LRCW3

Late Roman Coarse Wares,

Cooking Wares and Amphorae
in the Mediterranean
Archaeology and archaeometry
Comparison between western and eastern
Mediterranean
Edited by

Simonetta Menchelli, Sara Santoro,

Marinella Pasquinucci and Gabriella Guiducci

Volume II

BAR International Series 2185 (II)

2010

Published by
Archaeopress
Publishers of British Archaeological Reports
Gordon House
276 Banbury Road
Oxford OX2 7ED
England
bar@archaeopress.com
www.archaeopress.com

BAR S2185 (II)

LRCW3 Late Roman Coarse Wares, Cooking Wares and Amphorae in the Mediterranean: Archaeology
and archaeometry. Comparison between western and eastern Mediterranean. Volume II.

Archaeopress and the individual authors 2010

Cover illustration : Eratosthenes map (drawing by Giulia Picchi, Pisa, after G. Dragoni, Eratostene e l'apogeo della scienza

greca, Bologna 1979, p.110).

Papers editing: Giulia Picchi, Pisa

ISBN 978 1 4073 0736 7 (complete set of two volumes)

978 1 4073 0734 3 (volume I)
978 1 4073 0735 0 (this volume)

Printed in England by Blenheim Colour Ltd

All BAR titles are available from:
Hadrian Books Ltd
122 Banbury Road
Oxford
OX2 7BP
England
www.hadrianbooks.co.uk

The current BAR catalogue with details of all titles in print, prices and means of payment is available
free from Hadrian Books or may be downloaded from www.archaeopress.com

ON THE ORIGIN OF CARTHAGE LR AMPHORA 1

ANDREI OPAI
11 Burdock Lane, Toronto, Canada, M3C 2G6 (aopait@gmail.com)
During the first half of the 3rd century AD different pottery workshops of Cilicia started producing a new amphora type. The
emerging shape was probably created under the influence of Dr.30/Galloise type 4 as other variants, more related to the early
Roman Pompeii type 5, seem to have been abandoned around the mid 3rd century AD. Towards the end of the 3rd and the first half of
the 4th centuries, a new subtype occurs that lays the foundation for the well-known LRA 1.

KEYWORDS: AMPHORA, CILICIA, POMPEII TYPE 5, DRESSEL 30, GALLOISE TYPE 4, LRA 1, ATHENS, BEIRUT,
CAESAREA, NORTH SINAI, DOBRUDJA, CRIMEA.

LRA 1 is one of the amphora types most frequently found in the

western and northern Black Sea and especially around the
Mediterranean from the 5th century AD. The term LRA 1 was
coined by Riley for his Berenice and Carthage typologies (Riley
1979, 1981). Besides Riley there were more tentative typologies
made by Kuzmanov (1973), Scorpan (1977), Egloff (1977),
Keay (1984), Opai (1984), Peacock and Williams (1986) and
Van Alfen (1996). The most recent typological essay on this
form is by Piri, who rightly divides this type into two grandes
gnrations morphologiques, LRA 1A and LRA 1B, with a
transitional variant (Fig.1) (Piri 1998, 98-99; 2005, 69-85, fig.
26). However, a number of other variants are not included in his
typology.
The discoveries made in Dobrudja, the Athenian Agora and the
Crimea allow us to refine this typology, identifying at least six
sub-types (Opai 2004, 8-11). The evolution of this amphora
type is quite well established for the 5th through to the 7th
centuries AD. Starting with the 6th century, the shape of this
amphora can be seen to have adopted more of a canonical form,
and was made in many areas with a long tradition of wine
production and amphorae making, such as Cyprus, Kos and
Rhodes. This paper will only focus on the early varieties of
LRA 1.
THE PRECURSORS OF LRA 1
The origins of this very famous amphora type was suggested by
Piri to have been in Crete (2005, 70, fig. 26). However, his
suggestion that this type derives from Cretan amphorae
(Robinson 1959, G197 & K112) is quite difficult to accept
(Piri 2005, fig. 26).
I may agree with Piri that the cylindrical shape of the 6th
century AD subtype of LRA1 resembles the typical Cretan
amphora form but this shape becomes canonical during this
century and there is no transitional variant between an early
Roman Cretan amphora and LRA 1.
An origin for LRA 1 in eastern Cilicia was advanced by
Reynolds (2005, 565; 2007; 2008, 70-72) and this is probably
the closest to the real origins of this amphora type. However,
Reynolds suggestion that LRA 1 has roots in the type Pompeii
5 class is difficult to accept, as the morphology of this amphora
has nothing to do with the shape of LRA 1, although the fabric
of both types has many common points. Based on his research
at Beirut (His task was extremely difficult as he worked mainly
with fragments that, although providing a good dating for the
closed archaeological contexts, were difficult to be used for

typological classification. We all know that a good typological

seriation needs complete examples.), Reynolds pointed out three
transitional variants dated to the early 3rd century AD (2008, 70,
fig. 3.b, c, p). The fact that these variants are contemporary
suggests that they were produced by different pottery
workshops. The variant with bulbous neck that is very common
at Beirut (Reynolds 2008, fig. 3c), is connected to the sub-types
of the end of the 3rd and of the 4th centuries AD only by its
folded-bevelled handle (I have seen in the Athenian Agora
collection some 1st century AD examples of Pompeii 5 type - P
4502, P 25732 - that have the rim strongly pulled to the interior
creating a bulbous upper part of the neck; this seems to be a
common practice for some Cilician potters. Reynolds also
published similar rims discovered at Beirut, cf. Reynolds 2005,
plate 2, fig. 14). However, the shape of this handle points to the
strong influence of the Dr 30/Gauloise 4 amphora that seems to
have influenced the Cilician potter. The most suitable
forerunner of LRA 1 is probably the third variant mentioned by
Reynolds, which is the Cilician imitation of Dr 30/Galloise 4
(Reynolds 2005, 565, fig. 23; 2008, 72, fig. 3). In fact, the mid
3rd century variant is correctly presented by Reynolds but he
does not appear to connect it with the Dr 30/Galloise prototype
(Reynolds 2008, 70, figs. 3.d, f). Although his variant proposed
for the 4th century carries some weight, a few observations must
be made. Firstly, as can be seen below, the subtype of the 4th
century does not end in a small hollow foot (Reynolds 2005,
plate 4, figs. 27-28; 2008, 70, fig. 3.h) but instead in a massive
conical foot. This foot is not always preserved, as it is the case
with two amphorae from Athens (P 11726 & P 29339). The
conical foot of the former example came off perhaps because it
was attached when the body was already too dry and there was
not a good adherence between them. Secondly, it is difficult to
accept a dating to the last quarter of the 4th century AD, as
already the first variant of the classic LRA 1 was on the market
during this time, according to excavations made in Cyprus
(Williams 1987. We should not rule out the possibility that the
Cypriot variant was on the market a little bit earlier) and those
LRA 1 examples discovered on the shipwreck of Yassi Ada
(Bass & Doorninck 1971). However, more closed
archaeological contexts which contain these variants are
necessary before we have a precise dating of this subtype.
This shape was previously briefly mentioned by Arthur and
Oren (1998, 203, fig. 6.1, 2).
A study in the Athenian Agora (Stoa of Attalos), completed by
some discoveries made in Dobrudja and the Crimea, allows a
more complete picture of the development of this predecessor

1015

LRCW3
during the 3rd and the first half of the 4th century AD. The
forerunners of LRA 1 have been divided into three subtypes:
A
The archetype seems to be a Cilician imitation of Galloise
4/Dressel 30 (Figs. 2 and 3). It is difficult to say which one was
the source because this shape seems to enjoy a large popularity
during the 3rd century A.D. Certainly, it is a new shape for
Cilicia. The complete example discovered in the Athenian
Agora (P-11936) is a little bit taller than Dr. 30 but it also has
shallow grooves on the central area of the body (Figs. 4a and b).
The height is 67.5 cm, the maximum diameter 39 cm, the rim
diameter 10.8 cm and the base diameter 6.2 cm. The rim was
made by folding outwards and downwards the top of the rim
with a small undercut. This technique will be used for many
sub-types and variants through the next centuries, and it is an
important characteristic of this type. The handles are arched
with a wide central groove, folded-beveled to use the term
coined by Reynolds; their shape resembles more the handles of
Dr 30 than of Galloise 4. The shoulders are wide and the body is
plump, tapering towards a narrow ring base. However, this base
is similar to the base of Galloise 4. The fabric is typical Cilician
with grains of quartz, pyroxene and foraminifera. This amphora
comes from a cistern (C 14:2) which was filled between 240
A.D. and 275 A.D. (I am grateful for this information to J.
Hayes). On the other hand, a similar, unique, fragmentary
example discovered at Beirut was dated early 3rd century AD by
Reynolds (2005, 565, fig.23; 2008, 72, fig.3.p), therefore a date
around the first half of the 3rd century AD for the use of this
container is quite feasible. Excellent parallels have been found
on a shipwreck at Punta Mazza also dated to the first half of the
3rd century A.D. (Tigano 1997, 81, no. 9, 10, pl. III.9-10; 85,
no.17, pl.7.17. The author of this discovery considers the two
fragments as belonging to type Dressel 30 similis, while another
fragmentary amphora, which also has a Cilician origin, is an
Unknown type). At least three fragmentary amphoras belong
to this Cilician sub-type according to the fabric and morphology
(Williams 1997, 101, no.5). The fact that this sub-type is not so
abundant at Beirut may be due to specific trade connections
between Beirut and certain areas of Cilicia, while
Dr.30/Galloise 4 was made by different workshops and had
different markets as is attested by the Athens and Punta Mazza
finds.
B
The next sub-type continues the same developing line of subtype A. A complete example has been discovered in a well of
the Athenian Agora (P 14078) and was published by Robinson
(1959, 68, deposit P 18:2, K111, pl. 15). It has the same ovoid
body but tapers more sharply to the lower part; the base was
broken and reconstructed in plaster but it is certainly narrower
than that of the previous sub-type. The rim becomes flattened
while the handles are almost rounded in cross section preserving
a deep, beveled, longitudinal groove on the external part. This
type of handle will become typical for the next generation of
this type. The neck is narrow and starts to increase its height.
The body still preserves a very shallow grooving around the
middle (Figs. 5a and b). The dimensions of this amphora are
slightly reduced as its height is 64 cm, the maximum diameter is
34 cm, the rim diameter is 8.2cm, and the base diameter is
2.8cm. This deposit was dated initially first half of the 3rd
century (Robinson 1959, 68), but recently J. Hayes dates the
end of the first period of use during the Herulian invasion. As
can be seen, both sub-types came from deposits with similar
dating. However, using cross-dating and analyzing different
morphologic details, it can be confidently said that amphora P11936 (subtype A) was at the extreme of its data-range or

residual. Therefore, we can confidently date this sub-type to

between AD 260 and 280. A later date than that of the previous
sub-type is suggested also by the narrow handles and base. This
sub-type is a crucial link to the sub-type of the 4th century.
As mentioned above, a similar discovery comes from the Sinai
Peninsula (Arthur and Oren 1998, 203, fig. 6. 2) (Fig. 6). It
cannot be ruled out that these two amphorae were manufactured
at about the same time but in different workshops, although the
Athenian find has narrower handles that suggest a slightly later
date. The most striking difference is in the shape of the rim of
the latter example, which is slightly bent towards the interior
and not vertical as in the former variant. It may be that this find
has an intermediary position between sub-types A and B.
C
The next sub-type might be considered as the last sub-type of
these pre-cursors of LRA 1, as the dimensions and the shape of
the vessel display visible changes. According to some
differences, mostly in the form of the rim, it can be divided into
two variants.
C-1
Two amphorae discovered in the Athenian Agora (P 11726 and
P 29339) allow us to understand the major changes which occur
during the first half of the 4th century A.D. (Figs. 7a and b and
8a and b). The neck and the body become elongated, whilst the
handles, although preserving the deep groove, are not arched
but fall vertically onto the narrow shoulders. The base of the
Athenian example is rounded, although we cannot rule out the
existence of a conical spike which was detached when the
amphora went out of use. The rim is remarkable, which
continues to look like a band but its tip is slightly bent to the
interior, continuing perhaps the tradition of the previous form
discovered in the Sinai Peninsula (Arthur and Oren 1998, 203,
fig. 6. 2). The two examples discovered at Athens have different
dimensions suggesting that this sub-type has been manufactured
in different sizes. The largest example (P-11726) has a height of
64 cm, a maximum diameter of 29.6cm and a rim diameter of
8.5cm. It comes from the filling of a well (N 18:5) that is closed
around the middle of the 4th century AD. The find indicates that
this subtype was in vogue during the first half of the 4th century
AD. The second Athenian example (P-29339) has identical
morphological characteristics but more reduced dimensions and
the bottom is broken. It preserves a height of 47.2cm, a
maximum diameter of 19.2 cm and a rim diameter of 6.3cm. It
has perhaps half of the capacity of the former amphora. A
fragmentary rim of this type has been discovered at
Bezymyannaya, a fortified settlement of the Chersonesan chora
(Fig. 9a&b). Its fabric has a hackly fracture with ill sorted white
(lime?), brownish inclusions, and angular, opaque quartz. The
color vary between very pale brown (10YR 7/3) and light
yellowish brown (10YR 6/4).
C-2
The second variant, although it does not change much from C-1,
has minor differences which determined us to treat it separately,
being perhaps manufactured in a different workshop. The rim
was also folded outwards and downwards creating a band,
which is vertical and not bent towards the interior. The body
differs from variant C 1 as it is shaped more conically, with a
visible angle between shoulders and body, covered by shallow
grooves. It ends in a full conical spike. The complete example
discovered in Dobrudja (Opai 1980, 301, pl. VI.2; XIII.2) has a
height of 53.5 cm, a maximum diameter of 15.3 cm, and a rim
diameter of 5.8 cm (Figs. 10a and b).

1016

A. OPAI : ON THE ORIGIN OF CARTHAGE LR AMPHORA 1

A similar amphora has been discovered at Chersonesos
(museum storeroom, inv. 26a/36573), and fragments of rim and
handles come from the excavations made at Bezymyannaya.
Although the Chersonesan example has the bottom missing, the
morphology of the amphora is typical for variant C-2, with
some minor differences such as a deeper, sharpened
longitudinal groove of the handle (Fig. 11a and b). The fabric is
typical for LRA 1 with well-sorted, abundant inclusions of
pyroxen, and sparse small grains of quartz and reddish
inclusions (iron oxides?). The color is reddish-yellow (7.5 YR
6/6 to 7.5 YR 7/6). Another fragmentary rim discovered at
Bezymyannaya, attests to the existence of a further Cilician
workshops (Fig. 12a&b). It has a calcareous matrix with
abundant calcareous material (including foraminifera) with
many voids and reaction rims that create a spotted effect when
is freshly broken, together with small grains of quartz and
pyroxene. This fabric type looks similar to the LRA1 fabric
from Yumurtalik-Ouest (north-western shore of the Gulf of
Iskenderon-Cilicia) illustrated among the specimens
published on the Internet (Reynolds 2005), and described by
Williams (2005, a & b).
PROTO-LRA 1
Probably around the middle or the third quarter of the 4th
century AD, the amphora takes the shape of what we call LRA
1. The earliest variant is perhaps an amphora discovered in Italy
(Arthur 1998, 164, fig. 5.1) (Fig. 13). This variant may be
connected to an earlier fragment discovered at Beirut and dated
c. AD 325+ by Reynolds (2005, pl. 4, fig. 25). The morphology
of a small amphora, perhaps a fractionary variant, discovered in
Egypt suggests that it also belongs to this intermediary period
(Egloff 1977, 113, type 168, pl. 57.8) (Fig. 14). Both amphorae
still have a conical body ending in a prominent button. The
body is only partially ribbed, leaving the central area plain. To
these amphorae we should add least two variants of this protoLRA 1 that have been discovered in the inner harbor of
Caesarea (Tomber 1999, 313, fig. 5.82-84).
Probably, a little bit later, can be added to this intermediary
group another amphora discovered at Kourion, Cyprus
(Williams 1987), and later considered as proto-LR1 by the
same scholar (Williams 2005). The body of this amphora is
elongated, almost ovoid, and completely ribbed (Fig. 15). On
the other hand, we should not exclude that this amphora is a
local Cyprian sub-type. Only further laboratory analyses and the
discovery of workshops could elucidate this problem. Starting
probably with the end of the 4th century and the beginning of
the 5th century A.D., the new canonical shape, known as
Carthage LR Amphora 1, was well established and
manufactured by many workshops in Cilicia, Antioch area,
Cyprus, and even Kos.
In conclusion, I consider that during the first half of the 3rd
century AD different pottery workshops of Cilicia tried to create
a new amphora form. The emerging shape was under the
influence of Dr.30/Galloise 4 as other variants, more related to
the early Roman Pompeii type 5, were abandoned after the
middle of the 3rd century. Towards the end of the 3rd and the
first half of the 4th century AD this new sub-type will lay the
foundation for the well-known LRA 1.
AKNOWLEDGEMENTS
I am indebted to Paul Arthur and David Williams for reading
this paper and for their useful comments on it.
I am very grateful to Professor John Camp, director of the
American Excavations in the Athenian Agora, and to Paul
Reynolds who is working on early Roman amphorae from the

Agora, for allowing me to publish some of the Roman

amphorae discovered in the Agora that I consider as
predecessors of late Roman-early Byzantine types.
BIBLIOGRAPHY
Arthur, P., 1998, Eastern Mediterranean amphorae between 500
and 700: a view from Italy, in Ceramica in Italia: VI VII
secolo, Atti del Convegno in onore di J. W. Hayes (Roma
1995) (ed. L. Sagu), 157-184, Florence.
Arthur, P., and Oren, E. D., 1998, The North Sinai survey and
the evidence of transport amphorae for Roman and
Byzantine trading patterns, Journal of Roman Archaeology
11, 193-212.
Bass, G.F., and Doorninck, H., Van, 1971, A fourth century
shipwreck at Yassi Ada, American Journal of Archaeology
75, 27-37.
Egloff, M., 1977, La Poterie Copte, Genve.
Keay, S.J., 1984, Late Roman amphorae in the Western
Mediterranean. A typology and economic study: the
Catalan evidence, BAR Int. Ser. 196, Oxford.
Kuzmanov, G., 1973, Tipologia i kronologia na
rannozantiyskite anfori (iv-v v.), Archeologia, 14-21.
Opai , A., 1980, Considera ii preliminare asupra amforelor
romane i romano-bizantine din Dobrogea, Peuce VIII,
291-327.
Opai , A., 1984, Beobachtungen zur Entwicklung der zwei
Amphoratypen, Peuce IX, 311-327.
Opai , A., 2004, Local and Imported Ceramics in the Roman
Province of Scythia (4th-6th centuries AD), BAR Int. Ser.
1274. Oxford.
Peacock, D.P.S., and Williams, D.F., 1986, Amphorae and the
Roman Economy. An Introductory Guide. London.
Piri, D., 1998, Les importations damphores orientales en
Gaule mridionale durant lantiquit tardive et le hautmoyen age (IVe-VIIe sicles aprs J.-C.). Typologie,
chronologie et contenu, en SFECAG, Actes du Congrs
d'Istres, 98-106, Marseille.
Piri, D., 2005, Le commerce du vin oriental lpoque
Byzantine (Ve-VIIe sicles). Le tmoignage des amphores
en Gaule, Beyrouth.
Reynolds, P., 2005, Levantine amphorae from Cilicia to Gaza: a
typology and analysis of regional production trends from
the 1st to 7th centuries, in LRCW I, 563-612.
Reynolds, P., 2007, in S. Keay and D.F. Williams, Roman
amphorae: a digital resource, University of Southampton
http://ads.ahds.ac.uk/catalogue/archive/amphora_ahrb_200
5/details.cfm?id=268&CFID=146774&CFTOKEN=83880
127
Reynolds, P., 2008, Linear typologies and ceramic evolution,
Facta 2, 61-87.
Riley, J.A., 1979, The coarse pottery from Berenice, in
Excavations at sidi Khrebish-Benghazi (Berenice), (ed.
J.A. Lloyd), Lybia Antiqua Supp. V-2, 91-467, Tripoli.
Riley, J.A., 1981, The pottery from the cisterns 1977.1, 1977.2,
1977.3., in Excavations at Carthage conducted by the
University of Michigan VI (ed. J.H. Humphrey), 86-124,
Ann Arbor.
Robinson, H.S., 1959, The Athenian Agora. Pottery of the
Roman period. Princeton.
Scorpan, C., 1977, Contribution la connaissance de certains
types cramiques romano-byzantins (IV-VII sicles), dans
l'espace Istro-Pontique, Dacia XXI, 269-297.
Tigano, G., 1997, Rinvenimenti Subacquei a Milazzo e il relitto
di Punta Mazza, Milazzo.
Tomber, R., 1999, Pottery from the sediments of th Inner
Harbour (area I14), in Caesarea Papers 2. Herods
temple,the provincial governors praetorium and
granaries, the later harbor, a gold coin hoard, and other

1017

LRCW3
studies (eds. K. G. Holum, A. Raban and J. Patrich),
Journal of Roman Archaeology Suppl. 35, 295-322.
Van Alfen, P.G., 1996, New light on the 7th-c. Yassi Ada
shipwreck: capacities and standard sizes of LRA1, Journal
of Roman Archaeology 9, 189-213.
Williams, D.F., 1987, Roman amphorae from Kourion, Cyprus,
RDAC 32, 235-38.
Williams, D.F., 1997, A note on the fabrics of the amphorae
recovered from the imperial roman wreck near Milazzo,
Sicily, in Tigano 1997, 99-102.

Williams, D.F., 2005a, Late Roman amphora 1: a study of

diversification, in Trade Relations in the Eastern
Mediterranean from the Late Hellenistic Period to Late
Antiquity: the ceramic evidence (eds. M. Briese Berg and
L.E. Vaag), University Press of Southern Denmark, 15768.
Williams, D.F., 2005b, An integrated archaeometric approach to
ceramic fabric recognition. A study case on Late Roman
Amphora 1 from the eastern Mediterranean, in LRCW I,
613-24.

Fig. 1. (After Piri 2005, fig.25).

Fig. 2. Gauloise 4
(After Sciallano & Sibella 1994, 2
ed.) (sc. 1:10).

Fig. 3. Dressel 30 (After

Sciallano & Sibella 1994, 2
ed.) (sc. 1:10).

Fig.4a & b. Athenian Agora, P-11936: 4a sc.

1:2; 4b sc 1:10, courtesy of the American
School of Classical Archaeology at Athens
(drawings by the author).

1018

A. OPAI : ON THE ORIGIN OF CARTHAGE LR AMPHORA 1

Fig. 5a & b. Athenian Agora, P-14078: 5a

sc. 1:2; 5b sc. 1:10, courtesy of the
American School of Classical Archaeology
at Athens
(drawings by the author).

Fig. 6. (Apud Arthur and Oren 1998, 203, fig. 6.2) (sc. 1:5).

1019

LRCW3

Fig. 7: Athenian Agora, P-11726: 7a sc. 1:5; 7b

sc. 1:2, courtesy of the American School of
Classical Archaeology at Athens
(picture and drawings by the author).

Fig. 8a: Athenian Agora, P-29339: 8a sc. 1:5; 8b:

sc. 1:2, courtesy of the American School of
Classical Archaeology at Athens
(picture and drawings by the author).

1020

A. OPAI : ON THE ORIGIN OF CARTHAGE LR AMPHORA 1

a

Fig. 9. a: Bezymyannaya, sc. 1:1; b: close up of fabric (not to scale).

Fig. 10a&b. Plopu (Dobrudja), Opait 1980, pl.

VI.2; XIII.2: 10a: sc 1:4; 10b: sc 1:10 (drawings
by the author).

b
Fig. 11. a. Chersonesos, Cemetery Sovhoz No.10, urn XX-1, Inv. No. 2636573 (picture and drawings by the author) (sc. 1:5);
b. close up of fabric (not to scale).

1021

LRCW3

Fig. 12a: Bezymyannaya, sc.1:1; 12b: close up of fabric (not to scale).

Fig.13. Apud Arthur 1998, fig.

5.1, sc. 1: 10.

Fig.14. Kellia, apud Egloff 1977,

pl. 57-8, sc 1:10.

1022

Fig. 15. Kourion, apud

Williams 1987, sc 1:10.