4

BRITISH MUSEUM

Celtic
1

Art

Ian Stead
/

vir

'

Boston Public Library

Celti1C

Art
In Britain before

the

Roman

I.

in

larvard

onquesl

'niversirj

Pi

ambridgc, Maisachusetti

BR BR
t

The Trustees
British Museum

1985, 1996

of the

First published 1985

Second edition 1996

Library of Congress Catalog

Card Number: 96-77539

ISBN 0-674-10472-2
Designed by Carroll Associates
Typeset in Van Dijck
Printed in China by Imago
Jacket illustration The central
panel of the Battersea shield,
raised in repousse

and with red

'enamel' decoration.

Right Bronze boar figurines:

the three on the left are from

Hounslow and

32 mm^

is

the other (height

from Camerton.

NK6443
.S7
1996

Contend

Introduction 4

Metalworking

\rt

tc<

hniq

M\lrs 20

Dress and jeweller}

Weapons and armoui

(

[earth and

home 52

hariota and harness

Ritual

Further reading 94

Acknov ledgemeni
InJi

Introduction

THIS BOOK

is

concerned with the British Iron Age, the

five

hun-

dred years or so before the birth of Christ, when England, Wales

and part of Scotland were inhabited bv the Celtic-speaking

was spoken but never written, so it is
hardly surprising that their written history is brief, comprising a few references in Greek and Latin mainly by writers who knew very little about
those remote islands at the edge of the world. But three Latin writers did
visit Britain, and the earliest and most important was Julius Caesar, who
Britons. Their language, British,

organised military expeditions here in 55 and 54 bc. Before Caesar histo1

Air photograph of

the settlement

rv has

little to

say about Britain, and not a single Briton

is

known bv

at Gussage All Saints in the course

name.

of excavation. Ditches define the

settlement (c.100 x 120 m) and

where one of the most important sources is Posidonius T 35-51 bc), a

Greek ethnographer whose lost work was used in the first centurv bc bv
Diodorus Siculus, Strabo and even Caesar. But most information about
the Britons has to come from the discipline of the prehistorian. Bv the
studv of artefacts, excavation, field-work and aerial photographv masses

some of the buildings; the other

prominent features are

pits,

one of

which produced an important

collection

of metal-working

debris.

little

can be gleaned from accounts of their relatives, the Gauls,

of facts can be accumulated about certain aspects of their

life;

but

in the

Introduction

much about

absence of the rutin word

Without chronicle*

time w

.1

be constructed, and this

Modern techniques

laborious process bedei illed by uncertainties.

little help:

arbon 14 dating

stages of prehistory,

ia

import. uit deposit

.1

in

estimated by radio

for instance,

which

vital lor

is

little

pit

excavated at Gussage All Saints

carbon analysis

t<r

ti>

this period.

lu-

margin of error

is

huge, and such dates

arc- in

n.

is

.1

are ol

the earlier

<>t

one very

Dorset

be between 155 and 2 n

chances of the true date falling within these limits

I

he date

of

use

know

the Britona will never be

ale haa to

em

|s

but

lu-

only 68 per cent.

is

any case only rarely

associated with significant artefacts. Dendrochronology, the counting ol

tree-rings,
.1

is

was shaped from

only

ogy

in

more

far

.1

exact technique:

has shown, tor instance, that

tree felled in 22'

.1

em

wood

Hut well-preserved

exceptional circumstances. For the material

their typology

derived from artefacts

is

it

shield found near the edge of Lake Neuchatel in Switzerland

wooden

survives

hook chronol-

in this

and associations.

chronology tor European prehisattempts to construct

nineteenth
century
and were based on the materiwere
made
in
the
tory
als used Tor basic tools: three Ages were defined, ol Stone, Broti/e and
he

first

.1

Iron. The latest, the Iron Age, was subdivided in 1872 into two periods
he first took
named alter important assemblages recently unearthed.
its name from a huge cemetery near the salt mines at
[allstatt in I'pper
I

Austria, and the second was called

l.a

ene after

on the shores

a site

ol

Lake Neuchatel, where an impressive collection of objects had been

found when the water-levels of the .Swiss lakes were re-aligned.

names are applied because,

in a

artefacts typical of their respective periods: they are no

he two

more than type-

no suggestion that the cultures they represent origithose sites, still less that those names would have meant any-

and there

sites,

very general way, those sites produced

nated

at

thing

at all

The

is

to the peoples thus labelled by archaeologists.

l.a

ene period, which

is

the

mam

concern of

subdivided into Early, Middle and Late as long ago as

this

book, was

on the basis

lXXiS

of the typologies of brooches, swords .md scabbards, which throughout

(

t-lt

Europe developed along roughly similar

ic

lines.

At the turn of the

century two parallel classifications were established: La

France, and La

two systems
itself:

it

Tene A-D

Germany With

in

ene

I - 1 1 1

in

various sub-divisions these

is not an end in
must be attached.

operate today But relative chronology

still

provides

.1

framework

Although dendrochronology

to w Inch absolute dates

already of

is

some help

here,

it

has vet to

supersede traditional approaches which rely on dates given b) contacts

with the literate civilisations
tories, the occasional
I,

<>l

discovery

and even the odd

eltic object

absolute dates to be applied to

I

La Tene

Roman

conquest).

classical

eltic antiquities.

context enable

he resulting chronol-

ene period can be stated only in the most general of

4S<> :^n
.11
and 111 100 em to the
250-100 a

terms

Greece and Italy Greek and Latin hisGreek and Italian objects in ( eltic

ol

(l ,

the centuries Iron

Age

artefacts

must have been found and

rded wherever the ground was tilled or otherwise disturbed. M\

tin-

eighteenth century; with the industrial and agricultural revolutions, the

Celtic Art

ort iron

sword with bronze

handle and bronze scabbard, found

in the

River Witham, but

now

lost.

This illustration was published by

Horae Ferales ("7563).

length said to be 380 mm.

Franks
Full

in

pace of those disturbances and consequent

increased and coincided with a

discoveries

growing interest

One

in

history and antiquities.

of the earliest recorded Iron Age artefacts

Britain

is

bronze carnyx (trumpet)

a Celtic

known

the most complete example

when

the River

being dredged

Witham

in

in

- still

found

Lincolnshire was

was acquired bv Sir

worthy and a scholar

N?

in 1768. It

Joseph Banks, a

local

with an international reputation,

a zealous scientist to destroy

who

it

in

allowed

order to

its composition. Other antiquities

dredged trom the Witham have also been lost,

determine

including a remarkable short sword

scabbard: the hilt

is

of bronze and

in a
its

bronze

pommel

was represented (perhaps misrepresented) as a

kind of Lincoln imp [2]. In the eighteenth century some antiquities tound their way into cabinets ot curiosities, but in the nineteenth cen-

tury collectors took to the

field:

in

1815 the

Revd E.W

Stillingfleet 'joined a party,

was formed

for the

purpose of opening

which
group

of barrows' at Arras (East Yorkshire) and came

Age skeletons with some impressive

Bv the middle of the century a considerable number ot Iron Age antiquities were known, mainlv chance
finds, including some remarkable pieces dredged from the Thames and
the Witham. Many were published bv A.W. Franks in an outstanding con-

across Iron

grave-goods.

tribution to Horae Ferales (1863)

[2, 3, 5]:

Franks saw the British antiqui-

European context, and was ahead of his Continental colleagues

in recognising them as Celtic. From the end of the nineteenth century
archaeological excavations became more sophisticated and recovered
artefacts in contexts that enabled them to provide ever more information
about the past. In recent years the hobby of metal detecting has produced a huge haul of artefacts, including one or two really fine pieces [4].
For every metal object that was buried, either deliberately or by
chance, there must have been many more that were used until they were
broken, worn or obsolete and then recycled. The surviving sample is
ties in a

minute. Caesar mentioned 4,000 chariots retained by the British king

Cassivellaunus, and each of those chariots would have been drawn by two
horses, each with a horse-bit and with shared harness using five terrets
(rein-rings).

3 The

Witham

by Franks in

shield as illustrated

Horae Ferales

(7S63J). This illustration shows

clearly the outlines

of a boar

which once decorated the

Length 1.13 m.

shield.

Of

those 8,000 horse-bits and 20,000 terrets

is

there a sin-

our museums today? Probably not. Even the small sample now
available for study may be distorted, because objects that were delibergle

one

in

ately buried
cal

may

well have been specially selected and need not be typi-

of the objects of the day With pottery the problems are not so

marked: pots are

fragile

and readily broken, but once they are buried

As well as

either complete or in sherds they are well-nigh indestructible.

metal and pottery

a vast

range of organic materials such as wood, skin

Introduction

and fabrics was much used by the Celts,

as

by

all

primitive peoples. These

materials gradually deteriorate in use and only a small percentage

would

be buried; unless they were deposited in an exceptional environment

their deterioration

would then be accelerated.

In

our climate only water-

logged conditions will preserve organic materials, and the sample available for studv

Most

is

negligible.

Celtic art takes the form of abstract decoration on func-

tional objects,

which would have appealed to the Celt because of

ing or usefulness but which

is

its

mean-

also in tune with current taste. Sensitive

and appreciative modern writers have made valiant efforts to interpret its
meaning, but the imagination of modern people is an unreliable guide to
the aims, beliefs and feelings of their primitive forebears. ( )nlv the Celtic
and their patrons could explain Celtic art, and as they never set
pen to paper their knowledge died with them. This book attempts to
approach the subject on fairly solid ground, starting with techniques of
artists

metalwork (because most surviving examples are of metal then following the development of certain patterns, and finally giving examples of
decorated artefacts used bv the Britons in various walks of life.
,

small bronze booked blade

with j fine decorated handle.

The decoration

in the

is

'Witbam-Wandswortb
(see p.
is

29), and

Style'

the overall shape

comparable with motifs on the

Wandsworth round-boss
Found by a metal

(fig.

Stephen (Herts). Length

80).

detectorist at St

110 mm.

Augustus Wollaston Franks

(1826-97) joined

the British

Museum's

1851, was

staff in

Keeper of the Department of

British

and Medieval Antiquities

1866-96 and

one of the

greatest benefactors.

Museum's

Chapter One

6 The central panel of the

Battersea shield, raised in repousse

and with red 'enamel'

Diameter

290 mm.

decoration.

Metal working .techniques

HAD already been worked
BRONZE
um before the
Age began but
Iron

in Britain for
it

was

tance, particularly for decorative work.

still

Most

over a millenni-

of prime impor-

soils are

gentler to

bronze than to iron, and apart from a usuallv greenish patina much
differs little

from the day when

alloy of copper

and

tin,

it

was

lost or discarded.

ot
is

it

an

it was carefully mixed

between the two. Copper was mined in the

and judging from analyses

to obtain a precise balance

Bronze

Metalworking techniques

south-west

ol England, in Wales, Scotland and Ireland, and Cornish

was well known in the ancient world and attracted explorers from as

afield as Greece. Bui

foreign ores were also used in Britain, for

records thai bronze was imported and analyses have

shown

practice went back into the Bronze Age. Sheet bronze w.is

ingan ingot and beating

it

orated

Some

in

various ways.

into

.1

thin sheet; then

it

by

hammering from
.1

Ix-d oi

resilient

made

tins

In cast

[6],

raising the design

resting

Relatively small pieces wire- mass-pro.1

he sheet bronze to be decorated would then be placed over the former

and beaten into the recessed Bhapes to reate a number of identical pal
I

<

Some
ol

twice

ol

made

in this

upper hand two occur lour times and one

each time the impressions .wi- identical. In ai least one plan, at

the side ol the

line

the decoration on the Aylesford bucket was

the three designs on

detail of

tbe decoration

on

atrimali ba\

fbaped

in

.1

its

'pantomime horses', there is an impression of a vertical

which may well have been made bj the edge ol a former |7|. Relief

form

it

r,

iilcnrn.il impression

formaguished,

specially

.in

on tbe op:

Tbe vertical

bucket.

.is

duced by using a 'former' into which master design had been cut: both
iron and bronze formers are known but wood could also have been
used.

terns

could be cut and de<

the underside, with the object presumably

pitch

far

aesar

of the more ambitious products, such

the famous shield-bosses, wen- decorated by repousse

on

thai

tin

1.111

on

tl

be distin-

xzy

Metalworking techniques

decoration was also achieved In

lowering the background

t<>

working the surface of sheet bronze,

leave the design standing proud.

ration mi the Ratcliffe shield boss

used technique

rarel)

is

an outstanding example of

Ipposiu

tli> s

|s|.

at Rat

Surface decoration was sometimes inscribed or scratched using

fine-pointed scriber to produce

,H

Thedeco

.1

sharp

line.

his

i<><>l

was used alone

.1

colour

)/'<<

in

some designs, bui in others il was employed i<>r the preliminary mapping
More pronounced lines could have been engraved with a graver,

inn.

;;;

r,li,

nurkn.

is pushed over the surface and held in the fingers rather .is one
would hold pencil; somewhat similar effect is produced by chasing, in
which
tracer is hammered forward across the metal. It the marks arc

which

.1

.1

.1

il
ma) be possible to identify the tools thai made them
some instances it has been possible to follow the development of
such .is the stages ai which its edge
the tool in the course of the work

will preserved

and

in

w.is

chipped

.nul

subsequent!) resharpened

tools themselves are difficult to identify.

er to recognise

prised an

mm

have served as
identified

[9],

More

Unfortunately the fine

suhst.uiti.il tools arc easi-

and a collection from a grave at Whitcombe Dorset comhammer-head and file .is well as a chalk disc which could
the flywheel for a pump-drill. Iron files can sometimes be

with the aid of radiography, and examples from Fiskerton

7 of

mould

nlicone rubber

.1

of decoration on tbt b.nk

mirror. Engraved with

rraver and
r

(possibly

10 /;";

'guide

jbor,

iron

.1

mo

lor

.1

(tin

common

lightly

mouldi

.1

round

fine

.1

the

In.

h ribrr.

link ol

./

bone-bit,

01

II

Celtic Art

The end of a bronze scabbard

(Lines) and Gussage All Saints (Dorset) had

the gro6ves, showing that

from Bugtborpe (East Torks^.

specks of bronze

The

thev had been used by bronze-workers.

sheet

bronze front-plate,

Sometimes the bronze-smith

decorated with a graver or

tracer,

is

tried out

the effects of a tool on part of an object that

attached to an iron

back-plate by the binding strips

in

of

would be hidden from view.

On

the Birdlip

a chape, also cut from sheet bronxe.

mirror, for instance, there are practice tool-

The chape-end has been cast-on

marks in the area subsequently covered bv the

bronze handle. Similarly a design had been
roughlv worked on the inner face of a scabbard-plate found in the River Bann in
iXorthern Ireland. Bone would have been an

the binding strips

is

and

its

to

decoration

part of the lost-wax casting.

Width of chape-end 43 mm.

ideal

medium on which

sketch

to

designs

intended for bronze, and a collection of bone

[lakes with compass-drawn ornament from
Lough Crew (Co. Meath) seems to have been

used

in

this

wav.

Their context

because thev were found

in

is

curious,

tomb which

would then have been some 3,000 years old.

Among the Lough Crew finds was what may
well have been a pair of compasses (now lost)
and it seems reasonable to interpret the collection as workshop debris. Compasses were
undoubtedly used bv some Celtic artists and
very complex designs were constructed. Detailed
studv of the decoration on the back of the
Holcombe (Devon) mirror has shown that it could
have been formed entirely from compass arcs, and
some grooves on its surface can only have been made bv heavy scratching
with compasses. Designs could have been laid out directly on the bronze
by

first

coating

it

with

a thin layer ol

wax and using something

transparent slice of horn below the centre-point to ensure that

it

like a

did not

mark the metal surface.

Another way of producing decorative bronze-work was bv lost-wax
The objeel was first modelled in wax, and sometimes elaborated decorated at this stage. hen the mould was made bv encasing the
object in clay, heating to melt and remove the wax, and firing to harden
the clay. Bronze of a slightly different allov from sheet bronze (lead was
added to increase the fluidity), poured into the clav mould, would take
casting.

on the exact form of the modelled wax. The fired clav would have to be
broken open so the mould could never be used again. Finally the bronze
object was finished bv filing, polishing and perhaps by the addition of

more
This

detail using the tools already described tor decorating sheet bronze.

method of production must have created

huge quantity of broken

moulds, but very few have been recognised. The best collection was discovered in [Mi 209 on the settlement at Gussage All Saints (Dorset) (see
fig.

[10].

where more than 7,0(1(1 fragments of moulds had been discarded

The Gussage bronze-smith made harness and chariot fittings, and

he would have had a workshop on the site but nothing of

12

it

survived:

it

Mctalworking techniques

is

know

pit.

some

n onl) because

Among

wax. Ironical!) these fragile

(Km, whereas the hard

reduced to

he debris w as swept up and dumped in

some <>i the < >K used for modelling tinbone implements are Mill in perfect condi
tools used b) the same raftsman have been

<>i

.1

the rubbish were

irtuall)

steel

t <

unrecognisable lengths

of

corrosion products.

,i>i bronze .is used to make some complete objc<

s, but ii .is also
component <>i more complex objects. Sword scabbards were sometimes
made <>t bronze: i" scabbard plates would be cui from the sheet, one
wrapped round the edges of the other, and their lower puis secured by
(

.1

.1

sheet bronze chape

on

t<>

the frame.

witli the

in

Bui the very end

t<>

the chape w.is usuall) cast-

the same workshop.

mm

to

make

Pil

ol

<

lose co operation

ai Gussage All Saints included

moulds tor casting bronze. Bronze

209

as well as

vehicle fittings and harness |12|, and some-

times iron u.in covered with bronze, either dipped

with the rings

in

sometimes perhaps the same craftsman carried oui

produced b) forging iron

w.is cast-on

<>l

he bronze-smith must have worked

blacksmith

both trades
scale

|1 1|.

many

in

molten bronze

horse-bits, encased in sheel bronze.

or.

.is

\ir 11I

thanks

from
/

.1

lincb-pins vritb iron

and cast brotnu

cart-burial
Tori

.it

terminals,

kirk bum
1

20 mm.

13

Celtic Art

13 Some of the blacksmiths'' tools

from

the

IVdltham Abbey

tongs, anvil,

head of a

hammer and file. The

and

the

head of the sledge-hammer have been

grooved

so that they

swages.

The

could be used as

232

file is

mm

first

long

Head of an

iron fire-dog found at Baldock; the

complete fire-dog

14

is

700

mm

in Britain in

the seventh century

high.

in close

land needed to provide charcoal tor smelting.

producing

sites

in

B<

More

in

might well have been

where a well-known deposit

the nineteenth century.

is

The

difficult to establish

artefacts are imprecisely dated, but

Hallstatt

proximity to the wood-

One

Britain

\Ve\bridge 'Surrey ),

being worked

ings at Brooklands
14 Opposite page

worked

obtained from shallow opencast workings

sledge-

anvil

Iron was

widespread and plentiful, and therefore cheaper, iron ores were usually

ho.

it

is

date

of

the earliest iron-

at

Brooklands, near

of

of

iron ore

the Iron

was

still

Age work-

because the only associated

tempting to link them with

bronze bucket found only 100

Remains of ironno more than a simple

away.

smelting furnaces excavated at Brooklands are

bowl which would have been surmounted by a fired clay shaft: there was
no provision for tapping the slag which would have collected in the bottom, so that the furnace had to be dismantled to remove both bloom and

^'*.
.

'

'

Celtic Arr

-'

~0

We
-" "

'

'

WTjM

mm

^
i

,;

v >

\,hmf^

a*
--

ft

15 Three silver torques from j

board found at Snettisbam

The torque on the

185

mm

in

could not be worked

slag. Iron

990.

same way

in the

does not melt at the temperatures achieved

>ppositc page The

of the

.is

< ;..

bronze; for instance,

it

the Iron Age. Instead, the

smelted bloom was forged, which means that the iron was repeatedly

It

brought to

across externally.

red heat and

hammered

to

smelting and forging were carried out

16

in

The

centr.il boss

With.im shield (see

fig.

3J).

Age blacksmith seems

Iron

produce the finished object. Both

at

Brooklands.

to have

had most of the

skills

and

most of the tools used by village blacksmiths until recent times. His hallmark is the long-handled tongs needed to hold the red-hot iron at a comfortable distance. A small hoard of ironwork from Waltham Abbey
Essex included the tools of a blacksmith; those that could be broken
had been deliberately smashed, presumably as part of a ritual, before the
hoard was deposited in the River Lea. As well as five pairs ot tongs, the
blacksmith's tools included three small anvils, the head of a sledge-hammer,

a file

\n [13].

It

and

poker, in a collection dating from the

Two of the an\iK could be reversed

first

for use as

century W or

mandrels

over

Celtic Art

which bars and rods could be bent) and were also grooved for rounding
The rods would be shaped between an upper and a lower

metal rods.

groove, or swage: in the

Waltham Abbey

collection the anvils served as

the lower swages and the head of the sledge-hammer had been grooved
to double as an upper swage. Multi-purpose tools like these suggest that

the blacksmith was itinerant because he seems to have been anxious to

keep the number of heavv tools to a minimum. Ancient tools are bv no

means common because thev would have been highlv prized and passed
from one generation to another, and when worn out thev would have
been recvcled. Unassociated iron tools are

same forms remain

Iron

Age

The

Iron

and

in

difficult to date

because the

use tor centuries^ so most blacksmiths' tools of the

known onlv trom deliberate deposits in hoards or graves.

Age blacksmith also had hammers, set-hammers, hot-chisels,

are

slices 'a

long poker-like tool w ith a spatulate end used for controlling

the hot fuel).

One of

Age blacksmith

the finest products of the Iron

dog, and the head from one found at Baldock (Herts)

impressive piece of work

The

tall

the

is

fire-

an especially

upright has been bent outwards

top to form the basis of the head, from which the snout has been

at the

forged, the nostrils

elled.

[14].

is

punched and the mouth and prominent jaw-line chisforged separately, welded on top of the

The horns would have been

head, and their ends shaped into protruding eves. Iron could also be

engraved or chased, provided the graver or tracer was hard enough, but
surface corrosion has

Of

lett

other metals

us tew good examples.

silver, listed

bv Strabo

was

as a British export,

sometimes alloyed with gold but was rarely used as a predominant metal.
The most ancient silver artefact from Britain is a finger-ring from Park
Brow ^Sussex) belonging to a distinctive type found mainly in
Switzerland and dating from the third century BC. Otherwise until

Age were coins

recently the earliest silver objects from the British Iron

and

a few

But

in

brooches dating from the second half of the

kilos of silver

were found

and

known

as

at

Snettisham

Norfolk

a similar quantity of gold

an alloy of gold, and

rallv as
is

century

B(

1990 the picture was transformed when hvc hoards of torques

buried about 70
1

first

electrum, but

when

at

it is

they contained

15]. Silver

a significant

occurs natu-

component the

alloy

Snettisham some torques were relatively

up to 89 per cent silver.

dold was used much more frequently than silver, and had

pure, w ith alloys including

longer history: the earliest gold artefacts

ond millennium

BC.

in Britain

much

date back to the sec-

Always valuable, gold objects suffered especially

a work of art must have been consigned to the

from recycling and many

crucible to produce Britain's gold coinage. Alone

among

the metals, gold

does not corrode and comes out of the ground as bright as

last

when

seen in the Iron Age. Although thev would have worked

same way

as

it

it

was

in

the

bronze, the Britons hardly ever used gold tor brooches, and

onlv very rarely for bracelets, but gold torques feature prominently

among

British antiquities.

In

the

first

century

B(

British metalworkers experimented with

means of making base metals appear more

18

valuable. Bronze

was some-

Mctalworking techniques

times plated widi silver-copper alloy to give

and sometimes gilded by coating

and then heating

it

silver,

<>!

gold and mercur)

<>!

mercury

drive off excess

t<>

the appearance

it

with an amalgam

fitting

from the Polden llilh board.

Length 151 mm.

coloured ornaments. Precious coral from the Mediterranean

<>t

was applied

the form of knobs or snips to

in

attached by bronze puis or runs.

amounts

graves have huge

from pink to white as

Witham

Brom

kcorated with red cbampt

/.

Metalwork, especially bronze, was occasionally enhanced by the

addition

17

shield

Jeep, colour |lo].

ol coral, tin-

a result

)n tin-

thai

the earth

the Yorkshire

sec-

fig 40

rarely used alter

is

The

retain their original, very

still

onnneiit coral

variety ol objects, often

colour of which has been reduced

of centuries

knobs of coral

lias

.1

Some brooches from

Tcne

I. a

I,

continued into La Tene III: indeed coral in tin- Polden

Hills Somerset
hoard shows that ii was still employed at the time of
the Roman conquest Shell, amber and stone ornament s are also know n,

but

in Britain

it

but the most

common

alternative to coral was red

'_;lass

or 'enamel'.

The

opaque 'sealing-wax red' colour which is given by

cuprous oxide. It was used in small lumps which could be soft

glass usually has an

crystals oi

ened h\ heating and then shaped into small pellets to be attached b\

bronze pins; secured onto roughly keyed surfaces; or held by cut out
bronze frames

made

pared either

enamel
1

sec fig

some quant
is

in

it

field

rounding metal

By the

who

powder and fused

.1

|17|.

Britain

live

in

the

was

made

an oven.

in

is

famous

Ocean pour

sunken

slightly

field

is

pre-

subsequent cutting, and the

for

[these]

hud

he effect

is

to pro-

flush with that of tin- Mir

the third century

bronze and thai the) adhere-, and grow

that are

century u> champleve enamel was

of enamel whose surface

PhilostratUS recorded carb

rians

first

with this technique

the original casting or In

applied as

ll.it

<>

v:

its

enamel work,

U>: 'they

S3)

as

that the

colours on to heated

as stone,

keeping

tlu-

designs

them'

/^

Two

Chapter

Art

Styles

INSULAR

CELTIC

(or

La Tene)

art

must be studied

context, for in the early stages Britain

is

Continental tradition. But from the third century

recei\es a

new impetus,

takes

its

own

in a

original direction,

bc

and

pieces outclass the products of Continental workshops. La

first classified

ogist

who

left

by Paul Jacobsthal
Nazi Germany

in

[18], a

European

an outlying proyince of the

British art
its

Tene

masterart

was

distinguished classical archaeol-

the 1930s and settled in England.

published a detailed study of Continental Celtic art

never completed his work on the British material.

On

in

He

1944, but he

the Continent he

recognised three styles: an 'Early Style' strongly influenced by Greek art

but with some 'Oriental' and native traits; followed by the

'Waldalgesheim Style', named after a rich grave in the Rhineland; and
then two contemporary sub-styles - the 'Sword Style' (though decorat-

Paul Ferdinand Jacobsthal

(1 880- 1 95 T)

Professor of

Classical Archaeology at

Marburg

University 1912-35.

19 Opposite page Bronze

chape on a sheath from

Wandsworth, a

lost-irax casting

featuring concentric

of chape 66 mm.

20

circles.

Length

ibbards and not swords

His study ended within La

Roman

to the

ene

and the
II

Plasti<

Style'.

and did not extend

conquest. Jacobsthal's classification

is

iimJ.iinrm.il, although subsequent scholars have

villi

indicated us imperfections and suggested improve

Hunts. For the 'Earl) Style' more material is no*

know

n from Eastern

ment
the

urope, the geometric native ele-

pronounced, and the precise source of

iimrc-

is

It us

petai

'orientalising'

influence

is

unclear

main

its

lotui prr.ili mi .in

mpo\

inspiration might well have been the imagination of

the

Celtic

artist

once seen

The 'Waldalgesheim

the creation of

.is

master, (lows readily

Style',

Waldalgesheim

.1

from the Early Style of

motij

Ecury-sur-C

eastern France, though there

may be further
Many more examples of

influence from Italy

Sword Style .ire now known.

'Swiss
Sword stxlc-' h. is been distinguished from the
'Hungarian Sword style', and there has been
the

an increase

in

number of decorated

the

scab-

hards from France.

was

British Celtic art

de Navarro,

classified by |.M.

Cambridge archaeologist,

short paper that started

life

one of

as

in

a series

popular lectures before being broadcast and

o!

eventual!)

published. Jacobsthal's

three styles

were numbered and de Navarro added a Style IV to

Cover British masterpieces of the third century i;<
bsthal disapproved of this simplification, but it does
.

Apparent!)

).u

have the merit

ot

providing

tlr ough the complexities of La

(

ontincntal scheme but

even

in

Tene

art.

guide the general reader

to

classification related to the

.\

the same time distinct from

the second and

in

first

ontincntal developments.

clarity,

framework

it

is

ideal

because

the early stages the British material seems to have been

made, and
to

at

centuries

u<

it

its

Stage

I.

home-

only remotely linked

has the merit of

validity as a relative chronology,

but the use of the word 'style' creates problems:

development

is

The numbered sequence

and recent research confirms

rate to label the

it

would be more accu-

Stages rather than Styles.

in

Jacobsthal's 'Early

Style',

represented by geometric

is

Hallstatt elements and designs ultimate!) derived from (.reek art, but
there are none of the ( ontmental 'orientalism...' human masks and ani-

mal

he Hallstatt designs are extremely simple, such as strings

-s hatched triangles, lozenges and compass drawn arcs
and dots,
engraved or chased on bronze or iron artefacts dating from the fourth
figures.

Century and the vnd of the

tri<

lines

[19J.

designs
I

is

fifth

centur\

no more than

he more elaborate

he most complex of the

of circles linked In diagonal

i;<

a series

ontincntal Early Style an relies heavily

on (.reek mollis that were not slavishl) copied, but adapted,

dissected.
and rearranged in distinctive ways. In the Rhineland distuned elements
from classical

Moral

designs

St I'ul

.1

were arranged

in

distinctive

repetitive

from Cerrig-y-Drudion

21

>tlll

Celtic Art

patterns, but in eastern France similar motifs were linked to form con-

tinuous flowing designs that influenced British Celtic

on

a floral frieze

art.

design based

of palmettes flanked bv lotus flowers was especially pop-

ular [20], with elaborate variations

engraved on helmets, harness and

even on an imported Etruscan flagon, objects that have survived because

2

Part of the decoration on

the flange

of a bronze helmet

from Cerrig-y-Drudion Qee also

fig.

20y

a version of the palmette

flanked by lotus petals.

of flange

28 mm.

thev were deliberately buried,

used on pots.

Work

in graves. In

in this style is

been widespread, but only a

in a

me

of the few burials with metal grave-goods, found

stone cist at Cerrig-y-Drudion (Clwvd), had been robbed and exca-

vation in
ot

in Britain, and it may have

metalwork survives because
form of ornament was never used

little early

graves of the period are rare and this

on British pottery.

Brittanv similar motifs were

represented

924 revealed onlv broken fragments of bronze. However, some

the fragments were from a decorated flange and sufficient survived to

piece together halt ot a design that features palmettes and 'lotus petals'
[21]. It

was once thought that the flange was from

hanging-bowl, but

now been disproved and it seems more likely to have been from
a helmet. There is another version of this design on cast bronze finials
from a remarkable sword-handle found bv a metal detectorist at
Fiskerton (Lines), on the banks of the River Witham [22]. A third
palmettes, part of a scabbard whose bronze
British artefact with Stage
this has

crude palmettes flanked by large S-shapes

was acquired more than L50 years ago bv 'Philosopher' Smith of
Wisbech ( ambs and presumed to be of local origin.
The principal motif of Stage II the 'Waldalgesheim Style') is influfront -plate has a series of small
[23],

enced bv the wave tendril

in

Greek

art

and takes the form of

a string

of

triangular shapes each linked at two corners and with a tendril sprouting

from the third. Simple

as
in

friezes of this

type decorate elongated

fields,

such

sword scabbards and the bows of brooches. There are typical examples
northern Italy, where the Celts came into close contact with classical

influences after their invasion early in the fourth century

are found throughout Celtic lands from

isc

lungary to England

but others

[24].

One

ot

the finest examples, bordering the spine of a shield from Ratclifte-on-Soar

Notts

22

is

very similar to one from Moscano di Fabriano (Italy), and

is

22 Tin bandh of 4

word found

tkerton. Fitted with In-

intuit ornamented with

m
'

(a

derived from (ireek

c).

ould

mt l^o

mm

Imi^.

Celtic Art

23 Upper part of a bronze

scabbard-plate from Wisbech with

a palmette

(c/., Jigs

20d and

flanked by lotus petals or

I Lite

brJ triangles down the

sides

are in the Halls tatt tradition.

Width 48 mm.

24

f)

lyres.

'

comparable with
tin- same type
tools

ontinental work

of

is

"ii

ol

lev

the highest quality. Another frieze

the antler handle of an iron rasp found with other

and weapons on an archaeological excavation .it Fiskerton lams

where the Stage
sword-handle was found. Bui the Fiskerton
.

the- site

example, cruder) executed

There are

slightly

is

in

.1

.1

pointille,

in

muddled and

is

grave

Newnham

at

band w rapped

.11

roft

ambs

<

he

Nc nham

roft

some hatching survives and there are hints

was so treated.

his version ol

where the

the design

ontinent,

bracelet

is

very worn, but

background

that the entire

is

also

dose

to

ontinental

models, such as the scabbard from Liter, Hungary, and there

similar

rendering on

Oxon

Standlake

the

found

scabbard

The scabbard must

inferior.

heav)

.1

an angle round the bod) of the bracelet

Such angled banded ornament can be matched on the

on sword scabbards.

\er\

Waldalgesheim tendrils on

different

bronze bracelet from

ornament
ciallj

the

in

have been

but <ml\ two decorative bronze panels survive.

made ol
I

is

.1

rather

Thames at
wood or leather,

River

he panel

the foot of

.it

the scabbard, within the chape, has the Waldalgesheim Style tendril,

mouth

while the
.1

is

decorated with

repoussl pelta-like motif framed by

.1

teat urt- much

The Standlake scabchape-end and houses a La Tene

similar tendril. Both panels have hatched backgrounds,

more common on
bard, which has

.1

sword, dates from

century n

British than on

typical
the-

l.i

.1

ontinental work.

Tene

sun

end of the fourth century or the

Sometimes Waldalgesheim Style

triangles and tendrils flow

from earlier motifs without any sharp break. They

rise

and

Filottrano

scabbard

from

Italy,

on

the

lanosa hel-

torque from

the

Waldalgesheim, and on the Fiskerton sword-handle. Another example

from England, from the Thames at Brentford Greater London
is the
,

design

ol

three linked palmettes on

tively British artefact not

a east

bronze 'horn-cap',

found on the( ontinent

[251.

24

II

tendrils

on Continental and British

from the sides of

palmettes on the 'Early Style' works of eastern France, on the

met

of the third

distinc-

urn
Ji

Fabriano

I-

{-boss,

Ratcliffe-on-Soar
die

iif

.1

rasp,

<f) scabbiirj.

bracelet,
/])

Fish rton
Lit

Newnham

II

>

Crofi

scabbard, StandJah

Celtic Art

Stage III accounts for Jacobs thal's third Continental style, which he
sub-divided into two contemporary developments, a 'Plastic Style' and
Sword Style'. The 'Plastic Style', three dimensional high-relief orna-

25 Bronze %om-cap' from

Brentford. Height

62 mm.

26 Opposite Terminal

of

the

Grotesque Torque from Stuttisbam

(see fig.

50).

ment,

in

contrast

to

the

or

linear

low-relief

work of the

Early,

Waldalgesheim and Sword Styles, is still virtually unknown in Britain

though it strongly influenced one major work, the 'Grotesque Torque
from Snettisham [26]
Likewise the influence of the Continental Sword Styles

is

seen in

subsequent British developments, though close links are few. One major
piece, the Ratclifte shield-boss, was found in three pieces in the bed ot
the River Trent.
a provincial

classified as

museum

for

many

horse-armour,

it

resided in the reserves

ot

years before being acclaimed as a master-

piece of Early Celtic art in 1994

century after

its

discovery!

The

bordering the spine have already been noted

in Stage II see fig. 24b), but the ornament on the central boss is an elaborate Sword Style creation, a complex of swirling tendrils emanating
from a couple of fantastic beasts see fig. 8). Of the four Waldalgesheim

Waldalgesheim Style

26

friezes

r
TV

Celtic Art

2~ I'pper part of an iron

scabbard with dragon-fair

ornament, from Hammersmith.

of scabbard 5 J mm.

28

Art styles

Style friezes,

two terminate

British pieces with close

hundred years before

in

scabbards, found

Thames

the River

had

t<>

in

the middle

<>i

other

more than

wail even

was recognised.

their significance

in

Two

very similar fantastic beasts.

in

<nitiiH-ni.il links

work

Ih-.inis

|27|.

like

Wandsworth round-bosi

Dragon-pairs are confronted

l\ilf-f

engraved or chased towards the top of the

scabbard.

he signifk ance of the design

of course-, but

it

is

intended merely
tury

in

across

difficult
.is

is

unknot

to believe that

it

n,

was

decoration. In the third cen-

dragon-pairs appear on
eltic

Europe

.is

figu\

./

the nineteenth

century, but their dragon-pair ornament was obscured until receni

scrvation

itself

.1

hey arc swords

far east

tample from
oj
(.///

.1

belmet, proh.ihh fn

"ttbam shield,

scabbards

.is

be sn
<

(Romania

and there are some quite remarkable similar-

between examples from West an J East.

ities

Despite their Oriental appearance the

dragon-pairs seem to occur

est

in

earli-

Western

Europe, where they maj well have originated

the lyre or confronted S-motif:

in

ons are

and an
1

little

more than an

some drag-

with an eye

ear.

he RatclifTe shield-boss and the

dragon-pairs sho\t that there was

Thames

a close- link

with Continental developments around 300

but thereafter British art pursued an independent course. Stage IV belongs t<> the third
and includes masterpieces such as
century
Ik

r.<

and scabbard from the River

Witham, and the two shield-bosses from the
River
hames at Wandsworth. An important
shield

the

motii

is

the halt palnii-ttc. found already on

works

ol

the Continental 'Earl) Style', which

features prominently on both pieces from the

River Wit ham [28d, e].

he Witham scabbard
wasmadeol wood or leather that has now per
I

ished, but

decorated
blade

|2

'|.

its

the magnificent bronze panel that

mouth

is

still

corroded onto the

Beautiful!) preserved in bright shining

29 Brotrzt ornament
scabbard fron

48 mm.

"//

./

'

/.

Celtic Art

'

&

I
vt.

30

bronze,

it

ii

shaped

in

repousse with an outline thai crosses the sword

someol the Hungarian Sword Style designs, rhe overall

shape of tins panel ma) itself be distorted half-palmette, and that motif
certainly inspired some <>i tin- engraving that adorns it. Another distim
diagonal!) like

.1

tive feature
tlu-

Witham

paralleled in
ly in

the tendril thai

rosses itself,
motif thai >< urs on hot
and the Wandsworth round-boss [28a, bl. It can be
Hungarian Sword Style ornament, and features prominent-

is

the central design on the Ratclifle shield-boss

The 'Witham-Wandsworth
IV
s

.1

shield

Style'

is

only

of the elments

<>iu-

he related Yorkshire and Irish 'Scabbard Styles'

iimnis and wave tendrils,

some

greatly

elaborated and

ire
.ill

based

<>i

<>n

enhanced

with varied filler-motifs including tightly coiled spirals, triangular and

lobe

sli.i|H-s.

hatching and stippling

(3o|.

Musi of the designs

.ire

essen-

symmetrical and repetitive, bui one of the Bann scabbards has

variety ol tendrils occupying every available space.
he overall wave
tially

'

.1

is

'.

'in'

from

II

31

Celtic Art

32 Decoration at

of

the

mouth

the iron scabbard from

Fovant

1 filts). Width 45 mm.

Photograph of a

replica.

33 Opposite page Bronze mirror,

decorated on the back, from Aston

QHerts). The mirror-plate was

found by a farmer
the handle

in

1979;

was discovered

in

subsequent archaeological
excavation.

Width 194 mm.

apparent despite the complexities, and there are bordering bands that

geometric borders of Late Hallstatt sheaths.

recall the

scabbards have La Tcne

The decorated

chape-ends, derived from the Continent before

the middle of the third century

But unlike Continental scabbards,

those found in Yorkshire have central suspension loops, and that

may

have been the Irish practice too, which would suggest that the two

Scabbard

There

that

not

is

diverged from an insular rather than

one remarkable artefact decorated

the Scabbard Style-

St vies

dition.

is

scabbard.

available surface,

it

Made

Wetwang

Slack

because

would take

open

it

it,

It

was found

in

tin-opener to get into

it

is

completely

the grave of a

Hast Yorks) and the excavators called

a

Continental tra-

of sheet bronze and ornamented on every

looks like a cvlindrical box, but

sealed and has no lid [31].

to

in

it: its

it

woman

at

the 'bean-tin'

owner never needed

perhaps because there was nothing inside.

There

are hints of other regional stvles in Stage IV, such as the iron

scabbard from Fovant

Wilts

with a confronted motif surely derived

But the graceful Fovant design is far removed

from the mainly limited repertoire of Continental dragons, and its sprials and filler-motifs recall the decoration on the Bann scabbard. In eastern England another scabbard-plate, from Sutton (Notts), and a crown
from

dragon-pair

[32].

from Deal (see fig KM)

are decorated in related designs in which motifs

are dissected and rearranged.

To Stage V

He
32

did not

call

it

belongs the art style studied especially by Sir

Stage Y; indeed, he gave

it

no overall

title,

lyril

Fox.

although one

'

vies

Celtic Art

34 Diagrams showing how a

and cusp design Q> and

from a half-palmette

cj),

(rf),

aspect of

it

discovery

in

could

give rise to a trumpet void

he called the 'mirror-style'. His interest was aroused by the

lobe

derived

("*/):

1943 of a huge collection of metalwork at Llvn Cerrig Bach

(Anglesey), acquired

of Wales.

The

when he was the Director of the National Museum

two magnificent decorated bronzes, a
whose art-work was clearly related. In a series

collection included

) Saulces-Champenoises (Ardennes,

plaque and

France^); b_) Sutton (NottsJ;

of publications Fox analysed the designs and traced the principal motifs

We twang

(East

c_)

Torks^).

a shield-boss,

He was keen

other works across Britain.

in

to identify regional art styles

(schools) and, by tracing the evolution and devolution of designs, he

attempted to organise the material

a tight chronological

in

order to

which he applied tentative absolute dates.

Stage V includes tendril designs

elongated

in

fields, as well as frag-

minor panels and more ambitious designs in circular

or rectangular frames, shapes are more curvilinear than in Stage IV, fillings are confined to hatching (compare fig. 62 with fig. 30), used within
the design or as background, and often the hatching is interrupted by a

ments of

circle.

35 Openwork

sheet

bronze cover for

part of a shield-boss, from a grave

at Deal

(KenQ. Length 99 mm.

tendrils in

Tendrils terminate not

shape which, with an adjoining

with huge eve and beak

but

in a spiral,

circle, gives

in a

distinctive 'trumpet'

the impression of a bird-head

sometimes an open beak

[33].

The

ciated with these designs can be as distinctive as the pattern

Fox drew attention to one particular shape,

ed from three

lines,

voids assoitself,

and

'trumpet' void construct-

one compound curve (concave/convex) and two sim-

ple curves (one concave and the other convex). This shape, seen in

Stage V designs, occurs already

in

many

the Yorkshire Scabbard Style (on the

Kirkburn scabbard) and could have been derived from lobe and cusp
m turn evolve from half-palmettes [34]. In origin it was a

designs which

negative shape, but

it

boss decorated with a

in a

came to occupv a positive role as well. A shieldrandom arrangement of trumpet voids was tound

grave at Mill Hill, Deal (Kent) and shows that the motif developed

a life

of

its

own

as early as

<

200

B(

[35].

Engraving and chasing were not the only means of producing Stage
V patterns. Repousse was popular, and the plaque from Llyn Cerrig Bach
is a good example, with a design in a small circular panel and the repousse
executed
-

two planes

(36].

perhaps best known on the

34

The motif

is

basically a tnskele or triquetra

three-limbed device used throughout the histon of Celtic art and

Isle

of

Man

coat of arms.

On

the plaque the

'

triskele

is

buili

from repousse lobes; each limb ends in

'trumpet' and
is framed by lobes.
>n other pieces Stage V ornament
.1

and the whole

l*>ss.

.1

<

hieved by lost-wax casting, which can produce relief lobes and

hatched backgrounds. Sometimes single pun- has ornameni in different
.1

techniques, presumably

V.rks

made

in

the same workshop:

scabbard has engraved ornameni

rated cast

chape-end

<>n

the Bugthorpe

the front-plate and

.1

see fig ll
whereas
Little Wittenham
combines lost-wax casting with repousse.
With Stage V, in the second and first centuries n the compara
isolated developmenl of British art comes to an end in southern
.1

ibard

lively

nd.

In the second half of the first centur)

the neighbouring
were under Roman control, and increasing!) Britain u.is drawn
into tlu- same sphere B> Vugustan nines
decorated Roman objects were
in use and for the
time in the Iron Age there arc undoubted imports
from the ( ontinent. In the north and in Scotland
distinctive
r.<

(..mis

Am

.1

ari style

nued to flourish, bui workshops in the south of England copied

in products, and British traditions were
influenced In Roman taste

16

R :'

plaque Irmn

voidi

refou

oration on a brotrtt

Uyn

Cerr

Chapter Three

Dress and jewellery

rHE

GAULS are

while their hair

means

wash

and

tall in stature

is

to the nape

their flesh

to increase this natural quality

their hair with lime-wash

differs in

and

of

the neck

is

very moist

and

...

and draw
the hair

is

it

of

colour.

For they continually

back from the forehead to the crown

so thickened by this

treatment that

it

no way from a horse's mane. Some shave off the beard, while others

cultivate a short beard; the nobles shave the cheeks but let the moustache
freely so that it covers the

The

white,

not only naturally blond, but they also use artificial

grow

(Diodorus Siculus)

mouth.

description by Diodorus Siculus contrasts with the

modern image of

the short dark Celt and illustrates the dangers of generalisations that

ignore chronology and geography. Caesar confirms that the Britons too

'wear their hair long, and shave the whole of their bodies except the head

and upper

them

lip'.

Very few British skeletons have been studied, and most of

are from Yorkshire, but they suggest that

men were on

average 1.69

and women 1.57 m (5 ft 2 in), while life expectancy was

about thirty years, w ith only 8 per cent of the population over the age of
forty-five. Representations of Britons include the bronze head on the
(5

ft 6'A

in) tall

handle of the North Grimston (North Yorks) sword, clean-shaven and

down to the neck [37]. The three bronze heads from a burWelwyn (Herts) have their hair drawn back and sport impressive

with long hair

ial

37 Part of

the bronze handle of a

sword from North Grimston (Sorth
Torks^).

36

Height of head,

28 mm.

at

moustaches

known

in

accord with

in Hallstatt times,

the end of the

first

aesar's description [38].

Bronze razors are

but there are no La Tene razors

century

B<

when

in Britain until

large triangular 'razor-knives' were

cllcrv

The

used.

shears with which they must have cut their hair arc rarely

found before the end of the

at

Hertford Heath

Herts

century

first

and Alkham

when they occur

in

graves

Kent).

the Britons dye their bodies with pitrmm, which produces a blue

'All

more terrifying appearance in battle.

them
expanded by Herodian, describing the names <>i
observation
north Britain in the third century \i>: 'they mark their bodies with various figures of all kinds of animals and wear no clothes lor fear ol con-

colour, and
(

this

gives

.ics.ir's

.1

is

cealing these figures'.

aes.ir's

word

woad, an important source of blue dye

bv no means certain.

It

'\ it

rum'

in

more

usually translated as

is

recent tunes, hut that

is

has been argued that two ancient bodies from

landow Moss see p. 86 were painted with a copper-based pigment that

might have been ( aes.ir's 'vitrum'. But even it the landow bodies were
painted, no patterns can be distinguished now. What might have been a

common

British art-form has disappeared without trace.

irding

clothing
they

call

t<>

Diodorus Siculus, the Gauls 'wear

striking kind of

tunics dved and stained in various colours, and trousers, which

bv the name of bracae; and they wear striped cloaks, fastened

with buckles, thick

in

winter and light

in

summer, picked out with

a var-

iegated small check pattern". Very occasional]) fabric has been preserved,
either in waterlogged conditions or
o!

where the structure of small puces

cloth has been replaced bv corrosion products from adjoining metal

artefacts.

Vorks

some

Replaced fabric on an iron brooch from Burton Fleming

details

last

complex construction of stripes and diamond twill with

added by needle, making it one of the earliest attempts at

showed

embroiders known from England.

the dress described by

archaeologist can expect to find

is

Diodorus Siculus the most that the

the buckle, or brooch, which fastened

the cloak. Although then- are Hal I Stat I brooches

Britain, types that

tnmonb found in Italy, not one comes from an undoubtedly ancient

\t and they may be Comparative!) recent imports Instead, bronze
or iron pint were used

But from about 400

ene brooches are fair
La
K frequent: some were perhaps imported, though the vast majoritj
must have been manufactured locally. Such brooches were usuall)
bronze, and the pro tot v |X-s were made in one pio c.
he- del Orative bod)
i;<

Celtic Art

would be cast; then a projection from the head would be hammered and
drawn into a long wire to farm the spring and pin. The spring was coiled
first to the right of the bow and then to the left, always in the same way
so that the pin was engaged in a catch-plate on the left side of the
brooch. From the catch-plate extends a foot which turns back to the
bow both foot and bow are sometimes decorated in the original casting
-;

foot, provision is made for an applied

knob of coral or 'enamel' inlaw Iron brooches were made to the
same pattern, but were entirely forged and not cast. This La
Tene I type ot brooch was popular for a couple of centuries,

and occasionally, especially on the

and then the design was improved by lengthening the free

of the toot, which was liable to get bent and broken, and
clasping it to the bow with a separate collar, a development distinguishing the La Tene II brooch. It was then a short step, although it

end

took about

centurv to achieve

of the foot together

La Tene

III

in

it,

to cast or forge the

bow and the end

one piece, the distinctive feature of the

brooch. This classic typological sequence

is

used to

Tene chronology [39], but

La Tene brooches, and Britain in

distinguish the three stages of La

it

does not accommodate

all

particular has several peculiarities.

The manufacturers

39

typological sequence

La Tern bronze
Wood EatOi

British
J,

II.

Wetwang

III

of

brooches:

(^East TorksJ;

in

of

British La

ious hinge mechanisms. In one ot the

minated

Tene brooches

occasionally used

the Continental fashion, but they also experimented with var-

in a single

more popular forms the bow

ter-

ring which superficially resembles the coil of a spring;

the pin was manufactured separately with two linked coils to

fit on either
was secured by a rivet. Other British
pin simply pivoted between two projecting lugs. But the

side of that ring and the junction

'nprovenanced.

Lengths 47, 69 an J

springs

67 mm.

brooches had

hinge was not the only British peculiarity, for at a comparatively early
stage the foot was cast in one with the bow. This development, which dis-

tinguishes the La

40 Bronze and coral brooch

from the Queen's Barron; Arras
~East Torks^

38

Length 66 mm.

Tene

III

brooch on the Continent,

is

seen in the

much

.in J

41 Broi

nm.

earlier

Queen's Barron

Arras

at

East Yorks

where the brooch

is

other-

Lai ene shape and has elaborate ornament suggesting influence

from the La Tene
ontinenl [4(i|. Man)
'Munsingen' brooch <>n the
British La Tene II bronze brooches have the fool and bow cast in one
foot which had to he
piece, although iron brooches were still made with
vise of

.1

secured to the bow by

The

collar.

.1

distinctive 'involuted' brooch developed in Britain in La

ene

he
and may have lasted into the early years of the first century w
how s on some British brooches of La cue form were much (latter than
II

those fashionable

<>n

the

ontinent, and the) seem to have given rise to

When
down on

the involuted brooch.

lx-

natural to press

securing the pin

Wetwang Slack

at

would
bow
Many brooches were

catch-plate

in its

the centre ol the bow, and a long

could easily become dow n-curved, or involuted |41

deliberately manufactured

graves

1.

it

tl.it

and an interesting sequence ol

has shown how the long involuted

this w.iv.

East Yorks

brooch was gradually superseded b)

shorter and

more curved

variety.

and especially after ( aesar's expeditions,

British brooches again came under the influence of the Continental tradition. New forms may have been imported, perhaps including some ol
century

In the first

the silver brooches found

i;<

cemeteries

south-eastern England: they

resemble Italian silver brooches, and were used

at

tune when other

imports were certainly reaching Britain. Mut other I. a line III

brooches in Britain are sufficiently distinctive to show that there must
Italian

have bee n

worn
(

in

flourishing native industry. Brooches wen-

sometimes linked

pairs,

ontinent since the

many types were

in

tilth

century

common

use

b\
u<

all

chain,

B) the

now occasionall)

wa\ known on the

centur\

first

\i>

brooches of

o\ir southern England: tew of them

have other than the simplest decoration, but occasion. ill\ an elaborate
(

eltic

design

found.

is

bronze brooch

from

rhumberland

The most ornate

Aesica

found

in

the
small

latin

is

the surprising!)

name

for

Great

hoard of jeweller)

in

large gill

Chesters,
1894 [42].

rdsol the discover) are unsatisfactory, but the hoard seems to have

been concealed

at

the end of the third centur)

id,

although the brooch

jeweller)

Celtic Art

40

.irul

was probabl) made two hundred years earlier '< i its kind
probably the
mosi fantastically beautiful creation thai has come to us from antiquity',
enthused Sir Arthur Evans, but toJ.M. de Navarro ii was 'rather flamboyant, not n>

s.i\

jewel len

u|

simpler form of dress-fastening, used in Britain before

the introduction of the brooch and noi complete!) ousted in La
Pins

arc- a

times, though the) were quite

44<. burials ai
lias

produced

rare-.

here were only lour pins from the

Wetwang Slack, two of them

more-.

in

one grave, and nocemeten

he finest pins are quite long and have ornamental

beads, often ring-heads, and a 'swan's neck' bend in the stem
Two
I

[43],

ornam

126 mm.

41

Celtic Art

ornate pins from Yorkshire graves were found immediately adjoining the
skulls,

suggesting that theyimay have been hair-pins, but because of the

way

which the skeletons had been bundled up,

in

a dress fastening

from

the upper part of the body could easily have fallen bv the skull.
Bracelets were occasionally worn, but thev were far less

than brooches: whereas up to

accompanied by

The

common

of the Yorkshire burials were

a third

no more than 5 per cent had a bracelet [44].

were made of bronze and some had decorative setThev fitted fairly closely round the wrist so various
a brooch,

finest bracelets

tings for inlav.

devices had to be used to allow them to be pushed over the hand:

had a simple opening

the break to
a third

fit in

in

one

had

side, others

a hole at the other (a

tvpe had overlapping terminals.

a projection at

some

one end of

mortice-and-tenon fitting) and

A few

shale or jet bracelets have

been found, and some made of iron belonged to the later stages of the

Wetwang

Slack cemetery.

The

bracelets in these Yorkshire cemeteries

were always worn bv women, but according to classical writers

bracelets were worn by both men and women in Gaul. The cremation at
Snailwell (Cambs) seems to have been that of a male; the grave-goods
included a shield-boss and a razor-knife, but one of the finest objects
found there was a spirally twisted bracelet with 'snake-head' terminals
[45].

(centre

(East

fronQ and Burton Fleming

Torks^): the one

on the

This

is

the only bracelet of

are others from Scotland.

44 Bracelets from Cowlam

left

made of jet (diameter 84 mm^)

and the others are bronze

only

in

its

tvpe from England, though there

related type

is

the 'massive armlet', found

Scotland and Ireland, cast by the lost-wax process and some-

times with enamel or glass ornament

mctalwork

in the

terminals [46].

The

decora-

consistent with other pieces from northern

is

tion of the

(diameters c.60 mm^).

England and Scotland dating from the end of the first century and the
second century ad. Thev have never been found on a skeleton and

is

and jeweller)

indeed these ungainly objects could perhaps have been intended

gods rather than people.

i<>r

onceivabl) they could have been worn round

the ankle-. Anklets were certainly

worn on he
t

ontinent and one

is

sup-

posed to have been found on one of the Arras skeletons.

In the-

Wetwang sink cemetery more than 500

found, mosi of them

glass beads

were

I" different necklaces: 80 per cent of the heads

in

weTe plain and only 6 per cent contained colours other than blue. Three
Yorkshire skeletons had head necklaces, including one from

other
(

owlam

that has

one large bead decorated with

ixty-nine with white

lar

and

circles

White ornament on

scrolls.

'stratified eve bead'

Yorkshire are re resented

p

All

[47].

when found, 67

he majority

<>l

still

)t'

rich.

ring,

Found

in

1816

<

oil

the

in a

is

shallow grave under a small bar-

adorned with the necklace ol '^lass beads, an

see fig, 40
two bronze bracelets, a

oi

Uliously, toe rin^s

>ne objet

made up

he Queen's Harrow group, however,

belt fitting

from Iron Age Britain, now unfon unately

v

greenish tinge decorated with

the 100 beads said to have

bronze and coral brooch

bronze and coral pendani

are rare,

in

at

the Yorkshire burials are w ithoul grave-goods and

row, the skeleton had been

amber

a central

survive.

the rest are but poorl) equipped.

comparatively

channelled

necklace from the Queen's Barrow

the-

in

in

superimposing

the types of beads (bund so tar

Arras, including translucent beads with

white or yellow scrolls

necklace

white rings and

blue base was popu-

were created either by insetting annulets

rings or by inserting a white disc in a hollow and

blue dot

inset
a

and the onl) gold finger ring

lost

Fingei

am

ings ol

metal

seem to have been more common.

en as sen i.ued with the

cits

is

the torque:

nun Honed

4S Bronzt

brat

(L'jmhf). Diameter

105 mm.

J
Celtic Art

46 Bronze armlets with enamel

ornament, from Castle
(Aberdeenshire^)

Drummond

Newe

and (right

Castle (Perthshire^.

Diameters 141 and

147 mm.

classical writers, it is also shown on representations, and

and hoards.' The torque is a collar, or neck-ring, and its
name comes from one of the more common varieties, the hoop of which

several times

found

by

in graves

Roman,

T. Manlius, took a collar from a

Celtic warrior and earned himself the

cognomen of Torquatus. At the

is

a twisted strand of metal: a

Battle of

Telemon

the warriors in the front ranks were adorned in

'all

gold necklaces and bracelets' (Polvbius), and that was not an isolated
occurrence. But in Celtic graves torques are usually associated with

women

rather than warriors, and are

hardlv ever

of gold.

On

made of

bronze, rarely of iron, but

the Continent thev are best

known from

graves

Champagne, where thev were extremely popular until La Tene II, but
then thev became rare and they are never found with La Tene III burials.
in

In Britain torques are absent

from graves. The Yorkshire inhuma-

tions have bead necklaces instead, and the La

Tene

III

cremations

south-eastern England resemble contemporary cremations

in

in

northern

France and the Rhineland and have no torques. But their absence from
graves does not
for there

is

mean

that they were not

a rich collection of material

worn by some of the

often gold

Britons,

from other sources.

Cold torques must have been valuable always, and thus vulnerable: when

Ilrrv

they were broken,

damaged

melted down, and u

or unfashionable they

would have been

hardly

surprising that they are noi found in

graves. They found their way into the metalsmith's crucible
in recenl as
well .is ancient times: the survh ing fragments from ( llevedon
Avon are
is

the remains of a find made before 1897 and 'mostly melted

by Parson
son. Bristol 1 [48]. A most unusual burial was
said to have been found

Mildenhall

&
ai

in is 12
'a human skeleton of large dimensions,
length between the skeletons of t\w> horses ... on one
the warrior lay a long iron sword, on the other his celt: he
had a

stretched
side ot

Suffolk

ai its full

torque of gold'

but the torque was immediately melted

versmith

st

Bury

.it

down

by a

sil-

Edmunds.

he gold torques thai do survive, however, are verj

impressive.
ai Broighter
o.
(
in 1896 is a magnificent pieo

Urn

One found

work which has

man with

.i

.i

somewhat chequered

history [49].

urious assemblage of other gold obja

fcteni types, a

model boat,

bowl and two

is.

Found In a plough
two torques of Jii

fine chain necklaces,

it

was

47 Necklace

oj glati

from the gpeeii't

/

Ba

Celtic Art

46

and jewellery

bought by the British

the Royal Irish

London

L903 decided

in

exhibited

Museum

ai

but then claimed

Academy A famous

Treasure [rove by

.is

the Royal

trial ai

ourts of Justice

favour of the Irish and the collection

in

Dublin. Subsequently

is

48 Opposite

in

no*

reputable archaeologist claimed

<>iu-

had been collected and buried in the nineteenth century,

ditch!
it had been found in an <>lil umbrella in
is no*
accepted,
and
general!)
association
the
of
authenticity
But the
genuine La ene antiquil
there is n<> doubt at .ill that the torque is

th.it

the

Im.iril

and another declared thai

.1

.1

swivel-joint
two hollo* tubes whose terminals are linked by
"i
and there
degrees,
<>iuthrough
In
half
turning
opened
tli. it can be
would have been a decorative 'muff' to secure the two ends at the back.

is

made

.1

<>t

The

rich chased decoration

separately applied, and the

seems

not repousse

before the tubes were shaped.

The

to have

been executed
1

have been

high-relief 'snail shells

background

the design

t<>

h.is

been covered by

which the compass-points can still be distinguished.

hree rather similar but less ornate tubular torques, one large and

fine .iris for

1

two small, were found with the remains

Snettisham

Norfolk

in

<>i'

1948. Each had

.1

.1

fourth torque

tubular body,

in

halves like the Broighter torque, with buffer terminals and

cover
in

the ]<>mt at the back.

the course

<>t"

That

deep-ploughing

field

.it

in the-

Snettisham produced

autumns

<>f

field at

.1

made
.1

in

t\\<>

band

five

to

hoards

'4S and 1950: four ol

was about 55 m away.
(

them were within 25 m <>i one another and the fifth

Between 1964 and 1973 tour isolated torques were found in the course <>t
agricultural work, and after that it seemed very likely that the- site- had
been completely wrecked by ploughing But in 1989 Charles Hodder, ol
Kings Lynn, started to survey the field with lus metal detector, and in
Ins

km

hoard of mainly brosecond season he struck gold. He- discovered

total weight of 9.2 kg. Clearly the
gold and silver artefacts, with
.1

.1

v> Goldtorqu
Diameter 195

mm.

::

t'

Mi

V.fc

and jewellery

had noi been complete!) wrecked In ploughing. With the enthusias-

lite
tic

support of the landowner, Sir Stephci

participation of

harles

discovered

made

"it

.1

more hoards

hoard of silver lumps, and

with

huge hoard

The 'Gold

Field' ai

sent

nt

100 more.

ild, is

surc-K

the

wealth

The

oi

in

locate any evidence of

launched an

>>r

s.i\

least

.it

some 20 kg

ings ol an

iridic
it

ol

idual

was

silver

about

In

.1

activity, but

ditch.

torques that were found near

it

more than

100,

torques had been deposited.

defensive work hut

51

Til

and 15 kg

.1

tribal

treasury.

the immediate vicinity ol the hoards failed to

contemporary

\i>

and per

and must repre-

It

they did identify

It

is

its

.1

tempting

was not possible to date

the construction of the enclosure, but the ditch had been

sili

12

complete torques and

less

entire treasure,

huge 8-hectarc enclosure defined

allowed to

imme

clandestine metal detectorisi

community. Perhaps

.1

>p|X>s|t.

Subsequently Hodder

Snettisham has produced

more than the

Extensive excavations

[50J.

coins.

<>t

haps 14 hoards, including 75 more

ments

Museum

(odder, the British

diatc excavation and found five

50

and the active

ireen,

century and

abandoned and
.1

half alter the

to relate the enclosure to the

centre; the ditch was not an impressive

miuht have defined an ana that had had more

rom
IHt. Pi. 1)1,

Celtic Art

formidible defenders, such as ghosts or gods. Whatever

Snettisham

site lost

its

function, the

significance in the first centurv AD and the

its

Britons never recovered their treasure.

The
Torque

multi-strand torque from Hoard E at Snettisham, the Great

is

[51],

one

Britain's finest antiquities. Its

of

hoop

is

made

of

eight strands twisted together, and each strand in turn comprises eight

lengths of swaged wire.

low terminals

made bv

The ends of

the wires have been secured in hol-

the lost-wax process.

The

decoration on the ter-

minals, which would have been modelled in the wax,

is formed bv lowwhich define trumpet voids with matted hatching.

Details, including the small knobs with triple dots, show a close relationship with the terminal of a similar torque from nearbv Sedgetord and

relief

lobes,

some

of

more surprisinglv with

.Scotland.

comparable terminal from Cairnmuir

in

small Gaulish coin trapped within the Snettisham torque

(but not necessarilv deliberately concealed by the manufacturer, as orig-

contemporary with other coins from the site and supall the torques were buried about the same time.
A hoard of five torques was found at Ipswich (Suffolk) in 1968,
when a machine was moving earth on a new housing estate: a sixth torque
inally reported)

is

ports that notion that

52 The Ipsrrub

181

to

torques. Diameters

197 mm.

found two years later

in a

the original hoard [52].

nearbv garden may have been displaced from

They were made

of a gold alloy 'on average 80

***%
{*,

^*s^^>

<*",

and jeweller)

PCTccni gold

andfiveol them are quite similar, with twisted hoops and

loop terminals, one undecoratcd and tour with relief
designs like those
ontheGreai Torque. Bui the sixth torque is different, with ring terminals

and

bund

more complex

twist,

and

it

may be no coincidence

apart from the others. Experiments

thai it was
showed ho* the torques with

l<>p

terminals would have been constructed: from a casi ingoi

a long
d wire was formed, beni in half, and the two strands twisted
together; the terminals were then cast-on using the lost-wax
process.
wo.,1 the decorated torques were left
'as cast' from the mould, bui the
other two were worked over with a tracer which has
obscured most of the
tool-marks modelled in the wax. East Anglia is not the
only source of
gold torques, bui the onl) other marked concentration is
m
I

where they have been found on lour

rdshire,
oi

one another.

Two

of the torques

in

this

sites within 20 miles

-roup, from Glascote and

o,,d Forest, have multi-strand

hoops onto which broad loop terminals have been cast
[53J. Like most surviving gold torques in England
they seem to date from the firsi centurj n
Bui torques were still used
.

the middle of the following century, according

to the description of
licca given In D... < assius: 'in stature
she was ver) tall, in
ippeai ince mosi terrifying, in the glance ol
hereyemosi fiercei and her
ras harsh; a greal mass ol the taw
niesi hair fell to her hips; around
in

k w.is

.,

large golden necklace'.

Si Gold torque from

Forest.

Diameter

I ?

\
I

Chapter Four

Hearth and hQme

HEIR HOUSES are

large

and

circular, built

of planks and

wickerwork, the roof being a dome of heavy thatch.

(Strabo)

Strabo's description of a Gallic house might well be applied to Britain,

but such structures leave little trace for the archaeologist. Remains of
domestic architecture are restricted to plans of circular huts usually

from 5 to 9

form of

in

some up

diameter, but

a ring ot post-holes or a

tion course,

to 15

m,

in

the

rough stone founda-

and the only refinements are the occa-

porch and a trench to divert rain-water. Very

sional

occasionally finds add a touch of colour, as in the hillfort

at

Hod

Hill (Dorset),

sling-stones,

ut

for action

where some huts had collections

presumably once

belong to a chieftain had

tion.

in bags, stored

ready

bv the doorway, and one hut thought to

a

spearhead

in a

similar posi-

Perhaps the buildings were decorated inside with

fabric wall-hangings,

but these and any other

ments have long since perished and

facts are limited to the latch-lifter that

door.

Of

opened the

furniture there was probably very

Gaul 'when dining they

sit

all

fit-

significant arte-

little: in

not in chairs, but on

the earth' and 'their custom is to sleep on the

ground upon the skins of wild animals" Diodorus Siculus). Animal skins
must include those ol the brown bear, because two cremations in southern England produced terminal phalanges - the claws that would have

been
n4 Iron fire-dog from Welwyn.
Height

970/985 mm.

left in a

treated skin.

The Gauls dined next to 'hearths blazing with

and spits containing large pieces ot meat' Diodorus
artefacts

55 Iron frame I row Welwyn.

fire-dogs

Height 1.43 m.

in

come more into the picture. Iron

are known especially from graves

fire,

with cauldrons

.Siculus);

^j

south-eastern England, a ritual deposi-

tion that gives

no idea of their original dis-

tribution [541.

They were used

pairs like

their recent counterparts, to contain the logs

of the

But the Iron Age hearth was

fire.

in

the

centre of the room, not against the wall, and

the lire-dogs could be viewed from

they had

head

at

all

sides so

each end. Their excep-

tionally long necks have never been satisfactorily explained: an elaborate

(

apel

darn-ion

(Gwynedd)

example from
has

loops

at

either end of the uprights which could have

held cross-bars to support spits, but no other

fire-dog has such attachments.

frame found

52

in a

La Tenc

III

A curious iron
at Welw \ n

grave

lerts) looks

here metal

fire-dogs linked together and has

like a pair ol
[55].

Perhaps h was used as

di-d walls
tin-.

and

a brazier,

floor; certainly

a grave at Baldock

Herts

know

in

n,

while a grave

ed an iron tripod from which

tery.

should

it

Several bronze cauldrons are

two

tiers

<t

cross-struts

although h would have needed grid-

some way be connected with

one associated with fire-dogs
ai

Stanfordbury

Beds

includ-

cauldron had been suspended.

hroughoul the Hntish Iron Age iars and bow K were made ot potof them must have been made in the home, but some of the

Many

wares were produced professionally

finer

usually unambitious but

scrolls

some

ol

and traded. Decoration

and curvilinear patterns with shapes

Italian
tall

wme

is

the professional products have incised

infilled

with hatching or

Stippling l^'l Alter ( aesar's expeditions to lint am trade with

increased and fine table-ware came onto tin- British market.

a
in

dan was
I

was imported from the end oi the second century

in sherds on settlements and sometimes
l'.(

potter] amphorae, found

complete

graves |^7|. Diodorus Siculus reported thai the Gauls were

fond oi wme and sate themselves with the unmixed wine

imported b\ merchants; their desire makes them drink it greedily and
!.

when they Income drunk thc\ fall into a stupor or into a maniacal disposition'. Doubtless the Unions were allcctcd in the same way, but as time
went on thes adopted some oi the refinements of the Romans.
he new
drmk could be Utter appreciated in an Italian silver cup: several were
I

li-.irtli

.mil hm<

Celtic Art

57 The

t-xi.ir.it ion

Tene III burial at

City in 1965.

o\

rub

L.i

Welwyn Garden

A gas-pipe trench

had destroyed part of

the grave but

most of the grave-goods were

recovered

and a complete plan was

reconstructed. The grave measured

3.2 by 2.2

and contained

amphorae, two of them

five

still in

position when this photograph n.is

taken.

For glass game-pieces from

this burial see (ig.

99.

imported towards the end

vive.

ol

the

lirst

century w

handed with bronze, furnished with

a cast

halt a

means

serving

ol

it:

Avlcslord

Kent

The

prin-

who

preferred wine with the

bronze jug but also a longcould be warmed. The Roman drinking service
not

only

in which it
would not have been complete without a bucket
mixed, but it seems that this item was not traded

handled pan

dozen sur-

bronze handle, and hetter suit-

ed to native beer. Merchants provided those

in

which the wine was

to Britain. In a grave at

an imported bronze jug and pan were accompanied by

Celtic bronze-bound

wooden bucket whose

depository for the cremated bones

is

and

time was the tankard, made of wood

cipal native drinking-vessel of the

But British metalworkers did not attempt to copy them.

[58].

final

pair of

comparable vessels from

grave at Baldock (Herts) was associated with an Italian amphora, and

tempting

use had been as the

to see the type as the native equivalent of a

it

wine-mixing

in a gravel pit in 1886 and the

Arthur Evans, who was visiting the site with his
search of Stone Age implements. The bucket had not been exca-

bucket. The Avlesford grave w as discovered

finds

were shown

father in

to

vated under the best of circumstances, but the rim, uppermost bronze

band, internal bronze band, handle-mounts and handle held together

one piece, with fragments

bands. 'The arrangement

54

ol

ot

wooden

staves trapped

in

between the bronze

the lower part ot the bucket

is

not so certain,

bui on

(Ik-

strength

other Unt ish and

feet.

ol

.1

decorated fragmeni and on the analo

tin)

ontinental examples

had been dismantled and reattached

the bucket before

duced

formers

ii

u.is buried.

holes in the ba< k ot

in

the form of

in

can be equipped with three

ii

he bronze-bound iron handle pivots

bronze handle-mounts, each

Icarth .miJ

i\\ ice,

two cast

helmeted human head; the)

suggesting some antiquit)

he upper band

h.is reliel

i<>r

ornament pro

and the most interesting design iv

pair of
<>r perhaps stags, 1 reated by an .trust
not unduly worried b) details of anatomy.
hese fantasti< animals have
in

see fig 7

.1

confronted animals based on horses,

antlers, curling lips, bifurcating devil-like tails,

Marlborough
had capacit)

Wilts

.1

remation

w.is

found

in

.1

and human knees. At

t'.ir

grander vessel, which

about eight times that of the Aylesford bucket. Ii u.is

recorded and lifted by the Revd ( harles Francis about 1807:
'drawing
ol

58 Ti

.1

was made on the spot while

smallest jar or shake, and

it

it

fell

was entire', but 'it would not In-.ir the

puns'. According to the original draw-

uppi-r

band

\a

to

ing there were three decorated bands, but only fragments survive.

bandit-mount

fig

7.

II

home

Celtic Art

Fantastic animals have been seen on sword scab-

bards of the third centurv BC and there are faces

Wandsworth mask

designs of the

Witham
boar, see

shield

'which once carried

shield

but

fig. 3),

animals, like humans, are rare before the

the

long-legged

fairly naturalistic

in

and the

renderings of
first

centurv

From Felmersham (Beds) comes a pair of handlemounts in the form of cows' heads [59]. Thev were cast
bc

bv the lost-wax process and are

another: the difference

slightlv different from

one

quite deliberate in that onlv one of

is

the cows' heads has an outstretched tongue licking

its

muzzle.

From the back of the casting a stout rivet projects to attach the
mount to a wooden bucket, and on top of the head is a ring to take
the end of the handle. There are several other bovine handle fittings,
and Fox suggested that thev might have belonged to milk pails: all the
for the helmeted headv from Aylesford, Baldock and Alkham

more reason

licking

its

muzzle, a

scutcheon from a bucket,

fou n d

of the

.it

Feline rsham.

bona

is

The span

46 mm.

(see figs 74 and 75] to have presided over a much more potent brew.
There are few representations of animals other than cows and bulls, but
a fine pair of rams' heads, also bucket escutcheons, was found in what
seems to have been a disturbed grave at Harpenden (Herts) [60]. Thev

are powerfully modelled, each with hollow oval eves perhaps once inlaid

with 'enamel',

long bonv snout and large circular settings for 'enamel'

at the nostrils, the sides

head
60

of the

mouth and perhaps

as well.

ram^s head, a bronze

escutcheon from a bucket, found at

Harpenden.

!.

nm.

^^^

56

'

'

at the

back of the

61 Decoration
of

Icarth and

>

bnu

Collet tiot

::.

lassical writers speak of the vanity of the<

elts, and Strabo comments on the beauty of the women. Their houses maj
not have been
equipped with much furniture, but they had the
luxury of admiring
themselves in mirrors, [ron mirrors were in use .is
early as the third centurj ft
then bronze mirrors became fashionable, Information is
restrict
ed because the surviving sample depends on
burial practices. Mirrors are
usuallj found in graves, and their absence
from the south of England
before the end of the first centur) ft
isdueinpari totherarity of
(

bun

..Is

there before

aesar's expeditions.

Ik-

bronze mirror was

luxury

item, offering scope for decoration not only

in itscasi handle bui also on
the il.n field of the mirror plate itself
he t iewing surface was polished
plain, bui its ba< k was often covered
with elaborate
I

*w

line

uscii m,v well have been

I

side

hung on the

work; when

which case the de<

would normally have been seen with the handle ai
the top
wall, in

II

home

Celtic Art

Mirror decoration was studied and perceptively analysed by Sir

who saw the development of the decoration in terms of a

62 Opposite page The

Desborough mirror.
Length 350 mm.

Cyril Fox,

typological sequence evolved from the triskele within a circle as seen on

fig. 36). The designs were
sometimes enclosed in two or three adjoining circles; more often the circles merged into a scroll but a tripartite arrangement could be distinguished in all but the most devolved examples. Fox saw 'a familiar evolution of art forms' spanning less than a century: 'if Mayer [61a] may be

the Llyn Cerrig Bach repousse plaque (see

regarded as archaic, Colchester with

[61b]

is

classic; Birdlip [61c]

may be

its

severe and regular structure

held to correspond to the 'decorat-

phase of medieval, the 'baroque' phase of renaissance art;

Desborough [62] is clearlv to be defined as 'flamboyant', less justly perhaps as 'rococo'.' But he treated his sequence too seriously when it came
to assigning dates, suggesting limits of AD 5-20 for the Great Chesterford
mirror [611] and AD 1-15 for the one from Colchester [61b]. The few mirrors that can be dated seem to belong to the century after Caesar's expe-

ed'

ditions, but

it is

difficult to justify

anv sequence of evolution.

Detailed studv of the construction of the designs, bv a team led by

Richard Savage, has enabled the marks of various tools to be identified

Only on the Mayer mirror was the standard ot craftsmanship

it was impossible to tell whether the design had been chased
or engraved. The Mayer design was constructed with compasses, but
see

fig. 9).

so high that

free-hand scratched 'guidelines' can be distinguished, and

understand their function. .Some of the work,

to

as

it is

difficult

on the Holcombe

mirror, was meticulous and time-consuming, but other examples, such

as

)ld

Warden

[6le],

have uneven outlines and rough hatching carried

out quickly using a tool with

broken edge. From

a technical

point of

view the mirrors seem to group geographically, w ith a series of related

along the Jurassic belt from Devon via Gloucestershire to

Northamptonshire, and a second group to the south-east from
Buckinghamshire and Bedfordshire across to Essex. Detailed study of
the marks of tools is an approach which falls short of identifying the
work of individual craftsmen, but it reveals a great deal about the conStniCtion and execution ot designs.

pieces

The decorated mirror

development which seems

is

one of the highlights of Celtic

to be entirely insular.

and

art,

The Yorkshire iron mir-

rors may be compared with some from Celtic lands abroad, but the
Continent has nothing to match the decorated bronze mirrors. In some ot
the svmmetrv, as on the Desborough mirror [62], Roman influence may

be suspected, but other desp^ns

.ire far

removed from

classical taste.

Such

with an overall network of trumpet voids (see p

and Great Chesterford [61f], which Jope describes so vividly that we are
is

)ld

Warden

[61e],

obliged for ever to see

it

through his eyes: 'an unsteady lurch and

a leer-

ing face, with wicked eves running straight out into blunt-pointed ears,

and spiderv arms

space to end

in

like tentacles

wandering

One wonders how

through the available

the British craftsman would have described his design,

and which way up he would have viewed

58

cra/ily

keeled-volute derivatives that look like ghoulish suckers'.

it,

it

that mattered to him.

Hearth and home

Chapter Five

Weapons and armour

HE WHOLE

FL-iCE, which

now

is

called Gallic or Galatic,

fond of war, high-spirited and quick to

Celtic warriors were

is

madly

(Strabo)

battle...

armed with spears and swords, wore helmets and

defended themselves with shields. Reconstruction drawings are well

known, and usually such warriors wear the unique Thames helmet and
carry the one and only Battersea shield. In order to establish the warrior's

equipment, and to see how

varied from

it

man

to

man, time to time

and place to place, archaeologists relv on representations, historical

accounts and collections
all

these scores

On

is

grave-goods, but for Britain information on

ot

sadlv defective.

the Continent in the Late Hallstatt period and at the beginning

of La Tene

I it

seems that the spear was the most popular weapon, some-

times accompanied by a dagger or short

sword. The long sword was introduced in La

Tene

I,
and bv La Tene II warriors were
equipped w ith a single spear or lance, a long
sword and a shield. In Britain there are only a

few burials of armed warriors so

sible to generalise

A grave excavated

ment.

at

in

Ow

imposequip-

sleburv

trio of spear,

sword and

much disturbed grave

at Great
sword and

(Hants) included the

shield,

it is

about changes

Brackstead

had two

Essex

spe.irs,

shield, and a warrior burial at North

Grimston 'North Yorks" had two swords and
a shield. Otherwise the only associations ol
weapons is the sword and spear from
Whitcombe Dorset), while swords were

found with the remains of shields at Deal

Kent

and

mations

in four

it

Yorkshire burials. In cre-

was sometimes the practice

make do with only

to

a representative object, or

indeed part of an object: the rich Snailwell

(

ambs)

burial, for instance, included an iron

which had been deliberately

removed from the shield. No other weapon
shield-boss

was found but it may well be that the bos-,

was intended to indicate the dead man's status as a warrior.

me weapon commonly

bv the Britons but never found

used

graves

in

is

the sling: hoards of sling-stones have been

63 Iron spearhead with bronze

decoration, from tbe Rirer Thames
at London. Length

60

302 mm.

excavated

in several British hillfbrts.

Spears figure prominently

of the Celts

in battle, are

in

accounts

the only weapon-.

w capons and armour

aesar's description ol his invasion ol Britain, and are

mentioned in
Ik- remains are limited to the
sometimes depicted on Hntish coins.
iron missile-head, and it is usually impossible t" determine whether the
(

weapon had been spear or javelin to be throw n, or lance i<> be thrust.

Must spearheads were simply forged from iron, bul very occasional I)
they are decorated. An iron spearhead found in the River Thames .it
.1

London

is

.1

quite exceptional, having

chased decoration riveted

t<>

each side

.1

bronze openwork shape with

i>t

each w m^-. the four shapes and

ornameni arc .ill slightly different [63]. It is hard t< believe thai
weapon would have been thrown .it the enemy and it is more likely
have been the head of chieftain's ceremonial spear. A curious

their
tins

to

rice

.1

involving spears has been recorded from Beveral Easl

burials.

Spearheads were discovered among

not neatly
trary,

arranged

as part

the-

bones, bui they

of the warrior's equipment.

they had been used to

'kill'

Yorkshire

<

>n

were

the con-

the corpse: the spears had been

thrown into the grave, around and into the body, with souk- actually

64

/'

penetrating the bom-.

ing

Daggers and swords were doubtless more prestigious weapons, and

were certainly more complex in construction. As on the ( ontinent, daggers were used in Late Hallstatf times and
[*ene

I,

but

in

the fourth century

.1

typol
!

lefi

the very beginning of La

their place w.is taken by the long

sword, whose- arrival presumably indicates

The

ai

.1

change

in

warriors' tactics.

blades themselves were- undecorated, Inn scabbards and sheaths

centre,

Barn Elms; n.

irwortb.

295, and

Full lengtbi 341,

M 2 mm.

Celtic Art

65 Geometric decoration on
front-plate

of a

La

the

Tene I bronze

The earliest:
made bf wood, some-

ottered great scope to the artist.

dagger-sheaths were

times wrapped with strips of bronze and

sheath from Richmond.

sealed at the

bottom with

Late Hallstatt sheaths

metal.

entirely of
plates, a

a ferrule,

but most

were made

in Britain

They have two shaped

bronze front-plate (often decorated),

the edges of which are wrapped round an iron

which has

back-plate,

suspension-loop

towards the top. The tip of the two plates

is

secured at the bottom bv a chape, which gave

further scope for decoration and was subject
to quite rapid typological development. At
first it

like

was tubular, terminating

form;

then

in

an anchor-

vulnerable projecting

the

arms of the anchor were curved back and

attached to the bottom of the sheath to form
an open ring-like ending [64]. The very fine cast chape from Wandsworth
stands aside from this sequence, although its decoration and that of the
sheath

is

quite

When

keeping with the Hallstatt tradition (see

in

they are decorated Late Hallstatt and La Tene

fig.

19).

sheaths have

simple geometric motifs, such as lozenges, triangles and sometimes com-

pass-drawn arcs and

position and torm of

came from

circles

some

down

ot this

the borders of the front-plate.

ornament suggests

The

that the inspiration

stitching along the sides of leather sheaths [65].

Scabbards, used to house the long sword introduced

were often made of wood or

leather, but the best surviving

of metal. Like the earlier sheaths, they were

ends clasped by

chape

in

27"

down

iron scabbards are

the edges

the Late

two from the Thames decorated with

Hallstatt fashion and there are

fig.

I,

of two plates, their

whose top is bridged at

the back and clamped at the front [66]. Few La Tene

dragon-pairs 'see

La Tene

examples are

made

the form of a frame

decorated, but one has chased decoration

in

and one from Fovant

Wilts

see

fig.

32) with

a related design.

Sonic

I. a

Tene

and the River W'ltham

blades, such as those from Standlake (see

see

fig.

29),

The back of

with metal panels and chapes.

fig.

24

had wood or leather scabbards fitted

a

bronze chape from Little

Wittenham Oxon) sports a cut-out design including trumpet voids, but

the linear decoration is more reminiscent of Mage IV [67]. The scabbard
found
66 Opposite page,
bronze

La

Tene

in

chape and

let":

chape from

Northern Ireland. Length 104 mm.

If'irrenham, with openwork

and

chape-end. Length

62

165 mm.

grave

at

is

Kent

Deal

bronze panel

at

follows this tradition, with a bronze

the top [68].

The design executed

fillings

of Scabbard Style

trumpet voids typical

B(

ot

in

repousse

two interlocking S-shapes, formed by lobe and

essentially

cusp motifs embellished with repousse details very

bronze chape from Little

engraved decoration and a i.nt-on

on the panel
engraved

67 Opposite page, right Sheet

Stage

art.

like

some

of

the

But central to the three nodes are

Y.

The decorated Yorkshire and Irish scabbards of the third century

have front-plates made entirely of bron/e. In Yorkshire the back-plates

and chapes were

iron,

but

in Ireland

some, perhaps

all,

were of bron/e.

Three of the Yorkshire pieces were recovered complete from graves. The

>ns

and armour

^rs-?-

m>
-*-:/

\/M&ML

gf

-.

ms and armour
fourth, from Ferrybridge
imilar but fragmentary:

Wesi

Yorks

is

had been purposefully beni and broken before it was dis

carded, and onlj p.irt of n has been ro
it

ered. Such deliberate mutilation of

some<

w.is practised bj

communities,
Britain.

In-

but

is

was

the ditch of

in

left

in

.1

rel-

monument, which mighi seem very

appropriate, except thai the

Neolithic, erected

The mx decorated

monument was

some 2,000

years earlier

Irish scabbards, too,

probably deposited

in

were

the course of religious

ceremonies. Three came from

Bann,

the-

.1

river

comparable with the Witham and so

ond only to the hames in terms of Iron
I

The other

artefacts.

end

of

Lisnacrogher
years

about

.1

Antrim

o.

seventy

recovered from the

been

three were found

the nineteenth century

and

it

the
at

several

artefacts

inanity, and

were

may have

votive deposit comparable with Llyn

errig Bach and indeed La

at

bog

.1

Over

metal

in

ene itself
he overall decoration on the Yorkshire

Irish

scabbards

is in the form of waves or

one of the Lisnacrogher
designs combines the two [69]. S-motifs on

S-motifs,

and

either side of the central ridge an- paired,

alternately facing and backing; in each row

S-motifs arc- adjacent, not linked as on

Wisbech scabbard sec fig 23). The- overbalanced waves and symmetrical tendrils, but in the filling
the-

the-

all

ot

effect

is

oi

upper tendrils there

the-

no attempt at symmetry, with

and hatching all mixed together.

is

lobes, dots, concentric fillings

he La

ene

and early

blades between 550 and 650

were

lor

even shorter ones. But

marked increase
bronze were made

in

11

little

lor blades

Tene

II scabbards held weapons with

and the decorated Irish scabbards
the second and first centuries
there

in

r.<

some

line

mm

from 700 to870

is

spirals,

Ion-,

length, and

Wittenham Oxon
repousse ornament in the top

>t

l.a

mm

scabbards entirely of

Ion-.

>ne

found

in

1982

excellent condition: decorated with

panel, featuring trumpet shapes and

also has fine chased 'laddering' tor the length

<t

the scabbard

at

v., ids,

either

sideol the central rib and cast relief

ornament on the chape [70]. Related
rds have panels of engraved or chased ornament
at the top. \ quite
d. Herein
ene III type is represented In the bron/e sc.ibbard
a
found at
I

ambs

jn
1976 and
cd or inscribed ornament
(

front-plate

.s

in

probabl)
panels

Iron,

the

the top

River

md

lark

|7I|.

bottom of the
near pristine condition and the scabbard must
have been
in

.,,

.;

Tene

recorded

Yorkshire scabbard thus ritual-

istically killed

gious

;//

weapons

ontinental La
rarely

mm

Lnn.i,

;/.

plat r

J ,,

Celtic Art

when

nearly new

Lark.

It

it

was. dropped, or thrown, into the River

has a squared instead ot campanulate mouth, a round-

ed tip without a chape, and on the back

has an appendage stretching the

lull

its

suspension-loop

length of the scabbard:

others ot this type have been found in both the

Witham and

the Thames.

The Isleham sword, removed from

its

scabbard

in

the

laboratory, proved to be in very poor condition, but towards

the top ot the blade was an armourer's mark.

man

specialised as an

armourer rather than

How

tar a cratts-

as a general black-

smith is unknown, but certainlv the production ot swords

must have been a highlv skilled branch ot the trade. While
some seem to have been torged from a single piece of iron, others have

been constructed trom several different strips and

some very hard blades were achieved. Eight other

British

swords have armourers' marks, including one, found

in

the

West Row Suffolk), little more than a mile from

Isleham, which was stamped twice on one side and once on the
other [72]. The West Row blade is in excellent condition: 'its
River Lark

at

suppleness

is

itself

try?

extraordinary and

it

could be bent back upon

without breaking' wrote T.C. Lethbridge

Lethbridge,

who

in

1932 (did he

frequently brought drv archaeologv to

life, went on to speculate about the loss of the handle: 'no

doubt the whole weapon flew out of the owner's hand as he
was striking a blow and the unfortunate warrior was left gripping the hilt only. It is to be presumed that he did not long
survive this mischance'. More mundane archaeologists would
argue that the handle is likely to have perished after the sword

had been discarded, because very few survive

sword

is

still

wood, often

in its

even when the

scabbard. Handles were usuallv

made of

(pommel, grip and guard) separated by iron washers and slotted over the tang. The most
impressive is on a sword found in a decorated scabbard in a
grave at Kirkburn (Last Yorkshire) [73]. The pommel and
in three parts

guard are made

ot

horn framed with iron inset with red "enam-

and decorated front and back with domed "enamelled"

discs. The ynp, probably made ot horn too, is encased in an

el",

70 Upper part

of a bronze scabbard

from Little tl'irteiib.wu with

repousse ornament.

Photograph of a

Width 63 mm.

replica.

iron tube decorated with 'enamelled" panels.

British warriors seem to have worn little in the way ot
body armour. Tacitus comments that in their encounters with

the

Roman army

thev 'lacked the protection of breast-plates

and helmets', but archaeologv shows that iron mail was not
unknown. Indeed, Varro, writing in the first century i
implies that the Romans adopted mail from the Celts. A second
,

grave

at

Kirkburn included

complete mail tunic that had been

draped over the buried corpse. Made of thousands of small iron

71 Right Bronze scabbard from

lings, each

Isleham, with inscribed ornament.

it

Length 767 mm.

than the third century

66

butt-joint and each linked to four other rings,

in a

small group of burials unlikely to date later

with

was found

The Kirkburn mail

is

as early as

any

us .uul

from Europe, but such armour

It

is

u.is

never

common:

tour other English

have produced examples dating from the hrst centuries

sites

some of

their

<

iallic

counterparts

.ire-

said to have

and there are ethnographical parallels

being en< umbered In clot
nor. although he

rhere
again

it

1-

may

lu-s.

Some

represent

evidence thai

.1

Brii

.it

gone naked into

\i>.

all, for

battle,

warriors stripping to avoid

tor
i^li

and

ft

men won- nothing

conceivable that mosi British fighting

ai

oin-i ypes

show

.1

naked

.ir

god or mythical ancestor.

elti<

warriors protected the head, but

comes more from representations and the writings of

classical

authors than from archaeological remains. Several British coins seem to

show warriors with leather helmets, and Diodorus Siculus records th.it
tin-

Gauls won- 'bronze helmets which possess large projecting figures

lending the appearance

ol

horns form one piece with

enormous stature
tin-

helmet'.

to the wearer; in

some

cases

he bronze heads which serve as

handle-mounts on the Aylcsford bucket wear large crested helmets,

those on the Baldock buckets have what seem to be leather helmets \\ nli

drooping horns [74|,.mkI their counterparts on the Alkham buckhave curling ram-like horns [75]. Horned helmets are shown on the

flabb)
el

Gundestrup cauldron from Denmark, and on

but this tradition

is

stone- reliefs

from France,

represented In only one surviving helmei

in

the

72 Stan

.m iron

aj

from the River Lark at

imp, showing

(truck once on

and

.1

'>

<>>;

fa

photograph
II "tdth ol

tmpi
blade 41

mm.

Celtic Art

73 The handle of a sword (length

137 mmj) and the top of its
decorated scabbard, from a grave
at Kirk burn (East Torks^).

74 Opposite page Cast bronze

head, a handle-mount from a bucket

found at Baldock. Height 51

mm.

whole of Europe. Found

some time before

once regarded

asymmetrical

in

the River

Thames

near Waterloo Bridge

at

1866, this unique object with short conical horns was

as a

jester's

cap

whose

design

[76]

It

is

covered with a meandering

relief-work

Wandsworth round-boss and which

reminded

Fox

of

the

also has an affinity with the decora-

tion on the Torrs chamfrein, but the small repousse lobes and shapes

with hatched background are more reminiscent of the style of the Great

Torque from Snettisham.

from the

first

The
68

century

B<

It

seems

likelv that the

Thames helmet

dates

only other helmet from Britain

is

of

unknown provenance and

Weapons and armour

*f-

was formerly
bin
.1

wr\

in

.1

ollection [77].

It

Its

too

ii

form, a

made
'jot

bronze,

ol

kr\

ap\

Opposite

p.i

lias

conquest.

In the

absence ol bod) armour, the ( eltic warrior defended himscll

which was usuall) made ol wood <>r leather. The typical

a shield

known from

shield,

ontinent,

the- (

is

and decoration.

on the ( ontineni and the British example, whose long

which would have been worn at the back, to proteci the
lymmetrical repousse* design, probably dates from shortly after

Roman

ered

Meyrick

lu-

bears

with

history

loiii;

'peak

the-

in

different in slu|x-

<>n

is

representations and from some waterlogged finds

oval in shape, with

the from by

.1

.i

<>n

central circular or oval holi

wooden spindle-shaped boss. >n the back the hole

a strip ol woo J or iron which forms the handle,
<

crossed horizontally by

hand was accommodated in the central hole and protected by the

rhe only wooden Iron Age shield from the British Isk-s is from
covered in leather, it is rectangular with
Clonoura Co. Tipperary
so the
boss,

rounded corners,

unlike- the typical Celtic shield.

shield of the classic La

Tine

old watercourse of the River

shape from

Thames and made

but entirely of bronze, the only bronze La

The

Ihertsey

However, there
Surrey

found

is

in

.i

an

not ol organic materials

ene shield from Europe

[7S|.

hertsey shield was discovered in 1985 by the driver of a drag-line,

76 Hi
the
I

Kn

InuiiJ in
'

Celtic Art

77 Bronze helmet, provenance

who dredged

unknown. Height 165 mm.

searched tor
to have

from

it

its

waterlogged gravel

pit

and then successfully

missing handle. This unique bronze shield

been made

for display or votive

purposes than

78 Opposite page The back

and the same explanation may account

of the bronze Cbertsey shield,

shields also found in English rivers,

showing the kindle across the cavity

Europe, the YVitham shield and the Battersea shield.

under the

boss.

The

Length 836 mm.

shield

found

for the

more

likely

two complete bronze-faced

and also unmatched elsewhere

the River YVitham

in

is

for use in battle,

near YVashingborough

(Lines) about 1826 has the buss, spine and two terminal roundels covered

by

a single

piece of bronze while the rest of the face of the shield

structed from two sheets

see

fig.

3).

above the centre-line of the shield,

design springs from the spine,

seems to be based on

is

is

The boss

which

itself,

exceptionally wide.

symmetrical across

palmette motif. At the centre

is

Its

is

con-

slightly

repousse

diagonal line and

is

a roundel holding

three oval knobs of deep-coloured coral with two similar pieces, but circular, at either side (see fig.

16).

Each terminal roundel had

a central

petalled boss (only one survives) ringed hv an engraved scroll teaturing

the half-palmette, and supported bv what looks like the head of a fantastic

animal with large close-set eves, petalled ears, and an engraved pal-

mette on

snout

orated across

72

its

[79].
full

This magnificent shield has been

still

width with wh.it seems to have been

further deca

boar with

Weapons and

Celtic Art

~4

iin .irul

incredibl) spindl)

legs.

<

>nl\

the outline of the creature and the rivet-

holes which were once used to attach

\

\c-r\

similar shield

is

ii

'!<posite paj

car be distinguished

represented In

bronze boss found

in

the

Ii.iiihs.ii Wandsworth: it too may have been

entire!) faced with bronze,
but only the boss survives [81]. Much shorter than the corresponding
piece on tin- Wnli. mi shield it differs in having had separate
terminal
roundels. Ilu- one surviving end of the spine-cover expands i<>
mask
I

SI)

Ik-low

left

.1

from which

name, the Wandsworth mask boss, and tins would

have supported tin- roundel as on the Wit ham shield. Ilu- arrangement
of decoration, too, resembles that on the- Witham shield, in that it
sprmp from the spun- and is diagonal!) balanced, Inn the repousse here
is
c

in

it

takes

its

higher relief and recalls

ontinent. There

is

on!)

some of the cast

little

'Plastic Style'

works

,,

the

engraving, featuring typical Stage IV

tightly coiled spirals.

It ma) be that another form of shield is
represented by a circular
bronze boss also found in the Thames at Wandsworth [801. The
Wandsworth round boss has engraving distinctive of Stage IV, much <)" it

the form oi disjointed fragments occupying voids within a

repousse
It has
central dished hemisphere whose broad decorated flange
roughly finished at the edge and was obviously intended to be covered

design.
is

.1

by another sheet

<>r

sheets

of bronze.
he shield could have been cirdoes not necessarily imply a circular shield, and
the Wandsworth round boss is on a long shield -

cular, but a circular boss

the closest parallel tor

HI Ik-low right

mask
.it

Wandtwo\

armour

*T

m
J

A U

Weapons and armour

the most famous of

..II British Iron Age

antiquities, the Battersea shield.
Like the Wit ham shield, the Battersea shield is in faci the bronze
face and binding of
shield probably made of wood
Despite their

82

<

>pposii

.1

[82J.

markedly difTereni bosses the i\\<> shields are in some way related, for
both have circular terminal panels which are linked to the
central element b) features which remind us of animal heads. On the Witham
shield, and the Wandsworth mask boss, these heads face inwards
b
they are supporting broad roundels and narrowing through the snout
to
on the Battersea shield the position is
reversed
the broader element is at the centre and the animals face outthe spine of the shield; bui

wards. These Battersea animals have wide spreading antlers, and

are
in one piece with the terminal roundels but quite
separately from

made

the central boss.

The

three panels,

.ill

with highly accomplished, steeph

profiled repousse decoration, form the central

pan of
shield whose
with lour shaped bronze sheets, each occupying
quadrant, attached by rivets that pass through panel, sheet bronze and
then the underlying wood. The repousse" design on the central panel is
background

is

.1

filled

.1

based on an enclosed palmette which gives

either side. Strands from the other

rise to triangular

two corners of the

shapes on

triangles then meet

to form
circle.
his motif occupies one half of the panel and is almost
mirrored by the design in the other half; almost mirrored, but not
quite,
.1

because there are slight differences in some of the infillings. The end
panels earn similar but not exactly identical designs based on
interlocking S-motifs. Prominent on both end and central panels
are a series of
roundels built from east bronze frames into which

a soft

and malleable

red glass, or 'enamel', has been pressed from the underside.

All the shields mentioned hitherto are oval with
rounded ends, but
a shield ol very different shape was popular in
the second and first centuries it.
It was equally long, but in-curved
at the ends, so that it had

pointed corners and resembled the shape of a hide.

Fragments from the
bindings of the distinctive corners ,.| these shields have been

known

many

years, but their correct identification

was

mystery until

for

a collec-

tion ot bronze miniature shields appeared

on the antiquities market in

1988 [83J. Research showed that they had belonged to a huge collection

ot

Bronze Age and [ion Age antiquities found

hoard near Salisbury

Wilts

a few years previously m a

Miniatures were often made for votive

reasons, to represent the full-size originals

ai temples; the) are faithful
Copies, and in the case of the Salisbury shields
they even have tiny
handles riveted across the space behind the boss.
he face of

the finest
partitioned into eight compartments, alternate!) decorated
|

"I

them

and

is

Doubtless this engraved ornament accurate!) represents

the
decoration that would have been painted on the wooden
or leather original, lor shields 'were decorated m
individual fashion* according to
plain.

Diodoms

Within months of the appearance of the Salisbur)

1.19 in Ion-, was
found in a grave at Deal Ken.
lis organic parts had rotted,
but Us
shape was preserved by the bronze binding, ...id there
were fragments
iron, decorative bronze panels,
including a piece with openwork orna
men, that had covered par, ol the boss see fig
\5
Siculus,

shields the remains of a lull si/e hide shaped

shield,

83

Bm

sslhkn

Chapter Six
|

Chariots and harness

FOR

THEIR JOURNEYS and

in kittle they use

the chariot carrying both charioteer

and

two-hone

chieftain.

chariots,

If 'hen they

meet with

cavalry in the battle they cast their javelins at the enemy and then
descending from the chariot join battle with their swords.

The account bv Diodorus

(Diodorus Siculus)

Siculus refers to Gaul, where the war-chariot

became obsolete by the time of the Gallic Wars, but when Caesar invaded Britain he found chariots used in the same way, and more than a century later some tribes in northern Britain were still employing them to
resist Agricola. Some idea of the large number of chariots in Britain is
given by Caesar's claim that after the British king Cassivellaunus had disbanded most ot his troops 'he retained only some tour thousand charioteers, with whom he watched our line of march'.
It may be that the Britons used the same vehicle tor journeys and

s4

hariots and h

Ipposii

tup

ill

iron

".

and mi

the left

./

mi

lit

and a

ion

If

bad

roti

in the toil

i. iririii

.'

cavitu
.in,! tpi

'/<ii

the archaeological evidence

tor war. bui

One day
1968

ia

and no! very helpful.

slight

waterlogged context; in
complete chariot maj be found in
complete wooden wheel was found below the water-table .11 lolme
a

85

An

Notts

Pierreponi

and the excavator did wonder

moreol the

il

vehicle

artist's impression oj

my

.1

It

is

more

at a Torksbin cart-burial.
likely th.it tb

frame "/ .nlr and pok

was there, hut circumstances prevented complete excavation. For the

be constructed from scam details.

moment models and drawings must

Vehicles were occasional!)

hut

graves

in

in

eastern Yorkshire,

conditions preclude the preservation of wood, so only the metal

soil

have been (bund.

fittings
(.arts

buried

In

two burials near Pickering

North Yorks

had been buried complete: both wire found with wheels upright

and the dear

line of a central pole

could be traced

in

the sand of one of

the harrows, Sadrj those graves were excavated a long tune ago, so they
did not provide the

amount of information

that

would have been recoi

ered today. Between 1984 and 1987 five cart-burials were excavated
adjoining parishes of Wetwang, Garton and

Kirkbum

in

East Yorks

there the carts had been dismantled before burial, which was the
Usual pr.u tice
side by side,

in

flat

ontopol them

on the
|S4|.

floor ol the grave,

of

in all

the graves

rein rings
it

had been

the corpse, obi iously in accordance with

>r

the occasional cavity.

slightly different: the wheels

and

indi<

some

preserved mual. But the wood of the yoke had disappeared

without trace, and the othei woodwork had been reduced to mere
(

more

and the corpse had been placed

Alongside each skeleton a line of terrets

placed on the same side

marks

the
but

Yorkshire. In four of the burials the wheels had been set

indicated where the yoke had been buried:

i.ireiulK

.11

ions

<ii

had

he

mm

Ken

h grave, at

Garton Mat

soil

ion. w

.is

leant against the wall of the grave,

woodwork were clearer.

he

wooden

p.irts ol this

vehi

tn

form

.1

(.iimpy.

Celtic Art

cle

had rotted leaving cavities

ot

the grave.

It

in the

compact

filling

had then been temporarilv water-

logged, and clay was washed into the cavities.

Thus

on excavation, areas of clay were found where some

ot the wood had been buried, and the positions of
felloes,

spokes and hubs of the wheels could be dis-

tinguished, as well as parts of the pole and axle of

the cart. But in the centre the filling had been less
compact and of the bodv of the cart only a rectangular outline

could be seen.

It

seems that the standard

practice was to remove the T-shaped frame (pole and

axle)

and lower

it

into the grave after the wheels,

fol-

lowed bv the bodvwork of the vehicle which was

inverted to form a canopv over the corpse [85].
precise details of that

The

86 Bronze and iron harness from

Barrow, Arras: a

the King's

link horse-bit

and

rings

three-

with cast bronze links

of iron encased

in bronze;

a Imch-pin whose iron shank has

corroded and broken, but the cast-on

bronze terminals survive

condition;

and two

cast-on bronze

is

in

good

terrets - the

well preserved,

but the iron bars have corroded and

almost disappeared. The horse-bit
is

272

mm

long.

vehicle in the Yorkshire graves,

The

bodvwork remain a mvsterv.

with two wheels and a central

would have been drawn bv a pair of horses. In the so-called King's

Barrow at Arras both horses had been buried as well: one was a surprisinglv old animal, of no more use in this life, and its burial is consistent
with the discoverv of a defective horse-bit in the same grave [86]. Two
ot the human skeletons in the Yorkshire graven were accompanied bv
swords, and one had been buried with a mail tunic, so it might be supposed that the two-wheeled vehicle had been a war-chariot. But eleven
other vehicle-burials in the area had neither weapons nor armour, and
two had skeletons identified as female and accompanied bv iron mirrors.
The vehicle is best regarded as an all-purpose cart, which perhaps also
served as a hearse, and it may have been placed in the grave to indicate
pole,

the status of the deceased or perhaps to speed his journey to the other
world. There was no point
bolic journey, hence the

in wasting serviceable material on this symworn-out nag and useless horse-bit in the

King's Barrow.
Iron

Age and Roman wheels were made

with either a composite or a one-piece

87

set

of bronze harness from

Polden Hills: five terrets and two

two-link horse-bits. The horse-bits

are

80

220 and 223

mm

long

to

two different patterns,

felloe (the felloe

is

the

o
o o o <>

wooden

'

circumference bound In the iron tyre

Pierreponi wheel
associated with

in

is
.1

Holme

dug-out canoe whose wood has

time of the Yorkshire burials. The

is

composed

I.

excellent condition, and w.is

radiocarbon analysis to about

been dated In
wheel

lu-

hariots and

felloe

the

ol

lu-

segments dowel led togeth.ikes two spokes: ilm method

six

ol

er,

and each segment

oi

construction, which has remained unchanged t"

the present day, requires an iron tyre to be heated

and then shrunk, onto the wheel to clamp .ill the

he wheel has an ash
components tightly together.
1

with

felloe

<>ak

dowells, oak spokes and a birch nave

luor huh
a modern wheelwright would choose
same woods for felloe and spokes but he would useelm tor the nave. Km the very little surviving evidence from the Yorkshire graves suggests that some
of their wheels were made- by the other method, in
which the felloe was made ol a single piece ol ash
1

bent to form

hoop.

The wheel was secured

to the axle by a linch-

sometimes have been hardwood and

on one occasion at least was antler, but the best
pin which ma)

known examples

are

made ol

metal

.The

see fig 12

simplest form has bent shank and ring-head forged

from

two other

iron, hut there arc-

single piece ol

types whose- straight iron shanks have bronze terminals.

he one has

capped

In

ring:

.1

moulded head,

and the end of the

on top or

tlat

sometimes both the top

the head

ol

foot are decorated, cither in rcliel

or with enamel, and several examples are markedly

worn because they have rubbed against the nave.

he second type of iron and bronze linch-pin has
crescent-shaped head which provided an ideal field
!<>r enamel ornament |,HN|.

.1

tion

lone-bits

ol

the

made

iron,

ol

bronze or

two are found hoth

Yorkshire graves and

in

hoards.

in

combina

pairs

the

in

here are two

mam

designs, one with two

between the rings and the other with three; a prototype ol" the
three Imk hit comes from a French burial dated c. 4<io
but in Britain
the t\|x- was still in use in the hrst century
he two horses would
links

i;<

i'.<

have been harnessed one

wooden yoke

tin-

at

lour tcrrcts

rem

likel)

spued

equidistant!) so that

rings,

llrtt isli

In iron

Iron

ol

collar as

seems

that

each side

padded horse
tin-

rings

much

later

development

I) sh.i|H-d,

pu-d

central position,

hnud

bfOWZJt

llllil'-pllli

Sr.iiin !

/;//<

cast in

'saddle sh.i|H-d' har;

somewhere on the hue of the

it

is

always

must have

cart pole,

and

it

ncl.

two

bronze or with

h-pm

It

were strapped to the yoke,

thej oiur in srts of five, not four, and the fifth tenet
.1

.111J

the central pole, linked by a

a

bar onto which a decorative an of bronze had been vast. Hut

than the otlu-rs and has

from

reins of each horse passed through

Age tenets were

SS Iron

ma)

132

mm

I.

Celtic Art

have helped to secure the .strapping attaching the voice to the pole

The

[87].

Gussage All Saints (see fig. 10) suggest

that horse-bits, terrets and linch-pins were all manufactured by the same
craftsman at the same time, so thev were probably acquired in full sets.
One such set is seen in the King's Barrow where an undecorated threefinds from pit 2(f) at

link horse-bit

associated with a

is

cast bronze head

and foot

[86].

knobbed

terret

The products of

and

a linch-pin

with

the Gussage smithy are

more elaborate, because some of the side-links of the bits have lobed
ornament in relief, the heads of linch-pins are similarly decorated, and no
fewer than fourteen quite different types of terret were made.

The

Polden Hills (Somerset) hoard, dating, from the middle of the

cen-

first

tury AD has matching two-link horse-bits and terrets, but no linch-pins,

whereas the Stanwick (North Yorks) hoard has sets of bits, terrets and
linchpins. Some ot the Stanwick linch-pins were surmounted by shaped
rings

which closely resemble the terrets

rary with Polden Hills and Stanwick,

where several enamelled

terrets

is

[88].

third type,

contempo-

represented at Westhall (Suffolk)

were found. The deep decorative arc of

the terret provided an ideal field for champleve enamel, and there are

matching linch-pins with enamelled heads though not found

in the

same

hoard. Horse-bits are occasionally enamelled, but the fields available for

ornament arc much smaller than those on terrets and linch-pins [89].
Although buckles do not seem to have been used, harness straps must
have been fastened and linked in a variety ot ways and there is a wide
range of strap-links and ornamental fittings, especially from contexts in
the

first

century ad

One unique
Found
(

in

ralloway

a
.

[90].

piece of horse equipment remains to be discussed.

peat bog, possibly once a loch, at Torrs (Dumfries and

this

remarkable antiquity once belonged to

Sir

Walter Scott

now one of the treasures in the National Museum of Antiquities

at Edinburgh [91]. It used to be regarded as a chamfrein, the piece of
armour that covered the frontal of a horse, but detailed study by Stuart
PiggOtt and Richard Atkinson showed that it was more complex than
and

89 Bronze bone-bit, with 'enamel'

ornament, from Rise (East

Torks^).

Tbis variety was derived from the

three-link horse-bit (V/. fig-

#6),

but here the side-links are cast in

one piece with the rein-nn^s.

82

is

llillt.

It seems thai the horns, though approximately contemporary with the head-piece, were attached to it in relatively recent

had been supposed.

times, but before 1*2

is

made from two

whin

>

sheets

was

it

perhaps the damaged remains

another

ol

Experiments have shown

forations
al

thai the

k-ss

and

would be

lor the

i-.irs

and not

central hole could have taken

though only one retains

.is

.1

.1

crack
is

with

those

Wandsworth round-boss
see h'^s 10 mil 69

too sm.ill to have been

lor the eyes,

oi

.1

the

see fig 80

With. mi

and

.1

and the conjectur-

plume. The two horns are

They

fillings

s.>mc-

pair,

.1

are decorat-

of which may

scabbard

sic-

fig.

'

of the Irish scabbards

he original function of the horns

could have belonged to

but

on

There are three

the sheet bronze.

pony -cap: thus the two per-

ed with different engraved designs, the motifs and

be compared

in

bird-head terminal.

its cist

fine,

the centre.

.1

headpiece

could have been used

it

itself

jx-rtor.u ions at the sides

in

engraved repair patches, each disguising

more

<>r

symmetrical repousse which respects two

chamfrein, but

The headpiece

illustrated.

first

bronze decorated with

<>t

is

obscure: the)

horned helmet or have been the terminals of

more attractive notion is thai the) were mounts lor pair
.1

.1

.1

drinking horns.

Models

"i

British chariots

bronze handholds
iroin

Brentford

at

.iri-

often

shown with

the back, the type represented by

!5

.t

pair ol

the 'horn-cap'

but the identification of these objects

is

guesswork because none h.is been found in

context exclusively
linked with harness or vehicle fittings.
hey must have been attached to

only

.1

lhariots

and

li

Celtic Art

84

91

wooden stem and the simples, explanation is

thai they were maceIt ma) In- thai some carta had
metal plaques to decorate the woodwork am lit is tempting to see some of the
Stamvick
heads.

bronzes

in this light.

rhedoWul-loobng horse-head whose face is created from

abstract trum-

[x-t-mnnK Kerns an
dered by

.i

pair

<i

ideal

candidate

face-masks

Mosi reconstructions

for a char,,,,

ornament, perhaps bor-

similar style [921.

si,,,*

Celtic chariots open ai the front

the British charioteer running up and
the pole and partly because skeletons
in some French graves ue

partly because

down

in

aesar refers

...

outstretched on .In- chassis, uhich would

be imp
blc if the
chicle had a front. Bui charioteers
would have- had no difficult) in
Springing ovc , | u framC| 1IU
|u much t ir|u r prench u hiJ
uJd
well have- been modified ... order
tocarr) .lu- corpse to the grave
illic
coins sho chariots with a pair of
rounded frames side In side md
similar vehicle ii represented on a
stone carving ,., north I, ah. One
British coin type shows a char,.., In.,
onlj two specimens ... the coin
'K.vcn .nu\ neither .s in
good condition.
hill)

^^
I

chariots

and hai

)p[*.s,..

92 Bran,
Stanwick

tsk

North Torh

Height 101

mm

fhm

Chapter Seven
|

Ritual

/I

GROVE THERE

Aw

.^_ _X_

untouched b\ men's hands from ancient

/CTS",

times,

whose interlacing boughs enclosed a space of darkness and cold

shade,

and banished

the sunlight from above

gods were worshipped

...

and

there with savage rites, the altars were heaped with hideous offerings,
tree

was sprinkled with human gore

were uncouth blocks, formed of

....

<r

The images of the gods, grim and rude,

felled tree

trunks

....

The people never resorted

thither to worship at close quarters, but left the place to the gods.

(Lucan

Roman

poet, writing about a

sacred grove destroyed bv Caesar in Gaul]

Ritual

is

ot

the greatest importance to the study ot Celtic art: most ot

book were probably deposited

in accor-

and the designs which here are coldly

classified

the objects illustrated in this

dance with some

ritual,

and ordered could have been

full

beholders.

The Romans

terms:

the gods they most ot

'ot

also distinguished

two

tried to

of symbolic meaning to the original

understand Celtic religion

'classes of

in their

own
He

worship Mercurv', noted Caesar.

all

men

of some dignity and importance'

the knights and the druids. 'The druids are concerned with the worship

of the gods, look after public and private sacrifice, and expound religious
matters'. Druids were also philosophers and teachers, but their activities

were deliberated shrouded

secrecy and their teachings and traditions

in

were transmitted orallv and never committed to writing.

In Britain places of

that described bv

Lucan

worship certainly included sacred woods,

in

Caul. Tacitus

tells

how

Paulinus, desecrated druidic sites in Mon.i 'Anglesey

Roman

like

governor,

'the groves devot-

ed to Mona's barbarous superstitions he demolished. lor

it

was their

reli-

gion to drench their altars in the blood of prisoners and consult their

gods by means of

human

At the same time

entrails'.

in the

eastern part

of Britain Boudicca's troops were celebrating their rebellion with 'sacrifices,

banquets and wanton behaviour, not only

places, hut particularly in the grove ot

groves might well leave very

has been identified.

Iron

where

built of stone.

Age antecedents. As

may have

in all their

Dio

other sacred

assius).

trace for the archaeologist,

But the Britons also worshipped

Havling Island Hants

Roman temple

little

Andate'

in

Sacred

and none

temples such as

wooden building was superseded by

It

may be

for the

that other

blood and

contribution to make.

In

(
1

human

Roman

temples had

entrails, archaeology

'S4 workers at a peat

bog

at

I.mdow Moss ^Cheshire discovered a human body, well preserved until

it was sliced bv the peat-cutting machinery. Lindow Man had been killed
in the middle of the first century AD: he had been stunned, garrotted,
and then his throat was cut and he was bled. In his stomach there was a
little

mistletoe pollen

a rare archaeological find),

records that the druids 'hold nothing

(

oincidence, perhaps, but archaeology

date for

86

druidic victim.

and of course Pliny

more sacred than the

is

mistletoe'.

unlikely to find a better candi-

ruder)

carved wooden gods reminiscent

have been found

at

the sites

springs

<>i

dated wooden figures from the British

period.

More

seem

omplete examples range

tish

to
in

concealed b)

skin

are distinguished, and often

the carvings depicts not

in

Lucan's description

Mi---

sue earlier than the La rene

some small chalk carvings (bund in

date to the first centuries n and id [931.

height from

almost triangular blocks

feel .in-

<>i

Gaul, but with one exception

relevant, perhaps, are

East Yorkshire thai

(

in

.1

ol

<>r lull
.1

7<i t<>

7<

mm

and most

carved chalk, depicting

lengi h belted

sword

figure bui

.1

is

loak.

whose
arms

leads and

suspended on the back.

shield,

.in- il.n

figure

<

>iu-

of

and two others were found

the same layer as a miniature bronze shield. Perhaps like the shields

they had
ures,

'>r

a religious

function and represent warrior gods, mythical

fig-

ancestors.

Hundreds of crude

'Celtic' stone heads

Inn not one has an Iron Age context and

all

arc-

known from

must be regarded

Britain,

critically;

Yorkshire.

mi the

I'lt

ilk 1 1 ii

nun frm

Height ol
I !<>

mm.

ti

mp*

r***^**

Ritual

one thai was accepted In

most scholars has now been identified as

Romanesque! Massive carved stones provide surer ground, definitely

eltic

and surely

famous

the

is

ritual.

here are

Turoe Mom-,

five,

.ill

Ireland, ol which the mosi

in

granite erratic

''4

)ppos|t,

H
lerel

covered with carvini

elaborate tendrils [94],

POsidonius, quoted

Romans

ai

and part

in

In

Toulouse: 'pan of
the sacred lakes

for tlu-ir treasures'.

bilitj

The

Strabo,
it

being

of treasure found

tells

laid

the lakes
type-site

up

by

tin-

the temple enclosures

particular provided inviola-

in
ai

in

La Tine was interpreted by

de Navarro

.is '.i place where votive offerings

were thrown into the water'
and Fox used the same explanation for the metahvork from Llyn ( errig
Bach in Anglesey. To these examples mighi be added the artefacts r<

ercd from certain rivers

especial 1) the weapons and armour from the

Wit ham and the Thames. Rivers are natural boundaries, and river-cross
ings could also be battle-sites; once a weapon has been dropped in deep
water in the course- of battle us ow ner even if he was lucky enough to
Mirviw stands little chance of recovering it. )n the other hand, swords
. t

<

which are

scabbards are not so likely to have been dropped

by accident and the concentration of finds in selected river beds does
support a ritual explanation. It has been suggested that weapons were
still in

dropped into

tlu-ir

connection with a burial rite perhaps in the way

King Arthur, on the point of death, instructed Sir
Bedivere: 'take thou Excalibur, m\ good sword, and go with it to yonder
water side, and when thou coniest then- charge thee throw my sword
that

rivers in

Malory's

in that

rite in

water'. This idea

is

attractive in the absence of a

much of southern England

before the

first

century

known burial
when ere
lie

mat ion was introduced, but rivers have produced several swords which
are contemporary with the cremation rite, and win are mirrors
never
found in rivers?
Ritual, too, mighl explain some of the figurines and a few other
mysterious bronzes of the British Iron Age. A group of three bronze boar

/i

from

Celtic Art

96

A pair of

bronze spoons

from Crosby Ravenswortb

Length

18 mm.

97 Bronze

disc

from

Diameter 280 mm.

Ireland.

figures

was discovered bv labourers

at

Hounslow

in

1X64 along with two

model wheel [95]. Two of the boars

have high pierced crests (now broken) and the third has the remains of a
stand. Perhaps thev were originally on stands, like a more Roman-looking example from Camcrton Avon
which resembles a tov from a child's
farmyard. Model wheels are known from votive contexts, and a boar is
carved on the side of a famous representation of a Gallic god, so this
other figurines, possibly dogs, and

small collection from

Hounslow

is

best interpreted as ritual.

Curious spoons with short decorated handles are also likely to

have had a ritual function: otten tound in pairs, one of which is
pierced and the other marked with a cross, they are hardly
likely to

have been functional. Could they have been used

feeding the gods? The pair from Crosby Ravensworth

for
(

umbria is typical, though with uninspired decoraon the handles [96]. Thev were tound some seven
or eight metres apart in boggy ground around a
spring, 'well known for its copious supply'. As tor

tion

the bronze discs from Ireland, could they have been

other than ritual? Seven such discs are known, each

with

a circular hollow, slightlv off-centre,

bv high-relief

The ultimate

According
the souls of

number of
es into

to

bordered

scrolls [97].
ritual

was burial of the dead.

Diodorus Siculus the Gauls believed 'that

men

are immortal, and that after a definite

years thev live a second

life

when

the soul pass-

another body'. Beliefs must have varied considerably

across the Celtic world, and in Britain several different burial

practices were observed. In Yorkshire the corpses were buried and

90

Riiu.il

small barrows raised In

then covered In

around the grave

[98J.

cutting square-plan ditches

Such barrows were grouped

most of them have been flattened In centuries

recentl)

crouched

identified In
<>r

aerial

photography.

In

in

brooch

.1

In the rest

ol

<>r

joint ol

.1

cemeteries but

ploughing and onl)

Yorkshire skeletons arc

contracted, and some were accompanied In

goods, such as

simple

meat.

Britain burials are rare until the

when cremation was introduced: before

that

century

first

the lew graves

known

u<

arc

a m in< >n \ of the population and the normal rite has

no trace whatsoever. Most often cremations comprise onl) an urn to

uln iousl) those of

left

"t

house the burnt bones, bui some are more elaborate and have accessor)
vessels

and metal grave-goods.

he richest burials,

ber of artefacts deposited, are centred

in

in

terms

<>i

the

num-

Hertfordshire and have large

equipped with objects connected with eating and drinking.

Some went t<> the grave with fire-dogs for the hearth and a cauldron for
preparing food, and most faced the life hereafter with Dutch couraf
graves

at

Welwyn Garden

n\

in

1965 had five amphorae which

together would have held more than loo lures of Italian wine. That burial

also had an Italian silver

floor ol the grave.

cup and no fewer than thirty pots arranged

he most spectacular item, however, was

a set

mi

conn

full)

grave found

on the

98

Tori

Kit

.it

ultivation have

<

ompl
ut below

the plougbsoil the ftlled-in

ditcba remain, distim th

pi. ill.

:il

./1

11

'II

barm

Celtic Art

99 Four

glass game-pieces from

24 found

set of

in

Welwyn Garden City

Height 20-22 mm.
100 Opposite page
found

in

a grave at
Qsee

fig.

57^).

skeleton

a grave at Deal was

wearing a bronze crown: here a

replica

of

the

crown

is

shown on

the original skull.

ot

unique glass game-pieces: divided bv colour into tour sets of six

were intended for a race-game such as ludo [99]. Grave-goods

pieces, thev

may have been deposited to indicate the status of the dead, or to provide
them for a journev or with equipment needed in the afterlife.
Archaeologists are

to speculate

left

about the

beliefs

which have pro-

vided such an important source of artefacts.

One
at

Deal

ties,

of the most fascinating of British Iron Age burials was found

Kent^

in 1988.

The

skeleton was that of a

man

in his earlv thir-

of slightlv feminine build, but buried with a shield, a sword, scab-

bard and two strap-rings,

[100].

No

coral-ornamented brooch and

other British grave has included so

The crown was

a simple

manv

head-band of bronze with

bronze crown

pieces of Celtic art.

finelv

engraved deco-

band that went over the top of the head. Too

flimsv to provide protection, it must have been a status symbol, and its
resemblance to Roman pricstlv crowns may be significant. In an earlier
ration, riveted to a plain

age such

distinguished assemblage

sight of the Continent

artefacts buried almost within

ot

would have been accepted casuallv

ment of an immigrant. But every

The Celtic art that remains
what was made and used bv the

piece
for

is

study today

Britons.

unenlightening circumstances, because

archaeological excavations

as the equip-

distinctively British.

Much
less

is

ot

fair sample of
was recovered in

not a
it

than a third came trom

and that includes nineteenth-century exca-

vations. Nearly ninetv per cent of the objects illustrated in this book
were deliberated buried: about thirty-five per cent were in graves, and
a similar percentage in rivers or other watery deposits, while almost

twenty per cent came from hoards. Of the rest, most are isolated finds
whose precise context is unknown. Metalwork is represented disproportionately because of its high rate of survival: wood, leather and even
skin was probablv decorated but hardly any of these materials has been
pre-erved. However, in spite of its limitations, this selection is more
than enough to show that the British contribution to Celtic art was sec-

ond

to none, and to establish Celtic art as one of the outstanding

abstract arts in world historv.

92

Ritu.il

Further reading
The

present text

(1985), but

is

based on

Celtic

Art

in

Bnr.nn

before the

has been thoroughly revised in the light of

it

Roman Conquest

new

discover-

and recent research. Through the kindness of Professor lope have

read the proofs of the standard work, Early Celtic Art in the British Isles hv
P. Jacobsthal and E.M. Jope), soon to be published bv
)xford University
ies

most recent surveys devoted to the British material

and Purpose by Sir Cyril Fox (1958) and Early Celtic Art

Press; otherwise the

alone arc Pattern

in

Britain

and Ireland'by R. and

V Megaw

(1986, 1994). For the Continent,

Art (1944, reprinted 1969) is still fundamental,

while more recent volumes dealing with both British and Continental
Celtic art are P.- M. Duval's Les Celtes (1977) and R. and V Megaw's Celtic
Art (1989). A full survey of artefacts in Scotland and northern England
i> given bv M. MacGregor in Early C-lric Art in North Britain (1976); the
P.

Jacobsthal's Early

Celtic

Yorkshire burials are dealt with

in

Arms

I.M. Stead's The

Culture (1979),

and there are excellent accounts of the Irish material by 15. Rafterv in A
Catalogue of Irish Iron Age Antiquities (1983) and L.i Tine in Ireland (1984
A useful text book of the British Iron Age. with full bibliography, has
.

been written by

B.W

unliffe {Iron

Age Communities

in Britain,

3rd edn,

1991).

Coins rank among the

finest

examples

of

Celtic art and provide a

wealth of information about the Britons, but they have been excluded

from the present book because they bear

forms. For an excellent introduction see

little

I).

relationship to other art

Allen's

An

Introduction to Celtic

Coins (1978).

Books, however, are no substitute tor looking

selves,

and the British

Museum

at the objects

them-

has an incomparable collection.

Acknowledgements
The author and publishers are grateful
following photographs;
l.M.

Mead:

IS,

of Wales; 42,

1,

P.M. [ope;

Museum

lor

Aerofilms Ltd;
25,

Museum

permission to reproduce the

9, R.I..

Wilkins; 14, 57 and

of London;

36, National

^8,

Museum

of Antiquities of the University and Society

of

Newcastle upon Tvnc; 45, Cambridge University

Museum of Archaeology and Anthropology; 4 >, National Museum of
Ireland; 67, Reading Museum; 84, A.L. Pacitto; 91, National Museums of
Scotland; 94, Commissioners of Public Works in Ireland. All other phoAntiquaries of

tographs are copyright of the British

burial, 85,
line

94

is

Museum. The painting of

reproduced by kind permission

drawings are bv Karen

Indies.

of"

a cart-

Peter Connolly, and the

Indc:
ex
Alklum

53, 54,

4\

Km^

Arthur.
Nt

82

Aylesford 9, 54, 55, 56,

18, 53, 54, 56,

helsea

Sii

hertsq

levedon 45, 46

lonoura 71

712

30,31,

12,

olehester

Barn Elms

<>l

''I

mps,

6,

57, SS

Glascote
gold

42, 43

Battersea

liirdlip

boars

53,

.nil Brai kstead 60

.nut

<><>.

51,86

42-3,

39, 41

iraves

Broighter 45, 47

bronze-working

17-41,42

I5nK.kl.inds

buckets

9,

14,

54-6

12,

35

58, 59

\hn

SI

,|ss|l|s

B6

4,

-,.

9,

16, 53,

airnmuir 50

amerton

apel

arbon-14

2.

S2

Hill

16,

12, 57, ss

lolmc Pierreponl

'horn-cap'

horse bits

6, so. 81,

llounslou

ruin

'enamel'

Ipswich
44

.istlr

<

50

iron-working

Evans,
ellaunus

'..

Sii

14-is

<<

IS
8, 19,

m^

HI

I,

A.

77

2'',

Jacobsthal,

65

Jope,

20, 21. 26

I'

E.M

58

41, S4

7H

Felmersham 56

Kirkburn

si

<. <<..

44
i

auldrons 53, 91

figurines

82

9, '.2

druids 86

clfi

52

79, SI

Islcham 65

..irmnn
5,

Herodian 37

houses S2

Drummond
(

4.

90

dragon pairs 2s
.u-vir

Hayling Island 86

77,78,

discs

Holcombe

90

Buriitn Fleming 37, 42

4, 5,

13,82

[arpenden 56

Hod

Desborough
(

s. unts

tussagc All

Hertford Heath 37

Diodorus Siculus

If.

14.

Bugthorpe

21,41,

89

dendrochronology
S- 13

1,

helmets o7-71

de Navarro, J.M.

25, 26, S3

61

77, 92

brooches

hesterford

Great Chesters 39
(

14,60,62,64,65,

Deal

Brentford

ll.illst.itt

Danes

58

bracelets

SI

IS

II. 12.

Boudii

'l

^2

12, 57,

52,67,

ST. SS

daggers, sheaths

31

Ixur skins

''2

>

76,77

8, 60,

'bean-tin'

silver

4, 37,

12, 21

Ravensworth 90

n.sln

6, 7

78, 8

I"

n lam

barren

AW

pieces

(..mis

85

18, 50, '.7,

coral

game

Garton l

cremations 91

12, 34, 58,

'.I

compasses

Bann, River

Sii

Franks, Sir

coins

Banks,

Fox,

11,21

chasing

Atkinson, k |(

Baldock

2S

IS, 23,

formers 9

B5

20

chariots 6, 12, 7s

89

\2

\>t.>ti

19,

is, 52, 53, 91

ton

si, 82, B3

80

4=..

43,

J,

.irt

champleve enamel

dogs

fire

chamfrein 83

''I

Angles*
.inkltt

21, 22

idion

71

'.7,

amphorae

*.

87,

irk.

Rivet 65,

Ind(

3 9999 03376 567 6

La Tene

Polden Hills

Lethbridge, T.C. 66

13,81,82

linch-pins

Posidonius

65

Wittenham

Little

temples 86

pony-cap 83, 84

2 l>,

35, 62,

potterv

terrets

58, 65, 89

lotus flower

Ratcliffe-on-Soar
razors

Lough Crew

11,

26

36

rings

12

IS.

ritual

Margate 53

olleetion

71,

St

Stephen

Mildenhall 45

Seott, Sir

Sedgeford

12

Walt

82

YV.

shields 71-7

Welwyn

silver

Welw

18

Needwood

sling-stones

rot't

25
36,

Snailwell

60

Warden

>ld

>u sleburj

57, 58

60

16, is. 26,

.arden City

54, 91,

66, 67

Wetwang Slack

30, 31,

79

wheels 78, 79, 90

'Mi

Standlake 25

Whitcombe 11,60

palmettes 21, 22, 24

Stanfordbury 53

wine 53,

Park Brow

Slanwiek 81, 82, 85

Wisbech

22, 34, 65

IS

Paulinus 86
Philostratus

Pickering
Piggott,

96

37, 52
(

38, 39, 41,42. 43,78,

spears 60-1

spoons

West Row

42, 43, 60

47-50, 68
(

West hall 81,82

52. 60

Snettisham

17-18

21, 29, 31,

56, 61, 62, 68, 75, 77,

skeletons 36, 79, 91, 92

51

ham Abbey

83

50

necklaces 43, 44, 45

North Grimston

37

Wandsworth

58

age, R.

mirrors 57-9

New nham

35

trousers

77

Salisbury
S.i\

Forest

triskeles

[uroe ss, 89
57, 5s

72

moulds

43-51

18,

trumpet shapes 34

91

Marlborough 55

Mayer Collection

11, 16, 18

Torrs 68, 82-4

45

86-91

Rudston

<>,

torques

Rise S2

Lucan 86, 87

29,60,

89
tools

Richmond 62

22

82

6,

(.2,6d, 68,71, 75,

(.I,

Llyn Cerrig Bach 34, 35,

6, 80, 81,

Thames, River

53

6,

22, 25

tendrils

89

4,

65, 66

Me\ nek

Tacitus 66, 86

tankards 54

Polybius 44

Lindow Man 86
Lisnacrogher

19, 80, 82,

83

S.

pins

41-2

Plim

86

Stillingfleet, E.W.

19

7')

82

Wit ham. River

Stone heads 87

Strabo

4.

strap links

24, 2

is. 52, 57.

89

20-52, 60-6, 89

6, 12,

>,

6, 17, 19,

56, 65, 66, 72.

74, 77, 83, 89

wo. id

82

swords, scabbards

37

Wood Eaton

38

Brighton Branch Library

40 Academy Hill Road
Bnghton MA 02135-3316

RA( TAUOt

Celtic Art
he Celtic-speaking Britons who
inhabited England, Wales and part ot
Scotland

in the

fne hundred years

left no

before the birth of Christ

,

written history. However,

archaeology has revealed some ot their artistic

achievements, and every year more objects
are unearthed. Jewellery, weapons,

armour

and
harness are magnificently decorated with
fascinating and powerful abstract designs.

and the metal

fittings of chariots

In this fully revised

and updated edition

of the successful 1985 publication, Dr

Stead examines the craftsmen's techniques
then follows the development of certain
patterns, before finally describing a

number of the surviving masterpieces

ot

Celtic artistic achievement, such as the

Battersea shield and the Aylcsford bucket.

Dr Stead is an expert on the Iron Age

and a former Deputy Keeper ot the
Department of Prehistoric and RomanoBritish Antiquities at the British

With 100

illustrations,

43

in

Museum.

colour

HARVARD UNIVERSITY PRESS

Cambridge, Massachusetts