From suburb to port: the rise (and fall) of

Classe as a centre of trade and

redistribution
Andrea Augenti & Enrico Cirelli

INTRODUCTION

‘W hen I want to understand Italian history, I catch a train and go to Ravenna.’ This famous
statement was written by the great historian Arnaldo Momigliano (Marcone 2002). He
sought to illustrate the enormous potential offered by Ravenna for understanding broader trends in
the urban history of late antiquity. But Momigliano would have done well to add one further stop
to his route — the city of Classe, situated just south of Ravenna, where indeed there is a railway
station today. Classe was one of the most important ports of the Mediterranean from the fifth to
seventh centuries AD, as is reflected by the extraordinarily rich assemblage of pottery discovered
there. Thus in the same way that Ravenna offers us unique insights into late antiquity, so Classe is
especially important for the contemporary study of production and commerce. Furthermore, the
study of Classe in its broader urban context offers us the opportunity to further our understanding
of the transformation of cities between late antiquity and the Early Middle Ages.
At Classe this involved the transformation of a settlement that was born from the urban infra-
structure of nearby Ravenna, into one of the most important capitals of the late antique Mediterranean
and a ‘privileged centre’ in the sense understood by Saguı̀ (2002), but which soon became a sizeable
town in its own right, with defensive walls, monuments and houses. What follows is a synthesis of the
main results of investigations begun in 2001, mainly in the area of the port of Classe. They were
carried out by the Dipartimento di Archeologia of the Università di Bologna in collaboration with
the Soprintendenza per i Beni Archeologici dell’Emilia Romagna and the Fondazione Ravennantica.

THE TOPOGRAPHY

THE BIRTH OF A TOWN

Our conception of the area later occupied by Classe is that of a typical suburban landscape of the
Imperial period. Some cemeteries are attested in this area from the first century AD, situated on the
sandy dunes that run close to the coastline. A number of stelae belonging to classiarii, the sailors
of the military fleet based in Ravenna at that time, have been found here. Their barracks have not
been found yet, although one cannot rule out the possibility that villas further inland, some of
which have been partially excavated, belonged to high-ranking military officials.
This situation seems to change completely from the first decades of the fifth century AD. While we
lack precise evidence for the chronology of the walls of both Ravenna and Classe, it is probable that
both of them were built up at the very beginning of the fifth century, when Ravenna was chosen as one
of the Imperial capitals in the West (AD 402). It is precisely at this time that Ravenna underwent a
major expansion of its urban core, growing from an original surface area of 33 ha to 166 ha. It is
also at the beginning of the fifth century that a larger conurbation emerges, one composed of the
nuclei of Ravenna, Caesarea and Classe (Fig. 10.1).
From this moment onwards, Ravenna became the focus of an intense building programme. This
firstly comprised the construction of the imperial palace and the mint, both of which were located
206 AUGENTI & CIRELLI

FIG. 10.1. Ravenna and Classe: satellite image showing the location of the principal areas discussed in the text. (A. Fiorini.)

in the southern sector of the city. It also included the centuries basilicas proliferated in the new capital, not
building of a number of churches, a new and funda- solely in Ravenna but also in the suburbs of Caesarea
mental element that conditioned the distinctive form and Classe, and were often monumental in form. For
of the late antique town. Between the fifth and sixth its part, Classe witnessed the construction of the huge
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 207

Building Function Size Size with

portico
Building close to the port canal (zone 1)
1 warehouse 323 m2 493 m2
2
2 warehouse 432 m 696 m2
2
3 warehouse 400 m 580 m2
2
4 warehouse 360 m 522 m2
11 workshop 132 m2
14 warehouse 420 m2 660 m2
Buildings south of the paved road (zone 2)
6 warehouse & 260 m2 325 m2
glass kiln
6b courtyard 180 m2
17 warehouse 247 m2 299 m2
17b courtyard 141 m2
8 warehouse 190 m2 240 m2
18 warehouse & 166.5 m2 247.5 m2
workshop
9 warehouse & 380 m2 560 m2
workshop
10 warehouse 399 m2 588 m2
Buildings in zone 3
7 warehouse 85 m2
‘Island’
12 warehouse & 496 m2
thermopolium?

FIG. 10.2. Plan of the excavated area of the port of Classe, showing the location of the buildings discussed in the text. (Drawing: authors.)

Basilica Petriana (AD 432–50), Sant’Apollinare, built in the fifth and seventh centuries, suggests that the princi-
the mid-sixth century and one of the most important pal commercial contacts with Classe came from the east
places of worship in the whole conurbation, and the Mediterranean, with wine as a key commodity. North
Basilica di San Severo (AD 570–95). Thus between African imports, in contrast, largely took the form of
the fifth and the sixth centuries Ravenna was re- olive oil for lamps, cereals and fish sauce, but also
fashioned as a characteristic late antique city, while included fine- and coarse-wares. In quantitative terms,
the landscapes of Caesarea and Classe were dictated north Africa was the principal commercial partner of
by contemporary new ideas of town planning. Classe between the fifth century and the middle of the
But the most important feature of Classe — indeed sixth. Subsequently, the east became more important,
the driving factor behind its creation — is without although north African sources were not eclipsed.
doubt the port (Fig. 10.2). This was located along a Additionally, foodstuffs, especially wine, were
canal that formed the northern edge of the settlement imported from southern Italy as demonstrated by the
and connected the old Augustan inner basin to the huge quantities of Calabrian Keay 52 wine amphorae
sea. The port was composed of a series of warehouses (Cirelli 2007; Augenti et al. 2007a). Indeed, an inten-
organized in at least two rows, one of which faced the sive programme of pottery quantification, the first for
canal and the other bordering a paved road that ran Classe, has provided us with the basis for a better
parallel to it; in the centre of the canal there was an understanding of the key role played by this port and
island. These storage buildings would have held an by Ravenna as redistribution centres for foodstuffs for
impressive volume of goods (mostly foodstuffs), traces almost the whole of northern Italy. It suggests that the
of which can be recognized in the thousands of pottery capacity of the warehouses excavated between 2001
sherds that cover the surface of the archaeological site, and 2005 more than exceeded the needs of the inhabi-
and whose number sharply increases whenever a new tants of both centres.
part of the site is excavated. The topography of the port suggests that it was
The evidence so far, particularly that of the planned from the beginning, with streets, drainage
amphorae, many of which can be dated to between and a minimum of eighteen buildings. There is also
208 AUGENTI & CIRELLI

evidence for craft production at Classe, for the exca- literary reasons, it is difficult to ignore the likely
vations that were carried out between 2001 and 2005 effect of the sacking by the Lombard Duke Faroald in
brought to light glass wasters in each phase. This AD 576 or 584 (Historia Langobardorum 3.13), or the
activity took place between the fifth and sixth centuries. conquest by a Byzantine army lead by Duke Droctulf
So far, the most likely hypothesis is that the small kiln in AD 585–6 (3.19). In any event, Classe still had a
discovered in the 1970s was in fact a glass-furnace: a long life ahead of it; in fact it was after these traumatic
huge concentration of wasters was found there recently events that the city witnessed the establishment of a final
(Augenti, Cirelli and Tontini forthcoming). Addition- huge basilica dedicated to Saint Severus. It is only at the
ally, slag and traces of iron forging were found close beginning of the seventh century that we start to witness
to warehouse 9 and can be dated to the fifth century. the first signs of urban transformation, particularly at the
Finally, semi-worked pieces of bone were found close port, for which our evidence derives almost exclusively
to warehouses 8, 17 and 18, in layers dated to the from recent archaeological excavations.
sixth and seventh centuries. Although no evidence of Disruption of the port along the canal is shown in
a workshop was brought to light, these finds do suggest many ways. The most striking development is that,
that there was one nearby. starting from the seventh century, the warehouses pro-
Life in the port lasted for a long time: evidence spans gressively ceased to function as such. Walls began to
the period between the beginning of the fifth century fall down, and robbing activity spread rapidly across
and at least the ninth. Nevertheless, the associated land- the site. In the same period, parts of these buildings
scape did not remain unchanged over this period. The were occupied by small houses. Until a few years ago
main road, for instance, was remade on more than only two examples of these were known. One was
one occasion, and the paved phase that is now visible added in the porch of warehouse 2, which was built
dates back to the time of Justinian (Brogiolo and reusing some of the walls remaining from the original
Gelichi 1998: 81, 121). Some buildings were the building and with rubble infilling the spaces between
object of restoration as well; this is the case, for pillars (Maioli and Stoppioni 1987: 41; Ortalli 1991:
example, of warehouses 2 and 6, both of which were 179–81); another was established within warehouse
modified during the sixth century with the addition of 12 on the island that was located in the centre of the
an inner courtyard. Part of the building material used canal (Maioli 1986: 162; Ortalli 1991: 179). Recent
in this work was probably produced by a limekiln excavations have revealed many more buildings that
that was built in warehouse 17 during the mid-sixth succeeded the late antique warehouses, often estab-
century. Moreover, some other buildings went out of lished within their confines. So far it has been possible
use quite quickly: this is what happened, for instance, to locate at least two houses built on the remains of
to warehouse 17, which was burnt down in a fire at warehouse 8, one in warehouse 17, and another in
the beginning of the sixth century and was never warehouse 7. They are all relatively small (ranging
rebuilt, possibly because an open area, free of build- from 25 to 33 m2), often built from clay and with
ings, was needed at that time.1 The fate of warehouses wooden frames; they also had brick sockets, some of
6 and 8 was different: both of them were damaged in which were created from the remains of pre-existing
the same fire, but they were restored promptly and late antique buildings.
used as warehouses again (Augenti et al. 2007b: 36; According to the evidence gathered so far, it is
Cirelli 2007). possible to trace a first phase in the transformation of
Thus the topography of Classe had begun to incor- the port, which saw houses being built with robbed
porate open areas some 100 years after the establish- material, and in some cases incorporating the remains
ment of the port. Nevertheless, its infrastructure was of still standing structures — some of which were in
fully functional until at least the beginning of the ruins. This building activity and subsequent occupation
seventh century, after which we start to see the first, can be ascribed to the seventh century, and more
unequivocal signs of urban disruption. precisely the second quarter of that century, that is
during the reign of Emperor Heraclius (AD 610–41).
A DOWNWARD TRAJECTORY Another feature that strongly points towards a radical
transformation of the landscape is the progressive
Classe was already experiencing difficulties during the spread of urban burials (Cantino Wataghin 1999; Cirelli
second half of the sixth century. While Paul the Deacon 2008: 114–30). This is particularly true of the port,
was probably exaggerating the picture somewhat for where isolated tombs and relatively small cemeteries
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 209

were discovered. Recent excavations brought to light exchange for globular amphorae of eighth-century
50 burials, either in small clusters of up to three individ- AD date (Negrelli 2007b: 454–68; Murialdo 2007:
uals or in more extended groups of between thirteen 18–20; Cirelli 2009), a process that implies that the
and seventeen graves. In this area of the town, none town still had a fleet and was integrated into maritime
of these burials can be dated before the seventh century. exchange networks. Excavations at Classe brought to
The cemetery around the Basilica di San Severo, how- light surprisingly few of these containers, possibly
ever, is slightly different: here burials started by the end because they were used in a very localized context. In
of the sixth century (Augenti 2007: 28–30; Barbiera any event, it is clear that Classe had ceased to be the
and Ferreri 2007). This earlier date is probably to be main port of Ravenna by this date (Giovannini and
explained by the establishment of the church that Ricci 1985: 67–71) and that it had lost its key trading
housed the tomb of the holy martyr, and which would role in the interconnected mediterranean world
have acted as a strong stimulus for burial. (McCormick 2001; Horden and Purcell 2000: 123–
Afterwards, during the eighth century, the topo- 72; Augenti 2010), which some scholars today define
graphy and internal organization of the urban area as the ‘Commonwealth of Late Antiquity’ (Harris
seem to have been altered radically. In the port, for 2003; 2007). Thus the port of Classe gradually dis-
instance, some new wooden structures were built on appeared until, at the beginning of the ninth century,
an alignment that was completely different from that the local historian Agnellus tells us that the whole
of pre-existing remains. This is the case with seven city had become a ghost from the past.
small sunken warehouses with wooden frames infilled Andrea Augenti
with clay that came to light within the area previously
occupied by warehouse 7. Two of them were used as A ‘POTTED’ HISTORY
granary stores, and yielded the remains of sacks with
burnt remains of barley, peas and lentils. Another CLASSE BEFORE CLASSE
wooden building was discovered in the western sector
of the port, where late antique warehouse 14 was Archaeological deposits of the early Imperial period
located. Structures such as these continued in use were found during the excavations carried out during
throughout the ninth century at least. During this the twentieth century. They related mainly to residential
second phase, the port maintained its role as a place buildings and burial areas in what were probably
of production. This is attested by the remains of an suburbs on the edges of the inhabited area of Ravenna
installation for clay-working, which included a tank and its new Augustan port. Since the excavated
connected to a small brick channel and a platform material from these has not been subject to any quanti-
where raw material could be worked in order to get rid tative analysis, it is difficult to establish the economic
of air bubbles (Augenti et al. 2006) (Fig. 10.3). What dynamics of the settlement at this period, and we
exactly was produced here remains enigmatic, although have access only to qualitative observations about the
the discovery of a mould for so-called ‘Sicilian lamps’ ceramic corpus. However comparison of these data
provides us with an important clue (Augenti et al. (Maioli 1990; Stoppioni 1990) with the known econ-
2007a: 274). omic trends of late antique Classe makes it possible
One more sign of the changed situation at Classe is to put forward some hypotheses about key aspects of
the gradual decrease in imported foodstuffs (such as this earlier settlement.
wheat, oil and wine) and other goods, a phenomenon The earliest occupation dates to the Republican
that is reflected largely in a decrease in the volume of period, with some fragments of black-gloss pottery of
ceramic containers. This situation is stark in compari- the second century BC coming from the areas later occu-
son to that for the fifth and sixth centuries, and is a pied by the late antique Basilica di San Severo and the
reflection of the progressive breakdown of the networks associated port. Although the nature of the settlement at
of mediterranean commerce (McCormick 2001: 27–63; this period is unclear, it is probable that the area of the
Wickham 2005: 708–824; Murialdo 2007). It did not, future Augustan port was occupied by villas and open
however, coincide with the sudden decline of farms. It is also likely that one of these structures,
Ravenna’s economy. Its Church remained second which was found below the large Roman building
only to that of Rome in terms of the size of its estates that was destroyed by the basilica’s construction, pre-
in the Italian context (Cosentino 2005: 431–3), and dates the first century AD (Augenti 2007). First-century
the produce of its estates was probably exported in AD structures, however, are better documented, with
210 AUGENTI & CIRELLI

FIG. 10.3. Reconstruction of an eighth-century pottery workshop adjacent to small sunken huts. (Drawing: G. Albertini.)

residential buildings and burial areas being attested cargo including local products. Some second- and
across the whole site. The most important pottery third-century AD contexts from Classe included luxur-
assemblages associated with this phase are composed ious ceramic products from the hinterland, particularly
mainly of local and regional productions. We are red-painted or glazed flat dishes (Maioli 2008: 699),
talking mainly about pareti sottili and other fine- that imitated metal dishes, such as those found in the
wares, such as the Sarius Surius moulded cups and Rethel hoard (Baratte 2000: 175–6). There is also a
many forms of north Italian sigillata, whose production strong presence of Italian sigillata from Arezzo. In
is attested widely at Ravenna between the second half of the archaeological literature there are also reports of
the first century and the second century AD (Fig. 10.4). small quantities of eastern sigillata B and pareti sottili
Sarius cup production is well known in northern Italy, pottery, even if we have little idea as yet of the quan-
examples being distributed across the region of tities involved. By contrast, early forms of African
Venice (Veronese, Vigoni and Annibaletto 2007: 199, red slip ware from the area of Carthage and vessels
fig. 1) and all the territories of the Po delta (Pupillo produced in the territory of Lepcis Magna in Tripoli-
2007: 227–8). This can be inferred also from the tania (Felici 2004) are almost completely absent. As
cargo of the Comacchio (Berti 1990), for this boat far as amphorae are concerned, there was a sharp
probably set sail from the port of Ravenna, with a increase in the quantity of local Romagna (especially
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 211

FIG. 10.4. North Italian red slip ware mould

found at Ravenna. (From Maioli 1990: fig. 00.) Q1

Forlimpopoli and Santarcangelo types) and middle FROM SUBURB TO CITY: CLASSE IN THE
Adriatic productions, which slowly replaced Dressel FIFTH CENTURY
2–4 amphorae (Stoppioni 1990: 458). To these can be
added some fragments of amphorae from the Aegean The analysis and quantification of the pottery from the
area, especially from Rhodes, as well as from north recent excavations (2001–7) conducted in the port and
Africa. Thus all these first- and second-century AD close to the Basilica di San Severo provide us with the
materials reveal an increasing diversification of the opportunity to write a new ‘potted’ history of the
commodities at Ravenna, which may have had a commercial dynamics of the port between the late
more active commercial harbour than that at Classe. antique period and the Early Middle Ages. Although
At the beginning of the third century AD, Classe work on the material is not yet complete, enough is
witnessed a significant increase in the volume and now known to provide us with a relatively detailed
range of overseas imports, deriving in particular from picture. In this section we shall primarily discuss the
the Iberian peninsula and north Africa, together with pottery from warehouse 6 in the southern zone of the
local and regional sources — a situation that continued archaeological site (Augenti et al. 2007a), as well as
until the end of the fourth century. Assemblages also from contexts from two other important buildings, ware-
often included east mediterranean amphorae, mainly houses 17 and 18, which were located in the second row
from the Aegean region, although most of the amphorae, of warehouses facing the channel and which remained in
such as the Africana I type, came from Tunisia, although use down to the first half of the seventh century, before
also from Mauretania and Tripolitania. being refashioned as private houses.
212 AUGENTI & CIRELLI

TABLE 10.1. Quantitative analysis of 10,736 vessels from warehouse 6 at Classe calculated on the basis of the EVEs.

Pottery category AD 425–75 AD 475–525 AD 525–610

Amphorae from the east Mediterranean 70 100 101
Amphorae from Tunisia and southern Italy 18 59 75
African red slip ware 90 1,122 283
Phocaean red slip ware 25 47 308
Late antique Italic red slip ware 18 48 196
Local painted ware (closed forms) 5 79 11
Imported painted ware (LR 1 — Greece) 4 72 24
Late antique glazed ware (local) 10 121 8
African lamps 38 1,850 100
Local lamps 8 39 60
Plain ware (local and imported) 50 1,324 458
Coarse-ware (local and imported) 96 1,965 321

The greatest obstacle to the study of these contexts is imports were present in higher volumes than elsewhere
the huge quantity of material recovered. During the five at Ravenna, with lesser quantities of imports from the
years of excavation that began in the summer of 2001, east Mediterranean, and from local and other Italian
over a period of effectively nineteen months, some sources (Maioli 1990: 426; Stoppioni 1990: 463). Our
800,000 pottery sherds were collected. Because the evidence is constituted mainly by finds of amphorae
excavated contexts are all quite different, we cannot and fine-wares, even though we lack hard numerical
yet say how many vessels these represented in most data. This situation continues into the fourth, fifth and
cases. The exception to this is warehouse 6, where a the first half of the sixth centuries AD, as has been
definitive calculation based on the EVE (Estimated shown by finds from the excavations of 2001 (Augenti
Vessel Equivalent) method gives us a total of 8,400 et al. 2007a: 264; Augenti, Cirelli and Tontini forth-
vessels, from 50,000 fragments (Augenti et al. 2007a: coming). It is especially true for fine-wares, with
262), although it should be noted that this context more than 90 per cent coming from Tunisia between
was exceptionally well preserved. Another context, 425 and 475 AD. Inside one of the rooms in warehouse
datable to between the end of the fifth century and 6, for example, some 1,122 African red slip (ARS) ware
the beginning of the sixth, was smaller, with 300 vessels2 were unearthed (Table 10.1). It should be
vessels and 11,000 fragments, and is still being studied. remembered that the excavated area represents only a
More work thus certainly needs to be completed before quarter of the entire building, which was only partially
we have a clear idea of the total volume of pottery in investigated during the 1980s. The remaining material
use at the site. Overall, however, we can begin to from this context originated from Phocaea and pro-
detect a uniform trend in the pattern of imports to the duction centres in the Adriatic region, notably the
site. This provides us with an indication of the type area around Rimini, but also from eastern Tuscany. A
of cargoes being carried by ships arriving at Classe third group of red painted ware, with a repertoire of
and destined for the new capital of Ravenna between forms different from Italic or African productions,
late antiquity and the Early Middle Ages. cannot be associated yet with a particular geographical
source. Its fabric is similar to that observed in some east
COMMERCIAL TRENDS mediterranean ceramics and, according to Ferrazzoli
and Ricci (2007), the samples recovered at Classe can
From the beginning of the third century, prior to the be compared to some as yet unpublished red painted
development of an urban focus at Classe, north African wares from the city port of Elaiussa Sebaste.
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 213

The many forms of ARS that reached Classe

were produced in the two main Tunisian ateliers,
namely the northern Tunisian D1 and D2 pro-
ductions (Mackensen 1998: 30–3) and those from
central Tunisia (Mackensen and Schneider 2002:
131). The most frequent and representative fifth-
century ARS form found at Classe is Hayes 61b, a
type of dish that was popular and thus widely
imitated, as in the case of local red slip ware
productions (Negrelli 2007a: 301) and at Arezzo
(Molinari and Nespoli 2005: 305). In the layers
dating to the last quarter of the fifth century many
Hayes 81b bowls were also recovered, identifiable
as products from Zeugitana. The great quantity of
Hayes 85b cups of C5 production, which are gener-
ally rare in Italy (Tortorella 1998: 41), should also
be noted, a provisional total of c. 500 examples
(EVEs) having been recovered from warehouse 17,
which is one of the best preserved in the whole
port area. Examples have been found also in flooring
deposits: in the case of warehouse 6, a sample area of
c. 60 m2 yielded a total of 1,182 vessels. The same
contexts also yielded many moulded oil lamps from
the same atelier: 1,850 examples (EVEs) of types
Hayes IIa and IIb were found inside warehouse 6
— one of the largest assemblages of this type of
vessels ever found outside its production site.
The recovery of so large a quantity of central
Tunisian fine-wares is indicative of the intense
volume of other imports that do not survive in the
archaeological record. Thus Tunisian fine-wares,
together with the many coarse-wares found at ware-
houses at Classe (Cavalazzi and Fabbri 2010), may
reflect indirectly the import of north African grain
(Reynolds 1995: 128). Tunisian products main-
tained their primary importance at Classe in the
period between the end of the fifth and the middle
of the sixth centuries AD. A similar trend has been
observed at other major centres of redistribution in
the west Mediterranean, notably at Rome, Naples,
Marseilles and Tarragona, with forms of commerce
focused on centralized authority and the annona
system (Keay 1984: 417–27), but achieved by
means of commercial routes and modes of distri-
bution quite similar to those of earlier centuries.
In the fifth and the first half of the sixth centuries,
huge quantities of north African and east mediter-
ranean amphorae reached the port of Classe (Cirelli
2007: 46), amongst which cylindrical amphorae
(Fig. 10.5, 1–2) and spatheia (Fig. 10.5, 3–4) were FIG. 10.5. Tunisian amphorae found inside warehouse 17. (Drawing:
particularly important. Very significant in this E. Cirelli.)
214 AUGENTI & CIRELLI

FIG. 10.6. Warehouse 17 during the excavation season of 2005. (Photo: authors.)

regard is the large deposit found inside warehouse 17, a Bien and Bonifay 2007: fig. 8, no. 16) that was
building that was destroyed by fire between the last produced in southern Byzacena (Bonifay 2004: 143).
decade of the fifth and the beginning of the sixth centu- This whole context can be dated to between the end
ries AD, and in which 380 north African amphorae have of the fifth and the beginnings of the sixth centuries on
been discovered in situ (Fig. 10.6), along with many the basis of coin evidence. Despite the great quantity of
others vessels that have yet to be quantified. Among African amphorae recovered in warehouse 17, it never-
the containers that were found in situ were 200 theless should be noted that in most contexts analysed
spatheia, which are smaller and more slender than for this period Tunisian amphorae represent 31% of
variants of the amphora Keay XXVI-f/Bonifay type all material (Cirelli 2008: 135). In the same period
31 (Bonifay 2004: 124–5, fig. 67). The production the number of eastern amphorae is greater even if
centres for these have not been identified yet, although their storage capacity was less — approximately half
a preliminary analysis of their fabric suggests that they that of north African types. Their content was com-
may have been from the region of Nabeul3 and are simi- pletely different as well, with most of them being
lar to those found on the Dramont wreck in southern used to transport wine. The most frequently represented
France (Bonifay, Capelli and Long 2002; Bonifay type is the Late Roman 4 (32%) and other Palestinian
2004: 124, fig. 67). It perhaps may be identified as a amphorae (Cirelli 2007: 49, 152, VII.32). There are
variant of the Spatheion 1, production of which has also many Cilician Late Roman 1 (20%), Aegean
been attested also in the workshops of Sidi Zahruni Late Roman 3 (7%) and Late Roman 2 (2%) amphorae
(Ghalia, Bonifay and Capelli 2005: fig. 3, n. 2). One present. The distribution of these containers probably
hundred and eighty large4 oil amphorae have been was directed toward the Danubian countries (Kara-
recovered from the same context. They were character- giorgou 2001: 129, 146), and the Tyrrhenian coast,
ized by having pointed bases inserted inside a wooden notably at Rome (Panella and Saguı̀ 2001: 780) and
boarding, a practice different from that used at ware- Naples (Arthur 1998: 168; Carsana, D’Amico and
house 7, where the amphorae were stored inside holes Del Vecchio 2007: 424). Although these types of
cut into the clay surface.5 Convincing parallels have amphora had been found only rarely at the site hitherto,
not been found yet for this type of amphora, although recent work suggests that they were present elsewhere,
the closest was found at the basilica in Rue Malaval notably the Basilica Petriana, even though we do not
in Marseilles. The most common variant of this has a have quantified data yet. This same period also saw
thickened rim (Moliner, Bien and Bonifay 2007: fig. 6, the importation of very small quantities of south Italian
no. 6); a second and smaller variant with a deep amphorae (3%), while local productions were absent.
collar rim was found in the same contexts and is prob- At the same time there is evidence for the trans-
ably a variant of the type Keay 34/Bonifay 53 (Moliner, formation of agriculture in the surrounding territory,
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 215

which in the early Imperial period had been a key area

for the production of wine carried in the so-called
Forlimpopoli amphorae.
Between the fifth and seventh centuries, the port of
Ravenna imported wine from the east Mediterranean,
with smaller amounts from southern Italy and Egypt,
as well as oil for lighting, wheat and salsamenta (fish
sauce and other types of products) from north Africa.
The range of amphora types represented is quite differ-
ent to those observed in cities and fortifications in the
Balkans and the east Mediterranean.6 At Classe the
balance of imports changed in the second half of the
sixth century, probably following the Justinianic
conquest and the capture of many western markets by
east Mediterranean products. Thus eastern goods were
prevalent at the port. This is true of the Late Roman 1
(25%) and Late Roman 4 amphorae with later variants,
together with the Agora M334 and the Late Roman 5
(36%) — a trend that is repeated elsewhere in the
Adriatic region (Auriemma and Quiri 2007: 38–40,
46–7; Cirelli 2007: 49). The Aegean amphora Late
Roman 3 also continued to be imported (4%), together
with a rarer form (Fig. 10.7) and the Samos amphora
type, which is generally associated with privileged
centres of commerce (Saguı̀ 2002: 15), particularly
during the seventh century. During the same period
there was a huge increase in the quantity of Phocaean
red slip ware at Classe, a development that is probably
related to the redistribution of wheat from east FIG. 10.7. Aegean amphora found in a pit at Classe
Mediterranean sources (Augenti et al. 2007a: 274–5, (GRA excavation). (Reproduced courtesy of M.G. Maioli.)
fig. 25). It is also important to note that north African
products continued to be imported, both in terms of have been used by the inhabitants of Ravenna, whose
fine-wares and amphorae, the latter mainly comprising number had increased exponentially on account of the
large cylindrical vessels produced between the end of presence of the imperial entourage: recent population
the sixth and the seventh centuries (32%). The presence estimates suggest that its population may have been
of locally-produced red slip wares also increased, and as high as 10,000 people (Cosentino 2005: 411–12).
was redistributed through its hinterland to many northern Despite this extraordinary growth, however, Ravenna
Italian settlements (Cirelli 2008: 134–5, fig. 15). was still a city of middling size compared to Con-
stantinople and Rome. Their populations were in the
RAVENNA: A CENTRE OF region of 300,000–500,000 and 300,000–350,000
REDISTRIBUTION inhabitants respectively at this time (Dagron 1974:
526–49; Lo Cascio 1997: 40), while Alexandria,
From the first half of the third century onwards, Antioch, and Carthage hosted populations of over
Ravenna became a centre of redistribution, paving the 100,000 inhabitants (Hurst 1993: 336). Indeed,
way for its subsequent transformation into one of the according to some estimates, c. 624 ships were necess-
greatest ports of the late antique Mediterranean (Stop- ary each year to supply wheat for 80,000 people at
pioni 1990: 466). The zenith of this phenomenon was Constantinople (Kingsley and Decker 2001: 2), each
at the beginning of the fifth century AD, when it was vessel holding c. 31,200 tonnes.7
established as an Imperial capital and Classe became The approximate quantity of goods stored inside the
its commercial port. But to whom were the goods warehouses at Classe can be gauged from the evidence
found in its warehouses destined? Some of them must recorded for warehouse 17. This building, which was
216 AUGENTI & CIRELLI

destroyed by fire at the end of the fifth century AD, figure is equivalent to about ten ships of the same
yielded 180 intact north African oil amphorae, each tonnage as Yassıada A (Fig. 10.8). It has been observed
one of which could hold c. 60 litres, and 100 spatheia, that warehouse 17 stored c. 10 tonnes of grain prior to
equal to about one modius per amphora (6.5 kg). That its destruction by fire. To this figure should be added
means that the warehouse could have held c. 11,000 c. 13 tonnes of oil and other goods contained in the
litres of oil that could have been used for lighting and spatheia located in the front rooms10 and a quantity
unguents, together with c. 1,170 kg of salsamenta. of fine-wares that has yet to be calculated. Considering
Recent studies have shown that the large oil amphora the capacity of the upper floor, the entire storage
found here contained castor oil (Pecci et al. 2010), space in this building was roughly equivalent to that
whose production used the same pressing system as of a ship of the dimensions of the Yassıada A wreck
used for olive oil, but with lower costs and a higher (7,692 modii).
index of productivity, but whose distribution has not At the end of the fifth century, therefore, warehouse
been studied yet. On average, however, shelled castor 17 contained sufficient goods to satisfy a tenth of
seeds yielded c. 1,500–1,600 kg a hectare in com- Ravenna’s population for a month. If we use this
parison with the 2–3 quintals produced by olive trees building as a yardstick for measuring the capacity of
(Duke 1981). Spatheia, by contrast, are often associ- the other buildings discovered at Classe, we can
ated with the transport of fish sauce (Reynolds 1995: argue that the port area was able to harbour at least
50; Bonifay 2004: 129; Bonifay 2007: 24), although 28 ships contemporaneously. At the end of the fifth
no evidence for this has been found in the residue century, by contrast, Constantinople would have
from the five samples analyzed to date (Pecci et al. needed at least 256 ships of similar size to satisfy the
2010). demands of its c. 300,000 inhabitants, while Portus at
The assemblage of amphorae described above was the same period would have needed more than 85.
found close to the northwest corner of the warehouse, This calculation, based on the capacity of the buildings
occupying just over a sixth of its ground floor area of of the port area of Classe, allows us to argue that they
c. 225 m2. On the opposite side of the building were had been built to hold food for more than 30,000
found African red slip ware cups and other fine- inhabitants. However, this is certainly a conservative
wares, along with Hayes IIa moulded central Tunisian estimate, because it is based on known archaeological
oil lamps; in the middle of the building there were evidence. It needs to be remembered that the buildings
thick ropes that were used for lifting the amphorae excavated to date represent only a fraction (1.2 ha) of
and other heavy materials. A third of the ground floor the warehouses that must have constituted the port
of warehouse 17 was used to store wheat, particularly area between the fifth and seventh centuries: indeed,
the back rooms on its southern side. For every square recent topographic studies allow us to estimate that
metre of excavated area, a quantity equal to c. 130 kg the warehouse sector of the port covered c. 4 ha.
of grain (20 modii) has been calculated, which if Since the excavated area is thus less than a quarter of
averaged up to three rooms with an area of 24 m2 the whole, the late antique port could have stored
comes to an equivalent of 1,440 modii of grain.8 It goods for more than 120,000 individuals. Since this
should be borne in mind, however, that these calcu- exceeded the needs of the population, it suggests that
lations exclude the upper level of the warehouse, for Classe served as a centre for the redistribution of goods.
which we have remarkable archaeological evidence,
as well as three other rooms in the middle of the CRISIS OF THE PORT AND
building that were the same size as those previously TRANSFORMATIONS OF THE MATERIAL
calculated. The overall building capacity, therefore, CULTURE IN THE SEVENTH CENTURY
would have been some three times as great (225 m2)
as that of the storage area where wheat traces have The seventh century saw a change in the quantity of
been found (72 m2). The quantity of grain that could goods imported by Classe and thus in its commercial
be stored in this building would have been sufficient dynamics. The ceramic samples studied to date reflect
to satisfy the everyday needs of between 6,000 and trends similar to those observed at other contemporary
18,000 individuals.9 This calculation suggests that the late antique ports along the coasts of the northern
needs of late antique and early medieval Ravenna, Adriatic and Tyrrhenian seas, where the movement of
with a population of 10,000 inhabitants, could have cargo ships and the distribution of goods were mediated
been met by c. 450 tonnes of grain per month; this through and coordinated by Carthage, the centre of the
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 217

FIG. 10.8. Reconstruction of warehouse 17. (Drawing: G. Albertini.)

Byzantine economic system in the west Mediterranean were often reused for burials. An 8% increase in the
(Murialdo 2007: 10). They suggest that Classe was import of south Italian wine amphorae (Keay 52 and
transformed from being a centre of storage and trans- other later forms) has been observed also.
shipment into one of consumption, with excavations In the following centuries the settlement underwent a
showing that many buildings were converted into slow process of abandonment. Early medieval pottery
private houses and smaller warehouses. still appears in later contexts in the port area, including
The most meaningful information is provided by the eighth-century amphorae produced in southern Italy
growth of local and regional productions, such as the and the east Mediterranean. The same types also have
locally-produced ceramic oil lamps (Romagna) that been identified close to the Basilica di San Severo,
almost entirely replaced imported varieties, and which where a monastic community was installed at the end
can also be inferred from the discovery of ceramic of the ninth century (Augenti 2007; Laszlovszky
moulds. There was also a new local red slip ware 2007). This complex was also characterized by the
production that was slight by comparison with the existence of a prestigious building that hosted Emperor
imported variety of the previous century and with a Otto during his first visit to Ravenna in the tenth
smaller repertoire of forms. However, contexts of century (Cirelli 2008: 145). The pottery found during
seventh-century date do include many north African the last season of excavation shows that Classe,
amphorae, particularly variants of Keay 62 and 56. Ravenna and its territory continued to import east
Eastern Mediterranean amphorae also are well rep- Mediterranean products, such as wine amphorae and
resented, including the latest variants of the Gazan Byzantine glazed wares. But we should note also that
Late Roman 4, the Agora M344, and the Egyptian prevailing cultural influences gradually shifted toward
Late Roman 7, while from the Aegean Samos amphorae central Italy and the Ravenna–Rome axis, as the
218 AUGENTI & CIRELLI

presence of many vessels similar to forms produced NOTES

in Lazio suggests. The analysis of new excavation
contexts thus demonstrates a quite dynamic history 1. These particular circumstances allowed us to find almost the
for Classe. It shows how the settlement grew from whole content of the warehouse in situ and in a remarkable
state of conservation: see below, p. 000. Q1
being a small suburban area into a great commercial
2. Over a total of 1,217 fine ware vessels (EVEs), representing
urban port, and then suffered a slow decline, becoming
one-fifth of the whole ceramic assemblage (¼ 6,826 vessels
in turn a suburb and finally an open quarry. overall).
Enrico Cirelli 3. Michel Bonifay, pers. comm.
4. Height 90–110 cm; width 24–40 cm.
The new excavations have shed new light on the role of 5. For a similar practice, see building 2 at Dichin in Bulgaria
Classe, and by implication Ravenna, in the wider (Poulter 2007: 87, 91, fig. 19).
mediterranean setting during the late antique period. 6. At the Dichin warehouse, for instance, all the amphorae
If we compare its situation with that of Portus — one derived from eastern sources (Swan 2007: 253)
of the main port cities on the western coast of Italy 7. Equivalent to 4,800,000 modii. For the relationship between
— it is clear that both became increasingly important the modius and the kilogram, see: Rickman 1980: 261, 263.
as centres for the arrival and redistribution of goods. Calculation of the number of ships is based upon vessels the
But while Portus was mainly a conduit for supplies size and capacity of that recovered near the Yassıada A
wreck (7,692 modii) (Parker 1992: 454–5), which could hold
destined for Rome and parts of Latium, Classe, as the
a cargo equal to 50 tonnes.
city-port of the sedes regia at Ravenna, had always
8. An estimated value equal to 9,360 kg of grain.
played a broader role, redistributing goods to contig- 9. This is based on the assumption that the needs in wheat and
uous areas of northern Italy. The quick upsurge of forage of a first-century Roman legion (5,000 soldiers)
Classe as a major commercial node and the establish- consisted of about 7,500 kg per day, that is 1.5 kg a head ¼
ment of new trading-routes across the Adriatic ensured c. 0.2 modius (Elton 1996: 66–9; Heather 2005: 80).
that Ravenna was connected not only with Tunisia, but 10. Over 2,000 modii excluding the weight of the amphorae
also with the east. These were the outcome of political themselves.
and strategic decisions that were taken initially during 11. A complete overview of the archaeology of Classe is given in
the early Empire and subsequently reaffirmed by the the recent volume edited by Augenti (2011).
Ostrogothic kings and again by the Byzantine adminis-
tration. Although the urban trajectories of Classe and REFERENCES
Portus were different, they both bear witness to the
Arthur, P. (1998) Eastern mediterranean amphorae between 500 and
importance of large port cities in the history of the
700: a view from Italy. In L. Saguı̀ (ed.), Ceramica in Italia;
Mediterranean. Indeed, they were the outcome of
VI–VII secolo. Atti del convegno in onore di John W. Hayes
major political investments, as can be seen in their (Roma, 11–13 maggio 1995): 157–83. Florence, All’Insegna
increasingly monumental topography during the del Giglio.
fourth century at Portus (Paroli 2004; Keay et al. Augenti, A. (2007) (ed.) La basilica e il monastero di San Severo a
2005) and the fifth century at Classe.11 A thorough Classe. La storia, gli scavi. Ravenna, Elios.
understanding of the archaeology of both sites, which Augenti, A. (2010) Città e porti dalla tarda antichità all’alto
is greater now, will certainly help us gain a deeper medioevo. Rome, Carocci.
understanding of changing patterns of urbanism Augenti, A. (2011) (ed.) Classe. Indagini sul potenziale archeo-
between the late antique and early medieval periods. logico di una città scomparsa. Bologna, Ante Quem.
Andrea Augenti Augenti, A., Bondi, M., Carra, M.L., Cirelli, E., Malaguti, C. and
Rizzi, M. (2006) Indagini archeologiche a Classe (scavi
2004): primi risultati sulle fasi di età altomedievale e dati
archeobotanici. In R. Francovich and M. Valenti (eds), Atti
del IV congresso di archeologia medievale (Abbazia di San
Galgano, Chiusdino–Siena, 26–30 settembre 2006): 124–31.
Florence, All’Insegna del Giglio.
Augenti, A., Cirelli, E., Sabetta, T., Savini, E., Zantedeschi, E. and
Nannetti, M.C. (2007a) Classe, Podere Chiavichetta: dati dalla
campagna di scavo 2001. In S. Gelichi and C. Negrelli (eds),
La circolazione delle ceramiche nell’Adriatico tra tarda anti-
chità ed alto medioevo. III incontro di studio sulle ceramiche
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 219

tardoantiche ed altomedievali (Venezia, 24–25 giugno 2004): Archaeometry (British Archaeological Reports, International
257–94. Mantua, Società Archeologica Padana. Series 1,662): 423–37. Oxford, Archaeopress.
Augenti, A., Maioli, M.G., Malnati, L. and Sassatelli, G. (2007b) Cavalazzi, M. and Fabbri, E. (2010) Cooking ware from the
Nuovi scavi archeologici a Classe: campagne 2004–2005. In excavation of a 5th–7th century context in Classe (Ravenna,
A. Augenti and C. Bertelli (eds), Felix Ravenna. La croce, Italy). In S. Menchelli, S. Santoro, M. Pasquinucci and G.
la spada, la vela: l’alto Adriatico fra V e VI secolo. Catalogo Guiducci (eds), LRCW3. Late Roman Coarse Wares, Cooking
della mostra (Ravenna 10 marzo–7 ottobre 2007): 33–8. Wares and Amphorae in the Mediterranean. Archaeology and
Milan, Skira. Archaeometry (British Archaeological Reports, International
Augenti, A., Cirelli, E. and Tontini, S. (forthcoming) Commercio e Series 2,185): 623–33. Oxford, Archaeopress.
circolazione monetaria a Classe e Ravenna tra VI e VII secolo. Cirelli, E. (2007) Ravenna e il commercio nell’Adriatico in età tardo
In Uomini, merci e commerci nel Mediterraneo da Giustiniano antica. In A. Augenti and C. Bertelli (eds), Felix Ravenna. La
all’Islam (VI–X sec.). Atti del convegno (Bordighera 3–4 croce, la spada, la vela: l’alto Adriatico fra V e VI secolo.
dicembre 2004). Catalogo della mostra (Ravenna 10 marzo–7 ottobre 2007):
Auriemma, R. and Quiri, E. (2007) La circolazione delle anfore in 45–50. Milan, Skira.
Adriatico tra V e VIII sec. d.C. In S. Gelichi and C. Negrelli Cirelli, E. (2008) Ravenna: archeologia di una città. Florence,
(eds), La circolazione delle ceramiche nell’Adriatico tra All’Insegna del Giglio.
tarda antichità ed alto medioevo. III incontro di studio sulle Cosentino, S. (2005) L’approvvigionamento annonario di Ravenna
ceramiche tardoantiche ed altomedievali (Venezia, 24–25 dal V all’VIII secolo: l’organizzazione e i riflessi socio-
giugno 2004): 31–64. Mantua, Società Archeologica Padana. economici. In Ravenna da capitale imperiale a capitale
Baratte, F. (2000) Forme quotidiane a confronto: la ceramica d’uso esarcale. Atti del XVII congresso internazionale di studio
comune e il vasellame d’argento. I tesori tardoantichi. In S. sull’alto medioevo (Ravenna, 6–12 giugno 2004): 405–34.
Ensoli and E. La Rocca (eds), Aurea Roma. Dalla città Spoleto, Centro Italiano di Studi sull’Alto Medioevo.
pagana alla città cristiana. Catalogo della mostra (Roma, Dagron, G. (1974) Naissance d’une capitale. Constantinople et ses
22 dicembre 2000—20 aprile 2001): 174–8. Rome, ‘L’Erma’ institutions 330–451. Paris, Presses Universitaires de France.
di Bretschneider. Duke, J.A. (1981) Ricinus communis. In J.A. Duke and K.K. Wain
Barbiera, I. and Ferreri, D. (2007) Placing bodies and constructing (eds), Medicinal Plants of the World. Computer Index with
memory at San Severo. Annual of Medieval Studies at CEU more than 85,000 Entries. London, Longman.
13: 187–96. Elton, H. (1996) Warfare in Roman Europe, AD 350–425. Oxford,
Berti, F. (1990) (ed.) Fortuna Maris. La nave romana di Comac- Clarendon Press.
chio. Catalogo della mostra. Bologna, Nuova Alfa. Felici, F. (2004) La ceramica fine in Tripolitania nella prima e media
Bonifay, M. (2004) Études sur la céramique romaine tardive età imperiale: un quadro d’insieme. Africa Romana 15 (2002)
d’Afrique (British Archaeological Reports, International (2): 945–58.
Series 1,301). Oxford, Archaeopress. Ferrazzoli, A.F. and Ricci, M. (2007) Elaiussa Sebaste: produzioni e
Bonifay, M. (2007) Que transportaient donc les amphores africaines? consumi di una città della Cilicia tra V e VII secolo. In M.
In E. Papi (ed.), Supplying Rome and the Empire (Journal of Bonifay and J.C. Tréglia (eds), LRCW2. Late Roman Coarse
Roman Archaeology Supplementary Series 69): 8–32. Ports- Wares, Cooking Wares and Amphorae in the Mediterranean.
mouth (RI), Journal of Roman Archaeology. Archaeology and Archaeometry (British Archaeological
Bonifay, M., Capelli, C. and Long, L. (2002) Recherches sur Reports, International Series 1,662): 671–88. Oxford,
l’origine des cargaisons africaines de quelques épaves du Archaeopress.
littoral français. In L. Rivet and M. Sciallano (eds), Vivre, Ghalia, T., Bonifay, M. and Capelli, C. (2005) L’atelier de Sidi-
produire et échanger: reflets méditerranéens. Mélanges offerts Zahruni: mise en évidence d’une production d’amphores de
à Bernard Liou: 195–200. Montagnac, EMM. l’antiquité tardive sur le territoire de la cité de Neapolis
Brogiolo, G.P. and Gelichi, S. (1998) La città nell’altomedioevo (Nabeul, Tunisie). In J.M. Gurt i Esparraguera, J. Buxeda
italiano. Archeologia e storia. Bari, Laterza. and M.A. Cau Ontiveros (eds), LRCW1. Late Roman Coarse
Cantino Wataghin, G. (1999) The ideology of urban burials. In G.P. Wares, Cooking Wares and Amphorae in the Mediterranean.
Brogiolo and B. Ward-Perkins (eds), The Idea and Ideal of the Archaeology and Archaeometry (British Archaeological
Town between Late Antiquity and the Early Middle Ages: 147– Reports, International Series 1,340): 495–507. Oxford,
79. Leiden, Brill. Archaeopress.
Capo, L. (1992) Paolo Diacono. Historia Langobardorum. Milan, Giovannini, C. and Ricci, G. (1985) Ravenna. Le città nella storia
Fondazione Lorenzo Valla-Mondadori. d’Italia. Bari, Laterza.
Carsana, V., D’Amico, V. and Del Vecchio, F. (2007) Nuovi dati Harris, A. (2003) Byzantium, Britain and the West. The Archae-
ceramologici per la storia economica di Napoli tra tarda ology of Cultural Identity, 400–650. Stroud, Tempus.
antichità ed alto medioevo. In M. Bonifay and J.C. Tréglia Harris, A. (2007) Britain and China at the ends of the world?
(eds), LRCW2. Late Roman Coarse Wares, Cooking Wares Archaeological methodology and long-distance contacts in
and Amphorae in the Mediterranean. Archaeology and the sixth century. In A. Harris (ed.), Incipient Globalization?
220 AUGENTI & CIRELLI

Long-distance Contacts in the Sixth Century (British Archaeo- Marcone, A. (2002) Un treno per Ravenna. Il contributo di Arnaldo
logical Reports, International Series 1,644): 91–104. Oxford, Momigliano agli studi tardo antichi. Antiquité Tardive 10:
Archaeopress. 291–8.
Heather, P. (2005) The Fall of the Roman Empire. A New History. McCormick, M. (2001) Origins of the European Economy. Com-
London, Macmillan. munications and Commerce AD 300–900. Cambridge,
Horden, P. and Purcell, N. (2000) The Corrupting Sea. A Study of Cambridge University Press.
Mediterranean History. Oxford, Blackwell. Molinari, A. and Nespoli, C. (2005) Arezzo in età longobarda: dati
Hurst, H. (1993) Cartagine, la nuova Alessandria. In A. Carandini, L. inediti e nuove prospettive di ricerca. Archeologia Medievale
Cracco Ruggini and A. Giardina (eds), Storia di Roma 3. L’età 32: 305–16.
tardoantica II. I luoghi e le culture: 327–37. Turin, Laterza. Moliner, M., Bien, S. and Bonifay, M. (2007) La céramique de la
Karagiorgou, O. (2001) LR2: a container for the military annona on basilique paléochrétienne de la rue Malaval à Marseille:
the Danubian border? In S. Kingsley and M. Decker (eds), approche préliminaire. In M. Bonifay and J.C. Tréglia (eds),
Economy and Exchange in the East Mediterranean during LRCW2. Late Roman Coarse Wares, Cooking Wares and
Late Antiquity: 129–66. Oxford, Oxbow. Amphorae in the Mediterranean. Archaeology and Archaeo-
Keay, S.J. (1984) Late Roman Amphorae in the Western metry (British Archaeological Reports, International Series
Mediterranean. A Typology and Economic Study: the Catalan 1,662): 275–85. Oxford, Archaeopress.
Evidence (British Archaeological Reports, International Murialdo, G. (2007) Alto-Adriatico e alto-Tirreno nel mondo
Series 196). Oxford, British Archaeological Reports. mediterraneo: due mari a confronto tra VI e X secolo. In
Keay, S., Millett, M., Paroli, L. and Strutt, K. (2005) Portus. An S. Gelichi and C. Negrelli (eds), La circolazione delle ceramiche
Archaeological Survey of the Port of Imperial Rome (Archaeo- nell’Adriatico tra tarda antichità ed altomedioevo. Atti del III
logical Monographs of the British School at Rome 15). incontro di studio CER.AM.IS. Sulle ceramiche tardoantiche
London, British School at Rome. ed altomedievale (Venezia, 24–25 giugno 2004): 9–29.
Kingsley, S. and Decker, M. (2001) New Rome, new theories on Mantua, SAP.
inter-regional exchange. An introduction to the east Negrelli, C. (2007a) Vasellame e contenitori da trasporto tra tarda
Mediterranean economy in late antiquity. In S. Kingsley and antichità ed altomedioevo: l’Emilia Romagna e l’area medio-
M. Decker (eds), Economy and Exchange in the East adriatica. In S. Gelichi and C. Negrelli (eds), La circolazione
Mediterranean during Late Antiquity: 1–27. Oxford, Oxbow. delle ceramiche nell’Adriatico tra tarda antichità ed alto-
Laszlovszky, J. (2007) The monastery of St. Severo in Classe/ medioevo. Atti del III incontro di studio CER.AM.IS. Sulle
Ravenna. Annual of Medieval Studies at CEU 13: 197–211. ceramiche tardoantiche ed altomedievale (Venezia, 24–25
Lo Cascio, E. (1997) Le procedure di recensus dalla tarda giugno 2004): 297–330. Mantua, SAP.
repubblica al tardo antico e il calcolo della popolazione di Negrelli, C. (2007b) Produzione, circolazione e consumo tra VI e
Roma. In La Rome impériale. Démographie et logistique IX secolo: dal territorio del Padovetere a Comacchio. In F.
(Collection de l’École Française de Rome 230): 2–76. Berti, M. Bollini, S. Gelichi and J. Ortalli (eds), Genti nel
Rome, École Française de Rome. delta da Spina a Comacchio. Uomini, territorio e culto
Mackensen, M. (1998) Centres of African red slip ware production dall’antichità all’alto medioevo. Catalogo della mostra
in Tunisia from the late 5th to the 7th century. In L. Saguı̀ (ed.), (Comacchio, 16 dicembre 2006–14 ottobre 2007): 437–71.
Ceramica in Italia; VI–VII secolo. Atti del convegno in onore Ferrara, Corbo Editore.
di John W. Hayes (Roma, 11–13 maggio 1995): 23–39. Ortalli, J. (1991) L’edilizia abitativa. In A. Carile (ed.), Storia di
Florence, All’Insegna del Giglio. Ravenna. Dall’età bizantina all’età ottoniana II. 1. Territorio,
Mackensen, M. and Schneider, G. (2002) Production centres of economia e società: 167–92. Venice, Marsilio Editori.
African red slip ware (3rd–7th c.) in northern and central Tunisia: Panella, C. and Saguı̀, L. (2001) Consumo e produzione a Roma tra
archaeological provenance and reference groups based on tardoantico e altomedioevo: le merci, i contesti. In Roma
chemical analysis. Journal of Roman Archaeology 15: 121–59. nell’alto medioevo (Settimane di studio CISAM (Spoleto, 27
Maioli, M.G. (1986) Ravenna, loc. Classe, podere Chiavichetta, aprile–1 maggio 2000) 48 (2)): 757–820. Spoleto, Centro
quartiere portuale tardoromano e bizantino. Studi e Documenti Italiano per i Studi sull’Alto Medioevo.
di Archeologia 2: 161–4. Parker, A.J. (1992) Ancient Shipwrecks of the Mediterranean and
Maioli, M.G. (1990) Classe: la cultura materiale. In G. Susini (ed.), the Mediterranean Provinces (British Archaeological Reports,
Storia di Ravenna I. L’evo antico: 416–55. Venice, Marsilio International Series 580). Oxford, Tempus Reparatum.
Editori. Paroli, L. (2004) Il porto di Roma nella tarda antichità. In A. Gallina
Maioli, M.G. (2008) La ceramica invetriata a Sarsina. In A. Donati Zevi and R. Turchetti (eds), Le strutture dei porti e degli
(ed.), Storia di Sarsina I. L’età antica: 693–705. Cesena, approdi antichi: II seminario ANSER. Roma–Ostia Antica,
Stilgraf. 16–17 aprile di 2004: 247–66. Rome, Rubbettino Editore.
Maioli, M.G. and Stoppioni, M.L. (1987) Classe e Ravenna tra Pecci, A., Salvini, L., Cirelli, E. and Augenti, A. (2010) Castor oil at
terra e mare: città, necropoli, monumenti. Ravenna, Edizioni Classe (Ravenna–Italy): residue analysis of some late Roman
Sirri. amphorae coming from the port of Classe (Ravenna–Italy):
 RISE (AND FALL) OF CLASSE AS A CENTRE OF TRADE AND REDISTRIBUTION 221

relationship between form and function. In S. Menchelli, S.

Santoro, M. Pasquinucci and G. Guiducci (eds), LRCW3.
Late Roman Coarse Wares, Cooking Wares and Amphorae
in the Mediterranean. Archaeology and Archaeometry: 117–
622. Oxford, Archaeopress.
Poulter, A.G. (2007) The transition to late antiquity on the lower
Danube: the city, a fort and the countryside. In A.G. Poulter
(ed.), The Transition to Late Antiquity. On the Danube and
Beyond (Proceedings of the British Academy 141): 51–97.
Oxford, Oxford University Press.
Pupillo, D. (2007) Economia, produzione e commerci nel delta di
età romana. In F. Berti, M. Bollini, S. Gelichi and J. Ortalli
(eds), Genti nel delta da Spina a Comacchio. Uomini, terri-
torio e culto dall’antichità all’alto medioevo. Catalogo della
mostra (Comacchio, 16 dicembre 2006–14 ottobre 2007):
213–31. Ferrara, Corbo Editore.
Reynolds, P. (1995) Trade in the Western Mediterranean, A.D. 400–
700. The Ceramic Evidence (British Archaeological Reports,
International Series 604). Oxford, Tempus Reparatum.
Rickman, G.E. (1980) The grain trade under the Roman empire. In
J.H. D’Arms and E.C. Kopff (eds), The Seaborne Commerce
of Ancient Rome: Studies in Archaeology and History
(Memoirs of the American Academy in Rome 36): 136–58.
Rome, American Academy in Rome.
Saguı̀, L. (2002) Roma, i centri privilegiati e la lunga durata della
tarda antichità. Dati archeologici dal deposito di VII secolo
della Crypta Balbi. Archeologia Medievale 29: 7–42.
Stoppioni, M.L. (1990) Le anfore. In G. Susini (ed.), Storia di
Ravenna I. L’evo antico: 457–67. Venice, Marsilio Editori.
Swan, V.G. (2007) Dichin (Bulgaria): interpreting the ceramic
evidence in its wider context. In A.G. Poulter (ed.), The
Transition to Late Antiquity. On the Danube and Beyond
(Proceedings of the British Academy 141): 251–80. Oxford,
Oxford University Press.
Tortorella, S. (1998) La sigillata africana in Italia nel VI e nel VII
secolo d.C.: problemi di cronologia e distribuzione. In L.
Saguı̀ (ed.), Ceramica in Italia; VI–VII secolo. Atti del convegno
in onore di John W. Hayes (Roma, 11–13 maggio 1995): 41–69.
Florence, All’Insegna del Giglio.
Veronese, F., Vigoni, A. and Annibaletto, M. (2007) Il progetto
Vasa Rubra. Quaderni di Archeologia del Veneto 23: 198–203.
Wickham, C. (2005) Framing the Early Middle Ages. Europe and
the Mediterranean 400–800. Oxford, Oxford University Press.