1818. She was
more impressed by the ‘very pretty old parsonage
newly done up with good taste’ than she was by the
‘little shrubbery, stuck full of inscriptions and
grottos’.*° Slightly later, as Pevsner noted,*! Bowles
also added small Gothic turrets and pinnacles to the
vicarage, one being dated 1820. He wrote to
Thomas Moore on 1 July 1820: ‘I am making quite a
priory here; Gothic arches, turrets, pinnacles etc. .. ..*
The Parochial History of Bremhill has an engraving of
the vicarage and outbuildings viewed from the
front®’ with, on one side, a glimpse of the distant hills
showing the Cherhill horse. There is also a plan of
the rooms indicating a conservatory next to the
drawing room.

The Bremhill garden covered only 2: acres and,
even so, Grigson suggested that Bowles might have
included in it part of his glebe. The descriptions of it
are imprecise and give no idea of proportions. A
green lawn and gravel path went round the vicarage
on the garden side, the carriage entrance being on
the north. As the ground fell away sharply, the few
trees and shrubs at a little distance from the house
would not have obstructed the views. The garden
consisted of alternating bands of shrubbery and
flowers, including rose beds, descending finally to a
field of sheep, the bells of which, we are told by
Thomas Moore, were tuned in thirds and fifths.**
According to The Gentleman’s Magazine, the terrace
of flowers some way down the slope contained small
flower beds and trellis-work arbours, a Reptonesque

29. Bremhill, p.249.

30. C. Colvin (ed.), Maria Edgeworth. Letters from England 1813-44
(Oxford, 1971), p.98.

31. N. Pevsner, revised B.Cherry, The Buildings of England: Wiltshire
(Harmondsworth, 1981),pp. 140-41.

32. Lord J. Russell (ed.), Memoirs, Journal and Correspondence of
Thomas Moore (London,1860), p.261.

33. Bremhill, opp. p. 245.

34. Moore Memoirs, p. 161: 1 Sept. 1816.

35. The Revd Edward Duke ( 1779-1852) of Lake House, Wilsford
near Amesbury, elected a Fellow of the Linnean Society in
1809, A.B. Lambert being one of the proposers. I am indebted
to Miss G. Douglas, Librarian of the Society, for this infor-

103

feature. The shrubs and flowers were connected by
winding paths, with opportunities to pause at various
seats or larger features. Near these vantage points
were inscriptions (published with Bowles’ poems)
disposed to assist visitors to contemplate the view in
a proper mood. Bowles’ interest seems to have been
in feelings only and there is no indication that he had
any concern with botany or cultivation technique as
had his friends Colt Hoare, the Revd Edward Duke*®
and A.B.Lambert. He casually mentioned a few trees
— thorns, hazels and poplars and an ‘old ivied elm’,
described for its picturesqueness.

The most important of the garden features was
the hermitage or oratory, created by Bowles with a
distinctive ecclesiastical if not monastic character. It
was a late example of a building which had been
constructed in large and small gardens in the latter
part of the 18th century, sometimes made more
realistic by an actual inmate. The hermitage at
Stourhead, known as the Druid’s cell, was built of
timber with its bark left on. It was put up in 1771 but
taken down in 1814.*° The 1814 account of Bremhill
refers to a ‘root house hermitage’ but Bowles’
description suggests that later there were two
buildings. Neither description specified the materials
used for the walls; both concentrated on the fittings.
Bowles. mentions a window filled with stained
glass;*’ and according to The Gentleman’s Magazine,
the interior held ‘a rude stone table’ and a wooden
chair. Presumably just outside was a small sundial on
a fragment of twisted column and probably on top a
‘rustic cross which St Bruno, the Hermit, is
supposed to have erected’. Bowles gave more
information about the sundial, saying the plate was
dated 1688.°° An anecdote of Thomas Moore’s
relates that, on the arrival of visitors, Bowles would
send a servant to start the fountain and place a
‘missal and crucifix’ in the oratory.*”

Also in the oratory were ‘shattered fragments’ of
pillars from Stanley Abbey.*? It seems that Bowles
must have carried away a number of objects from the
site, even if he may not have undertaken an actual

mation. Duke was also a diligent investigator of prehistoric
remains. He exchanged letters with Bowles about the Wansdyke
and Avebury in The Gentleman’s Magazine, between 1827 and
1829. Bowles published his views in chapter II of his Parochial
History of Bremhill. Duke published The Druidical Temples of the
County of Wiltshire (London, 1846), maintaining the existence
of a planetarium on the downs.

36. K. Woodbridge, The Stourhead Landscape (The National Trust,
1991), p. 58; plan and section, p. 54.

37. Bremhill, p. 251.

38. Ibid., loc. cit.

39. Moore Memoirs, p. 161.

40. Bremhill, p. 251.

104 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

excavation. Knowing the care taken by Colt Hoare
to have plans and drawings made of the sites he
excavated, it is disappointing that there are no such
plans and drawings in Bowles’ chapter on Stanley
Abbey in his Parochial History of Bremhill.*' His
research was otherwise detailed, devoting much
attention to original medieval documents in local
custody, which were transcribed with the assistance
of Charles Bowles. It would be interesting to know
whether William Lisle’s interest in Stanley Abbey
predated or followed that of his brother in
Shaftesbury Abbey. Charles Bowles undertook a
small excavation at Shaftesbury of which he made a
brief report in The Gentleman’s Magazine (1817).*”
William Lisle published an article about Stanley
Abbey in The Gentleman’s Magazine (1823), his
findings being used by the editors of the revised
edition of Dugdale’s Monasticon.* Perhaps his work
on the abbey caused him to construct a new
hermitage, the original one becoming the root
house? According to the description of 1828, the
root house contained an ‘old carved chair’, in which
visitors were placed to admire the view. There were
also two specially built ‘rural seats’ in the garden.
Bowles mentioned a Gothic stone seat at the end of
the terrace of flowers. It was presumably over this
seat that the 1814 account gives the inscription:

Rest, stranger, in this decorated scene

That hangs its beds of flowers, its slopes of green:
So from the walks of life the weeds remove,

But fix thy better hopes on scenes above.

In another part of the book, Bowles mentions a
second seat incorporating painted Norman tiles and
pillars of Portland stone.*® Only the 1814 description
refers to ‘a small neat obelisk’ inscribed ‘Anno Pacis
1814’.

Bowles was already known to the third Lord
Lansdowne before he took up residence at Bowood.

41. Ibid., pp. 83-123.

42. The Gentleman’s Magazine vol. 87 (1817), pt. 1, p. 209. Charles
Bowles employed a workman to dig and a floor with remains
was found at a depth of c.6ft. A plate of the objects discovered
faces the report.

43. The Gentleman’s Magazine vol.93 (1823), pt. 1, pp. 24-6.
Mainly a summary of documents.

44. Sir W. Dugdale, Monasticon Anglicanum (London, 1825), new
edn. by J. Caley, H. Ellis and the Revd B. Bandinel (8 vols.,
London 1817-30) repr. Gregg International Publishers Ltd.,
1970, vol.5, pp. 563-64. Bowles’ article cited above was used,
followed by additional extracts from charters etc. Bowles
acknowledged the help of Bandinel and Cayley in his Parochial
History of Bremhill.

45. Bremhill, p.252.

Thereafter he was regularly invited to join the house
guests and eventually became a friend of the family
there and at Lacock, the home of Lord Lansdowne’s
nephew.*’ Guests at Bowood included not only
Wiltshire neighbours, especially Lady Lansdowne’s
sister and her children from Lacock, but also men
and women nationally and internationally famous in
the fields of literature, art and science. Yet it is only
possible to identify those who went to see the garden
and to hear Bowles take a service at Bremhill church
when they are named in the letters and memoirs of
others. Bowles proudly listed those who had sat in
the ‘old carved chair’ as Sir Samuel Romilly, Sir
George Beaumont, Sir Humphrey Davy — poets as
well as philosophers, Mme de Staél, Dugald Stewart,
and Christopher North (John Wilson), Esq.*®
Samuel Rogers, Maria Edgeworth, and Thomas
Moore have already been mentioned and it is
possible to add a few more names. Sir Thomas
Phillipps, the collector, visited Bremhill in order to
examine a pedigree of the Lisle family.*? It may be
presumed that friends and colleagues in the church
came there, including Archdeacon Nares while he
was vicar of St Mary’s church, Reading. Hannah
More and John Rutter, printer and local historian,
were invited.°° A charming memory of visits to
Bremhill was described in a poem by the young
writer Louisa Costello, who published a volume
dedicated to Bowles in 1825.?! Her poem begins:

Sweet Bremhill! When last in thy garden I stray'd

The trees were all green and thy skies were bright;
The spray of the fountain ’midst roses that play’d,
Reflected their colours and glistened with light.

Coleridge, a greater poet, visited Bowles in 1815
while he was staying with his friends, the Morgans,
in Calne. In a letter to Wordsworth, he wrote
enthusiastically that Bowles had ‘a paradise of a
place’ at Bremhill.°? Bowles had remained in touch

46. Ibid., p. 122.

47. Lady Lansdowne’s eldest sister, Elizabeth, married firstly,
William Davenport Talbot of Lacock and secondly, Captain
William (later Rear Admiral) Charles Fielding. She was the
mother of William Henry Fox Talbot and had two daughters by
Captain Fielding. Caroline, the elder girl, married Ernest,
Viscount Valletort (later Earl Mount Edgcumbe) at Bowood in
1831.

48. Bremhill., p.253.

49. Bodleian Library, MS: Phillipps - Robinson, d 68 f.56.

50.I am indebted for this information to Mr F.C.Hopton of
Shaftesbury.

51. L.S. Costello, Songs of a Stranger (London, 1825), p. 132.

52. Coleridge to Wordsworth, 30 May 1815: H.J.Jackson (ed.),
Samuel Taylor Coleridge, Selected Letters (Oxford, 1988), p. 190.

WILLIAM LISLE BOWLES

with Wordsworth and Southey, who like Coleridge,
had admired his early poetry. In the course of a
tour in the West Country, Southey, with his son
Cuthbert, visited Bowles in November 1836 and
parted from him in his garden.’ Perhaps one of the
last visits of a nationally famous person to Bremhill
was made in 1837 when Lord Lansdowne brought
the Prime Miuinister, Lord Melbourne, with Lord
John Russell, over there from Bowood.”*

There was a peculiar aptness about Bowles’ arrival
at Bremhill. As his friend Nares wrote, ‘Who will not
wish that the Poet may long enjoy the Place, and the
Place the Poet, so worthy of each other’. In fact,
Bowles held the living for over forty years and was an
acknowledged celebrity until the infirmities of age
overtook him. His church, house and garden made a
sympathetic background from which he would

53. Kenneth Curry (ed.), New Letters of Robert Southey 1811-38
(New York and London, 1965), vol.2, p.457 f.n. His visit was
referred to by Bowles in Scenes and Shadows...(1837), p-xlv.

54. Moore Memoirs, p. 635. The visit took place on 16 Oct. 1837.

55. An epitaph for Benjamin Tremblin, d. 1822, aged 92, can still
be seen on the church wall.

56. Revd J.Hunter, “The Topographical Gatherings at Stourhead
1825-31 in Memoirs Illustrative of the History and Antiquities of
Wiltshire ... (London, 1851). The names mentioned include
Richard Fenton, George Matcham, Robert Benson, the Revd
Thomas Leman, the Revd Richard Warner, Sir Thomas
Phillipps, the brothers Bowles and others.

105

emerge to move into different social spheres. There
was his parish work as a clergyman and a magistrate
and his interest in simple rustic characters, exempli-
fied in epitaphs for aged parishioners.? At a higher
social level, he and his brother belonged to the small
group of scholars and antiquarians who enjoyed the
hospitality of Sir Richard Colt Hoare at Stourhead
towards the end of his life.** At the same time Bowles
was able to enjoy the wider and more brilliant circle
of the guests at Bowood. The three historical works of
Bowles, Hermes Britannicus .. . (1828), The Parochial
History of Bremhill ... (1828), and Annals and
Antiquities of Lacock Abbey. . . (1835), may readily be
connected with people and places that have been
mentioned. Work could still be done, however, to
expand and explain in greater detail the interest of
Bowles’ career and to identify more of his friends.”’

57. My attention has only recently been drawn to the journals of the
Revd John Skinner, rector of Camerton, near Bath. Skinner
knew Bowles, Colt Hoare and others of the Stourhead circle.
He drew the map of Bremhill and its vicinity facing p.1 of
Bowles’ Bremhill. His journal for the autumn of 1812 mentions
two visits to Bowles: on 29 Oct., when he was shown the garden,
and on 31 Oct., when he dined at the Parsonage and enjoyed a
musical evening in which Mr Lewis West, pastor of the
Moravian Church at Tytherton, and Mr and Mrs Bowles took
part (B.L. Add. MS. 33645 f.36). Bowles’ interest in music is a
subject capable of further study.

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 106-115

A Provisional Checklist of Fossil Insects from the
Purbeck Group of Wiltshire

by A.J.ROSS and E.A.JARZEMBOWSKI

Over seventy species of insects are known from the Wiltshire Purbecks; the fauna ts listed and a selection

is photographically illustrated. The fossil localities in this historically important area, and the work of

the Revd PB. Brodie are summarised.

INTRODUCTION

Just over 140 million years ago, the late Jurassic sea
retreated from southern England and, with the
exception of some brief encroachments, did not
return for some 20 million years until the late Lower
Cretaceous. The geological deposits that accumu-
lated during this interval are predominantly non-
marine and are known as the Purbeck and Wealden
Groups (‘Beds’ of earlier authors). Whilst Dr
Gideon Mantell was looking for dinosaurs in the
Wealden of Sussex, the Revd P.B. Brodie was
collecting much smaller terrestrial animals in the
Purbecks of Wiltshire. Brodie (1839) is the first
account of fossil insects from the Purbecks and these
were found in the Vale of Wardour. Later, Brodie
(1847) recorded insect remains in the Purbecks of
Swindon. The Vicar then went on to become the
most important British collector of fossil insects in
the 19th century. Early in his career, he published a
book on the subject, A History of the Fossil Insects in
the Secondary Rocks of England (1845), in which
many species from the Vale of Wardour are described
and illustrated. Goss (1878: 285) pointed out that
Brodie ‘produced the only book [sic] on fossil insects
which has appeared in this country’ and his remark
could be repeated today.

Nowadays, the student of British palaeoen-
tomology is more likely to find modern information
on the subject dispersed in the foreign literature. We
have therefore attempted to bring together current
information on this historically important material
and summarise it in the form of a species checklist.
Following common practice in geology, we have
tried to include figured as well as named and type
material, based on literature survey and examination
of Brodie’s collection at the Natural History
Museum, London. The study has revealed that
published figures are not necessarily accurate and, in
some cases, species have even been described twice

(Scudder 1886). More exact stratigraphic and
provenance data have been incorporated wherever
possible. In consideration of the former, the listings
by Woodward (1895) have been rejected. This is
because of varying opinions by different collectors as
to the location of the Middle Purbeck/Lower
Purbeck boundary; also, some of Woodward’s ranges
are enigmatic (e.g. references to the Wealden) or
even wrong (attribution of species from the
Archaeoniscus Bed at Dinton to the Lower Purbeck).
However, examination of fossil insect collections at
the British Geological Survey, Keyworth, may well
add useful data in the future.

A more general stratigraphical problem is that
the Purbeck insect fauna straddles the Jurassic/
Cretaceous boundary if the latter is taken convention-
ally at the Cinder Bed in the Middle Purbeck. On this
basis, Purbeck insects from the ‘Lias’ beds and
Archaeoniscus Bed are referable to the late Jurassic and
early Cretaceous respectively (see localities, below).
Allen and Wimbledon (1991) have, however, pro-
posed a neat solution by attributing the whole of the
Purbeck Group to the early Cretaceous (Berriasian) as
a formation, but it remains to be seen if this will be
generally accepted. A likely consensus is that the
boundary will be drawn low in the Lower Purbeck.

Much work remains to be done on the Purbeck
insects, and systematic revision of the cockroaches
(Blattodea) is under way (A.J.R.). This year (1995)
is the 150th anniversary of the publication of
Brodie’s classic book, and we hope this compilation
will help stimulate new interest in his pioneer work.

THE REVD P.B.BRODIE IN WILTSHIRE

Peter Bellinger Brodie (1815-1897) developed an
interest in geology as a teenager, becoming a Fellow
of the Geological Society at the age of nineteen,
prior to entering the church ([Besterman] 1992).

FOSSIL INSECTS FROM THE PURBECK GROUP

107

Figure 1. Reproduction from Brodie (1845), plate 3. Scale c. 6: 5. (For identifications, see checklist)

His first appointment as curate of Wylye was
short- lived (1838-1840); in that time, however, he
was first to discover insects in the Purbecks and
the characteristic sea slater Archaeoniscus brodiei
Milne Edwards which is named after him. He
subsequently served as curate and rector in
Buckinghamshire and Gloucestershire, gathering
further material on fossil insects to produce his
unique book of 1845 which established many
Mesozoic insect genera and species. Privately
published, the book was dedicated to the Revd

Adam Sedgwick, his unofficial instructor at
Cambridge University. Brodie’s early work was
significantly facilitated by Professor J.O. Westwood,
an eminent 19th-century entomologist, who drew
the illustrations (see Figure 1, this paper). It seems
that Brodie had some disagreement with Dr Gideon
Mantell of Sussex who described some Liassic
insect material whilst Brodie was preparing the text
of his book; it may thus be no coincidence that
useful study of Cretaceous insects in southeast
England was delayed until the present century.

108 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

LOCALITIES

Brodie discovered two major and five minor
localities in the Purbecks of the Vale of Wardour. A
field survey by Ms J.B.E. Jarzembowski and E.A.J. in
1978 and 1980 showed that, with one possible
exception, they have long since disappeared. Brodie
(1845:18) described one quarry site as ‘about two
miles south-east of Dinton’. Comparison of his
section with the more detailed accounts of later
geologists (Andrews and Jukes-Browne, 1894)
shows clearly that ‘south-east’ is a misprint for
‘south-west’ and that the site was near Teffont. With
the permission of Sir Edgar Keatinge, E.A. and
J.B.E.J. were able to visit the present exposures in
that area and, like Brodie and Andrews, found insect
fossils in the ‘Jurassic’ part of the Purbecks. These
were in fine-grained, grey limestone called ‘Lias’ by
the old quarrymen. Subsequently, E.A. and J.B.E.J.
found insects in a similar lithology near Ridge.
Another horizon which yielded insects during the
last century was the Archaeoniscus Bed (Isopod
Limestone) at Dinton, in the Cretaceous part of the
Purbecks. E.A. and J.B.E.J. found a few insect
remains in this bed near Dashlet and on Ladydown
but, like the Victorians, concluded that they are rare
at this level.

The first fossil insects from the Purbeck Group
were found by Brodie at Dinton but later workers
failed to find the main horizon or ‘Insect Limestone’
(Andrews and Jukes-Browne, 1894). An excavation
at Dinton sponsored by the Nature Conservancy
Council in 1983 showed that the horizon must le
below the Cinder Bed and was probably a local
development of ‘Lias’. ‘Tantalisingly, Mr A.A.
Mitchell (Gillingham, Kent) has, in 1995, found
insect remains here in weathered limestone. It is
hoped that with the assistance of English Nature, the
old site at Dinton may be reopened following the
withdrawal of the Ministry of Defence.

CHECKLIST

Explanation

The Odonata checklist updates Jarzembowski
(1988). The family classification follows Carpenter
(1992), Clifford, Coram, Jarzembowski and Ross
(1994) and Rohdendorf and Davis (1991). An
asterisk (*) indicates genus excluded from Carpenter
(and not discussed by Clifford er a/., or Rohdendorf
and Davis). Locality (in round brackets) indicates
record by Woodward (1895). Specimen registration
numbers, stratigraphical and additional locality and
systematic data are given in square brackets.

Abbreviations

AB Archaeoniscus Bed (Isopod Limestone)

B’45 Brodie, 1845

D Dinton

det. WRD_ determined by Mr W.R. Dolling

Ii specimen registration number, Natural
History Museum, London

EL; ‘Insect Limestone’ of Brodie

Sc’86 Scudder, 1886

ao Teffont

Vw Vale of Wardour

Order Odonata (Dragonflies)

Aeschnidium antiquum (Brodie);
[IL D] B’45: pl. 5, fig. 10 [I. 3526] Aeschnidiidae

Aeschnopsis perampla (Brodie);
T B’45: pl. 5, fig. 7 [I. 12780] ?Family

?Necrogomphus jurassicus (Giebel)
(T] B’45: pl. 5, fig. 9 [I. 12782, I. 12778] ?Family

Necrogomphus petrificatus (Hagen)
(T] B’45: pl. 5, fig. 8 [I. 12779] ?Family

Order Blattodea (Cockroaches)

Blattula disjuncta (Scudder);
{IL D] Sc’86: pl. 46, fig. 14 [I. 12791] Blattulidae

Ctenoblattina arcta Scudder;
[IL] D Sc’86: pl. 46, figs. 1,2 [I. 12789, I. 12695]
Mesoblattinidae

Elisama kneri Giebel; Figure 2A;
(IL D] B’45: pl. 5, fig. 1 [I. 3528] Mesoblattinidae

Elisama minor Giebel; Figure 2B;
{IL D] VW B’45: pl. 5, fig. 20 [I. 12805]
Mesoblattinidae

?Mesoblattina kollari (Giebel) (=?Mesoblattina eatoni
(Scudder)); Figure 2C;

[IL D] B’45: pl. 5, fig. 14; Sc’86: pl. 48, fig. 19 [I.
12812] Mesoblattinidae

?Mesoblattina recta (Giebel); Figure 2D;
[D] B’45: pl. 5, fig. 3 [I. 12734] Mesoblattinidae

?Mesoblattina scudderiana Handlirsch;
{IL D]'Se’?86: \ pl. ..46, © fig. 13 xf
Mesoblattinidae

12763]

?Mesoblattina sp. Handlirsch;
[D] B’45: pl. 3, fig. 7 [I. 3507] Mesoblattinidae

?Mesoblattina sp. Handlirsch;
[D] Sc’86: pl. 46, fig. 8 [I. 12031] Mesoblattinidae

FOSSIL INSECTS FROM THE PURBECK GROUP 109

Figure 2. Blattodea: A. Elisama kneri Giebel, holotype x8; B. Elisama minor Giebel, holotype x10; C. ?Mesoblattina kollari
: (Giebel), holotype x5; and D. ?Mesoblattina recta (Giebel), holotype x 15

(Jaqety) wmgnp wnipidsouny ‘dD :B12

cx adMoypoy ‘(erporg) 2ys12spas snyJeusoloig ‘q pue £g°9x edAiopoy “Aou “quI0s
idouio {G1 X edMoypoy “({aqary) syrs DUIIDIQOUUD NY “A pure ‘9 [x edAlojoy “([aqety) puurd vuiwjqouunny “Y :eepone[g °¢€ amn3ty

FOSSIL INSECTS FROM THE PURBECK GROUP

?Mesoblattina sp. Handlirsch;
[D] Sc’86: pl. 47, fig. 6 [I. 12744] Mesoblattinidae

?Mesoblattina stricklandi (Brodie);
(IL D] B’45: pl. 4, fig. 11 [I. 3497, I. 3977]
Mesoblattinidae

?Nannoblattina brodiet Handlirsch;
[D] B’45: pl. 5, fig. 16 [I. 12732] Mesoblattinidae

Nannoblattina pinna (Giebel); Figure 3A;
[D] B’45: pl. 5, fig. 5 [I. 11977] Mesoblattinidae

Nannoblattina similis (Giebel) (=Nannoblattina
prestwichu (Scudder)); Figure 3B;

111
[IL D] B’45: pl. 5, fig. 2; Sc’86: pl. 48, fig. 3.
[I. 12810] Mesoblattinidae

?Nannoblattina woodwardi Scudder;
{IL D] Sc’8o: pl. 48, fig. 6 [I. 3501] Mesoblattinidae

Unnamed:

[D] B’45: pl. 5, fig. 6 [I. 12750] Mesoblattinidae
[D] B’45: pl. 5, fig. 11 [I. 12733] Mesoblattinidae

Order Orthoptera (Grasshoppers and Crickets)

Panorpidium dubium (Giebel) comb. nov.;
Figure 3C;

| Figure 4. Phasmatodea: A. Ensiferorum sp., I. 12724, x5; Hemiptera: B. ?Cixioides maculatus (Brodie), holotype x10; and C.

Ricaniites fulgens (Brodie), holotype x5

112 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

[AB D] B’45: pl. 5, fig. 21 [I. 3496] Elcanidae

Protogryllus sedgwicki (Brodie); Figure 3D;
[AB D] B’45: pl. 2, fig. 4 [I. 11990, I. 3525]
Protogryllidae

Zalmona brodiei Giebel;

[IL] Dinton B’45: pl. 5, fig. 13 [I. 3533] ?PHaglidae
{‘Order Phasmatodea (Stick insects)’]
‘Ensiferorum’ sp.; Figure 4A;

[D] B’45: pl. 5, fig. 19 [I. 12724, I. 12723] ?Family
Order Hemiptera (True bugs)

?Aphis plana Brodie;
[D] B’45: pl. 2, fig. 10 [I. 12702] ?Family

Cercopidium lanceolata Heer;
(Wilts) 2Family

Cicadellium pulcher (Brodie);
VW B’45: pl. 5, fig. 17 [I. 12729] ?Family

?Cixioides maculatus (Brodie); Figure 4B;
(IL D] B’45: pl. 2, fig. 8 [I. 3498, I. 3984] ?Family

Genaphis valdensis (Brodie);
{IL D] B’45: pl. 4, fig. 3 [I. 3522] Genaphididae

Fassites punctatus (Brodie);
{IL D] B’45: pl. 5, fig. 4 [I. 3510] Cicadellidae

Lygaeid sp. det. WRD;
[D] B’45: pl. 2, fig. 11 [1. 3517]

Psychodites egertom (Brodie);
{IL D] B’45: pl. 4, fig. 7 [I. 3502, I. 12623] ?Family

Psychodites kenngotti (Giebel);
[D] B’45: pl. 4, fig. 8 [I. 12691] ?Family

Figure 5.Neuroptera:
unnamed kalligrammatid, Lulworth Formation, Teffont, Revd. W.R. Andrews’ coll. no. 15, Devizes Museum, x3

FOSSIL INSECTS FROM THE PURBECK GROUP

Ricanutes fulgens (Brodie); Figure 4C;
{IL D] B’45: pl. 4, fig. 12 [I. 3505, I. 3995] Ricaniidae

Order Neuroptera (Lacewings); Figure 5

Order Coleoptera (Beetles)

Anapiptus brodiet Handlirsch;
{IL D] B’45: pl. 3, fig. 4 [I. 3515] 2?Family

Apistotes purbeccensis (Giebel);
VW B’45: pl. 6, fig. 6 [I. 11963] ?Family

Cerylonopsis striata (Brodie);
{IL D] B’45: pl. 3, fig. 1 [I. 3514] ?Family

Coleopteron rugostriatus (Giebel);
[T] B’45: pl. 6, fig. 2 [I. 11959] ?Family

Coleopteron spp.;

[D] B’45: pl. 6, fig. 1 [I. 3534] ?Family;
[D] B’45: pl. 3, fig. 5 [I. 3529] ?Family
Coleopteron vetustus Giebel;

[IL D] B’45: pl. 3, fig. 3 [I. 3516] 2Family
Diaperidium mithrax Westwood;

(VW) ?Family

Helophoropsis brodiet (Giebel);

{IL D] B’45: pl. 3, fig. 2 [I. 3524] ?Family
?Helopidium brodiei (Giebel);

VW B’45: pl. 6, fig. 5 [I. 11962] ?Family

Helopidium westwoodi (Giebel);
{IL D] VW B’45: pl. 6, fig. 3 [I. 3500, I. 3946]
?Family

HAydrobuites purbeccensis (Giebel);
[IL D] B’45: pl. 6, fig. 12 [I. 11966] ?Family

Hyperomima antiqua (Giebel);

[IL D] B’45: pl. 6, fig. 4 [I. 3504] ?Family
Kakoselia angliae (Giebel);

[IL D] B’45: pl. 6, fig. 8 [I. 3528] 2Family

Kamaroma breve Handlirsch;
{IL D] B’45: pl. 6, fig. 14 [I. 11968] ?Family

| Katapontisus brodiei (Giebel);

VW B’45: pl. 6, fig. 9 [I. 11964] ?Family

Omma elongata (Brodie);

| [ML D) B45: pl. 2, fig. 1 [I. 3527, 1 12149]

Cupedidae

Pseudocymindis antiqua (Giebel);
VW B’45: pl. 6, fig. 10 [I. 11965] ?Family

Suctulus brodiet Handlirsch;
{IL D] B’45: pl. 6, fig. 11 [I. 3519] ?Family

Tychon antiquum (Giebel);
VW B’45: pl. 6, fig. 13 [I. 11967] ?Family

Order Mecoptera (Scorpionflies)

Orthophlebia bifurcata Giebel; Figure 6A;
{IL D] B’45: pl. 5, fig. 12 [I. 3532] Orthophlebiidae

Stenopanorpa gracilis (Giebel); Figure 6B;
[D] B’45: pl. 5, fig. 18 [I. 12721] ?Family

Order Diptera (True flies)

‘Aphis’ dubia Giebel;

{IL D] B’45: pl. 2, fig. 9 [I. 3530] ?Diptera
*Asuba dubia (Brodie);

{IL D] B’45: pl. 3, fig. 10 [I. 3545] ?Family
*Bibionites prisca (Giebel);

VW B’45: pl. 5, fig. 15 ‘Fungivoritidae’
*Bria prisca (Brodie)

(IL D] B’45: pl. 4, fig. 10 [I. 3503] ?Family

Chironomopsis arrogans (Giebel);
[IL D] B’45: pl. 3, fig. 14 [1. 3493] ?Family

Chironomopsis extinctus (Brodie);

[IL D] B’45: pl. 4, fig. 5 [I. 3520, I. 12757] 2?Family
*Dara fossilis (Brodie); Figure 6C;

(IL D] B’45: pl. 3, fig. 15 [I. 3509] ?Family
*Hasmona leo Giebel;

[D] B’45: pl. 3, fig. 11 [I. 12751] 2Family

Olbiogaster fittoni (Brodie); Figure 6D;
[IL D] B’45: pl. 3, fig. 9 [I. 12753] Anisopidae

*Pseudosimulium humidum (Brodie);

{IL D] B’45: pl. 3, fig. 8 [I. 3952] ?Family
Remala sphinx Giebel;

{IL D] B’45: pl. 4, fig. 4 [I. 12711] ?Family

*Sama rustica (Brodie);
[IL D] B’45: pl. 3, fig. 13 [I. 3495] ?Family

*Sciophilopsis brodiet Handlirsch;
{IL D] B’45: pl. 4, fig. 2 [I. 3975] ‘Fungivoritidae’

‘Termes’ grandaevus Brodie;

WEeDIP B45 plse2, fig: 5, [Ey 3512, 112703]
“‘Bibionidae’

*Thimna defossa (Brodie);

[IL D] B’45: pl. 3, fig. 12 [I. 3586] ‘Fungivoritidae’

gx adAlojoy “yosippuey 2a7po.ig vigajydoyluopnasy “Ay versidoysizy, pue SQ [x edMojoy “(arporg) sor uaispso1giO ‘q pue
‘zx adAojoy ‘(e1porg) sypssof Dawg “> setardiq Sox adArojoy “(jaqaty) syiDus vdoundouary “gq pure écx adMojoy “Jagan vmwoinfig viqajydoyWOQ “WY reraidosayy °9 o1n317

THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

114

FOSSIL INSECTS FROM THE PURBECK GROUP

Order Trichoptera (Caddisflies)

Pseudorthophlebia brodiet Handlirsch; Figure 6E;
{IL D] B’45: pl. 2, fig. 7 [I. 3551, I. 4002] ?Family

‘Flata’ haidingeri Giebel;
{IL D] B’45: pl. 2, fig. 6 [1.3494] ?Trichoptera
Insecta incertae sedis

‘Diechoblattina’ ungeri (Giebel)

‘Meloe’ hoernesi Giebel;
{IL D] B’45: pl. 2, fig. 12 [I. 3508]

‘Philonthus’ kneri Giebel;
{IL D] B’45: pl. 2, fig. 2 [I. 3511]

Prognatha crassa Giebel;
[D] B’45: pl. 2, fig. 3 [I. 3978]

Unnamed
VW B’45: pl. 3, fig. 6 [I. 3521, 2>Hemiptera]
[D] B’45: pl. 4, fig. 1 [I. 3506, ?Diptera]

[D] B45: pl. 4, fig. 6 [I 3513, I.
?Hemiptera]

12759,

VW B’45: pl. 4, fig. 9 [I. 3531, ?Hemiptera]

Acknowledgements. We would like to thank Messrs Crabb, Richens
and Taylor (NHM) for photographic help. This is P. R. I. S.
Contribution No. 436 for E. A. J. ;

Addendum. Since this paper was prepared, the BGS collections have
been examined by AJR and contain no Wiltshire Purbeck insects;
the unsupported records by Woodward (1895) are therefore open to
doubt. The Purbeck and Wealden Groups are now referred to as
Purbeck Limestone Group and Wealden Supergroup, respectively.
Zalmona is ?Prophalangopsidae following A. V. Gorokhov and
Psychodites is a protopsyllidiid Simopsocidium according to D. Y.
Shcherbakov. The last entry in Insecta incertae sedis above was
originally referred to ‘?Cercopis larva’ Figure 3D is I. 3525; 4B,
I. 3498; 4C, I. 3505; 6E, I. 3551.

IMS:
REFERENCES

ALLEN, P. and WIMBLEDON, W.A., 1991 Correlation of
NW European Purbeck-Wealden (nonmarine Lower
Cretaceous) as seen from the English type-areas,
Cretaceous Research 12, 511-526

ANDREWS, W.R. and JUKES-BROWNE, A.J., 1894 On the
Purbeck Beds of the Vale of Wardour, Quarterly Journal
of the Geological Society of London 50, 44-71

[BESTERMAN, T.P.], 1992 The Reverend Peter Bellinger
Brodie: Geologist, pamphlet, The Warwickshire
Museum, Warwick

BRODIE, P.B., 1839 A notice on the discovery of the remains
of insects, and a new genus of isopodous Crustacea
belonging to the family Cymothoidae in the Wealden

{Purbeck] Formation in the Vale of Wardour, Wilts,
Proceedings of the Geological Society of London 3, 134-135

1845 A History of the Fossil Insects in the Secondary Rocks of
England, xvii + 130pp., 11pls., Van Voorst, London

1847 Notice on the existence of Purbeck strata with the
remains of insects and other fossils, at Swindon, Wilts,
Quarterly Fournal of the Geological Society of London 3,
53-54

CARPENTER, F.M., 1992 Hexapoda. Treatise on Invertebrate
Paleontology, Part R, Arthropoda 4, 3 and 4: 677 pp.

CLIFFORD, E., CORAM, R., JARZEMBOWSKI, E. A., and ROSS,
A.J., 1994 A supplement to the insect fauna from the
Purbeck Group of Dorset, Proceedings of the Dorset
Natural History and Archaeological Society 115, 143-146

GOSS, H., 1878 The insect fauna of the Recent and Tertiary
periods, and the British and foreign formations of these
periods in which insect remains have been detected,
Proceedings of the Geologists’ Association 5, 282-343

JARZEMBOWSKI, E.A., 1988 British dragonflies in the latter
part of the age of dinosaurs, Journal of the British
Dragonfly Society 4, 1-8

ROHDENDORF, B. B., and DAVIS, D.R. (eds.), 1991
Arthropoda, Tracheata, Chelicerata, Fundamentals of
Paleontology 9: xxx1 + 894 pp.

SCUDDER, S.H., 1886 A review of Mesozoic cockroaches,
Memoirs of the Boston Society of Natural History 3,
439-485, pls. 46-48

WOODWARD, H.B., 1895 The Jurassic rocks of Britain, V.
Middle and Upper Oolitic rocks of England (Yorkshire
excepted), Memoirs of the Geological Survey of the United
Kingdom, London

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 116-129

The Common Stinging Nettle (Urtica dioica L.) in
Wiltshire, 1994

by JACK OLIVER

Nettles have always been common in Wiltshire, but are now more ubiquitous and abundant than 40

years ago, when they were the seventh commonest flowering plant in the county. Urtica dioica is now

possibly Wiltshire’s second most common flowering plant. The results of four main sets of investigations

by the author and members of the Wiltshire Botanical Society on the abundance of common stinging

nettles (Urtica dioica) in Wiltshire are detailed. Special aspects and implications are discussed in

relation to land use and biology.

INTRODUCTION

Two species of stinging nettle are detailed in The
Wiltshire Flora (Gillam 1993), the common nettle
(Urtica dioica) and the small nettle (UL urens).
The first is an abundant plant in Wiltshire, and
was the subject of widespread media coverage
nationally in October and November 1994. This had
followed an article on Wiltshire nettles in New
Scientist which was concerned with the frequency,
size, spread, some aspects of the biology, and the
probable ecological significance of nettle abundance
in Wiltshire (Pearce 1994). Widespread over-
enrichment with phosphates was implicated.

The New Scientist special feature was based on
work by the Wiltshire Flora Mapping Project
(WEFMP), the Wiltshire Botanical Society (WBS)
the Wiltshire River Monitoring Scheme (WRMS),
and the author. This article brings together the
relevant publications and the combined evidence of
increased prevalence for one of England’s, and
certainly Wiltshire’s, most ubiquitous, successful
and frequently monopolistic flowering plants.

BIOLOGY

Wiltshire has five members of the Urticaceae
(Nettle family). These are the two aforementioned
species of stinging nettle, pellitory-of-the-wall
(Parietaria judaica), mind-your-own-business/baby’s
tears/hundreds of thousands (Soletrolia/Helxine
soleiroiz) introduced from Corsica or Sardinia, and a
third Urtica, the very rare stingless nettle (Urtica
galeopsifolia, recently found at new sites) which may
occur in only three or four places in Wiltshire,

perhaps mainly in hybridised forms back-crossed
with U. dioica (Last 1995).

U. dioica is ‘A coarse hispid perennial 30-150 cm.
Roots much branched, very tough, yellow. Stems
creeping and rooting at the nodes .... Leaves
opposite, 4-8 cm... coarsely serrate . . . inflorescence
up to 10cm... flowering June—Aug.. .’. (Clapham,
Tutin and Moore (CTM) 1987). It spreads by roots,
rhizomes, stolons, rooting stems and seeds. The
flowers are small, green, unisexual with four
perianth segments, with male and female usually in
different plants (dioecious). Dioecy in the Urticaceae
has evolutionary significance (Lahav-Ginott and
Cronk 1993). U. dioica is nearly always tetraploid (2n
= 48 or 52), and may not be the original native
British plant. Our most vigorous nettles may have
evolved from complex hybridisations involving U.
dioica and U. galeopsifolia (2n = 26) (Q.C.B. Cronk
pers. comm.; Geltman 1992). Wheeler (1995) refers
to 30 species of insect specifically associated with U.
dioica, which is also the food plant for the larvae of
five beautiful Vanessid butterflies: Red Admiral,
Painted Lady, Small Tortoiseshell, Peacock and
Comma. Nettles are so resistant to fungal infections
that they are being used in genetic engineering to
confer fungal resistance on the taxonomically related
English Elm, to protect against Dutch Elm Disease.
Nettles are, however, associated with 50 species of
microfungi, 11 of which are specific to U. dioica (K.
Wheeler pers. comm.).

U. dwica is a typical monopolistic plant, one
which excludes other species (Crawley 1989). Whilst
all authors agree on the importance of the dense
leafy shading created by nettlebeds, the emphasis
varies between tough branched roots (CTM 1987)

COMMON STINGING NETTLE IN WILTS 1994

and rhizomes (underground or sub-surface stems)
(Stace 1991), or stolons/creeping surface stems
(CTM 1987) as the reason for out-competing other
plants at and below ground level.

The nutritive importance of nitrates has in recent
years given way to emphasis on phosphates as
highly influential in the spread and vigour of nettles.
New patterns of agricultural, roadside, riverside
and woodland management also may be crucial
considerations, as are the influences of soil disturbance
on seedling survival and the scatter of viable shoot,
root, rhizome and stolon fragments.

INVESTIGATIONS IN WILTSHIRE

The main studies undertaken concerned measures
of frequency in relation to biology and land use.
Study I was an intensive series of counts and

LUT,

observations within 9sq.km; II involved measures
along 250 miles of Wiltshire roads; HI took
numerous river sites; and IV, sites along the entire
length of the Kennet and Avon Canal. Study V
highlights some special features contributing to the
vigour, spread and frequent dominance of nettles.

Study I: all routes in a 9km square, and
surrounding areas

Every path, track, road and riverbank was walked in
July 1994 within the 9 x 1km square centred on
Lockeridge, west of Marlborough (SU 145675; see
maps, Figures 1 and 2). Actual nettle counts were
made per 100 metres, applying to either or both
sides, including nettles insinuating flanking fences,
ditches, riverbanks and hedges (Oliver 1995).

Lockeridge
Dene NT

Lockeridge

= =~. ef :
abide Zegiees

River Kennet x > \
oe . Clatford

Jy
i!
Wh
eo
\ W,
\ M7
la
\
‘
\
\
\
\
\
N
\
BAN S|
Weta Gara} /
\ V4
\ /
\ /
\ /
1 /
/
a
eu H A f
Signs \ J} /

i Vg? Yes

: \
iy Sy

Figure 1. Nettle survey route map, centred on Lockeridge, west of Marlborough (SU 145675). Scale: 4mm = 100m

Key: roads; ------

tracks or rights of way

118 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Lockeridge
Dene NT
5

Figure 2. Nettle density route map, July 1994. Scale: 4mm = 100m
Key: no coded numbers on routes: no nettles; 1: 1 nettle (or 2 seedlings) per 100m (either side); 2: 2-10 nettles per 100m;
3: 11-100 stems per 100m; 4: 101-1000 stems per 100m: (semi-interrupted) fringes on or near both sides of route;
5: 1001+ stems per 100m: (semi-interrupted) bands 1.5m thick on or near either side (or sometimes obstructing route or
invading river bed), or substantial adjacent nettlebeds

Horton (1975) refers to three roadside habitats: the
1—10-metre wide cut grassy verge, uncut vegetation,
and the hedge, scrub or fence borders on the far
side. Roadside habitats could be more complex than
this; for instance, the A4 has a cycle path on the
south side, and cutting can be close, roughly
strimmed or intermittent. The general rule applied
was to consider 5 metres out from road verge, path
or river edge. More than 1000 stems per 100 metres
indicates either (semi-)interrupted nettle bands 2 or
more metres thick, and/or adjacent nettlebeds.
From 100 to 1000 stems per 100 metres indicates
either the (semi-)continuous fringing curtain of
nettles by hedge/fence/path edge; or (semi-)

continuous infiltration by nettles of coarse-cut grass
or coarse vegetation; or interrupted but fairly
frequent denser clumps, or re-sprouting strimmed
stalks and stolons.

Forty-six out of the total 53km were nettled to a
greater or lesser extent (Table 1, map, Figure 2 and
Oliver 1995). Most flanking fences had between 10
and 1,000 nettle shoots per 100 metres, and the
same applied to the roadside hedges (usually
hawthorn, compare Grose 1957, 700-03). Some
tracks and riverbanks were so overgrown with nettles
that stinging was inevitable, without protection,
from some of the 2-metre high riverside and wet
woodland nettles.

COMMON STINGING NETTLE IN WILTS 1994

119

Table 1. Prevalence of U. dioica along all routes within 9 x 1 km square centred on Lockeridge, west of Marlborough
(SU145675). Lengths in kilometres. (Reproduced with permission from B.S.B.I. News 68.17 (Oliver 1995))

COUNTED, PER 100 METRES FOR ENTIRE LENGTHS

Nettle-free 1 nettle or
seedling

River Kennet banksides 0 0
A4, fringes to verges 0 0
Minor roads, fringes to verges 0.25 0.50
Track-sides, rights of way 1.25 0.35
(villages, farmland, N. Trust)
Track-sides, rights of way 5.75 4.25
(West Woods)
All routes lend 5.10

Wayside nettles did not appear in the following four
habitats:

1. deep-shaded paths through the bluebell expanses
under dense beech (Fagus sylvatica), Thwa or
spruce plantations in West Woods;

. some paths across (in sharp contrast to alongside)
fields, in the absence of sarsens, fences, ditches,
dumps, wire or obstructions where nettle
rhizomes and stolons abound;

. sections of intensively mown, strimmed,
weeded and/or weed-killed verges; and

. undisturbed mixed woodland with a ground
cover of bracken, bramble, Dryopteris spp., or
large woodland grasses, as alongside the ancient
Wansdyke West Woods path.

nettle

In general, the dense

concentrations were:

most wayside

1. where paths, roads or river ran next to agricul-
tural land, especially near cattle, farm buildings,
etc.;

. on paths and roads between fields (compare item
2 of the preceding list);

. in, alongside and near ditches: the ground
between ditch and track/hedge/path/road was
usually dominated by nettles; ditches and hedges
functioned as permanent reservoirs from which
nettle stolons and rhizomes re-colonised mown,
strimmed, grazed and weed-killed areas;

. on paths near dumps and waste areas; dumps
also served as permanent reservoirs for nettle

2-10

0.50
0

0.90
0.50

3.00

4.90

up to 100 101-1000 1000+ All nettled Total

plants stems stems lengths lengths
0.75 E25 2:25 4.75 4.75
0 2.00 1:25 3.25 3:25
L225 6.00 3.10 11.75 12.00
1.50 3.85 6.40 12.60 13.85
2.00 1:75 2.25 13.25 19.00
5.50 14.85 15.25 45.60 52.85
rhizomes and stolons, from which nettles

colonised adjacent areas;

. ON many riverbanks, quite often fringing the
immediate water’s edge; again, nettlebeds were
most dense where cattle fences or ditches were
close to the river;

. on some damp tracks through and alongside
disturbed areas in and around West Woods;
often seedlings were seen after tree felling, but
the dense nettlebeds could appear in subsequent
years before the beech or spruce canopies
became dense again (compare items | and 4 in
the preceding list);

. on paths near rabbit warrens and badger setts
(compare Grose 1957, 714); and

. on paths through or alongside Piggledene
National Trust (sarsen) reserve (at least in July
1994).

Nettle stolons can be found under rough grass from
late December to March, but vertical nettle growth
starts to dominate other vegetation from late May,
and is usually obvious until early December unless
there have been a number of early frosts. Competing
herbaceous vegetation conspicuous in this study
included: cow parsley (Anthriscus sylvestris, to
150cm), which dominated stretches of roadside (but
seldom tracks or paths) from April to early June;
false oat-grass (Arrhenatherum elatius, 50-150cm
high when uncut), the dominant rough roadside
verge grass; rough meadow-grass (Poa trivialis, 70-
90cm high); cocksfoot grass (Dactylis glomerata,
150cm uncut); and large vigorous agricultural
strains of a perennial rye-grass (Lolium perenne).

120 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Sometimes any of these, including nettles, could be
invaded by brambles, docks or thistles, or become
festooned in high summer by cleavers (Galium
aparine) and/or hedge bindweed (Calystegia sepium).

Small grasses (such as annual meadow-grass, Poa
annua) and small or rosette wayside herbs such as
dandelion, daisy, creeping buttercup, clovers, knot-
grasses (Polygonum aviculare and P arenastrum),
plantains (Plantago major and P lanceolata), may-
weed, silverweed and shepherd’s purse (Capsella
bursa-pastoris) were still common on stony edges and
in closely and regularly manicured grass verges.
However, these little wayside plants were always
pushed to the edges by the tall wayside species listed
above and, in these nine square kilometres, were no
competition for either the big wayside and
agricultural weeds in general, or nettles in particular.

Both literally and metaphorically, it would have been
narrow to ignore the abundant nettling away from
the routes in these nine square kilometres, even
though no systematic counts were made. Nettles
were frequent, abundant or continuous in most
hedges, by most field-sides and ditches, and usually
abundant on and around farmland, waste areas,
dumps and rabbit warrens in general. There were
huge nettlebeds on National Trust land, especially
Piggledene, and nettles surrounded or infiltrated
sarsens and fencing. Some downland localities were
invaded by nettles, especially when cattle-trodden.

The dense beech tree cover in the bluebell areas
of West Woods precluded nettles, as did the spruce
and Thuja plantation shading, but nettles surround
West Woods and often became abundant in tree-
felled areas as well as dominating many edges,
clearings and open forestry tracks. Bluebells rely on
rapid spring photosynthesis before the beech trees
come into leaf and cast too deep a shade for nettles
to persist.

White dead-nettle is very common in Wiltshire as
a whole, and tends to persist (and even flower)
throughout the winter in this 9 km square, including
West Woods. Another two members of the mint
family, ground ivy (Glechoma hederacea) and the
naturalised silver-leafed archangel (Lamuastrum
galeobdolon ssp. argentatum) covered large areas of
ground in and at the edges of West Woods. These
three plants are mentioned because they all can
persist as an under-storey beneath the summer
canopies of stinging nettles. They all seem to have a
five-week window of rapid photosynthesis and/or

growth in March and April before becoming
overtopped by May or June. They therefore can
seem to be the dominant ground cover in winter and
spring but disappear entirely from view beneath
stinging nettles in summer and autumn. A similar
seasonal transition occurs with various grass species
seemingly dominant in winter, but overtopped in
high summer and autumn by the nettles.

Some fields had fairly pure crops, and other
pasture fields were dominated by almost pure mono-
cultures of vigorous strains of perennial rye grass
(Lolium perenne, and the hybrid Lolium x boucheanum),
which thrive on soil enrichment. However, many
fields had numerous blebs of nettles within the
pasture grass. This feature is illustrated, without
comment, for Sheepscombe in the Cotswolds on the
front cover of the journal British Wildhfe (du Feu
1994). In other fields the nettles were dominant and
TV cameras in October 1994 tracked two of the four
vistas south of the A4, down embankments into the
Kennet Valley. Fringing nettlebands south of the A4
were linked by vast agglomerations of pasture
nettlebeds, and pastures heavily infiltrated by cattle-
tramped nettles, to the dense ribbons of nettles
fringing the River Kennet. Many hectares at each
vista were severely affected.

Study II: 250-mile Wiltshire roadside nettle
survey

Study I was very intensive but limited to a small part
of Wiltshire. As the Lockeridge area could have been
rather exceptional, an area was surveyed further
afield, and this time by car. Horton (1975) referred
to 2,365 miles of metalled roads (excluding the M4
motorway) in Wiltshire. He estimated that the
associated verges represented 5,000 acres of roadside
habitat, mainly grassland. One-tenth of the total
mileage seemed a reasonable sample, providing that
measurements were continuous, unselective and
geographically diverse.

Some verges, especially on certain stretches of A-
roads and bypasses, could be very broad, so 10
metres out from the road edges or roadside paths
were allowed. As in Study I, lengths with the thin
fringe of nettles in boundary fences or hedges and/or
nettles infiltrating the grassy verges, as well as the
more conspicuous roadside clumps were counted.
Two mileocmeters, continuous and resetting, were
used to measure road verge-side nettling along a
spread of Wiltshire’s main and some minor roads,
but not including the M4. Actually, outside urban
areas, the reverse process was easiest: measurements
of any one-tenth of a mile stretch without nettles on

COMMON STINGING NETTLE IN WILTS 1994

ld
V {
os Sarna
Swindon
Malmesbury Ss;
V
a
J
i
a
ast
q ie R.Kennet
Yap D y “ar avg 4
aS Hes

ie Marlborough ay
ty ‘ oe sites Sy W
a Melksham vV ay

Ww vi q Wey

« ww LA Voy, ¥
K&A Canal G Wy 44 y

Trowbridge

) Warminster

Figure 3: Map showing nettle survey sites in Wiltshire, including 9km square for Study I (west of Marlborough).
Key: V sites; V including subsites surveyed by Dec. 1994

122 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

either side. It was not difficult in October 1994, even
when some nettles had been blackened by an early
frost, and others had been strimmed; no mistakes
were made even at 20-40 miles an hour, in
discriminating between the very common white
dead-nettle (Lamium album) and occasional patches
of hedge woundwort (Stachys sylvatica), even when
these vegetatively-similar plants were not in flower.
Annual dog’s mercury (Mercurialis annua) on
roadside stonework or in gutters could look like a
small frosted nettle if not seen clearly, and caused
several awkward stops in towns and villages.

‘There were six main sets of data gathered from
the following routes (the towns being shown on the
map, Figure 3):

1. A4 road from east of Savernake Forest (east of
Marlborough) to near the county boundary west
of Chippenham. Contiguous minor roads east of
Marlborough and west of Chippenham.

. A363 north-west of Bradford-on-Avon, south-
east through Trowbridge, Westbury, Warminster
(bypass); A36 to Wilton west of Salisbury.
Minor roads around Bradford-on-Avon.

3. A3094 south-west of Salisbury. Minor roads
north, south and west of Salisbury.

4. Main roads around Devizes: A360, A361 (to
the A4), part of A342. Minor roads linking
these.

5. A429 Chippenham to Malmesbury; A420
towards Swindon. Minor roads linking these.

6. A345 and A338, both linking Marlborough and
Salisbury; A338 and A36 south and south-east
of Salisbury. Contiguous minor roads south of
Marlborough, and south-east of Salisbury and
to south Wiltshire borders.

‘Table 2 shows the results. Only the A363/A36 route
(item 2 above) had less than two-thirds total roadside
lengths nettled to a greater or lesser extent. There

were usually considerable and varied lengths of
dense roadside nettling, especially on the minor
roads. Nettle bands and fringes tended to be set back
beyond the grass verges on most of the A-roads. For
these, and along B-roads, the most continuous plant
was false oat-grass (Arrhenatherum elatius). Also very
common were perennial rye-grass (Lolium perenne)
and cocksfoot (Dactylis glomerata). Occasionally,
fescue and meadow grasses (Festuca and Poa spp.)
dominated the most carefully manicured verges. This
was an autumn survey, but non-flowering shoots of
cow parsley (Anthriscus sylvestris), a conspicuous
spring-time road verge plant, were easily identifiable
from the car in October. Nettles were often dominant
right up to the tarmac on some minor roads and
lanes and were, in 1994, more continuous on all
roadsides than ribwort plantain, cocksfoot grass,
dandelion, smooth meadow-grass (Poa pratensis) and
cow parsley (compare Grose 1957, 723).

The five main nettled roadside situations were:
1. infiltration of coarse grass verges;

2. fringing verge edges, hedges and fences;

3. clumps and bands semi-continuously mono-
polising other roadside vegetation;

4. roadside ditches, nettles spreading out to
dominate continuously all other species
together; and

5. scattered or dotted individual roadside nettle
plants.

Nettle-free lengths were mainly:

1. urban/suburban/village locations (although nettle
seedlings and plants could often be seen on
stonework, by neglected gardens and in streets);

2. past industrial estates with close-mown lawns (as
for parts of route 2 preceding, A363 industrial
estates);

‘Table 2. Wiltshire road verge-side nettle survey. Length in miles for A-roads (A) and minor roads (BC)

IA IBC 2A 2BC 3A 3BC 4d 4BC 5A SBC 6A

ROUTES
6BC AllA Al BC All
Roads

Lengths studied 32.25 6.75 35.5 7.5 4 17 26

Nettled lengths 28.25 6.5 22 6
% nettled 88 96

17.25 13 8.25 65
3.515.5 21.5 15.25 11 7.5 56.25 16.5 142.5. 67.25 209.75

62 80 88 91 83 88 85 91 87 93 81 90 84

17.75 175.75 74.5 250.25

COMMON STINGING NETTLE IN WILTS 1994

3. some bypasses or other embankments with thin
soil, or thin scrub showing chalk, away from
agriculture (again mainly route 2); and

4. a few dark-shaded roadsides dominated by ivy.

Study III: river surveys
IIIA

In 1994 Wiltshire River Monitoring Scheme (WRMS)
workers completed simple plant checklists for 32
riverside sites. The WRMS had been set up because
of concern about the state of Wiltshire rivers,
especially those drying out. Measurements included
flow rates, depths, chemical analyses and invertebrate
counts, together with a plant survey for each site.

The common nettle was found to occur at 31 out
of the 32 sites (97%). It was not only more prevalent
than any other single riverbank/riverside species, but
surpassed combined counts for groups such as
‘thistles’ which comprise four common and several
less common riverbank species. This 97% corre-
sponded exactly to the much more detailed results of
the 1992-4 Wiltshire Botanical Society (WBS)
quantitative research covering 119 sites and 253
subsites throughout Wiltshire (see map, Figure 3
and Study IJIB ensuing).

IIIB

The barriers of nettles around. village and farm
ponds, lake margins, canal banks and even dewponds
were familiar obstructions during the Wiltshire Flora
Mapping Project (WFMP) surveys in the six years
before publication of the 1993 Wiltshire Flora. For
each of the 119 sites and 253 associated subsites (see
Figure 3) the WBS survey (1992-4) used the simple
5-point scale as used by Grose (1957, 677-79) to
assess the relative frequencies of all riverbank and
channel vascular plant species: A = Abundant (10), F
= Frequent (6), O = Occasional (3), P = Present (1)
and Absent. Areas and methods were comparable to
Grose’s studies, but WBS workers usually credited
more A scores than Grose. Reasons for this include
four seasonal visits and the three habitats (at least) at
every site: water (or channel), water-margin, and

| vertical, sloping or marshy riverbanks. Water-margin

emergent aquatics such as water mint (Mentha
aquatica) or water forget-me-not (Myosotis scorpioides)

_ might only become dominant in autumn, and like any

of the common reed or flote-grass species, could
invade channel and/or river banks depending on

_ varying water levels. Nettles do not like their roots

123

continuously in water, when the leaves go yellow, but
thrive on intermittent inundation. They could
therefore infiltrate river margin vegetation in cow-
poached meadows or down steep or vertical slopes,
and invade the channels of dried-out winterbournes,
but did less well in riverside marshland. Further up
the banks, there could be a seasonal sequence of
dominant grasses flowering from June to September:
rough meadow-grass (Poa trivialis), yorkshire fog
(Holcus lanatus), couch (Elytrigia repens) and creeping
bent (Agrostis stolonifera). Any of these, or any one of
the three common reeds could dominate one bank at
different times in summer or autumn. Another bank
might be a grazed perennial rye-grass (Lolium
perenne) river margin. On the other side of the bridge,
one bank might be 90% nettles draped with cleavers
(Gahum aparine) or bellbind (Calystegia sepium). The
fourth bank might be in deep shade with ivy ground
cover. However, the same site in January to April
could have one or more banks dominated by
snowdrop, celandine or cow parsley in turn.

From May to November, nettles were ubiquitous,
poking through riverside brambles, competing with
ivy, terrestrial or aquatic grasses, infiltrating reeds,
amongst stands of great willow-herb (Epilobium
hirsutum), or branched bur-reed (Sparganium erectum),
on stonework, lining water margins, dominating slopes
and tops of banks, forming narrow semi-continuous
fringes, broad parallel bands or dense nettlebeds
between field fences and water margins (Oliver 1993
atb, 1994 a+b, 1995). Table 3 shows that U. dioica is
now the dominant riverbank and riverside species in
Wiltshire, both in distribution and abundance. This
is true for little (sometimes ditch-like) tributaries,
and also beside the Thames, both Avons and the
Kennet where these are greater than 10 metres
across. Some 97% of the sites and 97% of the
subsites were nettled to a greater or lesser extent, and
in 67% of the sites, the stinging nettle was either the
most common, or equally most common riverbank/
riverside species. No other riverside or riverbank
plant, however abundant over limited stretches, now
approaches the repeated frequencies or abundance of
nettles. Second to eighth places in frequency go to
cleavers, reed canary-grass, great willow-herb, rough
meadow-grass, false oat-grass, creeping bent and cow
parsley (cf. also Grime ez al. 1989 for nutrient-loving
species associated with nettles).

Table 3 and map, Figure 3, show that there were
Many upstream sites studied where the channel was
less than 10 metres across, including winterbournes.
Three types are here mentioned:

124 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

‘Table 3. Nettles at river sites (and subsites) throughout Wiltshire, and their frequency relative to the other
riverbank/riverside species, 1992-1994. See also map, Figure 3

River systems Widths Nos. of Nettles
(metres) sites Present Absent Ist

Salisbury Avon

Orders of nettle commonness per site

equal

and tributaries <10 40 38 2 10
Bristol Avon 10+ 15 15 0) 6
and tributaries <10 20 18 2, 5
Kennet and 10+ 19 19 0) 10
tributaries <10 20 20 0 12
Thames 10+ 3 3 0) 1
Totals 119 115 4 45

Subsites
Nos. of — Nos. %
subsites nettled _nettled

2nd 3rd 4th Less
frequent

3 100
58 55 95

31 30 97
33 30 91

1. channels traversing old-fashioned water meadows
or marshes - such channels are now very
uncommon;

. steeper-sided channels or ditches, created to
improve agricultural drainage and intermittently
flowing hard only after storms or heavy winter
rains or snow melts (very common); and

3. grassy winterbournes, often cattle-poached,
mapped as rivers but dry for half to three-
quarters of the year (very common).

Nettles were only occasional alongside 1, common in
the grassy banks of 3, but invariably superabundant
for 2. The uppermost 13? miles of the River Kennet
in summer and autumn have been described as
converted to ‘... huge, almost continuous, double
ribbons (or broad single ribbons where the channel is
invaded) of dense stinging nettles. Broken ribbons of
riverside nettle growth even survive on pasture treated
with selective weedkillers favouring grasses (Oliver
1991a). This was the case for nearly all ditch-like
upper tributaries, even when moderately (but not
totally) shaded.

Study IV: canal survey

The Kennet and Avon canal is 65km (40 miles) long
in Wiltshire; 44 sites and several lengths between
them were studied in 1994 (see map, Figure 3). All
sites but one were divided into four subsites, making
173 in all, each about 200 yards long. Thirty-seven

of the 44 sites were by bridges or wharfs, and seven
were simple lengths.

On the non-towpath side, between May and late
November, nettles often fringed the water and
commonly insinuated other vegetation up the bank
to the fences or field boundaries. On the towpath
side, nettles again often fringed the water and were
common up the bank, but were usually separated by
the grassy or hard towpath (and grassy verges) from
the strip of rough ground with coarse vegetation
(often nettled, even if strimmed) and another fringe
of taller vegetation around fences, hedges or other
boundaries, where nettles were again frequent or
often abundant. Bridges, locks and wharfs were
often closely manicured for 100 yards or so, but with
nettles frequent beyond and outside the mown or
weed-killed walks and working areas. The very
common mown grass on towpaths and by locks,
wharfs and bridges was usually perennial rye-grass
(Lolium perenne). Common water’s edge vegetation
included, amongst other species, reed sweet-grass
(Glyceria maxima), reed canary-grass (Phalarts
arundinacea), great willow-herb (Epilobium hirsutum),
branched bur-reed (Sparganium erectum) and some
sedges (Carex acutiformis, C. riparia, C. paniculata, C.
hirta and C. otrubae). Nettles could be interspersed
amongst any of these except in continuous standing
water. On the whole, they were denser up the banks
in competition with bramble, coarse grasses like
cocksfoot, couch and false oat-grass, and other
common weeds such as cow parsley, docks and
dandelion.

COMMON STINGING NETTLE IN WILTS 1994

No site was free from nettles. One hundred and
sixty-three out of 173 subsites (94%) were nettled, at
least beyond and outside mown and weed-killed
wharf and lock areas. Nettles were abundant at 63,
and frequent at 65, of these subsites. The fringes of
nettles between chosen sites were more continuous
than those ar the (mainly bridge and wharf) sites in
1994, and water margin fringes were frequent. In
contrast to the findings of Grose in the 1950s, nettles
now form fringes along the Kennet and Avon canal
by waterside, banks and towpath in a 40-mile
waistband across the middle of Wiltshire from
eastern to western county boundaries.

Study V: special features

Roadside and hedgerow nettles in Wiltshire generally
grow within the 2—4 (occasionally 5) ft height limits
given in national floras. However, many riverside,
ditch and woodland nettles in Wiltshire are well over
6ft. The writer has recorded a 3.38m (11ft 32in)
nettle from Bodenham in south Wiltshire which was
growing from the rising waters of the Salisbury Avon
with others of comparable height (Oliver 1995).
Nettles from the Kennet and Bristol Avon were often
taller than 6ft, and some members of the WBS were
discouraged from river surveys because of the height
and density of nettle beds.

Nettles in many agricultural, riverine, woodland
and waste sites showed remarkable regenerative
powers both by seedlings and vegetative reproduction.
In particular, stolons (not mentioned in most British
floras) could grow over 1 metre in length in the
winter months. These aspects are detailed in three
papers (Oliver 1993 atb, 1994 a+b and 1995).
Formidable underground networks of rhizomes were

| frequently so dense that few or no other plants could

compete. These, which are usually 3-45cm below the
surface, were not measured. Wheeler (pers. comm.),

_ however, sampled the nettle rhizomes under Isq.

metre of riverbank: the total length was an astonishing
63.41 metres!

SUMMARY OF PREVALENCE IN 1994

_ The four main studies show repeated or sometimes

continuous abundance of stinging nettles in summer
and autumn months on Wiltshire river and canal
; banks, often fringing water, in and by ditches, along
‘field-sides and hedgerows and fences, amongst
\dumps and rabbit warrens and badger setts, and as
‘usually the major component of track-side
vegetation. On roadsides, nettles now form the usual
fringing curtains behind the cut grass verges, but

125

frequently form thicker bands and usually infiltrate
rough-cut roadside grasses (mainly Arrhenatherum
elatius, false oat-grass) to a greater or lesser extent.
The findings from Studies I], III and IV show
county-wide networks of nettles, and therefore
represent more than patchy abundance. Such net-
works are most continuous in agricultural areas but
thin out and can almost disappear in unenriched
habitats. Wiltshire is however mostly heavily farmed.

Fields of grass, the plants of open downland, or
the masses of grasses and buttercups conspicuous
in early summer might preclude consideration of
U. dioica as a main contender for one of the top three
most common of Wiltshire’s flowering plants. Fields
cover greater areas than field margins. However,
Study I noted the invasion of some fields by nettles
in the Kennet valley, either as numerous ‘blebs’
amongst the rye-grasses, or as more continuous
heavy infiltration obvious in mid summer and
autumn. Some National Trust grassland was even
more heavily nettled than adjacent farmland.
Fences, obstructions, margins, sarsens, depressions
and, above all, ditches create ‘reservoirs’ for nettle
rhizomes and stolons to invade and_ re-invade
grassland and downland. Perhaps nettles were well
controlled in three-quarters of the fields, but could
be seen to be occasional, frequent or abundant in the
remainder, covering more ground than any one
species of grass, buttercup, dock or clover when such
infiltration occurred.

In woodlands, the nettles showed equal oppor-
tunism. The deep summer shade of the mature
beech canopies allowed large areas of ground to be
almost exclusively covered by bluebells in West
Woods. However, wood margins, tracks and clearings
could become densely filled with nettles within two
years, and nettles often thrived in lesser shade as a
frequent or even dominant summer and autumn
ground cover over wider areas of woodland. After
tree-felling, the few nettle seedlings could progress
to immense nettlebeds in due course.

From 1992, the WBS has had numerous field
days, and the writer cannot recall one in Wiltshire
when nettles were not seen. Even when visiting
heathland, marsh or Salisbury Plain chalk downland,
nettles can be found along roads, watersides, parking
areas, tank ruts, cattle troughs, areas with rabbits or
badgers and wherever there has been farming. Thus
nettles are now found in virtually all of the 65
Wiltshire habitats described by Grose (1957), even if
mainly confined to margins such as _ track-sides,
around cattle troughs or car-parking areas. This
ubiquity of nettles throughout the county is backed

126 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

by the findings of the WFMP that U. dioica is now
found in at least 96% of the Wiltshire tetrads
(Gillam 1993). This county coverage is matched by
six other species: hawthorn, elder, cleavers, dande-
lion, yarrow and the creeping thistle (Cirsium
arvense). None of these six species covers the ground
as densely as nettles frequently do. The four studies
in this paper demonstrate that in 1994 U. dioica, as
well as being ubiquitous, is frequent, continuous or
abundant as a network throughout the county of
Wiltshire, and that this has been obvious from May to
early December throughout the 1990s.

DISCUSSION I: INCREASED PREVALENCE
BETWEEN THE 1950s AND 1990s

‘... I’ve lived here 50 years [Bodenham, S. Wilts]...
every year more nettles, especially these past few
years .... Unprompted comments by farmers and
others should be heeded, but actual measurements
are preferable.

The Wiltshire Flora (Gillam 1993) data showing
U. dioica to be (equally) the most widely recorded
single species merely indicates ubiquity over rather
than density within the county tetrads. The 96% for
nettle and the other six very common species may
just mean feeble coverage of the remaining 4% of the
tetrads rather than having specific significance. From
Wiltshire’s 30 common grass species, only cocksfoot
(Dactylis glomerata) was recorded as reaching 95%
tetrad coverage; but this is, in any season, the most
easily identifiable grass!

Donald Grose (1957, 671-767) made a deter-
mined and comprehensive quantitative study of the
vegetation of Wiltshire in the 1950s. Nettle did not
feature very high on most of his 65 habitat lists, and
consequently he seemed surprised that it came as
high as seventh in his overall final frequency list
(zbid., 705). He explained this by its occurrence ‘.. . in
a very wide range of habitats’, but by this he meant
23, just over one-third of the total. Combining the
findings of the preceding four studies with the
numerous WBS and WFMP field days over the last
five years in varied habitats throughout the county, it
can be said with confidence that U. dioica is at least
occasional, if not frequent or abundant in at least
three-quarters of the Wiltshire habitats. It is seen on
every outing, usually in quantity.

Grose (ibid., 734-41) never recorded nettles in
any of the top 30 species in any of his water/riverside
lists. Even if allowance is made for his main attention
being directed at channel and water-fringe rather
than bank-side species (a hard enough discrim-

ination), it is inconceivable that he would not have
mentioned nettles in the 1950s, even if they were
only one-tenth as abundant as they now are in the
1990s. They now, after all, often fringe the water’s
edges as well as sometimes invading the channel and
frequently dominating the sloping banks.

Study III was concerned with rivers rather than
lakes or pools, but the latter in the early 1990s were
often just as densely nettle-fringed as the rivers, if
not more so. This often applied even to dew ponds
on the downs. Grose (1b7d., 737-41) included water
margin and bank-side species in his 12 canal lists,
including the Kennet and Avon Canal. Nettles again
were not mentioned in Grose’s top 30 species. Study
IV shows their ubiquity and frequency now at all canal
sites, and often they were even more continuous
and/or abundant between (rather than at) the sites.
As with the rivers’ immediate waterside fringes,
nettles commonly occurred at the canal water edge.
Substantial stretches of the canal in October 1994
had nettles touching the water (although not usually
growing from it). Direct inter-species quantitative
comparisons were not made for Study IV, as had been
done in Study III, but the writer’s firm impression was
that nettles were the commonest canal bank species,
and the commonest fringing species in the fences
and hedges bounding the canal. U. dioica also
seemed the strongest contender for the commonest
single species amongst the canal immediate waterside
vegetation (see Study IV _ for other possible
contenders). Only on the mown areas, by bridges,
locks and short-cut grassy verges was perennial rye-
grass (Lolium perenne) dominant, the most
continuous flowering plant on the manicured parts of
the canal towpath. Once again, the contrast between
current findings and Grose’s studies is extremely
pronounced. Nettles frequently reign in the 1990s
and even when controls are attempted, they are
awaiting their opportunities to colonise and
recolonise most waterside habitats.

Grose had nettles as his most common single
species for waste ground (op.cit., 756), but even so,
averaging just over ‘occasional’ status, and first-equal
for rabbit warrens (with ‘abundant’ status, zbid.,
714). Other high levels were third for scrub (p. 697),
equal third for much disturbed chalk grassland
(‘occasional’ status, p. 717), eighth for woods
(‘occasional’ status on average, p. 692) and ninth for
hedgerows (less than ‘occasional’ status, p.703).
Except for the last two, these are in line with current
findings. With respect to hedgerows and woods,
nettles are now much more common. Many, perhaps
most, hedges are now nettled. The only hedgerow

}
|
i
|

COMMON STINGING NETTLE IN WILTS 1994

species now likely to exceed nettle in frequency is
hawthorn, the usual main hedging plant.

The most striking differences between the 1950s
and 1990s are illustrated by the four studies here.
Little wayside weeds, annual meadow-grass (Poa
annua) and ground-rosette plants are still very
common, but the dominant wayside weeds are bigger
in the 1990s than they seem to have been in the 1950s
(compare Grose 1957, 720-29). Lush enriched meadows
nurture large or vigorous and rapidly seeding strains of
perennial rye-grass which colonise waysides whether
cut, mown or left to seed in the fields. In such
agricultural areas, other coarse grasses such as rough
meadow-grass (Poa trivialis), cocksfoot, tall or false
oat-grass (Arrhenatherum elatius) and couch also
coexist with other big nutrient-loving farmyard weeds
such as cow parsley, docks, thistles and, above all,
nettles. Tall oat-grass formed continuous main
roadside verges in the 1950s (Grose 1957, 721-23).
This is still so, but all these large wayside weeds,
especially nettles, are squeezing the mosaics of smaller
ones to the margins. The little weeds, plantains,
dandelions, daisy, creeping buttercup, annual meadow-
grass etc., mostly still do very well on stony and mown
margins and verges, and in some _ well-grazed
meadows, but stand less chance in competition with
the mass of the taller wayside weeds, especially when
these are blanketed by the (also often abundant)
bindweeds and cleavers in high summer.

In short, observations in the 1990s _ indicate
frequent summer and autumn nettling in most
Wiltshire habitats, rather than the occasional
occurrences in the 35% found by Grose in the 1950s.
Even within Grose’s 35%, nettle frequency levels
seem to have very much increased. Fields are invaded
by nettles in the absence of effective controls, and
nettles abound in farmland. U. dioica was not noted
to be a dominant roadside species in Wiltshire in the
early 1970s by Horton (1975), although he did refer

| to its increased spread at one of the verges under
| study. Specific measurements in the studies I-IV
| preceding prove nettle ubiquity and vastly increased

abundances in the 1990s by A-roads, lesser roads,
tracks, footpaths, rivers and canals.

References to problems caused by the spread of
dominance of big grasses like creeping bent (Agrostis
stolonifera), upright brome (Bromopsis erecta) and
false oat-grass (Arrhenatherum elatius) in the new
Wiltshire Flora (Gillam 1993, 22, 28, 85, 87, 91, and
341) makes the present writer hesitate to place nettle
\ firmly first in abundance amongst all flowering plants
‘in Wiltshire in the 1990s, but it is now the most

jubiquitous, and has certainly overtaken ribwort

127

plantain (Plantago lanceolata), Grose’s number one.
On the basis of studies I-IV, U. diozca might be even
more abundant than amy one of Wiltshire’s 30 or so
very common grass species, with the probable
exception of the agriculturally abundant perennial
rye-grass. There is therefore a case to be made for
U. dioica being, in 1994, Wiltshire’s second, rather
than seventh, commonest wild flowering plant — or
first if agricultural strains of grass (especially
perennial rye-grass) are excluded.

DISCUSSION II: POSSIBLE REASONS FOR
INCREASED NETTLE PREVALENCE

The ensuing points are tentative. They are linked as
far as possible to direct observations.

1. Chemical enrichment

Phosphate (with nitrate) excess from farming,
sewage, animals and man has been blamed for the
spread of nettles in Southern and Central England in
the 1990s (Pearce 1994). Grant (1994) compared
this abundance to marine or other aquatic ‘algal
blooms’ but implied permanency rather than a
transient phenomenon.

These studies I-IV certainly confirm heavy
nettling to be associated with agriculture. The
habitats emphasised by Donald Grose with thin soils,
the swards of little rosette and low ground cover
plants, old-style hay pastures, or hedges and woods
with only sparse, occasional or no nettling, all only
occur away from enriched farmland and top-dressed
fields. The problem is that measurements of
phosphates and nitrates by the National Rivers
Authority and others do not show excess levels. Is it
possible that phosphates are held in organic chemical
reservoirs or that nettles (like lichens) provide a more
reliable measure as biological indicators than direct
chemical testing?

Intensive testings of the physiochemical environ-
ments of lichens on the Fyfield Down sarsens were
made in 1986 and 1987 (Dillon et al. 1992).
Lockeridge Dene and parts of Fyfield Down and
Piggledene sarsen National Trust reserve fell within
the area of Study I. Nettles abound around the
sarsens on these reserves, but were kept under some
control in Lockeridge Dene in 1994. In Piggledene,
numbers of sarsens were submerged in nettlebeds,
presumably influencing lichen survival. Comparable
studies on nettles show such facilitating or competing
variables for field-side nettles to be much more
violent than for terricolous lichens. These are
outlined by Grime et al. (1989) and detailed by

128 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Wheeler (pers. comm., 1995). In short, phosphates
encourage nettle seedling establishment, and nitrate
enrichment stimulates the very vigorous growth
rates.

2. Ditching and drainage

Study III showed nettles to be abundant by the
Wiltshire rivers. Most rivers, like the Kennet, have
been much changed since the 1950s by greatly
increased surrounding drainage, causing rapid
agricultural and roadside run-off. There are now
ditches rather than flattened streams in marshy
meadows in the upper reaches. This change favours
nettles rather than the older patterns of sedges,
rushes and reeds because of the rapid inundation
and drying-out sequences. ‘Meargealle’, the old
English marsh marigold, which may have given its
name to Merleberge in the Domesday Book (Grose
1957, 97), now stands little chance along the Upper
Kennet in competition with nettles in the environs of
Marlborough today. The river Bourne, also very
heavily nettled in many stretches, can be, over many
miles, either quite dry or heavily flooded (Delair
1991).

These rapidly alternating rzver water levels, dry >
wet —> dry — flooding — dry, do not explain why
nettles so often also fringe the canal watersides
(Study IV) in such abundance, although there may
be a bank-side zone of rising and falling canal water
in the summer months, which favours the spread of
nettle rhizomes and the health of nettle roots.
Human activity, whether fishing or enrichment or
disturbances such as dredging and wash from boats,
will also favour canal-side nettle spread.

3. Disturbed ground

This is given in floras and ecology articles as a (or
the) main cause of nettle spread, but there are
contradictory elements. On the one hand, the
tranquil ancient West Woods Wansdyke path has a
mixed flora free from nettles (see Study I), whereas
nettles can rapidly colonise clearings in logging
areas of West Woods. Also, nettles abounded around
badger setts, rabbit warrens, neglected gardens,
neglected urban and village stonework and often
where earth had been recently dumped. However
nettles also formed permanent fringing curtains to
hedges, roadsides and track-sides well away from
hooves (Studies I and II); they often dominated dry
ditches, ancient dumps, sarsens, fences and wire
entanglements a quarter of a century old or more.
Nettles have, of course, long been associated with
human settlement and stock keeping, as evidenced

by their persistence on a number of archaeological
sites.

Seedlings were most often seen in gardens, dry
riverbeds, cleared forestry areas (second year),
around farm out-buildings and on bare margins. Re-
growing vegetative fragments, stolons and rhizome
sections seemed to be moved around by cows (Oliver
1993a) and abounded at field edges, riverbanks,
some downland depressions and on clods of earth in
field centres. These competed well with grass. By
contrast, other nettlebeds were permanent, excluding
all other vegetation until very deeply shaded by trees.
The vigour and persistence of these spreading
nettlebeds owed nothing to disturbance. However,
Wheeler (pers. comm.) considers disturbed ground
crucial to the initial seedling establishments.

4. Genetic vigour

Under the heading ‘Special Features’ (and see Oliver
1993 a+b, 1994 a+b, 1995), attention is drawn to
variations from standard flora descriptions. Some
colonies had huge specimens. Others produced very
long stolons, especially in winter months. Recent
evolution of successful strains seems likely, with the
possibility of other advantages than those already
highlighted in competition with other species. These
could, for instance, include the very tough and rapidly
expanding root and rhizome networks, persistence in
shade, seedling vigour, frost-resistant stolons and
numerous other possible improved potentials.
Wheeler (pers. comm.) attributes the success of U.
dioica in diverse English habitats to the highly
successful genetical outbreeding system. This permits
extremely plastic responsiveness and consequently
great modifications in growth: for instance, large thin
leaves as well as giant stems in partial shade. The
present study strongly supports Wheeler’s arguments.

5. Changed grazing patterns

Grazing patterns, timings and preferences of stock
may have changed since the 1950s, and it is possible
that past patterns were more effective in controlling
nettles than the combined onslaughts of herbicides,
cutters and strimmers in the 1990s.

6. Expanding reserves of propagules

One theory is simply that nettles (like sparrows and
house-mice) are suited to man and his animals. Once
established, seeds, stolons, rhizomes, root-networks
or any vegetative fragments persist. Each year men,
machines and cattle all create new colonies. Once
established, these persist, with long term potential
for further colonisations.

COMMON STINGING NETTLE IN WILTS 1994

None of these categories is exclusive. It is probable
that there are subtle but strong interactions between
them, perhaps also involving other unknown factors,
to explain the increased May to November domi-
nance of Urtica dioica in Wiltshire in the 1990s.

Acknowledgements. Appreciation is given to the following workers:
Jennifer Acornley, David Blackford, Brenda Chadwick, Tony and
Stella Dale, Daphne Graiff, Malcolm Hardstaff, Diana Hodgson,
Barbara Last, Christine McQuitty, Phillipa Parker, Maureen Ponting,
Judith Robinson, Mary Robinson, Jean Wall,Winifred White, Civil
Williamson and Gwyneth Yerrington. I am grateful for the
comments and suggestions of Dr Keith Wheeler who would
welcome communications for his forthcoming book on any aspects
of the nettle, including uses, folklore, ecology etc.

REFERENCES

CRAWLEY, M.J., 1989 Invaders, Plants Today ii, 5, 152-158

CLAPHAM, A.R., TUTIN, T.G., and MOORE, D.M. (CTM)
1987 Flora of the British Isles, Cambridge: CUP

DELAIR, J.B., 1991 Notes on the flooding of the River
Bourne (Wiltshire) Jan—Feb 1990, WAM 83, 137-140

DILLON, P., SKEGGS, S., and GOODEY, C., 1992 Some
investigations on habitats of lichens on sarsen stones at
Fyfield Down, Wiltshire, WAM 85, 128-139

DU FEU, G., 1994 Cover photograph, British Wildlife v, 4

GELTMAN, D.V., 1992 Urtica galeopsifolia in Wicken Fen,
Watsonia 19, 127-129

GILLAM, B., 1993 The Wiltshire Flora, Oxford: Pisces
Publications

129

GRANT, M., 1994 Nettles and phosphates, British Wildlife v,
6, 407

GRIME, J.P., HODGSON, J.G. and HUNT, R., 1989

Comparative Plant Ecology: A Functional Approach to
Common British Species, London: Unwin Hyman Ltd

GROSE, D., 1957 The Flora of Wiltshire, Wakefield: EP
Publishing Ltd, republished by WANHS 1979

HORTON, P.J., 1975 Some notes on the vegetation of
Wiltshire roadside verges Supplement to The Flora of
Wiltshire, ed. L.F., Stearn, Devizes: WANHS, 137-143

LAHAV-GINOTT, S., and CRONK, Q.C.B., 1993 The mating
system of Elastostema (Urticaceae), Plant Systematics
and Evolution 186, 135-145

LAST, B., 1995 Another site for the non-stinging Nettle,
Botanical Society of the British Isles News 68, 10

OLIVER, J.E., 1993a Stinging nettles along the River
Kennet, Botanical Society of the British Isles News 63, 16

— 1993b Stinging nettles, IT Midwinter growth, Botanical
Society of the British Isles News 64, 29

— 1994a Sunging nettles, III Winter growth, Botanical
Society of the British Isles News 67, 18-19

— 1994b Increased abundance of stinging nettles, British
Wildhfe v, 5, 339

— 1995 Stinging nettles, IV 28 miles of nettles, Botanical
Society of the British Isles News 68, 17

PEARCE, F., 1994 Giant nettles stalk polluted Britain, Nez
Scientist (29/10/94) 1949, 9

STACE, C., 1991 New Flora of the British Isles (3rd edn.),
Cambridge: CUP

WHEELER, K.G.R., 1975 Personal communications

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp.130—143

Notes

An Early Anglo-Saxon Wrist-clasp from the Parish of Baydon
by JOHN HINES

A simple copper-alloy object found by Douglas
Wilson just inside the parish of Baydon, close to the
village of Aldbourne, is an early Anglo-Saxon
artefact that is of considerable interest for having
been found here in the north of Wiltshire. It is the
catch-piece of a wrist-clasp, a metal dress-accessory
used to fasten the slit cuff of a sleeved woman’s
garment on the hook-and-eye principle (Figure 1).
This artefact-type was first developed in southern
Scandinavia during the third century A.D. and was
eventually introduced into eastern England, appar-
ently from Norway, late in the fifth century (Hines
1993).

The example here belongs to form B 7, the
simplest and most common type found in England
(Hines 1993, 39-43). This form consists of clasp-
halves that are simple copper-alloy plates with two
holes for them to be sewn to the garment. These
plates are usually decorated, on this form only with
simple surface ornament of repoussé moulding,
punch marks or incised lines, if not some combina-
tion of these types of decoration. The Baydon
example carries only stamped ornament, one of the
less common decorative schemes, occurring on less
than one in ten of the 280 or so examples of form B 7
now recorded in England. Examples combining
punch marks with repoussé bosses are at least three
times more numerous. While it is possible that this
item was made very late in the fifth century, it was
most probably produced in the earlier or mid-sixth
century.

Wrist-clasps are of particular importance in Early
Anglo-Saxon archaeology because their distribution
defines a distinct zone of material culture in the
Midlands and north-east of England very sharply
(see Figure 2). This province coincides closely with
the early area of the Anglian English kingdoms as
identified by historical sources. It has consequently
been argued that the use of wrist-clasps was adopted
as an identity-marking feature of the costume of
Anglian English women in England (Hines 1993,
76-93). The southern boundary of this province of
material culture is a line running due east of the
mouth of the River Stour in Suffolk; several hundred
examples of wrist-clasps have been found north of
this line and very few south of it. Apart from the

G g e
G Oo Ye) QO
Bro Os —)
Bo OG 6
a ;
e 2 4
: oR 8
Es 3 ‘
¢ O 5
3 & ne)
sy i)
a

Figure 1. a: The clasp-half from Baydon, Wiltshire
(Devizes Museum 1992.192), drawn by Nick Griffiths
(scale 1:1); b. a reconstruction of the complete pair of

clasps represented at Baydon, drawn by the author

(nominal scale 1:1)

Baydon find there are only three southern outliers,
all of them of a form different from that of the
Baydon example, an elaborate cast type classified as
Class C. Two of these are from the cemetery site of
Bifrons, Patrixbourne, Kent and the third from a
grave at Saxonbury, East Sussex. It seems likely that
in all three of these cases the bodies of the clasp-
pieces had been refitted for use as brooches. There is
no reason to believe that this was so in the case of the
Baydon clasp-half, which is the only example of a
Class B clasp found outside the Anglian area in
England. It is therefore reasonable to conclude that
this item, very unusually, had been worn as a wrist-
clasp in northern Wiltshire in the sixth century.

The Baydon clasp’s closest known neighbour to
the north is a form B 7 clasp found at the cemetery
of Marston St Lawrence in the south of North-
amptonshire, some 42 miles away; not a huge
distance even by Early Anglo-Saxon standards,
although one must appreciate the strength of the
cultural boundary the clasp must have passed
through. The Baydon clasp appears a little less
strange in its local context, however, in the light of
other finds that provide clear evidence of connexions
between this area and the Anglian culture province in
the southern Midlands. There are no other Early
Anglo-Saxon finds from Baydon, and just a handful
from the neighbouring parish of Aldbourne. Just to
the east, however, in the valley of the River
Lambourn in Berkshire, there are some known Early

131

NOTES

(€66I SOUTH] Woy poidepy) ‘ar0ur 10 »
‘sojdurexs 7—] WIM setts Juasasdas SID [JBUIS :puel[suq

‘aB1e] pue 6g—-¢ ‘pozis-wInIpaut
ul sdsejo 2 g Wo] Jo uOnNqMsIp ayy, °Z% dInSr7

132 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Anglo-Saxon cemetery sites. Finds from these sites
include great square-headed brooches — an expensive
and elaborate brooch-type — from East Garston
Warren and East Shefford that have practically
identical counterparts at Norton in Northamptonshire
and Bidford-on-Avon, Warwickshire (1990 excava-
tions, grave 26), respectively (all to be illustrated in
Hines forthcoming; East Garston Warren: MacGregor
and Bolick 1993, 116-17; Norton: Leeds 1949, fig.
90; East Shefford: Leeds 1949, fig. 123; Bidford-on-
Avon: not yet published).

It is consequently quite reasonable to postulate
that there was a line of communication of some
significance from Saxon North Wiltshire through
the Upper Thames area to the southern central
Midlands in the sixth century, presumably via the
Roman road system known as Akeman Street and its
branches (Margary 1973, 155-70). It is interesting
that the cluster of sites at the southern end of this
line lies close to the likely route of the Roman road
(Margary op. cit., 170, Route 164) which almost
certainly branched off Akeman Street at Alchester,
Oxfordshire, and ran south-westwards to Mildenhall
(Cunetio), near Marlborough. The Baydon clasp
ought to have been made in the Anglian Midlands. It
is difficult to imagine that such a humble object as
this entered into a system of material exchange for its

own sake; that it could have done so as part of what
may have been a fine garment is more credible.
However, while the Baydon wrist-clasp cannot be
taken as automatic evidence of the ethnic identity of
its owner, the use of wrist-clasps is strongly marked
as an Anglian characteristic, and it is a realistic
possibility that this find represents the presence, and
death, of an Anglian woman in the area of Baydon in
the sixth century. In that case exogamy would
provide an obvious underlying explanation.

REFERENCES

HINES, J., 1993 Clasps: Hektespenner: Agraffen. Anglo-
Scandinavian Clasps of Classes A-C of the 3rd to 6th
centuries A.D.: Typology, Diffusion and Function, Kung}.
Vitterhets Historie och Antikvitets Akademien,
Stockholm

HINES, J., forthcoming A New Corpus of Anglo-Saxon
Great Square-Headed Brooches, Society of Antiquaries
of London, Research Report 51

LEEDS, E.T., 1949 A Corpus of Early Anglo-Saxon Great
Square-Headed Brooches, Oxford

MACGREGOR, A., and BOLICK, E., 1993 Ashmolean Museum,
Oxford: A Summary Catalogue of the Anglo-Saxon
Collections (Non-Ferrous Metals), British Archaeological
Reports, British Series 230, Oxford

MARGARY, I.D., 1973 Roman Roads in Britain, 3rd. ed.,
London

The Identification of Alentone/Allentone in Wiltshire Domesday

by JASON ST. JOHN NICOLLE

Among the Wiltshire lands held by the nuns of
Amesbury abbey in 1086 was a small estate, assessed
at 4 hides, which is described by Domesday as
Allentone.' This place-name is identical to Alentone, a
royal estate which was also assessed at 4 hides, for
the Domesday entry for Alentone includes the
following note:

im eadem villa sunt uu hidae terrae quas imjuste
abstraxit Heraldus ab aecclesia Ambresberie.

. Domesday Book [D.B.], f. 68b.

. D.B., f. 69a.

. R.C. Hoare, History of Modern Wiltshire. Hundreds of Everley,
Ambresbury, and Underditch (London 1826), p. 107; W.H. Jones,
Domesday for Wiltshire [Jones] (Bath 1865), p. 196; R.R.
Darlington, “Translation of the Text of the Wiltshire Domesday’
in R.B. Pugh, er al., eds., The Victoria History of the Counties of
England. A History of Wiltshire [VC.H. Wilts] (14 vols., London
1957-1991), ii, pp. 131, 134; H.C. Darby, G.R. Versey,

BQN

testimonio tainorum scirae. Modo tamen habet
aecclesia.?

Alentune had previously been held by Earl Aubrey,
and by Harold before him. But where was
Allentone/Alentone?

Allentone/Alentone has usually been identified as
Allington, about 4 miles to the east of Amesbury in
the valley of the river Bourne.’ However, one scholar
has expressed some doubt over this identification,

Domesday Gazetteer [Darby] (Cambridge 1975), p. 44;
Domesday Book. Wiltshire, C. Thorn, F. Thorn (Chichester
1979) n.p., entries 16.3, 23.7; J.E.B. Gover, A. Mawer, F.M.
Stenton, The Place-names of Wiltshire, English Place-Name
Society, xvi (1939) [E.P.N.S. Wilts], p. 90 identifies the place-
name as Allington in Chippenham, but provides no authority to
support this identification. They are followed by E. Ekwall, The
Concise Oxford Dictionary of English Place-names (4th. edn.,
Oxford 1960) [Ekwall], s.v. Allington.

NOTES

suggesting that the correct identification might
possibly be Alton in the parish of Figheldean, about
4 miles due north of Amesbury.* The purpose of this
note is to develop this suggestion further, and to
argue that Allentone/Alentone can indeed be identified
with Alton and, furthermore, that such an identifi-
cation is preferable to the traditional identification of
this place-name as Allington.

First, there is the evidence from toponymy. All
the medieval forms of the Allington in question
begin with the letters ‘Ald’; no form without
the letter ‘d’ is recorded before modern times.’
Although the letter ‘d’ might have been added to the
spelling of the place-name at some date between
Domesday and 1179, the earliest datable spelling,
it is hard to see any philological reason why this
might have happened. On purely philological
grounds, if the Allington in question were being
sought in Domesday, Aldintone or Ellatune, not
Allentone/ Alentone, would be expected.°

Eltone does occur in Wiltshire Domesday. John
the Doorkeeper held an estate there, assessed at 5
hides.’ This has been identified as Alton in
Figheldean, apparently solely on the grounds that
the Geld Rolls indicate that John’s estate lay in the
Hundred of Amesbury, and that Alton lay in this
Hundred in the later Middle Ages.* However, there
is no early form of the Alton in question with an
initial ‘e’.? Philologically, Eltone is much more likely
to be Allington, which also lay in the Hundred of
Amesbury in the later Middle Ages. The form is
close to Ellatune, the Domesday form of Allington in

Hampshire, which appears to have the same
philological history as the Allington near
Amesbury.!°

Secondly, there are arguments based on

Domesday geography. The first is merely suggestive.
If Allentone/Alentone were to be identified with Alton,
rather than Allington, it might explain why Harold
had held an estate there, and why he had also taken
the nuns’ estate, for Alton is contiguous with
Netheravon, where Harold held a substantial estate.!!

4. VC.H. Wilts, iti, p. 243n (R.B. Pugh).

5. E.P.N.S. Wilts, p. 358.

6. This argument is based on the Domesday forms of Aldington
(Kent) and Allington (Hampshire): Ekwall, s.v. Aldington,
Allington; R. Coates, The Place-Names of Hampshire [Coates]
(London 1989), p. 21.

7. D.B., f. 74b.

8. Jones, p. 214; VC.H. Wilts, ii, pp. 166, 194, 195; E.P.N.S.
Wilts, p. 366; Darby, p. 449; Ekwall s.v. Alton. For Amesbury
Hundred in 1334, see The Lay Subsidy of 1334 ed. R.E.
Glasscock, British Academy Records of Social and Economic
History N.S. ii (1975) [Glasscock], pp. 345-346.

9. E.P.N.S. Wilts, p.366.

133

Domesday does not record Harold holding land
contiguous with Allington. The second argument
based on Domesday geography is _ stronger.
Domesday notes that Earl Aubrey had held 42 of the
11; hides which Harding held at Figheldean. in
1086.'7 As noted above, Alton lay in Figheldean
parish. If Alentune is ignored, the only estate which
Domesday states that Earl Aubrey had held in or
near Figheldean was Ablington; but Ablington was
assessed for Geld at 12 hides short of the required 42
hides.!? The difficulty disappears if Aletone is
identified as Alton, because Alton was assessed at 4
hides, and thus Alton and Ablington together would
be sufficient to provide the unidentified 4: hides
which Earl Aubrey is said to have held at Figheldean.

Finally, there are arguments based on _ the
subsequent estate history of Alton. The nuns of
Amesbury certainly held the manor of Alton
(Alletona) by 1179, when it was confirmed to them
by Henry II. There is nothing in the confirmation
charter, or elsewhere, to suggest that it had been
acquired since 1086.'* It was retained until the
Dissolution.!? The charter of 1179 distinguishes
clearly between the manerium de Alletona, and
Aldintona (Allington), where the nuns held 4 acres.!°
The only other reference to the nuns _ holding
property at Allington comes from the Taxatio of
c.1291, which records a pension of 2s. due to the
nuns from the parish church.'!’ The evidence
therefore presents two possibilities. Either the nuns
had lost an estate at Allington and gained one at
Alton, at some date between 1086 and 1179; or they
had never held an estate at Allington at all, and had
simply retained their estate at Alton, which they had
held since at least 1086. In the absence of any
evidence to support the former explanation, and
given the fact that the nuns retained all the other
estates which they had held in 1086 until the
Dissolution, the latter explanation would seem
preferable.'®

It has also proved possible to trace the descent of
the estate which Earl Aubrey had held at Alentone.

10. D.B., f. 48b; Glasscock, p. 345; Coates, p. 21; Ekwall s.v.
Allington.

11. D.B., f. 65a.

12. D.B., f. 74a.

13. D.B., f. 69a.

14. The charter survives in the form of an Inspeximus from 1270:
Calendar of the Charter Rolls Preserved in the Public Record Office
[Cal. Chart. Rolls] (6 vols., London 1903-1927), ii, p.158.

15. Valor Ecclestasticus Temp. Henr. VII. Auctoritate Regia Institutus
[Valor] (6 vols., London 1810-1834), ii, p. 94.

16. Cal. Chart. Rolls, ii, p. 158; E.P.N.S. Wilts.

17. Taxatio Ecclestastica Angliae et Walhiae Auctoritate RP Nicholai IV
circa A.D, 1291 (London 1802), p. 180.

18. Compare D.B., ff. 60a, 68b with Valor, ii, 93-95.

134 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

The Earl’s lands appear to have been acquired by the
king in c.1080. They were then granted to Hugh de
Grentmesnil, whose lands were later appropriated by
Robert, Count of Meulan and Earl of Leicester
(d.1118).'° Thus the fee of Earl Aubrey, or at least
part of it, became incorporated into the great Honor
of Leicester. The Honor of Leicester was later
partitioned, in 1206-7, to form two new Honors,
known respectively as the Honor of Leicester and the
Honor of Winchester.*° Of the 9 estates which Earl
Aubrey had held in Wiltshire, 4 appear in 1242-3 as
fees of the Honor of Leicester, and 4 as fees of the
Honour of Winchester. Among the fees of the
Honour of Leicester was 1 knight’s fee at Aletune,

19. VC.H. Wilts, xi, pp. 120, 240; L. Fox, ‘The Honor and Earldom
of Leicester: Origin and Descent’, English Historical Review liv
(1939), 385-402.

20. Fox, Joc. cit., p. 391.

which is clearly to be identified with the estate which
Earl Aubrey had held at = Alentone.*! The
identification of Aleton as Alton is put beyond doubt
by two 15th-century surveys of the fees of the Duchy
of Lancaster, which by that date incorporated much
of the old Honor of Leicester.”

The cumulative weight of these arguments, based
as they are on three different types of evidence, must
surely put the identification of Allentone/Alentone
beyond doubt. The suggestion that this place-name
may represent Alton, rather than Allington, is more
than a mere possibility. It is a certainty, or as near to
a certainty as it is possible to come, in the perplexing
world of Domesday studies.

21. Liber Feodorum: The Book of Fees commonly called Testa de Nevill (2
vols. in 3, London 1920-31), ii, pp. 730, 731, 732, 746; D.B., f. 69a.

22. Inquisitions and Assessments Relating to Feudal Aids (6 vols.,
London 1899-1920), v, pp. 240, 242; vi, p.627.

The Silver Seal-matrix of Geve of Calstone
by JOHN CHERRY

This silver seal-matrix was discovered by Mrs N.
Taylor at Manor Farm, Calstone Wellington, a
village two and a half miles south-east of Calne, in
1992. Since the name of the owner of the seal, Geve
of Calstone, is engraved on the matrix, it is a
remarkable find of a seal-matrix of a lady on her
home manor. Its acquisition by Devizes Museum
(accession no. 1992. 371) was made possible by the
aid of grants from the Victoria and Albert Museum/
Museums and Galleries Commission Purchase
Grant Fund and the Beecroft Bequest.

The matrix (38 by 24mm), pointed oval in shape,
was cast and then engraved with a figure and
inscription on the front. The back is plain. There is a
loop at the back for the attachment of the matrix to a
chain (see Figure opposite).

On the front there is a lady, standing on a
bracket, dressed in a wimple and a cloak which
was fastened by a cord across the breast and which,
by the hatching attempting to show vair, was
probably lined with fur. In each hand she holds
a shield by its point. In her right hand (dexter)
the shield is barry of two with in chief two
lions rampant. In her left hand (sinister) she holds
a blank shield. The figure is divided from the legend
by a pearled border which extends to the outer
pearled border to include the two shields. The legend
reads * SIG JTLLUM.GEVE * DE * CALESTON’.

This inscription identifies the owner of the seal-
matrix as Geve of Calstone. Calstone in Domesday
Book includes three so-called manors, and Manor
Farm at Calstone occupies the site of the capital
messuage of the manor.

It is clear from the representation of the lady that
Geve is here used as a feminine name. While Geva or
Geve can be used as a masculine name, here it is an
Anglo-Norman version of the Anglo-Saxon feminine
terminal element -gefu or -gifa (Searle 1897).

The shield in her right hand is that of the
Wiltshire family of Calston or Calstone. Two other
seals are known with these arms. They were used by
Thomas Calston of Wilts in 1392 and Thomas
Calston in 1409 (Chesshyre and Woodcock 1992,
36-7, from New Sarum City Deeds seals and from a
deed in the Button Walker Heneage Muniments).
The Wiltshire Record Office also has a manuscript
of 1392 of Thomas Calstone ‘of Wiltshire’ con-
cerning land at Enford and his attached seal has the
Calstone arms. The sinister blank shield indicates
that Geve’s father did not bear arms.

Ladies are usually shown standing on medieval
seals. Some of the finest and earliest representations
of ladies on seals are those of the Queens of England
and of France in the 12th century, who often hold a
sceptre, an orb or a fleur-de-lis. One of the finest is
the silver seal-matrix of Isabella of Hainault (Johnes

NOTES

135

Yi” le

Silver seal-matrix of Geve of Calstone, with impression. Scale 2:1

1960, 73-6). In the 13th century noble ladies are
often shown with a hawk, a fleur-de-lis or a shield or
shields. At the same time, they begin to be clothed in
their husband’s armorial bearings. On her seal,
Margaret de Quincy, widow of Saher de Quincy,
Earl of Winchester, of c.1220, is robed in the
lozenges of de Quincy and stands under an arch,
next to a tree on which hang the shields of de Quincy
and Fitzwalter (Hunter Blair 1943, pl. XVc).
Sometimes the lady is shown holding one or,
occasionally, two shields with the other, or others,
suspended in air or hanging on a tree. The develop-
ment of the design of ladies’ seals in the middle of
the 13th century is illustrated by the two seals of Ela,
the daughter of William Longespée, Earl of
Salisbury (died 1226), who was the illegitimate son
of Henry II. Both seals are two sided. On her first
seal as Countess of Warwick she is shown standing
holding a hawk while the arms of Longespée are on
the reverse within a sexfoil (Bowles 1835, pl. III). On
her subsequent seal, as the wife of Philip Basset, she
is shown holding her father’s shield of Longespée in
her left hand while that of her former husband hangs
in the air to her right (Birch 1892, no. 6579). On the
reverse is the shield of Basset, her current husband.
She married Basset after 1242 and before 1271
(Bowles 1835, pl. III)

The earliest lady to hold two shields appears to be
Agnes de Percy who, in 1244, holds two shields by
cords from above (Antiquaries seal collection,
drawer F32). The earliest appearance of a lady
holding two shields by their points is on the seal of
Emmeline, wife of Stephen Longespée, in 1250
(Hunter Blair 1943, pl. XVI and Birch 1892, no.
6680). She is followed in 1274 by Ela, daughter of
William Longespée, Earl of Salisbury, widow of
James Audley (Bowles 1835, pl. II and Birch 1892,
no. 6573 from BL Add. Ch. 10, 619 dated 1274).
Other examples are the seals of Philippa of Lancaster
1284 (Hunter Blair 1912, no. 1536), Joan Achard
1292 (Antiquaries seal collection, drawer F1),
Margaret Basset (Birch 1892, no. 6586, from BL
Cotton ch. xxiii. 2 dated to the time of Henry III or
Edward I) and Eleanor de Zouche, wife of Alan,
Lord Zouche, of 1294 (Birch 1892, no. 6742, and
also Antiquaries seal collection, drawer F44). The
latter is one of the closest comparisons to the seal of
Geve, since she holds her husband’s shield in her
right hand and her father’s shield in her left.
Although the material of these seals is not known
they were probably silver.

The only surviving comparable silver seal-matrix
of a lady is of Hawys, Lady of Cyfeiliog, which was
found in Oswestry in 1853, and is now in the

136 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Shrewsbury Museum. Hawys is shown holding a
shield with a rampant lion for Powys in her right
hand and a shield with two lions passant for
VEstrange in her left. In this case the two shields
represent her two husbands (Siddons 1991, 291 and
fig. 93; casts are in the Antiquaries seal collection,
drawer F25, and in the British Library, Birch 1892,
no. 6670).

Geve is shown with her husband’s shield on her
right, and if her father had been armigerous, that
would have been on the left. One of the grandest
displays of heraldry and genealogy is on the two-
sided seal of Dervorguile, wife of William of
Balliol, of 1284. Standing in her widow’s weeds,
she holds in her right hand her husband’s shield of
Balliol and in her left her father’s shield. From a
tree on her right there hangs a shield of the
earldom of Chester while on the tree on the left
there is the shield of her maternal grandfather. On
the reverse there hangs from a tree a shield which
shows her father’s and husband’s arms dimidiated,
with smaller shields of Chester and her maternal
grandfather hanging from branches above it. It is
interesting that on the obverse of the seal she gives
the precedence to her husband but on the reverse
to her father (Hunter Blair 1943, pl. XVg from
Balliol Deeds 565).

Since there are no impressions of the seal-matrix
found at Calstone attached to documents, the only
way of dating it is by epigraphy. It is most likely that
the seal of Geve of Calstone dates from the second
half of the 13th century, and the shape of the letters,
especially the form of the letter G which has a long
upper arm overlapping the lower part of the letter

(Kingsford 1929, 155, 168), would support this. It is
possible that Geve may have been a widow when the
seal was engraved.

The value of the seal is in the insight it gives into
the portrayal of women on their seals in the 13th
century. One of the most famous later medieval
representations of a lady holding a shield occurs in
the Luttrell Psalter of 1340-50, where Sir Geoffrey’s
daughter-in-law holds his shield. This representation
has recently been studied by Professor Richard
Marks (Marks 1993-4, 343-355).

It may simply be the coincidence of survival but a
large number of the surviving ladies’ seals of this
type are connected with the Longespée family and it
may be that they had set a fashion in Wiltshire which
Geve of Calstone was imitating.

Acknowledgements. I am grateful to Paul Robinson for suggesting
that I write this; and to Steven Hobbs, the Wiltshire County
archivist; to Loveday Gee; and to my colleague, Leslie Webster, for
their help.

REFERENCES

BIRCH, W. DE GRAY, 1892 Catalogue of Seals in the Department of
Manuscripts in the British Museum, vol. 2, London

BOWLES, W.L., 1835 Annals and Antiquities of Lacock Abbey, London

CHESSHYRE, H., and WOODCOCK, T., 1992 Dictionary of British
Arms, London

HUNTER BLAIR, C.H., 1912 ‘Durham Seals IV’ Archaeologia Aeliana,
3rd series, vol. VIII

— 1943 ‘Armorials upon English Seals from the Twelfth to the
Sixteenth Centuries’, Archaeologia 89, 1-27

JOHNES, R., 1960 “The Seal Matrix of Queen Isabel of Hainault and
Some Contemporary Seals’, Antiquaries Fournal 40, 73-6

MARKS, R., 1993-4 ‘Sir Geoffrey Luttrell and Some Companions:
Images of Chivalry c.1320-50’, Wiener Jahrbuch fiir
Kunstgeschichte xlvi/xlvii, 343-355

SEARLE, W.G., 1897 Onomasticon Anglo-Saxonicum, Cambridge

SIDDONS, M.P., 1991 The Development of Welsh Heraldry, Aberystwyth

A Medieval Heraldic Roundel from Potterne
by PAUL ROBINSON

An heraldic roundel made of copper alloy with a flat
back and a hammered surface was found at Ryeleaze
in Potterne in 1983 and subsequently acquired by
Devizes Museum (accession number 1983.90;
Figure 1). With a very slightly convex upper side, it
measures between 43 and 45mm in diameter and
1.7mm in thickness. The heraldic design is inlaid
with enamel over a gilded base; the decorated border
and frame of three groups of a straight-sided, pointed
trefoil flanked by two leaves are similarly gilded.
Because of corrosion caused by long burial, the
detail of the design cannot be properly seen on an

ordinary photograph but it does appear very clearly
on the x-radiograph, kindly supplied by the Wiltshire
County Council Conservation Officer, as a cross
engrailed ermine with a crescent in dexter chief
(Figure 2). Degradation of the enamel makes its
original colour uncertain. The arms shown are,
however, certainly those of Robert Hallam, Bishop
of Salisbury (1407-1417), whose arms also appear
on his tomb at Konstanz Cathedral (The Right Rev.
the Bishop of Salisbury (john Wordsworth) 1889,
234ff., which supersedes Kite 1860, 97ff. and
p1.32) as well as upon his seal attached to British

NOTES

Library Add. Charter 19648 (Birch 1887, vol. 1,
344). Since at least 1254 the Bishop of Salisbury had
been Lord of the Manor of Potterne, prebendary,
rector and owner of the advowson. A manor house
was built in the village for the bishops’ use and
Bishop Robert is known to have resided in the village
on a number of occasions.

Originally the roundel would have been enclosed
in a circular rim attached by lugs or studs to the
surface of another object ‘to identify its owner.
Heraldic roundels clearly marked prestigious objects
and have been discussed most recently by Egan and
Pritchard (1991, 181-84) who suggested a number
of alternative functions for them. They may have
been attached to a case carrying a travelling chalice
and paten, as proposed by Dunning (1965, 54ff.);
they might be from morses, i.e. the clasps for
ecclesiastical copes; they may have been set in the
base of a maplewood mazer; or they may have
embellished an elaborate sword belt of a type similar
to that illustrated on the effigy of the Black Prince at
Canterbury Cathedral.

The roundel from Potterne is one of a small group
which shows the arms of important churchmen of the
first half of the 15th century. Other examples include
that from Rievaulx Abbey with the arms, it is believed,
of Abbot John III (Dunning, op. cit.); and an unprove-

137

nanced roundel with the arms of Henry Beaufort,
Bishop of Lincoln and Winchester (Catalogue, Spinks
coin auction 40, 6—7 Dec. 1984, lot 665). These may
have come from objects which had an ecclesiastical
function. In passing it may be noted that the only
other medieval heraldic roundels from Wiltshire are
those found at Minety (Devizes Museum Day book
807) apparently showing the arms of a member of the
Bassett family, and at Norton Bavant (zbid., 1679)
showing the arms of a member of the family of
Berkeley of Beverstone, both of which possibly came
from objects which had a secular function.

What seem to be the same arms as those on the
Potterne roundel also appear on a late medieval floor
tile found at Milton Lilbourne and presented to
Devizes Museum by Mr Hungerford Penruddocke
(accession number M 158c). This object is at
present unique; it has not previously been illustrated
and is shown below in its present worn state and with
the design reconstructed (Figures 3 and 4). The
arms were originally identified as those of William
Wotton, Abbot of Cirencester Abbey who died in
1440 (WAM 10, 328). There is, however, no
evidence to show that William Wotton bore the arms
shown on this tile, although Burke (1884) does cite
one Wotton coat of arms as ‘sable, a cross engrailed
ermine’, or that Cirencester Abbey held Milton

Figure 1. a: enamelled roundel from Potterne, scale 1:1; b: x-radiograph; c: inlaid floor tile from Milton Lilbourne,
scale 1:4; d: reconstruction of the arms of Bishop Hallam, based on c, scale 1:4

138 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Lilbourne in the later Middle Ages. It is much more
likely that the tile from Milton Lilbourne also depicts
the arms of Robert Hallam, Bishop of Salisbury, and
that it belongs to a small group of early 15th-century
tiles with the arms of senior churchmen of the
Salisbury diocese. Other tiles in this group display
the arms of William Alnewyke, Archdeacon of
Sarum from 1420 to 1426 and Simon Sydenham,
Dean from 1418 to 1431.

Acknowledgements. I am very grateful to Dr Penelope Rundle and to
Nick Griffiths for their help in writing this note and to Nick
Griffiths for also providing the illustrations.

BIBLIOGRAPHY

BIRCH, W. DE G., 1888 Catalogue of Seals in the British
Museum, London

BURKE, J., 1884 A General Armory of England Scotland and
Ireland, London

DUNNING, G.C., 1965 ‘Heraldic and Decorated Metalwork
and Other Finds from Rievaulx Abbey, Yorkshire’
Antiquaries Fournal 45, 53-63

KITE E. 1860 The Monumental Brasses of Wiltshire, London
and Oxford

THE RIGHT REV. THE BISHOP OF SALISBURY, (i.e. John
Wordsworth) 1889 ‘On the Seals of The Bishops of
Salisbury’, WAM 24, 220-243

EGAN, G., and PRITCHARD, F., 1991 Medieval Finds from
Excavations in London: 3 Dress Accessories c.1150 —
c. 1450, London

An Addition to the Stourhead Canon
by ROBERT J. MAYHEW

The links between Samuel Johnson and the Hoares
of Stourhead are few. Johnson’s only reference to the
Hoare family comes in a letter to Mrs Thrale and
refers to their bank: ‘a draught on Hoare for one
hundred and twenty six pounds’.' It is almost
certainly the case that Johnson never visited
Stourhead. His only visit to Wiltshire came in 1783
when he stayed with William Bowles at Heale
House. Johnson was impressed with Heale: ‘A good
house is it, but rather too modern and _ too
convenient to serve the imagination, but the lawn
and the hill, and the thickets, and the water, are
almost equal to the fancy of a TROUBADOUR.”’ This
stay also included trips to Salisbury and Stonehenge.

If Johnson did not visit Stourhead, the recipient
of his letter referring to the Hoare family, Mrs
Thrale, did, subsequent to his death and her second
marriage to Gabriel Piozzi. Mrs Piozzi mentions
briefly her ‘lean accounts of Stourhead, Wilton,
Southampton, &c.’ in her letters,’ a reference which
becomes more significant in the lght of her
Observations and Reflections made in the course of a
journey through France, Italy, and Germany. Mrs
Piozzi had gone abroad partly to avoid the criticism
Gncluding Johnson’s) of her marriage, and partly
to see Italy, the birthplace of her new husband.

1. Bruce Redford (ed.), The Letters of Samuel Fohnson (5 vols.,
Oxford 1992-4), iii, p. 86.

2. Ibid, iv, p. 194.

3. Edward A. and Lillian D. Bloom (eds.), The Piozzi Letters:
Correspondence of Hester Lynch Piozzi, 1784-1821 (3 vols. to
date, Newark, N.J. 1989-) i, p. 238.

After her return to England she published her
Observations and Reflections, which she tidied up for
publication in 1789, just after the tour which took
her to Stourhead (the above cited letter is dated 17
July 1787). Mrs Piozzi ‘loved prospects’* and used
her recent visit to Stourhead in two places in her
travel account. Reflecting on the view from the tower
at Cremona, she was led to a comparison with the
view from Stourhead:

Prospects, however, and high towers I have
seen; that in Mr. Hoare’s grounds, dedicated
to King Alfred, is a much finer structure than
this, and the view from it much more
variegated certainly; I think of greater extent;
though there 1s more dignity in these objects,
while the Po twists through them, and distant
mountains mingle with the sky at the end of a
lengthened horizon.’

Writing of Verona, Piozzi is led to another such
comparison in ‘an agreeable garden belonging to
some man of fashion’. She continues, ‘the grotto
disappointed me: they had not taken such advantages
of the situation ... and I recollected the tasteful
creations in my own country, Pains Hill and Stour
Head.°

4. James L. Clifford, Hester Lynch Piozzi (Mrs. Thrale) (2nd. edn,
corrected, Oxford 1968), p.115.

5. Hester Piozzi, Observations and Reflections made in the course of a
journey through France, Italy, and Germany (2 vols., London
1789), i, p. 116.

6. Ibid., i, p. 130.

|
|
!
|

}

NOTES

These references seem to have eluded Kenneth
Woodbridge, who does not cite Mrs. Piozzi in
Landscape and Antiquity or The Stourhead Landscape.’
Piozzi has a further claim to the attention of those
interested in the Hoare family. Her daughter, Sophia
Thrale, married Henry Merrick Hoare, brother to
Henry Hugh, 3rd Baronet.* Given the sensitivity to

landscape aesthetics Mrs. Piozzi demonstrated

7, Kenneth Woodbridge, Landscape and Antiquity: Aspects of
English Culture at Stourhead, 1718 to 1838 (Oxford 1970), and
idem; The Stourhead Landscape (no place, 1982).

139

throughout her journals and published writings, this
seems a worthy if minor addition to the Stourhead
canon. As Johnson himself put it, “There is nothing,
Sir, too little for so little a creature as man. It is by
studying little things that we attain the great art of
having as littke misery and as much happiness as
possible’.° Lovers of Stourhead and of Johnson
would probably agree.

8. Clifford, op. cit., p. 419.

9. James Boswell, Life of Johnson, G.B. Hill (ed.) and L.F. Powell
(rev.) (6 vols., Oxford 1934—50), i, p. 433.

Pen Pits and Sir Arthur Bliss
by RAYMOND J. SKINNER

Writing over a century ago about Selwood Forest,
which has now disappeared from most maps, Canon
J.E. Jackson observed:

A mile or two beyond Stourhead, there is. . . on
the high ground thereabout, a large square-
shaped piece of table-land, a sort of platform,
the sides of which are steep declivities. On this
platform stands the little scattered village of
Penselwood. Pen is a very commonly-found
Welsh word, meaning head, and so the name
signifies, not improperly, the head of Selwood.
On the slope of this platform, facing east, lie
the celebrated Pen Pits.!

Geographically, Penselwood in Somerset is adjacent
to the point at which the three counties of Wiltshire,
Somerset and Dorset meet, and the neighbourhood
contains much of interest, not only for antiquaries,
but also for those interested in music and musicians.
Around the village of Penselwood on Pen Common
and partly in Wiltshire, there is a series of curious
mostly conical pits in the ground which it is now
thought, may have been excavations made by those
in search of Greensand which could be used for
querns or millstones. Jackson described Pen Pits
rather prosaically:

The surface of . . . the common is scooped out
very irregularly into hollows or pits — some
large, some small, some roundish, basin-like,

1. Revd. Canon J.E. Jackson, ‘Selwood Forest’, WAM 23 (1887),
p. 269.
2. Ibid., p. 270.

others more of a square or oblong shape. They
are in no sort of order, but occur at intervals;
some are close together, divided by a partition
bank, along which you may find your way
about from one to the other.... The pits on this
ridge are said to be spaced over 100 acres.
But they did, within memory, spread also over
the platform at the top, covering altogether
700 acres; a vast number have been filled in
and levelled for agricultural use.?

Sir Richard Colt Hoare of Stourhead was perhaps
the first antiquary to write on Pen Pits and to ponder
over their purpose. He never quite made up his mind
what they were:

either 1) excavations for simple purpose of
procuring stone, or
2) that the ancient Britons may have
made them in searching for mill or
grind-stones, or
3) they were inhabited as places of
refuge in times of danger.’

That Colt Hoare was not the first to advance the
habitation theory, however, can be seen in the
following excerpts from a paper read to the Society
of Antiquaries on 29 April 1784. On this occasion
the Hon. Daines Barrington was speaking of ‘certain
remarkable Pits or Caverns in the Earth, in the
County of Berkshire’. These were situated just

3. Sir Richard Colt Hoare, Ancient Wiltshire, vol.
1810), p. 35.

1 (London

140 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

across the Wiltshire border, about half a mile west of
Little Coxwell near Faringdon. They were, it is said,
referred to in a survey of 1687 as Cole’s Pits and
extended over 14 acres, being 273 in number and
from 7 to 22 feet deep and up to 40 feet in diameter.
Barrington continued:

... having thus endeavoured to show what
could not have been the cause of digging so
many pits i.e. mining, as suggested by the
name, clay for bricks or tiles, or stone quarries.
I shall now risque [sic] a conjecture as to what
may have been the original inducement for
removing so many thousand cubical yards of
sand. I conceive this area to have been a
considerable city of the Britons in the time of
the earliest inhabitants of this island, which at
an average of five souls (to be accommodated
in each pit) would amount to nearly 1,400."

In the later 19th century, excavations were
undertaken at Pen Pits by members of the Somerset
Archaeological and Natural History society, as well
as by the celebrated archaeologist General Pitt
Rivers. Members of the Somerset society disagreed
about the purpose of the pits: some inclined to the
view that they were quarries, whilst others found this
theory untenable for, as was observed, ‘would
anyone have worked a bed of stone in this way’.
Jackson, too, found the quarry theory to be dubious,
and inclined to an argument propounded in great
length and detail by T. Kerslake of Bristol, that the
pits were used in some way for the accommodation
of inhabitants, particularly those taking refuge
from an invader, and would have been protected
with a superstructure of wood. Kerslake associated
Penselwood and its pits with ‘A Long Lost
UnRomanised British Metropolis’, Caer Pensauelcoit,
the latter part ‘coit’ being Welsh for wood, and the
whole having, as he noted, a striking similarity to the
present-day name of Penselwood.?

In the year 1016 there was a battle between
Edmund Ironside, King of the English, and Cnut,
who had invaded England in the previous year. On
Aethelred’s death in 1016, both became rivals for the
English crown, and after several battles made peace,
Cnut taking Mercia and the North and Edmund the
South of the country. When Edmund died shortly
after, Cnut became King of the whole of England.

4. The Hon. Daines-Barrington, ‘An Account of Certain

Remarkable Pits’, Archaeologia 7 (1785), pp. 235-243.

5. T. Kerslake, Caer Pensauelcoit —A Long Lost UnRomanised British

Metropolis (London 1882), Wiltshire Tracts 89/30, WANHS

Library.
6. J. Hutchins, The History and Antiquities of the County of Dorset

Writing on the parish of Gillingham, Dorset, J.
Hutchins observed:

The first mention ... of this place is in the
Saxon Chronicle, A.D. 1016, on account of
the battle fought between Edmund Ironside
and Canute, at Peona, or Penn, co. Somerset....
This action happened so near this place, that
some less exact historians style it the battle of
Gillingham; in which the Danes were entirely
defeated.... At Penn are very remarkable pits,
where the field of battle is supposed to have
been; they are very numerous and regular,
made for offence and defence; some for the
main body, some for advanced guards. Tradition
says they were made by Canute, which is
confirmed by an old MS. in the hands of the
late Mr. Biggen, one of the lords of the manor.

A footnote by a subsequent editor of this work
observed that ‘Hutchins’ account of the Pen-pits,
which were probably much more ancient, is very
doubtful. The most generally accepted opinion is,
that they are the remains of early habitations’.®
Notwithstanding Pitt Rivers’ failure to find evidence
of human habitation, this does seem to be the theory
favoured by most early writers and investigators on
the subject. Canon Jackson, particularly, enjoyed the
idea of ‘half-subterranean wigwams with conical
roofs’, and ‘the idea of our ancestors having shown a
preference for burrowing, like rabbits, in dry chalk
and soft sand’.’ Though recent researches support
the suggestion that they are quern stone quarry pits,°®
it is not impossible to suggest that some of the Pen
Pits may actually date back to the prehistoric period
as a site of human activity, probably therefore being
in existence over the long period stretching from
early antiquity into the centuries of Anglo-Saxon
England. Pen Pits may thus have seen both the
warriors of Caractacus and the Roman legions, as
well as the men of Edmund Ironside and the
invading Danes.

In 1934, about nineteen years before his
appointment as Master of the Queen’s Music, Sir
Arthur Bliss (1891-1975) bought a_ thirty-acre
wooded site on the south-eastern edge of the
plateau, and had a house built there in the then
avant-garde ‘International Style’, for which the
architect was a close friend, Peter Harland. Bliss

(3rd edn., 4 vols., Westminster 1861-70), 3, p. 615.

7. Rev. Canon J.E., Jackson, ‘Notes on the Border of Wilts and

Hants’, WAM 21 (1885), p. 340.

8. M. Rawlings, ‘Archaeological Sites Along the Wiltshire Section

of the Codford—Ilchester Water Pipeline’, WAM 88 (1995), pp.
42-44.

NOTES

had, for some time, wished for a country retreat, and
here the composer with his wife and two daughters
escaped the London scene for many springs and
summers to live there for about twenty years. It was
at Pen Pits — or rather in a small studio set in the
midst of the pits in the adjacent woods — where some
of his most effective music was created: scores for the
ballets The Miracle of the Gorbals (1944), and Adam
Zero (1946) and an opera, The Olympians (1949). As
Bliss explained in his autobiography, the Pits were
very evocative:

In some places the rims of the craters touched,
giving the ground the appearance of having
been shelled by howitzers.’

This comparison was an understandably poignant
one for Bliss, who had indelible memories of his
experiences not twenty years before, during the
Battle of the Somme, in which he had seen his
brother, Kennard, as well as many of his comrades,
killed. The time at Pen Pits was a very productive
period in Bliss’s life, and the strange and beautiful
surroundings of the house, to which his thoughts
often turned when on his many foreign tours, no
doubt inspired much of his work. Understandably,
Sir Arthur Bliss seems to have kept his semi-retreat
at Pen Pits something of a secret.

The studio, about fifty yards from the house, still
exists — an idyllic spot — although the timber-framed
structure today wears a rather forlorn and neglected
air.!° As the composer himself wrote:

From my windows I saw nothing but trees, and
the only sounds were those made by the wind
passing through them. Pheasants would make
their rough nests within a few yards, and quite
likely a fox, unconscious of my presence, would
lollop by; badgers had their setts within view.!!

The studio is built on the edge of one of the pits,
supported by a brick-built column which rises from
the bottom of the pit. Because of its sylvan situation
and the building materials used, the interior suffered
from damp, which badly affected the composer’s
piano, strings tending to rust and then to snap with
an alarmingly sudden noise. An open grate and a
chimney only partially remedied matters. In late
May and early June, the ground is still covered by a
mass of bluebells, which supersede the earlier
9. Sir Arthur, Bliss, As I Remember (London 1970), p. 103.

10. The present owners of the house have informed the writer that

the interior of the studio was restored in 1991, for the 100th
anniversary of the composers’ birth, and that it is hoped to

141

prolific snowdrops and daffodils.

It is paradoxical that Pen Pits, with a history
stretching back into obscure antiquity, should have
become Bliss’s home just at the time when he was
asked by H.G. Wells if he would be prepared to
collaborate by writing the score for a projected film
based on the author’s futuristic vision of the world,
The Shape of Things to Come. As with many film
projects, the financial pressures of having to appeal
to mass audiences caused a dilution of the power-
ful vision of its creator, and the film eventually
became just an exciting entertainment. As Bliss
explained:

I knew he (H.G. Wells) wanted his story of the
probable future to be an educative lesson to
mankind, to emphasise the horror and
uselessness of war, the inevitable destruction
of civilised life, the rise of gangster
dictatorship and oppression. He felt that only
the direction of far-sighted planners with the
use of scientific inventions in the cause of
peace could lift the world into a new era of
prosperity and enlightened leisure.!”

Wells’ vision of the future has come to pass very
much as he prophesised, but the film did not have
the influence on world events for which its author
had sincerely hoped.

Bliss later formed a suite of pieces from the film
music (1936), including the famous March and the
Ballet for Children, The March in particular, is
exciting, highly-coloured music with a fine swagger
to its Elgarian tune and perhaps serves as the best
introduction to Bliss’s music. Like Elgar, too, he
could write successful ‘popular’ music; a popular
note, however, has little place in the major works of
either composer. Further parallels with Elgar also
exist in the rich texture of both composers’ works
and these elements, together with a mastery of
orchestration and a fine musical craftsmanship,
suggest reasons for the choice of Bliss to succeed Sir
Arnold Bax as Master of the Queen’s Music in 1953.
This post, often perhaps an ‘establishment’ sinecure,
found, however, in Bliss, a composer willing to
expend much care on ceremonial works. His
achievement, exemplified in a work such as the
March of Homage in Honour of a Great Man — for Sir
Winston Churchill’s funeral — is an outstanding
piece of such occasional music.

restore the exterior in due course (letter, 23 June 1994).

11. Bliss, op. cit., p. 103.
12. Ibid., p. 105.

142 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Another interesting collaboration occurred
during the period that Bliss was living at Pen Pits
after the last war when, as a result of a meeting with
J.B. Priestley at the 1945 Cheltenham Festival, the
idea of an opera was born. This, later to be known as
The Olympians, was composed between 1945 and
1948. As Bliss stated:

To compose a full-length opera is a long and
fatiguing task, and it took me the best part of
the next two and a half years to complete the
three-act score. Most of it was composed in
the quiet remote music-room that has been
built for me in the woods of Pen Pits.'?

The work’s evolution was the subject of many
interesting and detailed letters between the composer
and librettist — those of Bliss written from Pen Pits
and Priestley’s from his house on the Isle of Wight.
This correspondence was initiated by Priestley in a
letter dated 1 August 1945:

Dear Arthur,

I enclose a very rough synopsis of the opera
plot. I don’t want to work out the story in any
greater detail until you feel fairly satisfied about
the rough outline. I am still uncertain about the
exact form of the finales to Acts I & III, and it is
possible that you might have a musical idea you
favour that would help me to find the kind of
‘curtain’ you need musically. ... I see a great
deal more in these scenes than I have put down
here, but I want to get the rough outline first,
particularly as I dislike typing out synopses and
have no secretary here.!?

In a reply dated 18 August, Bliss commented at
length on Priestley’s suggestions; his concluding
paragraph reads:

I am back in London September Ist. If by any
chance you go to Russia earlier, wire me, and
I will come up to London for a couple of
hours’ discussion before you go. I think it 1s
important that we meet before you get
filled with new visions inspired by the Bolshoi
theatre!"

A recent biography of Priestley observed that:

Priestley’s cantankerous, grumbling image
dissolved into that of the perfect collaborator
13. Ibid., p. 170.

14. Ibid., p. 171.
15. Ibid., p. 172.

described by Bliss as ‘generous and sensitive’.
Nine months were to elapse before music and
libretto began to fuse into the first act, and in
that interval Priestley visited Russia.!°

Back in England, and about a year later, composer
and librettist were concerned about the title of the
opera. Priestley’s letter of 8 April 1947 read:

Dear Arthur,

I am still worrying about the title. “The Gods
Grow Old’ has much to recommend it. ...
But... it has a rather melancholy ring that in
fact does not suit our piece... . I have spent so
much time — an unusual thing with me, by the
way — on this title, that I seem to have gone
stale on it, and perhaps we need some fresh
minds on the problem.

Yours ever, J.B.P."’

A further year of hard work continued until on 5
May 1948 Bliss wrote:

Dear Jack,

‘Today at 2.30 I brought down the final slow
Curtain on a beautifully poised A major chord.
I am very pleased with this last Act. ... IT am
going up to London for a week on Monday
next, and after that re-immerse myself in the
orchestral scoring. I am determined to cut
nothing from Acts II and III, and am with you
over all the love scenes.

Perhaps shortly I could spend a couple of days
at Billingham, [Priestley’s house on the Isle of
Wight] or you come here, for a final
appraisement.!®

Arthur

The opera was accepted for production by Covent
Garden in 1949, but a less propitious time for the
mounting of a large-scale work such as The
Olympians could hardly have been chosen. Just
after the war, when austerity was a major factor and
when conditions at Covent Garden, as Bliss
observed, were not really favourable to the prod-
uction of a work of these dimensions, it was hardly
surprising that the opera was in general poorly
received. Apportionment of blame for this is now a
redundant exercise, but at the time there was
apparently dissension between the musical director
16. V. Brome, 7B. Priestley (London 1988), p.294.

17. Bliss, op. cit., p. 178.
18. Ibid., p. 179.

NOTES

and the production and design team. It is difficult
enough, particularly in the 20th century, to bring to
birth a successful opera; and many are the examples
of works which failed at their first performances.
Some, like Puccini’s Madama Butterfly, later became
successful after revision, but unfortunately The
Olympians was not accorded a second chance. The
composer’s feelings at seeing the result of two and a
half years’ work nullified can only be imagined.
Bliss, however, bore the very mixed criticisms of the
work with great fortitude: He later wrote resignedly:

... I had urged that the final dress rehearsal
should be open to the critics. This was
disastrous. What they saw was a well rehearsed
and produced first Act, even an exciting one, a
second Act that obviously needed a few more
days’ polish, and then a third Act which
looked exactly, in its raw state, like some
village charade. Difficulties in this final Act
were not lessened by the mysterious absence
of the scene designer.

I knew we were lost, and on the first night
sat gloomily in the Garrick Club, only going to
the Opera House in time to thank those to
whom under difficult conditions, genuine
thanks were due.!”

Another major work conceived during the years
spent at Pen Pits had been the Piano Concerto
commissioned for the 1939 World Fair in New York.
The Concerto was dedicated to the people of the
U.S.A. and there is much of Bliss’s own ancestry
moulding his musical thoughts in this work for,
although born in England, Bliss was American on his
father’s side. Furthermore his wife, Trudy Hoffman,
was American, having been born in Santa Barbara,
California. The music of the Concerto reflects a
certain transatlantic verve and ‘get up and go’ and,
like its composer, was energetic and forthright. It
has proved to be one of Bliss’s most enduring works.

In general, it is valid to observe that no artist can
live for long in any environment without at least
some of its ambience colouring his works. Numerous
examples of such influence might be cited from

19. Ibid., p. 179-180.

20. Sir Arthur Bliss died 26 March 1975: The Times, obituary, 29
March 1975. At a memorial service in Westminster Abbey held
in the following May, at which the Queen was represented, one

143

widely differing sources. In Sir Arthur Bliss’s case,
the so-called abstract works, such as certain parts of
the Piano Concerto, and some of the chamber and
string works seem, to one listener at least, to
encapsulate something of the English countryside of
Hardy’s Wessex, particularly in the bitter-sweet
feeling of nostalgia which pervades certain passages
in these works.

There exist many more works than could be listed
or discussed here. The Times, in the composer’s
obituary, mentioned one hundred and thirty from
his prolific pen, with the Colour Symphony (1922) —
inspired by a book on heraldry — marking the start of
the mature works. Sir Arthur Bliss conducted his last
concert in January 1975, about two months before
his death.”°

At present, Sir Arthur Bliss’s music has apparently
temporarily fallen out of fashion, as the paucity of
current performances demonstrates. No doubt this
trend will change, however, for there is too much of
value in his work for it to be permanently neglected.
As he wrote of his predecessor as Master of the
Queen’s Music, Sir Arnold Bax (1883-1953):

The rapidity with which one musical fashion
succeeds another has for the moment relegated
Bax’s music to some lumber-room, where it lies
awaiting a new generation that will admire its
uninhibited musical flow and romantic expres-
sion. ... I have seen many reputations rise and
sink, and some which before my birth seem
buried for ever now exhumed with full honours.
Musical reputations seem to move around like
the slats on a water mill, first ascending to a
peak of admiration, then descending to a depth
of neglect, before once more climbing the
ascent towards renewed appreciation.”!

It is to be hoped that the twentieth anniversary of
Bliss’s death in 1995 will herald the ascent of this
mill-wheel to a position more commensurate with
his musical achievements.

Acknowledgements. The writer is indebted to Mrs Pamela Colman,
Sandell Librarian, WANHS, for the suggestion that Sir Arthur
Bliss’s house and its surroundings deserved closer investigation.

of the lessons was read by ex-Prime Minister Edward Heath:
The Times 21 May 1975.

21. Bliss, op. cit., p. 192. Bax had been Master of the King’s Music
(1942-52), and of the Queen’s (1952-3).

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp.144—153

Excavation and Fieldwork in Wiltshire 1994

Amesbury: Countess Road (SU 1542);
Multiperiod

A field evaluation programme was undertaken
between October and December 1994 over an area
of approximately 30ha between the River Avon and
Countess Road to the north of the A303, and along
a 40m wide corridor immediately north of the A303
between Countess Roundabout and King Barrow
Ridge. The work was undertaken by Wessex
Archaeology and Alastair Bartlett Associates on
behalf of English Heritage and The National Trust
as part of the assessment of a series of possible sites
on which the Stonehenge visitor centre and
associated facilities could be relocated.

On the Countess Road site 415 test-pits, each 1m
square, were hand excavated and the content of the
topsoil quantified. Magnetometry was also applied
here and along the corridor linking Countess Road
with King Barrow Ridge. None of the areas investi-
gated were found to contain extensive archaeological
deposits of such importance as to preclude devel-
opment; all, however, contain some archaeological
remains.

At Countess Road evidence of Saxon and
medieval activity was found in the southern and
southeastern sector of the investigated area as might
be expected near the town of Amesbury. An area of
ill-defined prehistoric setthement represented by a
low-density flint scatter was identified south of the
former military railway running through the site.
North of the old railway the lower ground was found
to be badly disturbed by an extensive gravel pit
exploited during the construction of the A303 north
of Amesbury. At the far north end of the site, on
slightly elevated ground, there was an area of low-
density flint scatter. Augering suggested that any
prehistoric land surfaces beside the Avon were
deeply buried and below the present water-table.

The corridor linking Countess Road with King
Barrow Ridge was found to contain a number of
geophysical anomalies towards the western end,
including some linear features, part of the encircling
ditch of a round barrow, and the parallel earthworks
of the Stonehenge Avenue. At the eastern end very
few anomalies were revealed, most being explained
as old field boundaries and agricultural features.

A full report, which also contains the results of
earlier evaluations on King Barrow Ridge, has been
submitted to the County Sites and Monuments
Record.

Amesbury to Berwick Down: A303
improvement;

Prehistoric and Romano-British

As part of the examination for a suitable route for
upgrading the A303, additional field evaluation
using fieldwalking (surface artefact collection) was
undertaken by Wessex Archaeology. The work was
commissioned by Sir William Halcrow and Partners
on behalf of the Highways Agency. Three areas were
examined amounting to 43 hectares, centred on SU
103435, SU 093424 and SU 064041, respectively.
Concentrations of worked and burnt flint were noted
within each of the three areas. The majority of
datable finds were flint artefacts of Bronze Age date.
Other finds included a small quantity of pottery and
ceramic building material, and a concentration of
Romano-British sherds occurred at the western end
of the study area. Geophysical surveys of nine areas
were also undertaken by Geophysical Surveys of
Bradford. These showed, in the case of the
westernmost area (SU 067410) that amongst other
features, a settlement identified earlier on aerial
photographs was larger than previously known. The
responses in the other survey areas were, in general,
poorly defined.

Amesbury: Stonehenge (SU 123 422);

Prehistoric (Neolithic and Bronze Age)

The Ancient Monuments Laboratory of English
Heritage carried out a series of geophysical surveys
at Stonehenge during 1994 as part of a project
commissioned by English Heritage from Wessex
Archaeology to publish an account of the structural
development of Stonehenge based on the primary
records of all the 20th-century excavations. A
magnetometer survey of the whole triangular area of
land containing Stonehenge was undertaken and this
was supplemented by a detailed resistivity survey of
the monument itself. The aim of the magnetometer
survey was to provide evidence of archaeological
features in the immediate landscape setting of
Stonehenge, while the resistivity survey aimed to
verify details of previously excavated features and the
location of former archaeological interventions at the
monument.

The magnetometer survey provided valuable new
detail of the 7 barrows comprising the Amesbury
4-10 group, west of Stonehenge, and mapped the
course of a linear ditch crossing the north-west
corner of the Triangle. The latter feature is a

EXCAVATION AND FIELDWORK IN WILTSHIRE 1994

continuation of the palisade ditch (probably of later
Neolithic date) excavated by Vatcher and Vatcher in
1967 at SU 1217 4228 in the area now occupied by
the pedestrian underpass (Vatcher, H. and L, 1968).
The resistivity survey did not produce any
momentous new discoveries about Stonehenge itself
but was nonetheless able to confirm the presence
and arrangement of several previously poorly
understood features of the monument. These include
entrance causeways across the ditch to the south of
the stone-settings, the circuits of Y and Z holes and
the counterscarp bank. The resistivity data also
provides a useful and complete plan of the current
condition of the sub-surface monument, which will
act as an aid to its management in the future. The
geophysical surveys are to be fully published in the
forthcoming English Heritage Archaeological
Report ‘Stonehenge in its Landscape’ (R.M.J. Cleal
et al., 1995).

REFERENCES

CLEAL, R.M.J.. WALKER, K.E., and MONTAGUE, R., 1995,
Stonehenge in its Landscape: Twentieth-century Excava-
tions, English Heritage Archaeological Report 10

VATCHER, F. and L, 1968 ‘Excavation and Fieldwork in
Wiltshire 1967’, WAM 63, 108

Ashton Keynes: Cotswold Community School
(SU 033 962); Late Bronze Age/Romano-British;
Medieval and Post-medieval

In advance of proposed sand and gravel extraction,
an archaeological evaluation was made on land to
the north of the school at Somerford Keynes by
Wessex Archaeology in May and June 1994.

The archaeological significance of the site was
indicated by an extensive set of cropmarks. Located
within the proposed development area, these
cropmarks appeared to represent a series of
rectilinear enclosures and associated trackways,
with one particular concentration representing a
possible Romano-British settlement. In addition, a
ring-ditch of possibly prehistoric date was located
in the western portion of the proposed extraction
area. In 1983, a field-walking project recovered a
scatter of Romano-British pottery from an area of
possible settlement. In 1988, a machine trench
evaluation of this area confirmed that the
concentration of cropmarks was a small Romano-
British nucleated settlement, consisting of a series
of adjacent small enclosures, surrounded by a
network of larger field enclosure ditches. To the
north and north-west of the current evaluation
area, a Late Bronze Age settkement was excavated

145

in 1992. Several post-built round-houses were
recorded, along with a circular gully and several
other circular and rectilinear structures. There was
also a group of very large pits, closely associated
with the occupation areas. Along with an important
ceramic assemblage, the finds from the site
provided evidence for metalworking, textile
production, and crop processing.

The most recent evaluation revealed numerous
archaeological features, predominantly ditches, pits,
and post holes, with dating evidence suggesting both
Late Bronze Age and Romano British activity.
Although recorded across the entire area, these
features were generally concentrated within the
northern and central southern part of the site. Many
of the linear features were already recorded as
cropmarks during the 1988 evaluation which
examined the northern half of the proposed
extraction area. The remains of extensive medieval
and post-medieval ‘ridge and furrow’ were found
cutting the natural subsoil, particularly within the
western portion of the evaluation area.

Avebury and its environs

Preparatory work by RCHME on this project began
during the year. The study area includes much of
the north Wiltshire chalk downs, as well as the greensands
and clays of the Pewsey Vale and much of north
Wiltshire south of the Upper Thames Valley. A
desk-top study involving limited field
reconnaissance, aerial photographic assessment and
documentary research has been undertaken to
assess the potential of the region. This has revealed
a staggeringly rich earthwork environment with
considerable potential for ground survey. The
prehistoric monuments on the chalk, such as
Avebury, Windmill Hill and West Kennet have
already been surveyed by RCHME; additional work
will record for the first time the well-preserved
medieval and post-medieval settlement pattern,
including the remains of a number of monastic houses. So
far, after the initial sample reconnaissance, a number of
new sites have emerged, including prehistoric and Roman
enclosures and settlements, extensive areas of shrunken
medieval settlement and in at least one case a completely
deserted example to the west of Rowde. Intensive
fieldwork will begin in the autumn of 1995.

Bradford-on-Avon: Greenland Mills

(ST 83106060); Post-medieval

An archaeological evaluation by AC archaeology in
advance of potential redevelopment of the site.
Several trial pits revealed deposits of archaeologically

146 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

sterile flood silts adjacent to the river Avon and
confirmed documentary evidence for former
quarrying on the south side of the site.

Bradford-on-Avon: Winsley Road

(ST82056113); Romano-British

An archaeological evaluation by AC archaeology in
advance of a proposed housing development. A
natural, largely undisturbed, soil sequence was
recorded across the site and no subsoil features of
archaeological interest were present. A single sherd
of Romano-British pottery and a modern coin of
Arabic origin were recovered from the topsoil.

Broad Chalke: Bury Orchard (SU 041252);
Prehistoric

An archaeological field evaluation was carried out by
AC archaeology in November 1994 on the site of a
proposed development. Four machine-excavated
trenches were dug south of scheduled ancient
monument AM451 (the presumed medieval
boundary of the village or manor). The investiga-
tions revealed no evidence for medieval activity
within the application area, but did locate a linear
ditch (tentatively dated to the Bronze Age), and an
area of formerly waterlogged deposits which may be
the remains of a (natural) pond, also of prehistoric
date. These lower-lying parts of the site were sealed
by colluvial (hillwash) deposits which contained very
small quantities of late prehistoric worked flint.

Bulkington: land at Lawn Farm (ST 3941 1584);
Medieval

A field evaluation was carried out by the Oxford
Archaeological Unit in January 1994 on a develop-
ment site adjacent to Christ Church, Bulkington,
north of the Bulkington to Keevil road. Three
trenches were excavated and revealed a number of
ditches and pits containing medieval pottery. The
majority of the pottery was early medieval, dating
from the 11th to 13th centuries. Most features were
concentrated away from the road, towards the centre
of the development site.

In April 1994 Chris Bell excavated an area
extending 65m back from the main village street. A
ditch ran most of the length of the site, evidently
marking out a plot parallel to the east side of the
churchyard. There was no clear evidence of a building
on the street frontage, although a stone surface may
have included remains of a building. To the rear were
cross ditches dividing the plot, one of which extended
across the adjoining plot. The most intensive area of
medieval activity was well. back on the plot and

included pits and curving ditches, with glazed roof
tile; one ditch had poorly preserved organic deposits.
The distribution of features confirms the evaluation
results, with most features identified in the trench
lying across the centre of the site.

Calne: land adjacent to Oxford Road

(SU 0049 7280); Roman and Medieval

The Oxford Archaeological Unit undertook a field
evaluation at the beginning of July 1994. The site lies
west of the road and 2 miles north of Calne. The
work was done on behalf of Dalton Warner Davis
Associates, in connection with the proposed
construction of a roundabout, a new section of road
and a straw-to-energy plant. Five trenches were
excavated.

A particular concern of the evaluation was to
locate any activity contemporary with the deserted
medieval village of Beversbrook, which survives as
an area of extensive earthworks immediately north of
the site. The only archaeological features found were
two shallow ditches, one of which contained a piece
of Roman tile, along with several sherds of Roman
pottery and the remains of some medieval ridge and
furrow ploughing. The absence of medieval features
indicates that the village of Beversbrook did not
extend into the area of investigation but that this area
was used for cultivation.

Cherhill: Cherhill Down, (around SU 04806975);
Modern

A watching brief was carried out by AC archaeology
during the laying of a new water pipeline across
Cherhill Down. The work revealed archaeologically
sterile colluvium or modern deposits associated with
previous pipe trenches and former military buildings.

Chute: Tibbs Meadow (SU 2984 5392);
Late Medieval, Post-medieval
Wessex Archaeology was commissioned by Halsall
Construction Ltd to conduct an archaeological
evaluation of land situated on the south-eastern edge
of the village in advance of proposed housing devel-
opment. Work consisted of a detailed earthwork
survey, five hand-dug test pits and three machine
trenches. The archaeological features revealed were
directly associated with the earthworks in the eastern
area of the site. These comprised a ‘platform’ to the
south which was linked, by an area of amorphous,
shallow earthworks in the north, to a linear feature.
Late prehistoric, late medieval, post-medieval
and modern artefacts were recovered from the
topsoil in all five test pits. Late medieval and post-

EXCAVATION AND FIELDWORK IN WILTSHIRE 1994

medieval finds were recovered from archaeological
features associated with the linear earthwork.

Compton Bassett: Compton Farm (SU 036722);
Prehistoric to Post-medieval

A programme of test pitting, by the Compton Bassett
Area Research Project (CBARP), was undertaken to
complement earlier exploratory fieldwork (see WAM
88 (1995), 149). The area investigated lies at the foot
of the scarp of the lower chalk shelf and has
undergone continual geomorphological change. In
certain trenches, up to four buried land surfaces were
recorded separated by deposits of hillwash material.
Finds from buried land surfaces included pottery of
probable Early Bronze Age date and worked flints,
including a barbed and tanged arrowhead.

Compton Bassett: Dugdales Farm (SU 029736);
Post-medieval and Modern

Trial excavations were undertaken by CBARP to test
for the presence of an oval feature observed as a soil
mark on post-war RAF aerial photographs. No
features were present and finds were limited to post-
medieval and modern material.

Compton Bassett: Freeth Farm, Mill Pond

(SU 029727); ?>Medieval

A section was cut by CBARP through the more
north-easterly of the two probable mill dams
identified in 1986 (Currie 1994). The dam was
largely composed of sand and clay and the section
indicated episodes of repair and consolidation. No
dating evidence was recovered.

REFERENCE

CURRIE, C.K., 1994 ‘Earthworks at Compton Bassett, with
a Note on Wiltshire Fishponds’ WAM 87, 96-101

Compton Bassett: Freeth Farm, Oak Bed

(SU 022725); Roman

A programme of test pitting was undertaken by
CBARP with the aim of providing a context for
material collected casually by Mr J. Henly. Although
no features were located a ceramic scatter was
isolated and may represent the site of a small
farmstead.

Compton Bassett: Manor Farm (SU 034731);
Medieval

Earthworks to the south of Manor Farm were
surveyed by CBARP as part of the study of the
development of the road network of the locality. The

147

survey revealed further evidence for the course of the
former route parallel and to the west of the present
village street (see WAM 88 (1995), 149).

Corsham: Easton village (ST 891706); Medieval
Planning of a group of earthworks opposite Easton
Court Farm was undertaken by the Chippenham
College Practical Archaeology Group (CCPAG).
The remains appear to be of one of the medieval
farms of the village which, unlike Easton Court
Farm, has not survived to the present day.

Cricklade: Prior Park School (SU 101935);
Medieval

Field evaluation by the Cotswold Archaeological
Trust on behalf of Prior Park School, identified two
12th—13th-century pits.

Devizes: Drews Pond (SU 006 597); Post-
medieval

An archaeological excavation of the supposed site of
an old hostelry was requested by the Drews Pond
Wood Project, prior to the area being converted into
a sensory garden. A thatched cottage which occupied
the site was demolished in 1955.

The excavation produced no evidence of this
having been the site of a hostelry and indeed
produced very little in the way of small finds or
building material from any period.

The work was carried out by the Archaeology
Field Group of WANHS as a training excavation for
its members, under the leadership of Chris Chandler.

Donhead St Mary: Church of St Mary

(ST 90672445); Romano-British/Medieval

A limited archaeological watching brief was carried
out by AC archaeology during the installation of a
new heating system in the church. The work revealed
the line of the original north wall of the nave and
produced a number of stray finds from beneath the
nave floor, including Romano-British pottery of
second to third century date.

Everleigh: Beach’s Barn (SU 184 510);
?Prehistoric, Roman

Geophysical survey was undertaken by the Ancient
Monuments Laboratory of English Heritage (EH) at
Beach’s Barn in response to a request from the EH
Monuments Protection Programme (MPP). The
villa (discovered by William Cunnington in the late
19th century) was under consideration for protection
as a Scheduled Ancient Monument (SAM) but the
full extent of the site was unclear. Test pitting work,

148 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

undertaken by Reading University in 1993
(Entwistle et a/., in press) as part of the EH funded
‘Salisbury Plain Project’, located the edge of a
potentially substantial building beneath approxi-
mately 0.8m of flinty colluvium. Considerable
quantities of Iron Age and late Roman pottery as
well as fragments of limestone were also recovered
from an arable field to the west of the site in a spread
which extended about 100m beyond the main
excavations. Resistivity and magnetometer surveys
were carried out in March 1994 with the aim of
placing the partially excavated building in its wider
context and also investigating the artefact scatter in
the field to the west.

The results of the resistivity survey were rather
disappointing and, although three different mobile
probe separations (each looking progressively deeper)
were employed, no buried walls were detected. The
lack of response was presumably due to a poor
contrast between the foundations and _ their
surroundings, and/or to their depth of burial. The
magnetometry, which covered an area in excess of
3ha, proved far more successful and a range of
interesting features was mapped. The most striking
of these was a 5m-wide ditch running straight
through the survey, dividing it into an area of
intensive occupation-type activity to the north
(including pits and enclosures) and a_ distinctly
‘quieter’ area to the south. A rectilinear pattern of
negative anomalies was also detected within the
northern area, in a region of generalised magnetic
disturbance, which may well represent the found-
ations of a former building. Further geophysical
survey would help to confirm this.

REFERENCE

ENTWISTLE, R., FULFORD, M.G., and RAYMOND, F. (in
press) The work of the Salisbury Plain Project, English
Heritage monograph

Grafton: Batts Farm (SU 268616)

Field evaluation near the centre of Wilton village
by the Cotswold Archaeological Trust, on behalf
of Michael Fowler Architects, obtained negative
results.

Horningsham: Longleat Estate (ST 830430);
archaeological field inspection

In February 1994 Wessex Archaeology was
commssioned by the Longleat Estate to conduct a
survey of all archaeological monuments on the
Estate. This survey involved field visits to all 28
surviving sites; comment was made on_ their

condition, state of repair, state of land management
and on potential threats from modern activities. The
survey concluded that the Estate’s sites fell into three
broad categories: first, those which were well
preserved and were under no imminent threat;
secondly, those which were under direct threat
through the type of land management currently
practised; and thirdly, those sites which have been
largely destroyed or eroded by agricultural activity.
Recommendations for future management aimed at
preserving the sites were defined.

Kington St Michael: Churchyard (ST 903772)

At the request of the parish churchwardens, a grave
plan was made by CCPAG, in advance of building
work.

Latton: Wharf (SU 101942); Roman, Post-medieval
Field evaluation, by Cotswold Archaeological Trust
on behalf of the Highways Agency; identified Roman
Ermin Street in the expected location. It was 7m
wide, although little of its metalling survived above
the clay mound upon which it was founded. The
ditch on the north-east side of the road proved to
have been recut on a number of occasions: beyond it
lay a low bank and another ditch and fence line.
Spreading out for 9m beyond this was a humic
deposit containing 3 coins (latest: AD 350-60), late
3rd—4th-century pottery, and fragments of tile and
oyster shell. This might represent midden material
redeposited from the settlement at Cricklade. ‘To the
north of the present road linear cropmarks proved to
be associated with the post-medieval Fairford road.

Ludgershall: Adjutant’s Press, Butt Street

(SU 263510); 2>Medieval

An evaluation was undertaken by Wessex Archaeo-
logy in November 1994 within the Ludgershall
Conservation Area, close to the standing earthworks
of Ludgershall Castle and St James’ Church. The
work was commissioned by L.J. Sturges Esq.

Two trenches, measuring 10m in length, were
excavated by machine. Trench 2 revealed part of a
single inhumation burial which rested at approx-
imately 0.90m below the present ground level in a
flat-bottomed grave. The inhumation was identified
by the presence of a pair of tibiae, probably of an
adult in a supine position aligned roughly east-west.
No upper leg or torso bones were visible, and the feet
were inaccessible under the trench edge. The tibiae
were resting in the base of a sub-rectangular grave,
the full dimensions of which lay outside the edge of
the trench. Detailed cleaning of the trench face

EXCAVATION AND FIELDWORK IN WILTSHIRE 1994

revealed that the grave was in fact cut from
immediately below the modern footings, a level
higher than that at which it had become visible. The
alignment of the inhumation might suggest a
medieval date, although neither a Romano-British
nor Saxon date can be excluded.

Malmesbury: Market Place, (ST 933873);
Post-medieval
An archaeological watching brief was maintained
during the later stages of a resurfacing programme in
November 1994. Prior to the involvement of AC
archaeology, deposits up to 500mm _ below the
existing ground surface had been removed from the
area north-west of the Market Cross monument;
previous archaeological investigations suggest that
disturbances at these depths have removed various
post-medieval resurfacings of the market place and
may have impinged upon medieval levels.

Observations in the remaining north-east area of
the Market Cross revealed a homogeneous layer
which contained building and resurfacing debris,
animal bone and pottery of probable 18th-century or
later date. A cobbled surface butting up to the kerb
stones of the Market Cross was also recorded; this is
believed to be a modern feature.

No features were recorded in the area beneath the
road on the east side of Market Cross as this had
been heavily disturbed by service trenches.

Melksham: The Hurn Site (ST 902639);
?Prehistoric

An auger survey was carried out, by Cotswold
Archaeological Trust on behalf of Meyer Ltd, to
determine whether archaeological deposits or
sediments of palaeoenvironmental interest existed
on land adjacent to the river Avon. A possible
palaeochannel was identified running parallel to the
present course of the river, while on the higher
ground a series of terrestrial erosional deposits were
found to have originated from near the church.

Minety: Upper Minety (SU 013911 ); Medieval

A second season of excavation by CCPAG was
undertaken at the medieval pottery production site.
No kiln was found but the foundations of a timber
building and quantities of wasters were discovered to
the west of the road uncovered in 1993.

Orcheston: Tilshead barrow (SU 053482);
Prehistoric

A survey of a damaged round barrow on West Down,
near Tilshead, was carried out by AC archaeology.

149

The survey involved a photographic survey, contour
survey of the existing land surface and scanning to
record the presence of artefacts on the eroded
barrow mound. Sufficient of the mound and ditch
remained to be recorded by the contour survey and a
low density of worked flint was recorded across the
site.

Salisbury: Bishop Wordsworth’s School

(SU 145 269); Medieval

An archaeological evaluation by Wessex Archaeology
was undertaken in March 1994 prior to an
application for planning permission for redevelop-
ment within the medieval walled precinct of the
Cathedral Close. The work was commissioned by
Bishop Wordsworth’s School on behalf of Wiltshire
County Council Education Service.

A total of six hand-dug and five machine-dug
trenches were excavated, and a number of wall
footings were located, including the north wall and
chalk floor of a building of probable medieval date.
Material recovered from excavated layers was
predominantly of post-medieval date, although
small quantities of medieval material were also
present.

Salisbury: Downton Road (SU 147284);

Bronze Age

The land subject to a development proposal covers
an area of roughly seven hectares and is located
approximately 1.5km south of Salisbury city centre.
Overall, the site is positioned on a moderately steep
north-facing slope which leads down to _ the
floodplain of the River: Avon. An evaluation was
undertaken by Wessex Archaeology at the request of
the Bell Cornwell Partnership. Twenty-nine machine
trenches were dug revealing a number of archaeo-
logical features.

A ring-ditch recorded on the SMR was identified
and partially excavated. A series of ditches was also
revealed and the recovered artefacts suggest activity
in the area during the Bronze Age, contemporary
with or later than the barrow cemetery indicated by
the presence of ring-ditches and an extant barrow
mound. Trenching in the eastern part of the
proposed development area revealed colluvial
deposits, over a metre in depth. An assessment of the
land snails from this colluvium indicated high
potential for detailed analysis of the environmental
background and thus a greater understanding of the
landscape history. A similar potential can be
ascribed to samples extracted from the excavated
ring-ditch.

150 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Salisbury: Ivy Street/Brown Street (SU 146298);
Medieval

Ivy Street and Brown Street in Salisbury form the
south-east corner of Antelope Chequer, one of the
blocks of land (or ‘chequers’) which resulted from
the laying out of the new city of Salisbury on a regular
grid pattern in or around 1219 AD. Although no
previous archaeological fieldwork has been carried
out within Antelope Chequer, the site is adjacent to
the continuation of New Street (now known in this
area as Ivy Street) which was probably one of the
earliest streets to be laid out in the new city.

Over a period of four and a half weeks
commencing 12 July 1994, Wessex Archaeology
conducted the excavation of 320 square metres.
These works were commissioned by Salisbury
District Council in advance of proposed residential
development.

The excavation mainly identified archaeological
deposits of the medieval period, specifically those of
the 13th and 14th centuries. Buildings of this period
lay along the Brown Street frontage with a burgage
wall running from the frontage into the backlands.
Rubbish pits and cess pits of this period were also
found within the backlands, one of the latter being
contained within an outhouse at the rear of the
buildings.

Very few artefacts of the later medieval period
were recovered and activity at this time appears to
have been very limited. In the post-medieval period
larger pits were dug into the earlier deposits in the
backlands. The burgage boundary wall appears to
have been superseded and a large building along the
frontage spread into the neighbouring burghal plot.
A stone-flagged yard may be related to this building.
In modern times, much of the area was levelled and
used as the forecourt of a commercial garage. A
vehicle inspection pit was one of a number of large
intrusive features which cut into the underlying
archaeological deposits.

Salisbury: Old George Mall (SU 144298);
Medieval/Saxon
Three stages of archaeological work were carried out
by Wessex Archaeology prior to redevelopment of
the southern half of the Old George Mall shopping
centre within the city of Salisbury; these included a
desk-based assessment, site evaluation, and area
excavation between Nos. 60 and 76 New Street.
The work was commissioned by Trafalgar House
Construction Management Ltd.

The excavation, within a 17x13m_ perimeter,
revealed good sequences of stratified deposits,

including substantial 13th-century buildings set
broad side on to the street frontage, with associated
floors and hearths. There is evidence to suggest that
the buildings are, in part at least, commercial rather
than simply domestic and are accompanied by
external features and structures of an industrial
nature. Large quantities of stratified artefacts and
ecofacts have been recovered, including organic-
tempered Saxon pottery, and an_ extensive
programme of bulk soil sampling has been conducted
on im situ primary deposits rich in_ palaeo-
environmental materials. The bulk of the structural
and depositional evidence survived at the front of the
site where nearly 400 separate contexts were
identified within an area less than 10m square and
0.40m deep. The rear of the site (where an evaluation
trench was located) revealed little other than 17th- or
18th-century cess pits and cellars with evidence of
substantial soil reworking and deposit attrition from
19th-century gardening. In summary, medieval
deposits and structures survived almost completely
undisturbed immediately below the modern ground
level at the front of the site but had been almost
completely truncated at the rear. Full post-excavation
assessment will not commence until all fieldwork is
complete, probably in the late spring of 1995.

Salisbury Plain Training Area: Multiperiod
Fieldwork by the RCHME has now been completed
on the SPTA. During the year survey work focused
on the Bulford Ranges where an important
prehistoric landscape was investigated. A number of
small enclosures of presumed Late Bronze Age date,
an associated field system, linear ditches and round
barrows were surveyed.

Other major sites surveyed during the year
include the deserted settlement of Imber Coney
which is located immediately west of the better
known, and larger, Imber village which was
evacuated this century. The survey of Imber Coney
showed it to have been a small planned settlement
comprising up to 6 domestic units belonging to a
manorial complex which complements that at the
larger sister village.

Other field targets during the year have included
the landscape assessment of the Chapperton Down
area. This not only included detailed survey of the
Romano-British village there, but also looked at the
linear ditch which was subsequently used as the main
street for the settlement which spread over a distance
of 1.5km. Elsewhere, in the vicinity of Chapperton
Down, work was carried out on the extensive remains
of prehistoric and Roman field systems.

EXCAVATION AND FIELDWORK IN WILTSHIRE 1994

Further to the east, detailed ground examination
also took place on the Thornham Down field system
which surrounds the Charlton Down Romano-
British settlement complex reported last year. As
expected, the field system was well-developed and
long-lived. It incorporates an earlier, prehistoric core
subsequently remodelled in the Romano-British
period. Overlying this there is the superimposed
weave of ridge and furrow ploughing of medieval or
post-medieval date.

The remains of a number of post-inclosure
farmsteads on SPTA were also examined. These
were tenant farms, built by landowners and
following ‘model’ plans as part of the period of
agricultural improvement in the 19th century. Most
failed, lasting only one or two generations before
abandonment.

The RCHME, in association with Geophysical
Surveys of Bradford, has also continued with its
programme of geophysical survey on SPTA and has
recorded the following sites.

1) The remains of an aisled villa lying immediately
to the north of the Salisbury Plain escarpment at
Edington.

The site of the destroyed 14th-century chapel at

St Joan a Gore. This proved to be a double-

celled building 9m wide and 20m in length.

ili) Sheer Barrow, a plough-flattened mound of
Neolithic date, was investigated. Geophysical
survey confirmed the presence of a mound and
encircling ditch, recalling the morphology of a
number of long barrows recorded in Cranborne
Chase. The survey picked up traces of an
internal mortuary enclosure and at least one
other oval-shaped enclosure immediately to the
north of Sheer Barrow.

iv) Further geophysical survey work was undertaken
on the western boundary of the enclosure and
villa complex at Figheldean. (See WAM 86
(1993), 14). The enclosure can now be seen to
follow a rather eccentrically ‘zig-zagging’ course,
suggesting that it is mirroring an_ earlier
landscape feature.

The final component of the RCHME project, the
aerial transcription, which has involved the comp-
uterised recording of all visible archaeology on the
SPTA at a scale of 1:10000 is, similarly, reaching a
conclusion. The work has been highly productive,
locating a number of previously unknown prehistoric
and Roman settlements, as well as adding new detail
to those archaeological complexes already known.

151

Swindon: Haydon Wick (SU 1220-8820);
Medieval

A watching brief, by Cotswold Archaeological Trust
on behalf of Thames Water, was conducted during
the cutting of a new sewer main over a distance of
2km on the northern fringe of Swindon. A pottery
scatter of broadly 10th—12th-century date was
recovered to the east of an earthwork south of Park
Farm Cottage, and known earthworks through which
the pipeline cut were recorded in section. The sewer
cut deep into geological levels and numerous fossil
finds of ammonites and bivalves in the Upper Oxford
Clays proved to be of palaeontological interest.

Swindon: Rushey Platt (SU 136837); Bronze Age
The prehistoric barrow at Rushey Platt was
investigated in 1922 by A.D. Passmore who located a
possible cist slab. The barrow was subsequently
scheduled as an ancient monument (Wilts. 668) but
it had been assumed that it had been entirely
destroyed in recent times during the preparation of
this low lying area for development. Accordingly, a
field evaluation by Cotswold Archaeological Trust
was commissioned by J.J.H. Homes Ltd and
Wiltshire County Council to clarify whether the
barrow did indeed survive. Trenching revealed a clay
mound, representing the damaged remains of the
barrow, in the anticipated position; the considerable
depth of modern land-fill across the site had
protected the mound to some degree.

Swindon: The Hermitage, Old Town (SU 1598375);
Roman, Saxon, and medieval

An archaeological excavation was undertaken by
Wessex Archaeology for Jephson Homes Housing
Association in advance of construction at The
Hermitage north of Dammas Lane. The site
comprised an area of c.725 square metres. A buried
topsoil, containing mainly Roman pottery of Ist to
2nd century date, together with nine sherds of Saxon
pottery, 34 sherds of medieval pottery, and 100 pieces
of worked flint, covered much of the site. Some
features, including a north-south ditch at the western
side of the site and a possible backfilled quarry to the
north-east, may have cut this soil. Several shallow
intercutting gullies, possibly associated with buildings,
were recorded at the southern side of the site, all
truncated by or lying below the soil. Truncated pits
and post holes were also recorded, one of which
contained an incomplete infant burial. Another
incomplete infant burial, possibly marked by
displaced natural stone, was found near the base of
the buried soil.

152 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Tidworth: Dunch Hill (SU 205486);

Early/Late Bronze Age

The area of Dunch Hill is situated near two extant
barrows and to the south of a possible Late Bronze
Age settlement. The Defence Land Agent commis-
sioned Wessex Archaeology to evaluate a section of
earth trackway, on the western slope of the hill, prior
to the construction of a new hard-surfaced track. In
the evaluation area, a number of important sites are
documented on the County’s Sites and Monuments
Record, including the two round barrows, one of
which is Scheduled, and a classic ‘Celtic’ field
system, ploughed but still surviving as earthworks.
Within the same field, Reading University had
discovered, during field walking, a scatter of later
Bronze Age pottery, burnt flint, and bone. At the
same time they had excavated a small pit which was
exposed within the trackway and contained Late
Bronze Age pottery. This led to the assumption that
the remains of a Late Bronze Age settlement had
been disturbed by the use of the track.

Wessex Archaeology excavated a single trench on
the course of the trackway. The evaluation established
the presence of a number of archaeological features
along the line of the proposed hard surfaced track,
seven of which were examined detail; six of them
produced artefacts of Early to Late Bronze Age date.

Trowbridge: Court Street (ST 856576);
Post-medieval

Four test-pits excavated by Cotswold Archaeological
Trust on behalf of Alder King revealed only modern
cellar walls, foundations and pipe-trenches.

West Overton: Park Farm (SU 155653):
Mesolithic

A scatter of Mesolithic flint debris was found during
fieldwalking by CCPAG in a field to the south of
West Woods long barrow.

West Tisbury: Tisbury Golf Course (centred on
ST 937288); Prehistoric/Romano-British

A fieldwalking survey was carried out by AC
archaeology at Wick Farm, Tisbury, in advance of a
proposed golf course development. A concentration
of pottery and worked flint was identified around ST
935289 and is considered to be the southerly
extension of an area of Iron Age and Romano-British
activity recorded around ST 935291 (Wilts County
Council SMR ST92NW550). The survey failed to
provide any surface evidence for activity associated
with the medieval settlement of Wyck (Scheduled
Ancient Monument 838).

Wilton: 13a Russell Street (SU 0976 3126);
Medieval

In September 1994 the OAU undertook a field
evaluation where it was proposed to build three
terraced houses. The work was carried out on behalf
of Peter Borchert Architectural Design Consultants.
The site lies within the Saxon town of Wilton which,
during the 8th century, was the capital of Wessex and
remained an important town in the region until the
late 13th century.

Four small trenches were excavated to a
maximum depth of 1.2m. Alluvial clay was located at
the bottom of all of the trenches, and this was
overlaid by a uniform layer of slightly organic clay
containing quite large quantities of early medieval
pottery, oyster shells, and small fragments of animal
bone; this appeared to indicate the dumping of
domestic refuse. The only archaeological features
located were an undated pit and a post-medieval
ditch. No Saxon remains were found during the
evaluation, and none of the medieval deposits were
related to structures. The presence of alluvium
perhaps indicates that this area was too prone to
flooding for occupation in the early medieval period.
The apparent dumping of domestic refuse at this
period may have been an attempt to raise the level of
the site, but more probably is to be seen as evidence
of manuring for agricultural or horticultural
purposes.

Winterbourne: Salt Lane, Winterbourne Gunner

(SU 18253521); Prehistoric and Anglo-Saxon

Two further phases of investigation of the Anglo-
Saxon cemetery have been undertaken by AC
archaeology. This has involved evaluation of the
west portion of a former house plot, known as
Camerton, and of adjacent land to the north.
An area excavation within the house plot site,
behind the Salt Lane frontage, was subsequently
undertaken.

The excavation revealed two further graves of
Anglo-Saxon date, along with a single, probably
prehistoric, crouched inhumation adjacent to a
previously unknown pond barrow. It can be shown
that the position of the pond barrow had been
respected and avoided by the Saxon cemetery. The
barrow was not further excavated and has now been
protected as a Scheduled Ancient Monument. To
the north of the pond barrow evidence of a Middle
Bronze Age urn field and a second, ploughed-out,
round barrow was recovered. Two cremation urns
were removed but the site was not further
disturbed. A full report is in preparation.

EXCAVATION AND FIELDWORK IN WILTSHIRE 1994

Yatesbury: All Saints Church (SU 063715);

?Late Anglo-Saxon and Medieval

Excavations by CBARP were undertaken to the
south of the church with the aim of investigating the
boundaries and interior of the sub-rectangular
enclosure whose north-west quarter is occupied by
the churchyard.

At its western end the interior of the enclosure is
up to 1.5m above the ground level outside. A cutting
through the western boundary revealed a series of
steps marking subsequent re-cutting of a holloway.
The latest holloway of the observed sequence was
surfaced with tightly packed pebbles placed directly
on top of the chalk bedrock. Two horseshoes of
14th-century date were found lying on this surface.
At the eastern end of the cutting approximately one
quarter of a substantial sub-rectangular pit, containing
domestic rubbish, was excavated and is provisionally
dated to between the 10th and 12th centuries.

A second cutting immediately to the south of the
present southern boundary of the churchyard
revealed a shallow gully running parallel to the
existing boundary and, further south, a sarsen stone
set into a circular pit which is interpreted as a post-
pad. Both features were sealed by a layer of chalk
rubble which contained medieval pottery with a 12th
— 14th-century date range. A fragment of glazed
floor tile dated to the first quarter of the 14th
century was recovered from the topsoil and it is
suggested that this and another plain glazed
fragment, from the holloway. sequence, may be
debris related to alterations to the adjacent church.

Yatesbury: Manor (SU 062715); Medieval

Two trenches were cut by CBARP across the bank
and ditch enclosure parallel and to the west of that
which contains the churchyard (WAM 88 (1995),
154). The composition of the bank and the filling of
the ditch indicated one major phase of construction.
Deposits within the enclosure comprised apparently
water-lain silts and the feature is interpreted as a
pond of medieval date, based on sherds recovered
from the ditch.

Yatesbury: Manor Farm (SU 065716);

Prehistoric to Post-medieval

Excavations by CBARP were continued within and
around the earthwork enclosure and other associated
earthworks first identified in 1992 (WAM 87 (1994),
157-8; 88 (1995), 154). A series of trial trenches was
cut with the aim of elucidating the nature of activity

155,

within the enclosure. Evidence for its northern limit
was found in the form of substantial intercutting
ditches similar, but fewer in number, to those
recognised in previous cuttings to the south and
west. The ditches appear to belong to the later part
of the sequence, c.1750 and earlier, Further trenches
revealed dense post-medieval occupation debris,
including two enigmatic rectangular chalk rafts,
measuring 1m x 2m, set side by side. A cutting was
made through the western tail of the east—west
mound, the eastern part of which has been shown to
be an Early Bronze Age round barrow (WAM 88
(1995), 154). This revealed a sequence of burnt
deposits, containing medieval and ?late Anglo-Saxon
pottery, within a ditch or pit. This feature was sealed
by ploughsoil in the form of a headland which has
given the mound its elongated appearance.

A cutting was made in the south entrance of the
enclosure as indicated by surviving earthworks. The
primary objective of locating ditch terminals, to
establish the presence of an entrance, was achieved.
The enclosure has been related to the early road
system of the region and research has indicated that
the Anglo-Saxon herepath (army way) which runs
into Avebury from Marlborough continued to
Yatesbury and beyond, following the line of the
Lower Chalk Shelf. The evidence is discussed in
greater detail elsewhere (Reynolds 1995).

A cutting through ridge and furrow to the south-
west of the enclosure was intended to provide a
broad date range for cultivation based on
examination of material deposited by manuring from
domestic middens. Very little medieval pottery was
found, but a ditch of late Roman date, running
east-west, was sectioned.

To the south-west of the enclosure, between the
church and the ridge and furrow, trial excavations
were undertaken to evaluate the archaeological
potential of the site. Substantial archaeological
remains were encountered, including groups of
intercutting pits, part of a masonry structure
apparently constructed upon a chalk raft and
concentrations of post holes. One 2m x 1m test
trench revealed eight post holes, three of which were
set along the bottom of a narrow trench. Excavations
in 1995 are to be concentrated in this area.

REFERENCES

REYNOLDS, A.J., 1995 ‘Avebury, Yatesbury and _ the
Archaeology of Communications’, Papers from the
Institute of Archaeology 6

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp.154-160

Reviews

Michael Aston and Carenza Lewis (editors).
The Medieval Landscape of Wessex. Oxbow
Monograph 46, 1994; viii + 280 pages; illustrations
and maps. £28, hardback. ISBN 0 946897 78 6.

This is a collection of thirteen papers by different
hands, with an admirable introduction by the editors,
an index, and a beautifully crafted preface by that
most eloquent of landscape historians, Maurice
Beresford. The Wessex described is essentially
Dorset, Hampshire, Somerset and Wiltshire. Five
chapters concern counties other than Wiltshire, and
are not considered here. Another (of which more
below) is exclusively about Wiltshire, and the rest are
regional in scope. The generous Oxbow format, of
double-column A4 pages, and the illustrated board
binding (rather than conventional casebound), make
this book not over-priced. But I should have
preferred to see fewer proof-reading errors and
confused citations, and a better standard of
indexing. For example, poor old Andrews and Dury
(creators of the 1773 map of Wiltshire) appear in
four different spellings in the course of one paper,
and the only index entries for the Mount at
Glastonbury, and the Ridgeway, are under “The’.

Two excellent surveys of pre-conquest archae-
ology in the region begin the proceedings. Bruce
Eagles tackles the earlier period, and suggests a late
Roman date for East Wansdyke. David Hinton’s
survey of the later period includes discussions of
recent findings from Trowbridge and Cowage, and
some timely scepticism about the theory of
specialised functions within multiple estates. Then
there is a masterly exposition of church organisation
from Roman times to the 19th century, by Patrick
Hase. This is based largely on Hampshire evidence
(as one would expect from the author’s previous
research), but is also of considerable relevance to
Wiltshire. Another highlight of the volume is a
comprehensive discussion of forests, chases, warrens
and parks in medieval Wessex, by James Bond.

Two shorter papers, by Della Hooke on the
administrative and settlement framework derived
from Saxon charters (drawing especially on evidence
from Teffont and Bradon), and by John Hare, on
medieval chalkland agriculture and settlement, are
also very stimulating. The latter has a particularly
good discussion of settlement shrinkage and
desertion. I am less happy about Michael Costen’s

attempt to analyse terms in Saxon boundary
charters. Apart from irritation that many of the Old
English terms are not translated, Dr Costen appears
not to have consulted the various papers written
about Wiltshire charters. Thus he speaks of the
Langley Burrell charter, which was shown
conclusively by Avice Wilson in this magazine (vol.
77) to relate to Kington St Michael and Kington
Langley; worse, he assumes that, because Wansdyke
occurs as a boundary feature in the East Overton
and Stanton St Bernard charters, that those estates
were divided after the construction of the
monument. In. fact, the boundaries here cross
Wansdyke and do not follow it, convincing G.M.
Young, Desmond Bonney and others that precisely
the opposite conclusion was appropriate.

Carenza Lewis writes on ‘patterns and processes in
the medieval settlement of Wiltshire’. Her starting
point is a map of the county, showing different
settlement forms, which she has laboriously compiled
by analysing the 1773 map of Andrews and Dury. It
shows, with striking clarity, the contrast between the
regular and compact settlements of the chalklands
and Cotswolds, and the irregular, dispersed pattern
found in the claylands. To account for this she
explores historical land use, medieval demography,
evidence of desertion and shrinkage, and _ the
distribution of moats; but none of these appears to her
to explain the cause and origin of the regional
differences. She therefore turns to the very problem-
atical evidence of Roman and early Saxon settlement
in Wiltshire, and suggests that, although there were
periods of change between 400 and 1500 AD, the
origins of the medieval distribution are, ‘set firmly in
preceding patterns of settlement and boundaries’.

Stimulating though such a discussion undoubtedly
is, Ms Lewis’s paper is open to criticism and
comment on a number of fronts. She exaggerates the
importance of clothmaking to west Wiltshire by the
early 14th century. In her discussion of moats she
ignores the explanation that many Wiltshire moats
(including the one she illustrates) are related to
medieval parks. She underplays the number of
regular linear settlements on the northern side of
Salisbury Plain — in the Vale of Pewsey, in fact, such
regularity is the norm. Her density maps, 1332 and
1377, appear to use civil, not ancient, parish
boundaries, so that (for example) the entire recorded
populations of the extensive medieval parishes of

REVIEWS

Calne, Melksham, Bradford and Malmesbury St
Paul are plotted in the urban portion, with ‘no data’
recorded for the rest.

Such observations do not affect the main drift of
her paper. But in another respect her argument, in
my view, is fatally flawed. She believes that the
pattern and form of settkements mapped in 1773
were virtually the same as those which existed at the
end of the middle ages, and that the map, ‘pre-dates
most of the changes in the landscape which occurred
with enclosure’. But this is true only of the
chalklands; on the claylands the process of enclosing
the former open fields had been proceeding apace
since the late middle ages, and was all but complete
by 1773. Thus what she has actually mapped are the
settlement patterns and forms of the chalklands
before enclosure, and the claylands after enclosure.
Is not the simplest explanation the correct one — that
the process of enclosure itself, between the 15th and
18th centuries, has created the form of dispersed
settlements to be found in 1773 on the claylands?

Aubrey recognised this process at work.
Commenting on the consequences for the poor of
the disafforestment and enclosure of Pewsham
Forest in the 1620s, he noted (Natural History of
Wiltshire, 1847, p.58), “Now the highwayes are
encombred with cottages, and the travellers with the
beggars that dwell in them’. Anyone who delves
into the evolution of the clayland villages (via VCH
Wilts., for example) will continually find examples
of roadside and common-edge squatter hamlets
springing up, especially during the 17th and early
18th centuries. It may well be, as Ms Lewis argues,
that clayland villages such as Bradenstoke and
Tockenham exhibit elements of continuity from
Roman and Saxon times; but the form in which
clayland settlements existed in 1773 was largely
post-medieval.

Indeed, at one point Ms Lewis presents evidence
in support of just such a view. Referring to deserted
settlements on the claylands, she writes: ‘The fact
that so many of these deserted sites take the form of
regular villages, which are generally uncommon in
this part of the county, is also very interesting:
furthermore very few dispersed [her _ italics]
settlements have recorded evidence for shrinkage
around them.’ Precisely. The effect of late- and post-
medieval enclosure on the claylands was to attack the
old-established (?late-Saxon, according to current
orthodoxy) regular villages surrounded by their open
fields, and to replace them with a dispersed pattern
of farmsteads and hamlets. No need for any Romans
in the equation, at all.

155

The length of my rebuttal of a single paper in this
volume will, I hope, be taken as a measure of the
intellectual stimulation that it fuels, rather than as an
indictment of the field in general. The landscape
history of Wessex is an important, fascinating and
rapidly developing subject, and Mick Aston and
Carenza Lewis have done it, and us, an immeasur-
able service by bringing together this collection of
papers. Long may the debates continue.

JOHN CHANDLER

Richard Bradley, Roy Entwistle, and Frances
Raymond. Prehistoric Land Divisions on
Salisbury Plain: The Work of the Wessex Linear
Ditches Project. English Heritage Archaeological
Report 2, 1994. 181 pages; 78 illustrations (line and
photographic); 58 tables. Price £28, paperback.
ISBN 185074 647 X.

Salisbury Plain is an area of great archaeological
value and sensitivity which has been the subject of
two major studies in recent years. The first, an
archaeological landscape analysis by RCHME,
derived from air and ground survey, covers the entire
area under military ownership and is currently being
prepared for publication. The second, a detailed
study of the later prehistoric land divisions to the
east of the River Avon is the subject of this review.
Conceived as a three year project funded by English
Heritage and executed by a small but skilled team
from the University of Reading, this volume has
been prepared and published with exemplary speed.
The results will be of considerable interest to all
students of prehistoric Wessex.

The opening chapter by Bradley gives a detailed
resumé of the study of the Wessex linear ditch system
from Pitt Rivers to the present day (an excellent
example of archaeological synthesis), and the criteria
governing the selection of the study area. This is
followed by Entwistle on the development and
application of the methodology. Fieldwalking
strategies were designed primarily to identify areas
along the linear ditches which could be targeted for
excavation. The lack of intensive modern cultivation
in the project area was largely responsible for the
recovery of large quantities of prehistoric ceramics
which contributed to the identification of many new
sites. In all thirty small scale excavations were
undertaken. The majority were across ditches,
although a number of lynchets adjacent to the ditch
system, notably on Dunch Hill (where an extensive

156 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

settlement was also located) were investigated. This
reviewer wishes that more work could have been
undertaken on field lynchets to gain a greater insight
into the chronological relationships between field
systems and linear ditches.

The results of the fieldwork have led the authors
to identify a broad morphological scheme and
chronological sequence which they use as the basis
for a detailed analysis of the social and economic
factors behind the division of the landscape between
c.1200BC and c.500BC. Two ‘core territories’,
defined by linear ditches, are postulated between
the Bourne Valley and Dunch Hill. The dating
evidence is convincing for an origin c.l200BC and a
ceramic association between unenclosed Late
Bronze Age settlements and linear ditches in the
study area is established. The authors argue for a
sequence of ditch system maintenance which
undergoes a radical alteration c.800BC-—500BC,
with the amalgamation of ‘territories’ culminating
with the emergence of the hillfort of Sidbury and
enclosed settlements associated with All Cannings
Cross ceramics.

Discussion of the finds and environmental data 1s
of great importance for the region, complementing
and expanding on the work of the Danebury Project.
Raymond’s pottery analysis is lengthy but well
argued, linking the primary phase of ditch
construction with post Deverel-Rimbury ‘Late
Bronze Age Plain Wares’ and noting the rarity of all
but the earliest All Cannings forms from the
unenclosed settlements. Molluscan analysis by
Entwistle has given a detailed picture (fig.70) of the
changing environment between the second and first
millennia BC and will be a welcome addition to the
data gathered by the Stonehenge Environs Project.

The arguments throughout this volume are
considered and well presented. How far they can be
applied across Wessex is more problematic. There is
little discussion of the remainder of Salisbury Plain
west of the Avon where a somewhat different
configuration of ditch systems has been recorded
which, contrary to the examples examined east of the
river, appear to retain an important landscape role
into the Romano-British period. Indeed the Late
Bronze Age and Early Iron Age archaeology of the
study area seems to have closer associations with the
chalklands of the Quarley/Andover area than central
and western Salisbury Plain. It is this reviewer’s
impression that there are highly localised and
specialised components within the Late Bronze Age
and Iron Age landscape of Wessex which desperately
require further research.

This volume has demonstrated the value of fixed-
term problem-specific projects and should be seen as
a model for similar investigations. As with all research
it has posed as many questions as it has provided
answers. The exceptional preservation of the field
archaeology on Salisbury Plain makes it ideal for a
wide range of research projects and the results of the
current investigation by the Reading University team
into the late Iron Age and Romano-British landscape
east of the Avon are awaited with interest.

The authors are to be congratulated on the
execution of the project and the speed with which it
has been made available in publication. The volume
has few typographical errors and the line illustrations
are crisp and clear although the photographs were
rather dark in the review copy.

MARK CORNEY

D.A. Crowley (editor). A History of Wiltshire,
Victoria History of the Counties of England
Series, Volume 15: Amesbury and Branch and
Dole Hundreds. Oxford University Press for
Institute of Historical Research, 1995; xxii + 338
pages; illustrated. Price £70, hardback. ISBN 0 19
722785 6

The publication of this volume brings the Wiltshire
series to three more than the next largest VC.H.
series (Oxfordshire), and at its launch there was
discussion about whether Wiltshire now has the
most copious county history ever produced. None of
those who took part offered to put the matter to the
test by making detailed word-counts to compare our
VC.H. with such giants as Nichols’s Leicestershire or
Hasted’s Kent, so the matter remains in doubt. What
is not in doubt is the quality and quantity of the
research embodied in these recent VC. H. volumes.
The two hundreds treated cover a compact area
to the north of Salisbury in the valleys of the Wylye,
the Avon and the Bourne, still largely rural. The only
towns, Ludgershall and Amesbury, were small, and
it was only the advent of the army from 1897
onwards that led to an increase in population around
Tidworth, Bulford and Larkhill camps. It is pleasing
to see that various official sources still in government
hands were made available to the editors so that the
growth and use of the various military establishments
could be described in detail. This volume also sets
out very clearly the growth of such places as
Amesbury and Durrington in the present century.
Few Wiltshire people, I guess, would be able to

REVIEWS

identify a settlement called Tin Town in Brimstone
Bottom; answer on page 156.

In detail a reviewer can only comment on points
that interest him or catch his eye. No new light is cast
on the mysterious burghal status of Tilshead in 1086
— the sources are lacking — though it is plausibly
suggested that its wide main street near the church
may have been the borough’s focal point. New and
unusual is the information on the development of
Wylye on a grid pattern, though I am sorry that
the origin of the name of Teapot Street there did
not come to light (nor, presumably, that of Coffee
Farm in Great Durnford). The Demon Drummer
of Tidworth is dismissed rather summarily as ‘a
poltergeist’ — what has the VC.H. to do with such
imponderables? — and it is a pity that it has not been
possible to identify the house which he troubled. It
could have been worth mentioning the tradition of
the seven children at one birth at Great Wishford
which, however implausible, has been current in the
village since at least the 17th century.

It is pointed out in the Editorial Note that the
support which Swindon Borough and _ then
Thamesdown District Councils have given over the
years (more than one fifth of the cost of compilation)
has now ceased. All interested in Wiltshire’s past
hope that this is only a temporary withdrawal.

K.H. ROGERS

A.P. Fitzpatrick and Elaine L. Morris (editors).
The Iron Age in Wessex: Recent Work. Published
by the Trust for Wessex Archaeology Ltd, 1994,
on behalf of the Association Frangaise D’Etude
de L’Age du Fer; 124 pages, line drawings and
photographs. Price £15, paperback. ISBN 1 874350
11 6.

The publication comprises papers presented in April
1994 at a conference of the Association Frang¢aise
D’Etude de L’Age du Fer, on the theme of the Iron
Age in Wessex. In the foreword, John Collis indicates
how the interest of archaeologists has shifted during
the course of this century from the mighty hill-top
sites such as Maiden Castle to the smaller enclosed
settlements visible both as earthworks and as
cropmarks and soil marks. Much of the detail of these
settlements has emerged through excavation, whilst
survey work by the RCHME has given a picture of the
landscape within which these sites were situated.

The conference report is in two parts. The first
consists of ten essays on various themes of Iron Age

HDi

archaeology, ranging from social organisation (Colin
Haselgrove) to the production and distribution of
Iron Age pottery (Elaine Morris). The second part,
with twenty-three contributions, presents informa-
tion on settlements and landscapes obtained from
archaeological excavation and survey.

In both sections, each report has been kept brief
and succinct and the references for all contributions
are presented in a single section at the end of the
volume. The publication is of particular value in the
range of themes and settlement types which are
represented. In part one, there is a particularly
interesting contribution by J.D. Hill which looks at
the evidence for ritual behaviour on Iron Age
settlements, examining such things as_ the
assemblages of bones and artefacts in pits and
ditches. The strength of part two is in the variety of
plans and photographs illustrating what has emerged
from recent excavation about the internal
arrangements on small settlements. Of particular
interest to the Wiltshire reader are the essays on two
midden sites: Andrew Lawson’s on Potterne and that
of a team from the RCHME reporting on East
Chisenbury.

The volume is well illustrated and well produced.
My single criticism is that the contributions have
been too severely limited in length, presumably
because of the editorial nightmare of overruns by
thirty-three different contributors! As an example,
Mark Corney, in reporting on the RCHME tasks in
Wessex, has been limited to about 1200 words and
one illustration. Nevertheless, the editors are to be
congratulated on extracting contributions speedily
from those who attended the conference and
presenting the results so professionally. At £15, this
is good value for money.

ROY CANHAM

J.L.Kirby (editor). The Hungerford Cartulary:
A Calendar of the Earl of Radnor’s Cartulary of
the Hungerford Family. Wiltshire Record Society,
Vol. 49, 1994; xix + 300 pages. Price £15, hardback.
ISBN 0 901333 26 4.

By the time of Walter, first Lord Hungerford (1378-
1449), the family of that name had risen within a
generation to become one of the most important
families in Wiltshire, having provided, in Thomas
Hungerford, the first Speaker of the House of
Commons. They held extensive estates in Wiltshire,
Somerset, Berkshire and London, and, not surpris-

158 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

ingly, in line with practice common among religious
houses and land-owning families, Walter had a
cartulary drawn up of the deeds of these estates in
order to record title and rights, thus protecting them
from challenge or threat. What is perhaps surprising,
and, in the light of the loss of so many medieval
cartularies, something of an extravagance, two
versions of the Hungerford cartulary, albeit with
some significant differences, survive.

It is to the version belonging to Lord Radnor of
Longford Castle and held with his archives in the
Wiltshire Record Office that Mr Kirby has brought
his considerable editorial skills, maintaining an
interest in the family which dates from his researches
at university in the 1930s. The Radnor cartulary
presents a clear view of the Hungerford estates
centred in Wiltshire at Heytesbury although the
ancestral home was established just over the border
in Somerset at Farleigh Hungerford. The deeds of
each estate, expertly calendared and _ indexed,
provide rich seams of topographical, genealogical
and toponymical material common to medieval
deeds although, of course, unique in each case. The
Wiltshire estates were dispersed throughout the
county, from Heytesbury in the west to Cricklade in
the north, and Hungerford (now entirely in
Berkshire) in the east. In the introduction, Mr Kirby
announces his intention to publish the Hobhouse
version, or more accurately the 45 folios of material
not in the Radnor version, deposited in the Somerset
Record Office together with the surviving section of
another family cartulary. It is unfortunate that it was
not considered feasible to produce a composite text
of all three documents. Furthermore, the decision
to omit the deeds of the chantry of Farleigh
Hungerford on the grounds that they were published
mm extenso by Canon Jackson in 1879 is regrettable. A
modern edition of the Hungerford material in a
single volume would have provided a complete and
readily accessible source.

The deeds document the ownership of the estates
up to and after their acquisition by the Hungerfords
and contain much of interest about other families.
Of particular note are three Inquisitiones Post Mortem,
two for Thomas Seyntomer in 1365 (51, 52) and one
for Hugh Wake held in 1312 (80) which do not
appear among the Public Records. That of
Bartholomew de Badelesmere held in 1329 (632)
lacks a footnote confirming that it exists among the
Chancery Inquisitiones Post Mortem Edw.IIl, file 9,
no.12. There is a little inconsistency in the handling
of undated documents. On occasions, for example,
in numbers 1—29 they are left unassigned for any

period whereas elsewhere, such as 463—475, several
approximate dates are offered. Similarities in witness
lists with dated documents and assessment of other
internal evidence might have allowed the editor to be
more positive. As it is, those undated deeds appear
rather exposed and adrift and would have benefited
from some form of anchorage however crude.

The cartulary contains a large number of
personal names which will allow many individuals to
be identified and placed more clearly, particularly
when their offices are stated. William Brygon, a clerk
of Exeter diocese who was also a public notary and
registrar of William Ayscough, Bishop of Salisbury,
is such an example.

The occurrence of Richard Whittington, in a
deed of 1435, not in his more familiar role as Lord
Mayor of London, but as Mayor of the Staple of
Westminster, is of wider interest to all of us who
received our first smatterings of history in fairy tales.

The most unusual documents in the cartulary are
pedigrees of the Hussey family, the maternal line of
Walter, first Lord Hungerford (737, 738) and of the
de Montfort family of Wellow and Farleigh (749).
The source of the former two pedigrees has not been
identified which makes their appearance in print of
particular note.

All three elements of the volume, introduction,
text and index, are dealt with skilfully by Mr Kirby.
The introduction is a little brief and would have
benefited from more of the content of the editor’s
unpublished thesis. By saving the Hobhouse cartu-
lary for a possible publication at a later date, the
opportunity to discuss and compare both versions is
avoided. Thus on both counts the reader is left
feeling slightly unsatisfied and possibly a little
confused. The most important element, the text, is
superbly presented by one whose editorial skills have
been developed over many years’ work on medieval
sources such as IJnquisitiones Post Mortem and
Wiltshire Feet of Fines 1377-1509, also published by
the Wiltshire Record Society. The 950 documents
have been calendared to a consistently high standard.
The text is more than adequately supported by an
excellent index, both of which received the critical
examination of Dr Roy Hunnisett — in itself a stamp
of quality if one were needed.

One typographical criticism is that the titles to
each group of deeds might have been emboldened,
and the index of estates printed as a separate section
in the introduction. Also, reading through the text or
following up references from the indexes, it is not
easy to identify the estate referred to. But this is a
minor quibble which must not detract from the

REVIEWS

notable achievement of the publication of this major
source for the medieval history of Wiltshire for which
both the editor and the Record Society deserve our
gratitude and praise.

STEVEN HOBBS

Michael Parsons. Farley with Pitton. L. M.
Parsons, 11 volumes to date, 1989-1995; various
paginations and prices; paperback.

It is unusual to find one person with the time,
energy, knowledge and enthusiasm to research
exhaustively the history of a parish and its environs.
Rarer still is it to find that person able to publish his
work. Such a rare bird is Michael Parsons who, in
addition to the eleven volumes reviewed here has
produced two other works on Pitton and Farley since
1985. His contribution to our knowledge of this
portion of south-east Wiltshire is especially impor-
tant as the parish has not yet been researched for the
Victoria History of Wiltshire.

The series began with four books on religious
matters, charities and benefactions. Included is a
history of Sir Stephen Fox’s beautiful hospital which
was begun in 1680 and which remains a notable
building for this area although almost all the trees
which were included in ‘landscaping the gardens and
planting trees’ have long disappeared. Sir Stephen,
who replaced Sir John Evelyn as lord of the manor,
aiso set out to build a fine new church to replace a
chapel in a sad state of repair; the church, of rosy red
brick with stone dressings, was completed in 1690.

Noble Achievements, Honest Men (1993) looks
further at the Fox family and their contemporaries in
the 17th century while A Certain Rule of Wisdom
(1993) covers the period 1700-1819 for manorial
history up to and including the Inclosure Award.
Earlier manorial history, 1100-1699, can be found in
the next volume, The Little Manor of Pitton and Farley
(1994) which contains useful descriptions of the
open fields, tenants and lessees. Michael Parsons is
descended from many generations of landowners
and farmers in the parish and has carried out good
comparisons of landowning and the state of
agriculture at different periods.

The Saxon Inheritance (1995) is a_ recently
published work on the earliest period of history in
the area which was a favoured site of the Romans
with several villas being built in the Dean Valley. A
theme running though many of the books is the land,
agriculture and the forest. In the latest volume,

159

Michael Parsons considers The Royal Forest of Pancet
(1995), later called Clarendon, its influence upon
Pitton and Farley and its economic importance to a
wider area.

This is a considerable body of work and one
which might normally be expected to emanate from
many members of a local history society. It is
valuable not only for the information on one partic-
ular parish but also as a comparative study and guide
for other parish historians. A good range of primary
sources, held at both a local and national level, has
been used, while much of the writing shows that
detailed fieldwork has been undertaken to amplify
the archival record. My only major criticism is the
lack of an index in any volume but it may well be that
an index volume will be produced when the saga is
completed.

MICHAEL MARSHMAN

F.E.Warneford (editor). Star Chamber Suits of
John and Thomas Warneford. Wiltshire Record
Society, Vol.48, 1993; xix + 108 pages. Price £15,
hardback. ISBN O 901333 24 7.

This volume comprises full transcripts of documents
pertaining to five cases involving John Warneford of
Sevenhampton in Highworth between 1539 and
1551 and his grandson, Thomas, in 1611, which the
editor unearthed during his exhaustive genealogical
researches. They are presented with little attention
to their administrative or archival background and
their interest lies exclusively in the glimpses they
provide into the lives of the central litigious
characters which reflect familiar themes of the mid
16th and early 17th centuries. Maurice Beresford, in
his pioneering work History on the Ground, warned of
the dangers of studying history from court records
which offer as distorted a picture of society as the
current tabloid press. This is emphasised in the cases
revealed in this volume since a verdict is only known
in one case, thus making it difficult to discern truth
from fiction and accuracy from distortion in the
depositions. Nevertheless, the ease with which the
parties apparently resorted to violent and
intimidating behaviour is not untypical for the
period.

The cases, although very specific and _ local,
nevertheless reflect tensions and issues that were
played out on the wider stage of 16th- and 17th-
century life. Thus, we are presented with disputes
over common rights, one involving the destruction of

160 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

hedges of newly enclosed land and another over the
creation of a rabbit warren on common land, each to
the alleged loss of rights by Warneford’s neighbours.
The violence in the latter case led to the death of the
ferret belonging to the Warnefords’ warrener. A
dispute in Highworth church between Warneford,
the lay rector and the churchwardens over the
ownership of the stone of the high altar, pulled down
as a result of the Protestant laws of Edward VI, is an
interesting example of the effects of the Reformation.
A case over fishing rights provides a rare if not
unique reference in Highworth to the name
Swannesnest which was, presumably, a stretch of the
river Cole near Sevenhampton.

The decision to publish the documents virtually
in extenso is regrettable on two counts: their verbosity
and repetition cry out for the editor’s red pen which,
judiciously exercised, need not have detracted from
the form and flavour of the originals yet would have
made their contents more accessible. The result is
that considerable powers of concentration are
required to follow through each case, and the reader
is required to adopt skills of summarising more often
associated with studying an original not an edited
text. Furthermore, the cases are adequately
summarised with extensive extracts in the editor’s
book The Warnefords (privately published 1991)
which can be used as a useful commentary alongside
this volume.

The index is small, due to the limited nature of
the text. It is sadly a little uneven: the unfortunate
ferret, killed in the dispute over the warren, is
included but not the watchmen called to respond to
an affray (pp. 8-9). Reference to the bailiff of
Highworth (p.54) is omitted yet is vital in making
sense of the indexed references to the method of his
election (p.60). It is regrettable that the index was
not extended to include the few occupations of
witnesses. Collectively the woollen draper, mercer,
victualler and shoemaker, all from Highworth,
provide evidence of the social make-up of that town
in the mid 16th century. Finally, the unidentified
Glebepeppar (p.58) must be a corrupt version of
Clyffe Pypard. Despite these reservations, the editor
does a good job in evaluating and analysing the
motives of the central characters and their evidence.
There is much of interest from the illustrations of
social and religious tensions to the earthy insults of
the protagonists, notably John Warneford, who
emerges as a devious and dangerous opponent, for
which historians of Highworth and areas further
afield will be grateful. This is a volume slight both in
size and content, and is not one of the Record
Society’s more significant publications. Nevertheless,
it fulfils the Society’s aim to present a wide variety of
Wiltshire material to as large a readership as
possible.

STEVEN HOBBS

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 161-164

Obituaries

Richard Atkinson, CBE (1920-1994). Normally
to be seen sporting a bow tie, Richard Atkinson was
a man of striking personality with a wide range of
interests and talents. His professional career in
archaeology spanned nearly forty years, beginning at
the Ashmolean Museum in 1944. This was followed
by a move to Edinburgh in 1949 to join Professor
Stuart Piggott. In 1958 he took up the Chair of
the newly created Department of Archaeology at
Cardiff, where he remained until his retirement in
1983.

Son of a Dorset farmer, Professor Atkinson was
educated at Sherborne and Oxford. His widely
acclaimed first book, Field Archaeology, was
published in 1946. It was followed ten years later
by Stonehenge, and it is his excavations at this
monument and Silbury Hill which are his best known
connection with Wiltshire. Less well known is his
involvement with the experimental earthwork on
Overton Down. Atkinson was keen on experimen-
tation to explore ideas and techniques, and his oft-
used words “I know because I’ve tried it’ in ancient
technology lectures became a catchphrase amongst
his students. As one of the founder members of the
Experimental Earthworks Committee he was deeply
involved with the design and building of the sites.

As his career at Cardiff developed so did the
amount of administration both within and outside the
University. Bodies with which he became involved
included the Council for British Archaeology, the
Ancient Monuments Board and the University
Grants Committee. The increasing load of this work,
together with ill-health, prevented him from
completing his work on his two Wiltshire sites,
though Stonehenge was revised in 1979.

Atkinson’s work on Stonehenge, including his
discovery of the carvings on some of the stones,
together with his ability as a communicator, led to
work with television which culminated in the BBC2
funded excavations at Silbury Hill in 1968-70. It
was not the most straightforward ‘dig’ to direct,
requiring the welding of a team which consisted of
archaeologists, mining engineers, caterers and
publicity staff with, in the background, one
television crew recording day-to-day events and
every now and then the descent of the Outside
Broadcast Unit to transmit live. With charm and
strength of personality this was achieved.

Though not an active member of the Society
in recent years, Professor Atkinson served as a
distinguished member of our Council from 1967
to 1971.

GILLIAN SWANTON

Grace Fairhurst (1928-1995) was the Society’s
Treasurer from 1989 until 1995. Born and brought
up in Accrington, Grace was known to her head-
master as the cleverest girl he had had in his school.
After taking her Higher School Certificate, she took
an office job and then came down to Wiltshire to
work for the Avon Rubber Company. By then, it
seems she had been a champion swimmer for
Lancashire, and considered for the English Olympic
team. During her time with Avon Rubber she became
a Chartered Secretary, taking her examinations by
correspondence course, and travelled abroad for the
firm.

Already a member of the Society and having
retired from Avon Rubber, Grace became our
Honorary Treasurer in 1989. The successful
running of a Society like ours depends upon the
expertise, commitment and sheer love of the
institution of its voluntary officers and helpers.
Grace Fairhurst was one of those extraordinarily
generous givers of their time and expertise. The time
she gave was really impressive. Hardly a day went by
when she was not at her desk, and while seemingly
oblivious of the discussions and chatter and
movement all around her, playing her own highly
individualistic part in it. And once back home, more
often than not she continued with some aspect of her
duties. The expertise, which was crucial to our
recovery from difficult times, can be seen at a glance
in any of her sets of accounts. It can also be
demonstrated through her briefing and amiable
bullying of our fund managers, Cazenoves, who have
given us such a reliable investment portfolio.

When word processors and micro-computers
reached the world of Long Street, Grace quickly
acquired her own equipment, seeing the advantages
it could bring to her work for the Society. Her last
accounts were visually impressive — as well as
masterful, in content — because she had learned
rapidly to exploit the presentational potential of the

162 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

new technology. To have this at her fingertips, which
she clearly did, delighted Grace. When curators
came to her to seek ways in which the Society could
find its half of the funding to re-equip the Museum
with the latest computer hardware, she cast innate
suspicion of new expenditure aside and made the
project possible. Its first two phases have been
completed now, and in many ways this equipment is
a monument to our late Treasurer’s wisdom and
guarded flexibility.

How she loved to talk! The driving voice fired
with enthusiasm for whatever the instant topic was,
and intoned as a true Lancastrian, often betrayed
great wisdom and an extraordinary breadth of
interests and general knowledge. These are reflected
clearly in her fine collection of books, now
bequeathed to us, containing volumes on travel,
history, archaeology, natural history, literature and
many other subjects. She regularly went abroad on
holiday, latterly with her friend Gladys Bland who
shared Grace’s love of travel, fine food and wine, and
who kindly supplied the details of Grace’s early life
which appear above.

I think that Grace saw Long Street and all its
people, young and old, as her family. She could be
critical. She could heave sighs of frustration, she
could even shout — as she did at me once because, so
far as I can remember, I had described the very
modest cost of something or other, unbudgeted for,
as a peanut. But she was straight, dedicated,
generous, humorous, loving to be amused, warm
hearted, and so successful in our service.

We can never forget Grace, not just because of
her reputation and the success of her reign as our
Honorary Treasurer. She has left us a most generous
legacy and we will use it to perpetuate her memory
in a way that would win her approval. A true
Northern lass, Grace loathed pomposity, so perhaps
she should be remembered by her own character-
istically pithy summary of herself: ‘I may be a Miss —
but I’ve missed nothing!’

NICHOLAS THOMAS
(Adapted from the obituary in the
Society’s Newsletter 32, August 1995)

Leslie Grinsell, OBE, MA, FSA, FMA, field archae-
ologist and museum curator, died on 28 February
1995 aged 88. He was born on 14 February 1907.
Leshe Grinsell grew up lonely and unfulfilled. It
was his discovery of prehistory among the ancient
field monuments of Sussex that changed his life,

starting him on a 69-year road of discovery and
publishing for which, in 1972, he was appointed
OBE, having received an honorary MA from the
University of Bristol a year earlier. A book of essays
in his honour, Archaeology and the Landscape, edited
by Peter Fowler, was published in the same year.

His life’s work was to record every upstanding
Neolithic and Bronze Age burial mound in southern
England, measuring, classifying and mapping them.
A huge corpus of meticulously published data is his
permanent memorial and a primary source for the
county sites and monuments records that have
become a normal adjunct of local government.

Starting in about 1926 and armed with rucksack,
timetables (he did not drive), notebook, tape and
folding rule, this self-taught archaeologist went into
the field every weekend from spring to autumn. The
admission that several months elapsed before he
discovered the tape’s retractable metal winding
handle showed what a loner he was. By 1944 he
had made more than 12,000 barrow visits and
surveyed and published the resulting barrows data
for ten counties. After retirement in 1971 he added
Somerset, Avon, most of Devon, Kent and
Herefordshire, with supplements to several earlier
surveys. The amazing fact is that apart from his three
Wiltshire years, all Grinsell’s fieldwork was done in
his spare time and, of course, at his expense.

At his death, the work extended from Norfolk to
the Tamar and the Welsh Marches. The records he
drew up were not just a matter of dimensions and
classifications (for which his early bank training was
priceless). He additionally included barrow names
and folklore, early published references, details of
finds and the physical state of the monuments. 7he
Ancient Burial Mounds of England (1936, 1953 and
1975) remains a classic work.

In the course of this pilgrimage Grinsell redis-
covered countless barrows and a Neolithic cause-
wayed camp. A visit to Stonehenge on foot, with the
writer, within hours of Atkinson’s recognition of
carvings there, soon led to his own astonishing
discovery of the series of engraved feet on a burial
slab in the barrow at Pool Farm, Mendip, which is
now a remarkable exhibit in Bristol Museum.

Leslie Valentine Grinsell was for 19 years a bank
clerk with Barclays and for 20 more a curator at
Bristol Museum. But it was his intensive study of
prehistoric burial mounds that made him pre-
eminent in British field archaeology. Having first met
LVG while a student in London, the writer took over
his prehistory class at the City Literary Institute after
Grinsell had been invited by the editor of the Victoria

OBITUARIES

County History, R.B.Pugh (soon to become WANHS
President), to leave London, turn professional
archaeologist and compile their gazetteer of Neolithic
and Bronze Age barrows in Wiltshire (VCH Wilts. I,
pt. 1 (1957)) — a work which was to become his
greatest accomplishment and a landmark in the
publication of field monuments.

When the writer became Curator at Devizes
Museum in 1952, it was Grinsell who took him, on
his first weekend at Devizes, to inspect the barrow
cemetery at Snail Down. They went by bus to
Upavon, then up the Avon’s valley side, past one of
the oldest flying schools in the world to Everleigh —
where, at the Crown Hotel, Hoare and Cunnington
had stayed in 1805 while digging those barrows —
then via Weather Hill Firs to Snail Down; and
afterwards, so characteristically of Leslie, to the Sally
Lunn at Collingbourne Ducis for cream tea.

Field archaeology was not a total preoccupation,
however. Other interests motivated his enthusiasm:
folklore, place names, numismatics, Egyptology, and
piano playing, teaching and lecturing, drawing,
sketching, photography, love of the countryside and
‘the tonic properties of the air’.

Grinsell’s fieldwork was not limited to these
islands. A wartime posting to the Air Photographic
Branch of the RAF in Cairo (he had been a pioneer
in this field) led inevitably to a Grinsellian survey
of the Pyramids (1947) and, during his last years,
his annual Christmas holiday to the larger
Mediterranean islands spawned Barrow, Pyramid
Tomb (1975, 1977; Italian edn. 1978).

Grinsell was fascinated by medieval British coins.
His studies of the mints at Bath and Bristol are major
contributions to a specialist field in which, like
Egyptology, he made himself expert.

Leslie Grinsell’s circle of friends included several
women but he eschewed personal female attach-
ments. Only when working for organisations such as
the Prehistoric Society, or in his painstaking help
given to anyone with similar interests, did he betray
that deep, selfless generosity which, with those
flashes of wit and the shy smile, made his friendship
so valued. Had he married, he might not have had
the freedom to achieve what he did and we would
know so much less about our past.

Reprinted by kind permission of The Times
Cecily Margaret (Peggy) Guido, FSA (1912-

1994) Vice-President and benefactor of the Society,
died on 8 September 1994, aged 82. She had lived in

163

Devizes since 1975, and been a most active
supporter of the Society, from her quiet assistance
with the Museum’s new galleries and acquisitions to
her lively and popular leadership of our outings to
places of interest. Her published work, some of it for
WAM, made a significant contribution to research in
the county.

Like many an archaeologist, Peggy was captivated
by the past from a young age. Given a Roman coin
when she was about 11, she was thrilled by the idea
of the ancient hands through which it could have
passed — perhaps those of a Caesar himself. She dug
with Mortimer Wheeler at Verulamium, and was
taught by the young Christopher Hawkes at
Wheeler’s field-school at Farnham, where she heard
stories from the country women who had known
General Pitt Rivers at Rushden. She had an
adventure in driving in a little car with a friend across
Europe towards Greece, quite innocent of how
rough the roads and life would be in the Balkans.
Ancient Greek helped them decipher the Cyrillic
road-signs, and drawings substituted for spoken
words: ‘Not being able to draw very convincingly, we
bought rather too many eggs!”

Once home, she resolved to become an archae-
ologist and took the diploma in western European
prehistory at University College London. Kenneth
Oakley was a fellow-student, and Robin Colling-
wood and Bernard Ashmole were among her
teachers. She dug at Whitehawk and met Stuart
Piggott. They were married in 1936, and set up
house at Rockbourne, on the downs south of
Salisbury. When C.W. Phillips called on colleagues
to help with the treasures in the famous ship-burial
at Sutton Hoo, under great pressure in the last
summer weeks before war, Peggy went to Suffolk
and did photography for the excavation.

Peggy stayed at Rockbourne during the war,
while her husband was on service in India. In the
winter of 1941-42, she excavated enormous
prehistoric barrows on Beaulieu Heath in the New
Forest, as rescue work in advance of war construc-
tion; the contractors’ workmen did not see the point
of it, so she dismissed them and faced the frozen
mass of the barrows instead with volunteer help. A
fine Bronze Age mortuary house was identified.
There were congenial friends in Rockbourne or near
by: David Cecil, L.P. Hartley the novelist, G.M.
Young the historian. Young suggested William
Cobbett’s domestic classic, Cottage Economy, as a
guide for those times of rationing; Peggy followed
that old advice, keeping geese and raising piglets.
Acting as billeting officer for the village, and getting

164 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

to know the travelling didikai, gave her a sense of
how other lives were lived.

In 1948 Stuart Piggott became Abercromby
Professor in Edinburgh, and they began dividing the
year between Rockbourne and Edinburgh. Peggy
dug in the Borders and the Western Isles. Her
excavation of a crannog remains a landmark in the
study of these puzzling Scots structures, and she
always remembered how miserably cold that dig was:
the complexities of the crannog’s stratigraphy were
explored in the shallow waters of a loch, which made
for very cold feet! She was not comfortable with the
social place of a ‘Mrs Professor’, but came to like
Edinburgh in winter, the Castle rock so black in the
darkness.

Her marriage ending, she moved to a little mill
cottage in Suffolk, with her niece whom she had
brought up from early childhood as if her daughter.
On marrying again, she moved to Sicily, and lived
in the Spanish-baroque old town of Syracuse.
Writing a guide-book to the ancient city was a
rewarding pleasure to compensate for personal
sadnesses during those years, and so was the work of
three other books commissioned by Glyn Daniel,
guide-books to the archaeology of Sicily and of
southern Italy, and the volume on Sardinia in the
Ancient People and Places series.

Returning to England, she continued to study,
working when time allowed on ancient beads. Beads
are one of those classes of small finds which are the
real stuff of archaeology — but only if the scattered
facts from many an excavation report or museum
accession are brought together with intelligence in a
structured form. Peggy joined in ordering the Beck
beads, the great reference collection that was safely
housed in the archaeology museum at Cambridge
but not given much curatorial attention there, and
wrote a fundamental account of British prehistoric
and Roman glass beads. This was published in 1978
as a Research Report of the Society of Antiquaries,
of which she had been a Fellow since 1944; a second

volume, on the glass beads of Saxon Britain,
was completed before her death.

Peggy Guido had a difficult personal life. She
grew up without a mother, and her father was
drowned on holiday when she was 8, ending all hope
of happiness or security for the rest of her childhood,
as she felt it. Perhaps a full confidence in herself
never came, and familiarity with the things of the
past instead offered a security. Introduced to
Mortimer Wheeler, she was terrified by his suavity
and assured comportment. She found no fulfilment
in either of her marriages. In a memoir, she wrote,
‘Feeling without intelligence to control it is inade-
quate; intelligence without feeling is frightening.’ Yet
when I came to know her in Devizes, first from
coming to the Society to explore Stonehenge matters
in the Library and picture collections, I enjoyed the
enthusiasm, hospitality and support Peggy showed
to me, as she did to everyone — and especially the
young — whom she discerned as attempting scholarly
work of the right quality and in the right spirit. Her
late years in Devizes were happy, I believe, but
friends saw flashes of the insecurity; and she was not
readily persuaded that her own work — whether the
old excavations or her bead corpus — was noticed and
could be of an enduring value.

For a few years an old friend, the classical
archaeologist A.W. Lawrence, came to share Peggy’s
house in Long Street when Lawrence became too
infirm to live on his own. I warmly remember an
evening at Peggy’s when I brought a group of
Cambridge students; the twilight in the handsome
long drawing-room sparkled when the freshness of
their youthful interest brought out the lively spirit in
both of them. One could say of Peggy, as Peggy said
of the young Tessa Wheeler, ‘she seemed to me... to
have the qualities I would like to have had — integrity,
charm, kindness, clarity of intelligence, and the
power to express her thoughts articulately’.

CHRISTOPHER CHIPPINDALE

Wiltshire Archaeological and Natural History Magazine, Vol. 89 (1996), pp. 165-170 165

Index

NOTE: Wiltshire places are indexed or referenced under civil parishes.

AC archaeology, 145-7, 149, 152

Achard, Joan, 135

aerial photography, 4, 9, 40, 151

Aethelred, King, 140

Aethelstan, King, 80, 89

Akeman Street, 132

Alchester, Oxon, 132

Aldbourne, 97, 130; Chase, 89; Snap, 97; Upham, 97

Allen, D.A., note on glass from Butterfield Down, 24

Allen, Michael J., notes on paleo-environmental material from
Butterfield Down, 35, 36; note on land mollusca from
Maddington, 67-8

Allington, 132-3

Allington, Hants, 133

Alnewyke, William, archdeacon of Sarum, 138

Alton, Figheldean, note on, 132-4

Alveston, Glos, 84

Amesbury, 9, 85, 86, 144, 156; abbey, 132, 133; barrow G71, 37;
Boscombe Down, 1, 4; Boscombe Down West, 34, 40;
Butterfield Down, report on excavations, 1-43, 56, 58; Chalk
Plaque Pit, 22-3, 38; church, 1; Countess Road, 144; Earl’s
Farm Down, 4, 36, 38; hundred, 133; King Barrow Ridge, 144;
New Covert, 20; Southmill Hill, 4; see also Stonehenge

Andover, Hants, 156

Andrews, W.R., 108

Ansty, preceptory, 75

antler, 36

archangel, silver-leafed, 120

Arundel, earl of, 97

Ashton Keynes, Cotswold Community School, 145

Aston, Michael, work reviewed, 154-5

Atkinson, Richard, obituary of, 161

Aubrey, earl, 132, 133; John, 93, 94, 155

Audley, James, 135

Avebury, 145, 153

Avon: see Bath; Bristol

Avon, River (north), 124, 125, 149

Avon, River (south), 56, 73, 76, 80, 81, 124, 125, 144

axes, Neolithic, 4, 22, 37

Ayscough, William, bishop of Salisbury, 158

Bacon, Peter, 75; family, 75

Badbury Rings, Dorset, 37

Badelesmere, Bartholomew de, 97, 158

Baldock, Herts, 52, 62, 63, 70

Balliol, William of, 136

Barford St Martin: Grovely, 89, 98

Barley Wood, Som, 101

Barrington, Daines, 139-40

barrows: long, 151, 152; round, 1, 4, 37, 144, 149-53

Barrows Hill, Oxon, 37

Basset, Margaret, 135; Philip, 89, 135; arms, 137; family, 97

Bath, Avon, 101

Bax, Sir Arnold, 141, 143

Baydon, note on wrist-clasp from, 130-2

Bayntun, Sir Henry, 95; family, 95, 97

Beach, William, 75

bead, Romano-British, 24

Beaufort, Henry, bishop of Lincoln, 137

Beaumont, Sir George, 102, 104

Bell, Chris, 146

Benson, Robert, 105

Beresford, Maurice, 154, 159

Berkeley, of Beverstone, arms, 137

Berkshire: see Cole’s Pits; Crookham; Denford; East Garston
Warren; East Shefford; Hampstead Marshall; Hidden-cum-
Eddington; Hungerford; Kintbury; Lambourn, River; Little
Coxwell; Newbury; Reading; Sandelford; Windsor

Berwick St James, 77; Asserton, 77; Berwick Down, 144

Berwick St John, 100; Barrow Pleck, 37; Rushmore, 98

Besylle, de, family, 97

Bidford on Avon, Warwicks, 132

Biggen, Mr., 140

Billingham, Isle of Wight, 142

Bingham, George, 100; Peregrine (sen), 100; Peregrine (jun), 101

Bishopstone: Faulstone, 97

Bliss, Sir Arthur, note on, 139-43; Kennard, 141

bluebells, 119-20

Bohun, de, family, 95, 97

Bond, James, 154

bone, 152, 157; animal, 11, 13, 15, 17, 18, 35-6, 47, 48, 52, 63-5,
68, 149, 152; human, 59-62; worked, 11, 34, 59

Bonney, Desmond, 154

boundaries, Saxon, 76, 77

Bourne, River, 128, 132, 156

Bowles, Mrs 105; Charles, 99-100, 104, 105; Henry, 101; William
Lisle, paper on, 99-105, 138

Bowood, 92, 95, 100-1, 102, 104, 105

Boyton, 100

Bradford on Avon: Greenland Mills, 145-6; Winsley Rd, 146

Bradley, Richard, work reviewed, 155-6

Bratton, 77

Braydon (Bradon), forest, 89, 90, 154

Breamore, John, 74

Bremhill, 95, 99-105; Stanley Abbey, 102, 103-4

bricks, Romano-British, 19

bridges, medieval, 85

Bristol, 140

Broad Chalke: Bury Orchard, 146

Brodie, Peter Bellinger, 106-8

Broke, Willoughby de, 97

Bromham, 92, 97; Spye Park, 95

brooches, Romano-British, 20

Brown, Graham, paper on West Chisenbury, 73-83

Browning, Robert, 74; family, 75

Brygon, William, 158

Brymmore, John, 80-1

Buckinghamshire, 107

building materials, buildings, 3, 144, 148, 151, 153; Iron Age, 68;
Romano-British, 9, 13-15, 18, 19, 34, 38-40, 47, 68, 73, 148;
medieval, 34, 80, 146, 147, 149, 150; post-medieval, 148; see also
quarries

Bulford, 156; Ranges, 150

Bulkington: Lawn Farm, 146

Burbage, 85

Burcombe Without: North Burcombe, 73; South Burcombe, 73

burials: see cemeteries; cremations; inhumations

Burnett, A., note on hoard from Butterfield Down, 19

Butterfield Down, Amesbury, excavation report, 1-43

Butterfield, William, 1

butterflies, 116

Calne, 102, 104; Beversbrook, 146; Oxford Rd, 146
Calne Without: see Bowood

Calston(e), Thomas, 134; family, 134
Calstone Wellington, seal-matrix from, 134-6
Cambridgeshire: see Willingham Fen
Camerton, Som, 105

Canham, Roy, review by, 157

Canterbury, cathedral, 137

Cantor, Leonard M., 90, 94

Caractacus, 140

carbonised grain, 35, 65-8

Carlisle, Cumbria, 32

Castle Combe, 95

castles, 94

166 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Cecil, Robert, 98

cemeteries: Romano-British, 4, 69-70; Anglo-Saxon, 152

ceramics: see building materials; clay, fired; pottery; tiles

Cerne, family, 95

chalk plaques, prehistoric, 13, 22-3, 38

Chandler, Chris, 147; John, review by, 154-5

chapel, medieval, 151

Charles I, 89, 93

Charlton (north), 95

Charlton (south): Charlton Down, 45, 81, 151

charters, Saxon, 154

chases, 89, 154

Cheltenham, Glos, 142

Cherhill: Cherhill Down, 146; white horse, 103

Cherry, John, note on Calstone seal-matrix, 134-6

Cheshire: see Leasowes, The

Chester, earldom of, 136

Chicklade, 99

Chilmark: Ladydown, 108; Ridge, 108

Chippenham, 95; forest, 89; Great Lodge Farm, 93; Lower Lodge
Farm, 93

Chippenham College Practical Archaeology Group, 147-9, 152

Chippindale, Christopher, obituary by, 163-4

Chiseldon: Burderop, 97

church archaeology, 147, 153

Churchill, Sir Winston, 141

Chute: Conholt, 97; forest, 89; Tibbs Meadow, 146-7

Cirencester, Glos, 20, 63; abbey, 95, 137

Clarendon, 97; forest, 89; Kennel Farm, 94; Park, 84, 85,
90, 94

clay, fired, Romano-British, 59

Cleal, Rosamund M.J., note on pottery from Butterfield Down, 24-
7; discussion of Butterfield Down, 37-40

Clyffe Pypard, 160

Cnut, King, 89, 140

coffin nails, Romano-British, 53

coffins, Romano-British, 50, 53, 70

coins, Romano-British, 4, 9, 10, 19-20, 38, 40, 50, 52, 69, 148

Colchester, Essex, 20, 21, 23

Cole, River, 160

Coleridge, Samuel Taylor, 104, 105

Colerne, 95

Cole’s Pits, Berks, 140

Collingbourne Kingston, 97; chase, 89; Lodge, 93

Collis, John, 157

Columbers, Matthew de, 97

Compton, family, 97

Compton Bassett, 147

Compton Bassett Archaeological Research Group, 147, 153

Compton Chamberlayne, 94, 98

Conygar Hill, Dorset, 37

corn driers, Romano-British, 15-18, 39, 40

Corney, Mark, 157; note on coins from Butterfield Down, 20;
review by, 155-6

Cornwall: see Falmouth

Cornwall, earl of, 95, 98

Corsham, 88, 95; Easton, 95, 147; Hartham, 95; Lypiatt, 93

Corsley, 97

Costello, Louisa, 104

Costen, Michael, 154

Cotswold Archaeological Trust, 147-9, 151-2

Crabbe, George, 100

Cranborne Chase, 89, 151

Crawford, O.G.S., 94

cremations: Bronze Age, 152; Romano-British, 52, 60, 62-3, 69-70

Cremona, Italy, 138

Cricklade, 148, 158; Hailstone, 94, 97; Prior Park School, 147

Crocker, Philip, 100

Crockett, Andrew, notes on finds from Butterfield Down, 22, 24,
34

Crookham, Berks, 86

cropmarks, 2, 4

crow, burial of, 13, 36, 39

Crowley, D.A., work reviewed, 156-7
Crux Easton, Hants, 95

Cumbria: see Carlisle

Cunetio (Mildenhall), 132
Cunnington, William, 100, 147
Cyfeiliog, Hawys, lady of, 135-6

Danebury, Hants, 156

Danvers, Sir John, 97

Darell, family, 97

Dauntsey, 95

Davies, John A., note on coins from Maddington, 52

Davy, Sir Humphrey, 104

deer, 88-98

deerleaps, 91-2

Denford, Berks, 86

Dervorguile, 136

Despenser, family, 95, 97

Devizes, 77, 88; Drews Pond, 147; Old Park, 90; parks, 97

Devonshire: see Exeter

Dinton, 86, 106, 108

ditches, 90, 151; prehistoric, 7, 9, 10, 155-6; Bronze Age, 4, 10, 36,
38, 145, 146, 149, 150; Iron Age, 157; Romano-British, 7, 11,
13, 15, 17, 47-8, 68, 145, 146, 153; Saxon, 77; medieval, 146;
post-medieval, 152

dogs, 50, 53, 63, 65, 69-70

Domesday Book, 73, 75, 77, 80, 83, 132-4

Donhead St Andrew, 99

Donhead St Mary: church, 147; Wincombe, 98

Dorchester, Dorset, 53, 70

Dorchester on Thames, Oxon, 37

Dorset: parks, 90; see also Badbury Rings; Conygar Hill; Cranborne
Chase; Dorchester; Edmondsham; Gillingham; Poundbury;
Shaftesbury

Downton, 86, 87, 88, 97

Druids, 100

Duggan, S.M., note on metalwork from Butterfield Down, 20-1

Duke, Edward, 103

Dumbleton, Glos, 99

Dun, River, 85

Dunstanville, family, 95

Durrington, 56, 156; Durrington Walls, 34, 40; Larkhill, 156

Dyonys, Philip, 74

Eagles, Bruce, 154

earthworks, medieval, 78-80, 146, 147

East Garston Warren, Berks, 132

East Knoyle, 88, 93, 98, 99

East Shefford, Berks, 132

Edgeworth, Maria, 101, 103, 104

Edington, 77, 151

Edmondsham, Dorset, 100

Edmund Ironside, 140

Edward I, 85, 86

Edward II, 85

Edward III, 85

Edward, Black Prince, 137

Egerton, J., note on animal bones from Butterfield Down, 35-6

Eire: see Lough Gur

Elgar, Sir Edward, 141

Elizabeth I, 89, 99

enclosures, 155, 160; prehistoric, 151; Bronze Age, 150; Saxon, 89;
medieval or post-medieval, 153

Enford, 75, 76, 77, 80, 134; Chisenbury de la Folly, 75; Chisenbury
Down, 77, 81; Chisenbury Warren, 38; Compton, 76; Compton
Down, 45, 77; Coombe, 75, 97; East Chisenbury, 73, 75, 77, 79,
80, 81, 157; Lavington Way Down, 77, 81; Saucers Meadow, 77;
Slay Down, 77; West Chisenbury, paper on, 73-83

Enford, ?Hants, 76

English Heritage, 144-5, 147-8

Entwistle, Roy, work reviewed, 155-6

INDEX

Erlestoke, 97

Ermin Street, 148

Essex: see Colchester

Esso Petroleum Co., 44

Evelyn, Sir John, 159

Everleigh, 97; Beach’s Barn, 147-8
Exeter, Devon, 20, 87

Fairbrother, M., note on ironwork from Butterfield Down, 22

Fairhurst, Grace, obituary of, 161-2

Falmouth, Cornwall, 22

Farleigh Hungerford, Som, 94, 97, 158

farmsteads: Romano-British, 147; model, 151

Felmingham, Norfolk, 21

Fenton, Richard, 100, 105

field systems, 152, 156; prehistoric, 151; Bronze Age, 150;
Romano-British, 151

Fielding, Charles, 104; family, 104

fieldwalking, 4, 9

Figheldean, 32, 34, 38, 56, 58, 70, 151; Ablington, 133; Alton, note
on, 132-4; Knighton, 80; Sheer Barrow, 151

figurines, Romano-British, 21, 40

fishponds, 80, 90

Fitzpatrick, A.P., report on Butterfield Down excavations, 1-43;
work reviewed, 157

Fitzwalter, arms, 135

Fitzwarin, family, 97

flintwork, 3, 48, 52, 53, 151, 152; prehistoric, 10, 69, 144, 146,
147, 149; Mesolithic, 152; Neolithic, 10; Beaker, 22, 25; Bronze
Age, 9, 10; Romano-British, 47; see also axes

Flower, John, 75

Folie: see Foyle

Fonthill Gifford: Fonthill, 98, 102

forests, 89, 154; law, 89; see also named forests

fossil insects, paper on, 106-15

Fox, Sir Stephen 159; family, 159

Foxley, Herefs, 101

Foyle (Folie), de la, Henry, 74, 75; Richard, 75; Roger, 75; family,
74

Fyfield: Fyfield Down, 127

gaming counter, Romano-British, 24

gardens, 81, 88

geology, 106-15

geophysics, 1, 6-7, 144-5, 147-8, 151

Germany: see Konstanz

Geve of Calstone, note on, 134-6

Gifford, family, 97

Gillingham, Dorset, 99, 140

glass, Romano-British, 24, 38, 39

Glastonbury, Som, abbots, 95

Gleeson Group ple, 1

Gloucestershire, 107; see also Alveston; Cheltenham; Cirencester;
Dumbleton; Somerford Keynes

Grafton: East Grafton, 92, 97; Suddene Park, 97;

Wilton, Batts Farm, 148

granges, monastic, 80

grasses, 119-20, 125, 126, 127

Great Barton, Suffolk, 37

Great Bedwyn: ‘brail’ names, 94; Chisbury, 97; parks, 97;
Tottenham Park, 97

Great Durnford, 157

Great Somerford, 95

Great Wishford, 157

Grentmesnil, Hugh de, 134

Grenville, Adam de, 97

Grigson, Geoffrey, 101, 103

Grinsell, Leslie, obituary of, 162-3

Grose, Donald, 125-7

ground ivy, 120

Grovely Forest, 89; see also Barford St Martin

Guido, Cecily Margaret (Peggy), obituary of, 163-4

167

Hagley, Worcs, 101

Hague, The, Netherlands, 89

Hallam, Robert, bishop of Salisbury, 136, 138

Hamilton-Dyer, S., note on animal bone from Maddington, 63-5

Hampstead Marshall, Berks, 86

Hampshire: see Allington; Andover; Crux Easton; Danebury;
?Enford; Lankhills; Oakridge; Owlesbury; Portsmouth; Quarley;
Silchester; Southampton; Winchester; Winnall Down

Harding, 133; Philip A., 53, note on flintwork from Butterfield
Down, 22

Hare, John, 154

Harland, Peter, 141

Harold II, 132, 133

Hase, Patrick, 154

Haselgrove, Colin, 157

Havering, John de, 92

Hawys, lady of Cyfeiliog, 135-6

Haydon Wick, 151

hearths: Romano-British, 13, 37, 68; medieval, 150

Heaton, Michael, report on excavations at Maddington Farm, 44-72

hedgerows, 126-7

Henig, M., note on sceptre head from Butterfield Down, 21

Henry I, 84

Henry I, 133, 135

Henry III, 84, 85, 86

Henry IV, 89

Henry VIII, 97

Herbert, family, 97

Herefordshire: see Foxley

Hertfordshire: see Baldock; St Albans; Skeleton Green

Heytesbury and Imber: Chapperton Down, 150; Heytesbury, 97,
100, 158; Imber Coney, 150

Hicke, (porter), 87

Hidden, Norman, paper on royal itineraries, 84-7

Hidden-cum-Eddington, Berks, 85

Highworth: church, 160; Sevenhampton, 159-60

Hill, J.D., 157

Hines, John, note on wrist-clasp from Baydon, 130-2

Hinton, David, 154; P., note on charred plant remains from
Maddington, 65-7

hoards, Romano-British, 1, 2, 19, 20

Hoare, Henry Hugh, 139; Henry Merrick, 139; Sir Richard Colt,
99-100, 103, 104, 105, 139; family, 138-9

Hobbs, Steven, reviews by, 157-60

hobnails, Romano-British, 52, 70; see also nails

Hoffman, Trudy, 143

Holt, 95

Holte, John de, 95

Hooke, Della, 154

Horningsham, 97; see also Longleat

horseshoes, medieval, 153

Hoskins, W.G., 88, 94

Hungerford, Berks, 84, 85, 86, 87, 158; Charnham Street, 85

Hunnisett, Roy, 158

Hussey, family, 158

Hutchins, John, 100, 140

Idmiston: Gomeldon, 80

Ilchester, Som, 69

Imber: see Heytesbury and Imber

incense cup, Bronze Age, 10, 29, 31, 38

inhumations, 48, 148-9, 151; animal, 50, 68, 69; prehistoric, 152;
Bronze Age, 10-11, 38; Romano-British, 17-18, 39, 50-1, 59-62,
69-70; Anglo-Saxon, 76, 152; see also cemeteries; cremations

insects, Mesozoic, paper on, 106-15

Isabella of Hainault, 134

Italy: see Cremona; Verona

ivory, 34

ivy, ground, 120

Jackson, J.E., 139, 140, 158
James I, 89, 92

168 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Jarzembowski, E.A., paper on fossil insects, 106-15; J.B.E., 108
John, King, 85, 89

John, the Doorkeeper, 133

Johnson, Samuel, 138-9

Keatinge, Sir Edgar, 108

Keevil, 97

Kennet, River, 84, 85, 86, 120, 124, 125, 128
Kennet and Avon Canal, 124-5

Kent: see Canterbury; Patrixbourne

Kerslake, T., 140

Keyworth, Notts, British Geological Survey, 106
kilns, Romano-British, 15-18, 40

King, Mr, 101

Kington Langley, 154

Kington St Michael, 95, 154; church, 148
Kintbury, Berks, 86

Kirby, J.L., work reviewed, 157-9

knives, Romano-British, 22, 39

Knook: Knook Down, 40

Konstanz, cathedral, 136

Lacock, 95, 104; abbey, 103; Bowden Park, 95

Lambert, A.B., 100, 103

Lambourn, River (Berks), 130

Lancaster: Duchy of, 95, 97, 134; Philippa of, 135

Lane, Josiah, 100

Langford: Hanging Langford, 73; Steeple Langford, 73

Langley Burrell, 154

Lankhills, Hants, 52, 53, 70

Lansdowne, Lord (third Marquess), 100-1, 103, 104, 105

Latton: Wharf, 148

Lawson, Andrew, 157

Lea and Cleverton: Garsdon, 88, 95

Leasowes, The, Cheshire, 101

Leicester: Honor of, 134; Robert, earl of, 134

Leland, John, 94

Leman, Thomas, 105

lEstrange, arms, 136

Lewis, Carenza, work reviewed, 154-5

lichens, 127

Lincoln, earl of, 97

Lincolnshire: see Willoughton

linears: see ditches

Lisle, family, 104

Little Coxwell, Berks, 139-40

Lodsworth, West Sussex, 58

London, 101, 142; Covent Garden, 142-3; East, 52; Garrick Club,
143; Hooper St, 63; Natural History Museum, 106

Long, family, 95

Longespée, Ela, 135; Emmeline, 135; family, 95, 136

Longleat, 97, 148; Stalls Farm, 94

Lough Gur, Eire, 27

Lovell, family, 97

Ludgershall, 84, 85, 97, 156; Butt Street, 148-9

Luttrell, Sir Geoffrey, 136; psalter, 136

Lydiard Tregoze, 97

Lyneham: Bradenstoke, 95, 155; Clack, 95

McKinley, Jacqueline I., report on excavations at Maddington
Farm, 44-72

Maddington Farm, report of excavations, 44-72

Maiden Bradley, priory, 75

Malmesbury, 88, 95; abbots, 95; Cole Park, 94; market place, 149

Malmesbury St Paul: Corston, 95; Rodbourne Rail Farm, 94

Mantell, Gideon, 106, 107

maps, 93

Market Lavington, 77

Marks, Richard, 136

Marlborough, 84-6, 128, 153; castle, 90

Marshman, Michael, review by, 159

Marston St Lawrence, Northants, 130

Matcham, George, 105

Mayhew, Robert J., note on Stourhead, 138-9

meadows, water, 77, 79-80, 81, 82

Melbourne, Lord, 105

Melchet, forest, 89

Melksham, 95; forest, 89; Hurn site, 149

Melksham Without: Bowerhill, 93

Mere, 98

Merewether, John, 75

metalwork: Iron Age, 4; Romano-British, 4, 17, 20-2, 40, 52-3, 70;
medieval, 4; post-medieval, 21

middens, Romano-British, 4

Middlesex: see Twickenham

Mildenhall (Cunetio), 132

Millard, J.I., notes on finds from Butterfield Down, 22-3, 34

Miller, Sanderson, 103

mills, medieval, 75, 81; dams, 147; stones, 1393; see also quernstones

Milton Lilbourne, 137-8

Minety, 137; Upper Minety, 149

Mitchell, A.A., 108

moats, 80. 90, 154

molluscan analysis, 36, 38, 47, 50, 67-9, 149, 156

Monkton Farleigh, 95

Montfort, de, Simon, 85; family, 97, 158

Moody, Richard, 95

Moore, Thomas, 100, 103, 104

Morden, Robert, 93

More, Hannah, 101, 104

Morgan, family, 104

Morris, Elaine L., work reviewed, 157

Murdie, David, report on deposits at Maddington Farm, 47-52

music, 139-43

nails, Romano-British, 15, 22, 47, 48, 50-1, 53, 70; see also coffin
nails; hobnails

Nares, Robert, 101, 104, 105

National Rivers Authority, 127

Netheravon, 73, 75, 77, 80, 133; prebend, 74

Netherlands: see Hague, The

nettles, common stinging, paper on, 116-29

Newbury, Berks, 84, 85, 86

Nichols, J.B., 102

Nicolle, Jason St John, note on Alton in Domesday, 132-4

Nigel the Physician, 73

nitrates, 116-17, 127

Norfolk: see Felmingham

Normanton, Lord, 75

North, Christopher, 104

Northamptonshire: see Marston St Lawrence; Norton

Norton: Cowage, 154;

Norton, Northants, 132

Norton Bavant, 137

Norway, 130

Nottinghamshire: see Keyworth

Oakridge, Hants, 36

Oaksey, 93, 95

Odstock: Bodenham, 125, 126

Ogbourne St George, proctor, 75

Oliver, Jack, paper on stinging nettles, 116-29

Orcheston: Church Pits, 50; West Down, Tilshead, 149

Oswestry, Salop, 153

ovens, Romano-British, 13, 18-19, 40

Owlesbury, Hants, 65

Oxford, 84, 99

Oxford Archaeological Unit, 146, 152

Oxfordshire, 58; see also Alchester; Barrows Hill; Dorchester-on-
Thames; Oxford; Ramsden; Woodeaton; Woodstock

oyster shells, oysters, 13, 36, 148, 152

Pains Hill, Surrey, 138
palaeoentomology, 106-15

INDEX

palaeontology, 151

parks, 154; deer parks, paper on, 88-98; pales, 90-2

Parsons, Michael, works reviewed, 159

Passmore, A.D., 151

Patrixbourne, Kent, 130

Pembroke, earl of, 97, 98

Penruddocke, Hungerford, 137

Penselwood, Som, 139, 140

Percy, Agnes de, 135

Pevsner, Nikolaus, 103

Pewsey, Vale of, 154

Pewsham, 95; forest, 155

Phillipps, Sir Thomas, 104, 105

phosphates, 116-17, 127-8

pins, bone, 47, 59; Romano-British, 34, 59

Piozzi, Gabriel, 138; Hester, 138-9

pits, 152; marling, 69; prehistoric, 10, 37, 140; Neolithic, 37;
Beaker, 1, 7, 10, 37; Bronze Age, 145, 152; Iron Age, 157;
Romano-British, 11, 13, 17, 18, 47, 68-9, 145; Saxon, 153;
medieval, 146, 147, 153; see also pyre debris pits

Pitton and Farley, 159

Pitt’s Wood, Hants, 56

place-names, 93-4, 128

plant remains, charred, 65-8

plaque, chalk, prehistoric, 13, 22-3, 38

ponds: prehistoric, 146; medieval, 153

Pope, Alexander, 101

Porteus, Bishop Beilby, 101

Portsmouth, Hants, 85

Potterne, 88, 97, 157; roundel from, note, 136-8; Ryeleaze, 136

pottery, 144; prehistoric, 10, 69, 155; Neolithic, 29, 37;
Peterborough Ware, 10, 29, 37; Deverel-Rimbury, 10, 29;
Beaker, 10, 22, 24-7, 37, 38, 53-6; Bronze Age, 10, 11, 29, 145,
147, 152, 156; Iron Age, 10, 52, 53-6, 148, 152, 157; Romano-
British, 3, 4, 9, 10, 11, 13, 15, 17-19, 27-34, 38-40, 47, 48, 50,
52, 53-8, 69, 145-8, 151, 152; Saxon, 150, 151; medieval, 146,
149, 151, 152, 153; post-medieval, 149

Poundbury, Dorset, 52

Powys, arms, 136

Preshute: Elcombe Park, 90

Price, Sir Uvedale, 101-2

Priestley, J.B., 142

Puccini, Giacomo, 143

Purbeck Group, 106-15

Purton, 33

pyre debris pits, 50, 52, 62-3, 69, 70

Quarley, Hants, 156

quarries, quarrying, 48, 50, 68, 79, 140, 146, 151; see also building
materials

quernstones, 9, 23, 39, 47, 58, 139, 1403 see also mills

Quincy, Margaret de, 135

Radnor, earl of, 158

Ramsbury, 84, 88; bishops of, 97; Littlecote, 97; parks, 90, 97

Ramsden, Oxon, 21

Rawlings, Mick, report on Butterfield Down excavations, 1-43

Raymond, Frances, 155-6

Reading, Berks, 85, 104; University, 148, 152, 155-6

Redlynch: Loosehanger, 97

Reynolds, Sir Joshua, 102

Richard I, 84, 85, 89

ridge and furrow, 145, 146, 151, 153

Rievaulx, Yorks, abbey, 137

ring-ditches, 4, 10, 36, 38, 145, 149

rings, Romano-British, 20

ritual: Iron Age, 157; Romano-British, 21, 39-40

- Rivers, A.H.L.F. Pitt, 140, 155

roads: Roman, 70, 132, 148; Anglo-Saxon, 73, 153; medieval,
paper on, 84-7, 147; see also trackways; vegetation, roadside

Robinson, Paul, note on Potterne roundel, 136-8

Roches, John de la, 75

169

Rogers, K.H., review by, 156-7; Samuel, 101-2, 104

Rolfe, William, 75

Romilly, Sir Samuel, 104

Ross, A.J., paper on fossil insects, 106-15

roundel, heraldic, note on, 136-8

Roundway, 97

Rowde, 145

Royal Commission on the Historical Monuments of England, 45,
150-1, 157

royal itineraries, paper on, 84-7

Rushall, 75; Thornham Down, 81, 151

Russell, Lord John, 105

Russia, 142

Rutter, John, 104

St Albans, Herts, 39

St John, family, 97

Salisbury, 84-6, 138; Bishop Wordsworth School, 149; bishops of,
97, 136-8; Brown Street, 150; cathedral, 74; Downton Road,
149; Ivy Street, 150; New Street, 150; Old George Mall, 150; see
also named bishops and deans

Salisbury, Ela, countess of, 135; William Longespée, earl of, 135

Salisbury Plain, 75, 76, 125, 155-6; Roman, 40, 45, 68; Training
Area, 150-1

Sandelford, Berks, priory, 84, 85

Savernake: Brimslade Park, 93, 97; forest, 84, 89; Great Lodge
Farm, 93; Great Park, 92, 94; parks, 97

Saxonbury, East Sussex, 130

Saxton, Christopher, 90, 93, 97

sceptre head, Romano-British, 21, 39

Seager Smith, Rachael, notes on finds from Maddington Farm,
53-9

seal-matrix, note on, 134-6

Sedgwick, Adam, 107

Seend, 93, 97

Selwood Forest, 89, 139

settlements: prehistoric, 151; Bronze Age, 145, 152, 156; Iron Age,
4, 157; Romano-British, 4, 11-19, 36, 37, 38-40, 45, 68-9, 70,
144, 145, 150, 151, 154-5; Saxon, 154-5; medieval, 73-83, 145,
146, 150, 152, 154-5; post-medieval, 145

Seyntomer, Thomas, 158

Shaftesbury, Dorset, 99; abbey, 104; Barton Hill House, 102

shale, Romano-British, 24

shell, 15

Sherrington, 97

Shirley, E.P., 94

Shrewsbury, Salop, museum, 136

Shrewton, 77; Maddington Farm, report of excavations, 44-72

Shropshire: see Oswestry; Shrewsbury

Silchester, Hants, 24

Skeleton Green, Herts, 39

Skinner, John, 105; Raymond J., note on Pen Pits, 139-43

slag, Romano-British, 22, 39

Slatter, Doreen, paper on William Lisle Bowles, 99-105

Somerford Keynes, Glos, 145

Somerset: see Barley Wood; Camerton; Farleigh Hungerford;
Glastonbury; Ilchester; Penselwood; Uphill

Somerset Archaeological and Natural History Society, 140

Southampton, Hants, 138

Southey, Cuthbert, 105; Robert, 105

Southwick, 94, 97

Speed, John, 89, 90, 93

spindle whorls, 11, 34, 39

spoons, Romano-British, 21, 40

Staél, Mme de, 104

Stanton St Bernard, 154

Stanton St Quintin, 88, 95

Stewart, Dugald, 104

Stonehenge, 1, 138, 144-5; Avenue, 144; see also Amesbury

stonework, Romano-British, 58-9

Stour, River (Suffolk), 130

170 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE

Stourton, 97; church, 103; Pen Common, 139; Pen Pits, note on,
139-43; Stourhead, 100, 102, 103, 105; note on, 138-9

Stourton, Lord, 97

Stratton St Margaret, 73

Stukeley, William, 100

stylus, Romano-British, 22, 39

Suffolk: see Great Barton; Stour, River

Surrey: see Pains Hill

Sussex, 1063; see also Lodsworth; Saxonbury

Sutton Benger: Draycot Cerne, 85

Sutton Mandeville: Dashlet, 108

Swanton, Gillian, obituary by, 161

Swindon, 97, 106; Old Town, 151; Rushey Platt, 151; Toothill
Farm, 33; West; 32-3, 58; Whitehill Farm, 33

Sydenham, Simon, dean of Salisbury, 138

Talbot, William Davenport, 104; family, 104

Taylor, Christopher, 94; Isaac, 89

Teffont, 108, 154

temple, Romano-British, 38-40

Thames, River, 124

Thomas, Nicholas, obituaries by, 161-3

Thrale, Hester see Piozzi; Sophia, 139

Thynne, Sir John, 97

Tidworth, 156, 157; Dunch Hill, 152, 155-6; Sidbury, 156

tiles; Romano-British, 15, 39, 58, 59; medieval, 153; see also
building materials; clay, fired

Tilshead, 77, 157

Tisbury, 98; Lawn Farm, 93; Wardour, 92, 94, 98; see also Wardour,
Vale of

Tockenham, 155

Tollard Royal, 98

tools: bone, 36, 39; Romano-British, 22, 53

trackways, Romano-British, 11, 145; see also roads

tree cover, 119-20

Tregoze, family, 97

Tremblin, Benjamin, 105

Trowbridge, 97, 154; Court Street, 152

Twickenham, Middlesex, 101

United States, 143

Upavon, 75, 77, 85, 86

Uphill, Som, 102

Urtica dioica L., paper on, 116-29
Urtica urens, 116

Urticaceae, 116-29

Valence, William de, 97

Valletort, Viscount, 104

vegetation: herbaceous, 119-20; roadside, 120-3, 125, 127;
waterside, 123-6

verges, road, 120-3, 125

Vernditch, chase, 89

Verona, Italy, 138

villas, Romano-British, 40, 45, 147-8, 151

Wake, Hugh, 158

Wanborough, 39

Wansdyke, 154; East, 154

War Department, 75

Wardour, Vale of, 106, 108; see also Tisbury

Warminster, 97

Warneford, F.E., work reviewed, 159-60; John, 159-60; Thomas,
159-60

Warner, Richard, 105

warrens, 88, 154, 160

Warton, Joseph, 99; Thomas, 99

Warwick, countess of, 135

Warwickshire: see Bidford on Avon

water meadows, 77, 79-80, 81, 82

waterside vegetation, 128

Watts, Kenneth, paper on deer parks, 88-98

Wealden Group, 106

weeds, agricultural, 119-20, 127

well, Romano-British, 17

Wells, H.G., 141

Wessex Archaeology, 1, 144-6, 148-52

West Ashton: Rood Ashton, 97

West Chisenbury, paper on, 73-83

West Lavington, 93, 97; St Joan a Gore, 151

West Overton: East Overton, 154; Lockeridge, 117-20; Lockeridge
Dene, 127; Park Farm, 152; Piggledene, 119-20, 127

West Tisbury: West Hatch, 98; Wick Farm, 152

West Woods, 119-20, 125, 128

West, Lewis, 105; Sir Thomas, 98

Westbury, 33, 58, 97; Brook, 97

Westwood, J.O., 107

Wheeler, K.G.R., 128

Whitaker, J., 94

Whiteparish, 75; Newton, 97

Whittington, Richard, 158

Wight, Isle of, 142; see also Billingham

Willingham Fen, Cambs, 21

Willoughton, Lincs, 80

Wilsford cum Lake, 103

Wilson, Avice, 154; Douglas, 130; J.D., 90; John, 104

Wilton, 97, 138; Russell Street, 152

Wiltshire Archaeological and Natural History Society Archaeology
Field Group, 147

Wiltshire Botanical Society, 123-4, 125, 126

Wiltshire Flora Mapping Project, 123, 126

Wiltshire River Monitoring Scheme, 123-4

Winchester, Hants, 85; bishops, 97, 98; College, 99; honor of, 134;
St Swithun’s priory, 80

Winchester, Margaret, countess of, 135; Saher de Quincy, earl of,
135

Windsor, Berks, 84, 85

Wingfield: Farleigh (Wiltshire) Park, 94, 97

Winnall Down, Hants, 65

Winterbourne: Winterbourne Gunner, 152

Woodbridge, Kenneth, 139

Woodeaton, Oxon, 21

Woodford: Heale House, 138

Woodhill, family, 97

Woodstock, Oxon, 84

Woodward, H.B., 106

Wootton Bassett, 88, 90, 97; Lawn Farm, 93; Vastern, 90, 93, 94,
97

Worcestershire: see Hagley

Wordsworth, John, bishop of Salisbury, 136; William, 104, 105

working hollows, Romano-British, 15

Wotton, William, abbot of Cirencester, 137

wrist-clasp, Anglo-Saxon, note on, 130-2

Wroughton: Elcombe, 97

Wyles, Sarah F., notes on mollusca from Butterfield Down, 36

Wylye, 107, 157; valley, 81

Yatesbury, 153; church, 153
Yatton Keynell, 95
Yorkshire: see Rievaulx
Young, G.M., 154

Zeals, 98
Zeals, Geoffrey de, 98
Zouche, Alan, lord, 135; Eleanor de, 135

Na FE SN PR a PATE Ue ae OV

ern tas

poy

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The Wiltshire Archaeological and Natural History Magazine"

See other formats