HANDBOOK OF

JOHN W. HAYES
 Handbook of
Mediterranean Rom
Pottery

John W. Hayes

IISTITUI 74" lIST!

A G RE

lor the
BRITISH
 OTE 0 SPELL! G OF NAMES
Contents
Where ancient names are cited in the te t and appear on the map,
the Latin form is used for Italian sites, but the Greek spelling
(transliterated) for most sites in the East of the Empire, where Greek
ienlained the common language for written texts. (Greek k counted
as the equal of the Latin c, and the Greek ending -os, -on as the Latin
-1lS, -um.) Anglicised versions appear where these are familiar: e.g.
Map of production sites and findspots
Corinth. Rhodes.
Acknowledgements
Colour plates

INTRODUcnON
Distribution
Identification of sources
Functions and shapes
Production
Sizes
@ tfJln The Trustees of the British Museum Quantities
Firs~ ~Iished in 1m by British Museum Press
A diV1SlOl\ of The British Museum Company Ltd
46 B\oomsbury Street, London welB 3QQ AMPHORAE
Dolia
ISBN 00-71.41-221~5
A catalogue record for this book is available
from the British library
FINE WARES
line illustrations by John W. Hayes Forerunners
Pholography by the British Museum Photographic Service Red-gioss sigiIIata wares
. Map by Technical Art Services
~ b): Behram Kapadia in Paialino and Gill Italian and western terra sigillata
Printed m Great Britain by The Bath Press Eastern and . si&iJIaIa
African Red Slip ware and other Late
Lead-and
Thin-waIled wares
Other 'fine' wares
 75
COARSE WARES
75
Buff wares: closed fonns
76
Cooking wares
80
Mortaria
83
Local Egyptian wares
We wish to record the most noble and miraculous artefact
VESSELS WITH SPECIAL FUNCTIONS
that was made of it. From this vases were made for many 0cca-
sions by the most noble and skilled Craftsmen of old 01
Unguentaria
noble city of Arezzo where we were born.... When, in our
Ritual vessels and religious souverurs
one for some reason dug within the city or round about" up
Toys two miles away, great quantities of these pieces of v _
Various products found... buried for well over a thousand years, and found
colourful and fresh, as if just made, so as to confound the
with pleasure... When one of these pieces came into the
Appendix: A brief history of the British Museum's collection 90 of sculptors or draughtsmen or other experts, they wouIll
them as if they were sacred objects, amazed that ~
Endnotes 95
could rise to such delicacy. . . and would say
Select bibliography 99
Craftsmen were divine or that those vases came
Index of items illustrated 103
heaven.
 Acknowledgements
he initial suggestion to me to write this Handbook came from Dr
T Susan Walker, Deputy Keeper of the Department of Greek and
Roman Antiquities in the British Museum, who has supported the
project during the various stages of preparalton (col. pI. II and pI. 14
reproduce illustrations from her handbook Roman Art). It ~as she
who encouraged me to contribute the necessary lme drawmgs -. a
number already have a thirty-year history, and have appeared m
print elsewhere. Her colleagues Dr Donald Bailey and, more
recently, Dr Paul Roberts (who has provided drawings for figs 8.2
and 25.9) have offered advice and practical assistance on numerous
occasions. The staff members of the five British Museum depart-
ments holding material illustrated here (see p. 95, n. 1) are thanked
for arranging study facilities (including long sessions in the
Museum's basements and reserves), searching inventories, ordering
photographs and the like. All the final photographs reproduced
have been supplied by the BM Photographic Service. The cheerful
collaboration and encouragement of Nina Shandloff (a Senior Editor
at British Museum Press), who has had the unenviable task of chas-
ing an author more often absent from Britain than at home, is much
appreciated; I thank all at the Press, including the designer Behram
Kapadia and copy editors Anne Marriott and John Banks, for co-
ordinating the fina1 stages of a rather complex publication within
pressing time limits.
J. W. HAVES
Oxford
16 February 1997
10
INTRODUCTION
"T"his short survey offers a cross-section of the pottery IIIlIde
.1 marketed.in an~ around the Mediterranean (the _ ~
the Romans) m anoent Roman times, here illustrated by llIIIleal
many items in the collections of the British Museum. While
wares from Britain and nearby are perhaps more famiIiu to
British visitors, and are the stuff of many museum display-.
mates from the provinces at the heart of the Roman ~
ing Rome itself, have often been less easy to visualile. The
of the new Wolfson Gallery in the Museum now at 1alIt permiIa
sampling of these Mediterranean wares to be viewed in ClIlt
The Museum, by dint of its long history of collecting. .
well placed to illustrate the products of these other pull
Roman Empire.'
For the better part of a century, Walters' 0IttII0gw'
vademecum to those seeking information in English OIl
tery from Britain and the Roman world at large. Since
the study of Roman Britain and its neighbours in the
- Gaul and the Rhineland - has made great ~
field of pottery research. Indeed, a specia1ilt joumII
Romano-British pottery.' Until fairly recently, dloup.
Roman-period pottery in Mediterranean 1andI
For many years, it was treated in .
indeed, in its native regions) as the poor
which were more suited to the dominant
research. It is only since about 1970 that 1oc:aI'
in Mediterranean lands has adopted modem
niques and come to terms with the full ranee aI
products, the result being a plethon aI
languages. The first sections of Walters' 0IflIIaIw
good deal of value.
Some of the pots illustrated below appIM'
few are listed in Dalton's CIIttIlope tf EMf
lIODIe of the Egyptian items are pubUIhed
reports, but others have hitherto
hen! to present a series of new
m-iy available elsewhere. The
eeIective, with an emphasis OIl
For the sake of convenience,
 de and used within Italy and the Roman provinces
as that mba t BC and AD 600 and its stylistic precursors and
between a ou 100
,( ,h' h 'n the East
'"
lasted unllithe nse
f ,I
0 Is am,
) In
po
n-
I
successors " IC , I , " th
, I t ms the history of Rome and its empIre compnses e
Ica e r , 'I 'd( 27
Re ublican period (c. 50<)-3' BC) and the Impena pe~lO 3 ' or
.p 10
or ,-6 depending on whIch cntena are used),
Be-AD 324, 4 .." d f h . t
Culturally, it overlapped in its earh wars the en 0 t e ancIen
Greek world - the 'Hellenisllc' period - and, later on, the emergence
f new Chnsllan 'Byzantine' ci\ ilisatlOn centred on Plate I Dec"""", ~ ...'" on
~oll-~antinople, orth of the Alps, the beginning and end of lem "& la. AppJ.ed mobls. ....
'Roman' culture are convenient Iv marked by the conquest of the stamped by L Gel o..m 75
regions by Julius Caesar and his ,uccessors and then by ,~elr loss to haJ'd-roIl<d ( apploed spectode lpnl
plate Barootme ornament 8cawI.
Germanic 10\ aders 10 the years after AD 400, both of \\ hich events o..m
d sh from Tharros. 65
led to marked changes in life-Sl) les, In the Mediterranean, however, RouIettJng. 8oaom '9't C\4>. tom
Rome took 0\ er from earlier civilisations, and the Byzantme and Own I] 2 ern
Islamic cultures which followed adopted many 'Roman' features.
Hence what we term 'Roman' pottery cannot be bracketed within
hard-and.fast dates, Its principal technical features - red (rather
than black) glossy surfaces, the use of stamps and rouletting.' bar-
botine and mould-formed applied ornament (pI. 1), manufacture
within moulds or owr forming devices - are in fact all in evidence
by c. 200 BC somewhere in the Mediterranean world, whether in the
'Hellenistic' East or in Italy. Only the use of lead glazes (not parti-
cularly common and in any case intermittent) strikes a new note: see
col. pI. V. In common with other craft products, the pottery of
Roman times exhibits certain stylistic trends, some of Italian origin,
others resulting from the exchange of ideas within a single vast eco-
nomic system - the Empire. By later Imperial times (third-fourth
centuries AD) various provincial centres, from Britain to Egypt, were
purveying their own 'Roman' styles of pottery. Each of these was Roman pottery in the same way as one d
distinct, but all hared some common features. though the two are often related in terms of
Our knowledge of Roman pottery comes almost entirely from Greek vessels are appreciated mainly for the
archaeological finds, especially the complete vessels found in graves them, which may reveal an individual artist' hind.
and in ancient shipwrecks. Specific references to pots are rare in however - with some notable e cepti
~ literary texts and documents (a few can be gleaned from outer provinces - exhibit the potter' art.
papyn), and types of Roman pots recognisable to an archaeologist pleasing combination of form with simp
do not ~ figure on Roman artwork uch as sculpture, painting. rarely reveals the personal touch. growins
or _IC floors. 1be vessels depicted on the latter are either more duction (always a matter of Ii elihood, for the
up-ma;rket (silverware, for instance), or are conventional renderings potter) is in evidence, with increased
of ~t falUts like the Greek mtn', or again are shown in very oration mechanically. Where figured
~ - one might say 'cartoon' - form. to be in the form of cartoon-like indi
. a~ can ~ taken to this abundant material. rang- known motifs (standard Imagery 01
In8 from pure artistic apprUItion to chemical analysis of the mater- tiler than co-ordinated b!lIinI
ials uaed and tiItical. study of trading patterrts;o others may yet GIftk art. The exception to this QII
Proal an artistic standpoint, one cannot easily approach raft lIIId Iead-glazed Is 01 the
 . (c 25 BC-AD 50), The delicate relief ornament on F1eure 2 Lathe-tumed bronze bowl from
and his successoflrsects the 'official' court aesthetic of the period, Nubl' (scale 1:3).
these vessels re ed' ( II . tin'
'ch in a very wide range of other m 1a wa -pam g,
whi recurs , . d )
. are ivory-earving, cams, gem-engravmg an so on .
silverw Pots in common with Greek, also provide evidence for
~t~teraeY: makers' stamps and shipping instructions quite
ften pear On them and pots (or broken fragments) can also serve
aas thea p, l'th'A'
medium for longer written texts (an examp elSe swan
dish noted below, pI. 2).'

DISTRIBUTION
A major feature of the Roman wares of the Mediterranean is the very
wide distribution of a relatively small number of fine wares. As a
result of easy sea links, the whole Mediterranean coastline served as
a potential market, as it had done preViously for Athenian products
in the Classical Greek period of the fifth century BC. These wares -
the Campana black-gloss wares of the Naples region and Etruria
c. 2Olr50 Be, the Italian and Eastern SigiUata wares in the first and
second centuries AD, and the African Red Slip wares and their
derivatives thereafter - were particularly influential and were
widely copied (pl. 14, fig. ).3). Some pieces of the same fifth- and
sixth-eentury Asia Minor and North African vessel-types found in
excavations in Istanbul and at Tmtagel in Cornwall (fig. 1.1) may
serve as an illustration from later times. Some cooking wares
(chiefly Italian and North African) were similarly traded.' Rarely, a
fine-ware pot from far away may have inspired metalware, down to
the maker's signature, as seems evidenced in Nubia (fig. 2 and pI. 16
bottom - possibly a 10caI copy of Gaulish samian ware).
1 located, but only one or two have been
2
4
1
3
fiaure 3 (I) N.batean bowl from Petra. gift of P,"1Cess Mochael d KsIlII"''''
(2. 3) Eastem copies of italian cup-shape~ (2) from Kntdos (Newlan. . . . . . . .
(3) from Ephesus (Wood's excavatJon~ 1864) (all scale 1:3)
IDENTIFICATION OF SOURCES
In contrast to the situation in Roman Britain lind
duction centres of Roman Mediterranean wues
documented, as is also the case with their Greek
sors. Where kiln-sites are known, they are rueIy
Among the workshops producing red-p.
those making Arretine ware are documented by
products (chiefly mould fragments, pL 4),
century ago. The workshops of PeJpmon
Loeschcke's work carried out over eighty
by more recent finds. Most other Ealem
tres are not even known. Sonlt! of the
is known of the factories Cor lead-glued
duetion centres of amphorae are ~
IitenIIy hundreds remain to be
lIOUI'CeII do exist but they
~ brief mention 01 production
papyrus with deWIs 01.
pols (mostly amphorae _lied
P. 'Z/ Stamps on fine-ware
, together with prod
PI.,e 2

latter (such a, mked or pamted inscriptIon, and stamped stoppers)

may provide more information on both sources and original con-
tents. In contrast, those on bulky storage-jars (do/ia), which are
common in Italy, often indicate the owners of workshops as much as
their physical locations, since in this case it IS the potters, rather than
their pots, who may have moved around. Quite unusual is the late
bowl (pI. 2) of EgyptIan Red Slip A ware. This ware is generally
assigned an Aswan source, here proved by an ink inscription in
COptIC, translatable as 'the plate [pillax - a Greek word) of Souan'.
This bowl was found downstream at Nagada near Luxor, hence the

Plate 3 Moo made imp f P d dn

1 ')'pc I Egyp ,n
'. '. 'n e ~elta fae': re h g--
qUG! A~ t ~d on the om as
g., lOI'g
bali e1.r Lnd tur AD
engtr 78 rr "" dth 6 c

A, ph n.
d
'0 I
Egyp lte 4tr)
C'~n,
t An ,
 wares , ed decorated wares made In the Aswan reg ~n. Upper (~ Egyp
13 Minor two-tone parnted decoration (scale e I' 3) Left '.0 ngll two-handled...,. WI
eaker painted leaf-sprays (black) and bands (red), shal ow cup WIth paJnled "/>an
ergamon, ntenor; two-handled Jar WIth two-colour barbotlne omamel'lt AI teo I or
>rontle(t: AD.
dn. stamped
) AD I 30
relief-ware
c uol estored),
Ar line e 25 I Be.
beaker. signed ed cups, early to mrd-I st century AD .eft probably Sc
WRVS,
o\IVS d Paphos), Cyprus. RIght: Mytdene pnoduet W1th whne (a
AD C RI r platter. nm) 9.5 cm; (with handles) 152 em He gr-t 7 b em
AretJ war'e stamped
CN A;-E e ') BC ADIO
DMTl 327c~

III f'aJnted Egyp an

2nd entury AD
II Abo,e; Red-gloss wares IV Thin-walled decorated wares made In the Aswan eg ..m Jpper ~ n
from italian and Asia Minor barbotlne or two-tone paInted decoratl:ln 'scale 3 eft gh tv.-c hand~ "~:
workshops Left beaker attached lid, pa,nted leaf-sprays (black) and bands (red) sha ow c p pal t "l
with appliques, Pergamon. (black) on Intenor; two-handled Jar with twocohur b;\l1>otlne or Iment AI te
c. 100 SO BC Front left 2nd century AD
small bowl, ItalIan. stamped
INAC / HV5, c. AD I 30,
Top centre: relief-ware
chake (foot restored),
Arret,ne, c. 25 I BC
Front fight beaker. signed V Lead-glazed cups, early to m,d-I st century AD Le~ p",bab/y c;.
L5ARIV5 LL5VRV5, Kouklia (Old Paphos), Cyprus Right Myt, ene produ,. w<th w te
c AD I 30 fOght: planer glaze D,.m, (nm) 9S cm: (w,th handles) 52 cm He ght 6 ~
ArretJne ware, stamped
CN ATEl.c 10 BC ADIO
D'am 321 cm

we Pa cd I EgypL
C<:'llu AD
 need for a label. Stamps may, however, give unreliable SOurces _ for
example, the 'Arretina' stamps on some eastern finds usually adver-
tise 'Arretine-type' products made locally, rather than the genuine
article. The clay lamp (pI. 3) which calls itself a '[product] of
SegaIassos' (a fine-ware production centre in Asia Minor) was aclu-
8lly made in Egypt, in the western Nile Delta - presumably in a new
workshop staffed by foreign craftsmen intent on advertising their
wares.

UNCTIONS AND SHAPES

Mediterranean lands, where wheel made pottery had been the

VI Two small contaIners. for perfumes or medications, Nile Delta, perhaps 2nd century AD.
Left bust (bottom mISSing) of Hermes w~h lotus leaf flanked by wings on brow Height 9.6 em.
Right head of man. bald. Height 86 em
VII Plate wM serving compartments (small Inset bowls), Aswan (Egyplian) ware. copying
Afncan Red Slip ware. Pink clay. yellow slip with black painting, c 7th century AD.
D,am (nm) 33 em Height SA em.
for centuries, if not millennia, the role of pottery vessels as
rs either for on-site storage or for shipping, as cheap _
than luxury - kitchen- and tablewares, as burial urns and
in graves, was well established. These basic functions
little during Roman times, though some differences in
. may be seen. Changes in burial practices led to an
use of glass vessels and clay lamps at the expense of pot-
vessels, and eventually, under Christian influence, to the aban-
t of all grave offerings, though this was not uniform in all
. Glass vessels came to challenge pots as cheap drinking vessels
containers for liqUids, to the extent that some pottery vessels
glassware shapes. The cheap pottery perfume flasks (unguen-
which were often placed in graves in Hellenistic times had
effectively ousted by glass versions by AD 100. A fashion for
. 1 urns is also to be noted.
COmmon functions for pottery vessels may be observed,
use as clay beehives, lanterns, and hot-water compresses
purposes. Some of these originated in pre-Roman limes;
examples cited are commonest in the Roman perIod,
only in certain regions. Some pots were frankly made as sou-
inscriptions on them referring to spectator sports. (The
was not the prize, as the Greek Panathenaic vases had

(and also Jewish) culinary practices demanded specific

, examples of this being the flat baking pans (patInae)
COvers noted below (P.78). However, the traditional
. basins for kitchen use (mortaria, fig. 34) were n<;,w
by stone versions. Common Mediterr~eanpracti~e
from that in the northern Roman provInces, as dId
for drinking vessels with or without handles. Further
on these points are made in the text below (pp. 78--80).
of Mediterranean Roman pottery, particularly the

17
 . 'tate metalware of their own and occa-
tablewares, ~~tantJy urn I. 1 ). This reflects the natural desire of
sionally earlier times (seea~ket ~ silverware was, after all, the basIc ,, ~

the consumer to g~p-~e example of this is the flat platters noted

Roman status sym fi' 8) which were difficult to make.
below (pp. 2()-1, see gs. 14, 1 ,

PRODUCTIO I
I
I --
Most Roman pots were simply thrown on a potter's wheel and then , I
--
--
off the still-rotating wheel with a Wire or stnng, leavmg the typ- I
I
---
.-
: 'fin rin!' marks on the bottom. A neater base (or low foot) -
~as th!~rma1Iy produced by inverting the vessel, haU-dned, on \ .--....-
~

the wheel and shaving off excess clay as both revolved. Marks of ----/
this turning process are often visible on the finished vessel m the ~
form of horizontal scratches and spatula-marks (see fig. 3 centre, pI. 2 1
and others). Necks and feet might be formed a~ separate elements
and be added afterwards - slight bulges, seam-lmes or rough edges
on interiors may be evidence of this. The cos/rei shape, particularly
common in Egypt and Palestine, comprises two wheelmade ~Ie
ments for the closed body and a third for a neck attached at the Side
r-
,
/' ,
-...,. -
;II
(fig. 4.3, cf. fig. 25.9), to which handles may be added by hand. Large , 'I( ,

tall vessels such as the commercial wine-amphorae tended to be " ... ..I,I ,"
assembled from three or more wheel-thrown sections, the joins "
I......
I...... JL"; _-""...... ,-
I --- -- I :
being smoothed over before the firing. An Egyptian jug (fig. 4-1) is , 3
unusual in having a body formed of two pieces which had been
rotated in different directions on the wheel, as is clear from the
ribbed treatment visible on them.
The finer Roman wares are generally more thinly potted and del- flaure .. Techniques of manufacture: wares from Egypt (scale 1:3),
icately finished than modem craft pottery (see especially fig. 27).
The basic methods of production can hardly have differed much
from those used by earlier Greek and Near Eastern potters. Indeed, of manufacture is seen among lead-g1azed wares. In other
almost identical pots (often ribbed in the Roman marmer) are individual motifs made in moulds were applied to the ..m-
known fron. Syria and . over two thousand years earl- wheelmade pot (see the beaker in col. pl. m.
ier. However, a significant proportion of Roman vessels were In fact, the old finds still tell us very little in detail.txJUt
formed within a mou1d mounted upon the potter's wheel; elements duction techniques of the potters of Arezzo, except by
such as rims might protrude above the encasing mould, and feet
their suc:cessors in Gaul. For many undecorated WIIeII
and ~: formed separately, might be added after the initial duction aided by moulds is probable, though di8il:ult
forming. Sigillata ware bowls with relief decoration were made thus Eutier generations of archaeologists may simply haft
(roL pL n top cmtre, and fig. 5), and a good number of broken pieces
the presence of plain moulds, especially if broJcen
of the ~ moulds front the Arretine worbhops (pl.... and fig. 5) Ie materials like plaster.
can be studied at the British MWIeUDL TheIle take the form of thick-
walled bowls in smooth hard-fin!d reddiIh ware. Punches (usually
the mass-produced sigillata wares of Italy and
looted
~ to by the French term~) wen! UIIed to impress the plates with wide horizontal fIoon, fJaD
profiled rims rise at the outer edge (figs. Lto J8).
individual motifs 01\ their 8IIlOOIh inner eurflIcea. The same method , measure 60 em or more in diameter
18
 the Arretine example (which is quite often observed) Presumably
occurred at a late stage, after the plate had been turned the right
way up in order to receive the potter's stamp. Slight traces of stac:k-
ing are seen at the bases of the feet and on the floors immediately
above them, showing how such vessels were positioned for firing.
Note the imprints of three rectangular support-pads seen in coL pL
n and pI. 5 These seem to have been coated with fine sand or the
like to prevent the clay of the pad from fusing with that of the plate
during the firing. The marks of such pads are especially COIIlDlOn on
the heavy dishes produced at <;:andarh (Pergamon). Particularly
elaborate stacking arrangements were needed for early Roman Ie8d-
g1azed wares (see pp. 64-5).
As in Hellenistic times, many vessels, including some of the fine
wares, bear a slip on the more visible parts only (Le. the outside 01'

Plate 4 Mould nghl for the product on of Arrebne rei ef ware and eft modem cast
(scale 1:3)

-- i
support must have been needed to prevent sagging and deformation --
of the slill-wet clay during their manufacture. Two possible ways, __.~h
involving moulds, for Eastern Sigillata A and for Italian (Arretine)
Sigillata respectively, are illustrated in figs. 6 and 7. In both instances ..-'----i ;---_ _..J_
one surface of the resulting pot is completely smooth, while the
other (that presumed to have been exposed dUring manufacture)
displays slight brush-marks. The gentle sagging of the heavy rim of
..... 6 Eastern Sigillata A ware plate: probable technique of manufactun! (. . I
fiaure 5 Cross-seclJon of the mould above (pI 4). and reconstruction of the type of
vessel produced from ~ (scale I J).
..... 7 ltaJoan terra sigtllata plate: technique of manufacture (scale 1:3)

..~ \

Plate 5 Production defects. Top: floor of
Arreline platter (see col. pI. II). shOWing the
Impnnts of pads (sanded, to aVOid fUSIOn)
from stacl(Jng for firing. Stamp of
en. Atelus, wrth letters combined.
and palmbranch omament (scale 2:3).
Left Jug, dented In the kJln, from Kerch,
Cnmea. Local early Roman product.
Height I S.9 em.
tripod stilt arrangem~nt used to sepa,rate them in the kiln. Some
were clearly fued upside-down, smce drops' of glaze have formed
on the rim, creatmg pronounced knobs (see fig. 26); on others the
glaze is of uniform thickness. Some recent finds of kiln-furniture
and debris from Perge in southern Turkey - now seen to have been
a significant production ce~tre .in the period from Augustus to
Claudius - help to c1anfy this pICture. The stacking arrangements
for the kilns evidenced there are in fact quite complicated.
It seems likely that some of the larger African Red Slip products
were made in plaster/stucco moulds (just like contemporary
lamps), though at present the only definite evidence of this is for
vessels which were clearly not formed on a wheel, such as rectan-
gular trays. Many Italian thin-walled bowls and beakers of eaJ1y
Imperial date, and not just those bearing relief patterns (col. pI. n
front right, and fig. 15. 2 ), may also have been made thus. It is diffi.
cult to imagine how eggshell-thin vessels of the type of fig. 27.2
could have been made in quantity (as they evidently were) without
some moulding process. The careful turning required to produce
their smooth exteriors would otherwise have been very risky, and
would have slowed down production to a point that was surely

top), the result of a quick partial immersion in a clay solution;

smudges (here visible on pI. 5 left and pI, 14 right) often appear
around or above the base to show where the vessel was grasped for
this purpose. On several Eastern fine wares (see below) one can
observe a transverse streak of slightly darker colour where two par-
tial applications of slip overlap. This 'double-dipping streak' is
already seen on some second-eentury Be black-gloss products from
Greece (pI. 6) including Corinthian relief-ware 'Megarian' bowls.
The application of the glaze material on the lead-g1azed wares
seems not to have been uniform. Some vessels bear the scars of a
.... , 8lack1!k>ss medallion bowl, shOWIng 'doubledlppmg streak', Diam. 16.2 an.

2)
 . m also to have been used in the pro-
uneconomic. Partial mou~ds see _ notably the classic varieties of
duction of large flat ~aking) p
ans
of which are stamped or signed
. Red ware (fig.}2 - some
Pompelan

Some.other wares, I
ih
in relief on their bottO:: fine and coarse, were made in the man-
b ressing clay into two or more
h
ner of figunn~ and amp~d Passembling the impressions into
moulds of vano~ : \ : ' cl~ shapes Gugs, bottles and so on)
hollow vessels. ff' y . es could be produced: examples here
d ases in the form 0 19unn . I VI)
an v 'd' .
are the Kni Ian senes . (pi 26) and the figurme vases
. (col.
R d Slip . .
A feature 0 f th .
e Slgt'lIata wares and the AfrIcan
. .. ehi h p wares
t'l d
noted below (p. 59 ) IS . the uniformitv .] of their fmng, w c. th en al e
kiln
(at least on some senes . ) the use of a special . flue
.. system m 'd e'fy .
Among the coarser, wares certain peculianties. serve to I enti
Roman-period wares over large parts of the MedIterranean:
1. Ribbing or wheel-ridging of the body, particularly on. closed
ha d on cooking vessels. The latter (seen also on certam earl- Plote 7 Three pots (made In Egypt) shOWing typical Roman treatments; rilll>ng'sIiced'
s pesan Pl' . d handles. rouletting. Height 22.2. 20.9, 170 cm.
ier Egyptian and Syrian wares) seems to start on . a estiman an
Syrian (?) cooking wares as early as c. 150 BC, ~ut IS not common
there and in Egypt until the first century AD (fig. }}, pI. 7). In the
Aegean and central Mediterranean areas it becomes common dur- the surface was exposed to changes in the kiln abnosphere durins
ing the second and third centunes (fig. }1.5-7). firing. Sometimes, especially on thin-walled beakers and mugs, this
2. Indented bases with an inset 'button' at the centre (fig. 4 top right). was exploited to produce a two-tone effect (figs. 27.3, 28 right, and
pis. 24 top. 25 right).
This seems to be a manufacturing trick, either to ensure proper cen-
tring of the ball of raw clay on the wheel, or to form the closur~ of
an element thrown upside-down. It is seen on some Levanttne SIZES
wares of later Hellenistic times and is Widespread thereafter, occur-
ring from Sardinia to Palestine.
It was noted long ago that certain fine-ware shapes tend to recur in
}. Sliced handles formed of a strip with one edge sliced off straight more or less uniform sizes. This observation was reinforced by the
with a knife, instead of being tapered or rounded off in the normal discovery of tally-lists from the South Gaulish Si~ta ~uctiaa
manner. These are found occasionally in Cyprus as early as the first sites noting specific sizes. While these sizes cannot m practice
century Be and are a feature of Aegean cooking pots from the first been applied industry-wide and at all tt,mes, a s~ p
century AD. Thereafter they are common on closed vessels and cook- (c. 16-18 em diameter. for individual servmgs) and two SIZeS 01 CIpeft
ing pots in other regions (figs. 4.1. 31.5 and pI. 7). Such handles are cups and bowls (c. 11-14 em and 6-8 em) do recur among the
not generally noted on ancient wares of other periods. and Gaulish Sigillata wares of around the first century AD (fig.
EVidence for the production of cooking and domestic ware can and these are more or less adopted on the imitatims, IIUIde
also be gleaned from Jewish (Talmudic) texts. In particular. a type of ~ factories (as in fig. 20). A series of ~ pla~ ~
casserole with close-fitting lid (p.80) may be identified there. wtth different treatment of the foot, occurs m matc:hins
Some production faults can be ken on the pots illustrated here. A these do not exhibit standard dimensions. The smaller the
jug froUl ~ (pl.. 5 left) has a deep gash on the body which may lllaldled, at first, in the African series (fig. ~1-4), but
~v~~l inflicted Just prior to firing. Often, firing is uneven. result- dishes and plates are larger, reproducing the SIZeS of
~~ .. in colour from one part of a vessel to another. 1IleIaIware (fig. 22.7, 10), and cups are few.
-..-........'lS on the way that the vl!llllels Wl!I'e stacked, i.e., whether A COUnter tendency, with commercial advantages, nay
developed in Asia Minor in the first century. Some
 . g gest that some early fine wares from here were made
shipwrecks su d ed Sizes,
w hich could be more closely
dl packed for
in several gra
shippmg
. than the Italian vessels with their awkwar y proJec g
. tin AMPHORAE
fee~erent considerations applied in the case of shipping contain-
h
ers suc as amphorae . While these vessels were made 'ffi as]far
. as pos-
.th
sible to a uniform size within each workshop (a dl cu t aun WI
handm d rod cts) indications of the actual contents tended to be
a ep u ,
ched " bb . d
serat or pam. ted on after filling', most are m cryptic
'. a revJate
form (col. pI. I, pI. ]0 left). In any case, amphora sIZes vaned greatly (Received:] 2 maurai, 3/uuirianai, 1 lagoinos, ... kmilat, 3 Icnidia
from region to region, and from century to century - the caveat emp- ... dichora, 4(?) aminnaiai (i.e. Mauretanian and northern '
tor rule applied here. Adriatic (jars], 'wine-jar', Cretan and I<nidian, (local] double-
sizers, Campanian)

QUANTITIES
From a papyrus management account from the Fayum, Egypt,
Mediterranean sites tend to produce far more pottery finds than dated to the period AD 249-68, quoted by D. Rathbone in
those in, say, Britain. A Roman-period urban Mediterranean site Economic Rationalism and Rural Society in Third-Century .4> Egypt
may yield over a million sherds per hectare; a medium-sized exca- (Cambridge 1991: 304). He further comments,
vation can come up with ]00,000 or more during a season. The chief 'When ... received ... , they probably no longer contained wine
reason for this is the very Widespread use of clay amphorae as ship- from their areas of production, but had been reused:
ping containers. Shipwrecks have produced them by the thousand,
and on the average urban excavation site amphora sherds comprise
'T'he normal commercial container for liquid (and sometimes dry)
between about 50 and 85 per cent of all pottery found, while fine
tablewares constitute c. 10 per cent or less. Hence, while the finer 1 goods in Mediterranean lands was a two-hancl1ed jar of elon-
gated or rounded form, the amphora, IJ which also gave its name to a
wares may have been used in similar quantities, on Mediterranean
sites they are only about a third as common in relative terms as in ~ measure of capacity. If The shape, originally bullet-shaped
Britain. WIth small handles on the body - a type which Survived into Roman
times in the Punic domains of North Africa _ was adopted by the
Greeks around 700 Be; Greek versions have a neck, often lang,
8InJced by handles. Designed for stowing on board ships, they taper
to a projecting solid toe, which aided hancl1ing and reduced the
chanCe of breakage. Roman examples, identifiable from c. 250 IIC
ClIlWards, generally follow the Greek pattern. A yellowish or Iipt
bIown.~loration, at least on the surface, was generally prefemd in
Hellenistic and Roman times.
Meet complete surviving ancient amphorae come from ship-
, of which hundreds are now documented from the
. ." Another rich source is burials, ~ in
Jegions like Gaul, Britain and Nubia - regions where the can-
ol the amphorae, not being locally produced, might rate a
l~ and olive-oil were the chief contents; pidded
.--r) and Various dry goods were also shipped in them.
of Roman amphorae (like those of modem boIde.)
Q!rtain extent to identify the type of contents and
Origins. Most distinctive are the heavy globuJar
lroin southern Spain (pI. 8 right shows a late
 F1aure 8 Amphorae: Dressel type I
(scale I: 10); Graeco-/taloc (upper
(scale 1:5) (onglnal drawmg by port)
P Roberts).
-

I .

Plate 8 leti: Palestinian nng-handled amphora LR Amphora Carthage 5 = Benghazi LR 4

(e! Peacock c ass 46) 6th-7th century AD. RIght Span,,~ baggy Oil-amphora. Dnessel 23
(denvat"'" of Dnesse! 20). c late 3rd-4th century AD

and the tall, extremely heavy wine-amphorae of the first century Be

from Etruria and central Italy (fig. 8.1). Both of these, the products
of highly commercialised systems, normally have producers'
stamps in Latin on their handles." Stamps are also present on some
other Roman amphorae (see fig. 9 for a selection). Alternative means 1 o 0 I
of identification were stamps on the stoppers, although these are not o I I
\ -1... .1
often found (pI. 10 right) and markings inscribed or scrawled in
paint or ink directly onto the vessel, indicating contents, shipper,
recipient and so on (col. pI. I)."
Hundreds of different varieties of amphorae were produced in
Roman times, some of which are found throughout the Empire. The
shapes found in Rome were classified a century ago by Dressel,
whose typology is still much cited." A recent reclassification in
English is that of Peacock and Williams;" this adds the results of
petrographic analysis of the grits in the clay of the pots, permitting
more prerioe pin-pointing of their source areas in cases where
maken' stamps and documented kiIn-sites are lackms.
In the later Republican period (late third to mid-first century Be),
~_~_~ aeries of amphorae typified the products of
.................,g Italy. All to a certain extent follow the 'Greek'
~) trends oeen elsewhere in this period: long necks,
pointed lioeI, handies bearing , a pale surface colour
2
28
,
2

I .... :
 --

, ..

~
~~~
one<!
e I
~ oj
5 flight Romano-
over pemaps
Found at
~
r \ -~
I

Rathtor mall amphorae, lacking tamps. are common m many parts 10 ~ scaJe 1:4) (I) Marseille type CGUoose 2 about 2S IIC-AD
of ltal , and elsewhere in thto and to fourth centuries AD. RhinelancI. (2) Spcntleon FYn (~ see pi. 38 lejt
An arl hybnd eJ'SIOIl made in the Marseille repon IS seen in
fig. 10.1, a find from the Rhineland.
Some mall amphorae were used for other commodities. Tmy
carrot-Maped Is WIth small handle-nnp on the body. made in
Ilerytus (Beirut) and helE, are known to hoi contained dates
and other ueh . IOIIleWhoItllimilar on the land
of Upan ( Ital ) rna hoi ppmg alum."
C D much of to be uper-
by prod of red)
 - --
Figure II Egypt,an amphora. from excavat,ons at TaM. Ist2nd century AD (scale 1:6).
latter, the so-called spatheion ('sword-shaped [jar)': see pI. 10 left, and
pI. 38 left with detail fig. 10.2), which was cornmon m the fifth and
sixth centuries, may have contained wine or some other product.
These Late Roman shapes are scarcely ever stamped, but many bea~
scrawled ink markings, either indications of contents and so on, or
protective symbols such as crosses.
While the original contents of these vessels have almostalways
vanished, a resinous coating (generally seen as a blackish starn) may
survive on the inner surface, indicating that this was how seepage
was prevented. Such staining is cornmon on a long-lived series of
slender micaceous brown jars from Asia Minor, perhaps the 'Tmolan
Plate 10 Left amphora neck with ong,nal plaster seal ,n place. North Afncan carrot-
shaped (spathe,on) type (made ,n TunIS"'). Red paont ,nd'cat'on of capacrty (1) In Greek
letters (XMG). 5t1>---6th century (scale 2:5). RJght amphora stopper with rehef letters.
Northern Adnabc' Found ,n Malta. I st century Be or AD D'am. 82 cm.
34
,,
,, 2
,,,
I
,I
,, I
,
Figu.... 12 (I) Amphora stopperi and
1 (2) flagon stopper. with shp on upper
part. I st-2nd century AD (scale 1:3).
jars' of the papyri, and the late Gaza products, among others.
Discarded amphorae could be used in various ways, most com-
monly as makeshift coffins for burials (which they may help to
date).
The essential stoppers and seals of ancient amphorae are mostly
lost, except for a number of finds from shipwrecks and some from
desert sites. Materials used for these included plaster or stucco, vol-
canic ~ wood and cork, sometimes sealed in place with another
material. Impressed, relief or branded stamps are occasionally pre-
Il!ftt on them, giving producers' or shippers' names. A series of day
disb with rough relief lettering or patterns (pI. 10 right) appears to
IIIIIlth the Adriatic Italian amphora series of c. 100 IIC-AD 100, but
otherwise day lids are abnormal. Some late amphora seals bRr
0uiItian motifs: see the plaster example (pI. 10 left) on a late North
Afrbn product found in Egypt. String-holes on the latter show ~
the IIDpper Was removed. An earlier series of small vases with
IIlIrpIj splayed rims (fig. 12.1) may have had a similar functim-
in a plaster seal to serve as a bung, the VlIIe would
a-n WJenched out to break the seal, which explains why
.n lIUCh finds lack their rims.
day jars (Latin dolium, Greek pitJros) were _ _
bulk storage of liquids either on the ann
produced or at the retail end of the procelL
e am les from italy, produced in sec~ons or built up from coils,
l have a heavy wheel-turned nm, and often bear a large FINE WARES
:~e~'s stamp on the rim or shoulder similar to those found on
stamped bricks and mortana (d. pI. 32 bottom). Those from the east-
ern prO\inces illustrate traditional types and are rarely stamped.
They are generally little studied.

FORERUNNERS

The ~r. pottery of Rome. and of. its . (IAdIum.

~ Etrurta) In the later Republican period is best Ill!en pm aI.
larger Mediterranean picture. Fine black-gloes wares in tile e-k
tradItion were dominant after about 400 IC, and in I!tnIIia IIIIl
Campania one sees the end of the painted Greek VlIIe IndIIian.
Then, while these Greek-derived wares Were in dedine in IDIIIMm
- Greek colonial - Italy during the third century, pncti&:aIly
pearing there after the Second Punic War (218-2011C),_
appeared in the Roman core area, normally bearing ItlImpI
than painting. Some have glossy black-slip surfac:ea, quite tile
of those typifying the best Greek wares, and may have been find
the complicated process used for Greek painted VIISl!lI. A
process, analogous to that used for some later RmlIIl '
coated' wares, was probably used for the more mecUoc:n!
One workshop in the vicinity of Rome, known to sc:hoIaal
Iltelier des petites estampilles (pI. 11 top), produced Rome's
ware (c. JOO-250 Be): fine-ware bowls (poculIl) with
often apparently impressed from signet-rings. SimiIIr
made during the third century in other Italian c:erdreo
figures painted in white over the black surface. 'IWo
ations later, products from Rome are represented by
of bowls with gem-impressions of Hen:uIes (
left). Plainer black-slipped vessels made in ~
served as the normal tablewares for the growing dty..
Late derivatives of the third-<entury types, in
finished, purplish-red ware with very bud
produced in and around Naples. This ao-caDed
made in great numbers in the period c. ~ Ie:,
ClI\ Roman Republican sites all over the 1ftIlIIDl
Examples are much rarer in the East. thoush. fIIr
at Delos. Its stamps, which date to the IllCOftd
while bands hark back to earlier Greek modeII.
are lOme much finer wares from the lonDIr
hera1d the classic Anetine ware allalIlr
badIecl Campana B, made mostly after the
very delicate rouIetted bands and .....
 In the first century BC, a series of reduction-fired grey-bodied
wares appears in certain regions. These, requiring a simpler firing
process, form a prelude to the red-bodied wares of the Imperial
period, Best known and perhaps most widely distributed are the
soft, somewhat micaceous Campana C wares of Sicily, consisting
largely of flattish plates; their eastern counterparts are a series of
large mould-formed ~latters with a ~lish~ b!'tck slip .produced
somewhere in the regIOn of Ephesos m Asia Minor. Regional grey
wares of significance are those from Metapontum (southern Italy)
and Knidos, and an early series of Arezzo products ('Black
Arretine') is also present. At least some of these survive throughout
Augustan times. The unique grey-bodied vase in pL 12 seems to be
related to them,

Plate II Above: poeu/um (dnnkJng' bowl),

with multIple stamps, from the atelier des
peutes estampilles, e. 300-260 BC (scale
e. 2:3). Left 'Hercules' dISh, so named from
the signet ImpressIon at its centre,
e. 200 BC (scale e. 1:2).

confined between grooves - features shared with the later red

Arretine wares, This ware introduces the large flat-floored plate as a
common shape (compare the strangely similar Eastern Sigillata A
ware type noted below, p. 57), A black gloss remains normal on this
ware, though a whole or partial red firing marks off some of its local
com~tors. By 5,0 Be these wares were generally in decline; the ...... 12 A rare funerary
transl,tion to their red-gloss successors in the region remains vase (burial urn), in black
Am!tine (1) relief ware,
patchily understood, Some of the later examples acquire potters' with gold-leaf aclditions,
stamps,
c. 25-1 Be (scale 3: I0).
 Plate 13
Top: mouldmade wares, p. 67) may be observed (pI. 13 top). Some of these are signed
relief bowl. by with 'Roman' names. They seem too early (perhaps c. 150-7511C)" to
Popillus. Plain have directly influenced the classic relief wares described below.
ware without At Pergamon in Asia Minor the transition froDl earlier Gn!ek to
slip, partly
blackened
Roman wares is marked by a class of drinking vessels with applied
where exposed plaques, impressed from ~oulds, most commonly bearins ivy
dunng the finng. motifs (col. pI. II left) or erotic encounters. These pots are often &r.d
Made In Etruria, black or dark brown on the outside and a sigillata-red 01\ the inIide
c 150 Be (scale and around the base (implying firing in stacks). A parallel
c 1.2).
Boaam: silver shows wispy incised leaves and flowers enlivened with dola of
relief bowl. white paint (a late version of the Hellenistic 'West Slope' dec:onIift
found In style). The kilns for these two series, which more or 1ess IlpIft 1M
Bulgaria, first century Be, have been located at Pergamon-KesteL 'Their'
c 200-1 50 Be.
gloss appearance is not matched on the later Pergamon Ii&iIIaIa
Dlam. 15.0 em,
height 8.4 em. wares, which were presumably produced in other workshope.
imitations of the Pergamon shapes and applied omamenIs are
at Knidos, perhaps at Ephesos, and in SOUthelll Italy ISicily ill
region of the Straits of Messina.

RED-GLOSS SIGILLATA WARES

The relief-ware ('Megarian') bowls characteristic of the
Hellenistic East>' continued to be made there for a while under
Ro~ rule. The Ephesos region (Asia Minor) was a major centre for
~ In ~ ~ecades ~und 100 IIC, and may have enjoyed certain
trading pnvileges - Its products are extremely common in the
Roman mtrrpM of ~Ios and occur sporadically in Italy, while a few
even reached ~ Some of the Ephesian bowls are fired grey;
other ~~ versions are aeen at Metapontum and in the
erode Uburruan c1ass from Dalmatia. Such relief bowls were not
often ~y ~de in central and northern Italy, though some small
cla... In a thin uncoated fabric (matching that of the 'thin-walled'
4D
The typical Roman fine wares are those with a glouy red
a category known to modem scholars as terra sigilJtd8 or
ware, both somewhat unsatisfactory terms, but ones made
by long usage. . .
The chief categories of these are the vanous ItdIn
wares, those from South Gaul and Spain, and
Eastern Sigillata wares (see below, pp. 53-4) (The
Eastern Gaulish wares found in Britain
Mediterranean customers.) Any MediterraneIJI
liable to reveal examples of some or most of
(pI. 14), and local imitations of them are scarcer II-.
10cked sites of the northern Empire. A vivid example
tion of these wares is provided by a aateful of.
the south of France still not unpacked after ahippin&
ruins of Pompeii."
lttdimt lind western terril sigilllltll
AIound 40 Ie a fine-quality red ware with
began to be made at Arretium (modem
y, where some b1ack-gloss w had
This 'Am!tine' ware (llOIN! euIy
1) developed rapidly into one of the
Roman international culture of the

F,CU'" 13 iLlhan terra s,gdlata. early products with mult'ple stamps (scale 1:3).
(I) ~ate (Conspectus form 2) wrth stamps A TITII FIGVL.(,] (see pi 38 bottom).
(2) Base of platter. stamps LVMBR['c ]1 SEXTIO (see pl. 18.1).
within pottery-using circles, became enormously influential, espe-
cially in the newly acquired northern provinces of the expanding
Roman Empire, like Gaul. Vessels bearing relief decoration, pro-
duced within moulds (such as pI. 4), were soon added to the ware's
repertoire; these, along with some other features, betray Hellenistic
Greek influence. The delicacy of the decoration and sharp angles of
the developed forms ape contemporary metalware, especially silver
plate. In general, sigillata shapes tend not to repeat those of con-
temporary glassware and the 'thin-walled' wares noted below
(pp. 67--'70). Vessel-shapes are distinctive and standardised, and a
potter's name-stamp is almost always impressed on the floor (pI.
18.1, 3~) or sometimes, rather incongruously, appears as a feature
of the relief decoration (pI. 18.2).
Arretine ware is mentioned in ancient texts, and as a result of
disc.overies of the ~ebris of several workshops at Arezzo, mainly
dunng the later nmeteenth century, the term became - in the
predominan~lyno~ernE~peanscholarly literature - almost syn-
~nymous WIth Italian SlgiIlata wares in general. Attention at that
time was mainly focused on the relief-decorated wares, which
appealed to contemporary Victorian taste, to the extent that a good
ma~y COpies were produced, reconstituted from fragmentary
ano~t moul~s or v~ls (such as pI. 4). These, especially the com-
plete mo~ds, may display slight 'seams' where breaks occurred on
the oogmal elements. The works of the German scholars
Dra~endorff,~cke and Ox~ (see bibliography), which not only
deta~ the decorative styles but also list potters' stamps and furnish
~= typologychf the plainer wares, provide models for this par_
. approa. A good number of recent studies, main! b
Italians, follow the same lines, concentrating on particular JorJ-
shops or museum collections, though publication of the . I
Arezzo collections is still patchy. cruoa
42
Until recently, however, what was lacking in most of these
studies was an interest in the procedures of production and a full
appreciation of the geographical spread of sigillata production in
Italy at large. Broken moulds, decorative punches and the 0cca-
sional waster furnish almost all the published evidence from the
AJezzO potteries; for other aspects of production (see figs. 5--'7 Em
some possible solutions) we rely on the analogy of finds from Gaul
and the German frontier. The identification of other regional Italian
fabrics by, for example, chemical analysis has come into its own
rather recently, Haltern in Germany (one of the major 'old' sites Em
the ware) being again a key point of reference. However. the
regional production sites remain little studied." The known time-
span of the Italian wares has also been expanded by more c0ncen-
trated work on Mediterranean finds, which include many late types
not seen on the northern European sites.
Three regional patterns of production and consumption. rather
than one, now seem to emerge. The best known is that seen on the
northern and western market (between c. 25 Be and AD 25), supplied
from Arezzo and its subsidiaries at Pisa and, for a while, Lyon in
France. At a later date different sources, in and around the Po Valley,
marketed their wares to the north-east, towards the nortIlem
Adriatic area and Pannonia; the most common wares here are tile
relief-decorated cups of Sarius Surus and his associates (col. pL n
front right, and fig. 15-2), and vessels with applied motifI. In
third area, comprising the central and eastern MediterraneUl
lands in general, Italian Sigillata wares are not dominant until
reign of TIberius (AD 14-37), and survived in drc:u1ation
second century; finds of relief wares are rather few, pedlapI
per cent of the total. The products of various centres are hem
gled: Arezzo products in Augustan times but less 10 ~
Wares of Pozzuoli (Puteoli) and other centres in Camp8Ilia
dominant in the period AD 100SO), vessels from Rome itIIeIf
where and, at a late stage, the so-called Late Italiln
were made mostly in the Pisa region. , thin
South GauIish Sigillata," mostly of decorated lonnIliIIe
re.c:hed distant Mediterranean markets after c. AD 50'
are scan:e1y recorded outside SpUt Iftli
The new veseel-shapes of Italian 5igiData
.... Ieries of footed plates with abIolute1y Sat fIoon,
)Ial6Iect rima. Such plates regularly come in I.upr Iftli
with its own treabnent of the Coot: bl'OIId md
1aJpr ones (now usually termed 'platten"'), Iftli
beveUed lower part, on the smaller 'pJaW
cups (fig. 17), many with rim Iftli
thc.e of the plates, aecompUlY tt.e. n.o
Figure 14 OasSIC ltahan terra Slg' ata shapes ('Serv,ce II) (scale I 3). (I) Arretlne platter Plate 15 Arret,ne cup, beanng nouletlJng and applted 'spectacle spiral' ~
(Conspectus form IB) stamped CNATEI (pI. S) See also col pi II nght. (2) Arretlne plate (diam. 7.6 em) Stamp: AVCTVS (see pl. IB4).
(Conspectus .B), stamped VMBR (pI. IB.3) (3) Cup (Conspeaus 17), stamped CN ATEI, a
vanant of the common form 22 (compare fig. 174).

matching plates and cups, normally termed 'services' in the special- fi&ure 15 Arretine neltef wane (scale 1:3),
ist literature, typify the Augustan age. In basic terms, one series is
early and the other late, though there is a fair overlap, The later one, r ------------T-----------
~l~~.
,, ,,
displaying two rim mouldings separated by a hollow marked off by ., ,,.
,
I

,
---

fine grooves (see fig. 14, and variant pI. 15), was enormously influ- \ \1------------- -------------4
ential on Mediterranean products at large (pI. 14; d. figs. 3 and 29.1) , ,
- it can be considered the Roman shape par excellence. Its distinctive
features, and those of its successors, were copied (with more or less
~----'F==:-=; -
-._.

i>- }=~......
3 "..--'~ -- ----~
C...r-_.:..____ _ .:::J,

I
Plate 14 SIX cups of a typocaI Roman tern I1ata 4
From lett clockwise: 1taI.." tern IIata; SIgI. shape, In vanous wares (scale c. 1:6).

.....,= N......,. ttDm: ~

'Pontx:' ,.--- v_..__ 80 Eastern SigiIlata A ware: GaIIo-l!elgic letlll rubro;
<4>",", SigiIIata B2 ware,


- 2

3 - ---------

I-~
8

f1aure 16 Plate and dosh forms (scale I 3) CDnspectus forms 2 base as 12 or 18. 203
(platter) 21.204 (two) 21 (1ate).4 (late) 32 respecl/IIely See Indelc forswnps.

flaure 17 Oppos,lt page Cup and bowl fooiilS (sale 12)

Ccnspeaus forms 14 (large. smaIl). I ~ (large). 22, 32, 26. 27. 33 36.4, 23 2 (bottom /eli). 10
34 (bottom nght) respectively See Indelc for stMnps.
success) by potters as far afield as southern Egypt (see fig. 25. 1 ) and
the Crimea (fig. 21 bottolll and pI. 20 rlgllt - Italian-style potter's
stamp). The earlier senes, with elaborate profiling on the inside of
the rim (see fig. 17-2), had less influence, since it dates to before the
peak in Italian exports. It was more in vogue in the northern Roman
provinces.
Among the decorated vessels, the typical shape is a medium-
sized bowl set on a pedestal foot (now to be called a 'chalice', in
preference to the older term 'crater'). Early examples (see col. pI. II
top centre and fig. 15.1) have an elegant im'erted bell-shaped form,
i
but those from the reign of Tiberius and later (fig. 15.3-4) add a high
vertical wall above the decoration, which becomes more stereo-
typed. A large cylindrical 'beer-mug' shape with a handle at the side
(termed a 'modiolus', fig. 5 right) is another well-known type.
Closed shapes such as jugs are rather rare; the black burial urn
I
(pI. 12) is quite exceptional- it may have been a special commission.
Two centuries of collecting permit many of these regional prod-
ucts, as well as Arezzo finds, to be illustrated from the collections of
the British Museum. Attribution of the plain pieces is tentative,

Plate 17 'Late italian' Top LR.P1S

Plate 16 Top: South Gauhsh rehef-
decorated bowl by MONTANVS,
Dragendorff type 29, a type present on
Medrterranean sites, though this bowl was
found In London. Dlam. 26.0 em.
left Signature ('Philo' In Greek) on
bottom (e><tenor) of Gaulish-style bronze
bo", (see fig. 2) (scale c I: I).
(for L Ras,n,us P,sanuS. a very large
manufacturer), typICal coarse rouIettirc.
P,sa region 'Found In K61n - duboous!
Around AD 100 Length of stiIrr4> 2.4
Lefi: Crescent-shaped (Ulate) ~
SEX MP on Intenor of a decor iIIed
(see pi 19 5) Wodth of sliImp 2.0
~ for none ha
8laDlps, I'epft!8en . ve been chemicaUy tested Some
lent in the ~about a hundred different ~or 150
the.e_trueArretine(~1s 5, 17, 18, 19.1). Probably ~
IICIIdhem ItIly lind ~e the pattern of acquisitiOllll
PI'oducta. In COntrast th' ~ ensured a good mix of
, e pIeces of decorated v
 ,
.-
"
......

" ,
'. .~ -,
~::' '. ,
.-
,

..r:.
.. , . . , '.. r
"'.... ".

, , .. ' . . '.
'.

, 3

... "

, - ,,:. ~ ~~< "./'

,
I"
'

.<

3
, - '~""-~ / .r;Ji " / !!-,
,

~'-
,
-."r:, #'
,
r'
-
.--- , ~

, , -~- 5

4 6

Plato! 18 IWtan ler'uglllata S1aIT'P' (I) Early 1wo-hne example LVMBR[ICII) 50010
(names of pt'OIlnelQr and ...... ""'"""an), c 30-15 Be Length of stamps 1.7 cm.
(2) On reItef wore S1MT'4l '" re,.( panel on decorated Arre1Jne,ware chalice (fig ISJ) by
PComei.... The Ieners RNE form a monogram, Widlh of stamp 155 cm. (3) Normal
rectangular shlpe:VMBI\ (for Umbn(lus; d (I) above). Length of stamp 1.15 cm. .
(1) ~ Yir1On1, IIlden1ed ends: AVcrvs ('Increase' or 'profit' ~ a typical slave s
1Wne) c 108C-AQ 25l.ength of S1amp 12 cm (5) ZOEU (=20111), In three,lobed of
slamp Scale (. I I. (6) Fool'shaped (plonra pedis) form SEXM.CL A typical lale stamp,
23 cmW1tI1 ful111ne-port Roman Cl1lZen's name (as often, abbreviated). Length of
pt'OIlnelQr
S1amp
 .-- --
moulds for them are generally Arezzo finds, from old digs in the
original kiln-dumps. These show that diverse decorative styles --
existed side by side (for examples of this see pI. 19) Free-flowing
.....1 &Ill"' SiflltaA pIatler.found ,n Nubia (scale 1:4).
compositions, mainly with scenes from mythology, are outnum-
bered by decorative arrangements of stock motifs, normally
repeated to create a rough symmetry. Some later products show that
individual motifs were copied by moulding until they were almost mocUfied _(or insfanCe, potters' stamps appear in a more random
unrecognisable. A very debased Late Italian version is shown in fIIhian, IIld aJmo&t always in Greek (pI. left) These eastern
20

pI. 19-5.
wues. whole fabrics can be distinguished by an expert eye, show
After about AD 75, the main export centre was the region of Pisa, IOIIle peeuBar leatures. For instance, the treatment of feet and bases
where the so-called Late Italian Sigillata was produced until the tends to vary from ware to ware: low feet, often marked off by tiny
mid-second century. A debased mouldrnade relief ware was made groove IIld ridges, on Eastern Sigillata B, stepped insets on the
here, influenced by the technically superior South Gaulish deco- bottaDIoi Eutem Sigillata A, and so on (figs. 18-20).
rated wares of the period, such as pI. 16 top. Many of the figure- Five widely exported red-gloss wares embody the sigillata
motifs, which are repeated as appliques on the walls of the much trIditIaIl in the eutem half of the Roman Empire. Since only one of
more numerous plain-ware cups and dishes (pI. 39 bottom), derive
from earlier Arretine patterns of a couple of generations earlier.
A few large proprietors c?ntrolled this production, notably one
L. Rasm(lUs) Pls(anus); theIr stamps are either foot-shaped (pI. 17
top) or of crescent fmm (pI. .17./eft), as on contemporary stamped 1
bncks. indIVIdual artisans WIthin these undertakings are no longer
named.

Eastern and ,!,.iscellaneous sigillata wares

~e mam tradItions of sigillata wares in the East, those of Asia I
Minor and of Syria, are each represented in the British Museum by
a fau numb~r of vessels. Both began around 150 Be, much earlier
than the Italian wares; early specimens reflect different decorative
styles, and are not always red-surfaced. However, starting from
an;nmd 10 Be-AD 10, they widely imitate the Italian wares. The Asia
Minor workshops, some of which were Italian-owned, whole
~Iy accepted the new fashions for angular shapes with fine
mouldings, ~d the practice of adding makers' stamps. In Syria and
the more dIstant centres the new influences are somewhat

lI"wc..t
 -
these has a clearly documented source - the Pergamon area - they
are generally known by the anonymous tagsEastern Sigillata'A',
'B', 'C', etc. Here I use regional names - even if they are not stnctly
justified _ wherever possible, in order to overcome the different 2
usages of various authors. The true sources of the common Eastern I
Sigillata A ware (once wrongly named 'P~rgamene'),so~ewhere in
the Syrian region, remain elusive - the regIOn o~ e~stern CiliCia north
of Iskenderun is now mooted. Those of the B ware, In western
Turkey, should include Tralles (mentioned in a text of Pliny the
Elder), with Ephesos apparently its main export outlet. The
4
Pergamon regional series was made at Pergamon itself and on the 3
nearby coast at <;:andarit, which has served as a useful tag for the ........................
ware at large; along with its forerunners it comprises three or four
parallel or overlapping series. Cypriot Sigillata is so named because
of its prevalence in Cyprus, though some nearby regions have
yielded much of it. Pontic Sigillata here denotes one or more related
wares from the shores of the Black Sea, whether made in the Istanbul
region, the Crimea or elsewhere. The kiln-sites remain unclear.
These five wares, with their interlocking but distinct patterns of 5 8
I I ,
export, share some common traits. There is, for instance, the copy- tJ 1.. 1

ing of Italian Sigillata shapes and, less consistently, of their potters'

stamps - features mostly confined to the first century AD. Also com-
mon, though not on all wares, is the so-called 'double-dipping
streak' noted on p. 22 above (see pI. 6). Other features, such as
applied motifs on the rim, or barbotine ornaments, here copied from
the Italian 'thin-walled' wares, are much rarer. All five wares have
their own versions of the plate and matching cup with upright
moulded rim (E. Ettlinger et al., Conspectus forms 17-19, 22; pI. 14),
which are quintessentially Arretine. However, earlier and later ves-
sel-shapes in these wares owe less to the Italian potters; indeed, the
later ones are often closer to known metalware types. Some poor ,
COpies of Italian relief-decorated bowls are present in Eastern
Slglllata A, overlapping the end of a production of relief bowls of I

Hellenistic shape; better ones are seen rarely at Pergamon (compare
,, I

the vanous lead-glazed wares of Asia Minor, for example col. pI. V ,
right).
The Asi~ Mino,r wares of the early first century AD are the most
blatantly It.alian In treatment, with their small, neat stamps,
sharply articulated mouldings, and penchant (though not at
P~~amon) for rouletted ornament (fig. 19). Two of the Eastern
Sigillata B( 1) workshops seem to be Italian foundations: C. SENT -

flauro 10 Oppow pogo Eastern ~ A. ViI'1OUS leo If" (sale I 3)

 from 1W, RENI - probably from Pozzuoli. A few v 01
- ..l""'-- be r the (Latin) tamp ARRE/TINA proclaiming their
though not their actual place of origin!
In the faIhIcn
currenl III ia Minor, Latin and Greek were \lied .
The m on the other Eastern wares, more variable in form, .
-. I a1wa in Greek lettering. though the actual IIIIIlea b
(which ~y ~Iude thoee ~f slaves and freedman) . .
me Roman III ongm (e.g. Loukios for Lucill$). TheIe-eem
lamps are almosl aU more or less square or rectanpIar; the
,, ;>ed (plan/II Pftlis) type of stamp currenl on the later ltd.
al I rare, paraUeling a decline in the importation of the
m Is after c. AD 30 When the Ialler picked up again later
tury, the eastern fashion for poIIers' Slamps had all but
F l-shaped tamps, while they ocxur, are generaI1y 1IIID-
(th ugh for an early signed one, see pl. 20 rig/lt).
tern SlgiUata A ware (figs. 6, 18,20) illustrates three IIUCXIIIift
lis trends, in pari matched on the other wares. Until the time 01
3 Quist, I most characteristic shapes are platten with 1lUIOW, wq
thick which provide stability to the finished product. n- ill
put nurror metalware, in part late Italian black-glola ware typa
Groups of rouletted circles, often flanking a band with IIIIIIl ...
melle tamps, decorate their floors. During the first oentury AD, the
~21 flat-floored Italian shapes are in fashion; the bottoms of theIe
manly feature one or more 'steps'. In a fina1 phaae, the IhapeI
to copy fashions in metalware, with a series of f1at-baled cIiIh8
bowls supported on smaU ledges at the outer edge. CcJmpIIIIlIt
developments occur in the other eastern wares, thou8h CJIlly
the time of Augustus onwards. The properties of the YadauI
and local traditions, are seen in differing foot-pIOfiIa 'l1le
quIar ring-feet of Italian products are copied at PeIpmIlD
. minor regional wares; thoee of Eutem ~ A
limpler (replacing an earlier foot-treatment with a
IIIOUIding at the base), while on Eastern 5igiIIata B
- Iedp. moulded on the inner fact!, is substituted ( ...
heavy pedestal-like bales). The late
forms, relatively little copied ~
Pergamon (<;andarh) repertoire. The
and p- appearance to the ItaIiID
by extImle1y thick and -vy
ClIl their exteriors ( fig. ,.,). The
(ldhHepualon) often appear an
my decoration. .
Italian are the 1IUIJOI'
lnfl aJIl
an SiSUJataB
 Sigillata A (pI. 20 left), and one bowl-form of the latter has a stepped
pedestal base of 'B' type. Several features of Cypriot Sigillata (fig. 21
top) seem to copy the 'A' ware - the stepped bottoms of the latter are
rendered in the former by gouged-out hollows, often ringed by a
deep groove, and the small 'saucers' of the Cypriot ware should
derive from a flat-based 'A' ware type. However, a round-bow1ed
'crater'-form peculiar to Cypriot Sigillata is more likely to be
directly modelled on a late Hellenistic metalware shape; its larger
versions rest on three moulded 'knucklebone' feet."

AFRICAN RED SLIP WARE AND OTHER LATE

ROMAN FINE WARES

From the late first century AD onwards, the red-g1oss terra si8iJlata
wares were progressively replaced in the Mediterranean Romm
market by less sophisticated wares with smooth red-slipped lUI'-
faces. North African products, made at first in the Carthage JegioD.
and soon followed by the products of related factories in vuiDuI
parts of (modem) Tunisia and eastern Algeria (the ancient ~
of Africa, Byzacena and Numidia), were the leaders in this deve1op-
ment. These utilised clays of fairly simple composition (with
iron compounds, quartz sand and lime impurities, and little
- which did not require such closely controlled firing as thole
sigillata. In addition, the vessel-shapes produced (fig. 22 tGp),
those of the later Eastern Sigillata B ware (fig. 193, H), had
and were easy to stack for shipping. Hence they could
be produced and marketed far more cheaply than the ......
-- African cooking wares (such as fig. 32.4) led the . . OIl
scale, in the first century, and during the lIeCXlIld md thbd
the red-slipped tablewares gradually ousted their
almost everywhere. For a time in the fourth century they
tively without rivals, until deliberate imitatiCllll of
proliferate.
-- The exported versions of African Red Slip wee
major IJeJ'ies - Terra Sigillata africana ' K, 'e:
Italian terminology (the former '8'
Pn!nch IJeJ'ies noted below, p. 62). The 'K
ad with a continuous slip coating <generaDy
products from the region of Carthage of the
13
- cIoIe copies. The 'D' wares an! their
14
number of Cl!Iltn!s, mainly in northem
coven CIIl1y the interior end the
~~~ Sip u,* n1 tK* 1:4). (II) 'Tripolitanian I\ecl Slip' dish. the 'c' fabric from central
~ (13) Cypriat. ( 4) (,Will). aatlns. INlBtly abMIrbed into
 ---

-
fiaure 2J Dtsh.late Roman. South France. found .t Con)ux (sc.le 1:3).

the examples - if not all- appear to have been produced in moul.ds.

'This series was made from about AD 220 to 500; a large production
of mouldmade lamps parallels the later types.
The shapes of the early African ware comprise smallish dishes
and bowls with low feet. Some of these, such as fig. 22.1-3, were
modelled on common Italian and South Gaulish Sigillata forms, for
instance the wide-rimmed dish - Dragendorff type 36 - which here,
as in the Spanish Sigillata wares, retams its typical barbotine leaf
patterns. A distinctive version of the carinated bowl (Dragendorff
29), bearing rouletting only, was also created (fig. 224) Some mugs
and small jugs, often derived from the western 'thin-walled' wares
(see below, p. 68; pI. 24, fig. 28.1) accompany them. The later ver-
sions of all these became larger, losing their decoration and their
polished surfaces. Around AD 200 a new range of broader, flatter
dishes came in, some of which were highly polished. The 'C' wares
of the third century and later (here fig. 22.6) consciously ape metal- -.
ware in both thinness and their choice of shapes: large platters, a
1arge dish with straight sloping sides, small bowls with wide, flat
rims. Comparable shapes (mostly wide, flat-bottomed dishes) occur
in the parallel '0' ware (pl. 22 top, fig. 22.7, and regional variant fig. --
----
22.11), indicating a general shift in the function of these tablewares.
The 1arge dish/sma11 bowl contrast lasts until the demise of the ---
---
-

._- --

-

P10te 21 Afnean Red Slip ware fragments With applied or moulded relief omament
e. AD 360-400+ (scale e :3).

ware in the seventh century. During the fifth century higher

foot-rings - some thrown separately and a~ached - reappear, as
they do on metalware (fig. 229, to). Applied 0n:'~ents, some
arranged to form continuous scenes, appear on the C ware of the
third and fourth centuries (pI. 21); a few relief-decorated vessels
could even be cast-off metal originals. Stamped patterns, at first flo-
ral but increasingly figurative (pI. 22), were introduced on both
wares around AD 325 and lasted until the later sixth century. Crosses
and other Christian motifs adorn fifth- and sixth-century products.
Almost all these features and trends are matched on metalware of
the period.
As noted above, fine-ware imitations of these wares appear from
the fourth century onwards. Some of the closest copies are found in
Egypt, an area which had remained outside the mainstream of
Roman pottery fashions until this period. The large potteries of the
Aswan area produced and exported African-style stamped wares in
quantity; fifth-eentury African shapes persisted here until the eighth
century and later (fig. 25.5--<i, 8). Some of these bear painted texts
(pI. 2). Other, less delicate versions of these types were made else-
where in Egypt (figs. 22.12 and 24) and Cyprus (fig. 22.13). At the
other end of the Mediterranean, African and Phocean (see below)
shapes were grafted on to the pre-existing Hispanic sigillata
tradition, which enjoyed a final flourish in the late fourth and fifth
centuries. A South Gaulish counterpart of simpler manufacture is
illustrated in fig. 23.
More independent of the African wares, and in terms of fabric
and technique a continuation of the Pergamon/<;andarli terra sigil- __. _ Top' boWl from ......
Plate 22 Afncan Red SlIp ware: stamped un -,~,.. . of dish. c..AD lO
lata tradition, is the Phocean Red Slip ware from the west coast of
Turkey (Asia Minor), dating from c. 350-550 and continuing on a
iefl::=
Found in Egypt. e. AD '!OO. Dlam. 41.0 em. Bottom
(scale c. 2:5). Bottom nght plate fragment WIth chHho m ,
c..AD ~500 (_

reduced scale until after AD 650. Its initial fortunes seem to be linked
to the f~urth-eentu.ry ~wth of the eastern Roman capital of
Constantinople, which It was well placed to service; thereafter, it
served much of the eastern and central Mediterranean. This ware, of
which complete preserved examples are few - they do not occur in

62
 Plate 23 Lead-glazed dnnlmg-vessel.
ehr' h' burials - may have been part-fabricated in found at TharroS, Sardinia (tomb 33).
Byzanhne IS an s are a marked .'
thmrung au t a f the Sl'd es late 1st century AD (scale 1:3).
moulds'' thnotabl'de fea ture f
o'mt and 'scratched' outer sur aces. Th"
elf rims
towards e mlI ured -p (black
, 'd th
for inStance) on the OutSI e, e 0 y nl
are 0 ft en d ISCOO ' . thi d .
part of the vessel exposed during firing, smce s was one m
stacks.
Early examples of this ware imitate the shapes an d fl I
ora st~mps
of the African wares, but after about AD :l20 these develop differ-
ently. The classic vessel-shape is a deep dish or shallow bowl With
an upright projecting rim, which in the courseof a century develops
a more and more pronounced 0\ erhang (see fig. 3. 2 , 3 for the de~el
oped shape). The broad bands of rouletting often seen on the rims
and floors of these dishes are impressed With an Implement With
three or more rows of teeth - a technique copied on some other ASia
Minor wares of the period. The stamps, mostly a simple impressed
flcure 26 Lead-glazed Jar. 'Rome' senes.
outline have a 'childish' quality.
The Cypriot Sigillata series noted abo~e revived in ~e later sixth 2nd century AD (scale 1:3).
and seventh centuries, producing a series of dlshnchve flat-based
dishes, some bearing a simple stamped cross at the centre.
The varying fortunes of these later Roman stamped wares are
perhaps best seen on the Palestinian andNile Delta sites, .where no
local good-quality tablewares were available as competitors. The
Phocean ware, and some African products, are on occasion found as
far afield as western Britain (see sherds in fig. 3)
Derivatives of these wares, generally with painted decoration
(e.g. co\. p\. VII), continued to be made in Egypt, and especially in
Nubia, well into the Islamic period. Another painted red-slipped
ware of high quality comes from the Jerash region of Jordan during
the sixth and seventh centuries; it derives from a small regional class
of close imitations of the stamped African wares.

LEAD- AND ALKALINE-GLAZED WARES

- for standard types of vessels. Most appeK to
Lead-glazed pottery appears intermittently throughout the Roman stacks within closed containers known ......
Empire from the later first century IIC onwards. In the IItilIll which separated the vessels during firinI fM'/
Mediterranean world it was never common, though some large- floors and elsewhere. Remains of such
scale production, not relevant he,e, occurred in Pannonia and in the with various waste materials, are known m.n
far north of Italy in the fourth and ear\ier fifth centuries. Only in the Pep In Pamphylia (southern lUrkey). The
seventh and eighth centuries, in Constantinople, did lead-glazed pnenIly fairly thick. tending to crack UId to run.
tablewares begin to come into regular \lie. at the rim or base indicate which
The earliest such wares can be Ren one of a number of alter- for firing. Certain Asia Minor
natives to luxury silver and gold plate, WN.e forms and ornamen- (aee col. pI. V rigltt), UId
tation they mimic. like g\uIware, and the glua-frit vessels noted dilplaya thinner, even gJaze UId
below, they offer a new range of coIoun - green, yellow and brown how their gJaze material

64
 THIN-WALLED WARES
. the priming of the hollows in the moulds with white clay
senes IS e vessels were formed, to create highlIg h te d motifs on the
b e fore th . f dd d The plates and bowls in .terra sigilla~a were coml:'lemented, for table
finished, glazed, pots (col. pI. V); white barbohne moti s, a e Just use, by a class of drinking vessels m a very thin and light-weight
before the glazing, were an alternative. . . fabric, generally going under the name of 'thin-walled' wares (for
The early drinking vessels carried mouldmade relief decoration, examples, see fig. 27 and pI. 24) Many of these can be seen as con-
comparable in style and motifs to Italian Slgl1lata, and are repre- scious copies of vessels in other ma~~als: in particular silver plate
sented in the British Museum's collections by specimens from vari- and, at a later stage, glassware. Origmating probably in northern
ous places. Most were probably made in western Asia Minor; the Italy in the second century AD, they are typical of the new wave of
wares of the two important export centres at Perge and Mytilene, Italian 'Roman' tablewares of the period 100 BC to AD 100. The loca-
mentioned above, are almost certainly represented (col. pI. V right), tions of several of the workshopslfactories are known either from
along with others less well documented. In o~er regions we know literary sources or through excavation. These Italian products were
of a workshop located by excavation at Tarsus m the east, while cen-
iJnitated, especially during the first century AD, by a number of
tres in Italy and possibly south Gaul- mostly unlocated - also made
provincial workshops, in north-western Asia Minor, southern and
the standard types: a shallow cup ('scyphus') with spurred handles,
central Gaul, Spain (particularly Baetica in the south) and the Alpine
a deeper version of the same ('cantharus') and a chalice comparable
to Arretine ware products (col. pI. II top cmlre). These generally date regions, extending to the Danube."
While some of these are wheel-thrown in the normal way, others
from the first century AD - some of the latest dated examples are fur-
nished by Pompeii. A thin-walled class with small, applied, decora-
tive plaques appears to stem from southern Italy (col. pI. V left)
although the production centre is not known; from the same source
come small 'askos' jugs with relief vine or ivy motifs. Barbotine
ornament is used elsewhere; spine-like motifs and leaf patterns are
seen on a rather later (second- to early third-century) ware from the
region of Rome (fig. 26). A plain glazed vessel from the same source
is shown at pI. 23. Late Roman specimens from Mediterranean
countries are rare, though offshoots of the north Italian products of
this period are noted from Rome.
Mention must be made of a category of glazed vessels made in
glass hit (commonly but erroneously termed 'faience') current in
Egypt. This self-glazing material had been used there for millennia
to produce small moulded ornaments and figurines, but in
Hellenistic and early Roman times was put to use for vessels of
some size. The Roman series of the first to second centuries AD has
a brighter, more glassy glaze than before, resulting from the addi-
tion of lead oxide to the glaze mix (copying lead-glazed pottery).
The angular shapes of plain terra sigillata (fig. 14) are imitated here,
in a heavier, thicker form.'1
A similar alkaline glaze mix was favoured on the buff-ware pot-
tery of Mesopotamia in the Parthian period (roughly from the first
century Be to the third century AD). Though present on Roman sites
on t~e eastern frontier, such wares hardly ever reached the
MedIterranean. Their shapes are more Hellenistic than Roman; early
versIOns dating to the Seleucid period occur at Antioch. '-14 SIrdinia
"'*,.... Thin-waJled wares. Top: late Ist<entury Be types 60m Grave 53 (1I1iliII1ufIO"
IIldbowl (~saIe 1:5). 8oltOm.:Two-handled cup (dernied 6om'll1in-'Mlld'~
~) Scuthemon Gaul,sh 518t11ata) In slip<oated ware (red slip, rauIIla8d
France, 2nd century AD and later (scale 1:4).
66
 :::::: ;;;;;:;;;;;;;;;;;;r~r.::;p:::,-- --.-.

(and perhaps not just those bearing relief patterns) were made in c;-:
moulds like the decorated sigillata wares which they can resemble
in fabri~. Attempts to emulate the thinness of metalware and glass
,

led to the production in quite large numbers of very delicate and

fragile bowls and beakers. These were around a millimetre in thick-
ness (fig 27. 2, for instance) and It IS difficult to Imagme them bemg
made freehand on a potter's wheel m the numbers eVidently
demanded. The delicatelv formed handles, however, were generally
added freehand. as was the ornament. which was \ aried: rouletting
(fig. 27. ) and barbotine (fig. 27.10) were the commonest. B~rbo~e
2
and 'rough-cast' decoration (the scattenng or brushmg of fine gnts
on to the pot, to produce a rough surface) are particularly typical of
these wares. Some variants, such as roller-stamping, are also illus-
trated here. Like the vessels in more precious materials that they 4
copy, some (mainly the earlier types) have a plain body, with the
ornament restricted to the handles, which can be elaborately treated,
1
sometimes with decorative thumb-rests and spurs (fig. 27. ; d. col.
pI. V right). The bases are usually flat and without projecting feet, --- ........
thus avoiding problems if encasing moulds were used. Many ves-
sels were evidently fired in vertical stacks, leaving most of the outer
surfaces to be exposed to changing conditions in the kiln, which .
---:t=~7' . '

turned them grey. This two-eolour effect (grey and yellow /brown)
was sometimes exploited to provide a contrast with added ornament
on the wall, as on the Egyptian (Nubian) specimens, (col. pI. IV and
-- -- .. .



.. .... .,- ,


fig. 28 bottom).
The typical shapes are beakers and small drinking bowls without



. . .. .
o.

handles, as well as versions of these - cups - provided with two

handles (fig. 27). These latter could be cupped in the hand of a per-
son taking a meal or drinking while reclining on a couch according
to approved Mediterranean custom; the handles may have served
for setting the cup down on a side-table as much as for actual use for 7
drinking. The ornamented body-surface would provide a firm grip
for the user. Mugs with handles, more suited to users who were
seated or standing, are typical of one or two series only; their han-
dles are generally attached to the body (fig. 28.1). In Gallic and
northern ~ds, the shallow cups give way to mugs and larger
beakers which copy the decorative treatments of the Mediterranean
10
wares while pointing to different habits of use.
After C. AD 60-80, the Italian products were succeeded by two
series of thin-walled mugs hODi the eastern Aegean. with handles
set cl~ to the rim, and two-colour firing after the Italian manner.
~, ~th.a high ~ffset rim (see pI. ~51eft). was produced at Phocea
u:' Asia Minor, while the other (fig. z8.~ and pI. ~5 right), with a low
nm marked off by a smal1 external .lA_ (collarino in Italian par-
lance) comes from saDIE unlocated centre' - 01\ the '
coast of Thrace or

68
 - -- -

.-

. -
----

--
--
I iill'i'MI'IIIIIIIWIri', .'- , .

11111111/11/11111111111'
IlImm 111111 1/1fllft , J , ..... 25 Thin-walled mugs. late Ist to early 3nd century AD Lefl: .~' (I'M->
UlIlIUlmmnmnrm- . ' ~.llncian (I) 'a collanno' type. late version with white painted insaiplillIlll
-

.. -
-. ',' . ...._.'
..
-

'.
' ' -, ..
,.- ....-.. ,-
c - .- , . - . --I GNlktuluchi& ('Good Fortune'). Height 12.6 em, 8.9 em,
.... '.
) ..'

FiCU
re 18 Th,n-walled ware vanants/denvat,ves (scale 1:2). (3. 4) Nub,an denvatives.
. . . 2t ICridiIn, grey and coarse wares (scale 1:3).
around the Dardanelles. This appears sometimes to have been given
a salt-water wash before firing; when overfired it occasionally
appears salt-glazed. During the second and early third centuries
both series were exported en masse to the south and west, as far as
the northern Adriatic and Rome, in company with the Phocean
cooking wares. Some late examples of the second type bear white
painted ornaments or motto-like inscriptions in Greek (pi. 25 right);
the type and the tags are copied on slip-coated mugs made in
Athens after c. AD 200."
A separate, earlier tradition of thin-walled vessels, deriving from
Hellenistic black-gloss wares, is seen at Knidos in Asia Minor
('Knidian Grey Ware'). The typical shape here is an angular ('cari-
nated') bowl with thin handles, twisted upwards to form cross-bars
flanking the rim (fig. 2CJ.2). Such bowls, in IIOIN! cases stamped on
-
the inside in the Hellenistic manner, were exported a good deal in
Augustan times, and survived until about AD 100. --
 OTHER 'FINE' WARES

Two distinctive series of containers, found thinly scattered all over

the Mediterranean (and on occasions further afield), may have owed
their popularity to their contents, perhaps prized as medicine or
even aphrodisiacs. The mouldmade Knidian relief ware (pI. 26)
comes in a variety of closed shapes, ranging from ornamental jugs
to vessels in the form of anatorrucal parts (especially phallus-shaped

Platll 27 Top Connthoan relief-ware bowl

battle WJlh Amazons. 3rd century AD .
Dwn I I 9 em Right: small mouldmade
double flas~ Harpokrates and swags.
Egypban. probably Memphos ware. early
Roman' (scale 35).

Plate 26 Top: Mouldmade

table-amphora (o,nophorosj.
Kntdlan relief ware. front and
bad<. DIOOY"ac scenes (satyr'
and maenads). 2nd century
AD He,ght 21 3 em.
Left hollow handles of shallow
bowls (poteroe) foc ntual use
Mouldmade. wrth an,malhead
term<nals. Knod,an fabnc.
2nd century AD (scale c. I 2~
~ ~ongside these are versions of the JUP and
IIlOn1 animal-headed handles (normally metaJw-
y used for religiOUS rituals. In contrast to the
JIlIeantare orange or brown, with a lip of simiIIr
Carin ~ rec~lI bronze. They pan the period c
bow ~ relief ware, typically in the form 01 SIIlIII
the ::th mould-formed relief friezes (fig )0 and pi. %'/
. century; no lid are known, but
likeJy. The Dionysiac and other 011

72
 Figure 30 COflnthlan relief bowl (see
pi 27 lopl (scale IJ). COARSE WARES
--

BUFF WARES: CLOSED FORMS

poorly modelled, hardly relate closely to their presumed contents.
n Roman times, following earlier G k f hi
Some of the Knidian types are revived In orth Africa (in red-slip
ware) around AD 300.
I cream or buff tint were used widel~ ~ on, clays firing to.
serving vessels. The surface of these y or JUgs, flagons and Olher
Some other vessels U\ the form of figurines are known, mostly in . . was generally left .._'-'
early Roman times. A few from south RUSSIan sites seem derived
allowmg a certam amount of evaporation' wh the --
be retained (as in amphorae for shipmen't) aendeliberatr
. COIItmlI had
coaq cI to
from Hellenistic forebears. In Egypt the tradition of small mould-
made figured vases and figurines functioning as lamps continued
more vigorously than elsewhere. A double unguent-vase with fig-
ures (pI. 27 right) here illustrates a class of small vessels made
around 100 BC-AD 100 in the Memphis and Bubastis areas. Two
examples of head-shaped form, in a hard-fired red ware with a
'metallic' sJjp, from an unidentified source in the western or central
part of the Nile Delta (col. pI. VI), may be influenced by the Knidian
wares; they are matched by a class of second-century clay lamps.

74
Plate 29 PI..n krtchen-ware JUg, --
nbbed Found at Carthage. 4th-5th
century AD (scale c 1:3)

--

...
-- -
- - --
- -- -
resin or the like was applied to the inner surface. Some finer ver- - -- - - --
sions, for table use, were coated with a slip of similar colour. ----
Recurrent shapes include narrow-necked flagons, derived from the ---
-'~-~~
Hellenistic lagynos type and mostly with a globular or ovoid body
(pI. 28), which were popular in the first centuries Be and AD. -'0=-
Similar shapes are seen in the sigillata wares. . ---
-----
Ribbed bodies, handles of the 'sliced' variety, and indented bases - --
--
with central 'buttons' are recullent features of the later varieties, in
common with the cooking wares of the period (see fig. 4 and pI. 7;
5

also the late Carthage example, pI. 29).
Heavy-duty kitchen wares, principally basins and rough lids,
were also regularly made in uncoated buff wares.

COOKI C WARES 7

The buic cooking-ware shaps in Mediterranean lands in
Roman fur... were round-boltmned moking pots and casseroles II c:aaa. pols (sale: 1:4).
(pI. }O and fig. 31), IBIaIly with close to the rUn,
often wi~ proviIIm b normally of shallow conical
pe WIth a small made in both oxidised
 ----------- --
1

--- -
---

flcure 32 Coolang dishes, ltahan and denvat"",s (scale 1:4).

Plate 30 Cooking pots, Egyptian. Ist-3rd century AD (scale 1:4).

from the side were produced from the first century Be onwards. 111
(red, brown) and reduced (grey) fabrics. Like the Greek and other products include mugs in thin fabric (see above), flat-rimmed
Hellenistic types from which they derive, these are generally rather cooking dishes with a red wash modelled on the 'Pompeian Red'
thin-walled, allowing rapid transmission of heat, and are made of imports, and a large series of finely ribbed cooking Wan!I (deep
refractive (heat-resistant) fabrics, containing an admixture of sand pots, casseroles, jugs: see fig. 31.5-7). The latter (e.g. fig. 31.S) haft
or similar material. In this they contrast with the wares in the Celtic smaII handles of the 'sliced' variety noted above (p. 24). Such
tradition which were made in the northern Roman provinces. are common in the period c. AD 60-250 on sites around the AepID
To the basic 'Greek' types were added, from about the second
century Be, flat-bottomed pans suitable for baking (fig. 32). The best f\ure 33 Coolang pots, Levant (scale 1:3).
known of these bear a smooth 'non-stick' red coating on the interior,
and go by the name of 'Pompeian Red ware'. (The name describes
the colour..of the slip, matching the common red tint of wall plaster ----I
at ~ompeu; th~ fact that one series actually comes from the same
regIon IS fortuitous.) Such vessels, made in the Naples/Pompeii
area and elsewhere in south-eentral Italy, were widely exported --
after c. 100 Be. In Augustan times they reached both southern
En~d ~d Petra in Jordan (d. the Palestinian find, fig. 32.1). Some
matching lids are known. Along with these travelled deeper pans
~th ~ up~ght rim bearing a groove on the upper edge (named the
OlIo bifido type; see fig. 32.3) and shallow lids used as baking cov-
ers. The clay body of these contains dark volcanic grits, indicating
~ on the west coast of Italy (Campania, etc.). Some later
--
~ products.derive from these types (e.g. fig. 32 .4). =--
Rival centres In the East include Phocea on the west coast of
Turkey, where 'frying pans' with distinctive tubular handles rising I 3

78
and up the Adriatic, where th~y 'enerally supplanted the earlier
Italian exports. d bod' d kin
Rath hallower versions of the roun Ie coo g pot are
current e~ sthe Levant (includmg C) pru ) throughout Roman times
(fi ) These often of e treme thinn ,hare the nbbed treatment
an~ ~;Iiced' h:mdles. Casseroles Vlth cl sely fitted lids (actually i== = = = j .--------
'-. ". ".

Single vessel and then liced apart) are also common here. --- b~-----=-" --...
ma d easa d . f' I .
Th seem to reflect Jewish culinary nee , In terms 0 ntua punty,
.'.
.
-
'.. . . .

ese y later ha\'e been adopted by lui tians and others. The type _
t
b u~t . . Anth . ..
.'. :. . .--
ars by AD 100 and is current f r orne SIX centunes. 0 er


,

----I
::::: of casserole, flatrimmed, ha a flat 'saucepan' handle folded 2 .'
back in 'wishbone' shape.

MORTARIA

Buff. and cream-ware morlaria (bowIs for grinding and preparing

food, for kitchen use), often bearing impressed makers' stamps, ,,,
were made in Italy and some western centres in the first century AD, , I
, I
continuing a long tradition of the use of such vessels in the Greek
world. However, the type soon disappeared from most
Mediterranean kitchens, being replaced by similar vessels in stone
and marble. In the northern parts of the Empire, though, vessels
based on the early Imperial Italian model remained in use through-
out Roman times. The basic Italian type (first to second century: fi&ure 34 Mortana (scale 1:4). (I. 2) Italian and related. (I) Stamped SA1VRNINL
fig. 34.2) is larger and heavier than most of the imitations, perhaps c.AD I~. (2) No stamp. c. 2nd century AD. (3) Synan. stamped 'HertnclpnOuS'.1l
indicating the use of rough moulds to accommodate its unwieldy Greek. c. AD 25~ 350. Found at Kn,dos. ,n the 'Gymnasium' (Newton's l'lO'IOIIOIw. 1119).
projecting features (now an integral part of the vessel, and not
added, as formerly). Most examples bear stamps (e.g. pI. 32 bottom
left) similar to those found on Roman bricks and tiles, suggesting ...... 31 Small flanged bowl with gritted ,nterior as on mortana. The ..., ..., lin be
manufacture in the same workshops, with some techniques shared. mere decoration. North African. found at or near Carthage. Sth or eoriy 6Ih CIIlWYAD
A revival of the type, retaining the same shape but with stamps in Diam. 19.5 em.
Greek (fig. 34.3, pI. 32 bottom right: detail) is seen in northern Syria
in the later third and fourth centuries. The latter class, made in a dis-
tinctive dark brown ware containing black volcanic grits, had some
success on markets in the Levant, surviving in modified forms into
the seventh century. In North Africa a flanged version is found in
late ~man times; here pieces of imported lava millstones were
sometimes recycled to form the grits. Small non-functional versions
~ the type oc~ (pI. 31) after AD 400. Cooking pots, along with their
lids, were occasIonally (re)used in Roman times to serve as the urns
for clelllation burials." Sometimes, in Italy, special versions were
made for this specific purpCIlIe, the name of the deceased being
scratched or painted on them (see fig. 31.2 and pI. 32 top, an urn
frolll Puteoli (pozzuoli.

80

Plate 32 Top: cooklng vessel used as a cremation urn, with
scratched name 'SICABIA CHRESTE'. From the W. Temple
Colledlon, found at Puteoli (modem POl'UOII), 1st-2nd
century AD' Height 22.8 em, dl3m. (nm) 22.2-22.6 em.
8onom: stamped fragments of mortana (scale 1:3).
left ~alian (see fig 34.1), stamped SA1VRNINI.
RJght North Synan (see fig. 34.3), stamped (In Greek) by
Henmogenes, 3rd--4th century AD.
82
LOCAL EGYPTIAN WARES
Until the fourth century AD, Egypt, home to some 10 per of
the popuIation ('I.e., pottery-users) of the Roman Em . cent
something of a world apart, with its own local pue, : :
which tended to keep imports out of the hands~ sys
. 'f . , I average
peasant. Except m orelgn enc aves such as Alexandria regional
self-sufficiency was prevalent. In pottery terms this is sea: in a muJ.
tiplicity of 'local' wares, as yet hardly studied in depth, and in 011
or two dominant 'regional' wares such as that of Aswan, which w:
shipped down the Nile in quantity. Most of these wares are based 01\
Nile silt, which fires to a distinctive deep brown and regularly
contains gold mica sp~ks. The ~ore mixed clays of the Fayum
depreSSIOn can be reddiSh, and m the eastern Nile Delta a deep
pinkish-red tint is present (see the figure-vase, col. pl. V1). Outside
the Nile floodplain sandy buff fabrics are found, and in some plaa!s
mined clays of glacial origin produce smooth whitish wares (seen
both in Nubia - col. pI. IV - and in the Luxor area). The Nile silt is
frequently tempered with straw or lime to produce a more durable
ware, but some fine-textured wares were also made of it, especially
in the Delta region (e.g. the figure-vase, pI. 27 right).
Painted decoration often appears on these Egyptian wares, u it
had earlier. It was generally applied over a reddish or whitish slip,
masking the coarseness of the clay body. An 'early' Roman style
emphasises silhouette florals arranged in scrolls, sometimes U.
nating with bands of cross-hatching (col. pI. III). Old dillcoveriII
from Nubia suggest a date around the second century fur II*-
which seems matched stylistically on a series of thin-walled
in pale brown (Upper Egyptian?) ware (col. pI. Iv. see COIIIIIlIIIlII
above, p. 68). Considerably later, and found from Upper Egypt
to the Delta, is a class of quite heavy jars bearing ~.
in thick black and orange-red paint, often depleting II1imI1
human subjects in a rather crude, lively style. Where dallble,
seem to belong to the fifth and sixth centuries. Related
wares, with somewhat schematic and abstract tJeatmenlll,
~ Upper Egypt and Nubia at .a~ut this time. Some 01 01
influenced by foreign styles, exhibIt a ye~ow-creem ~
Aswan fabric. An example of this is a cunous type 01
(col. pI. VII) wi~ inset circular bowls, matching a rue
Slip type of the sixth to seventh centuri~~ lIUCh
rise in turn to a long tradition of Nubian CI)ristian peinIIlI
which are beyond the scope of this volume. .
The generally rough nature of Egyptian WAJ8
the typical amphora shapes, which are heavy. ~
provided with small straw-tempered hand\eS
manufacture (pI. 9 right) and poorly attached. A ~ommon. feature of VESSELS WITH
th maller flagons for domestic use IS a stramer mserted m the base
ofe~e neck _ presumably to exclude flies and so on. This could be
t ed an Egyptian peculiarity, and gIVes a foretaste of the very
SPECIAL FUNCTIONS
e~rately patterned strainers developed in medieval Islamic times.
Cooking pots bear some resemblance to contemporary Palestinian
types (fig. 33.3 and pI. 30 top). Favourable soil conditions have often
preserved for us the ropes which were attached to the ha.n dles of
these Egyptian household wares (pi 35); some o~er finds are
cocooned in protective wrappmg (pI. 34 bottom). Quite large num UNGUENTARIA
bers of complete vessels are preserved in Egypt in domestic rather
mall narrow-necked flasks to contain perfumed oils 01'
than funerary contexts, as a result of bemg sealed under collapsed
mud-brick walls and the like. In fact, m Egypt proper (as against S (hence the usual name unguentaria) were CODUl1Oll inur::"*'
century AD. They were frequently deposited in burials
Nubia) recorded pots of Roman date from burial contexts are few -
many may simply have been removed m antiquity for reuse.
to create sweet smells. Their contents are rarely ~~
stopper were of organic materials which have perished. The
version of the first century AD has an egg-shaped body with
neck; an earlier variety with a spindly toe was COIIIIIlOIl
Hellenistic world (late Italian versions are 5n until about AD
 Classic 'Roman' versions of both of these shapes, in a smooth tan-
brown ware WIth a rather glossy black or brown slip, were mad .
Italy, perhaps specificall~ for the pe~fume industry of Capua.
enjoyed a certam vogue m the provmces (from Britain to Phoenicia
n:=
and Egypt), presumably as representatives of a Roman ~
and were much imitated (see pI. 33, from Byblos). By AD 100 theY
had been largely ousted by versions in glass, though in some
provinces, such as Cyprus, clay copies of successive gIasswue
shapes pers~st. Typical .specimens are ~hown in 'pI. 34, including 0Ile
with its ancIent wrappmg. A luxury silver versIon is also present in
the collections of the British Museum."

RITUAL VESSELS AND RELIGIOUS SOUVENIRS

Small vessels for domestic ritual use are common pottery finds in
Italy and the West, at least from the first and second centuries. The
most popular type is the so-called lazza, a footed bowl with frilled
clay bands applied to the exterior, probably used for the burning 01
incense. The type has also been noted on some eastern sites, such.

Place 34 Unguentana. Top centre and nght. standord (Italian) ~ Bottom left; wrapped
EgypIJan versK>n. Top left; late Asia Mnor vanant Hei&ht 20.8 em. Mostly Ist century AD;
the two small ones at Iowet- nght should date from the time of Augustus. .... 3$ Incense burner
(WIlh rope attachment).
Romano.Egyptian. found at
Hawara. Heoght 18.5 cm.
86

Plate 36 {)lnstJaI' ampullae
(small flasks for sanctJfied ad
or water). Right from the
51 MeNs shnne near
~ MotJf shoWS
51 MeNs. with cNneIs at hIS
feel [)oam. c. 105 em.
8e/ow left EpheSOS r<g!OO,
Tur1cer (ross. [)am. 29 em
8e/ow ngtrr. southern Turl<ey,
source ..,mown, NalMlY
scenes. [)am, 15,0 em,
88
Athens). Squ~re ~ays wi~ homed sides, set on pedestals,
similar function m Egyptian homes (see pl. 35). Votive ~ I
Greek tradition a~ gene~ally scarce, though the jugs and sa mlhe
like vessels associated With Roman cults are occasionall :
the East (e.g. pI. 26 bottom, Knidian). Vessels made 1ocally~"':
cults, bearing applied snakes and other symbols are 8CIIIletlms
found in Mediterranean contexts. '
In later, Christian times small clay vessels were produced in ar
near several holy places to serve as portable containers for IIIldi&ed
liquids, Most of these take the form of 'pilgrim-flasb' _
with a small spout and handles attached to a f1at1ided ~
made in moulds, bearing suitable patterns in relief. Such
dating between c. AD 400 and 650, come mostly from the
Mediterranean (for examples see pI. 36). A number bear C1riIIiaD
inscnptions; best known are those offering a blessing from I
Egyptian saint, Menas, whose shrine lay not far diIbInI
Alexandria.
Toys
Positive identification of toys is difficult, but llOIIle
miniature vessels found in children's graves may be
rather than as miniature cult vessels (see above). The IIIIIIt
tive are thin-walled vessels in the form of aniJnaIs -
instance - bearing applied glass 'jewels' (pI. 37 "ft); they
the second century and come from the . of
VARIOUS PRODUCTS
Some plain buff-ware 'tubs' with partially c:IoIed
Crete, Asia Minor and elsewhere, may have .....
toilets; others, common on Aegean sites, bem'inI
on their inner surfaces and fitted with di8c .
ceramic beehives. A kind of cy1indril:al jar in
culinary de1icacy) were raised has also been

ApPENDIX
A BRIEF HISTORY OF THE
BRITISH MUSEUM'S COLLECTION
The British Museum's collections of Roman pottery have long
1 been known to scholars through Walters' classic Catalogue of
1908. More recently, C. Johns' brief guide (Arretine and Samian
Pottery, London 1971) has served as a more popular introduction to
the subject. Both were mainly focused on the wares which can be
found in Roman Britain, whether made there or (in the case of the
well-known red samian ware) in France (Gaul). However, the
origins of the latter ware in Italy (at Arezzo and elsewhere) are duly
acknowledged in both works.
Plate 37 Early aceesSlon~ Sloane ColleclJon (the
angInal 1756 goft).le(t small pIg vase wllh glass Inlays.
Ita"an (Rome'), 2nd century AD. Length lOS em,
he,gttt 6.0 em. Right garum amphora. SpanIsh type
(Dressel 12 Beltran IIVIV; Mau-Schone XIV).
ISl century AD
The 24<H>dd years of the Museum's existence h- '
II ' (I )
the Sloane C0 ection p, 37 ," have spanned the , OD'llsori_
. --.....- Dl
Grand Tour collectors, before the age of modem ~ of the
(pl. 38), as well as that of the peak of the British Emp' ~ Iawa
an administrator or military man amassed his priv:' IIIIIly
mementoes (pI. 39 top). In due course many of these~
passed to the British Museum. Others (mostly the prizes of
aune to stock ~y of our provincial museums." ~
Museum, along WIth others, was adding to its collections by
lOring its own excavations abroad, or (when . .::::
ted) by purchase on the open market. It has not always had the pick
of what was available - for instance, personal rivalries led to ...
bulk of Sir Flinders Petrie's finds from Egypt going to Uniwalty
College, London, rather than to the British Museum.
The main Mediterranean sources of Roman pottery in ...
Museum are - from the early years of the collections - Italy. ....
with some items from Greece and elsewhere (e.g. pI. )II ;." . .
bottom right), then - from the 1830S onwards - Egypt, jcined
colonial times by Cyprus (e.g. pI. 39 top and fig. 21 lop) IIId a
amount from Palestine (figs. 32.1, 3 and 33-1) and Malta (sa.
Excavations at Tharros in Sardinia (pI. 23, 24 top), at ICnidaI
Ephesos (figs. 3.2-3, 29.1 and 34.3) in western Turkey, at
locations in Egypt and Nubia (e.g. col. pis. I, ill-Iv, VI-VII. pl.,
and fig. 25), and elsewhere have added their tally -1lIlIDlIIlJ
coJonial times, on the basis of a share-out between the 1lIlIt
IIId the excavators. In recent years the volume of ~
1Iackened, but host countries have often generously ......
pie finds, occasionally of some significance, to the
VItoIs. New, well-documented material has arrived. far
&om Sabratha and Benghazi in Libya (pI. 40 and sa. ,s.
in Jordan and Hermopolis in Egypt. The occasionIl
hu also accrued to the collections." As a result ol_
IIldent trading in such wares, the MIllleUDl's c:ol1edIaIII
nearly all the major wares of the period." Molt ol tbe
play are complete vessels, but a considerable coIIediaD
CIherds) is also held by the relevant
Among older acquisitions from private ~
8nd-epots are recorded, presumably invented by
CIDIIectOrs. Well-known Roman sites such II
1Ileet-names supplied) were favoured ,~,
places u KOIn acted as similar DJI8lIf:lI-
'~' appearance or dubious date may be
'Wmketer' cup (fig. 27.10), a Late Italian- '
Afrbn dish. both from 'Kaln' (pI. 11 "" ...
uly early terra sigillata stamp' fraIIl


Plate 38 b'pan n "Iv 19th century. Top: Burgoyne

Co lect on In-he'd 1I die .,f a bowl. Athenian.
4tr century AD Left Land Elg s CollecliOn: amphora,
North AllCan spathe on type 5th to early 6th century
AD He ght 90 em (See also fIg. 10.2 for detail.)
Bottom Towneley Collection early Arretlne-ware plate.
stamped ATiTI FIGV(L) four t,mes (see also fig. 13.1 ).The
whIte T marl<.Ing stands for'Towneley', and the printed
label refers to Wa~ers' pubhshed Catalogue.
About 30-20 BC (scale c 2:3. stamps 0.8 x 0.65 em).

Plate 39 Oppos,te page: Spo,ls of Empire and war.

Top: flagon, collected by (Earl) Krtchener ,n Cyprus.
Eastern S'g,lIata A ware. c. AD I-50. Height 15.8 em.
Bottom: plate. 'from the Semme', acquIred dunng the First
World War. An ~han Srgrllata plate of late type. shOWing
stamp (of planta pedIS form, here Q.TERV), rouletting. and
apphed ornaments on the outsIde. c. AD 30-70.
Dram. of nm 15.2 em

92
 )' Among alters I \en pI <'tltalian igillata stated to Endnotes
(Lond:, finds (his L H~) m re dubIous, others (e.g. fig. 154
be B"6 ) late products and l uld '" 11 han' arrived as isolated
un are 04 . uch pI
= L 1rtsI after n t L..
uC approac
hedth
WI an open
~,sinee properl e ca aled m t n I also mcludes the occasional
freak i1nport. Introduction
1 See pp. ~4 for more details. Roman and related . .
fact divided among five of the BM's Departments of=-:
Greek and Roman, Prehistoric and Romano-British, Egyptiml,
Western Asiatic, and Medieval and Later.
2 H.B. Walters, Catalogue of the Roman Pottery in the lJrpirrfJnmlI tf
Anlrqultlt5, British Museum (London 1908)
3 Journal of Roman Pottery Studies (1C)8S--, annually, iIIued bJ
Oxbow Books, Oxford). A new comprehensive survey 01 ~
British wares and other British finds is now available: P. 'I)m,
Roman Pottery in Britain (London 1996)
4 Rouletting is the term used for fine notches produced either bJ.
rotating cogged wheel or by holding a vibrating blade to the ......
of the pot.
S Barbotine decoration consists of patterns created by ~
liquid clay through a nozzle, in the manner of icing a cab.
6 Two contrasting approaches may be seen in R.J.
Roman Pottery (London 1955), and D.P.S. peaeoc:k, PotIftJ
Roman World: An Ethnoarchaeological Approach (Lorldoo 1_
7 See, for instance, D. WU!iams, Greek Vases {LondIlft
further aspects, see T. Rasmussen and N. Spivey (eels).
Greek Vases (Cambridge 1991 ).
8 See for instance S. Walker and A. Burnett. T1It I,." tf
PIaU! 40 Finds from (London 1981), and S. Walker, Roman Art (LoIldon 1991)-
modem excavations.
Mug/Jug of Athenian type 9 Another BM vessel from Egypt, a late amphora (EA
from Anglo-libyan dig al covered with Coptic texts.
BenghazI. gift of the
libyan government 10 I here exclude mention of commercia1 ~
c. AD 350--"150 traded on account of their contents - again. lIIlIIJe typII
Hetght 19.5 em lady widespread.
U Pliny, Natural History xxxv.12(46).dio. Molt cl
him remain uninvestigaled.
U H. Coclde, Journal of Roman Studits 71 (1981>''''''''

94
 1992)' This lies in a rural area between Arezzo and Chiusi.
Amphorae 'Z1 La Graufesenque and Banassac products.
. La":ft borrowing of the Greek term amphoreus.
13 Theword ISa .... '
L_ t under 26 litres. Most actual Roman vessels 28 Actually found in London.
14 One amprw'M = Jus .
fall well short of this capaoty. 29 The cullent terminology in various languages is PIe8elllled
.' f these and indications of their contents, see most Ettlinger et al., Conspectus, pp. 45 ff . 11\ B.
pOker A~cient ShIpwrecks of the Mediterranean and the
15 For lisAtiI\gSJ 30 A small version of this is seen in fig. 21.2. top.
recently ., ar , )
Roman Provinces, BAR 5-580 (Oxford 1992 .
Roman amphorae, see M.H. Callender, Roman
16 For stamPs on . McC TheR Lttul- and alkaline-glazed wares
Amphorae (London 1C}65); also E.L. Will, In A.M. ann, oman
Port and Fishery of COSII (Princeton 1987), 17Q-220. 31 The 'Aswan'-ware pot (fig. 25.1) apparently copies IUCh I
17 So good examples of these have been published from ware shape.
Masa:':~ Israel: see H.M. Cotton and J. Geiger, Masada n, The Latin
and Greek Documents ijenasa1em 1989)
Thin-walled wares, etc.
18 H. Dressel, Corpus inscriptionum latinarum XV.ii.1 (Berlin 1899)
32 Provincial products are here represented by pI. 24 bottom 11ft,
19 D.P.S. Peacock and D.F. Wtlliams, Amphorae and the Roman the Egyptian derivatives, fig. 28 bottom. Examples &om
Economy (London 1986) (Sudinia) in the British Museum's collections (including pi. 2411W
are now published in some detail in R.D. Barnett and C.
20 See SA Castle, Britannia 9 (1978), )8)-92 .
T1Imus: A Catalogue of Material in the British MllSellm (London
21 Berytus type: Peacock and Wtlliams 1986, class 12. Lip~ri typ.e: Compare (for these) M. Pinna, Studi Sardi 26 (1981-5), :l)9-)llIL
Peacock and Williams 1986, class 1) (see also RJ.A. Wilson In
33 A late version of these, with ribbed and slashed
Archaeological RqIorts 1995-96, 82, fig. 15 for new kiln finds). A few
examples of both of these types reached Britain.
appears in pI. 40.
3t For these, see I. Nicolaou, Bulletin de ,.",....
(1989), 301-18.
Fine wares
22 See P. Bernardini, MIIStO Nazionale Romano: I.e ceramiche V.1
(Rome 1986). CcGrse wares
35 liipolitanian, Sardinian and Cypriot (1) veniaN
23 These mimic decorated metalware: see for example the silver
bowl, pI. I) bottom. Most later Roman relief wares do the same. Ii Glass urns were, however, in more CQDUIlOII
24. The date of these has been a good deal diBc:us&ed; for the earlier rt Inv. 188S.1crl1.1, found at Bordighera (on
evidence,. - M.T. Marabini Moevs, in Memoirs of the American Gallery).
~ m Rome XXXIV (Rome 1!Jl1o), 161-227, with 2) plates fol-
lowmg. Examples are pn!IeIlt on the s.nt Jordi I shipwreck
(Majorca), which is dated c. lCJO-9) IIC. See Parker 1992. Further
comments are provided by 5.L Roboff, in loun,,11 of Roman
~ 9 (191}6), )16-20.
25 D. Atkinson, lDUnIIl If__ . " . 4 (1914), 2~
26 For a recent lIbId.y of cam, whidl produced other fine lID illustrate IIQIIle of theIe in I
w_ aIIo, lee G. " .... 11 /t1rrw iii Umbricio Corda (Florence
96
 . gift to Princess Michael of Kent.
40 See fig. p, a Jordanum
Turkey and Spain are under-represented; Select bibliography
41 In tenns of sources, French colonial areas of North Africa and
wares from the f?rmFer ch collections _ are here mostly illustrated
Syria - common In ren
by examples found elsewhere.

Gerrm1 works
M. Beltran Uoris, Cerdmica romalllJ. TipoiogiD y chrsi/ialdDn (2 va1L, ~
19']8). Reproduces drawings from classic works. UIefuJ ~
but many authors' names are misprinted.
R.J. Qlarleston, Ror,um Pottery (London 1955).
(Errcidopalia dell'arle antiCII,) AtJante delle forme cmnnidIe ro:: I k I, D
1981,1985/6).
lid em.riae RorMnae Fautorum Acta 1- (various imprinla, &wa
Specialist joumal on Roman pottery studies; Iarge1y CllIIiInIa
P.1Arique and ].-P. Morel (eds) Cbamiques hellinistittua d . . . . 1-
~ 1980, 1987, and forthcoming). Surveys of VIII_
Britisla Museum works
ox Dalton, Catalogue of Early Christian Antiquities in tIw BriIiIt
(London 1901).
C. JoIms, Armine and Samian Pol/ery (London 1971). Short pidI.
KB. Walters, Catalogue of the Roman PoI/try in the /JqIlItIJIkIIII If
British Museum (London 1908). The fundamental cata1ap&
S. w.Drer, RorMn Art (London 1991). A general overview.

Am"IIorIIe
D.P.s. Peacock and D.F. Williams, Amphorae arultlw Jlo:::8.
brtrorlUdory Guide (London 1986).
No ScI.umo and P. Sibella, Arrrphom. QH""kJlt III . . . . ,
Aix-el-Provence 1991; and ed. 1994).

Rome, Rqubliam fine wares

P.1Ieauudini. M_ Naziorulk R/nrMno: Le ali _ V.
_Halt.,. tIIrl Tevere (Rome 1986). Finds fleD'"
INIjor categories, with bibliography.

IWleriIIie IllrIres in tM &st

J 'FIne _ in the IfeI)eniIIk ~
(ella) LooIciJIg lit Gmt VIIIII' f
~ Perpmene cup with'--
~ (eel), Owicwl Art
1995), 22)-", pllr. 72-), Ilaodc
81"
Tma sigillata Pompeii, etc.
E. Ettlinger et al., Conspectus Jo"""rum ttrrat <igillatat italico modo confectae A. Carandini (ed.), L'instrumentum domesticu", d' E
. , . .I Q I rcolano e P"",-,
(Materwlim =ur romzsch.gtrmiJnL<chm Ktram.k 10) (Frankfurt and Bonn pnma etu .mpena e, uaderni di cultura mat . Ie -"T" IIrIII
trrII 1 <Rmne 19'77).
1')90). .
A. Oxe and H. Comfort, Corpus UJsorum arret",orum (Bonn 1968) Listings Late Roman wares
of potters' stamps on Italian Slgillata J.W. Hayes, lAte Roman Pottery (London 1972) th
C. Bemont and J.P, Jacob (eds), IA t=r IgJllee gallo-romaine (Documents P. Reynolds, Trade in tlw Western Mediterranea~ ~ SU/111Iement (1980).
d'Arclriologie Franfl"st' 6) (Paris 1986) A 'urvey of production centres Evidence, BAR 5-604 (Oxford 1995). ,400-700: T1It em-
in Gaul.
'Toy' pigs
Terra sigillata: relief decorated ves els and moulds G. Messineo, 'Puerilia crepitacula?', Rivista di Studi 11 '.
A. Vannini, Museo NlU.ionalt Romano: I.e ceramiche Y.2, Matrici di ceramica published 1994), 119-32. ompnm; S {tCJ9l'1,
amina decorata (Rome 1988)
C. Troso, II ceramista amino Publius CornelIus IA produ=ione decorata a rilievo Finds of Mediterranean wares from sites in Britain
(Pubblicazioni della Facolta di Lettere e Filosofia deU'Universita di P.'!Yers, Roman Pottery in Britain (London 1""')
-nu, 8<-10
J 5, 111-12, 121'1,
Pavia 63) (Florence 1991). Decorated products of one large workshop 152-a (amphorae, fine wares, mortaria).
(listing material in Arezw). P. ~ur, 'Roman Amphorae from Canterbury', Britannill 17 (lgl16),..,.
M. Medri, Ttrra sigillata tardo italica decorata (Rome 1992), and C. Rossetti (WIth references died).
Tella, IA ttrra sigillata tardo-italica decorata del Musto Nazionale Romano D. Wtlliams and C. Carreras, 'North African amphorae in Jlomm'"
(Rome 1996). Two studies of the decorated 'Late Italian' class, offering a re-appraisal', Britannia 26 (1995), 231-52.
rival classifications.
GRzetteer ofshipwreck finds
Fakes ofsame A.J. Parker, Ancient Shipwrecks of tlw Mediterranean and the Ron_
F.P. Porten Palange, 'FaIschungen in der arretinischen Reliefkeramik', BAR 5-580 (Oxford 1992).
Archiiologisches Korrespondenzblatt 19.1 (1989),91-<), pis. 2<>-3.
'Iransition to Middle Ages
Eastern Sigillata B ware versions of the above L Paroli (ed.), La ceramica invetriata tardoantica e 1Il~
(FIorellce 1992).
III'"
S. ZabehlickyScheffenegger, 'Subsidiary factories of Italian Sigillata
potters: the Ephesian evidence', in H. Koester (ed.), Ephesos, Metropolis
of Asia: An Interdisciplinary Approoch, Harvard Theological Studies 41 Clay analyses and identification of sources
(Valley Forge, PA 1995) 217-28. Summarises longer reports in German G. Olase (ed.), Ceramica romana e arclreometria: 10 stIdD ditgIi .....
by an Ausman team. conference papers (Montegufoni 1993) = QutImti""
ArchlOlogia, Universita di SimJI 37 (Florence 1994/95).
Lead-g~ wares (and associated kiln fittings)
A. Hochuli-Gysel, Kkinasiatische glasime Reliej/cmlmik (SO v. Chr. bis sOn.
Chr.) und ihrr oberitJllischen Nachahmungen (Acta Bemensia W) (Bern
1977).Identifications now in need of revision
N'Alii<.
. Die Ker:amik aus den Slidtlw"nen _ P,..,.,. lstanbuler Mitteilungen,
finds. ri:
Beiheft (T,;I.;ft- .,,-,
M-.u'f't~. 1995), esp. pp. xix-xx, 1-s1l, pis. 1-11 for perge
(l!xcavation site reports are extensively cited in the
Hochuli-G ~ finds,are lIliIl not lonna1ly published; most of
~ yseI s 'Smyrna category (see the work died above) should
COOle om this soun:e.

100
 Index of items illustrated
INVENTORY NUMBERS

British Museum Departmental codes: EA indicates

Antiquities, MLA Medieval and Later Antiquities, PRB
and Romano-British, WA Western Asiatic. Items withoutaJde
are in Greek and Roman Antiquities. Objects currently llIl
are marked with an asterisk. In recent years, when 11\ ab;Ict
been transferred, a new number indicating the trana6!r_
been assigned by the receiving Department. Old
Departmental numbers are also cited for Egyptian and
Asiatic objects; 'K', 'L', 'M', etc. refer to the series in the older
itive Museum publications (e.g. Walters' Catl/logut), oltIn
the specialist literature. For Dalton, see bibliography.
appearing in parentheses indicate known find-tlpots.

COLOUR PLATES

(" iniIicates that the object is on display in the WoHson c.u.,

.. in the new Hellenistic Gallery)

I EA54362

o 1902.10-12.2 [L 35); 1814.7-4.1554 [L 1)6]; 1~

~.)28 1M 1); 1866.6-8.2 [L 123) (= WaIIrI!r, fit.,..,

m "BA 35993 (1868.11-2.113)

IV Left: EA 65437 [195.5-20.2). Cmtrr. GR Ip.UHUo ...,
RipI: EA 5987I [1930.7-14-5)
Left: .1888.11-15.8 (K 6 - Leventis GIIIery). Jill':
)6J

Uft: EA 37540- Rigltt: EA 26lI2O


PLATES
1 Top: 1868.1-10.756. Lift: 19Bo. 1C>--145 Bottom left: 1856.12-23.404.
Bottom right: 1856.12-23.379
EA 27718 [1877. 11-12.86ff.)
2 " .
3 Lift: 1967-4-2.1 [Q 2086 h,,: S('(' Baile\. BM Catalogue Lamps IV (1996).
132. pis. 173. 1 79)
4 1900.7-26.5 [L 107): mould on riF;ht. cast on left
5 Top: '1866.6-8.2. Right: 1856.1C>--4 234
6 "189611-21.2
7 Lift: EA 22402 [1885-11-1.274). Cmtre: EA 22181 [1885. 11- 1.247).
Right: EA 22379 [1885-1 1- 1.295)
8 Lift: '1899.12-19.1. Right: "904. 2-4"45 '
9 Lift: '1859.2-16.12. Right: EA 223 64 [1885 11 - 1.156?)
10 Lift: EA 53955 [1915-2--7.5). Right: WA 48333 [1880.5-1.3 1)
11 Top: '1756.1-1.904. Lift: '19046-1.2
1.4 From left. clockwisr. '1856.12-23.378 [L 481; '1856.12-23.366 (Thanos);
'1961.12-18.2 (formerly Victoria &: Albert Museum 1929""1901);
'18571~7 CmlTe: '1859.12-26.327. Bottom: '1868.6-20.263 (= Walker.
fig. 78)
16 Top: 191512-a.53. Left: EA 51547 [1912.11-<)-401) (fig. 2)
17 Top: 1910.10-11+ Left: PRB, uninv. (= pl. 19-5)
18 (1) 1915. 12-29+ (2) 1847.11-6.62 (=fig. 15.3). (J) 1904.2-4.888. (4)
1980.10-14.8.(5) 1856.12-23.)92. (6) 1814-7-4-1557.
19 (1) 1873lh2O515. (2) 1S,J.5-U4 [L 1591. (J) PRB uninv. ('U 32').
(4) 1 693-5-24.8, L 701 (5) PRB,IIIlinv. (=pl. 17 left)
104
20 141: 1853.5-28.62 (Tarsus). Right: 1848.lh41v. ( 1;8
'''"1 see"6"21.3~
21 MLA 1986.7-1 . 1 2. 3.4
22 Top'MLA 1928.4-13.8. Bottom left: 0Ml.A 1928 -1
liIk Roman Pottery. pl. XIVb). Bottom rioht. 'Ml.A-4"'"J.9 (see a.,..
1,....7-1.11
231856.12-230406 [K 17) (=Bamett and MendIeson.1'IIruillll ...
pis. 21 and 137) ,-"'"
24 Top. left to right: 1856.12-23.345,344,359 and 339(= 1Im1ett...
Mendleson. Tharros, 243-4. pis. 20, 142). Bottom 1eft: 1905-5"'m1
Bottom nght: 1985-10-14.2 (ex 1888.12-18.714) '"
25141: 1868.6-20249 Right: 1963.12-22.1
26 Top: 1849.6-20.8 [G 186). Bottom: 1857.12-20.169; ll1li8..6-aa11
27 Top: 1933.6-1).1 (Rhodes). Bottom: EA 37631
29 1860.10-224
32 Top: 1856.12-26.392 (Puteoli). Bottom: ls".j-lJ.116;
1990.1- 27.1, 2, 4 (glass), 3, 5 (lamp)
34 From top left, clockwise: 1868.6-20.228; 1856-u--a""
(1IM1l); 1856.12-26.5)8; 1951.9--7.3; 1888~2lI.54
J5 1888'9""20-47
"Right: MLA 1875-10-12.16 [Dalton. no.lI6o). . . .
1I8a.l-<J.l [Dalton, no. 903). Bottom rigItt: -MIA
110. goB)
YI Uft: 1756.1-1.101.4 [K 631. Right: 1756-1-1.1132
39 Top: 198z.7-z9.59 (formerly Victoria c!c Albert Museum 33~1883;
from Salamis, Cyprus). Bottom: 19 1 7. 1 (>-15. 2
FIGURES
1 (1) 1949.5-1.5 (Tmtagel); (2) MLA 1985-1-1.8 3 (Istanbul);
U) 19495-1.9; (4) MLA 19857-1.75
Z EA 51547 [1912.11-9-411 (Faras)
3 (1) WA 139Oz5 [1977.11-2.Z) ('Petra'); (2) 1977- 10-11.9 (Knidos, old
find: Newton), 1864.1(}-'7-1359 (Ephesos)
u) 1857S-4Z0 [L 85); (4) 1868.1-10755 [L 41 (Corfu)' (5)
(Somme valley); (6) 1855.5-12.6 [L 1641 ('London')" 1917101
('Koln'); (8) EA 5 26z; (9) 1856.12-26.557 [L 147) rr;,.t;;l~~
Stamps: (1) CCRISPIN / PRJ; (Z) VETIl/OJ7TA' (4) eP;;;:~
(5) Q.TERV (see pI. 39 bottom); (7) LRPIS; (8) SIOCMcA:i"')~'"
(4-9 in plantQ pedis) 9 .......
11 (1) 19093-2O Z3; (Z) 1915 1Z-2910 (Rome); U) 1856-12-2). lL
(Tharros); (4) 19093-2OZ; (5) 1814-7-4.1553 [L 131); (6) t8ft&.:"
(7) 1814-7-4. 1554 [L 136 ); (8) 1904,z-4-748 [M 33) (VailOll)' - -
(9) 1856.12-Z3381 (Thanos); (10) 1868.1-10753 [L6) (~
(11) 1814.7-4. 1557 [L 142) Stamps: (1) CMEM[mi) I PRIMYS;
(Z) CRISPINI; (3) C.VIBIE[ni); (4) P01VS I PCOR[neliJ; (6)
(7) NAE I HVS?; (8) CRESTI; (10) LAVIL; (11) SEXMCL (6, 10, u
pltmta pedis)

4 (1) EAu)87 [1888.~2O551; (z) EA u181 [1885-11-1.2471, base

detail; u) EA 5110 [1837.7-14.158)
19 (1) 1848S-4 1 [L 8) (Kerch pen.); (Z) 1868.6-20 :161; (J)
5 1900.7-26.5 [L 1071 (Arezzo) (4) EA 22270 [1885-1 1- 1.361 ); (5) 1856.12-26.556 (L 135) cr-
Annunziata); (6) 1872.4-5.205; (7) 1868.6-20-431
6 1936.7-20.1 [Q 1009)
20 (1) 1848.S-4.15; (2) EA 22437 (1885.11-1.362); (J) 185),
1 19171(}-15-2 (Somme valley) (4) 1853.5-28 .6 3 (L 36) (Tarsus); (5) 1853.5-28.71; (6) 1~
(Cyprus); (7) EA 5255; (8) 1853.5-28.69; (9) 189+11-1.512
8 (1) 19'5-3-U.Z; (z) 1973.1-5.1

9 (1-3, 5) 1955~2O9O, 89, 19', 169; (4) 1955.~zo.80;

(6,7) 1886-4-1.1703 (Naucratis); 1955.~2O.91; (S-10): 1857.10-15.3, 1,6
(all Monte Testaccio, Rome)
U (1) Towneley Call., PRB not inv. 1M 2405); (z) ItpOoll
10 (1) 1846.1-10.10 (Eppelsheim, AIzey); (Z) 1816.6-10.Z57, rim (J) 1856.12-Z3-403 (Tharros); (4) 19'5-12-13.1; (5) 1856-
(= pl. )8 left) (1'harros); (6) 1859.11-Z9.10 (Karpathos); (7) 1!J2l1.4-1)1
(8) EA 37413 [19OZ.10-11.11Z); (9) 1851.12-18.164; (10)
11 EA uJ65 [1885.11- 1-459) (Egypt?); (11) EA 5Z58; (12) EA 69)81 [I!)1lz.r26-t3t. J
&: Albert M1IlleUIl\, 472-1891); (13) EA so654 (l86L7'"Q4oI
U (1) 1904-2-4-902; (Z) 11167.11-:u.)88 Sinai); (14) EA Z7718 (1877.11-U.86If.) (= pL 20."
13 (I) 1814-7-4-1559 [L 128); 1915.12-29-4 (Rome)

it ~~i~).2 [L 123); (Z) 1904-2-+888 [L 1261 (Loriol); 0) 749 24 1911.10-18,.

25 (1) EA 5129; (2) EA s8677 (1921.2-12.2); (J) 1JA
15 ~) 1873.8-20.515 [L 581 <-
pi. 19); (Z) 1865.12-14-40 [M Z); 14) EA 59555 (1929-10-16.1291; (5) EA so6J4 [1911.i
U) 847.8-6.62 [L 551 (Me pi. t8.z); 14) 1855.5-12.7 [L 161) ('London') (6) EA 27718; (7) EA 5149B (1912.11""9-)911
<9) ~1 (two views)
16 (I) 1904-2-+1)26 [L 17a1 (Onnp); (Z) PRB, uninv. base;

106
26 1956.2-19. 1
27 (1) 1839.11-<}.148; (2) 1839.11-12.2; (3) 1814.7-4.656; .
(4) 1873.8-20.333 [L 531; (5) 19 10.12- 19.7; (6) 19B<>.1<r146 (ex Victoria &
Albert Museum 1068-1905); (7) 1839.11-12.1; (8) EA 5252;
(9) 1935.&-103; (10) 1871.7-14.8 [M 25341 ('Wroxeter')

28 (1) 1961.2-13-4; (2) 19801()-149, (3) EA 51638 [19 12.11-<}496)

(Faras); (4) 1912.11-<}521 (Faras)

29 (1) 1859.12-26.323; (2) 1926.4-10.48

30 1933.&-1J.1 (Rhodes)

31 (1) 1977.1()-11.61; (2) 1856.12-26.392 (rim detail); (3) EA 5101;

(4) EA 5264; (5) 1975-2-73 (Benghazi); (6) 1975-2-75 (Benghazi);
(7) 1864.10-7.1936; (8) EA 5261; (9) EA 5257

J2 (1) 1927-4-11.130 (Tanturah); (2) 1814.7-4.1118; (3) 1927.4-11.97

(Tanturah); (4) EA 5260; (5) EA 22381 [1885.11-1.296)

33 (1) 1927.4-11.128 (Tanturah Dor; one side restored);

(2) 1927-4-11.100 (Tanturah); (3) EA 22179 [1885-11-1.272) (rim detail)

34 (1) 1893-7-13.86; (2) 1921.12-20.117; (3) 1859.12-26.556 (Knidos)

108
 I ~TlTUT

!, '''j k..Ef?4 NI ;(1/

HANDBOOK t,;cf,7q (Iu" 19<j:::;

MEDITERRANEAN
ROMAN POTTERY

Roman
I al pottery. defined for convenience as that made and d . h'
use WIt In
t Y and the Roman provinces between about 100 Be and AD 60
b h . d 0, can
e c araetense by a group of stylistic and technical developments
whICh bUIlt upon those of the Hellenistic Greeks and then led to those
of the Byzantine and Islamic worlds. Roman pottery can provide
evidence for ancient literacy. artistic trends and trading patterns within
the complex of Mediterranean lands which made up the empire, and
in recent years local archaeological activity in these regions has
produced a plethora of new finds and information.

Here John Hayes, the acknowledged authority in this field, explains the
particular features which mark Roman-period wares from the
Mediterranean area. concisely describing how vessels of various types
were manufactured, decorated, traded and used. An ideal identification
guide, the book is copiously illustrated with the author's own drawings,
and the photographs include many pieces never before published.

With 7 colour, 75 black & wllite alld 35 lille illHstratiorls

JOH W HAYES studied Classical Art and Archaeology at Cambridge

and specialises in Mediterranean and European archaeology, particularly
pottery and glass. Formerly a Curator in the. Greek & Roman
Department at the Royal Ontario Museum 111 Toronto, he IS now an
expert consultant,on excavations all over the Mediterranean regIOn as
well as the author or editor of numerous reports, catalogues, arncles
and monographs.
19541113
ISBN 0_7141-2216-S

1/8.60
1499
BRITISH MUSEUM PRESS