Delos Carthage Ampurias
Delos Carthage Ampurias
C~~Ζ~~-~~ GΕ
AMPURIAS
«L'ERMA» di BRETSCHNEIDER
DELOS, CARTHAGE, AMPURIAS
DELIS, CARTHAG 1,
AMPURIAS
The Housing of Three Mediterranean
Trading Centres
BY
BIRGIT TANG
«L'ERMA» di BRETSCHNEIDER
ROME MMV
CONTENTS
6. CONCLUSION 175
APPENDIX: TYPOLOGY AND TERMINOLOGY OF PAVE-
MENTS 181
Introduction 181
Lexica 181
The literary sources and the terminological debate 182
Pavements: typology and terminology 185
Conclusion 189
GLOSSARY 193
BIBLIOGRAPHY AND ABBREVIATIONS 195
CATALOGUE 221
Preface to catalogue 223
Delos 226
Carthage 287
Ampurias 310
Concordance between cat. nos. and houses 333
Tables 339
List of figures 395
List of tables 397
ABBREVIATIONS AND SYMBOLS
USED IN TABLES
Α = atrium
av = average
ha = bathroom
beg = beginning
bi = bicolour (black-and-white)
C = courtyard
ch = chip pavement
cor = corridor
cu = cult room
di = dining room
E =east
ff = figurative frieze
gf = ground floor
ki = kitchen
1 = litre(s)
la = latrine
LP = late Punic
m = metre(s)
ma = marble
me = monochrome
míp = mortar pavement with inset pieces
mi = mortar pavement
mot = mortar pavement with tessera design
MP = middle Punic
mr = main room
N = north
of = opus figlinum
os = opus sectile
of = opus tessellatum
itip = opus tessellatum with inset pieces
iv = opus vermiculatum
P = peristyle
pc = polychrome
pe = pebble mosaic
pi = portico/porticoes
rect = rectangular
S = south
SE = south-east
sh = shop
St = storeroom
tc = terracotta
us = upper storey
W = west
wo = workroom/workshop
ACKNOWLEDGEMENTS
This book is largely based on my Ph.D. dissertation, which was underta-
ken at the University of Copenhagen, Department of Archaeology and Eth-
nology 1997-2000 and defended in 2001. The research project has benefited
greatly from a number of people and institutions.
Thanks are due to the Board of the Danish Academy in Rome for granting
me accommodation at the academy and to the entire staff of the academy for
assistance during my stays.
I am greatly indebted to the curators at the Museu d'Arqueologia de Cata-
lunya-Empúries for permission to work on the site and for generously pro-
viding me with information: in particular, the museum's director Xavier
Aquilué, Pere Castanyer, Marta Santos and Joaquin Tremoleda. I owe spe-
cial thanks to Marta Santos for providing me with relevant literature and for
help with translation. I am particularly grateful also to Lurdes Burés for our
discussions of the cisterns at Ampurias and for allowing me to read her un-
published paper presented at the Cura Aquarum conference in Sicily.
My research in Rome was funded by the University of Copenhagen and by
the Doctor Else Nikolajsen Grant. Fieldwork at Ampurias and Carthage and
participation in conferences were made possible by financial support from
the University of Copenhagen and the Engineer Svend G. Fiedler and Wife
Foundation.
I would like to thank senior lecturer, Annette Rathje, who supervised my
work. For reading drafts, practical help, discussions and encouragement I
am indebted to a number a friends and colleagues: Carsten Jacobsen, Kristi-
na Winther Jacobsen, Mette Korsholm, John Lund, Alexandra Nilsson,
Christopher C. Parslow, Birte Poulsen, Hanne Thomasen and Christina
Trier. My thanks also go to the members of the Evaluation Committee, Prof.
Inge Nielsen, Prof. Eva Rystedt and senior lecturer Lone Wriedt Swrensen
for inspiring criticism.
A generous grant from the Elisabeth Munksgaard Foundation made it
possible for me to work intensively on the preparation of the book for publi-
cation. I wish to thank the Department of Archaeology and Ethnology at the
University of Copenhagen for providing me with facilities during the publi-
cation process. Thanks are also due to Peter Spring who revised the English.
The revision was funded by the Danish Research Council for the Human-
ities.
I am most grateful to the former director of the Danish Academy in Rome,
Dr. Gunver Skytte for accepting my work for publication in the ARID Sup-
plementum series. The publication was made possible through generous
grants from the G.E.C. Gad Foundation, the Elisabeth Munksgaard Founda-
tion, the New Carlsberg Foundation, the Politiken Foundation and the
Queen Margrethe and Prince Henrik Foundation. My special thanks to all of
them.
Finally, I would like to express my warmest gratitude to Kjeld de Fine
Licht for the drawing of Figs. 5, 14, 21 and 28, and to Kristina Winther Ja-
cobsen for her friendship and encouragement throughout.
"Una casa è come un organismo vivo.
Suo principio vitale è l'uomo che l'abita e che l'ama".
(Pesce 1957, 23)
1. INTRODUCTION
AIM AND HISTORICAL BACKGROUND
8
Buitrey et al. 1989, 34-67; Tsäl{írgi's Ι95
2 For the activítíes a n d doeub2ëtitatíoid of Ρiι~lostεratos ~Η belos, see Macemeik $apiamatia 1983;
tιι6heiäκ 1592, 85.
14
Judging from the epigraphical source material, three main groups lived on
the island in the period 167/166 - 69 BC: Greeks (Athenians in particular),
Italians/Romans and Syrians/Phoenicians. 8 Moreover, the abundant material
facilitates the tracing of families through generations. 9 From ancient Italy
came not only Romans and Italians, but also Ita liote Greeks. 10 The Delian in-
scriptions employ the following designations in Greek: Ρωµαιος/Ρωµαιο t,
haλικοt. In Latin the term Italicei οccurs. The question is whether the
Greeks on Delos distinguished between Romans and Italians and Ita liote
Greeks, meaning that the terms were specific, or whether the Ρωµαιοςterm
was used for anyone coming from Italy as advocated by some sch οlars. 11
Strab. 14.5.2.
Vial 1984, 3.
Epigraphical material documents the presence of Delíans after 166 BC (Couflloud 1974, 247).
6 88 BC: App. Míthr. 528; Paus. 3.23.3-4; Strab. 10.5.4. 69 BC: Phlegm of Tralles, quoted by
Italiote Greeks' are from the Greek colonies of Italy. For Italiote Greeks on Delos, see Appen-
dix in Lomas 1993, 191-194.
11 Hatzfeld 1912, 6 and 132; Zalesskij 1983, 30, note 51.
15
However, Sohn rightly suggests that the term Roman when used in inscrip-
tions was reserved for people with Roman citizenship or from Rome. ~2 In any
case, it is a complex situation, as some of these Rhomaioi were freedmen of
Greek or Oriental origin working on Delos for Italic/Roman patron famil-
ies. 13 The term Ιznλtκοι appears to be a collective term for peoples from the
Italian Peninsula and its islands. 14
Carthage: At Carthage the late Punic period, from the mid-3rd century
BC until 146 BC, provides the richest material. 15 Literary sources state that
Carthage, situated to the north-east of the Lake of Tunis in North Africa,
was founded by Tyre in 814/813 BC. 16 As a Punic metropolis, it was a super-
power of the western Mediterranean from the mid-6th century BC onwards.
Its area of influence overseas comprised North Africa, Malta, western Sicily,
Sardinia, Ibiza and parts of the Iberian Peninsula where a `New Carthage'
(modern Cartagena) was founded by Hasdrubal, the son-in-law of Hamilcar
Barca, in 221 BC. 17 The last confrontation with Rome (the Third Punic War,
149-146 BC) was fatal to Carthage which was conquered and destroyed.
The epigraphical material from the 3rd and 2nd centuries BC documents
the presence of Libyans, Egyptians, people from ancient Sardinia and a group
of Sidonians. 18 Several Etruscan inscriptions have been found at Carthage and
its environs. One is particularly interesting, as it shows that a Marco Unata re-
sided in the vicinity of the city. The inscription probably dates from the period
295-264 BC. 19 Furthermore, literary sources of different periods mention the
presence of Italíote Greeks, other Greeks, Etruscans and Italíans/Romans. 20
Ampurias (ancient Emporion/Emporiae): At Ampurias the chronological
framework covers the period from the 2nd century BC to the 3rd century
AD, with particular emphasis on the 2nd century BC to the late 1st centur 5y
AD. The two excavated nuclei of the city are Neapolis and the Roman city.
The houses of the former can be dated within a broad chronological period
from the 2nd century BC until its abandonment in the Flavian age. In the
Roman city, the development of the houses can be traced from around 100
BC to the 3rd century AD.
Ampurias is situated in the north-east zone of Catalonia in Spain, on the
south-west side of the Gulf of Rosas (Fig. 1). It is the only documented Grae-
co-Roman city on the Iberian Peninsula and thus represents a unique oppor-
12
Solin 1983, 114-116.
3 ΙD 2346: The freedman Lucius Spurius is a Rhomaios.
14
Poccetti 1984, 647.
s The Roman period of the city is beyond the scope of the present work.
16 According to the `early' textual tradition, the city was founded in 1215 BC. For the `early' and
`late' textual traditions, see Lancel 1995, 20-23, and for the foundation myth, Lancel 1995, 23-25.
17
For the Barcids in Spain, see for example Lancel 1995, 376-380; Richardson 1996, 16-24.
18
Halff 1965, 82. Sidonians: Grainger 1991, 203-205. Only one inscription is complete.
19
Pittau 1996, 1666-1672.
20 For instance Diod. 14.77.4-5; Polyb. 36.7. For references, see Lassére 1977, 37-42.
21
Palaiapolis, the initial settlement, is the third nucleus, but it is situated below the village of Sant
Marti d'Empúries.
16
Ι . Rosas (Rhode)
Ampnoion (Βmροι~οπ/Empcnae)
Uiiaabni. Puíg de. Sant Andrea
Caatell de Ια Foscα de Palamós
Puíg Castellet de Lincei de Ματ
Roman villa of Can Marti, Soecalda
Badolono (Banali)
Bacoelona (Boenino)
Alorda Pac k, Calafell
τη 5ο KM
~ ι
the West. The city was fö .nded at the beginning of the 6th century BC by
Phòceans who came e6t ~ex 22 directly from Phocéa ín Asia Minor or from 'Mar -
seules (ancient Massilía) The earliest evidence for the name of Emporion
cdiïsists of ínscriptíóns on lead tablets from the late 6th/early 5th century BC
and silver coinage from the last quarter of the 5th century BC onwards 23 In
21$ BC, at the beginning öf the Second Punic War, the Roman army dísem-
barked at Ampurias, marking the beginning of the Roman period of the pen-
ínsula. The war ended with the expulsion of the Carthaginians from Spain in
266 BC. Ìn 197 BC the provinces of Hispania Citerior and Hispania Ulterior
Strabo refers to the city as a Massiliote colony (3.4.8). Sa ~tiago, taking all sóurce materials into
coa µ
τsiςeτation, advocates a Phocean founding (1994, 63-64). Maiseilles was also a Phocean colony.
S~τιtiago 1988; Santiago 1994, 69.
17
were created. The status of the Roman city of Am?urias, established around
100 BC to the west of the Greek city, is unknown. a Shortly before, or in the
early years of, the Augustan age the different nuclei of the city were political-
ly unified in the Municipium Emporiae as evidenced by coinage (MUNICI
EMPORIΑ), 25
The archaeological and epigraphical source materials leave no doubt that
Iberians lived near and also inside the ancient city of Αmpurías. 26 Palaiapolis
(the initial settlement) was established atop a local settlement, and from the
beginning of the city's existence the presence of Iberians can be documen-
ted.27 From the 4th century BC onwards Iberians appear to have been living
inside the walls of Neapolis (cf. section on the habitation quarter extra-
muros below). Thus before the arrival of the Romans, Ampurias was already
a mixed community. The source material also attests that Iberians were in-
volved in the commercial life of the city. 28 According to Livy, the indigenous
inhabitants of Ampurias were granted Roman citizenship before the Greek
population. 29 A bilingual inscription in Greek and Latin mentions a man
named Noumas from Alexandria. He erected the temple, statues and porti-
coes of Sarapis and Isis. The inscription is dated to the mid-1st century BC. 30
24 Some scholars have suggested that it was a Latin colony. For the debate, see Pena Gimeno
1988.
Probably between 36-27 BC (Fabre et al. 1991, 17-18). The unification is not mentioned by lit-
erary sources.
zc
`Iberian' is used in a geographical sense to designate the various indigenous peoples living in
the coastal area stretching from southern France to southern Spain (for example Gusi/Olaria 1984,
14-16; Richardson 1996, 9-16; Aranegui Gasc ~~1998). Cf. also map in Los Iberos 1998, 40-41. This
term is, however, in dispute. Because of the variations within the Iberian culture and Iberian peoples,
Dominguez Monedero considers the term too general and thus unsuitable (1983).
ηΡ
Sant Marti d'Εmpúries 1998, 18-27. The presence of Iberians at Ampurias is documented by in-
scriptions and cremation graves (Almagro 1952, 63-83; Pena 1988, 17-19; Sanmartí-Grego, E. 1988a;
Sanmartí-Grego, E. 1993a, 21-22).
29
See for example Santiago Alvarez 1994.
29
Liv. 34.9.3.
30 Fabre et al. 1991, 46-48.
Grego, Ε. et al 1994.
18
HISTORY OF RESEARCH
The study of the Greek and the Roman house continues to be character-
ized by the problem of the relationship between archaeological and textual
evidence, especially in relation to the De Architectura written by the Roman
architect Vitruvíus3 8 . This work has been considered almost a lexicon' by
some scholars. Accordingly, his descriptions and terminology have been ap-
plied to the excavated houses in identifying room types and room functions
The uncritical application of the Vtruvian terminology is still upheld, bï ~t i~~
recent research it has been increasingly questioned.
In early scholarship purely hypothetical ground plans of houses were re-
constructed, and as houses were beginning to be excavated from the mid-
18th century (the Vesuvían cities) and the late 19th and early 20th centuries
(Greek cities: for'instance Delis and Priene), the remains were fitted into the
Vitruvian model39 The best and the worst example regarding theGreek
house is Rumpf's interpretation of the insula of the House of the Masks on
Delos.40 He identified Vitruvius' gynaikonitis,. or the women's quarter, in
House C (De115) the andronitis, or the men's quarter, in House B (D ~114),
and the hospitalia (rooms for guests) in Houses A and D (De113'' and
De1M6).4~~ Generally, scholars have been obsessed with the identification of
33
Texts referring to Ampurias are listed and commented on in Almagro 1951. See also Mar/Ruiz
de Arbulo 1993, 461-479.
34 ; Liv. 34.9; Strab.3A.8 ι; -
35 The texts .have. been studied by Pena especially (1985; 1988). Cf. also Domínguez Monedero
1986, 6-7.
36 Ruiz de Arbulo Bayons 1991, 476-477. The Roman city is divided into two parts by a transver-
se wall.
3' Cf. Aquilué 1997, 47-48. .
8
In ten books and written during th ~~reign of Augustus. The sections referring to layouts and
rooms in particular are: the Greek house:. 6.7; the Roman house: 6.3.
; 9 For an overview o£.the Roman house, i.e. the Pompeian house, and Vitruvíus, see Allison 1593,
1-2 and 6-7. For the Greek house, see Nevett 1999, 21-29.
40 Rumpf 1935:
42
For example Jameson 1990, 93 and 104; Nevett 1994; Nevett 1995.
43
Trümper 1998, 15-16; Dickm ann 1999, 23-39.
44
Leach 1997.
45
Goldberg 1999. She gives an example from Plato's Protagoras where a storeroom is converted
into a bedroom for visitors (1999, 150).
46 Priene: Wiegand/Schrader 1904, 289. Olynthos: Robinson/Graham 1938, 141-151.
Bulla 1970.
48
Krause 1977.
49 Hoepfner/Schwandner 1994,X111.
50 Knell 1988.
51 Jameson 1990.
20
sents the private life of the family and is a closed unit. The absence of pro-
nounced variation in size of the houses and the presence of only limited lux-
ury suggested a relative equality. Moreover, Jameson questioned the useful-
ness of the traditional pastas and prostas designations. 52
The nature of social relationships in Greek households, i.e. between men
and women, household members and outsiders, is the subject for the latest
comprehensive study of Greek houses during the Classical and Hellenistic
periods by Nevett. She also reassesses the traditional typology. 53 Instead of
focusing on one particular architectural element, such as the pastas, prostas
or peristyle, she makes the overall organization of space serve as a criterion
for classification 5 4 Thus her "single-entrance, courtyard house" encompasses
both the pastas, prostas and peristyle houses, as these share the following
characteristics: a centripetal plan, a single entrance, a central open space oc-
cupying the majority of the plot and often with a portico (or a full peristyle),
and also often a dining room identified by its raised borders and decoration
(the andron). This reassessment of the typology is most welcome, as the
problem of distinguishing between a pastas and a prostas is illustrated by the
houses from Halieis 55
The study of the Roman house is inevitably linked to the Vesuvian houses
and especially the Pompeian ones. Owing to Vitruvíus' description, the at-
rium house is regarded as the Roman house per se. Like the Greek house, re-
search has concentrated on typology and again an evolutionary approach has
predominated.56 This approach is also reflected in the studies on the relation
between the Greek house and the Roman house, i.e. whether the influence
was from the West to the East or vice versa5'
In Vitruvius' ideal atrium house several roof systems were employed, both
with openings (Tuscan, tetrastyle, Corinthian, displuviate) and without
openings (testudinate) 5 8 Moreover, he reports that the atrium is spatially re-
lated to a number of rooms (fauces, alae, tablinum, and peristyle). All these
elements combined are considered to constitute the typical Roman atrium
house. In relation to the archaeological material this has meant, as Wallace-
Hadrill has shown, that when a central circulation space without an implu-
hium is found, for example in the so-called row houses, it is automatically in-
terpreted as a testudinate atrium.59 Wallace-Hadrill has argued that these
spaces might as well have been unroofed, thus being open courtyards, and he
58 Vitr. 6.3.1.
59 Wallace-Hadrill 1997, especially 221-231. This covered space may also be called atrium (Ling
1997, 25). For the row houses, see Nappo 1997, especially 99-100 (roofed or unroofed circulation
space).
21
uses the term "open atrium" to designate them. This term, however only
adds to the terminological confusion. If they were open courtyards, then they
should be classified as courtyards and called such. Irrespective of the termi-
nology, it is important to recognize the variations within the Roman house.
This variety is also documented by the analyses of houses from Herculaneum
made by De Kind, which show that the Vitruvian atrium house was not the
standard. ó0
Leaving the textual evidence aside, detailed analyses of architectural de-
sign and interior decoration, wall-painting in particular, have been employed
in order to define room functions and the use of space in general, including
social relationships within the house. 61 In his studies of the social structure of
the Roman house, Wallace-Hadrill has stressed the role of the house as a tan-
gible symbol of the owner's status and social position. This was achieved by a
subtle interplay of architecture and decoration differentiating between
rooms and areas, for example between low status (service zones) and high
status areas.
Another way of identifying room functions is by analysing artefact dis-
tribution, and the conclusion reached by both Allison and Berry is that many
rooms were multifunctional, including atria.62 For example, common house-
hold objects have been found in atria, which calls into question the purely
ceremonial function of the room. The problem with this method is, however,
that portable objects may have ended up in any given room by coincidence.
Research on the Carthaginian house is relatively new. Prior to the interna-
tional campaign to save Carthage, hardly anything of the habitation was
known, and if this initiative had not been taken, we might still have been ig-
norant. There is a tendency in classical archaeology to focus on the Greek
and Roman world, forgetting the other actors on the Mediterranean scene,
such as the Carthaginians and indigenous peoples. G3
In recent years the interrelation between houses and people, between the
built environment and behavioural conventions, has been the focus, especial-
ly within interdisciplinary studies encompassing a number of dísciplines. 64
The collection of articles Domestic Architecture and the Use of Space brid-
ges the disci p6 lines of prehistoric and classical archaeology, ethnography and
architecture. 5 It focuses on the interaction between domestic structures and
spatial organization, and especially the role of culture in this interaction.
Many ways of viewing architecture and houses were presented in the book.
Wilk saw the house as a consumer good. In this way the human actors them-
selves and the processes by which people balance various options are in fo-
6°
De Kind 1998, 188-193. Examples of his eight types are shown in fig. 11.
61
For example Wallace-Hadrill 1988; Roman art 1991; Functional and Spatial Analysis of Wall
Painting 1993 (Part I); Wallace-Hadrill 1994.
62
Allison 1993; Berry 1997.
63 Studies on indigenous houses of southern Italy (Apulia and Luc ania) have been done by Russo
Tagliente (1992).
69 For example Housing 1989; Domestic Architecture 1990; About the house 1995.
65
Domestic Architecture 1990.
22
THEORETICAL APPROACH
66
Wilk 1990, 35.
6
S an ders 1990.
6a
"The house is an extension of the person" (Carsten/Hugh Jones in About the house 1995, 2).
69
Carsten/Hugh-Jones in About the house 1995, 1-6.
70 Allison 1993, 1-2.
23
tíva" or "Wohneinheit" because these terms are not associated with a certain
physical appearance (i.e. size, number of rooms, materials etc.)." Moreover,
such neutral terms allow shops that also served as habitations to be included.
Since the present work is in particular concerned with the occurrence of the
combination of different architectural designs, shops or workshops consist-
ing of only one or two rooms are not included in the material.
According to the Concise Oxford Dictionary, one meaning of `house' is a
building for human habitation, and this broad definition is employed in the
present work.' Z A house fulfils some basic needs for shelter, including a place
to dwell, sleep and eat, but other functions may also be attached to the
house, such as commercial activities. Moreover, a house is often more than
just a physical structure; it can also be a symbol of status and social position.
A house may be identified in the archaeological record by looking at a
number of criteria: typological criteria such as ground plans; functional cri-
teria including evidence of domestic activities (for instance finds of pottery,
hearths/braziers, spinning and weaving implements); furniture; location
within the urban setting. It is obvious that a minimum of elements have to be
present, but it is unrealistic to believe that it is possible to establish a defini-
tive check-list perfectly applicable to any ancient material of which we have
only fragments. Each case has to be viewed within its own context, and cases
of doubt will occur. Accordingly, the present work includes both fully and
partially excavated houses, houses with commercial purposes attached, but
not shops or workshops of one or two rooms with residential function at-
tached. Although the exact function of partially excavated houses is un-
known, they are included because of their location within residential areas.
METHOD
For convenience the terms `Phoenician' and `Punic' are defined in the fol-
lowing way: `Phoenician' refers to Phoenicians of the 1st millennium BC in
the East and of the 8th to 6th centuries BC in the West; `Punic' designates
the western Phoenicians from the middle of the 6th century BC onwards. 82
However, the terms are debated, and López Castro rejects the use of the
term `Punic' for the Iberian Peninsula because it implies the recognition of a
political and cultural unification effected by Carthage. 83
Pavements: Due to the lack of consensus on terminology and the need for
a re-evaluation of the material, a separate chapter has been dedicated to the
pavements (cf. Appendix). The terms and their definitions reached in the ap-
pendix can be found in the glossary.
For the problems of the four Pompeian Styles established by Mau, see A llroggen-Bedel 1992.
$' Ling 1991.
86 lieto Prieto 1977, 852 and 856-866.
2. DELOS. THE SACRED ISLAND IF THE EAST
HISTORY OF RESEARCH $'
n Focusing on the houses. For details of the history of research, see Chamonard 1922-1924, 9-14;
Laídlaw, W.A. 1933, 5-11; Plassart 1973; Kreeb 1988, 7-11; Trümper 1998, 4-8.
s
Paris 1884, especially 485-486 (De ΙI6 and De1M1).
n
Chamonard 1922-1924.
90
Bulard 1908.
91
Homolle 1884; Hatzfeld 1912; Roussel 1931.
92
Cf. Bruneau 1968; Bruneau/Ducat 1983.
93 Bulard 1926a; Bulard 1926b; Michalowski 1932; Laumonier 1956; Bruneau 1972.
ínhabítants.95 The rich material at hand should rightly encourage such stud-
ies. However, Rauh did not take into consideration the diachronic perspec-
tive of the houses. Based on the reconstruction of building phases (neglected
so far) and the recording of the architectural and decorative equipment,
Tríimρer's monograph aims to elucidate changes in the Delian "Wohnkul-
tur".9 The main conclusion reached by Triimper is that many houses origi-
nated as standard units (standard size and internal layout). 97 Moreover, the
houses tended to expand vertically often with independent habitation, units,
and thus the best comparative material is that of the Vesuvian cities where
the same insu la complexes are found. That there is still new knowledge to be
obtained is demonstrated by the recent, and final, publication of the houses
from the flot des bijoux (De1N12-De1N18), Îlot des bronzes (De117-
DelN11) and the Maison des sceaux (De1123). 98 The publication deals with
topography and architecture, but important issues, such as identification of
room functions and inhabitants, are also discussed.
The city is situated on the north-western part of the island and developed
fan-like around the Sanctuary of Apollo towards the north, south-east and
south with the so-called Sacred Harbour to the west of the sanctuary. 99 Apart
from the public buildings and monuments, our knowledge of the city prior
to the period of the Second Athenian Domination (167/166-69 BC) is poor.
Residential districts appear to have been confined to the north and the south
of the sanctuary, and the built-up area was in any case limited. 100 Traces of
habitation from the 4th and 3rd centuries BC have been located in the area
of the Agora of the Compitaliastai/Hermaistai, and some houses situated in
insula III of the Theatre Quarter may go back to the 3rd century BC. In
101
the Northern Quarter structures, interpreted as shops and workshops, ap-
pear to date from the period of independence. 102 Public buildings of the 3rd
century BC include three porticoes in the immediate vicinity of the Sanctu-
ary of Apollo, at least one palaestra to the north, a gymnasium to the north-
east and a theatre to the south of the sanctuary. Moreover, the Agora of the
Delian was begun. Sanctuaries for foreign gods, e.g. Sarapieion A, were also
erected in the 3rd century BC.
Great building activity took place in the 2nd century BC, for example new
palaestrae and no less than three agorae: the Agora of the Compitaliastai/
03 The earliest dedications date to the middle of the 2nd century BC and were made by the Her-
108 The division and order follow Bruneau 1972. One exception is that the Area of the Sanctuary
of Aphrodite forms a separate quarter in this work instead of being part of the Theatre Quarter. For
plans, see Bruneau/Ducat 1983, pls. I-IV and VI-VII.
32
Note on chronology: The years, 167/166 and 69 BC, do not provide a se-
cure terminus post quern and terminus ante quern respectively. Thus houses
may be ~lder than 167/166 BC, and some continued to be used after the last
attack on the island. As regards the time of construction, however, a fairly
rigid dating has been employed among scholars, especially with respect to
the Theatre Quarter: either before 167/166 BC or after 167/166 BC. 1~~ Strati-
graphical evidence for dating the time of construction, the building phases or
the time of abandonment is only documented in some cases. 113 In the North-
ern Quarter soundings beneath the Îlot des bijoux revealed not only earlier
structures, interpreted as shops and workshops, but also ceramic and numis-
matic material belonging to the late 3rd century BC and first half of the 2nd
century BC. 114 The year 128/127 BC is referred to in an inscription on a seal
from De1N23, and thus a dating of the house to the third quarter of the 2nd
century BC is most plausible. This dating might also point to an earlier date
of the flot des comédiens which is traditionally dated to the very end of the
last quarter of the 2nd century BC. 115 The houses of the Stadium Quarter are
dated no earlier than the second half of the 2nd century BC. Den can be
dated no earlier than 150 BC. The insula of the House of the Masks is dated
to the second half of the 2nd century BC.
ARCHITECTURE
ARCHτrεcτVRAL ANALYSτS
109 House V situated in the Ilot des bronzes is not included, as it was transformed into a shop
complex.
11° House B in insula I was a restaurant.
111 House V and Y from insula IIΙ are not included. The ground plan of the former is too unclear,
and the latter consists of two separate units, each having two rooms.
112 Before 167/166 BC: Raeder 1988, 327. After 167/166 BC: Hoepfner/Schwandner 1994.
113 For a summary of the chronology of quarters and houses, see Triimper 1998, 121-126.
114 Siebert 2001, 142-145.
115 Siebert 2001, 141-142.
33
Ground plan €(Table 1): Ninety houses can be analysed. 116 Three main
ground plans occur: rectangular, square and irregular. The irregular ground
plan is clearly the most common with fifty-three examples, followed by the
rectangular ground plan (thirty-two) and the square (five). Completely regu-
lar ground plans are rarely found, and there are different levels of irregularity
within this group, such as trapezoidal forms. Rectangular and square houses
with minor irregularities are for example De1N6, De1N9, De1N11, De1N13,
De1N17, De1N18.De1N20, De1121, De1123, De1ST2, 'De1ST3, DelST4,
DelST7, De1P5, Delll, DeΙI4, De1Τ7, De1T10, De1T20, De1T39 and De1T40.
It comes as no surprise that the majority of the houses .located in the crowd-
ed Theatre Quarter are irregular, but even a regular insula is no guarantee
for regular house plots, for instance the insula of the House of the Masks
where all four houses are irregular.
,
Dimensions (Table 1 and Table 2): 117 The area varies from 53 m 2 (DelT43)
to 866 m2 (Dell19). The largest group of houses has an area between
101-200 m2 (forty-two). The second largest is houses of 201-300 m 2 (fifteen),
followed by houses of 301-400 m 2 (nine). Only eight houses are smaller than
100 m2. Houses of more than 400 m2 are located in all quarters, apart from
the Stadium Quarter, the Peribolos Street and the Area of the Sanctuary of
Aphrodite. Some of the large houses owe their huge areas to the joining to-
gether of more than one house (for example De1T3 and De1T16), while oth-
ers had an original layout of such large dimensions (for instance De1N21,
De1122, De1129., De1Ml and DelS2). Due to their large areas, houses such
as Dell 19 and DelS2 may have had other purposes than habitation only,
and they may have functioned as clubhouses for associations. 118
Entrance room (Table 1): Eighty-four houses are provided with an en-
trance room. A house having two entrance rooms occurs twelve times. DelI2
had two entrance rooms as well, but here there appears to have been two
habitation units. De1S2 consisted of (at least) two units. Sometimes it is pos-
sible to distinguish between a primary and secondary entrance room, for ex-
ample in De1T41 where room a was the primary entrance, as it is broad and
provided with niches in contrast to the narrow room b with a drain running
through it. The entrance room generally takes the shape of a small rectangu-
lar room, but could also be a long narrow corridor, for instance De114. In
other houses an alley was privatized, for example De1T4. The frequent oc-
currence shows that it was a room that was appreciated.
Circulation space (Table 1 and Table 3): The most common circulation
space is the open courtyard without porticoes with a total of sixty-three ex-
116 Twenty-one houses cannot be included due to the small extent of excavation: De1N1, De112,
De113, De1115, De1124, De1126, De1128, De1129, DelST5, De1P3, Dell'4, De1P6, De127, De128,
De129, De1I6, De112, De1T19, De1T29, De1T49, De1T50.
117 All figures are taken from Tramper 1998.
118 De1119: Kreeb 1988, 155-157. De1S2: Marcadé 1969, 399-405; Bruneau 1970, 633.
34
amples and two possible courtyards. It occurs in sixty houses, and can prob-
ably be identified in two more houses. In the Theatre Quarter it is clearly the
predominant type of circulation space. The second largest group is the peri-
style occurring in twenty-three houses. ΥΥ9 The number of columns varies.
Most common is the peristyle with a column in each corner (2 x 2) (seven
houses). The largest peristyles have 4 x 4 columns and are naturally found in
large houses, De1Ν19, De1Ν21, De1Ν27 (main unit), DelI6 (partially excava-
ted), De1Ml, De1M4, De1T2 and DelS2 (main unit, special arrangement).
Two of these are so-called Rhodian peristyles, De1M4 and Del Ρ2, with high-
er columns on one síde. 12t In four houses there is no stylobate (DelST6,
De1ST7, De1T4, De1T5). Α courtyard with one portico occurs in eight houses
and possibly in two additional houses. Α courtyard with porticoes on two
sides is extant in four houses, while six houses have a courtyard with porti-
coes on three sides. 121 The Doric order is clearly prevalent with a few exam-
ples of Ionic. 122 Pillars occur in De1 Ν22 (combined with low wall), De1T4
and De1T40. Low walls and pillars with columns on top are used in De1113.
In some houses one of the porticoes Is deeper than the others and/or it may
be extended ín one or both ends into a kind of recess (De1 Ν19, De1Ν22,
.
119 For details on dimensions, materials etc. 0f courtyards with peristyles and porticoes, see Trüm-
per 1998, table 1.
~29 This type of peristyle is described by Vitruvíus (6.7.3).
121
De115 and De1T46: Stylobate on three sides, but only supports on two sides.
w
Cf. Trümper 1998, table 1. - -
~2; Small recesses about one metre deep or less are not regarded as rooms.
35
had the N1-arrangement, but either one of the rooms at the back was given
up or an extra room was added, thus originally there was a total of twenty-
eight examples, 124 The wish to have a more spacious room at the back, in
particular in relation to narrow plots, may explain the frequency of 14 (only
one room at the rear). There can be no doubt that the arrangement with a
large transverse front room and two smaller rooms at the back was stand-
ard. 125 This is emphasized by the fact that the arrangement occurs on all sides
of the circulation space (El, Si and W1), and with these added the number
of the complex reaches thirty-two, and including the altered arrangements
the total is thirty-six examples. 126
The symmetrical arrangement of 12 in De1121 should be noted. The
smaller room on either side is entered from the central room only. It is also
found in De1ST1 (E2), but here it is a more irregular version. Also the ar-
r an gement of De114 should be commented on (18). At the back of the main
room is a triple room complex consisting of a central room opening onto the
main room and giving access to a smaller room on either side.
In general, the Delian main rooms are broad-rooms of varying sizes. The
large ones are often provided with a central door and a window on either
side and may be on the axis of the entrance, for example De1N19, De1121,
De1122, De1M1, De1T2, De1T7 and DelT41. Del13 even has two, one on the
north side of the courtyard and one on the east side.
Other rooms (Table 1): Service rooms, í.e. latrines, bathrooms and kitch-
ens, are common in the Delian houses. Latrines are most frequent, occurring
in sixty-one houses, and ten houses may have had one. 127 Some houses appar-
ently shared latrines (De1N15 and De1N16, DelPl and De1P2). In most cases
they are located near the entrance area or in the entrance room itself. The
distinction between a bathroom and a kitchen is often blurred, especially in
those houses in which the presence of drains, waterproof pavement types
and/or hydraulic mortar on the walls is not supported by other finds or in-
stallations. This is the case in twelve houses, and in fact these rooms may
have served various purposes. Moreover, either a bathroom or a kitchen is
likely to have existed in seventeen houses. Separate bathrooms are present in
eight houses. 128 The majority of these are provided with an individual bath-
tub of the so-called "à sabot" type (De1122: room AL; De1 Τ2: room C;
De115: room j; DeiAl: room c; De1S1: room XVI Ι). In room b' in De1T31
and rooms f and i'in De1 Τ33 only the gap in the floor reserved for the bath-
tub is left. Circular structures, interpreted as sweat-baths, are found in two
houses (Del122: room AM; De1T6: room i).529 Two houses had two bath-
"" In De1Τ33 one of the rooms at the back was subdivided, and a bathroom (i') was installed.
'15 Chamonard considered it the most characteristic element of the Delian houses (1922-1924, 167).
126 The 1\11-arrangement is also present in De1T16 (not main room) and in the eastern part of
De1T25 (not main room).
n The latrine in De1113 belongs to the independent habitation unit on the first floor.
va De1122 (two), Deli7, DellS, DeΙT6, De1T31, De1Τ33 (two), DelAl, De1S1. See Trümper
1998, 347, fig. 82; 348, fig. 83; between 348 and 349, fig. 85.
129 The function of the circular structure in room EP in De117 is disputed: sweat-bath or decanta -
rooms, De1122 (bathtub and sweat-bath) and De1T33 (each room with a
bathtub), and in De1T6 the bathroom was provided with an anteroom (room
i'). A kitchen has been identified in four houses (De117, De1122, De1123 —
upper storey and De1I2).
Dining rooms have been identified in twenty houses, and there is one pos-
sible (Table 5). The identification is based on the pavement design marking
the position of the couches. As shown in Table 5, the rooms used for dining
can be categorized in two main groups. The first group consists of large
broad-rooms which often are identical with the main rooms of the houses
(De1N1, De114, De1N16, De1121, De1122, De1I4, De114, De115, De1T2,
De1T24, De1T25, De1T31). These broad-rooms are in most cases provided
with a central door or slightly off-centre door, and adjoining rooms are com-
mon. Furthermore, some of the rooms employ a pavement design with two
or three emblemata. All this points to a flexible use of the rooms, and alter-
native arrangements of the couches must be taken into consideration, for in-
stance two Pi-shaped sets of couches, one at either side of the room. 130 This is
especially evident in De1122 with two emblemata or in DelN4 where an exe-
dra with a room on either side opens onto the back wall of the dining room.
The second group includes square or squarish rooms, i.e. similar to the
andron from domestic contexts. Most are of modest size, but in De1N16 the
room has an area of c. 56 r2, thus resembling the ones found in Macedonian
palaces. Some of the rooms have a wide opening towards the circulation
space, for example room EE in Del Ν7 and room R in De112 1. Three-couch
rooms also occur, for instance rooms Q and R in De1121. The only room
with the characteristic raised platforms is room G in DelI2. The floor design
in room f/AN of DelS2 should be singled out, as it represents an early, or the
earliest, version of the characteristic T+U design employed in Roman trick-
nia. 13 " Here the length of the horizontal bar of the T corresponds to the
width of the door opening. Several houses have a combination of the two
room types (Dell 1, De1N4, De1N21, De1M4, De1T2, De1T31 and maybe
De1S2).
Dining might also have taken place in rooms where the position of the
couches is not detectable in the archaeological record. This goes for a num-
ber of square and squarish rooms often located close to the entrance, i.e. the
same location as many domestic andrones from the Classical period, for in-
stance room EW in De118 and room T in De1N14. Siebert suggests that this
room may be the only specialized room in the Delian houses. 13
A cult room has been identified to the right of the southern courtyard in
DelP9 due to the find of a relief. Also in De1123 there may have been a cult
room on the upper storey, but the context of the relief cannot be definitively
answered.
Shops are not common as an integrated part of houses. They occur in rela-
130
Dunbabín 1998, 84-86; Westgate 2000, 414; Siebert 2001, 123.
133
The design is especially known from the 1st century AD onwards (cf. Dunbabin 1998, 92). For
the design in this room, cf. Tríímper 1998, between 317 and 318, fig. 59; description by Bruneau
1972, 310.
132 His pecus minor (Siebert 2001, 122-124).
38
tun to twelve houses of which the majority are situated in the Theatre Quar-
ter. A terracotta workshop was installed in De1T34. Find complexes of sculp-
ture from within De1119, DelP1 and De1P2, including unfinished works,
have been interpreted as evidence for a sculptor's workshop. For De1 Ν19, it
has been suggested that the sculpture was carved on the spot to avoid dam-
age during transportation and therefore, the identification as a workshop
should be rejected.' 33 The most plausible explanation, however, is that the
house at the time of abandonment functioned as storage space for various
sculpture. The six basins located in room e of De1T21 have traditionally been
linked to a fullery or a dyer's shop. Trümper, however, suggests that this part
of the house may have been a bath complex for public use.' 34 If the basins
were bathtubs, three were for bathing in a seated position, and three for
bathing in a standing position. Hot water would have been provided by the
oven in the south-west corner of room d, and room b could have been the
changing room. It is not clear whether this part belonged to the rest of the
house. Room c in De1T23 has two entrances directly from the street and a
third internal door. The pavement design shows that the room functioned as
a dining room. Because of the openings towards the street, Trümper has sug-
gested that the room may have been a room which could be rented.' 35
Upper storeys (Table 1 and Table 6): Upstairs rooms were fairly common,
as they are documented in nearly eighty houses either by finds and/or by re-
mains or traces of staircases. Moreover, they are likely to have existed in
eight houses due to the presence of indicators such as downpipes and/or
beam-holes. Judging from criteria such as external staircases and internal
staircases with a separate access (for instance the presence of a double
threshold at the entrance), a number of independent habitation units existed
on the first floor (De1N7, De118, De1 Ν9, De1N13, De1N18, De1121,
De1122, DelST4, De1P9, De113, De1M4? De1M6, De1T1, De1T2, De1T3,
De1T5, De1T11, De1T18, De1T21, De1T39, De1T40, De1T41, De1T47). Con-
sequently, the first floor cannot uncritically be considered the private part of
the house.
Water supply (Table 1 and Table 7): The water supply of the Delian houses
was secured by means of cisterns and wells. In general, the cisterns are lined
with masonry and hydraulic mortar and the wells with masonry. The cover of
the cistern takes the shape of flat stone slabs supported by beams or arches.
According to ground plan and section, the following types of cisterns are
present:
Type I: Quadrangular, vertical walls.
Type 11: Circular, vertical walls.
Type III: Irregular.
DIACHRONIC PERSPECTIVE
that they were erected on two or three sides only. In De1T31 some of the
porticoes are only 0.8-1 m deep. The original peristyles and porticoes are
characterized by the use of marble for stylobate and columns, whereas the
added peristyles and porticoes often employ reused material, while marble is
rarely found. It is clear that the presence of peristyles or porticoes regardless
De1T3, De1T9, De1T16, De1T17, De1Τ18, De1Τ21, De1Τ25 (House S), De1T30, De1T31, De1T34,
De1T35, De1T36, De1T37, De1 Τ38, De1T45, DeΙT46, De1T47, DelAl, De1S2. Possible: DelΝ6, De117
(House II), De1N10, De1P1, De1T11, De1Α2 (Trumpet 1998, 34, note 185).
40
CONCLUSION
Clearly defined house types cannot be established due to the various com-
binations of architectural elements, i.e. a certain ground plan is not correlat-
ed with a certain circulation space etc.). 139 From Table 1, 3 and 4a it is clear
that the most common elements are an irregular ground plan, a vestibule, a
courtyard without porticoes as circulation space and the main room complex
Ni with its variations El, S1 and W1. Furthermore, the majority of the hous-
es are provided with a latrine, an upper storey and individual water supply.
As a result of enlargements, some houses had more than one circulation
space. Although the character of the circulation space may serve as a criter-
ion for a classification (i.e. courtyard houses, peristyle houses), this can only
be a general one because of the variations within the groups. The central
courtyard with or without porticoes is an omnipresent feature.
The peristyle court forms the second largest group as circulation space
(twenty-three examples). Peristyles as part of the original layout are found in
De1N19, De1N21, De1125, De1127, DelI6, DelMi, De114, De1T2 and
De1S2. In De1ST6, De1T3, De1T5, De1T7, De1T16 and De1T47 they were ad-
ded later, while seven examples are disputed (De1N6, DelST2, Den.,
De1T4, De1T31, De1T40, De1T41). 140 Courtyards with a portico on one, two
or three sides occur in nineteen, maybe twenty houses. The original layout of
De1122 and DelI2 included porticoes, while it is not clear whether the porti-
coes belong to the first phase or are a later addition in the followíng houses:
De1N13, De1129, De1ST4, De113 and De115. In eleven houses porticoes were
added in a later phase: De1N1, De1N7, De1N12, De1123, Delll, De115,
De1M6, De1T6, De1T17, De1T21 and De1T46.
The conclusion reached by Trümper is also that the houses cannot be di-
vided into a clear typology, but she has identified one type, the so-called "ka-
19 Already Chamonard concluded that the Delían houses do not constitute distinct house types
(1922-1924, 100-102).
14° Cf. Triimper 1998, table 1. The building history of Del12 has not been investigated (partially
excavated).
41
The traditionally defined Greek house types, i.e. the pastas, prostas and the
peristyle house, are all represented on Delos. However, the problem of dis-
tinguishing between a pastas and a prostas is also illustrated by the Delian
fl ouses. 145
i
141 Tramper 1998, 107. According to Kreeb, there is no Delian "Normalhaus" (1988, 13).
142
Tramper 1998, 107, note 544.
143 Tamper 1998, 109, note 553.
sets of rooms to the north and the courtyard to the south. The middle part of
the pastas opens onto the courtyard (one column) The present house is a re-
sult of the amalgamation of three houses and apart from the pastas, it does
not share characteristics with De1125. Originally both sets of rooms to the
north. were N 1-arrangements, but it only survived in the western suite
(slightly modified). Because of the great similarity between De1125 and the
Olynthian houses, the architect and/or the (first) owner of the house must
have been acquainted with the pastas houses of the Greek mainland It is in-
teresting to note that the Olynthian houses were destroyed in 348 BC, and
pastas houses outside Olynthos are not numerous. Isolated examples are
known from Attica (5th and 4th centuries BC) and Pella (Hellenistic peri-
od). ~4' On Delis a rural house was provided with a pastas in its second phase
dated to the 2nd century BC. 148 Moreover, a pastas like area is found in hous-
-
es from for instance Athens (5th century BC) and Phaistos (Hellenistic peri-
od). The internal layout of these houses has nothing to do with the well-plan-
ned houses of Olynthos, Attica, Pella and De1125. The perfect layout of
De1125 and its similarity with the Olynthian houses are, however, still re-
markable, and it might indicate that the use of pastas houses was more wide-
spread than the material so far has documented. This similarity applies both
to ground plan (square), size -(De1125: 307 m 2 . Olynthos, standard plot: c.
289 m2) and internal layout. In any case, De1125 documents a strong tradi-
tion of organizing space in the Greek house.
Several other houses have a more or less open-fronted area, traditionally
called a prostas, between the main room and courtyard. This applies to
De1N16, De1117, De1123, De1124 and De1T45. In the latter the arrange-
ment of the rooms is identical to the prostas houses of Priene and Abdera:
main room (room h) with an open-fronted anteroom giving access to a suite
of rooms on the northern side (a dining room and a shop) . 149 Similar room
arrangements can be found in De1 Ν16 and De1123 (column supporting the
gallery on the first floor).
Another characteristic Greek feature is the triple room complex consisting
of a large transverse room and two smaller at the back, the so-called
11-arrangement in the present work. According to Krause, it can be traced
back to Geometric times with continuity right down to the Hellenistic peri-
od. 150 In the late Classical age it occurs at Eretria in houses dated to the be-
ginning of the 4th century BC, for instance House II and the House of the
Mosaics. 151
Square or squarish dining rooms of varying sizes, well known from domes-
tic architecture and Macedonian palaces, are also common. The raised bor-
der, however, only occurs in De112 (room G). Three-couch rooms, which are
rare in general, for instance rooms Q and R in De1121, can be paralleled to
152
Room Q: 8.60 m2. Room R: 8.30 m2. Room 5 in the House of the Mosaics measures 3 m x 2.90
m and the andron in House Bu 7 3 x 2.65 m (Ducrey et al. 1993, 45; Robinson 1946, 128-129).
153
Nielsen 1994, 31-72 and 112-129; Nielsen 1998, 107-109.
54 For plans, see Nielsen 1994, 56, fig. 27 and 117, fig. 60.
ass For plans, see Westgate 1997-1998, 105, fig. 8; Nielsen 1994, 94, fig. 48.
156
Nielsen 1994, 82, fig. 41 and 84.
157
Bruneau 1978, 125-132.
158
Bruneau 1978, 138-145. In De1N4 there is no documentation for upstairs rooms, thus the por-
tico apparently had a purely prestigious function.
159
See also Wallace-Hadrill 1994, 108. Vitr. 5.1.1; Plín. nut. 35.37.113.
160
Raeder 1988, 344 and 344, note 61. Aristot. oik. 2.1347.4-7.
44
'62 Wall decoration: Chamonard 1922-1924, 357-391; Bezerra de Menses 1984; Bruno 1985;
Kreeb 1988; Alabe 1993; Alabe 1995. Pavements: Bruneau 1972; Baldassarre 1976; Joyce 1979; Kreeb
1988; Guimier-Sorbets/Henna 1992; Westgate 1997-1998, especially 108-115.
These rooms are not registered in the catalogue.
'64 Andreou considers it unsuitable to employ `First Style'on Greek wall-paintings because of the
differences between these and the Pompeian wall-paintings (1988, 214).
For categories, see Chamonard 1922-1924, 359; Alabe 1995, 195-196. For illustration, see also
Bruneau/Ducat 1983, 82, fig. 15.
166 Cf. Kreeb 1988, 55, table IX and 57, table X.
45
between the ground floor and the upper storey cannot be reconstructed.
Houses particularly rich in murals on the ground-floor are De1N1, De1121,
De1N22, De1123, De1125, De1127 (main unit), De1ST3, De1ST6, De112,
De1I3, DelIS, De114, De1M5, De1M6, De1T2, De1T3, De1T6, De1T31,
De1T40, De1T41 and De1T42. It is not a coincidence that they all happen to
be houses of a certain size, i.e. that they have many rooms. The smallest
houses are DelΜ6 and De1T42 belonging to the group of 101-200 m 2 . There
are also large houses where wall-paintings are few, but of course the state of
preservation must be taken into consideration, for example Dell 19 and
De1Ml.
The Delian examples of the First Style including the friezes correspond to
the paintings of Hellenistic Greece in general, for instance murals from
Athens and Knidos (scenes of everyday life and mythological scenes). 167
The only example of the Second Style came from the upper storey of
De1124. The find comprises more than eighty fragments depicting a Corin-
thian portico decorated with festoons, and birds shown in the lower zone. 168
167
Athens: Wirth 1931, 55-56, pl. XX.1. Knidos: Love 1972a, 65, pl. 16, fig. 9; Love 1972b, 397,
399, ill. 4.
168
Alabe 1991, 34, figs. 1-3.
169
Thus following Bruneau (1972, 32).
v° For example De114, De1T21, De1T24, De1T25 and De1T45 (Bruneau 1972, between 256 and
257, fig. 204; 276, fig. 229; 280, fig. 234; 282, fig. 237; 296, fig. 260).
46
with figural scenes or parts of such panels are documented in Dell 13 (up-
stairs room), De1 Ν16 (ground floor), De1122 (ground floor and upstairs
room), DeΙI4 (upstairs room), Del14 (ground floor), De1T31 (upstairs
room), DelΤ41 (ground floor) and DeΙS2 (ground floor). Moreover, emblem-
ata have been removed in the following houses in ancient time: De1121
(rooms Q and R), De1122 (room AE), Dell2 (room G); De1I4 (room F),
De1Τ2 (room k) and De1S2 (rooms h/ ΑΒ and f/AN). Both opus tesellatum
and opus vermiculatum mosaics are more common in upstairs rooms than on
the ground floor.
Common geometric motifs are waves, stepped triangles or meanders,
which often decorate the borders, cubes in perspective and the lozenge pat-
tern. 171 The latter design is employed in for instance De1N16 (doormat) and
De1T41 (black grid on a white ground) 172 On a fragment from an upstairs
room of De1T4 1 stepped triangles appear to form a field with four rows of
both bl~ck and white triangles. 1'3 Chequer-boards in black-and-white with
the tesserae laid obliquely to the walls of the room decorate central fields or
entire floors, for instance De1N1 and De114. 174 The same motif is also used
as a frame, for example De1M1. Among the floral motifs a rosette forms the
central motif three times: De1128 (peristyle), De1T24 (doormat) and De1T45
(doormat). 175 A wide range of figural motifs occur: maritime motifs (dol-
phins, anchor, anchor and dolphin, trident, Triteness and Eros), theatrical
ones (such as masks), animals (for instance pigeons, leopard/panther), am-
phora (or hydria) and palm, dancing, and centaurs. Also the so-called sign of
Tanit occurs (entrance room of Dell 1). Mythological scenes include Diony-
sos riding a leopard or a Dionysiac figure riding a tiger (De114: dining
room; De1T41: peristyle), Ambrosia and Lykourgos (Dell 13, upper storey),
Athena, Hermes and a seated woman (De1N16, main room). The wheel, fre-
quent in the Classical period and symbol of Tyche, is also found (DelI2). 176
The tessellated mosaics are mainly used in covered rooms, but they are
also found in areas exposed to the weather, such as peristyles (De1127,
DelI6, De1M1, De1Τ2, De1T41) and entire courtyards (DelNl, De114,
De1I5). 17 Judging from the numerous fragments fallen from upstairs rooms,
these were also commonly decorated with opus tessellatum and opus vermicu-
latum. 17 All categories of motifs occur on the ground floor as well as on the
upper storeys. Figural motifs are found predominantly in main rooms and re-
ception rooms, while the courtyards and peristyles are decorated with geo-
metric and floral motifs. The exceptions are the Dionysiac figure in the peri-
style of De1T4I and the Erotes riding dolphins located in the corners of the
Bruneau 1972, 146-149, no. 50, figs. 43-44; 295, no. 299, fig. 258.
173
174 Bruneau 1972, 127, nos. 10-12, figs. 12-15; 146, no. 45, fig. 42.
175 Bruneau 1972, 184, no. 93, figs. 102-104; 283, no. 267, figs. 234-236; 295 and 300, no. 306,
figs. 260-262.
176
For Olynthos, see Robinson/Graham 1938, 290.
177 Cf. Kreeb 1988, 53, table V-VIl.
mosaic from the peristyle in De1M1. Moreover, the two emblemata in the
porticoes of De1Τ2 depict an anchor with a dolphin and a trident.
Pebble mosaic: There are five examples of pebble mosaics: De114 (two
circles), De1N19 (plain), De1P9 (solid lozenge pattern in black and white), 179
De112 (wheel) and De1T45 (plain). The pebble mosaic and opus tessellαtum
are combined in De114 and De1T45 with the pebble mosaic decorating the
outer field. The pebble mosaic and chip pavement are combined in the peri-
style of Deli 19.
Chip pavement: This pavement type is the most common in the Delian
houses; it occurs in fifty-five houses on the ground floor, but only rarely in
upstairs rooms. Chip pavements are used in areas exposed to the weather
(for instance courtyards) as well as in prestigious rooms (main rooms and
other reception rooms). In the latter, chip pavements are most often em-
ployed in the outer field or in the border framing the central motif in opus
tessellαtum and thus emphasizing the carpet-like design. This combination of
chip pavements and opus tessellαtum is documented in several houses
(De114, De1N7, De1118 — upper storey, De1122, De1ST6 — upper storey,
DelIl — upper storey, De1I4 — the main room, De115, De1M4, De1Tl6,
De1T21, De1T23, De1T24, De1T25, De1T31).
Mortar pavement and mortar pavements with tessera design: Mortar pave-
ments are rare. There are five examples without design: De1ST2 (upper
floor), De1P1, DelI2, De1T16 and DelS2 (upper floor). The following mortar
pavements with tessera designs are documented:
Meander? Fragments from upper storey: probably from De1T2. 18o
Crosslets set in rows: DelS2, in two rooms. The crosslet is composed of
four white tesserae and one central black tessera. 181
Equally spaced rows obliquely set to the walls of the room?: DelS2. 182
The largest complexes of tessellated mosaics occur in Del Μ4 (rooms e, g,
h, í) and De1T2 (courtyard, rooms i j, k). Both are large houses with several
,
function of displaying the owner's wealth, the tessellated mosaics taking the
highest position, i.e. those with figural scenes and in opus vermiculatum. In
De1T2 the mosaics are used to guide the visitor towards the main room east
of the peristyle and to stress the importance of the room with a design in
front of the entrance. The emblema in the peristyle of De1T41 is orientated
towards the northern portico. Guests arriving in the house would inevitably
see it on their way to the main room north of the peristyle or as they made
their way round the peristyle, for example to the reception room (room í) lo-
cated on the eastern side of the peristyle.
According to Bruneau, the Delian pavements chiefly date to the late 2nd
and early 1st centuries BC. 183 Thus the mosaics are a relatively late phenom-
enon in the Delian houses.
The Delian houses were in general well equipped with murals and pave-
ments. There are, however, also houses without any interior architectural
decoration, such as the 'lot des bronzes in the Northern Quarter (De118,
De119, De1N11) and in the Theatre Quarter (for instance De1T8 and
De1T17). One might wonder why some of the .large houses have no pave-
ments in the main room. This is observed in De1119, De1N21, De112, De1M1
and De1T41. Possible explanations could be that the houses were in a state
of refurbishment when the island was attacked or that carpets were em-
ployed.
The conclusion is that the interior architectural decoration is homogeneous
and that the houses display common preferences and taste. The wall-paintings
including the figurative friezes represent the Greek Hellenistic tradition. All
motifs depicted on the friezes are Greek. Apart from the Phoenician/Punic
sign of T anit, the motifs of the pavements are traditional for the Hellenistic pe-
riod in the eastern Mediterranean area. Polychrome tessellated mosaics and
chip pavements are characteristic. In De1M1 the mosaic of the peristyle was
executed by Asklepiades of Arados (ID 2497). A foreign element in Delian
floors is the Italic/Roman mortar pavements with tessera design. 184
The majority of the figural friezes came from upstairs rooms. Opus vermi-
culαtum mosaics were also a little more frequent here than on the ground
floor. In the houses from the Stadium Quarter tessellated mosaics were re-
served for upstairs rooms. Since most of the houses here have internal stair-
cases, the ground floor and the upper storey formed one habitation unit, and
these cases together with De1I4 (tessellated mosaics and figural friezes) attest
that finer decoration was reserved for the rooms of the first floor (Table b).
Independent habitation units on upper storeys also display fine decoration
(for example De1113 and De1T47).
'm Simple mortar pavements without decoration are also a Greek phenomenon, for instance at
Olynthos (Westgate 1997-1998, 100).
49
FINDS
De1T41 and De1T44 appear to have been used as storage space for sculpture.
This may also have been the case with De1N19. 187 The head of a portrait stat-
ue was found in De1T41 and the body in De1T4. 188 It is impossible to deter-
mine in which of the houses the statue was displayed. Moreover, the body
was found together- with a complex of statues of gods and Muses which is
traditionally supposed to come from the theatre. ~89 Queyrel, however, has
proposed that all the sculptures may come from De1 Τ41, though his conclu-
sion is that only two pieces of sculpture belong to the house with certaínty. 190
Marble statuary is the most common. The majority of the houses only have
one or two pieces. The largest number (more than thirty) came from DelI2
which is also one of the largest houses on the island. Two sculptures in
bronze were found. They came from two houses situated in the Northern
Quarter, DelΝ2 1 and De1123. The low number of terracotta statuettes, only
85
Apotropaic reliefs are dealt with in the section on religious finds.
186
Catalogues: Harvard 1987, 152-192; Kreeb 1988, 105-333. They do not always coincide.
18' Lime kilns have been found in De1Τ41 and De1T46. For De1119, Jockey has suggested that the
sculpture was carved on the spot to avoid damage during transportation (1995, 88 and 93). See also
Harvard 1987, 113.
188 For this particular case, see Queyrel 1988.
in ten houses, maybe due to the fact that precise provenances are lacking for
the old excavations of the Stadium Quarter and the Theatre Qúarter. 191
The most common themes are mythological figures. Aphrodite and her
followers form the largest group, next -come Dionysos and Dionysiac image-
ry. The third largest group represents Kybele, and Herakles and herms are
also found widely. 192 Herms have been found in De1121, De1123, De1125,
DelST3, De1ST6, Delll, De112 (several), De114, De1T21, De1T34 and
De1T46.
Sculpture has been found in courtyards, peristyles, main rooms and other
rooms, and some were found in situ.193 Several finds came from upstairs
rooms including one in situ from DelI2. 194 in the ground floor most finds
came from the courtyard.
Sculpture in the private sphere appears to have originated from a purely
religious function which may explain why ancient authors do not mention it
in connection with the beginning of luxurious decoration of the Athenian
private houses in the late 5th century BC. 195 The distinction between a reli-
gious and a decorative function is difficult to determine, but in the late Hel-
lenistic period the function appears to have been primarily decorative,
though herms probably kept their religious function. In addition, other stat-
uary from the Delian houses apparently had a religious function, in the form
of apotropaia, votive offerings or cult images. This applies for instance to the
Artemis relief from De1123, the relief depicting Agathodaimon and prob-
ably Sarapis and Isis found in De1P9, the nymph statuette from the niche on
the south side of the courtyard in DelI2 and the Hermes herm from the sec-
ond floor of the same house. Furthermore, a base with a triangular hole was
found in situ in a niche in room c (main room) of De1 Ρ36. This points to the
existence of a Hekateion.
The location of display was often carefully chosen in order to impress visi-
tors to the house and to emphasize certain rooms. This is obvious in Dell2
where the statue was displayed in the main room (only the base is left). In
Del14 the herm stood in the north-east corner of the peristyle. Thus it
would be seen on the way to the reception rooms situated on the east, north
and west sides of the peristyle. In De ΙT25 the statue group of Artemis and
deer was exhibited on a raised position in the courtyard and was visible from
the entrance and from the street when the door was open. A statue of Posei-
don was displayed in the peristyle of De1T41, and it was meant to be seen
from the main room. 196
DelI2 should be commented on, as it was filled with statuary: Aphrodite,
191 The Stadium Quarter: Laumónier 1956, nos. 328, 387, 500, 592, 610, 716, 1050 and 1337; The
Stadium Quarter and the Theatre Quarter: Kreeb 1988, 178-179, 266-271, 276-278, 281, 289 and
304-309.
192 Cf. Harvard 1987, 119-120, table 1; Kreeb 1988, 58-60.
193 Cf. Kreeb 1988, 48, table I -Il and 49, table IlI.
194
Cf. Kreeb 1988, 50, table ΙΙ --.
195
Harvard 1987, especially 57-79 and 98-100. Sculpture in domestic context is referred to no
earlier than the late 4th century BC (Harvard 1987, 80).
196
For these lines of sight, see Kreeb 1988, 18, fig. 2.2; 20, fig. 2.3; 29, fig. 2.6; 34-35, figs. 2.7-2.8;
36, fig. 2.9; 41, fig. 2.11.
51
Artemis, a nymph, several herms (both on the ground floor and upper stor-
eys) and apparently an `antique' made by the famous Praxiteles in the 4th
century BC. Of the latter only the base is left; it was displayed in the main
room, while several herms appear to have been located in the courtyard.
Portraits in domestic setting: The portraits from the houses are singled out
as they represent an important element of the private display of the inhabi-
tants; they are an example of a public element transferred to the private
sphere.
Ten portraits, maybe twelve, came from secure contexts. Unfortunately,
the base is all that is left in some of the cases, the heads are lacking in other
cases and one portrait is very fragmentary. The five portraits from uncertain
contexts will not be commented on (De1119 and De1T41).
De1N11: Several fragments of a male head in marble came from the de-
struction layer. 197 The style appears to be identical with the portraits from
De1123. Fragments of shoulders and chest were also found.
De1123: Two male busts in marble. ~98 They were displayed on the first
floor, apparently reversed around a central axis. The portraits are executed
in the verist style.
De1127: Male head in marble. 199 The precise find-spot is unknown.
De1ST2: Inscribed marble base found in room g (ID 1802). It once sup-
ported a statue in bronze displayed on the first floor: The inscription states
that three freedmen with Greek cognomina erected the statue of their patron
Q. Tullius.
DeΙI6: Part of inscribed marble base for statue (ID 1724). The exact find-
spot is unknown. The inscription states that three members of the gens Eg-
natius erected the statue of their benefactor Philostatos from Askelon.
De114: Male statue in marble. It was found in room g or room e fallen
from the upper storey. He wears a himation. According to Michalowski, the
man is Greek or Oriental rather than Roman due to his bare feet . 200
De1T7: Male head in marble 2 01 It was found in room f. The hair is not fin-
ished at the back of the head, suggesting that the portrait was displayed in a
niche or set against the wall.
De1T16: Two statues in marble of the Athenian couple Kleopatra and Di-
oskourides, from the deme Myrrinous. 202 The inscribed base was found in
situ in the door opening to room f (ID 1987). The heads are not preserved.
The inscription reveals that Kleopatra had the statue of her husband (and
herself) erected. The offerings of her husband in the Sanctuary of Apollo are
also mentioned. The inscription is dated to 138/137 BC during the archon-
tate of Timarchos. Whether this date refers to the erection of the statues or
to the votive offerings must be left unanswered. The statue of Kleopatra rep-
203Kreeb 1985b, especially 55-59; Kreeb 1988, 20, fig. 2.3; Trümper 1998, 334, fig. 66.
204His name is written in the accusative which is the case used in relation to erections of statues
on Delos (ID 2011) (Poccetti 1984, 651).
203
Marcadé 1969, 273 and 308-336; Stewart 1979, especially 67-69 and 142-145; Smith 1981.
206 This was for instance the case with the colossal statue of C. Ofellíus Ferns who is presented in
heroic nudity. The statue came from the Agora of the Italians.
207 Harvard 1987, 140.
200 The different influences on Delian sculpture have been studied in detail by Marcadé (1%9,
307-354; 355-405; 407-467).
53
209
RELIGIOUS FINDS
Dedications: Dedicatory inscriptions from various objects such as statu-
ary, altars and a censer reveal that the Greek gods were predominant: Zeus
Kynthios in De1114; Zeus Pasios, Poseidon, Apollo, Herakles and Artemis
(De1123, censer); Artemis (De1P3, De1T6, DelSl); Artemis Soteira (De1P8)
and Hermes (De112).
Apotropaia: Apotropaia (reliefs and mosaic motif) have been found in re-
lation to eight houses (DelNS, De1119, De1127, De1129, De1I3, DelIl,
De1T41, De1S2) 21° De1N27 was provided with three apotropaia and De1S2
with four. The repertoire comprises phalli (most common), symbols of the
Dioskouroi and Herakles. The only motif symbolizing a foreign divinity is
the sign of Tank from De1Ml.
Religious paintings (Table 1 and Table 11): The group of religious paint-
ings provides important evidence for the religious life of the inhabitants of
Delos. They were painted directly on the house façades next to the entrance
door, and/or in niches and on altars located immediately next to the en-
trance door. A few examples were painted inside rooms. The paintings are in
fresco technique and dated to the period from c. 125/120 to 69 BC. 11 They
were renewed several times. There may be up to twelve layers. The motifs
belong to two categories: 212 1) Sacrificial scenes with altar and male partici-
pants dressed in toga with covered head or dressed in himations and
wreathed. Servant with the sacrificial animal (a pig). Games (boxers or wres-
tlers) and the prizes to win. 2) Various gods, heroes and mythical figures, es-
pecially Hermes and Herakles. Various divine attributes, such as the club of
Herakles and the caduceus of Hermes.
There is now consensus that the paintings from the first group depict
events taking place in relation with the Compitalia, the festival celebrated in
honour of the Roman Lares Compitales of the crossroads. 213 The second
group of motifs relates to the other gods and heroes worshipped by the
Compitaliastai, for example Hermes and Herakles as protectors of the home.
The name comes from the Latin compitum referring to the crossroad
where the lares were worshipped. On Delos the Lares Cómpitales were wor-
shipped by the Compitalíastai. Several dedicatory inscriptions are preserved
from this college. 214 The majority of these came from the Agora of the Com-
209
Finds from houses serving as storage space or a sculptor's workshop are not included.
2~° Bruneai 1970, 643-648; Marcadé 1973.
211
Based on investigations of the paintings from De1N26 (Bezerra de Meneses/Sarian 1973, 99
and 104; Bruneau 1970, 619).
212
See for instance Bulard 1926a, pl. XI ΙΙ.
213 The paintings were originally published by Bulard who misínterpretedthe sacrificial scenes as
relating to the Roman domestic cult introduced by the Roman merchant colony. He based the inter-
pretation on the fragmentary painted inscription AGAT (from De11112), accordingly reconstructed as
Agathodaímon; the Roman Genius. Bulard admitted, however, that AGAT could be a personal name
as suggested earlier by Plassart (Bulard 1926b, 10-11; Plassart 1916, 211). For the history of research
on this particular subject, see Bruneau 1970, 589-615.
2t4 ID 1760-1771.
54
The altar is most often placed immediately to the right of the entrance
door or to the left. Different locations also occur, for example at the corner
of the insulti (De1N10), opposite the entrance (De117), immediately inside
the alley/corridor leading to the house (DelT13) Finally, there is one altar
inside a room (De1T36), but it may have been intended for a domestic cult
instead of the Lares Compitales.
In eight houses there are altars or remains of altars without paintings
(De1113 — upper storey, De1ST4, De1ST6, De1I4, De11114, De115, De1M6,
De1T6). These altars are all located to the right of the door (at the corner of
the house in the case of DelI4), and thus it is reasonable to relate them to the
Lares Compitales cult as well? 18 This brings the number of houses up to thir-
ty-three, representing thirty-five units (30/32 per cent). In total almost one
third of the houses can be related to the Lares Compitales cult.
Moreover, niches without traces of paintings have been located to the
right of the entrance in six houses (De1121, De1122, De1T2, De1T31,
De1T42, De1 Α2). Here the identification is more problematic, as these niches
may have served a purely practical function, i.e. location for lamps. But if
these are included the number reaches a total of thirty-nine houses/forty-one
units (35/37 per cent).
The analysis has shown that the cult of the Lares Compitales was wide-
spread on the island, probably from the last quarter of the 2nd century BC
onwards. The question is whether the paintings can be used as evidence for
the origin of the inhabitants of the houses, since it is not known who decided
to have them painted, i.e. the owner of the house or the magistrates of the
Compitaliastai. If the latter were the case, it would mean that the paintings
did not necessarily have anything to do with the house and its inhabitants.
But the location of the paintings in the immediate vicinity of the entrance
speaks in favour of a link between them and the inhabitants. Instead of plac-
ing them at the crossroads that were the natural location for such images in
Italy, they were placed at the door of a member of the Compitaliastai.
The combination of the Roman element (sacrificial scene) and the Greek
element (Greek gods and heroes) in the paintings reflects the influence from
the Greek setting. This also explains the existence of sacrificial scenes both
according to Roman custom (toga and covered heads) and to Greek custom
(himations and wreathed). Seven of the depicted persons are named, all
bearing Greek names. Six of the names are written in Latin. The most telling
example is a painting from De1126 where the three magistrates sacrifice in
the Greek way, and their names are written in Latin: ΡΗΕΟGIΡΙΑSΟΝ. 219
The paintings and the cult of the Lares Compitales suggest that the inhab-
itants of the house were Compitaliastai, thus a high percentage of the inhabi-
tants of Delos must have been slaves or freedmen. The paintings cannot re-
veal whether the houses belonged to freedmen or if these lived together with
their patrons in the houses. Both cases are likely.
From the Delian material it appears that the cult of the Lares Compitales
changed on foreign ground. The administration was identical to the one in
Italy, but at Delos the Compitaliastai also worshipped other gods and sacri-
ficed in the Greek way. The paintings inside the houses should probably be
seen in the same light, i.e. as a local solution or respond to the situation on
Delos.
218 According to Siebert, a small conical column in granite located to the right of the entrance of
De1114 might have formed the core of an altar (2001, 31, note 60).
219 Bezerra de Meneses/Sarian 1973, 89-90, figs. 21-22.
56
22Ο
EPIGRAPHICAL MgTF.RT Ι,
Personal names inscribed on various objects may provide the names of the .
220 Finds from houses serving as a sculptor's workshop or as storage space are not included.
221 ID
1730 and ΙD 2628.
222 r
1761.
223
Bruneau 1982a, 467-475 and 499-502.
224 r
1687, ΙD 2155-2156, ID 2379, r 2446 and r 2499.
r Flambard 1983, 72-74; Sohn 1983, 101-104; Rauh 1993, 222-230.
226
For the activities of the family on Delos, see Kreeb 1985b, 43-45.
57
22' ID
1768 and ΙD 1769 (Bruneau 1970, 600).
228 For example De1127, De112, De114, De1T16, De1T25 and De1T41.
58
229 Siebert rightly questions the strict privacy of the Dehan houses (1998, 171-173).
230 Tríimper 1998, table 1.
59
cates that it was displayed in the open. Members of the household may have
worshipped the Lares Compitales (niche without paintings). An independent
habitation unit existed on the first floor.
In De1122 (415 m2) the courtyard was provided with two Campanian por-
ticoes. Two emblemata (one removed) decorated the central field of the mo-
saic in the dining room (room ΑΕ) north of the courtyard. The alignment of
the motif, a Triteness, in the remaining emblema and the door in the east
wall would seem to preclude a continuous row of dining couches. The large
square rooms AF and AG of c. 33-34 m2 may also have functioned as recep-
tion rooms. Furthermore, the house had two bathrooms (bathtub "à sabot"
and sweat-bath). The upstairs rooms belonged to a separate unit whose inte-
rior architectural decoration included an emblema depicting a bird and figu-
rative friezes showing Nikes.
The owners of De1123 (257 m2) were presumably the two portrayed men
(busts in marble). Judging from the verist portrait style, they can be identi-
fied as Italians/Romans. Their involvement in late Hellenistic commerce is
documented by the c. 16,000 seals coming from papyrus contracts kept in
the archive located on the upper storey and by the large amount of amphor-
ae found in ground-floor rooms. 231 According to Boussac, it is likely that the
archive was both private and public. The house has a clear internal division.
On the ground floor the western part constituted the living area with court-
yard (room 8), pastas — or prostas-like area (room 8') and main room (room
µ), while storerooms and workrooms were located in the eastern part. 232 Ac-
cess to the upper storey was by the staircase located immediately to the left
inside the entrance room (room η). On the upper storey the office with the
archive was situated to the west and living rooms to the east, including a
kitchen and a possible cult room. 233 Both ground-floor rooms and upstairs
rooms were decorated with opus tessellatum, chip pavements and murals.
Moreover, statuary was displayed both on the ground floor and upper storey.
A bronze herm (not preserved) was probably erected on the marble base in
front of the Doric column between the courtyard and pastas-like area. 234
Sculptural finds from the first floor include for instance the portraits, statu-
ary in bronze and terracotta, and a votive relief depicting a sacrifice to Arte-
mis. The marble censer with names of several Greek gods should also be sin-
gled out, as it demonstrates the mixed culture expressed by the inhabitants.
It may not be a coincidence that the Lares Compitales cult is not attested
in connection with the perfectly laid out pastas-peristyle house De1N25, and
it is likely that the house was inhabited, or at least built, by Greeks from the
mainland. Most rooms on the ground floor were provided with murals, and a
bicolour chequer-board mosaic decorated the central part of the peristyle.
The only statuary from the house is a Harpokrates herm.
De1ST2 is a slightly modified "kanonische Normalhaus" (190 m 2) with a
231 Boussac 1988; Boussac 1993. Rauh has suggested that the house belonged to the banker family
narrow, rectangular ground plan, entrance room (room a) on the west side,
latrine in the front part of the plot and main room (room g) with two rooms
at the back north of the peristyle (2 x 2 columns). It is not clear whether the
peristyle is original or was added later. The upstairs rooms (internal staircase
in the entrance room) were decorated with mosaics, figurative friezes
(Brotes, festoons) and statuary: the statue of Q. Tullius was displayed here.
Fragments of a mortar pavement also came from the first floor. It cannot be
determined whether the house belonged to Q. Tullius or his freedmen that
erected the statue. It is possible that the owner was Q. Tullius, but that most
of the time the freedmen resided in the house and took care of their patron's
business. In any case, there is documentation for the cult of the Lares Com-
pitales. On the south side paintings are located on walls (originally, the en-
trance was here), and on the west side on walls and altar.
The question is how the presence of both a migveh and a Lares Compi-
tales altar can be explained in the case of De1ST6. Rauh has suggested that
the Jews living in the house were members of the C οmpitaliastai. 235 The ex-
planation offered by Triimper is, however, more plausible 2 36 In the first
phase of the house there were two entrances directly from the street to the
courtyard on the north side. The migveh below the courtyard indicates that
the house functioned as a Jewish meeting house or that it was a private house
with facilities for the whole Jewish community on the island. In the second
phase the entrances on the northern side were closed as new inhabitants who
were members of the Compitaliastai moved in. The peristyle without a stylo-
bate also belongs to a later phase.
The statuette base of Artemis Soteira located in a niche in the entrance
room (room 28) documents that Sp. Stertinius lived in De1P8. The entrance
room, a long and narrow corridor, leads to the courtyard with the main room
on its southern side. First-Style murals are documented in relation to the
main room and upstairs rooms. Due to the row of shops facing the Peribolos
Street, most of the houses situated here were entered through long and nar-
row corridors, and upstairs rooms projecting into the street were common,
judging from the porticoes built in front of De1P6, De1P7 and De1P8.
De112 is one of the largest Delian houses with an area of c. 708 m2 . There
were four storeys, but owing to its split-level design no part of the house
comprised all four storeys. The question is whether the freedmen Dionysíos
Paconius Neoteros lived here together with other members of the college
mentioned in the dedicatory inscription of the Hermes herm or if freemen of
the gens Paconius inhabited it. The dedicatory inscription is inscribed on the
base of the Hermes herm erected in a niche located on the second floor of
the southern part of the house. According to Triimper, the first phase of the
house consisted of the upper southern part only. The two-storeyed northern
part with its service rooms, courtyard with three porticoes and a nym-
phaeum, a statue made by the famous Praxiteles, herms and dining room
(room G) was built in the second phase, and most likely the northern and
southern part constituted two separate habitation units. 237 The southern en-
trance is generally regarded as the main entrance of the house, but there was
no axis or vista from the southern entrance to the statue located in the main
room in the northern part. 238 Moreover, the use of marble for the doorposts
and threshold of the northern entrance defines it as a primary entrance. Reli-
gious paintings related to the cult of the Lares Compitales were found to the
right of the door. Maybe the patron resided in the northern part and his
freedmen and slaves in the southern part. 239 The northern part of the house is
particularly interesting, as it comprises Greek elements well known from
Classical domestic contexts: the 11-main room complex (rooms D, E, F),
dining room or andrοn (room G) with raised platforms and pebble mosaic
decorated with a wheel, and herms. The `antique' made by Praxiteles on dis-
play in the main room and the Archaizing style of the Hermes herm from the
second floor enhance this `Greekness'. It appears that the Italic/Roman in-
habitants created their version of an `ideal' Greek house. 240 The house was
richly decorated with murals and pavements, but tessellated mosaics are only
attested from the upper storeys. Sculptural finds also include statuettes of
Aphrodite and Artemis from the upper storeys.
The Phoenician Philostratos of Askelon or the three brothers from the
gens Egnatíí may have resided in DelI6. Although the house is only partially
excavated, it appears to have been of a certain extension, as the peristyle had
4 x 4 columns.
De1M1 of 404 m2 is noteworthy for its axiality from the southern entrance
(room a) through the large peristyle (4 x 4 columns) to the central door of
the main room (broad-room, room h) with two additional rooms (rooms í, j)
to the east. There were service rooms (bath/kitchen and latrine), additional
reception rooms (rooms f, g) east of the peristyle, and upstairs rooms pro-
jecting into the street. The sign of Tanit decorating the mosaic of the en-
trance room suggests that a Phoenician built the house and inhabited it for a
period of time. The mosaic of the peristyle was made by a Phoenician too,
Asklepiades from Arados. At a later time members of the Compitaliastai may
have moved in and installed the altars and niches. Various scenes with Erotes
decorated the figurative friezes in room g. A spacious peristyle was achieved
by the layout with rooms on three sides only.
Due to its area (c. 655 m2) and four dining rooms (rooms e, g, h, i) richly
decorated with mosaics, Del Μ4 may have functioned as a clubhouse for an
association. The courtyard has a Rhodian peristyle (4 x 4 columns), and all
the dining rooms are located around the northern portico. A herm was dis-
played to the right of the entrance to the largest of the rooms (room g). The
portrait fallen from the upper storey may have belonged to a separate unit or
to upstairs rooms of the house. There is, however, no evidence for an inter-
23' Trümper 1998, 238-240. The two parts have different alignments, and also the internal door
system speaks in favour of this interpretation.
2}8 Kreeb 1984, 325-326; Kreeb 1988, 37-40.
239 Regarding the question of inhabitants, see Harvard 1987, 132-133; Rauh 1993, 223-330; Pe-
nal staircase. Other sculptural finds from the first floor include an enthroned
god, female statuette, a youth, and herm base. Probably there were members
of the Compitaliastai residing in the house (altar, no paintings are pre-
served).
De1T2 of 286 m2 also has an axis from the entrance (room a on the west
side) through the peristyle to the main room and dining room (broad-room,
room k). Moreover, there is a Rhodian peristyle, additional dining rooms
(rooms i, j), many tessellated mosaics and figurative friezes with Erotes and
festoons from the peristyle. A fragment of a mortar pavement (with a mean-
der?) probably came from an upstairs room. Whether the Lares Compitales
were worshipped cannot be determined (niche without paintings). The
house has both an internal staircase and an internal one with separate access.
De1T2 may have been built by a Syrian or have had a Syrian owner at a cer-
tain time because of the lion and bull protomes decorating the consoles of
the eastern portico. 241 The lion and bull are attributes of the Syrian gods
Atargatis and Ηadad. 242 The lions are executed in a style similar to finds from
the area of the Sanctuary of the Syrian Gods. 243 The combination of lion and
bull protomes occurs on two temples in Ηeliopolis. 244 Moreover, stucco
masks depicting men with tall pointed hats have been found within the
house and may be identified as S~yrían priests 24s
De1T16 with an area of 405 m represents an example of plot extension by
purchasing the neighbouring house. A peristyle (3 χ 4 columns) was added,
although there was hardly room for ít. The northern and southern porticoes
are only 1 m-1.5 m deep. In the third phase two portrait statues of the own-
ers, the Athenian couple Kleopatra and Dioskourides, were erected and dis-
played in a naiskos-like arrangement in the former door opening to the main
room (room f). The statues were visible from the entrance and demonstrate
self-representation in the highest degree. The main room (room f) is to the
north of the peristyle, and there is an additional dining room (room e) to the
south. This eastern part was clearly the reception area of the house. The
western part had a small courtyard (room i) and the characteristic triple
room complex to the north (rooms k, 1, m). A terracotta statuette of a god-
dess came from the house.
De1T41 (451 m2) had a large peristyle with 4 x 3 columns. The porticoes
on the east, south and west sides are 1.4 m-1.6 m deep. A Poseidon statue
erected in the peristyle could be viewed from the main room (room f) locat-
ed to the north of the peristyle. There are no traces of pavements in the main
room. Additional reception rooms (rooms i, 1) are located around the peri-
style whose inner part is decorated with an emblema showing a Dionysiac fig-
ure on a tiger. A Kybele statuette was found in the peristyle. Furthermore,
there is evidence for the Lares Compitales cult. Upstairs rooms belonging to
243
Marcadé 1969, 382-385.
244
Seyrig 1929, 317.
245
Bruneau 1970, 473. For photo, see Marcadé 1952, 112, fig. 10.
63
6
14 Marcadé 1969, 399-405.
64
freemen and freedmen were united for the purpose of profit-making 2 47 Mar-
cus Minatius financed a part of the Building of the Poseidoniasts of Berytos,
and the Phoenician Philostratos paid for the ground-floor portico on the
northern side of the Agora of the Italians2 g5 Furthermore, the Egyptian and
Syrian sanctuaries were used by Greeks and Italians/Romans2 49 Other sanc-
tuaries were private and reserved for specific communities. This goes for the
Semitic sanctuaries located on Kynthos. They comprise for example a sanc-
tuary to the Gods of Askelon. The Jewish community was probably concen-
trated in the Stadium Quarter. Here a synagogue has been identified, and
less than 100 m further to the north two honorary stelae were found. They
were erected by the Israelites on Delos and dedicated to Garizim. 250 Moreo-
ver, a migveh existed in De1ST6 (first phase of the house).
The considerable Italic/Roman impact on the housing is mirrored in the
city. This is evidenced by the existence of colleges having both a religious
and commercial character, by their monuments, and by the epigraphical ma-
terial. The island's new political situation is attested by the establishment of
the Romaia, festivals in honour of the triad of Hestia, Demos and Roma, in
166 BC or immediately after. 251 Moreover, the Phoenician Poseidoniasts of
Berytos worshipped Dea Roma in their clubhouse situated in the Northern
Quarter. Also the motif decorating the pavement of the cella (cella 11) was
inspired by mosaic floor decoration in the West: an opus tessellatum mosaic
with a black-and-white chequer-board. 252
The earliest evidence for the college of Hermaistai dates to c. 150 BC.
They established an agora between the Sacred Harbour and the Sanctuary of
Apollo, or more appropriately, a compitum where various Italic/Roman cults
were worshipped, including the Lares Compitales from 150-125 BC on-
wards.253 The Hermaistai probably installed the cult of the Lares Compitales
on Delos, and later they assigned it to the C οmpitaliastai. 254 Accordingly, the
agora is called both the Agora of the Hermaistai and the Agora of the Com-
pitaliastai. The members of the Hermaistai were freemen and freedmen. This
was also the case with the other Italic/Roman colleges: the Apolloniastaí and
the Poseidoniastai.
The influence was particularly strong from the last quarter of the 2nd cen-
tury BC onwards 255 The inscriptions from the Boulé and Ecclesia stopped
about 145/144 BC. 256 Instead honorific decrees were made by the Athenians,
247
Vial 1984, 389. According to Strabo, thousands of slaves were sold daily on Delos (14.5.2). For
the use of slaves and freedmen as middlemen in commerce in general, see D'Arms 1980; Zalesskij
1983, 22-24.
298 Mancinettí Santamaria 1983, 88; Zalesskíj 1983, 38.
249
Cf. for instance Bruneau 1970, 472; Will 1985, 140.
250 Bruneau 1982a.
251
Bruneau 1970, 444.
252
For photo, see Bruneau 1972, 145, no. 40, fig. 38.
253
Mavrojannís 1995.
254
The college of the Compitaliastaí is documented no earlier than around 100 BC.
255
The province of Asia was created in 133 BC.
256
Wilson 1966, 113.
65
Romans and other foreigners. The first decree made by this group is dated to
126/125 BC, and from this time onwards a status as permanent residents was
assigned to the Romans in the inscriptions. 257 The role played by Rome is also
reflected in the fact that Delos sided with Rome in the Mithridatic Wars,
whereas Athens sided with Mithridates.
Another manifestation of the Italic/Roman colony on the island was the
Agora of the Italians built at the end of the 2nd century BC. The agora ap-
pears to have had several functions, for instance meeting place for the colony
and provision of office facilities. 258 But non-Italians/Romans were also do-
nors of architecture and statues. 259 Siebert's analysis of the inscriptions from
secure contexts within the agora shows that ten were written in Greek, six
were bilingual, and six were in Latin. 260 Siebert sees this as a sign of the mu-
tual influence taking place on the island where the Romans could allow
themselves to be influenced because of their political and financial control.
pet-like design. The motifs are mythological (for example Dionysos in a char-
iot drawn by leopards), geometric, floral and symbolic (wheel). The statuary
is primarily in terracotta with a few examples in marble.
The House of the Mosaics at Eretria, which was built during the first third
of the 4th century BC and destroyed c. 270 BC, illustrates the next step 268
The house consists of two parts, one to the west with a peristyle and three
dining rooms (andrones) and an eastern part with courtyard and the triple
room arrangement with a large transverse one and two smaller at the back. It
was decorated with wall-paintings, polychrome pebble mosaics, and statuary
in terracotta and marble was employed. The statue of a youth found in the
peristyle was probably displayed between the entrances to the andrones situ-
ated to the north (rooms 7, 9) 269 The mosaics are decorated with mythologi-
cal themes and a rosette.
At Pella the development is taken a step further in the last quarter of the
4th century BC. The House of Dionysos (I,1) and the House of the Rape of
Helen (I,5), which were probably private houses, have several dining rooms
located around the large peristyles. 270 Mythological motifs prevail in the mo-
saics, now polychrome and with perspective, but there is also a black-and-
white lozenge pattern (the House of Dionysos, room D). Again the motif
with Dionysos riding a leopard occurs.
Like the Delian houses, many houses at Priene (353/352 130/129 BC) — .
underwent alterations in the 2nd century BC, for instance House 33 that in-
corporated the neighbouring house. 271 While pavements are represented by a
,
few mortar pavements and beaten earth floors, wall decoration includes
First-Style murals, sometimes with entablature in relief. The statuary in terra-
cotta and marble depicts for instance Dionysos, Aphrodite and theatrical
masks.
A great variety of houses are present at Pergamon in the Hellenistic peri-
od: pastas, prostas and peristyle houses. 272 Apart from the palaces, many
houses were provided with porticoes or peristyles no earlier than the 1st cen-
tury BC. One example is the Komplex VI Ι with a peristyle court of 270 m 2
and a total area of c. 740 m2.273 A large house of the 2nd century is the Peri-
stylhaus I from the area west of the Lower Agora. 274 The peristyle has 8 x 8
columns and measures 22 m x 22 m. A characteristic feature of the houses,
especially of the peristyle houses, is the suite of three rooms lying next to
each other, also found in Macedonian palaces. Often the rooms intercon-
nect, and the central room may be open-fronted and provided with columns
towards the peristyle. The Italic/Roman influence is reflected in the pave-
268
Ducrey et al. 1993, 32, fig. 25.
269
Harvard 1987, 35-38.
270 Nielsen 1994, 86, fig. 44; Westgate 1997-1998, 105, pl. 8. The House of Dionysos has an area
century BC.
274 Wulf 1999, 184, fig. 74.5
67
rents combining a central field in opus sectile decorated with cubes in per-
spective and surrounded by a border in opus tessellatum. 2 'S
275
The house of Attalos, rooms 37 and 38, dated to the early 1st century BC (Salzmann 1991,
440-442).
276
According to Vial, the main concern of the Delíans during the period of independence was to
preserve the family, their identity and traditions such as democracy (1984, especially 385-390).
27 Cf. also Siebert 1998, 174-178.
3. CARTHAGE. THE PUNIC METROPOLIS
HISTORY OF RESEARCH
This section will only give a brief outline of the history of research, since
detailed accounts have already been provided 2 78
The map drawn by the Danish Consul-general in Tunisia, Christian Tuxen
Falbe, was a crucial factor for the outset of archaeological exploration of
Carthage. His topographical investigations took place in the 1820s, and the
map was published in his Recherches sur l'emplacement de Carthage, Paris
1833. Soon afterwards he participated in the exploration of three of the sites
recorded on the map2 79 Another pioneer from the early nineteenth century
was the Dutch engineer Jean Emile Humbert, the first to discover Punic ma-
terial at Carthage (four gravestones with inscriptions and two fragments) 28Ο
Humbert also contributed with a plan of the Carthaginian peninsula and
conducted excavations not only at Carthage, but also at other North African
sites. In the second half of the nineteenth century archaeological research in-
creased, and in 1882 the Service des Αntiquités de Tunisie was established2 81
The excavations carried out in this period concentrated on the Punic ceme-
teries and consequently, the early necropoleis of Carthage were to a large ex-
tent known by the turn of the century. On the Byrsa hill the first excavations
were carried out in 1859 bey Beulé, who was also the first to reach the de-
struction layer of 146 BC. 28 Work on the hill continued in 1880 when Père
Delattre, one of several White Fathers active in the excavation of Carthage,
began his investigations2 83
The twentieth century saw the discovery of the tophet (1921). Excavations
on the Byrsa hill were resumed in the 1920s and 1930s by Saumagne and La-
peyre, in 1947 by C. Picard, and in the 1950s by Père Ferron and Pínard. 284
The latter erroneously dated the structures to the time of the foundation of
the colonia Iunonia Carthago by the Gracchi.
New knowledge of Carthage has been obtained from the "Campagne In-
ternationale de Sauvegarde de Carthage" launched in 1972 by the Tunisian
government and UNESCO. The project is multifaceted: to excavate, pre-
serve and restore the Punic, Roman and Byzantine remains of Carthage. Ar-
chaeological teams from Europe, North America and Tunisia conducted
large-scale research in the period 1973-1979 under the direction of the Insti-
278
For example Deneauve in Byrsa 11979, 41-55 (Roman structures on the Byrsa hill); Lancet in
Byrsa I 1979, 13-39 (Punic remains on the Byrsa hill); Picard 1985; Lancet 1995, 25-26, 228-248 (the
tophet), and 438-446.
279
Lund 1986.
280 Halbertsma 1995, especially 23-28 and 39-58.
287 For a summary of the results from this project, see Ennabli ín Pour Sauver Carthage 1992,
16-24 and bibliography 203-227.
2" See for instance Rakob 1989 with bibliography.
287 In Semitic Carthage means `New Town' (Qart Ha dasht).
288 Lance! 1988. Also the name itself `Byrsa' is still a riddle. Cf. also Lance! in Byrsa 11 1982,
379-383.
289
Cf. Lance! 1985b, 729; Rakob 1992.
298 For the extent of the Archaic city, see also Rakob 1987, fig. 2; Lance! 1995, 39, fig. 22; 42, fig.
23.
71
structures situated on the central part of the coastal plain ran parallel to the
shoreline. Habitation structures, such as mudbrick walls on stone socles and
beaten earth floors, have been found in situ at various locations 291 Among
the earliest is the quarter situated below the intersection of Decumanus Maxi-
mus and Cardo X (cat. nos. CarDM Ι -CarDM2). It dates from the second
quarter of the 8th century onwards, and it is noteworthy that the street was
paved and constructed with a drain as early as the second half of the 8th cen-
tury BC. Apart from the cemetery on the southern slope of the Byrsa hill that
functioned only in the 7th century BC, the other cemeteries were used as ear-
ly as the end of the 8th century BC, but they date mainly to the 7th century
BC. Tombs exclusively containing Phoenician material could be older.
The middle Punic phase, 5th century BC to mid-3rd century BC: In this
phase the city underwent a major reorganization2 92 An orthogonal street grid
was employed on the coastal plain and a fan-shaped grid for the south and
east slopes of the Byrsa hill. On the coastal plain the so-called `Mago Quar-
ter' was laid out including a city wall towards the sea. The wall was built with
a monumental gate flanked by towers. The same layout from the Archaic
phase continued: a nucleus fringed by industrial zones, for example on the
south slope of Byrsa and in the southern zone towards the tophet, surroun-
ded by the cemeteries. At the end of the 5th century BC a sanctuary was
built on the site of the present-day Rue Ibn Chabâat. Votive stelae found
nearby indicate that the sanctuary was dedicated to the so-called young Baal
Hammon.293 A Roman basilica atop the Punic sanctuary points to the exist-
ence of the forum nearby, and thus the Punic agora is supposed to have been
in this area as well. To judge from literary sources, the city appears to have
expanded by the end of 4th century BC. A "Nea Po lis" is mentioned by Di-
odorus of Sicily in the passages referring to the riot of the general Bomilcar
who tried to seize power in 308 BC294 Probably the "Nea Polis" refers to a
kind of suburb situated beyond the cemetery zone towards the north.
the 3rd century BC or the beginning of the 2nd century BC the Punic sanc-
tuary was reorganized.
Note on the quarters: The `Hannibal Quarter' (Fig. 3): With the exception
of the north-western and western slopes of the Byrsa hill, investigations have
been carried out on all the other slopes. The so-called `Hannibal Quarter' is
situated on the southern slope of Byrsa on a NE-SW/NW-SE alignment and
was created as a new habitation quarter at the beginning of the 2nd century
BC on top of a workshop area2 9 Apparently it was just one of several resi-
dential zones on the slopes of Byrsa which was laid out according to a polyg-
oηal system adjusted to the sloping terrain and employing an orthogonal lay-
out within each area. The `Hannibal Quarter' is the only one of which a sub-
297 An unpublished Tunisian rescue excavation in the Rue Florus on the east slope of the Juno hill
is not included, since no information of the nature of these remains has been obtained, point 16 on
Lancel's map (Lancel 1995, 145, fig. 74).
298 The name of the quarter was given by the Tunisians in 1981 due to its chronology (Lancel
1985b, 750; Lancel 1988, 82). Hannibal held the office as suffete in 196/195 BC.
73
Ι . Car2
Carl
Car21
Car22
Car23
~ . Car24
7. Car25
S. Car7
Car20
Cα r6
Car19
CarHi - Car Η13
Car11
Car12
I5. Car13
Ι 6. Car14
17. CarS
Ιδ. Car15
C α r16
Car17
CariS
C αι-26
CarDM 1 - CarDM2
CarM1- CarM10
CarB
Cm-27
Car28
28 2 δ. Car9
29. Car10
Tophet
29
1 km
Fig. 2. Carthage, location of habitation in the middle Punic and late Punic phases (approximate loca-
tions) (based on Lancel 1995, 145, fig. 74).
74
ítií ,í.
~
~~~
α• / .: ~ ςφτlζβ
~ '~,• ,, ~ ~ \ι
;
ν
, ~ ~φc4
'~
6
~φτθ1
/
\ \~
`‚
Τ
,,τ :
~~
' ~` \
Γ~ Cφτ$Α e• ~ Ε. λ \.- "
G ~
\ ~ \~ ρ
~ G ~ )~/ ~ ~ b
„• ~,
% >> ~ , -
\~, α εΡ w
``
"„' ~~ ~~ n✓
b çsτg5 . Sλ v . ~ .
, ` , %~ $,η \,~;~À~I• ^ ,, . . - . _, ~
•\`•
,,
.η' \c
. Cφτa ~1 η. ~/ ~ i~~i i_
..i
'
~~~ ";.: - η . .
.:
D α~~,, ' η .ι . . , . .
~7~
0 20m
Fig. 3. Carthage, `Hannibal Quarter', plan (Lancel 1981, 159, fig. 1).
Ν Μ Ψ
~~\
ι
Ι
ι-- - ττ - τΓ-)—
\ ι . ..ς •.
_= 1 _~
D1
i , ;;: _
1Ι .ι α
~
ι ι
2 S
αι
ι~ 4ι ~ ~ _ ..—..-.-.— __.__.~
ο
aι '
υ
-Ι
r
οτ=
LL
L. i ~:,
Ι; ν
viΣ9+J-' 'rPr
:_,. ~
~ S
ι; ι αiσ.ι ' , . /
.~% ®✓
'~/~i ~‚)‚; \C~
Γ
ο
Ιν 'j ο
Ι
if*
rn
υ
ι
ro
11τΡ, ll r
;
Ι
i
ι
~-
Ε& ν Ι"
_ _.-..- ΓJ Μ Ill Ψ
75
stantial part remains, whereas the documentation of the other zones chiefly
consists of wall sections and/or cisterns (cf. Cari 1-Car19). Four building
phases are distinguishable: 299
The first years of the 2nd century BC: Layout of insulae C, E and prob-
ably A.
Stairway built between insulae C and E at the intersection of street II
and street III.
Stairway (A) built in street II along the short side of insula E and at the
intersection between street II and street V.
No earlier than the end of the second quarter of the 2nd century BC:
Layout of insulae B and D and stairway (B) in street I Ι, additional stair-
ways in street II.
The Mago Quarter' (Fig. 4): The quarter is situated on the coastal plain, 301
Although no house has been fully excavated, it is possible to follow the de-
velopment of the quarter from its initial layout in the second half of the 5th
century BC until its destruction in 146 BC. It is laid out with an orthogonal
plan on a NW-SE axis following the coastline, its north-south extension
measuring minimum 95 m. The reorganization of the area in the second half
of the 5th century BC also included a wall towards the sea. Judging from the
soundings carried out west of the modern Rue Septime Sévère, two more in-
sulae existed here. 302 Three main phases have been identified:
303
1) Mid-5th to mid-3rd century BC (Phase Ia, Ib, Ic): In the 5th century
299
Lancel 1981, 177, figs. ha-1d.
goo
Lancel 1981, 180-182.
301 The name of the quarter was suggested by M. Fantar (Rakob in Karthago I 1991, 252, note 232).
302 Rakob 1989, 158-160 and 184-189, especially "Sondage Kardo XVI" and "Sondage N".
303 The houses have a complicated building history with many phases. For details, see the publica-
tion (Karthago 11991, IL "Synoptische Tabe llen", between 92 and 93 (Car Μ2, CarM6, CarM7,
CarM9); 4. "Synoptische Tabe ll en", 133 (CarM1); 187-189 (Car Μ3 -CarΜS and CarM8); Karthago I
1991, plan 26-29.
76
3 4 Structures within the plot of House Ic are even older dating to the 6th and first half of the 5th
century BC.
305 Lance! in Byrsa II 1982, 376. The excavators reconstructed two insulae and considered wall
rn/I an outer wall (Carrié/Sanviti in Byrsa 1 1979, 130).
306 Morel in Byrsa II 1982, 205.
77
ARCHITECTURE
ARCHITECTURAL ANALYSIS
Ground plan (Table 13): Since those houses whose ground plan can be re-
constructed are situated in the carefully planned `Hannibal Quarter'wíth an
orthogonal street grid, it is hardly a surprise that the rectangular ground plan
is the most common. The number of such ground plans increases consider-
ably if the reconstructed house plots, mainly from the `Mago Quarter', are
included. No house in this quarter has been fully excavated and consequent-
ly, no ground plan is known with certainty. All plots are, however, recon-
structed as rectangular or square (CarM8). L-shaped ground plans occur in
CarH4 and CarHl1, due to the location of independent shops and work-
shops in one of the corners. Car5 may be irregular, but the house is only par-
tially excavated (Fig. 5).
Dimensions (Table 13): In general, the Carthaginian houses are of modest
dimensions. In the `Hannibal Quarter' the house-plot size was to some ex-
tent standardized. The layout of the insulae shows an intended ratio of one
to two with a basic module of 52.18 cm (identical with one Punic cubit). In-
sula C measures 31.30 m x 15.65 m. Same dimensions are reconstructed in
insulae A and E, i.e. 60 cubits for the length and 30 cubits for the width. The
dimensions of insula B are 30 cubits for the length and about 21-22 cubits
for the width which gives a more compact layout of 15.65 in x 10.50 m. Insu-
la D was probably identical with insula B. Houses of c. 75 m2 are most com-
mon (three units: CarH6, CarH7, Car Η8). The best preserved is Car Η8
measuring 15.65 m x 5.20 m (30 x 10 cubits). The largest houses are the two
reconstructed double units of c. 150 m2, i.e. the width is twice the standard
width, c. 15 m x 10 m (CarHl and CarΗ13). Finally, CarΗ9 and CarH10 ap-
pear to have been only half the standard unit, i.e. c. 7.5 m x 5 m. Because of
this standardization, it has been suggested that the houses were made for a
homogeneous population. 307 In the 'Mago Quarter', where space was avail-
able, CarΜ6 apparently occupied an area of 1440 m2: a fact that, acc οrdín~
to the excavators, could point to its identification as a public building. 3
However, it should be stressed that there is no definite evidence that this
huge area was occupied by only one house. The identification of boundary
walls between the individual houses makes the reconstruction of the other
plots more likely. CarM8 is the smallest house in this quarter (156 m 2). In the
original layout standard units of different sizes may have been used, but the
later extension and amalgamation of two or more houses resulted in houses
of varying dimensions. Car4 occupied an area of c. 200 m2, and in Car5 the
area was larger than 50 m 2.
Entrance room (Table 13): Normally, the entrance room takes the shape
of a narrow corridor, often fairly long and with a width of c. 1 m. They are
η
r
•Ι•
~
ο
best documented in the `Hannibal Quarter': Car Η6: 1.15 m; CarH7: 1.1 m;
CarΗ8: 6.6 m x 0.9 m; CarH10: 1.35 m. 309 In Car3 the width exceeds 1 m,
but not much. In Car4 it is 1.36 m and 1 m in Car8. In the `Mago Quarter'
there is only one certain entrance room, room P 100 of CarΜ5. Here the
width is 2 m. Room P 73 in CarM4 and room P 118 in CarM8 also appear to
be long and narrow entrance rooms, although there is no evidence of the en-
trance/doorway itself. CarH4 is provided with a cruciform corridor. In rela-
tion to CarM6 a former alley (rooms P 30, P 30a, P 49, P 49a) was privatized
and converted into an entrance. The entrance room can be centrally (CarH4,
Cari5) or laterally placed. The latter is the more common. A corridor is re-
corded in connection with Car6 and Car7, but whether it served as an en-
trance room is uncertain.
Circulation space: A courtyard without porticoes has been identified in six
houses: CarΗ8 (4 m x 3.5 m), CarΗ13, CarM4, Carl, CarS and Car6. In
Carl there were two. Moreover, courtyards are likely to have existed in the
following houses: CarH6, CarΗ7, CarM2 and Car3. In CarH6 and CarH7 it
is proposed because of the similarities with Car Η8. In CarΜ2 a section of a
cistern has been located in room P 54, and in room P 55 there is a drain. The
corridor in Car3 turns at a right angle in what appears to be a courtyard.
Courtyards provided with porticoes are documented in three houses:
CarM ι (at least two courtyards), CarM4 and CarM6. The original layouts of
CarΗ8 and CarH13 also included porticoes. CarΗ8 had one portico suppor-
ted by two rectangular pillars. The portico was closed at a later time, and in-
stead three rooms were built. Among the identified courtyards there is a
preference for porticoes on three sides, a so-called porticus triplex. This ap-
plies to the two courtyards in CarM ι and to the courtyard in CarM4. In
CarM6 it is not clear whether there were porticoes on two or three sides
(north and south sides, five supports on each). Moreover, the double unit
CarH13 in the `Hannibal Quarter' had a portico-is triplex in its first phase.
There is a column in the western corner and a pillar in the northern corner.
It is obvious that the limited space of the houses on the Byrsa slope did not
allow large portico arrangements, whereas the more spacious plots on the
coastal plain could provide the required space. A characteristic feature of the
portico-is triplex is the use of pillars instead of columns Pillars are found in
CarM ι, room P 9 (c. 4.9 m x 4.9 m) and room P 18 (c. 5.8 m x 5 m). In room
P 9 there were four supports on the eastern and western sides, judging from
the remains, and three on the south. In room P 18 the number of supports
on the north and east sides is reconstructed. As to a possible third courtyard
in room P 1, a block of El-Haouria sandstone may be interpreted as a base
supporting a square pillar. 310 It cannot be ruled out that room P 13 connect-
ing the two courtyards instead of a wall and doorway towards the courtyard
in room P 18 was provided with a row of pillars. In that case it would be a
true ρerístyle, 311 In CarΜ4 there are traces of eight square bases supporting
either pillars or columns. 312 Pillars also app ear to have been employed in-
stead of columns in the porticoes of CarM6. 13
The use of columns in the Carthaginian houses is documented by the fol-
lowing examples: Cari5 : a Doric capital on a half-column with smooth shaft
and stuccoed. 314
CarΜ4: Fragments of Doric columns and capitals.
Car4: Two Ionic capitals coated with stucco, but their ascription to the
house is uncertain.
Carla: Column drums, both fluted and smooth.
Main room (Table 13): The poor state of preservation and often partial
excavation make the identification of a main room problematic. Moreover,
the ground-floor rooms in the `Hannibal Quarter'appear partially to have
served commercial purposes. Main rooms have been identified in the follow -
ing houses:
CarH8: Room I. The room is provided with a pavement of opus tessella-
turn, a threshold made from a monolithic limestone block and a niche in the
(room G in both houses) and may best be described as shower baths: a small
room provided with a vertical water-supply pipe and a waste-water drain. 316
Furthermore, the bathroom in Car Η13 has two niches in the wall for bathing
accessories. The bathrooms in Car4 and Cars appear to be much alike too,
consisting of two interconnected rooms. In Car4 there is a room with a
bench (the changing room) and the washroom proper with two basins. In
Car5 the bench is located in room 3 which also has an outlet grating. Mr-
ever, both rooms have traces of what appears to be a hot-water tank or ba-
sin.317 Due to destruction the arrangement in CarM4 (room P 79) and CarM6
(room Ρ 41) is unknown or hardly detectable. A wash basin (50 cm x 34 cm)
is documented in CarΜ6 318 Traces of installations in room Ρ 52 of CarΜ2
might relate to a bath3 19 The use of the bathtub "à sabot" is attested at
Carthage. In Car10 the tub appears to be a built-up one, judging from the
photo and descriptíon320 Besides, a fragment of a portable tub in terracotta
was found in the L-shaped corridor of Car Η3 and in Car11321 According to
Cíntas, a bathtub was found in Carl and Car6, but the excavator of Carl
does not mention this322 It should be noted that the find or reports of the
presence of a bathtub is not necessarily evidence for a separate bathroom, as
the tub may have been placed in a room used for various purposes.
A kitchen occurs in Car3 (oven), Car4 and Car6 A latrine is mentioned in
connection with Carl. A room intended for a domestic cult may have existed
in CarS. This is suggested by the three niches ;in room 4 where a large central
niche is flanked by a smaller one on either side. Moreover, there is a bench
along the wall of the room3 23
The north-western area of Car Η8 towards street I is interpreted as store-
rooms. Room P 25 of CarM1 appears to have been turned into a cellar for
storae after a fire. Three levels of Punic amphorae set in rows were found
here. 24 In CarM6 there was a suite of storerooms (rooms P 31- Ρ 36). Work-
shops have been located in the following houses: Car Η13, Car5 (crushing of
pottery), Car4 (eastern wing with olive oil production). In Car Μ8 it is not
clear whether room P 111 served as kitchen or as a workshop, maybe it
served both functions. Amphorae and traces of standing vessels were found
here together with remains of an oven.
Upper storeys (Table 13): The existence of upper storeys, or at least terra-
316 The bathroom in CarΗ8 measures 1.46-1.78 m x 1.27-1.35 m, and in CarH13 the dimensions
are1.11mx0.80m.
3 17 Chélbi 1980, 31-32; Chelbi 1984, 23-24. Room 2 measures 2.15 m x 120m and room 3 2.10 m
x 1.10m.
318 Stanzi in Karthago I 1991, 17.
Water supply (Table 14 and Table 15): Installation of cisterns and wells
secured the water supply of the Carthaginian houses. With a few exceptions,
the cistern is an omnipresent feature, whereas wells rarely occur.
Characteristic of the construction of the Punic cistern is a double wall in
the form of an outer and inner wall, built of small stone blocks. On the in-
side the cistern is lined with hydraulic mortar, the colour ranging from grey
to black due to the admixture of ashes. The cover takes the shape of a pit-
ched roof made of sandstone slabs or they are placed horizontally. A third
method is a cover of amphorae fitted into one another. Based on ground
plan and section, the cisterns can be divided into the following two types:
Bathtub-shaped, vertical walls. The bathtub shape is defined as an
elongated rectangle finished off by two semicircles. This type also includes
variants, for instance slightly curved at one end or not perfectly semicircular.
It may be provided with a separate shaft for the wellhead, sometimes meet-
ing the main section at a right angle.
Circular, bottle-shaped, 327
325
According to Appian, there were houses six storeys high at Carthage (Pun. 128).
326
Lancel et al. 1980, 19. No details are given.
327 Also called bell-shaped (Vann 1981, 30).
328 No conclusive arguments for either a Punic or a Roman date can be given (Davis 1981, 45-49;
pre-existent structures, while on the Byrsa hill they were part of the original
plan 329
Type V is only represented four times twice in CarH4 and on the out-
,
skirts of the city: Carl and Car3. Two cisterns are situated in the same area
as Car2, but apparently they do not belong to the house and are not includ-
ed, since their date is more likely to be Roman than Ρunic.330
The shapes of several cisterns are unknown due to insufficient informa-
tion, partial excavation or destruction. In Car Η13 the presence of a cistern
beneath the courtyard is presumed because of the inflow hole in the centre
of the courtyard, but it has not been investigated. As regards the cisterns in
rooms P 9, Ρ 19 and P 21 of CarMi, it should be noted that no traces of the
cistern from the Punic period in room P 9 exist due to destruction of the
area by Roman structures. However, remains of a channel system confirm its
existence. Room P 19 is situated outside the excavation area. In room P 21
the cistern was rebuilt in the Roman period leaving no traces of the Punic
period 331 In CarDM2 the number of cisterns and their shapes are unknown.
The houses in the `Hannibal Quarter' are all provided with cisterns. Due
to the incomplete excavation of CarHl1, it is not clear whether this house
had a cistern or not. The maximum number of cisterns is two, found in five
houses, whereas the much larger houses on the coastal plain as a rule have
several cisterns, at least in the last phase. CarMl appears to have been pro-
vided with eight and CarM6 with five cisterns. Car Μ5 had four or five de-
pending on where the boundary between this house and Car Μ9 is drawn.
No cisterns have been located in CarM8.
It is characteristic that the cistern is located below the courtyard, or partly
below this (Table 15). Another frequent location is below the long and nar-
row corridors. With respect to the houses from the `Hannibal Quarter', the
large capacity of the cisterns has been used as an argument in favour of the
existence of upper storeys. The same applies to the `Mago Quarter'. Since no
house has been fully excavated here, it is not possible to get ~~complete íi~-
pression of the relation between the area of the house and the number of cis-
terns, but so far the largest number of cisterns have (logically) been found in
the largest houses. In the 'Mago Quarter' the width of the cisterns varies
from 0.8 to 1.2 m and the longest of the fully excavated is the one from
CarΜ4 room P 75 with a length of 8.5 m and a capacity of c. 14,300 litres.
The cistern in CarM6 room P 48 is the smallest with a capacity of c. 5,550 li-
tres. In general, the Carthaginian cisterns show a striking consistency in di-
mensions, especially as regards the width.
329 In the `Mago Quarter'at least eleven cisterns have only been excavated to a very small extent.
However, one semicircular end is documented or part of one end: Car Μ2: room P 54; CarM4: room
81•; CariS: rooms P 102, Ρ 106, room 120•; CarM6: rooms Ρ 42, Ρ 44, Ρ 46, Ρ 47a; CarM7: rooms
135•, 136•. For plans showing the water supply and disposal in the `Hannibal Quarter'and `Mago
Quarter', see Lancet 1995, 158, fig. 81 and 169, fig. 90; Karthago 1 1991, plan 30.
30 They are rectangular with rounded corners (Merlin 1919, 194). This type of cistern is common
Only a few cisterns date to the middle Punic period, while the majority are
late Punic. In Car10 they can be dated before the mid-3rd century BC. The
cistern in CarDMl is from the 4th century BC. In CarM10 and Car17 the
cisterns may be middle or late Ρunic3 32 All the cisterns in the `Hannibal
Quarter' date to the first half of the 2nd century BC (Table 15). This corre-
sponds well with the late installation of the cisterns in the houses on the
coastal plain from the second half of the 3rd century BC onwards (Table 15).
The majority of the cisterns here date to the second quarter of the 2nd cen-
tury BC, for example one of the cisterns in Car Μ2. In CarM3 the cisterns are
supposed to be contemporary with the securely dated ones from CarM4 and
CarIS, i.e. the second quarter as well. The cisterns of Car Μ6 are said to be-
long to the first quarter of the 2nd century BC.333 According to the publica-
tion, the cistern in room P 48 was constructed in "Phase 2" (after the middle
of the 3rd century BC), while the cisterns in rooms P 42 and P 46 cannot be
dated. However, the one in P 42 dates before the second quarter of the 2nd
century BC (before "Phase 4") 3 34
It should be stressed that if only the cisterns from the `Hannibal Quarter'
and the `Mago Quarter' were taken into consideration, one would get the im-
pression that the cistern is a late feature of the domestic architecture at
Carthage.
No wells have been located in the `Hannibal Quarter', but they exist in
three houses on the coastal plain (CarMi, CarM4, CarM6). The wells occur
in the largest houses suggesting the need for additional water supply in these
houses. Wells appear to have been the main water supply in the `Mago Quar-
ter' in the second half of the 5th century BC (Carli, CarM4). The well in
room P 75 of CarM4 was replaced by a new one at the end of the 3rd or dur-
ing the first quarter of the 2nd century BC, while the others located in rooms
P 72 and P 78 were converted into overflows for the nearby cisterns in the
second quarter of the 2nd century BC. Thus only one well was in use in this
house in the last phase. Furthermore, wells are mentioned in connection
with the following locations: Car2, Car4, Car6, Cara, Car10 and Car17. On
the Byrsa hill waste water from the houses in the `Hannibal Quarter' was dis-
charged by pipes and channels into soakaways situated beneath the streets or
allowed to flow directly down the slopes. 335 In the `Mago Quarter' the same
soakaway system existed, but here they are situated within the houses. Soak-
aways are also mentioned in connection with Carl and Car3.
DIACHONIC PERSPECTIVE
The housing of the middle Punic period is best documented in the `Mago
Quarter'3 36 The plots were narrow, the width ranging from 11 m to 30 m.
House IVa had a corridor leading to the courtyard at the rear of the house.
Two porticoes existed in the courtyard of House lib in the 4th century BC.
Wells appear to have prevailed over cisterns as systems of water provision.
CarDMl was provided with a cistern in the 4th century BC, and its southern
part served again as habitation after having been temporarily converted into
a Tanit sanctuary. An extension of the city northwards is likely to have taken
place in this period, i.e. the "Nea Po lis" mentioned by Diοdοrus.337 On the
heights of Bordj-Djedíd walls dating no earlier than the 4th century BC,
Car22, have been located and may document this expansion.
In the late Punic period two tendencies are clear: on the one hand, houses
were enlarged. On the other hand, there was a need for more housing. The
extensions of plots are particularly attested in the `Mago Quarter' where re-
planning and more luxurious furnishings took place from the second half of
the 3rd century BC onwards, including the amalgamation of two or more
houses and the installation of cisterns and porticoes. This applies for in-
stance to CarMi where three houses were joined: In Car Μ2 new pavements
were laid, and a bath appears to have been installed. Car Μ4 was created by
the amalgamation of two house units, and a new large, courtyard was built.
CarM6 was extended towards the east and was provided with porticoes in
the courtyard. The same happened in the Quarter of Decumanus Maximus
and Cardo X, according to the excavators. They refer to a room of 60 r 2, ex-
tension of plots, a "<Peristyl>-Haus" and the installation of cisterns in
CarDIl and House II. They also speak of a monumentalizatíon of the quar-
ter in the second half of the 3rd century BC, including the placing of hexago-
nal terracotta tiles in front of the entrance to one house. 338 This quarter is ex-
tremely important because of the documentation of a paved and sewered
street.
The need for more habitation is documented all over the city. The houses in
the `Hannibal Quarter' and also Car9 and Car10 were built on top of work-
shop areas or converted from workshops into houses. New houses were con-
structed in the `Mago Quarter' where all available ground became occupied.
CONCLUSION
Distinct house types are not present, but two features stand out: the use of
standard plots to a certain extent in the `Hannibal Quarter' and the prefer-
ence for porticus triplex supported by pillars. So far no true peristyles have
been documented in the Carthaginian houses. As circulation space the open
courtyard is most frequent, but is not omnipresent. The most common fea-
tures are the planned quarters with rectangular houses; the narrow, laterally
placed corridor serving as entrance room and thus protecting the privacy of
the inhabitants; the bathrooms and the characteristic bathtub-shaped cistern
often situated below the courtyard or corridors.
The internal layout of the best preserved house, Car Η8, can be described
as an U-plan with a laterally placed corridor leading to a central courtyard
and the main room in the front part of the house (the relationship between
entrance room, courtyard and main room). "' This plan may also have been
present in CarΗ6 and CarH7 if the rooms towards the street functioned as
main rooms and not as shops. In Car5 the courtyard also appears to have
been centrally placed, whereas the courtyards in the `Mago Quarter' were sit-
uated at the back of the house plots, judging from the excavated remains.
In addition to the long and narrow corridors, a characteristic succession of
narrow and wide rooms occurs in the 'Mago Quarter', for example in Car Μ2
and CarM3. In CarM3 there are two room complexes; each comprises one
large room with a small room on either side - (rooms P 50, P 51, P 52 and
rooms P 53, P 57, P 58).
wash basins are attested. The use of this type of bathtub must be seen as in-
fluence from the Greek world including Sicily and North Africa. Here porta-
ble and built-up bathtubs are well-documented in private houses and public
339
For this type of plan, see Daniels 1995.
340 Lézine 1968, 152.
;41 Braemer 1982.
342
Especially type Illl of Braemer (1982, 71-73, 84 and 72, fig. 24).
343
Durand/Saliby 1985.
344
For plans of these palaces, see Nielsen 1994, 35-72.
88
baths dating from Classical and Hellenistic times, for instance in houses at
Olynthos, the public baths at Gela (c. 310-280 BC) and the baths in the
Sanctuary . of Apollo in Cyrene (early Hellenistic period). 345 Immersion tubs,
representing the tradition of the Oriental world, are lacking so far. 346
The remains give a rather incomplete picture of the wall decoration in the
Carthaginian houses which is due to the poor state of preservation and often
only pàrtial excavation. Moreover, the publications of the `Hannibal Quar-
ter' and `Mago Quarter' include only a selection of the finds. Based on this
material, identification of different categories of-wall decoration is confined
to the First Style, while a group of miscellaneous remains is too fragmentary
to provide evidence for a particular decorative system.
Wall decoration in situ has been found in the following houses: Car Η3,
CarΗ4, CarH5, CarΗ8, CarH10, CarΗ l3, CarΜ4, CarM6, Carl, Car2, Car4
and CarS. In general, it consists of remains of plain white or grey stucco. 347 In
the corridor of 'CarΗ8 there is a slightly projecting socle 0.65 m high in hy-
draulic mortar. Above is a plain white zone. Hydraulic mortar is also attested
in Carl, Car4 and CarS (bathroom). Remains of a panel with drafted margins
are documented in CarM4.
Fragments of mouldings came from CarDM2, Car4, Cars and Car25. The
latter may include traces of painted decoration, especially in blue, red and
yellow. The finds from CarDM2 consist of two painted fragments of a Lesbi-
an cyrnatium and two painted fragments of egg and dart mouldings. 345 Wall
decoration from upper storeys occurs in CarH5 (many fragments of moulded
architectural stuccoes) and CarH13 (debris of stucco). Moreover, there are
finds of uncertain provenances from the `Hannibal Quarter' (Table 16). Al-
though they cannot be assigned to a specific house, they supplement the ma-
terial. As to nos. 1-6 in the table, Ferron and Pinard write that the majority
of the finds listed in this part of their publication came from the habitation
described earlier in the text. 349 The problem is, however, that more than one
house are mentioned in this section, thus it is not clear which house they re-
fer to. There are three options: Car Η3, CarΙ7 and CarΗ8 of which the latter
is the most likely because it is described in detail.. Fragments from upper
walls (no. 7 in the table) document the existence of more complex decora-
tion including painted mouldings with the following elements, reading from
majeure partie, de l'habitat décrit plus haut. Outre les fragments de décor architectural, il comprend
du mobilier, de la cYrarnique et d'autres objects usuels" (Ferron/Pínard 1960-1961, 127).
89
The analysis shows that wall decoration occurred in houses situated in the
very centre of Carthage as well as on the outskirts of the city. CarM4 and
CarM6 provide the richest material, but finds from Car Η5 (upper storey),
CarΗ8 (including the fragments of uncertain provenances) and CarH13 (up-
per storey) also point to decoration of a certain quantity.
PAVEMENTS
3θo Cf. Lance! 1995, 319, fig. 190. Nothing is said about the find-spot.
„1 Laidlaw, A. in Karthago Π 1997, 228.
,xi
Laidlaw, A. in Karthago 11 1997, 228.
,s3
For exact parallels, see Laidlaw, A. in Karthago 111997,215-225.
90
bal Quarter', showing the distribution of types. 354 Types and distribution are
shown in Table 17.
Opus tessellatum mosaics on the ground floor only occur to a modest extent
(nine catalogue numbers), and they are mainly monochrome white employing
marble and limestone tesserae. 355 In the courtyard of CarM4 the tesserae are
regularly set, whereas they are laid obliquely in the porticoes in order to distin-
guish the two different zones. 356 Sometimes the monochrome mosaics are em-
bellished with tesserae of dark coloured limestone (or marble?) randomly or
regularly set. This applies to Car Η8 (room G, threshold mosaic) (Fig. 6),
CarΜ2 and CarΜ3 357 Α bicolour mosaic depicting a black-and-white chequer-
board frames the opus figlinum pavement in one of the courtyards of Carl. 358
The only polychrome mosaic on ground-floor level documented so far is the
threshold mosaic from the bathroom of Cars. It has a geometric pattern con-
sisting of nine rows of stepped triangles in black, red and white surrounded by
a lead band. 359 It is the only more complex design found so far at Carthage. 36°
However, a fragmentary tessellated mosaic from CarDMl may depict a poly-
chrome or at least bicolour motif. Judging from a photo, it appears to be a
monochrome white field with a central design. 361 The latter consists of a frame
with what looks like a chequer-board and a dark square in the centre.
Fragments from upper storeys and/or destructions layers are both mono-
chrome and polychrome. Α selection from the `Mago Quarter'is shown in
the publicatíon3 62 Unfortunately, information of exact find-spots is lacking
except for one3 63 One of the fragments is decorated with a chequer-board
pattern in black, white and red. 3 The remaining polychrome fragments are
too small to give an idea of the design.
pl. 7021). According to the text, the mosaic came from room P 90 or House 111. However, in the sec-
tion on this room it is stated that no pavement was preserved here (Teschauer in Karthago I, 1991,
151). The mosaic therefore appears to have fallen from an upper storey.
358 For photo, see Renault 1912, 364, fig. 11.
359 For photos, See Chelbi 1980, 33, fig. 7; Chelbi 1984, 31, fig. 1 and 32; fig. 3; Dunbabin 1494,
Fig. 6. Carthage, Hannibal Quarter', Car Ηδ, room G (bathroom), opn.r tesselhtum and opus fighnui7
(author's photo).
Opus figlinum is widely used (fifteen catalogue numbers and possibly one
more). This type of pavement is laid in areas exposed to water and humidity:
Corridors: CarΙ5, CarH6, Cari- 112 (immediately inside the doorway) and
CarM6. Bathrooms: CarΗ8 (Fig. 6), CarΗ13, CarM4 : Carµ6 and Cars.
Basin: CarMi (room P 7).
92
Tile pavements: Opus sectile made of terracotta tiles occurs on four loca-
tions. In Car8 and Car17 the tiles are lozenge-shaped3 73 In CarDM2 they are
hexagonal and laid in front of the entrance to a house. Carl is the fourth lo-
cation.
Mortar pavements: This pavement type is the most common. Many sub-
types are distinguishable of which at least three exist in the `Hannibal Quar-
ter' (`h' refers to the `Hannibal Quarter'):
hi: with a dense scattering of white fragments and tesserae3 74
h2: with a sparse scattering of white fragments and tesserae. 375
365
Lancel/Thuillier in Byrsa 11979, 234.
366
According to the caption, the pavement is composed of "tessélles de poterie" (Fantar 1985, 52,
pl. VIIl.c and 53, pl. IX.b).
367 For CarDMl, see Niemeyer et al. 1996, 32, fig. 36.
37 ` Karthago 11991, pls. 50.j, 50.1 and 51.d-51.e; colour pls. 69.10-69.12 and 70.13-70.14.
372
Karthago 11991, pls. 50.j, 50.1 and 51.e; colour pl. 69.12.
373 Car8: 024 x 0.12 x 0.05 m, corridor. Car17: 0.195 x 0.090 m, black, yellow and red (Saumagne
1924, 186).
374 Lancel 1985 a, 171, no. 2.
373
Lancet 1985a, 171, no. 1.
93
36 Lancel 1985a, 174, no. 8. This sub-type includes the pavements termed lithostroton in the pre-
urinary reports.
377 Pavements, wall decoration and architectural finds will be published in a separate volume (Ra-
kob in Karthago I 1991, 223).
3'8
"...aus feinem, rotem Terrazzo..." (St an zi in Karthago I 1991, 10).
3'9
Karthago 1 1991, colour pl. 69.7.
380 Fantar 1985, 14.
Fig. 7. Carthage, `Hannibal Quarter', CarH13, room E (courtyard), detail of mortar pavement with
tessera design (author's photo).
containing only a small amount of fragments or tiny ones. Car8 has a mono-
chrome mortar pavement 3 83 In Car10 some of the pavements are termed
opus segmentatum and opus scutulatum, but as they are not described fur-
ther, their precise character remains unknown. These terms, however, often
relate to mortar pavements, and therefore mortar pavements are likely to be
present here. The pavements of Car11 are made of a "mortier friable" .384 In
the case of Car25, opus signinum and opus scutulatum allegedly occur. The
latter term is given in a photo caption. The photo a&ears to show a mortar
pavement with white fragments randomly scattered.
Mortar pavements with tessera design: At least two patterns occur:
1) Tesserae arranged in equally spaced rows laid parallel to the walls of a
room: CarH7 (corridor), Car Ηl3 (courtyard) and Car4 (corridor, red-
dish colour). 386 Ιn CarΗ7 and Car4 the individual tesserae are set obli-
quely to the walls and far apart, whereas the tesserae of the courtyard
in CarΗ13 are laid parallel to the walls and densely (Fig. 7)3 87 In gen-
eral, the Carthaginian mortar pavements have a polychrome effect due
to the many fragments of varying colours. The overall design of the
FINDS
Scr Trani
The sculptural finds include sculptures of all dimensions and materials.
Only finds that have been found within the house units are included in the
catalogue. Finds of uncertain provenance, but found within habitation' "regs,
are listed in tables in order to supplement the material.
Finds found within the houses (Table 1 and Table 18): Only statuary in
terracotta came from precise contexts. The most frequent mythological theme
is the enthroned Baal Hammon documented in four houses. A fragmentary
sphinx that most likely decorated the throne of the god was found in
svs
Rakob 1979, 26.
397 Chelbi 1985, 83.
97
CarΗ13. Another common type is the female figure, also attested in four
houses. There is one example of a male figure and of an animal (ram). Finally,
a group consisting of a man and a female idol appears to represent a scene
from comedy (the area of CarM1).
The largest amounts of finds came from CarHl and Car Η13, the two dou-
ble units of the `Hannibal Quarter'. Of course, reservations about quantity
and distribution are obligatory due to the inaccurate information about find-
spots and the selective ,publication of finds.
statuettes were even more frequent, and there were several Bes statuettes.
Less frequent are Herakles (Table 19, no. 7) and Pan (Table 19, no. 38? and
Table 20, no. 3). Masks are also common.
The finds within the area of Carl l should be singled out. Numerous terra-
cotta statuettes of Baal (enthroned) and Bes were found. 400 More finds of ter-
racotta are listed in Table 21. One is remarkable because it is almost life-size
(Table 21, no. 3). Unfortunately, only parts of the hair are left.
RELIGIOUS FINDS
Censers representing Demeter are documented in CarHl3 and Cars.
There are three examples of uncertain contexts from the `Hannibal Quarter'.
One was found within the coordinates C,D,E-1,2,3, meaning that it came
from CarΗ3 or just outside this house 401 Two came from coordinate I-7 and
may be ascribed to Car Η7 or CarH8 402 Both the cult of Demeter and.Kore
and the prototypes of the censer were introduced from Sicily 403
EPIGRAPHICAL MATERIAL/GRAFFITO
Two black glazed vases, probably Campanian A, are of special interest be-
cause the name of the owner is inscribed on them: SPASINII. 404 They came
from coordinate I-6 and can be ascribed to either Car Η6 or CarH7. The in-
scription may be reconstructed as SP[UR[I] ASINII. It is tempting to sug-
gest a direct connection between the owner of the vases and the inhabitant of
the house, but the vases themselves are not conclusive evidence, since they
may have ended up in the house by accident. In any case, the presence of
Italians/Romans right down to the outbreak of the Third Punic War is attest-
ed by literary sources 405 Moreover, trade continued between Carthage and
Rome despite hostilities. This is documented by the large quantities of Italic
amphorae imported to Carthage between the end of the 3rd century BC and
146 BC 406
402 Ferron/Pinard 1960-1961, 153, nos. 457 and 458, pl. LXXXI. Additional .fragments were
404 Ferron/Pinard 1960-1961,142, nos. 393 and 394, pls. LXVI and LXVIl. These vases appear to
be lost today, since it has not been possible to locate them in the museum (Lancel in Byrsa 11979, 34,
note 64).
4"
Polyb. 36.7.
406 Fulford 1983, 7-8.
99
only one of the pavements may indicate a hierarchy between the two. Fina lly,
the courtyard in Cars also appears to be of importance due to its opus tessel-
latum mosaic consisting of marble tesserae.
Already by the 4th century BC there is evidence for the use of porticoes,
the bathtub-shaped cistern, opus tessellatum (monochrome), mortar pave-
ments and murals of the First Style. In the late Punic period enrichment of
the architecture is evidenced by the extensions of plots and addition of porti-
coes in the courtyards. This led to the creation of palatial houses (especially
in the `Mago Quarter' and the Quarter of Decumanus Maximus and Cardo
X). In the `Hannibal Quarter', by contrast, the porticoes were closed. The
majority of the bathtub-shaped cisterns date to this period. Geometric pat-
terns were employed in both tessellated mosaics and mortar pavements. The
lack of figurative motifs may admit of several explanations: that they have
not yet been found; that there was no taste for them because the Carthagín-
íans preferred simpler decorations; that Carthage was cut off before the in-
fluence from the Hellenistic East reached its culmination in the West. The
statuary mainly depicts Punic gods, Baal Hammon in particular, but Greek
and Egyptian influences are also present (various gods and theatrical
themes). Judging from the amount of terracotta statuettes found in the 'Han-
nibal Quarter', these must have been quite common.
Despite the adoption of foreign elements, for instance the Doric order,
bathtub "à sabot", First-Style wall-paintings and Egyptian mouldings, the
Carthaginian houses remained essentially Punic as evidenced by the opus
africanum technique, the Punic cubit, the long narrow corridors, porticoes
supported by pillars and the predominant bathtub-shaped cistern. Moreo-
ver, opus figlinum (sometimes combined with opus tessellatúm), floors of ter-
racotta tiles and especially the mortar pavements with their wide range of
colours (such as red, green, grey, yellow, brown in various shades) are char-
acteristic of the houses. The flowered frieze from the wall decoration is also a
specific Punic element.
CarΗ13 has been reconstructed as a double unit of 150 m 2. Here the vesti-
bule (room Β) is a small room of 3.10 m x 1.76 m. Special attention was giv-
en to the courtyard and main room. Despite the limited space, the courtyard
(room E) originally had a pórticus triplex and was distinguished from the
portico zones by the regular tessera decoration of the mortar pavement
(equally spaced rows) (Fig. 7). A pilaster and half-column decorated either
side of the entrance to the main room (room F), again having a monolithic
threshold. Fragments of wall decoration and of both opus tessellatum and
opus figlinum provide evidence for a richly decorated upper storey that may
have been reached by a staircase next to the bathroom (room G). Other
finds from the house include terracotta statuettes, for instance a sphinx
(from an enthroned Baal statuette) and female heads, and a censer.
Before leaving this quarter the graffito SPASINII inscribed on two vases
should be mentioned, as it might refer to inhabitants of either CarH6 or
CarH7. If so, Italians/Romans were living in the very heart of the Punic
metropolis.
The `Mago Quarter' is characterized by its large houses in its last phase.
Despite the severe destruction and incomplete ground plans, it is clear that
these houses made use of porticoes, various pavement types, including poly-
chrome opus tessellatum, and murals of the First Style, whereas sculptural
finds are rare. Due to the destruction, however, room identification is almost
ímpossiblé and as to the bathrooms most details of the installations, whether
wash basins, immersion tubs etc., have not survived.
CarM1 had at least two courtyards (rooms P 9, P 18), each with a porticus
triplex employing pillars. Apparently these two courtyards were part of a
symmetrical arrangement centred round room P 13. Mortar pavements pre-
dominate. A terracotta group, maybe representing a comedy scene, came
from the Punic destruction layer.
CarM4 was divided into two zones of which the ' νestern with an open
courtyard (room P 72) is interpreted as a service area and the eastern as the
reception area of the house. The eastern courtyard (room P 75) was provided
with a porticus triplex and an opus tessellatum mosaic. There are traces of
eight square bases for either pillars or columns. Both a main room (room P
78) and a bathroom with opus figlinum (room P 79) have been identified in
the eastern zone. The murals of the First Style belonged to an earlier phase
of the house. Other important finds are fragments of Doric columns and
capitals and a fragment of a Baal statuette (filling beneath Roman pavement).
CarM6 may have been the largest house of Carthage (1440 m 2). A corridor
(rooms Ρ 30, P 30α, Ρ 49, P49a) paved with opus figlinum gave access to the
house from the east side of the insula. The following elements are documen-
ted: a courtyard with two or three porticoes (room P 40), a bathroom (room
P 41), storerooms (rooms P 31-P36), five cisterns and one well, murals of the
First Style, opus figlinum and various mortar pavements. Upstairs rooms
were decorated with opus tessellatum mosaics.
Unfortunately, the latest phases of the Quarter of Decumanus Maximus
and Care/o X have not been dealt with in detail. However, there are indica-
tions of rich material: a paved street, a "<Peristyl>-Haus", the extension of
101
houses, several cisterns (only one identified as type I), opus tessellatum with
motif (only shown in photo) and terracotta tiles.
In Caro a corridor served as circulation space partitioning the house of c.
200 m2 in two halves with habitation in the western part and production in
the eastern part. The house was provided with a kitchen and a bathroom.
The latter consisted of two rooms, a changing room and the bathroom prop-
er. Interior architectural decoration includes mortar pavements, some with
tessera design (equally spaced rows), opus figlinum and mouldings in stucco.
Despite incomplete excavation and irregular shapes, Cars is noteworthy
because of the monochrome opus tessellatum mosaic of marble and staircase
in its courtyard (Fig. 5). Furthermore, the polychrome opus tessellatum mo-
saic decorated with stepped triangles from the bathroom, i.e. from the door-
way between the two rooms forming the bathroom, is unique. The three ni-
ches and bench in room 4 suggest that it served as a cult room. Fragments of
wall decoration and terracotta statuettes together with a censer have been
found
Finally, Car10 should be mentioned, as it has the only built-up bathtub "à
sabot" so far documented in Carthaginian houses. This stresses the impor-
tance of including the scattered remains from the city.
Remains of public buildings at Carthage are rare, thus there is not much
comparative material.
The Punic sanctuary located at the Rue Ibn Chabâat dates to the end of the
5th century BC and was renovated at the end of the 3rd or beginning of the
2nd century BC. Judging from the votive stelae that were found nearby in the
19th century, this sanctuary was dedicated to the so-called young Baal Mam-
mon (assimilated with the Egyptian Harpokrates) 4 07 Greek influence is attested
in the architectural elements coated with white stucco, for example Doric capi-
tals, Doric half-capitals and Aeolic capitals. Also a spout in the shape of a lion's
head from the early phase of the sanctuary follows Greek models4 08
A Tanit sanctuary was installed in the southern part (room E) of CarDMl
in phase V (6th and early 5th century BC) and was given up again in the 4th
century BC The most remarkable feature of the sanctuary is the five sym-
. 409
bols set with irregular tesserae in the surface of the mortar pavement dating
to the late 5th century BC. The symbols depict a sign of Tanit, a cross set
within a circle (related to Baal Hammon), a rosette (which may represent
both Astarté and Tanit), and the last two symbols probably depict a fish and
a rosette. These symbols prove that the Carthaginians used figurative and
floral motifs in pavements, but apparently only in a religious context. The
most important thing to note is that the use of tesserae in mortar pavements
is documented as early as the end of the 5th century BC.
The question is how the housing of Carthage relates to other Punic sites in
North Africa. Phoenician/Punic sites are located in a zone stretching from
present-day Libya to Morocco, and according to literary sources, they were
numerous 414 For the purposes of this work only sites within present-day Tu-
nisia are included.415 Unfortunately, only a few cities provide comparative
material.
Punic remains are scant at Utica. However, soundings below the Roman
insulae 1-3 have documented that an orthogonal layout of houses was em-
ployed sometime between the 4th century BC and the middle of the 1st cen-
tury BC when a major reorganization took place 416 Remains of mudbrick
walls and mortar pavements date to the 3rd century BC. The use of white
tessellated mosaics was introduced at the beginning of the 2nd century BC.
Remains from the 2nd century BC comprise stone socles, mudbrick walls
and bathtub-shaped cisterns. Houses with walls entirely built of stone also
occur.
Different stuccoed fragments from architectural decoration (for instance
two Ionic capitals) have now been dated to the Hellenistic period, and some
of the fragments have parallels at Carthage (the Byrsa hill) 417 Moreover,
First-Style wall-painting appears to be present. As to pavements, the motif of
stepped triangles occurs at Utica, but only in the Roman period. 418
umns). Porticoes (one, two or three): Rue de l'Apotropalon no. 21, Quartier du Four nos. 1 and 2,
Quartier du Boulevard no. 1. See table Fantar 1985, 127-128.
424 Quartier du Four nos. 1 and 2 (Fantar 1985, 139).
427
Fantar 1985, 148.
426
Fantar 1985, 297, fig. 3.
427
Fantar 1985, 572.
428 Fantar 1985, 603-617, pls. I-XXI.
429
Fantar 1985, 401-402. According to Fantar, no cisterns have been documented (1975, 10, note5).
430 Fantar 1984b, 499-514.
104
and red is dated to the 5th century BC by Morel. 431 However, no details on
the stratigraphy are given. The opus sectile pavements are of terracotta, the
tiles being lozenge-shaped, hexagonal and quadrangular. Several sub-types
of mortar pavements exist 432 They may have a densely set scattering of mar-
ble or limestone fragments. Mortar pavements with tessera design also occur.
It should be stressed that both geometric designs, floral and figural motifs
are present. The motifs are set with white marble or limestone tesserae:
equally spaced rows laid parallel to the walls of the room, uncertain linear
design 433 The figural and floral motifs include the sign of Tanit, fish and
flower. 434 The sign of Tank occurs three times. It is located in the courtyard
in front of the threshold to the main room. 435
The most characteristic features of the houses at Kerkouane are the open
courtyards provided with a well, the bathtubs and mortar pavements. 436 Por-
ticoes in courtyards were in use before the first destruction in the late 4th
century BC. Features to be singled out are the modest, but true peristyles,
the installation of bathrooms in almost all houses, the sign of Tanit decorat-
ing domestic pavements and mortar pavements with geometric patterns be-
ing earlier than the mid-3rd century BC.
The fort situated at Ras ed-Drek at Cap Bon should be mentioned in this
section, as it apparently Rrovídes the earliest examples of the bathtub-shaped
cistern in North Africa. ' The fort dates to the 5th century BC and contin-
ued to be used until the fall of Carthage. However, no information on the
stratígraphy is given. The fort had five large cisterns with an average length
of 5.28 in and an average width of 1.04 m. A bathtub-shaped cistern
(L-shaped) is also present in the sanctuary at Ras ed-Drek. It dates no earlier
than the 3rd century BC. 438
Punic sites and Punic influenced sites outside North Africa: In order to
supplement the material, some sites in Sicily and Sardinia will be commented
on. A general problem of the sites in Sardinia is the difficulty in distinguish-
ing between Punic and Roman phases (from 238 BC onwards) due to the
persistence of Punic building techniques (opus africanum) and architectural
elements. 439 In Sicily many cities were alternately under Greek and Cartha-
ginian control prior to the Roman conquest.
In Casa II A at Herakleia Mínia in Sicily the bathtub-shaped cistern be-
low the courtyard is dated no later than the first half of the 3rd century BC
439 For example Mezzolani 1996. For sites on S ar dinia, see Barreca 1986, especially 185-194.
105
(as is the layout of the house), and probably it goes back to the 4th century
BC. The courtyard is paved with square tiles of terrac οtta.44° Casa II B at
Herakleia Minoa was built in the first half of the 3rd century BC and is de-
scribed as a square closed unit having a corridor leading to the central court-
yard or the "atriolo". Upstairs rooms were decorated with First-Style wall-
paintings, mosaics and red mortar pavements with white tesserae. 441
Figurative motifs set with white tesserae in mortar pavements are docu-
mented at Selinunte in Sicily. The sign of Tanit between two caducei occurs
in a house situated on the acropolis, and in Temple A there are several mo-
tifs: the sign of Tank, a bull's head in a wreath and a caduceus. They prob-
ably date to the first half of the 3rd century BC, i.e. before the destruction of
the city in the First Punic War.
From Monte Sirai in Sardinia two buildings should be commented on.
The so-called Mastío on the acropolis was provided with a bathtub-shaped
cistern in the 5th century BC. 442 The second is the so-called Casa Fantar situ-
ated in zone B on the acropolis and built before the end of the First Punic
War (264-241 BC). 443 The rooms are arranged around a central courtyard.
Like CarH8 and the houses from Kerkouane, a drain for waste water is locat-
ed in the entrance room (corridor). Room c served as a domestic shrine.
At Tharros the bathtub-shaped cistern is documented in twelve houses
(casa n. 40, 41, 52, 51, 49, 64, 66, 6, 17, 21, 34 and 79) and in one public
building (Tempio delle semicolonne doriche), 444 A variant has rounded cor-
ners (casa n. 58, 68). Some of the cisterns date to the Punic period. However,
it is difficult to determine the time of construction, and the repair of the hy-
draulic mortar shows that the cisterns were used fora long period. 445
A very detailed account of the research and excavation of the city has al-
ready been given by E. Sanmartf and Rípoll and later by Mar and Ruiz de Ar-
bulo.447 Thus this section presents only an overview of the investigation of
the city.
The identification of modern Ampurias, Empúries in Catalan, with the an-
cient city of Emporion-Emporiae is ascribed to the bishop of Gerona, Joan
Margarit (1421 1484). Unlike most of his contemporaries in the Renaissance,
-
whose studies were purely philological, he actually visited the site (Paralipo-
menon Hispaniae libri decem, Granada 1545).448 A description of the ruins,
however, did not appear until 1609 (Jeroni de Pujades, Cor~nica universal
del Principat de Cathalunya, Barcelona). The first plans of the site were pub-
lished by Jaubert de Pass in his Notice Historique sur la ville le comté d'Em-
púrias, Paris 1823. He correctly identified the village of Sant Marti d'Empú-
ries as Palaiapolis mentioned by Strabo. 449
The first excavations financed by public funds were carried out in the pe-
riod 1846-1848 in the north-eastern part of the Greek city, i.e. Neapolis, and
in the forum area of the Roman city. The most renowned find of the nine-
teenth century is the mosaic depicting the Sacrifice of Iphigeneia in Aulis. 45°
The architect and archaeologist Puig y Cadafalch presented his project for
the excavation of the site in 1907; work began at the southern gate of the Ro-
man city in the following year, before being transferred to Neapolis a little
later in the same year. The systematic excavation took place in the period
1908-1936 with Gandía as field director. 451 During these years Neapolis was
totally unearthed. The Spanish Civil War brought the fieldwork to a tempo-
rary halt, but excavations were resumed by Almagro in 1940. In the 1940s
and 1950s work concentrated on the Roman city (city wall, houses and fo-
rum) and the necropoleis. Moreover, soundings were made in Neapolis. In
the period from the 1960s to the 1980s special interest was devoted to strati-
graphical studies, especially in the forum area of the Roman city and the
northernmost and southernmost parts of Neapolís. 452 In 1983 a study of the
architecture of Neapolis was presented. 453 But it was not till 1993 that the
first monograph to include all structures of the city, also comprising descrip-
tions of all houses, was published by Mar and Ruiz de Arbulo, 454 This work
447
Sanmartf, E./Rfpoll 1981; Mar/Ruiz de Arbulo 1993, 49-102.
448
Sanmartf, E./Ripoll 1981, 21; Mar/Ruiz de Arbulo 1993, 50.
449 Strab.3.4.8.
450
The mosaic was found in 1848 in the Roman city to the north-west of the forum. For photo,
see Marcet/Sanmartí, E. 1990, 147; Mar/Ruiz de Arbulo 1993, 55; Dunbabin 1999, 146, fig. 150.
Gandía's excavation diaries (Diarios de Eχcavaci~n) are kept in the museum at Ampurias.
452
For the latest bibliography on the city, see Sanmartf-Grego, E. 1996. For the forum, see Aqui-
lué et al. 1984.
453 Aquilué et al. 1983.
454
Mar/Ruiz de Arbulo 1993.
108
The ancient city of Ampurias consisted of three urban nuclei. In the pres-
ent work they are named as follows: Palaiapolis, Neapolis and the Roman
city (Fig. 8) 457 The first harbour lay to the north of Neapolis. At the end of
the 2nd century BC a breakwater was built to protect it. Due to increased
trading activity in the 2nd century BC, new facilities were required, and a
second harbour was established further to the south-east in the area of Riells-
La Clota. 458
The initial settlement, Palaiapolis, was originally located on a small island
just off the coast at the mouth of the river Fluvíà. Today the island is joined to
the mainland, and the ancient city is situated beneath the image of Sant Martf
d'Emρ ries that has been inhabited continuously and thus inaccessible for
major archaeological investigations. However, recent excavations have re-
vealed a part of the earliest Greek settlement dating to 575/550 BC. The re-
mains include for instance rectangular houses with stone socles and beaten
earth floors. 459 Traces found below the church of Sant Marti d'Em ρ~ries have
been interpreted as belonging to a temple of Artemis of Ephesos, the goddess
of the Iοnians.460 According to Strabo, a cult of the Ephesian Artemis existed
at Emρorion.461 The traces date to the last third of the 6th century BC, and a
fragment of an Archaic frieze depicting two sphinxes is supposed to come
from this temple together with an Ionic capital and an altar. 462 The late Impe-
rial city wall is partly preserved. 463 After the abandonment of both Neapolis
and the Roman city, Palaiapolis was the only inhabited nucleus.
The name Neapolis is modern. It was given by Puig y Cadafalch in 1908 in
order to define the urban area south of Sant Marti d'Empúries (Fig. 9). 4'
Life in this part of the city lasted from the 6th century BC into the second
—~~
-_ — -.___ . __~~ -
~~r--
%_~ `
~~ ,
,
~
;
~~
~
~
`~r. '
...-
®\~^ξ
~~~ ~
/Ι
r~ ~ -,Neaιiö~1 5
α alaíaplis
~~ ~ ~
Fig. 8. The three urban nuclei of Ampurias (Mar/Ruiz de Arbulo 1993, 204).
half of the 1st century AD. In its last phase Neapolis occupied an area of c.
465
3.6 hectares or c. 250 m x 145 m The city grew from the north towards the
south, since the earliest material comes from the northern sector and can be
dated to the first half of the 6th century BC. In the southern part of the city
the oldest remains date to the first half of the 5th century BC. 466 The earliest
part of the city wall goes back to the third quarter of the 5th century BC. A
rectangular tower located behind the so-called Asklepios temple was part of
this wall. A suburban sanctuary dating to the second half of the 5th century
465
Domínguez Monedero 1986, 3.
466
Sanmartí-Grego, E. 1992b, 178-179.
110
461
Sanmarti- Grego, Ε. 1990, 399-405. -
468 Sanmartí- Grego, Ε. et al. 1986; Sanmarti, Ε. et al. 1988; Sanmarti- Grego ; E. 1988b; Sanmarti ~~
Grego, Ε . et al. 1991; Sanmartí- Grego, Ε . et al. 1992.
469
Sanmartí, E./Noila 1986. .
~ιι ~`τ""'tí■ιιιιΠ Ε . ~
·
Street 9 1 ~ Ü . .
1 ~~ ^y
r'. ~ Jr ιιi'~ ιM .νπl ~
..ιιΥ
ι ~± *
τl ατMtEwιιιTσ~rn..T~ιτuw λl
i Í θι σ.- υ7 ~πχι nm J ι { Π ι.fiΛσι.ιιηΝ g νι.ι
í αΡ1. ~~αW
~ι. , ι ιυιιιυπι ι
.η.ι~ α ~αι•
9~
Street 5
Ampurías, Neapolis, plan (Mat/Ruiz de Arbulo 1993, sep arate
υ.. η..• •ι ..
rr υ,..•.ω.~
1 11
Transverse wall
j l
~~ Γ `s-J1 Jt ~~
Amphitheatre Palaestra
100m.
Fig. 10. Ampurias, Roman city, plan (Aqullué et al. 1984, fig. 13).
112
~.
s~
ο + + + + + + + + + + + ε ο
o ο — σ
+
F ττ
σ
-50R1
i ❑
0 Ι❑
—
~
ε_ ~
.-.
σ Q D
loB
li
Fig. 11. Ampurias, Roman city, plan, forum and so-called praesidium (Mar/Ruiz de Arbulo 1993, 190).
113
ern city wall of the former and the eastern of the latter were dismantled, and
a new wall was erected between the two nuclei. This fusion of the two urban
nuclei is connected with the creation of the Municipium Emporiae which ap-
pears to have taken place between 36-27 BC 475
475
Fabre et at.. 1991, 17-18.
476
A fifth house has been located west of AmpRi (Mar/Ruiz de Arbulo 1993, 94).
47 Ampli and AmρΝ9.
Second phase
j 43 42 41
1,1
49.
Fourth phase
II 55.
50·
a
lo η.
Fig. 12. Ampurias, Roman city, AmpRi, plan, last phase with indication of building phases (Santos
Retolaza 1991, 30, fig. 13).
' Santos Retolaza 1991, 22; Mar/Ruiz de Arbulo 1993, 395; Carri ~n Masgrau!Santos Retolaza
1995, 113.
ο
Μ
~Φ
C)
Μ
η
Μ
J' ❑
Γ
η
M
Γ
Q
Ε
ο'-
η
α
~
Μ
Ν
~
Ν~')
tú
~~
~
Μ
~r
Fig. 14. Ampurias, Roman city, AmpR4, plan (drawing by Kjeld de Fine Licht).
area within the house plot. On the other hand, the cistern in a given room is
not necessarily evidence for a courtyard in that particular room, as the rain-
water could reach the cistern situated below a covered room by means of
pipes or channels. Consequently, when roof systems and water collection sys-
tems remain unknown, and other evidence is lacking, the cistern alone is not
conclusive evidence. At the same time some courtyards may not be detect-
able in the archaeological record. Apart from the obvious courtyards (pro-
vided with porticoes), courtyards have been identified in the present work
when more criteria appear to be present, for example cistern and inflow
channel and/or drainage channels.
ARCHITECTURE
ARCHITECTURAL ANALYSIS
Ground plan (Table 22): Four types of ground plans occur: rectangular,
square, irregular and L-shaped. Most frequent is the rectangular ground plan
with a total of sixteen examples. However, some of these have minor irregu-
larities and/or trapezoidal shapes ( ΑmρΝ3, ΑmρΝ8, Ampi 12, ΑmρΝΙ5,
ΑmρΝΙ ~, ΑmρΝΙ7). The second largest group consists of irregular ground
plans (nine houses). There are five more or less square houses. Finally, three
houses are L-shaped. The ground plan of five houses is unknown ( ΑmρΝ22,
ΑmpΝ25, ΑmpΝ27, ΑmpΝ31, ΑmρR4).
Dimensions (Table 22 and Table 23): The largest houses are situated in
the Roman city. AmpRi occupies an area of 3492 m 2 and AmpR3 an area of
2128 m2. These huge areas were achieved by several extensions. In leapolis
the areas range from c. 50 m2 (Ampli) to more than 868 m2 (Αmp134). 483
The majority of the houses where the area is known belong to the group
101 200 m2.
-
Entrance room (Table 22): Almost half the houses (seventeen) have a sep-
arate entrance room. Sometimes it is a long and narrow corridor. This is the
case when the house is situated in the middle of an insula (ΑmρΝ21,
Αmρ125, Αmρ127). Corridors also occur in Αmρi6, ΑmρΝi4 and
ΑmρR4. However, the majority of the entrance rooms are small rectangular
or square rooms. In ΑmpΝ34 there is an arrangement consisting of an
L-shaped room (room 2) which opens onto a rectangular room (room 3a)
and a corridor (room 3b). Both the rectangular room and the corridor lead
to the peristyle. The entrance door of Am ρ 3 is recessed from the street.
463 The eastern part of the house is situated below the modern coastal path.
118
The northern peristyle (room 35a•) of ΑmpR3 is provided with porticoes on three sides and a
wall on the fourth side.
119
as a shop is based primarily on the door opening onto the street. This argu-
ment, however, is not conclusive, since the doorway directly to the street
may be explained by the wish to have a separate entrance for guests.
The original layout of the houses in the north-eastern part of the Roman
city included a tablinum east of the atrium constituting the main room
(AmpRi: room 37; ΑmρR3: room 52•). On either side of the tablinum in
AmpRi is a smaller room opening onto the atrium. The present main room
in ΑmpR2, judging from the size, is room 9, which is not on alignment with
the atrium. The extensions in AmpRi and ΑmρR3 included numerous large
reception rooms and the old tabling probably became less important. The ta-
blinum was maintained in AmpRi, while in AmpR3 it was reduced to a pas-
sage. A hierarchy of reception rooms may have existed, but it cannot be de-
termined which one was the most important.
Ampi21 only consists of four rooms: the entrance room, the peristyle and
two large rooms of almost identical size situated to the north of the peristyle.
Room 3 is c. 38 m2 and room 4 c. 33 m2 (interior dimensions). Judging from
size alone, room 3 ought to be the main room. As to Am ρΝ29, it has been
suggested that room 2 served as the main room because of its ceramic pave-
ment 485 However, the relationship between this room and the rest of the
house is not clear, as the room is situated on a higher level. Furthermore, the
pavement is modern, and the identification must be rejected. 486 In AmρR4
the largest room is room 13 of c. 24 m2 situated to the north of the courtyard.
Another argument in favour of its identification as the main room is its mor-
tar pavement embellished with a central emblema measuring 1.2 m x 1.2 m.
The majority of the main rooms do not have additional rooms: In
AmpliO there is a smaller room at the rear (room 2). The main room of
AmρΝΙ3 is combined with a smaller room on either side, only accessible
from the main room.
The location of the main room is often to the north of the circulation
space or in the northern part of the house. This occurs in eight houses
(AmρΝ2, ΑmρΝ5, ΑmpN7, ΑmρN13, ΑmρN22, AmρΝ26, ΑmρΝ32,
AmρΝ34). In three houses the main room is situated to the west (AmpliO,
AmρΝ17, AmρΝ25) and in two houses to the south (AmρΝΙ4, AmpliS).
In the houses situated in the north-eastern part of the Roman city the main
rooms are situated east of the atria. The size ranges from c. 17 m2 to c. 29 m2.
One exception is AmpN34 where the main room is c. 57 m2.
Other rooms (Table 22): Only a few service rooms have been identified
comprising one bath and five kitchens. The bath complex (rooms 30, 31, 32)
belongs to ΑmρR3 where it was added in a later phase. It consists of a set of
three rooms built behind one another on a N-S alignment: the southernmost
room is the apodyterium (room 30), followed by the tepidarium (room 31)
and the caldarium (room 32). Room 33 is a combined kitchen-ρraefurnium. 48'
Four houses in Neapolis had a kitchen, judging from the presence of a
487 Ρa1auí/Viν~~1993b. The authors date the bath complex to the middle of the 1st century AD.
120
Upper storey (Table 22): Upper storeys are very difficult to detect ar-
chaeοlοgically when the site is poorly preserved. Conclusive evidence does
not exist, apart from the lowermost step of a staircase in room 17 of ΑmρR4.
Due to large differences in levels within the house plots of ΑmpΝ2, ΑmρΝ3,
ΑmρΝ8 and AmρΝ30, an upper storey is reconstructed above the area of
the houses situated on the lower part of the plot. Mar and Ruiz de Arbulo
suggest that ΑmpΝ9 also had an upper storey because of the thickness of the
walls 491 However, the thickness of the walls of this house (c. 0.5 m) is not ex-
traordinary. The thickness of walls of other houses ranges from approximate-
ly 0.45 to 0.95 m492 In ΑmρΝ7 the small corridor situated between rooms 3
488
Puíg y Cadafalch 1915 1920, 700; Mar/Ruiz de Arbulo 1993, 357, 369 and 403.
-
489
limos Rimera 1989, 50-52.
490 Proportions alone are not conclusive evidence (cf. Dunbabin 1996, 67-70).
Water supply (Table 24 and Table 25): 494 The cisterns are characterized by
walls built in either sandstone or limestone masonry lined with hydraulic
mortar. The cover takes the shape of slabs placed horizontally or of a vault. 495
Based on their ground plan and section, the cisterns can be divided into the
following types:
Quadrangular (rectangular and square), vertical walls.
Circular, vertical walls.
Irregular.
Bathtub-shaped, vertical walls (Figs. 15-16).
Circular, bottle-shaped.
As it appears from Table 24, type IV (bathtub-shaped) is predominant
with a total of twenty-three examples. It is predominant in Neapolis (seven-
teen out of eighteen). The only other type represented here is type I (quad-
rangular). The alignment of this cistern is different from the present house
(AmpΝ7), and the cistern is therefore ascribed to an earlier period. On the
other hand, Burés Vilaseca has suggested that the different orientation may
be explained by the presence of earlier structures. 49εΡ In the Roman city a
greater variety of shapes occurs, but still type IV is the most frequent. How-
ever, it should be stressed that this type has often been modified with respect
to shape (broader, for instance Am ρ 3 ), materials (limestone instead of
sandstone) and cover .(always a vault). Nevertheless, it is worth noting that
even here all houses made use of the bathtub-shaped cistern. Type II, III and
V are represented by one example each.
Only one house in Neapolis is provided with more than one cistern
(AmρN29). In the Roman city AmpRl, the largest house at Ampurias, has no
less than nine cisterns. There is documentation for cisterns in half the houses
in Neapolis and in all excavated houses of the Roman city. 497 The inhabitants
of houses without private water provision had to use the public cisterns. In
Neapolis these are situated for example in the stoa and the small market. It is
clear from Table 25 that the cistern in most cases is located below the atrium,
courtyard or peristyle. The cisterns in AmpNS, Am ρΝ26 and ΑmρΝ32 are
partly situated under covered rooms (main rooms). In AmpN 13 the cistern is
situated beneath a covered area. This shows that one cannot be sure that the
space occupied by a cistern is always an open courtyard, since the cistern sit-
uated below a covered room may be supplied via pipes from the uncovered
area. The average capacity of a cistern in Neapolis is 14.8 m 3 . In AmpR1 the
total capacity was 349.3 m3 and in AmρΕ3 106.5 m3 498 In Neapolis the cis-
terns often have a width of approximately 1 m, whereas they are broader in
the Roman city.
The question of the dating of the cisterns in Neapolis must be left open. It
cannot be determined whether they all belong to the late Republican/early
Imperial period or whether they were reused in the houses from this period.
Burés Vilaseca has suggested the following development of the cisterns
based on shape, materials and cover: The oldest is the bathtub-shaped type
built in sandstone and with a flat cover. Next comes the bathtub-shaped type
in sandstone and covered with a vault. The latest group consists of rectangu-
lar and bathtub-shaped cisterns built in limestone and with a vault in opus
caementicium4 99 In the Roman city the cisterns situated in the atrium zones
belong to the earliest phases of the houses, i.e. the beginning of the 1st
century BC.
The use of wells is documented in three houses only. All are situated in
Neapolis (AmρΝ21, ΑmpΝ31, AmρΝ34). These houses are also provided
with a cistern. Two of them, Am ρΝ21 and AmρΝ34, have large peristyles,
and ΑmρΝ31 is an atrium house.
It is hardly a coincidence that the cisterns of largest capacity, i.e. size or
number, or the combination of cistern and well occur in the largest houses:
ΑmρΝ34 (cistern and well), ΑmρΝ29 (two cisterns), ΑmρΝ25 (cistern with
three chambers), AmρΝΙ3 (cistern of 17.3 m3) and AmρΝ21 (cistern and
well). However, ΑmρΝ2 with its 370 m2 only has a cistern of 12.1 m 3 . The
opposite occurs in AmρN32 having an area of 84 m 2 and a cistern of 26.7 m 3 .
In the Roman city the largest capacities and numbers of cisterns occur in
AmpR1 and AmpR3 which also are the largest houses here. In Αmρ 3 there
was an extra need because of the bath.
Disposal of waste water occurred by means of drains and pipelines. In
Ampi 15 the rainwater collected in the impluvium was channelled directly to
the street via a drain.
Gardens: Urban gardens belonged to the plots of ΑmpΝ6 and AmρN7.
Both houses are situated at the western edge of the city, and the garden
zones are to the west of the houses, thus occupying the area bordering the
city wall. Also AmρN22 and ΑmρΝ25 at the western and northern edge re-
spectively appear to have been provided with,gardens or terraces. In the Ro-
man city the addition of large peristyles with gardens and other garden ar-
rangements in AmpR1 and AmpR3 is not surprising, since this was the
standard way to enrich atrium houses as documented in the Vesuvían cities.
Whether the gardens in Neapolis were utilitarian gardens or flower gardens
cannot be answered. However, it is important to note their presence, since
both the archaeological and literary sources show that in Greece gardens
were chiefly confined to suburban and rural habitation in the Classical and
Hellenistic periods 500
DIACHRONIC PERSPECΙIVÉ
sa
Santos Retolaza 1991, 22-30.
502 Sanmartí-Grego, E./Santos Retolaza 1986-1989, 297.
τΡο; Santos Retolaza 1991, 29. Mar and Ruiz de Arbulo suggest a testudínate atrium (1993, 395).
5Ó4 Mar/Ruiz de Arbulo 1993, 395-397.
sis
Mar/Ruiz de Arbulo 1993, 396-397.
126
CONCLUSION
From the analysis it can be concluded that the elements in consideration
do not constitute a distinctive typology or well-defined types of houses, í.e. a
certain ground plan is not related to a certain size, circulation space and
main room. Houses of all shapes and sizes are present, from the most simple
consisting of only three or four rooms to huge villa-like houses. This applies
both to Neapolis and the Roman city. The most common features are the
rectangular ground plans, the simple courtyards without porticoes and the
main rooms located north of the circulation space. As regards the circulation
space, the majority of the houses have courtyards followed by houses with an
atrium. Main rooms are oblong, broad or squarish. The water supply of the
houses depended mainly upon cisterns. Wells are rare. Upstairs rooms do
not appear to have been common, but this may be due to the poor state of
preservation.
A criterion for a general classification could be the nature of the circula-
tion space, i.e. atrium, courtyard, peristyle, but this cannot be a rigid classifi-
506
Santos Retolaza 1991, 31.
507
Santos Retolaza 1991, 31-33.
500
Mar/Ruiz de Arbúlo 1993, 391-394.
Fig. 17. Ampurias, Neapolis, atrium houses and atrium-peristyle house (plans of the individualhouses, /Ruiz de Arbulo 1993, sep arate plan
127
128
cation and sub-types or variations should be allowed to exist within this clas-
sification.
Atrium houses and atrium-peristyle houses (Fig. 17): There are five atrium
houses and three atrium-peristyle houses. The atrium houses situated in Nea-
polis display fairly individual layouts due to the fact that they had to be in-
stalled in a pre-existent urban setting. It was not possible to lay out a conven-
tional atrium house with fauces-atrium-tablinum sequence as it can be ob-
served in the first phase of the houses in the Roman city. In this connection it
should be noted that according to Zaccaria Ruggiu, there are no atrium
houses in Neapolis 509 Despite the individual solutions, some common fea-
tures also occur. One of the atrium houses in Neapolis, AmpNS, developed
into an atrium-peristyle house. In the Roman city AmpRl and AmpR3 had
more circulation spaces, including peristyles.
' AmρΝ7 has a more or less square ground plan, a laterally placed entrance
room, a Tuscan atrium (room 1) and a main room (room 2, also dining
room) north of the atrium. Moreover, the northern zone is divided into three
with a room on either side of the main room (rooms 4, 5). These side rooms
are of different sizes. The house is provided with a garden to the west.
In AmpI 13 the ground plan is irregular. There is a laterally placed en-
trance room (room 1), a tetrastyle atrium (room 2), a main room (room 8)
north of the atrium and tripartition of the northern zone. The main room has
a room on either side (rooms 9, 10). Access to these side rooms was gained
via the main room. The southern zone of the atrium extended towards the
west (room 3).
AmpI 15 is rectangular. There is a laterally placed entrance room on the
west side (room 5) and an entrance on the north side through room 2. A
double door divided by a column separates room 2 and the Tuscan atrium
(room 1). The main room (room 7) is situated south of the atrium.
In AmρΝ31 the tetrastyle atrium is entered directly from the street. The
rest of the house has disappeared.
AmpR2 of a rectangular ground plan has an entrance room (room 1) and a
tetrastyle atrium (room la.), but no tαblinum. There is a fauces-atrium se-
quence, but the main room (room 9) situated east of the atrium is not on
alignment with the fauces and atrium.
Amρi5 is irregular. There is no entrance room, but the-entrance is oppo-
site the tαblinum (room 5), however, it is not on alignment. The house has a
Tuscan atrium (room 1), a main room (room 5) north of the atrium, a dining
room (room 6) west of the atrium and a laterally placed peristyle (room 9).
In AmpRi the fauces-atrium-tablinum sequence is present. There is a Co-
rinthian atrium (room 30a•), a laterally placed peristyle (room 46), two
courtyards (rooms 12, 50•) of which one has three porticoes and a second at-
rium (room 18, Tuscan?).
ΑmpR3 also has the fauces-atrium (Tuscan)-tablinum sequence. Moreover,
the house has two peristyles (rooms 40a•, 35a·) of which the southern was
—
o
Ε
Ε
2:
οο
'-
Ε
Ε
129
added on the axis of this sequence. Further towards the east was a garden
zone with porticoes.
Houses with courtyard (Fig. 18): Although an open courtyard is present in
ten houses, these do not constitute a homogeneous group 5 1 The houses vary
in shape (rectangular, L-shaped, irregular) and number of rooms. Moreover,
the courtyard itself is located very differently:
At the front: Amp ΝΙ9.
At the back: Amp Ν3.
Centrally: ΑmρΝ8, AmpΝ25, and AmpR4.
Laterally: Ampl\i32.
Laterally, at the front: Am ρΝΙΙ.
Laterally, at the back: Amphi, AmρΝΙ2, and AmρΝΙ ~.
Two of the houses have an internal division with four rooms of almost
identical size: Ampiii (irregular ground plan) and Ampi16 (more elongat-
ed ground plan).. AmρΝ3 and AmpΝ8 are both rectangular but of a more
elongated and narrow shape. Ampi 17 only has three rooms: one large room
and two small ones.
ΑmρΝ2 may have had two courtyards: room 9 (porticoes on three sides)
and perhaps room 3 where the cistern is located. This may have been sup-
plied via a pipe from the room 9. Ampi28 perhaps had a courtyard due to
the presence of a cistern beneath room 1. Courtyard with three porticoes oc-
curs in AmpΝ2 -and AmpR1 (room 12).
A group of houses has no identifiable circulation space and no cisterns:
Ampli, Amρl4, AmρΝ6, Ampl9, Ampl10, AmpΝΙ8, Amρl20,
Amρ124, AmρΝ30 and Amρl33. Several of these are similar in ground
plan and internal division. The first type consists of houses with three rooms,
one large and two small ones: Ampli, ΑmρΝ18, Amρl20 and Amρ133.
The second type has four rooms of almost identical size: Ampi4 (plus front
corridor), Amp Ν6 (plus corridor down the middle of the plot), Am ρΝ9,
Amp124, Amp Ν30. Their layouts are identical with or similar to Amphi,
AmρΝ16, AmρΝ17 and Amp128 where a courtyard has been documented
(AmρΝ28 only likely), and thus it cannot be ruled out that they were provid-
ed with a courtyard, although the final evidence is lacking. More important,
however, is that all these houses are the most common at Ampurias.
Peristyle houses (Fig. 19): According to Zaccaria Ruggiu, the houses situ-
ated in ieapolis "si articolano alla «maniera greca» con al centro il peristi-
h ο ".511 However, only two peristyle houses are documented, thus her sen-
tence does not make sense. Maybe Zaccaria Ruggiu regards the tetrastyle at-
ria as small peristyles. This, however, would only add two houses to the
group (Amph13 and Am ρΝ31) and even with the peristyle of Am ρΝ5 and
the courtyard with three porticoes of Amp12 included, six out of thirty-four
houses hardly form a majority. Moreover, with respect to the latter two hous-
510 Since the open courtyard (room 50•) of AmpR1 was a secondary circulation space, it is not in-
cluded here.
51 Zaccaria Ruggiu 1995, 380.
130
Fig. 19. Ampurias, Neapolís, peristyle houses (plans of the individual houses, Mar/Ruiz de Arbulo
1993, separate plan).
es, the peristyle and the courtyard with porticoes are placed at the margin of
the plots.
The peristyle houses are: AmρΝ21 that has a rectangular ground plan, an
entrance room (narrow corridor), a peristyle a nd two rooms north of the
peristyle. Amp34 has an irregular ground plan, many rooms and an axial se-
quence of entrance room, peristyle and dining room. It belongs to the group
of so-called axial peristyle houses: a type which was widely used in the Impe-
rial period in the western empire 512 The main features are an entrance area, a
peristyle at the centre of the house (formerly placed at the rear or at the mar-
gin of the house plot) and a dining room. Moreover, there is a fountain in the
peristyle located opposite the main room.
Other houses (Fig. 20): Several of the houses from Neapolis do not fit into
the above categories. Am ρΝΙ4 is L-shaped and the rooms are arranged
around the corridor. In Am ρΝ23 a central covered room with a pillar func-
tioned as circulation space.
As mentioned earlier, a circulation space occurs in Am ρΝ22, AmρΝ26,
ΑmρΝ27 and AmρΝ29. Cisterns are present below this space in ΑmρN26
and Amρ129. According to Mar and Ruiz de Arbulo, AmρΝ26 and
ΑmρΝ27 should be interpreted as atrium houses, although there is no evi-
dence for an impluvium. The identification of Am ρΝ26 is based on the inter-
relationship between the circulation space and room 2 to the north, thus be-
512
Meyer 1999.
131
Fig. 20. Ampurias, Neapolis, other houses (plans of the individual houses, Mar/Ruiz de Arbulo 1993,
separate plan).
teen examples. Houses with a central circulation space or an atrium form the
second largest group (group 1). Six houses are not arranged around a court-
yard but defined by their corridor or covered central distributor. Finally,
there are four houses with a peristyle or porticoes in the courtyard.
Group 1 is problematic because it includes houses with an open courtyard
as well as atrium houses. An open courtyard and an impluvíate atrium are
two distinct architectural elements. In AmpI 12 Mar and Ruiz de Arbulo
suggest that room 3 was the courtyard. Room 5 is a more likely candidate be-
cause of the drainage and in that case, the courtyard is laterally placed at the
rear. The courtyard in Ampi Ι ~~(room 1) is not at the centre of the house,
but laterally placed at the back. Am ρΝ23 does have a central circulation
space (room 1), but it appears to have been covered, judging from the pillar
at the centre of the room. Thus it would be more appropriate to place the
house in group 4. At the same time Mar and Ruiz de Arbulo suggest to iden-
tify room 4 to the north as a courtyard. The ground plan of Am ρ125 is not
perfectly clear, but the courtyard (room 1) appears to be more or less cen-
trally placed. As regards group 2, Am ρΝS is an atrium-peristyle house and
may be placed in their group 1 as well. Of the fourteen houses ascribed to
group 3 there are only three (Ampli l, Αmρi17, Amp132) with evidence
for a courtyard 515 The rest of the houses are ascribed to the group because of
resemblance to Ampli i, AmpΝΙ7 and Αmp128 (possible courtyard in
room 1), i.e. a more or less square or rectangular ground plan with an inter-
nal division into three or four rooms of which one is supposed to be the
courtyard 516 However, some houses have more rooms and different internal
layouts (AmpliO, Αmp122). Moreover, there are examples of inconsisten-
cy. The ground plan of Amρ132 occurs together with the houses defined by
their corridor or covered central distributor. 517 ΑmpΝ14 is arranged around
a corridor and thus belongs to group 4. Am ρΝ3 and AmpΝ19 are also pro-
vided with a courtyard located at the back (Amp13) or at the front
(AmρΝ19). As to Amp18, the basis for the classification should be the
courtyard (room 2) rather than room 1 at the front. Am ρΝ29 has a circula-
tion space connecting the two wings of the house. The presence of cisterns
and drainage points to an open space rather than a covered one.
In summary, Mar and Ruiz de Arbulo do not distinguish clearly between
atria and central courtyards. Furthermore, they do not even consider that
there might have been houses without courtyards. Am ρΝ14 is the only house
where they have not tried to identify either a courtyard or an atrium. Houses
with courtyards not centrally placed are instead classified on the basis of: 1)
A covered room located centrally, but in Am ρΝ8 this covered room is not
central; 2) The lateral location of the courtyard. It appears, however, more
reasonable to keep the courtyard houses together and subdivide them within
the group if necessary. Despite the different classifications, it is clear that the
516 "La cοnserνación del sistema hídraulíco en la mayoría de los ejemplos emporitanos permíte
identificar uno de los cudrantes comp el patio de la casa" (Mar/Ruíz de Arbulo 1993, 353).
517 Mar/Ruiz de Arbulo 1993, 405.
133
Fig. 21. Alorda Park, houses in the western sector, 4th century BC (drawing by Kjeld de Fine Licht).
Because of the Greek origin of the city, the traditional typology and termi-
nology of Greek houses have been employed to some extent, but there are
no pastas, prostas or pastas-peristyle houses. 518 Greek features in the domestic
architecture of Ampurias are confined to courtyards with porticoes, the peri-
style court in AmρΝ21 and the preference for locating the main room to the
north of the circulation space. Moreover, the room complex in Am ρΝ13
with main room and a room on either side (rooms 8, 9, 10), accessible from
the main room, can be paralleled with palaces and houses from the Hellenis-
tic East, for instance at Pella, Pergamon and Delos.
Roman features are the vault used as cover for cisterns, impluviate atria, the
axial peristyle house where a centrally placed peristyle displaced the atrium
(AmρΝ34), peristyle gardens and peristyles added at the margin or at the rear
of the houses (AmρN5, AmpRi and AmpR3). There is a clear difference
between the atrium houses of the Roman city and those of Neapolis where they
were fitted into a pre-existent urban setting. The bath in AmpR3 follows a
518 Balil identifies the main room of Am ρΝ7 as a prostas and finds the layout similar to the houses
at Olynthos and Priene (1972a, 43).
134
model used in Roman houses and villas in Italy, both with respect to the layout
(the rooms built behind one another) and the location next to the kitchen, for
example House VI.11, 8-10 in Pompeii (Casa del Labirinto). 519
The Punic element is represented by the bathtub-shaped cistern. Both
shape, technique, dimensions and cover can be paralleled in the Punic world
(Table 15 and Table 25).
The internal layout of three or four rooms represent the most characteris-
tic element of the houses in Neapolis. Houses of an identical plan, with three
rooms, and size are documented at the Iberian site of Alorda Park (Fig. 21).
Two house units display this arrangement: C-D-O and L-S-V 5 20 There is one
large rectangular room occupying half the plot and two smaller ones, almost
square. The size of this house type is approximately 60-70 .m 2 . The largest
room is identified as "l'habitació «noble»" used as living room and reception
room521 Both units go back to the beginning of the 4th century BC and con-
tinued to be in use until e. 200 BC5 22 Ampli is composed of a transverse
front room with two rooms at the back. To some scholars the front room
would be a pastas, but this arrangement can also be found in House E3 at the
Iberian site of Puff Castellet de Lloret de Mar, though the ground plan here
is more elongated. 23
519
Fabbricotti 1976.
520 For plan, see also Sanmartí, J. 1998, 92.
521 Belarte/Barberà 1994, 45.
522
Sarimarti, J./Santacana 1992.
523
Lu rens í Rams et al. 1989, 112 and 113, fig. 2.
524
Description and sketch by Gandía (1918, 305).
525
Made of "morter" (Gandía 1935, 103-104).
526
lieto Prieto 1977; lieto Prieto 1979-1980; Carrión Masgrau 1992; Carrión Masgrau/Santos
Retolaza 1993; Carrión Masgrau/Santos Retolaza 1995.
135
527
There are five categories in all: esquema 1-5 (lieto Prieto 1977, 852 and 856-866). lieto Prie-
to 1979-1980, 329, fig. 64.
528
lieto Prieto 1977, 856-858; lieto Prieto 1979-1980, 283-313. For room 26, see lieto Prieto
1979-1980, 284, fig. 2.
529
lieto Prieto 1979-1980, figs. 3-14. It was found in the north-western zone of the house.
5"
lieto Prieto 1979-1980, figs. 15-56.
531
lieto Prieto 1977, 856; lieto Prieto 1979-1980, 319-322; Carrión Masgrau 1992.
532
Carrión Masgrau 1992, 82-83, figs. 2-3.
533
Carrión Masgrau 1992.
534 lieto Prieto 1977, 859-862; lieto Prieto 1979-1980, 322-335; Carrión Masgrau/Santos Retola-
556
The large collection of black-and-white mosaics in opus tessellatum and the opus sectile pave-
ments from the houses in the Roman city will not be dealt with in detail (for example motifs), since
they are part of the doctoral thesis being prepared by M. Santos on these houses. Furthermore, these
pavement types cannot be paralleled to material from Neapolis. Therefore, the focus is on the mortar
pavements with tessera design widely employed in both nuclei.
53' For instance Mar/Ruiz de Arbulo 1993, 386 (impluvium of Amp Ν5).
538 Mar/Ruiz de Arbulo 1993, 406.
539
Room 4 of AmρN34 might have had an opus tessellαtum pavement ("pavimentada en mosaí-
co"), but only the preparation layer is left (Mar/Ruiz de Arbulo 1993, 389).
540 AmρΝ2: Mar/Ruiz de Arbulo 1993, 387.
541
Ping y Cadafalch 1908, 191, fig. 40: "Casa romana en el recinte grec, ab el paument d'opus
testaceum en l'habitació, y ab el marxapeu de la porta decorat ab mosaích".
542
For photos, see Marcet/Sanmarti, E. 1990, 162.
137
Fig. 22. Ampurias, Roman city, AmpRi, room 49, opus tessellatum (author's photo).
543 The latter is very fragmentary. For photos, see Balíl 1961, 42, fig. 1; 49, fig. 2 and 51, fig. 3;
Marcet/Sanmarti, E. 1990, 142 and 157; Santos Retolaza 1991, 27, fig. 9.
544 There appears to be some confusion of the provenances of the emblemata. According to Balil
and Santos Retolaza, the three panels all came from AmpRl (Bali' 1961, 41; Santos Retolaza 1991,
27). Ma_cet and Sanmarti write that the one with the mask came from a house of uncertain location in
the Roman city, while the one with the partridge was found in the Roman city (1990, 142 and 157).
545
Balil 1961, 47 and 51-52; Santos Retolaza 1991, 27.
138
Fig. 23. Amρurías, Roman city, ΑmpR3, implu νiu;z, opus tessellatum with inset pieces (author's photo).
139
Fig. 24. ρLrias, Ν·mρ lis_ ΑmρΝΝT7, atrium, mortar pavement with inset pieces (anth οr's photo).
lican time (end Df the 2nd/early 1st century BC and the Sul=an period). 546
The twn other motif=_ also find counterparts in Pompeii. The Pompeian ex-
ample with the bird and box dates no later than 70-60 BC 54 '
Gpu_ se:tile: This type is only present in the Roman city where it is found
in Amp•Ri. Αmρ13 and ΑmρR4. As to the latter house, only the imprint of
sixteen slabs forming: an emblema of 1.2 m x 1.2 m is left in room 13. J48 Ac-
ccrding tc the e ζcaνator, the slabs were of terracotta, but marble or lime-
stone has also bin suggested 549
Mor_ar bavelnents: In general, the mortar pavements have a monochrome
reel or :eddish colour. Some contain a scatter of sma π irregular marble frag-
m: ts. An apρaηntht common variant is provided with a dense admixture of
broken terracotm fragments, probably to be identified with the opus testa-
certra or o3us testacic pavements mentioned by B:alíl in relatior_ to the Ro-
man city (Fig. 24). The term is also used by Puig y Cadafalch to describe a
pavement in ΑιπρΝ&755 This variant is often combined with inset pieces or
an emb'emr-a. In AmpR1 :t occurs in rooms 24, 34, the porticoes of the peri-
' Guimi.r-Sorbs 199). The Pompeian xenia comes from House VI.12, 2 (Ca del Fauno).
_i.
See
a1 FPM 1994, 105, 27.
Labirinto)
Βi:d stralirg j veller -i House VI.11, 8-10 (Casa del (PPM 1994, 49, fig. 78). Theat-
ri1 mask: House VIl 16, 1 ~~(PPM 1997, 908-9)9, figs. 40-41).
' - ΑΡL~nag:ο Βasc- 19cí211.
i4' BYIíl 1172
2 b, U~ .
5" Púg yiCadafa_ch 1915-1920, 701.
140
Fig. 25. Ampurias_ Nea_no lis, AmρΝ22, room 2, mortar pavement with tessera design and inscription
(author's phnto).
style and of the cryptoporticus. Moreover, these pavements are also present
in `Casa 2' and rooms 13 and 14 of AmρR4. 551
Mortar pavements with tessera design: Α number of patterns and compo-
sitions is employed:
1) Tesserae arranged in equally spaced rows.
Equally spaced rows running parallel to the walls of the room.
Equally spaced rows laid obliquely to the walls of the room.
The pattern of rows parallel to walls is used for the outer field in room 2
of Amp2 (Fig. 25). During on-site research a small part of rows parallel to
the walls was found in room 2 of Amp ΝΙ4. Α verification of the presence of
the same design in room 3 was not possible 552 The pattern of rows laid diag-
onally decorated the outer field of room 3 of Αmp127."
2) Lo2enge pattern
This is by far the most common pattern. It is predominant as the main de-
sign of th= floor in: Àmp Ν2 (room 1), AmρΝ7 (rooms 2, 3), ΑmpΝ17 (room
2), Amphl22 (room 2), Amp Ν25 (room 2), AmρΝ27 (room 3), AmpR1
(rooms 36, 38), ΑmρR3 (rooms 15, 17, 19, 26) and ΑmpR4 (room 14, eastern
half). In AmρΝΙ3 and Amρ 3 (room 8, and opening between rooms 31• and
32) only a small section of the pattern could be observed. Four main variants
of arrangement occur:
Covering the entire floor surface: Amp Ν7 (room 3) and AmρΝΙ7
(room 2). Probably also in AmρΝ2 and AmρΝ25. For the latter a bor-
der composed of two wavy lines encloses the lozenge pattern s54
Decorating the central rectangular field or panel only: Am ρΝ7 (room 2),
Amρ122 (room 2), AmpR1 (room 38) and AmρR3 (room 17) (Fig. 26).
_c) Decorating a circle forming a star-like composition: Am ρ127 (room
3), AmpR1 (room 36) and AmρΕ3 (room 26). In room 3 of Amp Ν27
there are two large juxtaposed circles of this kind, each is set within a
square enclosed by a meander 555
d) Doormat or entrance panel. In room 15 in ΑmpR3 two small rectangular
panels placed next to each other and with identical lozenge pattern (the
grid made by lines of white and black tesserae) are located immediately
inside the room between the doorway and the rest of the room (Fig. 27).
Rosettes
Judging from the plan of room 6 in AmpNS, there is a circle in the centre
decorated with a rosette motif. A 12-petalled(?) rosette decorates the implu-
vium in AmρΝ31. Small rosettes or flowers form the secondary decoration of
room 26 in AmpR3 (between the threshold and the rest of the room).
Meander
The typical meander is composed of alternating swastikas and squares and
serves as a border. Each square may have a single tessera placed in the cen-
tre. The pavement of room 21 in AmpR3 is described as having swastikas
and squares 556
Crosslets
Small crosslets composed of five tesserae (a white as the centre and the rest
being black, or vice versa) are used as secondary decoration or fill, for example
placed in the centre of each rhombus of the lozenge pattern (AmpR3: doormat
in room 15, opening between rooms 31 and 32), central field and or outer bor-
der (for example AmpR1: room 38), or band between the central field and the
meander border (for instance AmpR3: rooms 17, 26).
Others
The tessera design of room 15 in AmpR3 is special. The decorated field,
covering almost the entire room, is enclosed by a frame in opus tessellatum
(black-and-white) (Fig. 27). The area inside the frame is divided into squares
by lines of black and white tesserae, the north-south lines being black and
the east-west lines white. A crosslet occupies the centre of each square.
554
Mar/Ruiz de Arbulo 1993, 372. This can also be observed on the plan.
sss
For photo of the pavement and its design, see Marcet/Sanmartí, E. 1990, 109.
„6 Carri~n Masgrau/Santos Retolaza 1995, 117.
142
Fig. 26. Ampmias, Roman city, AmpR3, room 17, mortar pavement with tessera design (author's phom).
143
Fig. 27. Ampuurias, Roman city, ΑmpR3, room 15, mortar pavement with tessera deign and opus tesse/ta-
turn (authcr's ρhoto;.
144
7) Unknown design
The pavement of room 7 in AmpR2 is decorated with two different motifs,
thus marking a subdivision of the room. The motifs are, however, not descri-
bed.557 Whether the impluvium of ΑmρΝ5 and some rooms situated north of
the peristyle in AmpR1 are provided with tessera design cannot be an-
swered 558
Mortar pavements with inset pieces: Marble fragments densely set in a white
mortar are often used for impluvia. This applies to ΑmρΝ5, AmρΝ13 and
AmpR1. Because of the densely set pieces the impression may be that of an
opus sectile pavement rather than of a mortar pavement. 559 Moreover, the tabli-
num (room 37) in AmpR1 and two rooms belonging to the bath complex in
AmρR3 have this type of pavement (rooms 30,3 ι).5 To the left of the implu-
vium in AmρΝ7, a section of a mortar pavement with a dense admixture of
large terracotta fragments has been observed. Inset pieces of stone form two
lines that are more or less parallel (Fig. 24). In room 12 of Am ρR4 the frag-
ments are evenly set and form a pattern of interrupted lines Ss'
Inscriptions: Mortar pavements with inscriptions occur in three houses sit-
uated in Neapolis. The inscriptions are in Greek and set in white marble tes-
serae. All inscriptions are in situ. Several scholars have dealt with them, most
recently Gómez Pallarès in 1997. 362 Regarding the dating of the inscriptions,
there is consensus of the 1st century BC. 563
Key to symbols: [ Ι = restitution of the text
+ = the letter is illegible
564
Gómez Pallarès·1997, 93-95. For photo, see Almagro 1952, 38.
567 Room 7: For photo, see Almagro 1952, 39. Gómez Pallarès 1997, 95-96.
566
Gómez Pallarès 1997, 92-93.
567
For example Mar/Ruiz de Arbulo 1993, 359.
568 Olmos Romera1989, 46-52.
569
Aim agro 1-952, 38; Santiago 1993, 287.
57° Gómez Pallarès 1997, 90-92.
571 Almagro 1952, 37; Olmos Rimera 1989, 57.
572
Durand 1981, 277.
" Gomes Pallarès 1997, 91.
574 Schröder 1996.
575
For photo of the statue, see LIMC 1I.2, 647, no. 153.
576
Schröder 1996, 232.
146
The houses in the Roman city were lavishly decorated with murals resem-
bling the First, Second, Third and Fourth Styles. As far as the houses in Nea-
polis are concernëd, no conclusions can be reached due to the extremely
sparse remains. The mouldings from Amp Ν17 indicate the existence of more
complex wall decoration.
It is interesting to notice the widespread use of mortar pavements with
tessera design in Neapolis. They were preferred to figurative tessellated mo-
saics, the tradition of the eastern Mediterrranean area. Moreover, the same
geometric patterns are employed in Neapolis and the Roman city. The only
motif in opus tesse ilatum known from Neapolis is the lozenge pattern (not lo-
cated). The pavements in the houses of the Roman city closely follow the
57 Santos Retolaza 1991, 25. Only the pavement in room 38 is visible today.
578 Barn 1972b, 95 and 99-100.
579
Santos Retolaza 1991, 27.
580 Carrion Masgrau/Santos Retolaza 1995, 21.
147
trends of the homeland. 581 First the houses were provided with typical mor-
tar pavements, assuming that they date to the first phase as well, then they
made use of black-and-white opus tessellatum mosaics, emblemata in opus
vermiculatum and opus sectile. The earliest securely dated mortar pavements
with designs are those from the houses in the Roman city, i.e. around 100
BC/first quarter of the 1st century BC. Most likely they were immediately
imitated in the houses in Neapolis.
FINDS
SCULPTURE
The finds came from AmpR1 and include two herms with drapery (heads
not preserved) and a female head in marble. The portrait was found in one
of the rooms adjoining the large room 53• on the south side of the peri-
style. 58z Almagro dates it to early Imperial time on the basis of the hairstyle, a
nodus coiffure (variant of Octavia). Since this part of the house dates to the
end of the 1st century AD, this was not the original place of display. The
portrait shows clear signs of advanced age such as hollow cheeks, deep naso-
labial folds and a thin lipped mouth, and has resemblance to late Republican
matrona portraits. Fittschen and Zanker have categorized the portrait as a
provincial or an uncertain replica of Livias Marburg Hall type, whereas
Winkes advocates a certain ascription to the Marburg Hall type. 583 The por-
trait is the only one, if Livia, with such veristic facial features. According to
Winkes, this is due to a deliberate local concept. This is, however, always an
(too) easy way of explaining divergent elements.
RELIGIOUS FINDS
581 For instance House 11.11, 8-10 (Casa del Labirinto) and House 111.2, 45 (Casa de ll'Orso) in
EPIGRAPHICAL MAIΣRIAL
585
Almagro 1952, 160, no. 112; Fabre et al. 1991, 114-115, no. 90.
588
Fabre et al. 1991, 114-115.
587
Fabre et al. 1991, 115, note 295.
568 Santos Retolaza 1991, 32-34. For the development in Pompeii and Herculaneum, see for ex-
ample Dickmann 1997; George 1998 (also including the display of statuary).
589 Gusí/Olaria 1984, 19-20; Maluquer de Motes et al 1986, 15-17.
149
590 For description and photo, see Puig y Cadafalch 1908, 191, fig. 40.
150
The rooms in the northern zone are arranged around two nuclei: a courtyard
with three porticoes (room 12) and probably a second atrium (room 18).
Characteristic pavements here are black-and-white opus tessellatum mosaics
and opus sectile. A huge peristyle garden (room 46) dominates the southern
zone. Reception rooms of varying sizes are located on its northern and south-
ern side. The importance of this zone is evidenced by the opus tessellatum
and opus vermiculatum mosaics employed here. Three emblemata of opus
vermiculatum are documented; they depict a theatrical mask, a bird stealing
jewellery and a xenia. Besides, a female portrait (perhaps Livia) was dis-
played in one of the rooms adjoining the huge room 53·. The walls were dec-
orated with murals resembling the First and Third Style, for instance rooms
15, 26 and 47.
AmpR3 had an area of 2128 m z in the last phase and is made up of four
main zones (Fig. 13). Luxurious elements in the house are no less than three
garden areas, numerous reception rooms and a bath complex. The atrium
zone (Tuscan and the oldest part) is located to the west, and access to a peri-
style garden with a huge reception room (room 40) is achieved by a wide
passage made at the rear of the former tablinum (room 52•). This creates an
impressive line of sight from the entrance through the atrium and peristyle
garden to the reception room. The third zone is formed by a second peristyle
garden (room 35a•) with adjoining reception rooms and the bath complex
(rooms 30, 31, 32). Towards the east the house is provided with additional
rooms, including the largest room (room 46) of the house, and a huge garden
area with porticoes. The reception rooms are characterized by wide openings
towards the gardens. Interior architectural decoration includes murals of dif-
ferent decorative systems (close to the First, Third and Fourth Styles) and a
great variety of pavement types (opus tessellatum, opus tessellatum with inset
pieces (impluvium), opus sectile (reception room 35), mortar, mortar with
tessera design (rooms 15, 17, 19, 21, 26, and opening between rooms 31 and
32), mortar with inset pieces). An altar decorated with a cock and snakes was
found in the first peristyle garden (room 40a•).
Because of the Greek inscriptions, the inhabitants of Am ρΝ5, AmρN22
and AmρN27 should be expected to be Greeks or Hellenized Iberians, but
that only Greeks and Iberians lived in Neapolis, as assumed by Marcet and
Sanmartí, is unlíkely. 594 It cannot be ruled out that Italians/Romans bought a
pre-existing house in Neapolis or had one built, for instance Am ρΝ34. Fur-
thermore, Punic people may have settled here, though conclusive evidence is
lacking. Likewise, it would be wrong to assume that only Iberians lived in
the modest houses consisting of three rooms. Greeks, Iberians, Italians/Ro-
mans and others living in Neapolis may have imitated the atrium houses in
the Roman city. In fact, several of the inhabitants of Neapolis were living `à
la romaine', judging from the atria, cisterns covered with vaults and mortar
pavements with tessera design, but the language used in the inscriptions of
the mortar pavements was still Greek.
That Neapolis displays this mixture and non-Greek appearance is not sur-
prising. The interaction between Greeks and Iberians took place from the
beginning of the cítys history. In the 2nd century BC the Italic/Roman im-
pact began to be felt, and from the late 2nd century onwards there was a
massive Roman influence. The Italians/Romans not only introduced their
own traditions such as the atrium house, but also the traditions of the Hel-
lenistic :East, however modified by the Romans (for instance the axial peri-
style house), and Punic elements such as the building technique of opus afri-
canum.
In the 2nd to the 1st centuries BC different cultural influences are also
documented in the public sphere of Ampurias.
The Greek origin of the city is reflected in the new agora and stow con-
structed in the centre of Neapolis towards the middle of the 2nd century
BC 595 Domestic structures were dismantled to make way for the new com-
plex, but later on parts of the complex were privatized again. Two bathtub-
shaped double cisterns covered by vaults are located inside the stoa. A third
cistern with a flat cover is situated below the agora and thus antedates it. The
second major building activity took place in the Asklepieion whose area was
enlarged, raised and provided with a temenos wall to the east in the third
quarter of the 2nd century BC. 596 Also the public baths built around 100 BC
in the north-east area of the city had Greek elements such as a palaestra and
sweat-bath. 597
Simultaneously with the Greek manifestation in Neapolis the Romans in-
stalled a military camp on the hill west of Neapolis, represented by the so-
called praesidium building with a complex of four large bathtub-shaped cis-
terns placed next to each other. Their length is 11.5 m and width 1.3 m.
Their cover was most likely flat. They can be dated around 175 BC and are
the earliest securely dated examples of the Punic bathtub-shaped cistern.
However, there is no consensus on the interpretation of the earliest remains
from the area. Pena has suggested that Iberians lived here, resettled under
Roman control 5 98 This would have taken place in relation to the general or-
ganization of the territory by the Romans which resulted in the abandon-
ment of several local settlements from the late 3rd century BC onwards (for
instance Ullastret) 599 Pena has also stressed some peculiarities of the layout
of the later Roman city such as a dismantled tower located immediately out-
side the present southern gate and the slightly different alignments of the city
"s
Mar/Ruiz de Arbulo 1988.
596
Sanmartí i Grego, E. et al. 1990, 136-139 and 144. There is no consensus on the dating and
number of phases of the sanctuary. The second phase is also dated to the first half of the 2nd century
BC (Mar/Ruiz de Arbulo 1993, 173-179).
"' Ρalauf/Ιν~~1993x, 109-110. The alteration in Augustan time removed the Greek elements.
5" Pena 1989.
5"
For the subdivision of the territory of Ampurias, see Plana Mallart 1989; Plana Mallart 1994,
chapter Π.
153
wall and the street grid. 600 Furthermore, several silos, characteristic for Iber-
ian settlements, are located in the area around the building. This argument,
however, is not convincing, since the Romans may have adopted the local
custom. The Roman influence is also reflected by the importation of Italic
products in large quantities and by the building constructed in opus africa-
num south of Neapolis. This building was constructed directly on top of a
Greek necropolis. Moreover, the bronze coinage introduced at the beginning
of the 2nd century BC followed the Roman weiht standard, but it had Iber-
ían monetary magistrates and an Iberian legend. 01 Pena's interpretation is ar-
guably supported by the Las Corts necropolis containing only cremation
graves. It dates to the period from c. 200 to the late 1st century BC or early
1st century AD. G02 According to Almagro, Greeks were buried here. He con-
siders the cemetery too Hellenized and too monumental to be Iberian. In
view of the fact that Greeks and the indigenous population had been living
next to each other and together for a long time, this is only a natural conse-
quence. Nonetheless, it is more likely that the necropolis was Iber ~an than
Greek due to the cremation rite or it may have been used by both. If Iberians
were resettled here in the 2nd century BC, they might have been living in the
area that became the northern part of the later Roman city. Unfortunately,
this area has not been investigated. G03 In summary, both a Roman praesidium
and a settlement inhabited by Iberians may have existed in the area of the
later Roman city.
Iberian personal names inscribed on the bronze coinage of the 2nd cen-
tury BC show that Iberians held offices as monetary magistrates. 604 Further-
more, the coinage is inscribed with the Iberian UNTIKESKEN (the name of
the indigenous people in this particular area) and UNTICA which could be
the Iberian name for the Greek Emporion. 605 On the same bronze coinage of
the 2nd century BC Latin names in Iberian script occur: Tiberi and Luci.
These forms may be the Iber~an version of the Latin names or a dialect from
southern Italy. G06 In the case of the latter, Iberians apparently adopted the
Latin names as written in dialect because some of the Italians who settled at
Ampurias came from southern Italy.
In the 1st century BC the Roman impact is strong. The forum and the tem-
ple in the Roman city were built after late Republican models from Central and
South Italy, and in its second phase the temple here imitated the Temple of Di-
vine Julius in the Forum Romanum. 607 The sanctuaries situated in Neapolis also
followed models of Central and South Italy (inspired by the Hellenistic world).
600 Cf. Barberà/Μοrral 1982. Soundings made in relation to the "muralla Rubert" and "muralla
Roura" have revealed material dating no earlier than the late 2nd and early 1st century BC (Sanmarti-
Grego, E./Santos Retolaza 1986-1989, 304-306).
6°1 ViHaronga 1977.
602 Jones 1984, 238 and 240-241.
603 Ruiz de Arbulo Bayona 1991, 477-478; Aqui1ué 1997, 47-48.
604 Villarοnga 1977, 12; Pena 1988, 21.
224-228.
154
This applies for instance to the layout of the Asklepíeíon. 608 The bathtub-
shaped cistern and the mortar pavement decorated with a lozenge pattern
known from the houses are also found in this sanctuary. 6Ó9 The Egyptian divin-
ities Sarapis and Isis were worshipped in the Sarapíeion built around the mid-
dle of the century and financed by the Alexandrian louras.
The great building activity that took place in Neapolis in the 2nd century
BC is traditionally seen as a sign of the prosperous situation of the trading
centre. Recently Kaiser has offered a new and more convincing interpreta-
tion: the building projects were an attempt by the Greeks to create a strong
group identity because they felt under pressure both as a result of the Ro-
mans favouring the Iberians and their own loss of political and economic sta-
tus (Iberians as monetary magistrates and with citizenship before the Greeks
etc.). 610 Kaiser stresses the non-Greek character of the city prior to the 2nd
century BC: there was no theatre, only a small agora and no orthogonal street
grid. With the new projects the Greeks tried to make the city look like Mar-
seilles (Massalia) or a Greek city in Magna Graecia.
This interpretation is particularly interesting in relation to the houses of
Neapolis. As already attested, they display a lack of Greek elements. Howev-
er, there is the problem of chronology, since several houses in their last
phase, i.e. after the establishment of the Roman city and the political unifica-
tion of Neapolis and the Roman city in early Augustan time, were influenced
by Roman architecture and pavement types. Thus the houses of the 2nd cen-
tury may to a large extent be unknown to us. Nonetheless, following the in-
terpretation of Kaiser, one may conclude that the Greeks manifested them-
selves via traditional Greek public architecture and not in their homes. One
the other hand, it is also likely that they tried to be as Greek as possible in
the private sphere. This could have been done by ways not traceable in the
archaeological record, for instance clothing.
(third quarter of the 1st century BC). The mortar pavement with tessera design is employed in the cel-
la of Temple P.
610 Kaiser 2000. He follows the phases and datings suggested by Mar and Ruiz de Arbulo (1993)
which means that he places the second phase of the Asklepieion in the first half of the 2nd century BC.
155
ι Ι ι ι ι ι
ι ι ι ι ι Ι
I `
-----_-- ~__-
Γ'
ι
G F σ
~~ Α
ι
ι
ι
Ι
L ~~ Σ? ο
α
ε
5 10
Fig. 28. Rosas, House Α-II-3, plan (drawing by Kjeld de Fine Licht).
ber of Greek colonies, but only Ampurias and Rosas have been identified in
the archaeological record. 611
Rosas (ancient Rhode, Roses in Catalan) is situated on the northern side of
the Gulf of Rosas and was probably founded by iassíliotes, but literary
sources also refer to Rhodes as founder (Fig. 1). 612 The earliest material dates
to the 5th century BC. Remains of residential character go back to the begin-
ning of the 4th century BC, and in the 3rd century BC a new habitation area
was laid out on an orthogonal street grid, the so-called "Hellenistic Quar-
ter".613 The remains are sparse, but apparently the plots were of equal size, at
least at the beginning. Another characteristic element is the combination of
habitation and workshop in one and the same building. The following four
houses are the best preserved. House A-II-3 is, according to li d, a pastas
house with tendency to symmetry in the ground-plan (Fig. 28). The rooms
are arranged around the courtyard (room e) like a i. A portico may have ex-
isted on the northern side of the courtyard. Thus this northern portico or
part of the courtyard together with room b would have formed the pastas.
Vivd suggests that room d was the main room because of its size, and room a
61 For a summary of the Greek colonies, see Richardson 1996, 14-15 and especially note 17.
612
For the textual evidence, see Vivó 1996, 82.
613
For plan, see Vivó 1996, 84, fig. 4.
156
was the dining room (andron). In House A-II-1 the courtyard has been iden-
tified in room a and the main room in room c. 614 Inside room c there is a
smaller room, and this arrangement resembles the so-called "oecus unit" in
Olynthian houses. House B-II-1 is square and apparently divided into four
rooms of which the largest, room b, is identified as the main room. G15 The
workshop area (f) attached to the house is located to the west. House B-II-2
probably comprised a residential unit, a pottery and perhaps a shop where
the products were sold. 616
With its orthogonal grid, standardization in house-plot size, a possible
pastas and oecus unit the quarter is quite different from Neapolis and has af-
finities to the Greek world, for example the pastas houses of Olynthos and
Attica. Rosas was abandoned in the 2nd century BC, and thus it would be
more appropriate to compare it with the Ampurias of the 3rd century BC,
and not the city in its existing state, but the domestic architecture of Neapo-
lis prior to the 2nd century BC is unknown. When compared with Rosas the
Iberian impact on Ampurias is even more evident.
Ullastret is one of the best documented of the Iberian settlements. G17 It is sit-
uated on the hill of Puig de Sant Andreu c. 14 km from Ampurias and was
abandoned around the middle of the 2nd century BC (Fig. 1). The urbanism
of the settlement, including houses with rectangular ground plans, stone socles
and mudbrick walls of the mid-6th century BC and the city walls, is ascribed to
the Greek influence in the area. 618 In general, the houses are described as hav-
ing one or two rooms. The presence of three bathtub-shaped cisterns should
be noted. One is dated to the 3rd century BC and thus appears to antedate the
cisterns from Ampurias (earliest securely dated: around 175 BC). G19 However,
several of the cisterns in Neapolis cannot be dated precisely.
The Punic cistern is also documented from other local settlements. Two
have been found at Castell de la Fosca de Palam ~s (Fig. 1). 620 At Ensérune in
southern France both bathtub-shaped and rectangular cisterns occur. The
latter have rounded corners. They date to the 2nd and 1st centuries BC.
Apart from one cistern, the cover is flat and may be supported by pillars. 621
In the case of Roman colonies and villas, atria, the First Style, opus tessel-
lαtum mosaics and mortar pavements with tessera design are standard ele-
ments. A good example is the villa of Can Marti at Samalús built between the
end of the. 2nd and the first half of the 1st century BC (Fig. 1).622 A mono-
chrome white opus tessellatum mosaic (originally with an emblema) and a
614
Vivó 1996, 87, fig. 5 and 105, fig. 17.
615
Vivó 1996, 93, fig. 10 and 106, fig. 18.
616
Vivó 1996, 95, fig. 11 and 107, fig. 19.
617 Martin i Ortega 1997.
618
For example Padr ~/Sanmartí, J. 1987, 24-25.
619
Martin i Ortega 1997, 22-23. The dimensions are: length 5.3 m, width 1.25 m, depth 3.7 (Oliva
Prat 1960, 346-349).
620 Pericot/Oliva Prat 1949, 320-321.
621 Jannoray 1955, 108-109, 170-185 and 465-466; pls. XIII, XXVII.2 and =01.2.
622
Aquilué í Abadías/Pardo í Rodriguez 1990.
157
played. Two of the .houses from the Roman city of Ampurias developed into
villa-like structures (AmpR1, Am ρR3) with areas of 3492 m 2 and 2128 m2 re-
spectively. Moreover, AmpR1, ΑmpR2 and AmpR3 were lavishly decorated
with murals and various pavement types. In ieapolis ΑmρΝ5 had a peristyle
built at the margin of the plot, and the large axial peristyle house ΑmρΝ34
had an area of more than 868 m2 and was provided with a fountain in the
peristyle.
An element documented in all three cities is the room opening directly to
the street (not anentrance room). In De1T23 room c has two openings to-
wards the street and one internal doorway, and the pavement design leaves
no doubt that it functioned as a dining room. At Αmρurias room 2 of
AmρΝΙΙ is provided with one opening towards the street and one internal.
The mortar pavement with tessera design and wall decoration suggest that it
was the main room of the house. The main room (room I) of CarH8 may
have had an opening directly to the street as well. The same phenomenon
can also be observed in the house of insula VI at Olbia-de-Provence. In
phase lb, dated around AD 10, room 2 was converted from a shop into a re-
ception room with one opening towards the street. 633 Thus an opening di-
rectly to the street is not necessarily evidence for a shop or workshop. All as-
pects of the room must be taken into consideration.
Another shared feature is the presence of foreign divinities, especially
Egyptian gods. Some were recognized at an official level (for example the
cult of Demeter and Kore at Carthage), while others were private (for instan-
ces the Semitic sanctuaries on Delos).
Since Ampurias was a Greek colony, one would expect to find some char-
acteristic elements from Greek domestic architecture, at least a pastas or a
prostas, which, irrespective of the terminology, represent a characteristic way
of organizing space in the Greek world. This is not the case, but the eager-
ness to apply the traditional terminology to the Ampurian material is demon-
strated by Bahl who identifies the main room of ΑmρΝ7 as a prostas, and
Zaccaria Ruggiu who speaks of Greek peristyle houses in Νeaρolis.634 There
is no open-fronted anteroom in Am ρΝΙ (atrium house), and there are only
two houses with a peristyle court as central circulation space, AmρΝ21 and
ΑmρΝ34, of which the latter is a Roman axial peristyle house. According to
Bali], the tetrastyle atrium in AmρΝ31 is a tetrastyle peristyle because the in-
ternal layout does not share characteristics with Roman atrium houses. 635
Delos and Ampurias overlap in the 2nd to 1st centuries BC, and both
came under Roman control in the first half of the 2nd century BC, so it is
natural to look for similarities and differences. As regards ground plans, a
tendency towards trapezoidal shapes occurs in both sites. The central peri-
b3 Bouet 1997, 444-445 and 445, fig. 1. As a shop it had two doorways.
634
Bali11972a, 43; Zaccaria Ruggiu 1995, 380.
6i5 Bahl 1972b, 44-45. The northern and eastern parts of the house are situated below the modern
pathway.
161
636 For families with widespread activities, see D'Arms 1980, 85; Wíkander 1985; Rauh 1993, 218.
Service rooms commonly found on Delos, especially latrines, are rare in the
houses of Neapohs.
The use of polychrome figured mosaics, characteristic of the Hellenistic
world and in particular on Delos, is only documented in the Roman city of
Ampurias. The situation is reversed when it comes to mortar pavements with
tessera design. They are rare in Delian houses, but common in the houses of
Ampurias, both in the Roman city and Neapolis.
Unfortunately, Only AmpRi has provided sculptural finds (one female
portrait and two herms). Thus the occurrence of iconographic themes and
the display of statuary cannot be compared.
643
Dd/D 1962, 80-85.
Dunand/Duru 1962, 234.
645 Grainger 1991, 109.
696 Herbert 1994, 14.
64,
Herbert 1994, 62-72; pls. 38 and 40.
as
Millar 1983, 58.
649 Arr. anab. 2.24.5; Ροlyb. 31.12.11-12.
650
Liv. 33.48.3-8. Grainger 1991, 201-205.
651
Grainger 1991, 214.
165
Another early candidate is the cistern located below the pronaos of Temple P in the Askle-
píeíon. It is bathtub-shaped and with a flat cover. A dating before the 2nd century BC cannot be ruled
out (cf. Burés Vílaseca 1998, 281).
6" Lézine 1968, 101, note 3.
654
For Phoenician imports along the coast of Catalonia and Languedoc in the last quarter of the
7th to the first quarter of the 6th century BC, see Arteaga et al. 1986.
6" For example Gómez Bellard 1992, 385-386. According to Diodorus, Carthage founded its col-
ony on Ibiza in 654 BC (Dud. 5.16).
636
Villaronga 1977, 2.
657 García Bellido 1994.
658
Oliver Fox 1995; Sanmartí-Grego, E. et al. 1995.
166
first half of the 2nd century BC. Thus it is also possible that the cistern was
introduced, or more plausibly revived, in relation to the increased commer-
cial activities of Ibiza in this period. 659 Although the Carthaginians were de-
feated by the Romans in 206 BC, Punic people continued to live on the'Iber-
ían Peninsula and the Balearic Islands. This is evidenced by Phoenician and
Punic inscriptions of which more than half belong to the 4th to 1st centuries
BC.660 An indicator of the close trade connections between the Punic me-
tropolis and Ampurias is that outside the Iberian Peninsula the largest con-
text of Ampuritan grey ware has been found at Carthage. 661
As regards other water installations at Ampurias, three houses reused Pu-
nic amphorae as overflow drains for cisterns ( ΑmρΝ16, AmpΝ21,
Amp129).
The difference between Delos and Ampurias is not surprising. Late Hel-
lenistic housing as represented by Delos was in fashion, i.e. imitation of pala-
ces and palatial houses of the East and Macedonia. The housing of Neapolis
in Ampurias was a veritable mix of several cultural traditions owing to its
long history and location on Iberian ground. Some of the Punic elements
were the result of a direct contact with Punic merchants (and inhabitants?)
(bathtub-shaped cistern with flat cover), while others were transmitted
through the Romans (opus africanum, mortar pavements). Elements from the
eastern Hellenistic world were likewise transmitted through the Romans
from c. 100 BC onwards (peristyle gardens, emblemata). Here too the Italic/
Roman influence eventually became the most visible.
`HOUSE TYPES' AND THE MUTUAL INFLUENCES BETWEEN THE HELLENISTIC, PUNIC
AND ROMAN WORLDS
659
The Punic Wars did not effect the activities of the island, and these increased after the Roman
conquest (Gomez Bellard 1992,387-389).
660
Koch 1976.
662
Morel 1990, 88-91.
662
Nevett 1995; Nevett 1999.
167
es with two courtyards. Nevett concludes that despite variations, also within
the groups, the Greek house of the 4th and 3rd centuries BC is characterized
by the segmentation of space in order to regulate social relationships within
the domestic environment. This segmentation is most evident in the houses
with two courtyards. Nevett's reassessment is timely and reasonable because
it replaces the rigid and dubious typologies. Since there is no clear archaeo-
logical distinction between the pastas and prostas, and there is no consensus
on the use of terminology, the pastas/prostas should be considered a flexible
room as regards both length and width, as well as the degree of closeness and
openness. The criterion for its identification is the location within the inter-
nal layout, i.e. between the circulation space and the main room(s). Houses
in southern France dated to Augustan time testify to the persistence of this
way of organizing space.
The importance of the house as a status symbol is already attested in the
4th century BC in the Greek world and was a result of a self-centred trend
replacing the values of the Classical polis. This trend towards private ostenta-
tion escalated throughout the Hellenistic period with the elite and wealthy
imitating the royal architecture of the Hellenistic East in their homes. The
culmination of this development is represented by Delian and Pergamene
houses with their large peristyles, many reception rooms and demand for
decorative items. On Delos the main room used for reception and dining in
these large houses is the Oriental broad-room, while a suite of three rooms
similar in size and shape is typical at Pergamon. Often these rooms intercon-
nect, and the central room may be open-fronted with columns in antis, for
instance Bau Z. This room complex clearly imitates the palatial versions from
Macedonia. ó63 These late Hellenistic houses also testify to influence from the
West owing to the Roman involvement and presence in the East, especially
after the creation of the province of Asia in 133 BC. The Roman features are
for example parapets between the columns (for instance De112, De1122 and
Komplex VII) and various pavement types: mortar pavements with tessera
design (for example the House of Attalos and DelS2) and opus sectile of
stone tiles depicting perspective cubes with a surround of opus tessellatum
(the House of Attalos). On Delos there is a concentration of tessellated che-
quer-boards decorating larger fields or entire floor surfaces in the Northern
Quarter. This is no coincidence, since this quarter in particular displays Ita-
lic/Roman features, and also this motif can be ascribed to the Italic/Roman
influence. Unlike motifs such as the lozenge pattern and meanders, the che-
quer-board does not occur in Greek pebble mosaics probably because it de-
pends on regular tesserae. G64 In contrast, the motif used for both frames and
larger fields is widely documented on the Italian Peninsula (especially Pom-
peii), in Sicily and North Africa. Examples from the Italian Peninsula are for
instance: House 111.1, 40 at Pompeii (Casa di M. Caesius Blandus, late Re-
publican date, threshold mosaic in bath), Villa Prato at Sperlonga (third
quarter of the 2nd century BC, threshold mosaic in bath and also decorating
663
For instance the southern wing of the palace at Verging (Nielsen 1994, 82, fig. 41).
664
Tsakirgis concludes that the pattern must be late for the same reason (1989, 411).
168
the borders of the bathtub), a now disappeared villa at Ciampino (the bor-
ders of the bathtub) and the maison du Cryptoportique at Vulci (last quarter
of the 2nd century BC, threshold mosaic in bath). 665
665 Lafon 1991. For the Pompeian house, see PPM 1996, 380, 435, figs. 106-107 and 437, fig. 110.
666 Zanker 1995, especially 20-43; Dickmann 1997; Dickm an n 1999, especially 125-158; George
1998. The origin of the atrium house will not be discussed in the present work.
66, Nappo 1997; Wallace-Hadrill 1997.
668 For example Wallace-Hadrill 1994, 63-90. See also Wallace-Hadríll 1997, 219, note 2 for
references.
669
For canonical atrium houses, see for example Pompeii, Regio VI, insulae 13 and 14 (Wallace-
Hadríll 1994, 211, fig. Α 10).
670
De Kind 1998, 189-193. His classification of the houses from insula III and IV at Herculaneum
is based on the way in which the available space was organized and not on the atrium.
671 Coarellí 1995, 18; Coarelli in Fregellae 1998, 64-65.
699
Rakob 1989, 184.
693
For example Selinunte, first half of the 3rd century BC (Dunbabin 1994, 32); Morgantina, first
half of the 3rd century BC (Tsakírgis 1990, 441); Pompeii, 3rd century BC (Pesando 1997, 17); Fre-
gellae, middle of the 3rd century BC (Coarellí 1995, 19); Rome, late 3rd century BC (Papi 1995,
341-342).
169
674 Motifs such as the lozenge pattern and ctosslets are documented in the 'Mago Quarter' in the
16-17 and 33-35; House VIII. 2, 14-16: room e'and eastern part of portico f (PPM 1997, 958, fig. 19;
961-963, figs. 24-25; 999, fig. 94; 1000, fig. 98 and 1002, fig. 102; PPM 1998, 84-85, figs. 19 and 21).
678 For example House I.6, 4, tab//rim (f) and fauces (g) (PPM 1990, 306-307, figs. 44-45).
679
For example House 1X.2, 22.29 (PPM 1998,971, fig. 30).
680
For example HouseΝI.9, 2.13, room 28; House I.6, 4 (PPM 1993, 789, fig. 245; 1990, 306,
fig. 45).
661 For example House I.13, 2, ala (PPM 1990, 865, fig. 6).
682
For example House I.16, 4, atrium (PPM 1990, 999, fig. 1).
683
Rome and its environs: Morricone Matini 1967; Morricone Matini 1971; Papi 1995. Fregellae:
Coarelli 1995. Northern Italy: Conti 1996.
684 Coarelli 1995.
170
the first half of the 3rd century BC. The emblema is combined with a mortar
pavement. The impluvium of House 7 was provided with lozenge-shaped
tiles in terracotta, and other rooms had mortar pavements in its first phase. A
mortar pavement decorated with equally spaced rows parallel to the walls of
the room was laid in the tablinum in the middle of the 3rd century BC. All
houses underwent alterations in the second phase of the city, around 200 BC
or a little later. As far as pavements are concerned, there is documentation
for mortar pavements with various designs (equally spaced rows laid parallel
to the walls of the room, the lozenge pattern and meander), mortar pave-
ments decorated with fragments of coloured stone and white marble, and
monochrome white opus tessellatum mosaics.
The meeting and intermingling, of traditions from the Greek and Punic
worlds, and later on also the Italic/Roman world, is illustrated by the cities of
Sicily. Here Monte Iato and Morgantina will be commented on.
Eight houses dating to the Hellenistic period have been located at Monte
Iato (ancient Ietas) situated in north-west Sicily.8 5 While the general layout
and room complexes belong to the tradition of the Hellenistic East, the most
common pavements are mortar pavements with and without tessera designs.
At least four houses are provided with peristyles. Both Peristyle House 1 and
Peristyle House 2 had a two-storeyed peristyle and probably also House Si.
Peristyle House 1 built around 300 BC has a peristyle with 4 x 4 columns.
The suite of reception rooms is located to the north of the peristyle and finds
an exact counterpart in the palace at Verging dated to the second half of the
4th century BC: a central room with two columns in antis giving access to a
dining room on either side. The mortar pavement of the bathroom is decor-
ated with equally spaced rows laid parallel to the walls of the room. In Peri-
style House 2 there is documentation for mortar pavements decorated with
the lozenge pattern (porticoes and upper storey).
The houses from Morgantina situated in central Sicily cover the period
from the 3rd century BC to the mid-1st century AD. They display a similar
pattern, i.e. general layouts and architectural elements following models of
the Greek world and a predomination of mortar pavements with tessera de-
signs. In 211 BC the city came under Roman control and was given to Span-
ish mercenaries. In the archaeological record this new situation is documen-
ted by coinage inscribed HISPANORUM, but the houses do not testify to
the presence of people of Iberian origin. The majority of the houses are ar-
ranged around a courtyard, Reristyles are common, and there may be more
than one circulation space. 6 In the 2nd to 1st centuries BC some houses
were subdivided into two units, for instance the House of Ganymede and the
House of the Arched Cistern. Again three-room suites (central room, some-
times open-fronted, and one room on either side) are documented, for exam-
ple in the Southwest House (rooms 1, 2, 3).687 The Pappalardo House con-
structed in the 3rd century BC shoùld be singled out, since the ground plan
685
Fuchs 1997; Islet. 1997. See also Trumpet 1998, between 356 and 357, fig. 95; 357, fig. 96.
686
For plans, see Tsakirgis 1990.
687 For plan of the Southwest House, see Trümper 1998, 360, fig. 99.
171
(elongated rectangle) and the internal layout with off-centre peristyle (rooms
on three sides only), main room and entrance room located opposite each
other resemble houses on Delos, for instance De1N19 and De1M1. 688
The tessellated mosaics, some using irregular tesserae, are mainly decor-
ated with geometric motifs such as waves, meanders, cubes in perspective
and chequer-board (only frames). The earliest mosaics belong to the House
of Ganymede and the House of the Official and are dated to the 3rd century
BC.689 The combination of opus tessellatum (central panel) and mortar pave-
ment with tessera design occurs in the House of the 'Tuscan Capitals (room).
This house was rebuilt in the 2nd century BC. 690 The Roman impact is clearly
visible in the interior architectural decoration. The standard range of geo-
metric designs is employed in the mortar pavements: equally spaced rows,
lozenge patterns, meanders, crosslets and rosettes. 691 Different motifs are
used within the same room in order to differentiate between zones, for in-
stance the House of the Arched Cistern (rooms 9, 24 and 25) and the House
of the Tuscan Capitals (room 6). Moreover, there are traces of early Second-
Style murals in rooms 20 and 21 in the House of the Tuscan Capitals. 692
The most famous mosaic from Morgantína is beyond a doubt the Gany-
mede mosaic with an apparently secure dating to c. 260-250 BC.693 Based on
this dating the invention of the tessellated mosaic has been ascribed to Sicily.
Other places suggested are: the eastern Mediterranean (Greece and Asia Mi-
nor) and Alexandria. G94 But as Baldassarre and Dunbabin have clearly dem-
onstrated, the invention of the tessellated mosaic was not a single invention,
meaning that independent and different developments took place in the east-
ern and central Mediterranean area respectively. 695 This was the case with
pebble mosaics found both in the East (Asia Minor and Greece, 8th century
BC onwards) and on the Iberian Peninsula (7th to 4th centuries BC).6 6
Moreover, the evolutionary theory (first pebble, then irregular tesserae, and
finally regular tessellation) must be rejected, since different techniques were
used deliberately in one and the same mosaic in order to create specific ef-
fects. There can be no doubt that the tradition of the figured tessellated mo-
saic belongs to the East, but the examples from North Africa together with a
688
For plan of the Pappalardo House, see Tríimper 1998, 365, fig. 104.
689
Tsakirgis 1989, 408.
690 Tsakirgis 1989, 404, figs. 20-21. The pavement appears to be later than 150 BC.
group of finds from Sicily and southern Italy show that there was a inde-
pendent tradition in the central Mediterranean area. 697 The group includes
finds of tessellated mosaics from Gela (before or after 282 BC?), Herakleia
Minoa (terminus ante quern of c. 300 BC or terminus ante quern of the 2nd
century BC), Monte Iato (3rd century BC?), the baths at Megara Hyblaía
(terminus ante quern of 214/213 BC), Buccíno (ancient Volcei) (late 4th cen-
tury BC?) and Elea/Velia (first half of the 3rd century BC). 698 At Carthage
regular tessellation is documented as early as the second half of the 4th cen-
tury BC in the `Mago Quarter' (monochrome white) 699 To my knowledge
this is the earliest securely dated example of the technique. The mosaic frag-
ment from the Rue Didon/Rue Arnobe combining opus figlinum and a poly-
chrome chequer-board is dated to the 4th century BC by the excavator, and
a tessellated fragment (red and white) from Kerkouane has been dated to the
5th century BC. 70°
To sum up, judging from the evidence of North Africa, Sicily and the Ital-
ian Penisula, a Carthaginian-Sicilian-Roman connection can be traced with
respect to the development and spread of both mortar pavements with tes-
sera design and the tessellated mosaic in this area. At Carthage mortar pave-
ments with scattering of fragments and tessellation go back to the 4th cen-
tury BC. Moreover, irregular tesserae forming religious symbols and decora-
tive motifs are known from the city (the end of the 5th century BC) and Ker-
kouane (before 256 BC). Mortar pavements with tessera design (confined to
the pattern of equally spaced rows) covering whole floor surfaces existed in
the 3rd century in North Africa (again before 256 BC at Kerkouane), in Sici-
ly (for instance at Selinunte, first half of the 3rd century BC) and the Italian
Peninsula (Fregellae, mid-3rd century BC). In Sicily and the Italian Peninsu-
la regular tessellation goes back to the 3rd century BC, maybe the end of the
4th century BC. Influences from Sicily and Egypt are attested at Carthage,
and both areas have examples of figured mosaics owing to the interplay with
Greek culture. These mosaics date to the 3rd and first half of the 2nd cen-
tury BC, but as mentioned earlier this tradition was not adopted in the Punic
metropolis.701 Instead there was a preference for non-figurative pavement de-
signs. This preference is attested at Cyrene as well. Here the tessellated mo-
saics, dated to the 3rd to 1st centuries BC, are characterized by white central
69'
Baldassarre 1994; Dunbabín 1994.
698
For photos, see Dunbabín 1994. The 3rd-century dating of the baths at Megara Hyblaia only
applies to the structures (Lafon 1991, 113).
699
Rakοb 1989, 183.
' Rue Didon/Rue Arnobe: Chelbi 1985, 83. For colour photo, see Chelbi 1985, 78, fig. 9. Ker-
kouane: Morel 1990, 499-500.
70
Sicily: Besides the Ganymede mosaic, there is the pebble mosaic from Casa dei Mosaici at Mo-
tya (isolated example), dated to the 3rd century BC by Salzmann (1982, 97). Egypt: Various examples
from Alexandria (combination of tesserae and pebbles) and Thmuís (irregular tesserae). One is dated
to the late 4th century BC by Daszewskí (1985, cat. nos. 1-2 and 5-6, 101-104 and 111-113). Daszew-
skí's `high' chronology of the Egyptian mosaics has not found wide acceptance. See for example Hel-
lenkemper Salies 1986, 242-244.
173
fields and frames decorated with geometric patterns such as meanders and
wave bands. 702
A Carthaginian-Sicilian-Roman connection should also be advocated as
regards the chequer-board design. The Carthaginian examples with a termi-
nus ante quern of 146 BC comprise both bicolour and polychrome ones. The
earliest example may be the polychrome fragment from the Rue Didon/Rue
Arnobe (4th century BC?). From Sicily the pattern spread to the Italian Pen-
insula where the motif was favoured in threshold mosaics in baths. 703 The
Italians/Romans brought it to Delos where a diagonal design often occurs.
This diagonal version of the motif was brought even further east, namely to
the settlement at Tell Anafa, probably by Hellenized Phoenicians who knew
it from Delos. Also here it is found in relation to a bathroom. 704 Thus an orig-
inally Punic motif, but in Hellenistic/Roman disguise was brought to the
Phoenician homeland. This complexity of influences is reflected by the Ro-
man triclinium. In the late Republican period the Roman elite and wealthy
adopted the great longitudinal Hellenistic hall in their houses, sometimes in
a modest version (the three-couch triclinium), and in the 2nd century AD it
appears to have been brought back to the Greek East as a Roman fashion.' 05
706
Baldassarre 1994; Dunbabin 1994.
176
has not been challenged, as far as the figured tessellated mosaics are con-
cerned. A non-figurative tradition that developed in Punic North Africa is
evidenced by the early Carthaginian material comprising a fragment of a mo-
saic from the second half of the 4th century BC and the symbols set with ir-
regular tesserae in a mortar pavement dated to the end of the 5th century
BC. Moreover, there is the fragment (opus fig /mum and opus tessellatum)
from the Rue Didon/Arnobe dated to the 4th century BC and the mosaic
fragment from Kerkouane dated to the 5th century BC.
Different traditions took new forms as they met and were developed fur-
ther, as illustrated by the Punic mortar pavements brought to the Italian
Peninsula via Sicily. Judging from the early examples from houses at Fregel-
lae, the spread of this pavement type appears to have taken place almost in-
stantaneously. The Italians/Romans refined and widened the range of de-
signs and the introduction of these pavements with their standard designs in
the Mediterranean area followed the Roman expansion: the Iberian Peninsu-
la (such as Ampurias), the East (for example Delos and Pergamon), and back
to North Africa where it all began. The use of the Punic bathtub-shaped cis-
tern at Ampurias, however, should be ascribed to direct Punic influence,
though the Romans had been acquainted with the type since their conquest
of Sardinia and Sicily in the third quarter of the 3rd century BC. A hybrid
cistern, bathtub-shaped with vault, was created in the city due to the Ro-
mans.
The internationalization of the Mediterranean area in the period under in-
vestigation is reflected by the interior architectural decoration in particular
and the statuary (Dionysiac and theatrical themes). This applies to the range
of decorative motifs for pavements (both mortar pavements and tessellated
mosaics) and First-Style murals occurring simultaneously within a vast area.
In the period from the late 4th to the first half of the 3rd century BC this
decorative system is attested both at Carthage, in Sicily (for example Hera-
kleia Mínoa) and at Fregellae. The internationalization is seen in the religious
sphere too. Foreign divinities, especially Egyptian gods such as Sarapis and
Isis, were adopted owing to the immigration of different peoples and reli-
gious syncretism.
The function of the house as a symbol of the owner's status and wealth
was also an international phenomenon. The enrichment was achieved by
shared characteristics such as enlargements of plots, porticoes or full peri-
styles, reception rooms, wall-paintings, various pavement types and the dis-
play of sculpture.
It is also necessary to recognize a flexible use of space in the ancient
house. This is obvious in houses of modest size, but even the andron may
have been used for other social occasions than the symposium and in the
daytime as indicated by the presence of windows. 707
The direct correlation between the individual house and the inhabitants is
difficult to reconstruct. Though the name of the occupant has come down to
us, this person may not be the builder of the house, or maybe this person was
only partly responsible for its layout and interior decoration, and perhaps the
last occupant moved in without changing anything. Another question is also
whether a given house was considered a permanent residence, and how great
an effort the owner would invest in redecoration, if the house was just a tem-
porary or seasonal residence. The question whether the master of the house
was owner or tenant of the house must also be taken into consideration.
Moreover, once a foreign element was introduced, anybody could adopt it.
Finally, there is the question of specialized craftsmen offering a standard rep-
ertoire of motifs etc.
through interior decoration, and consequently mere size may not necessarily
reflect the status and wealth of the inhabitant.
Delian housing gives the impression of a high degree of cultural exchange,
the vivid activities taking place here and the need for more habitation space
reflected in the many upstairs units. The sacred island was turned into a
place for profit-making where wealthy inhabitants displayed their success in
their houses. The most obvious example of self-ostentation is in De1T16
where statues of the owners, the Athenian couple Kleopatra and Dioskour-
ides, were displayed in a divine or cultic setting.
Unfortunately, the life of Punic Carthage already ended in the mid-2nd
century BC, and only a fraction of the Carthaginian houses are preserved.
Thus the results must be taken with some reservations, but judging from the
material at hand, the cultural exchange was neither so intensive nor extensive
as documented at Delos and Ampurias. The presence of foreign residents
did not lead to the introduction of new house types or room types. The es-
sentially Punic character of the housing is represented by elements such as
opus africanum, the Punic cubit, the modules employed in the `Hannibal
Quarter' and the `Mago Quarter', the long and narrow corridors, the prefer-
ence for pillars rather than columns, the bathtub-shaped cistern, the wide
range of mortar pavements, opus fig/mum and the non-figurative pavement
designs. Apparently there was a continuous influence from the Phoenician
homeland, subject to Hellenistic Greek influence as well. Epigraphical and
textual evidence document Phoenician merchants in the vicinity of Carthage
and maintenance of contact with Tyre right down to the fall of the Punic me-
tropolis.
Foreign elements in the Carthaginian housing include for instance the
Doric order, the bathtub "à. sabot", murals of the First Style and Egyptian
mouldings. Both Greek and Egyptian influences are also present in public
buildings, but again it is difficult to judge to- what extent the city was Helle-
nized and Egyptianized due to the scanty remains. Likewise, the possible
presence of statuettes of Greek and Egyptian gods ín the houses may reflect
the presence of Greek and Egyptian inhabitants as well as Carthaginians
adopting foreign themes.
Before the fatal years of 149-146 BC, the Punic metropolis had influenced
a vast area in the central and western Mediterranean, and the important role
played by Carthage as regards the development of both tessellation and mor-
tar pavements with tessera design should be emphasized.
Different cultural influences are clearly reflected in the housing at Ampur-
ias. One could say that all the peoples who are known to have lived here,
judging from archaeological and epigraphical evidence and literary sources,
left their distinct fingerprints on the houses. However, the different cultural
traditions are combined in many ways in the houses in Neapolis. Thus it is
hardly reasonable to relate a specific house to a specific origin, and it would
be wrong to assume that only Iberians lived in the houses of three or four
rooms and that only Greeks lived in the houses with Greek inscriptions.
Moreover, it is possible that Italians/Romans settled here; perhaps they had
the atrium houses built. A Roman must at least have been responsible for the
179
LEXICA
708 For example Gauckler 1904; Morricone 1970; Fischer 1971, 147; Balmelle et al. 1985; Gi-
neau 1988.
710 Balmelle et al. 1985.
711 Balmelle et al. 1985, 4.
182
and the correspondence between the different languages valuable, the desig-
nations themselves sometimes tend to be too lengthy. 12 This applies for in-
stance to the well-known lozenge pattern which is described in the following
way in the English version: "Lattice pattern in dotted simple fillets of tesser-
ae" or "Lattice pattern in serrated simple fillets of tesserae". 713 More open to
criticism, however, is the fact that this study does not define a `mosaic'and is
in fact a mixture of motifs and different pavement types. Consequently, the
terms sometimes relate to technique and material and not geometric decora-
tive patterns: for example opus segmentatum, opus signinum, opus pseudo fäg-
linum.714 Definitions of these techniques are given in the "Lexique francais",
but without references.
The second work is the bilingual Glossario tecnico-storico del mosaico/
Technical-Historical Glossary of Mosaic Art.715 It has 258 entries of ancient
and modern terms related to mosaic typologies (comprising a large range of
different pavements), mosaic materials and execution techniques. As a refer-
ence book, it is extremely useful and particularly with regard to the listing
under each term of the different definitions and usages such as no. 12. Li-
th~strοtοn. 716 However, some minor uncertainties exist, for instance opus fig-
linum that occurs twice in the list.''
The main literary sources are Vitruvíus and Pliny the Elder. Both dedicate
long passages to the subject and thus provide important inf οrmatiοn.720
However, the textual evidence and the archaeological material are two dis-
72 The authors are well aware that there are different ways of reading a pattern, but at the same
time they consider their own descriptions simple and expressed in ordinary language (Balmelle et al.
1985, 4 and 7).
'l' Balnelle et al. 1985, 316 and 317, ρls. 201b-201c. Their terms have been adopted by Tsakirgis
in her work on the pavements from Morgantína (1989; 1990).
74 Balmelle et al. 1985, for example 158, pls. 103a and 103c; 160, pls. 104c-104d.
715
Farneti 1993.
76 Farneti 1993,134-135.
tinct source materials, and an exact correspondence between the two cannot
be taken for granted. Furthermore, the diachronic and polysemic aspect of
ancient terms should not be ignored.' 21
The purpose of this section is to give an example of the use of a term and
the corresponding terminological debate.' 22 The unfruitfulness of using an-
cient terms can be exemplified by the passage in Pliny the Elder on the word
scutulatum. 723 According to Pliny, this kind of pavement was employed for
the first time in the Temple of Jupiter Optimus Maximus on the Capitol af-
ter the outbreak of the Third Punic War in 149 BC. There are several inter-
pretations of the scutulatum term: 724
an opus sectile pavement whose surface is made from tiles cut into a
number of geometric shapes such as squares, rectangles, triangles,
hexagons and lozenges. 725
a sub-type of opus sectile where the term describes the motif: cubes in
perspective formed by lozenges of different colours. 726
an instrument for levelling the surface of the pavement. 727
a pavement with inset pieces of stone or marble of various shapes and
sizes . X28
a mortar pavement with tesserae forming a lozenge pattern. 729
It is beyond the scope of the present work to render the complex discus-
sion in detail. I will concentrate on the main arguments in favour of the two
most widely adopted interpretations: nos. 2 and 4.
721 Gioseffi is aware of this and suggests how the meaning of scutula might have developed (1955,
curely dated (according to her) buildings from Rome and its environs. She repeated the interpretation
in 1994 (Morricone latini 1994). Her definition of scutulatum is followed by Tsakirgis (lorgantína)
and Spanish scholars (Ampurías and Velílla de Ebro (Colonia Victrix Iulia Lepida-Celsa)).
731 Μorricbne latini 1994, 283.
184
than the second half of the 2nd century BC. Her arguments can be summa-
rized as follows: Scutulatum cannot refer to a single decorative motif, in this
case cubes in perspective (interpretation no. 2), as Pliny deals with pavements
as such in this passage and not their decoration, i.e. scutulatum refers to the
technique of the pavement. Nothing in the passage indicates that this pave-
ment type is to be considered a rarity. If the pavement in fact depicted cubes
in perspective, its rarity would have been alluded to in the text, as the motif
had gone out of use at the time of Pliny. Passages in Vitruvius and Palladíus
support the interpretation of scutula as a "piastrella" in Feneral of different ge-
ometric shapes and not only the shape of a lozenge. 73 Another set of argu-
ments is provided by Greek transcriptions of the Latin word scutula occurring
in inscriptions from the Greek area and dated to the Imperial period. 733 In one
case (Sardes) the inscription is accompanied by drawings of different geomet-
ric figures (triangles, squares, hexagons and curved ones). All this speaks in fa-
vour of the interpretation of scutulatum (adjective to scutula) as referring to a
pavement technique employing fragments.
Interpretation no. 2: Pesando has recently rejected the interpretation sug-
gested by Morricone. Instead he argues in favour of interpretation no.
His arguments can be summarized in the following way:
Pavements with inset pieces of coloured stone are documented before the
second half of the 2nd century BC. 735 They have been found in houses from
Fregellae and the Palatine in Rome dating to the 3rd and 2nd centuries BC.
When the term is being used in relation to the Temple of Jupiter it must re-
fer to an extraordinary pavement. It is hardly likely that a mortar pavement
with inset fragments or tessera decoration forming a lozenge pattern (inter-
pretation no. 5), already in common use, was employed in the temple and
furthermore commemorated. A passage in Vitruvius confirms that scutula-
turn is a sub-type of opus sectile as scutulis should be translated with lozen-
ges.736 Republican pavements in opus sectile depicting cubes in perspective
have been found in various buildings of prestige: Pompeii: the cella of the
Temple of Apollo (terminus ante quern of 80 BC); the tab/mum of Casa del
Fauno; perhaps the Temple of Jupiter at the forum (a copy of the Capitoline
one). Rome: the cella in the Temple of Castor and Pollux (beginning of the
1st century BC, only the imprints are left); an emblema in room B of Casa dei
Grifi (last decade of the 2nd century BC) . 737 Finally, the motif also appears to
decorate polychrome marble tiles from Greek buildings of Imperial time.
The motif is defined in the Greek inscriptions by transcriptions of the Latin
732
Vitt . 7.1.4; Pallad. 1.9.5.
733 An inscription from Tralles and on buildings at ~phesos (dated to early Hadrianic time) and
Sardes (first half of the 3rd century AD).
7i4 Pesando 1997, 221-234. Earlier Gaggiotti has argued against Morricone's interpretation (1988,
217).
733
Primarily found in mortar pavements.
736 Vitr. 7.1.3-4.
737 For buildings with this design, see also Guímíer-Sorbets 1994, 19-21.
185
DELIS
The Delían pavements have been classified by Bruneau. 739 According to
his typology, there are six categories of mosaic techniques: 1) "Pavements cé-
ramique"; 2) "Pavements de galets roulés "; 3) "Pavements en éclats de mar-
bre et d'autres pierres "; 4) "Pavements de ciment "; 5) Opus tessellatum; 6)
Opus vermiculatum.
The categories relevant for the present work are nos. 2-6. The pavements
of category 2 are termed pebble mosaics in English. Category 3 consists of
two sub-types that belong to the group of chip pavements: 1. "Pavements en
longs éclats" (6-24 cm). 2. "Pavements en éclats de dimensions moyenne"
(3-7 cm). The latter is the most common and has two layers of mortar, the
"e Same view held by Bruneau and Guídobaldí (Bruneau 1988, 52; Guídobaldi 1994, 455-456).
79 Bruneau 1972.
186
upper one being finer than the lower one, 740 The "pavements de ciment"
from category 4 are subdivided into two groups: 1. Those containing a dense
admixture of ceramic fragments and embellished with white tesserae. 2.
Opus signinum is defined in the following way: "un ciment très fin, formant
une surface lisse, et d'un rose vif".741 Both groups have inset tesserae forming
geometric patterns. Opus tessellatum and opus vermiculaturn are applied to
mosaics, the latter making use of tesserae with a maximum size of 0.4 cm
square. 742
CARTHAGE
teries de diverses couleurs, mêlés à de petits fragments de marbre blanc, sans aucun souci d'organisa-
tiοn géométrique... ".
746
Opus signinum: Lancel in Byrsa 11979, 81; terrazzo: Lancel in Byrsa 11979, 87 and 89; Thuilli-
er in Byrsa I1 1982, 71; terrazzo-signinum: Lancel in Byrsa 11 1982, 118 and 121.
747 Thuíllier 1986, 113 and 115.
748 It is a widespread definition, for instance Bruneau 1972, 23; Beltr ~n-Llοris et al. 1984, 81-82;
The pavements from the houses situated in the `Mago Quarter' on the
coastal plain are treated in chapter XI Ι of the more recent German excava-
tion report. 754 The mosaics are termed opus tessellatum and opus figlinum
(tesserae of terracotta). Regarding mortar pavements the German excavators
distinguish two types with the following terms:
The German excavators reject the term cement pavements because it can-
not be translated into German. Nor can they accept the opus signinum term
because of the variety of colours to be found in the Punic material. This term
751 Lancel 1985a, 160-161 (following the definition by Bruneau 1972, 23).
752
Lancel 1985 α, 163 -164; 172, figs. 3 and 5; 173, fig. 9.
753 Lancel 1985 α, 176, fig. 12.
754 Rakob in Karthago 11991, 220-223. Α separate publication of the pavements and fragments of
the excavators reserve for the pavements of red colour embellished with geo-
metric tessera decoration, all dating to the Roman period of Carthage. Their
definition of opus signinum is thus identical with Lancel's.
In the light of Lancets sub-type "opus tessellatum, tesserae not adjoining"
(i.e. chip pavements), it is worth reconsidering some of the illustrated pave-
ment fragments found in the `Mago Quarter'. 755 All these fragments are
termed "Terrazzo", but judging from their surface appearance they are simi-
lar to, if not identical with, the analysed fragments from Byrsa. 756 Only a clos-
er investigation, such as sections and/or description of the layers, of the frag-
ments could confirm this. In fact, some of these fragments ought to be re-de-
fined as chip pavements. This means that when the terrazzo term is used in
the publication of the `Mago Quarter'without further description, one can-
not be certain whether the pavement in question should be classified as a
mortar pavement or as a chip pavement.
The remains of houses situated at various locations form a somewhat het-
erogeneous group with respect to time of excavation and methods of excava-
tion. Descriptions and terms vary likewise. Sometimes the lack of informa-
tion, for instance descriptions or photos, makes it impossible to identify the
specific type of pavement the term refers to. The most frequent terms are:
opus signinum, signinum, opus segme ~tatum, opus tessellatum, opus figlinum.
Besides, opus scutulatum/scutulatum occurs and, at least in one case, it ap-
pears that the term is used to denote a mortar pavement containing frag-
ments.757 In one case opus segmentatum is described as a mortar pavement
with a scattering of small fragments of ceramic in different colours, white
marble and greyish limestone. 8
As
No publication has so far been dedicated to the pavements of Ampurias. 759
The terms used for classifying the pavements are as follows: opus signinum,
signinum, opus tessellatum, opus vermiculatum, opus sectile, sectilia, opus scu-
tulatum, scutulata, opus testaceum or opus testacio.
Opus tessellatum and opus vermiculatum are employed for mosaics and
opus sectile for tile pavements of marble. In general, details on the opus signi-
num p avements are not given. 760 One is described as a "simple opus signi-
num". 61 The use of the opus testaceum/opus testacio term in relation to pave-
ments should be rejected and reserved for wall construction only. The cha-
racter of the opus testacio itself is not described, but one should expect it to
CONCLUSION
The survey has shown that some of the terms are used with a high degree
of consistency. This applies especially to the mosaic pavements (Opus tessella-
tum, opus vermiculatum, opus figlinum). The system of combining opus with
an adjective of more or less ancient origin is convenient, as translation prob-
lems are avoided in this way. Of course, the system only functions as long as
scholars can agree on the definitions. On the basis of the above-mentioned
considerations, the following typology and terminology are suggested:
Mosaic pavements:
A mosaic is defined by its separate surface made up of small elements.
Opus tessellatum: The surface is made up of tesserae. These can be square,
oblong or irregular. Opus tessellatum may also be embellished with inset
fragments of marble or stone.
Opus vermiculatum: A variant of opus tessellatum consisting of tiny tesser-
ae (maximum size of 0.4 cm x 0.4 cm) used for emblemata and details.
Opus fig [mum: The surface is made from terracotta tesserae of square or
oblong shape. 765 White inset tesserae may occur.
Chip pavements:
Common for pavements of this type is the insertion of not-adjoining chips
into the mortar surface. The size and setting of the chips vary to a large ex-
tent, for example the Delian pavements that make use of larger chips than
the Carthaginian ones. The crucial point is how to classify these pavements:
762
Almagro Basch 1962, 10-11: rooms 7 and 8 in the insola situated to the south-east of the forum
in the Roman city, and room 13 of AmpR4 situated in the same insu la. These pavements are covered
today.
° Puig y Cadafalch 1915-1920, 701; Puig í Cadafalch 1934, 249.
64 The scutulatum term is used by Santos Retolaza (1991, 33), whereas Almagro Basch and Bald
call it opus signinum and signinum respectively (Almagro Basch 1962, 11; Balíl 1972b, 106).
765
Dunbabín 1978, 176.
190
Tile pavements:
These pavements should all belong to one category with the following
term and definition:
Opus sectile` Pavement whose surface is made from tiles cut into geomet-
ric and other shapes. The term covers the whole range of materials (such as
terracotta, glass, stone and marble) and designs, Subdivision may be done ac-
cording "to the material, but not the design, for instance pavements made of
terracotta tiles. However, the different materials are sometimes combined
within the same pavement.'G'
Mortar pavements:
This pavement type is the most problematic, reflected in the variety of
terms employed, for example opus signinum, signinum, cement pavements
and terrazzo. 768 As the general term for this pavement type, I °consider mortar
pavements the most suitable in English because the mortar is the basic com-
ponent and at the same time the most neutral term. It is the aggregate or ad-
mixture of various materials which gives the individual character to the sin-
gle pavement. The opus signinum term should be rejected because tradition-
ally it is defined as a red mortar pavement sometimes being embellished with
tesserae. The use of the term cement may give associations to modern ce-
ment and is thus an unsuitable term.
Of course, the general term `mortar' does not account for the great variety
present within the group, and subdivisions are required. But one should bear
in mind th~t this pavement type cannot be fitted into a rigid system. In order
to distinguish sub-types or variants, a detailed classification within each site
or even excavation is needed comprising descriptions, sections and photos of
the pavements or fragments. Standard elements included in such descrip-
tions could be: the colour of the mortar; materials, colours, shapes and sizes
%6
Cf. Lancel 1985a, 163.
767 Guidobaldi 1994. He operates with two main categories: "...que ll a marmorea..."; "que lla non
marmorea...". The latter comprises for instance various limestones and artificially made materials like
brick and glass.
768 In Italian the term cocciopesto is widespread.
191
769
Dunbabín 1976, 46; Dunbabin 1978, 172-176. She uses the opus signinum term.
GLOSSARY
Chip pavement: a surface made up of densely set, but not-adjoining,
chips of stone or marble. The term includes all variants of the type.
Column: a vertical support of circular section.
Emblema: a panel, worked separately, set into the centre of a mosaic or
other pavement type.
Mortar pavement: a mortar mixed with aggregates of various materials,
decorations with tesserae and/or inset pieces may occur.
Mudbrick: a sun-dried clay brick.
Opus africanum: vertical chains of large stone blocks in which upright
blocks alternate with horizontal ones, infillíng of rubble masonry
between the chains.
Opus figlinum: a separate surface consisting of tesserae (square, oblong)
made of terracotta, decorations with white tesserae may occur.
Opus sectile: a separate surface made from tiles cut into various geomet-
ric shapes and of various materials (stone, marble, terracotta, glass).
Opus tessellatum: a separate surface composed of tesserae (square, ob-
long or irregular), decorations with inset pieces may occur.
Opus vermiculatum: a surface made up of tiny tesserae (maximum size
of 0.4 cm x 0.4 cm), used for emblemata and details.
Peristyle: a four-sided portico surrounding a courtyard or garden.
Pilaster: an .engaged pillar.
Pillar: a vertical support of square or rectangular section.
Pisé: packed mud/clay mixed with straw and ceramic, kept in place
during the drying process by shuttering.
Porticus triplex: a three-sided portico surrounding a courtyard or gar-
den.
BIBLIOGRAPHY AND ABBREVIATIONS
Abbreviations of periodicals follow those of the Archäologische Bibliographie 1993.
Periodicals not found here are written in full. Translated works have the original year
of publication in square brackets. Regarding the alphabetization of Spanish authors
the hyphen, y (Castilian) and i (Catalan) between two surnames are ignored.
Abdallah et al. 1980 Abdallah, Z.B./Annabi, M.K./Chelbi, F., Découverte d'un
quartier punique au Kram, CEDAC 3, 1980, 17-18.
About the house 1995 Carsten, J./Hugh-Jones, S. (eds.), About the house. Lévi -
Strauss and beyond, Cambridge 1995.
Actes du τ Colloque Actes du τ Colloque international d'histoire et d'archéolo-
1984 gie de l'Afrique du Nord, Perpignan, 14-18 avril 1981, BA-
Paris, noue. ser. 17 B, année 1981, Paris 1984.
Actes III 1995 Actes du IIIe Congrès International des Études Phéniciennes
et Puniques. Tunis 11-16 novembre 1991, I-I1, Tunis 1995.
AISCOM Associazione Italiana per lo Studio e la Conservazione del
Mosaico.
AISCOM 11994 Atti del I Colloquio dell'AISCOM, Ravenna, 29 aprile — 3
maggio 1993, Ravenna 1994.
AISCOM 111995 Atti del II Collogio dell'AISCOM, Roma, 5 — 7 dicembre
1994, Bordighera 1995.
Alabe 1987 Alabe, F., TEE en 1986, BCH CXI, 1987, II, 642-644.
Alabe 1988 Alabe, F., TEF en 1987, BCH CXII, 1988, II, 768-778.
Alabe 1991 Alabe, F., Peintures apparentées au deuxième style pom-
péien découvertes à Délos, KölnJbVFrühGesch 24, 1991,
33-34.
Alabe 1993 Alabe, F., Technique, décor et espace à Délos, in: Functio-
nal and Spatial Analysis of Wall Painting 1993, 141-144.
Alabe 1995 Alabe, F., Intérieurs de maisons hellénistiques: les murs
peints à Délos, RA 1995, fast. 1, 191-197.
Allison 1993 Allison, P., How do we identify the use of space in Roman
housing? in: Functional and Spatial Analysis of Wall Pain-
ting 1993, 1-8.
Allroggen-Bedel 1992 A llroggen-Bedel, A., I quattro st ill pompeiani ed il loro
ruolo ne lle provincie, ín: Coloquio de pintura mural 1992,
25-34.
Almagro 1947-1948 Almagro, M., Dos buenos retratos femeninos romanos hal-
lados en Ampurias, Ampurias IX-X, 1947-1948, 43-53.
Almagro 1949 Almagro, M., Ceramica griega gris de los sighs VI y V a.de.
J.C. en Ampurias, RStLig XV, 1949, 62-122.
Almagro 1951 Almagro, M., Las fuentes escritas referentes a Ampurias,
(Mono grafías Amp uritanas 1), Barcelona 1951.
Almagro 1952 Almagro, M., Las inscripciones ampuritanas griegas, ibéricas
y latinas, (Monografías Ampuritanas 2), Barcelona 1952.
Almagro 1964 Almagro, M., Excavaciones en la Palaiapolis de Ampurias,
(Excavaciones Arqueologicas en Espana 27), Madrid 1964.
Almagro Basch 1962 Almagro Basch, M., Ampurias, (Excavaciones Arqueol ~gicas
en Espana 9), Madrid 1962.
Andreou 1988 Andreou, A., Griechische Wanddekorationen, Mainz 1988.
Annabi, C. 1978 Annabi, C., Activités à Carthage, CEDAC 1, 1978, 19-20.
Annabi, Μ.Κ. 1981 Annabi, M.K., Fouille du quartier punique au Kram
1980-1981, CEDAC 4, 1981, 26-27.
196
Barrai í Altet 1979 Barrai i Altet, X., Els mosaics d'Emρúries. Infοrmació pre-
liminar, Faventia 1, 1979, 71-76.
Barra' i Altet 1979- Barrai i Altet, X., La preparació del corpus de mosaics de
1980 la ciutat d'Empúries, Ampurias 41-42, 1979-1980, 463-465.
Barreca 1983 Barreca, F., Le fortificazioni puniche sul Capo Bon. II. Ras
ed-Drek, in: Prospezione archeologica al Capo Bon, II, Rome
1983, 17-28.
Barreca 1986 Barreca, F., La civilta' fenicio punica in Sardegna, (Sardegna
Archeologica. Studi e Monumenti 3), Sassari 1986.
Bartman 1999 Bartman, E., Portraits of Livia. Imaging the Imperial Wo-
man in Augustan Rome, Cambridge 1999.
Bartoloni 1989 Bartoloni, P., Monte Sirai, (Sardegna Archeologica, Guide e
Itinerari 10), Sassari 1989.
Bartoloni 1995 Bartoloni, P., L'insediamento dí Monte Sirai nei quadro
della Sardegna fenicia e punica, in: Actes III 1995, 99-108.
BEFAR Bibliothèque des Écoles françaises d?Athènes et Rome.
Belarte/Barberà Belarte, M.C./Barberà, J., La casa ibèrica a la Catalunya li -
1994 teral: els elements i la dístrib υció, Cota Zero 10, 1994,
38-48.
BeltrYi Lions 1991 BeltrY.n Lloris, M., La colonia Celse, in: La casa urbana
1991, 131-164.
BeltrYi Lions et al. Beltr~n Lions, M./Mostalac, A./Lasheras Corruchaga, J.A.,
1984 Colonia Victrix Zulia Lepida -Celse Melilla de Ebro, Zarago-
za), 1. La arquitectura de la Casa de los Delfines, Zaragoza
1984.
Benichou-Safar 1995 Benichou-Safer, H., Les fouilles du tophet de Salammbô
Carthage (première partie), AntAfr 31, 1995, 81-199.
Berry 1997 Berry, J., Household artefacts: towards a re-interpretation
of Roman domestic space, in: Domestic Space 1997,
183-195.
Beulé 1861 Beulé, C.-E., Fouilles à Carthage, Paris 1861.
Bezerra de Meneses Bezerra de Meneses, U., Essai de lecture sociologique de la
1984 décoration murale des maisons d'habitation hellénistiques
à Délos, DialA n.s., anno 2, 1984.1, 77-88.
Bezerra de Meneses/ Bezerra de Meneses, U./Sarian, H., Nouvelles peintures li-
Sarian 1973 turgiques de Délos, in: Études Déliennes 1973, 77-109.
Bizard 1907 Bizard, L., Fouilles de Délos, BCH XXXI,1907, 471-503.
Blake 1930 Blake, M.E., The Pavements of the Roman Buildings of the
Republic and Early Empire, MernAmAc VIΙΙ, 1930, 7 -159.
Boak 1916 Beak, A.ER., The Magistri of Campania and Delos, ClPhil
11, 1916, 25-45.
Börker 1975 Börker, C., Scutulatum in Iovis Capitolini cede, AA 90,
1975,371-377.
Rouet 1997 Rouet, A., Olbia-de-Provence (Hyères-les-Palmiers, Var):
La maison de Pilot VI et l'évolution de la maison a pastds
de type Olynthien, in: Mélanges dédiés à la mémoire de J.
Coupry, (REA 99.3-4), 1997, 443-457.
Boussac 1988 Boussac, M.-F., Sceaux déliens, RA, 1988.2, 307-340.
Boussac 1993 Boussac, M.-F., Archives personnelles à Délos, CRAI 1993,
677-693.
Braemer 1982 Braemer, F., L'architecture domestique du Levant à l'age du
Fer, Paris 1982.
198
Donderer 1987 Donderer, M., Die antiken Pavímenttypen und ihre Benen-
nungen, JdI 102, 1987, 365-377.
Ducrey/Ducrey 1973 Ducrey, P./Ducrey, N., Les supplicies de Fourni, in Études
Détiennes 1973, 173-181.
Ducrey et α 1. 1993 Ducrey, P./Metzger, I.R./Reber, K., Le Quartier de la Mai-
son aux mosaïques, (Eretria. Fouilles et recherches VI ΙΙ),
Lausanne 1993.
Duliere et al. 1974 Duliere, C./Baaziz, S.B./Gretzinger, J./Metraux, G.PR./
Soren, D., Utique, Les Mosaïques in situ en dehors des insu-
lae I-II-III, (Corpus des Mosaïques de Tuniesie I.2), Tunis
1974.
Dunand 1981 Dunand, F., Agathodaímon, in: LIMC I.1, Zurich-Munich
1981, 277-282.
Dunand/Duru 19 62 Dunand, M./Duru, R., Oumm el-`Amed: une ville de 1'épo-
que hellénistique aux échelles de Tyr, Paris 1962.
Dunand/Saliby 1985 Dunand, M./Saliby, N., Le temple d'Amrith dans la Pérée
d'Aradus, Paris 1985.
Dunbabin 1976 Dunbabin, K.M.D., 2. The Mosaics and Pavements, in:
Humphrey, J.H. (ed.), Excavations at Carthage 1975 con-
ducted by the University of Michigan I, Tunis 1976.
Dunbabin 1978 Dunbabin, KID., The Pavement Fragments and Their
Typology, in: Humphrey, J.H. (ed.), Excavations at Cartha-
ge 1976 conducted by the University of Michigan IV, Ann
Arbor 1978.
Dunbabin 1994 Dunbabin, K.M.D., Early pavement types in the West and
the invention of tessellation, in: Fifth International Collo-
quium 1994, 26-40.
Dunbabin 1996 Dunbabin, K., Convivial Spaces: Dining and Entertain-
ment in the Roman Villa; IRA 9, 1996, 66-80.
Dunbabin 1998 Dunbabin, K.M.D., Ut Graeco More Biberetur: Greeks
and Romans on the Dining Couch, in: Meals in a Social
Context 1998, 81-101.
Dunbabin 1999 Dunbabin, KID, Mosaics of the Greek and Roman
World, Cambridge 1999.
Dupont-Sommer 1968 Dupont-Sommer, A., Une nouvelle inscription punique de
Carthage, CRAI 1968, 116-133.
Durrbach 1921 Durrbach, F., Choix d'inscriptions de Délos avec traduction
et commentaire, Paris 1921.
EAA Enciclopedia dell'A rte Antica.
EAD Exploration archéologique de Délos.
Ellis 1987 Ellis, S., Brefs Rapports Archéologiques. Grande-Bretagne,
CΕDAC 8, 1987, 11-12.
Études Délieηηes 1973 Études Délieηηes publiées à l'occasion du centième anniver-
saire du début des fouilles de l'École française d'Athènes à
Délos, (BCH Supp1. I), Paris 1973.
Fabbricotti 1976 Fabbricotti, E., I bagni nelle prime ville romane, Cron-
Pomp II, 1976, 29-111.
Fabre etal. 1991 Fabre, G./Mayer, M./Roda, I., Inscriptions Romaines de
Catalogne, III. Gerone, Paris 1991.
Fama 1987 Fama, M.L., Solunto: Evoluzione e trasformazione dell'ar-
chitettura residentale. Il metodo della ricerca, SicA 65, XX
Anno, 1987, 73-76
Fantar 1973 Fantar, M., Récentes découvertes dans le domaine de l'ar-
203
Functional and Spatial Moormann, E.M. (ed.), Functional and Spatial Analysis of
Analysis of Wall Wall Painting. Proceedings of the Fifth International Con-
Painting 1993 gress on Ancient Wall Painting, Amsterdam, 8 —12 Septem-
ber 1992, (BABesch Suppl.3), Leiden 1993.
Gaggiotti 1987 Gaggiotti, M., L'importazione del marmo numidico a
Roma in epoca tardo-repubblicana, in: Atti del IV conve-
gno di studio su L'Africa romana, Sassari, 12-14 dicembre
1986, Sassari 1987, 201-213.
Gaggiotti 1988 Gaggiotti, M., Pavimenta Poenica marmore Numidico con-
strata, ín: Atti del V convegno di studio su L'Africa romana,
Sassari, 11-13 dicembre 1987, Sassari 1988, 215-221.
Gamer 1989 Gamer, G., Formen römischer Altäre auf der hispanischen
Halbinsel, Mainz 1989.
Gandía 1908-1937 Gandía, E., Diaris d'Εχιaναc~~ .
García-Bellido 1994 García-Bellido, M.P., Las relaciones econ ~micas entre
Massalia, Emporion y Gades a través de la moneda, ín: 'be-
rosy Griegos 1994, 115-149.
Gauckler 1902 Gauckler, P., Les fοι~lles de Tunisie, RA 41, 1902,
369-408.
Gauckler 1904 Gauckler, P., Musivum opus, ín: Daremberg, Ch./Sagho,
E./Pottier, E. (eds.), Dictionnaire des Antiquités Grecques
et Romaines d'après les textes et les monuments, 1II.2, Paris
1904, 2088-2129.
George 1998 George, M., Elements of the peristyle in Campanian atria,
JRA 11, 1998, 82-100.
Gimeno Pascual 1991 Gimeno Pascual, J., Estudios de arquitectura y urbanisme
en las ciudades romanas del nordeste de Hispania, Madrid
1991.
Ginouvès/Martin 1985 Ginouvès, R./Martin, R., Dictionnaire méthodique de l'ar-
chitecture grecque et romaine I, (Collection de l'École fran-
çaise de Rome 84.1), Rome-Athens 1985.
Gioseffi 1955 Gioseffi, D., La terminologia dei sistemi di pavimentazione
marmorea e una pagina della Naturalis Historia, RendLinc
X, serie VIΙΙ, 1955, 572-595.
Gioseffi 1975 Gioseffi, D., Terminologia dei sistemi di pavimentazione
nell'anticità, in: Rossi, R./Gioseffi, D./Stern, H. et al., Mo-
saici in Aquileia e nell'Alto Adriatico, Atti della Va settima-
na di Studi Aquileiesi, 25 ~prile —1 maggio 1974, (Antichità
Alto — adriatiche VIII), Udine 1975, 23-38.
Goldberg 1999 Goldberg, M.Y., Spatial and behavioural negotiation in
Classical Athenian city houses, in: Allison, P.M. (ed.), The
Archaeology of Household Activities, London-New York
1999, 142-161.
G~mez Bell~rd 1992 G~mez Bellard, C., La Isla de Tbiza en la época de las
Guerras Púnicas, ín: Estudios de Arqueologia Ibérica y Ro-
mana. Homenaje a Enrique Pia Ballester, (Servicio de Inve-
stigaci~n Prehist~rica, serie de trabajos varies 89), Valencia
1992.
Gómez Pallarès 1997 G~mez Pallarès, J., Εdici~n y comentario de las inscripciones
sobre mosaico de Hispania: inscripciones no cristianas, Rome
1997.
Graham 1966 Graham, J.W., Origins and Interrelations of the Greek
house and the Roman house, Phoenix 20.1, 1966, 3-31.
205
Kreeb 1985b Kreeb, M., Zur Basis der Kleopatra auf Delis, Horos 3,
1985, 41-61.
Kreeb 1988 Kreeb, M., Untersuchungen zur figürlichen Ausstattung de-
lischer Privathäuser, Chicago 1988.
Lafon 1991 Lafon, X., Les bains privés dans l'Italie romaine au IF sié-
cle ay. J.-C., in: Les Thermes Romains. Actes de la table ron-
de organisée par l'École française de Rome, Rome, 11-12 no-
vembre 1988, (Collection de l'École française de Rome 142),
Rome 1991, 97-114.
Laidlaw, W.A. 1933 Laidlaw, W.A., A History of Delos, Oxford 1933.
Lancel 1981 Lancel, S., Fouilles fr an çaises. Carthage. La colline de
Byrsa et l'occupation punique (VIF siècle — 146 ay. J.-C.).
Bilan de sept années de fouilles, CRAI 1981, 156-193.
Lancel 1984 Lancel, S., Remarques sur la topographie urbaine de la
Carthage punique, in: Actes du Ι" Colloque 1984, 35-53.
Lancel 1985 α Lancel, S., Les pauimenta punica du quartier punique tar-
dif de la colline de Byrsa, in: Carthage VΙΙ. Actes du Con-
grès international sur Carthage, Trois-Rivières, 10-13 octo-
bre 1984, deuxième partie, (Cahiers des études anciennes
XVΙΙ), Università du Québec à Trois - Rivières 1985,
157-177.
Lancel 1985b Lancel, S., La renaissance de la: Carthage punique. Réfle-
xions sur quelques enseignements de la campagne interna-
tionale patronée par l'Unesco, CRAI 1985, 727-751.
Lancel 1988 Lancel, S., Les fouilles de la mission archéologique fr an çai-
se à Carthage et le problème de Byrsa, in: Lipinski, E. (ed.),
Studia Phoenicia ΙΙ: Carthago, Louvain 1988, 61-89.
Lancel 1990 Lancel, S., Problèmes d'urbanisme de la Carthage punique
à la lumière des fouilles anciennes et récentes, in: Carthage
et son territoire 1990, 9-30.
Lancel 1995 Lancel, S., Carthage. A History, Oxford 1995 [1992].
Lancel/Morel 1992 Lancel, S./Morel, J.-P., La colline de Byrsa: Les vestiges
puniques, ín: Pour Sauver Carthage 1992, 43-68.
Lancel et al. 1980 Lancel, S./Robine, G./Thuillier, J.-P., II. Town Planning
and Domestic Architecture of the Early Second Century
B.C. on the Byrsa, Carthage, ín: Pedley, J.G. (ed.), New
Light on Ancient Carthage. Papers of a Symposium Spinse-
red by the Kelsey Museum of Archaeology, the University of
Michigan, Marking the Fiftieth Anniversary of the Museum,
Ann Arbor 1980, 13-27.
Lasheras Corruchaga Lasheras Corruchaga, J.A., Pavímentos de opus signinum
1984 en el νalle medic, del Ebro, Boletín del Museo de Zaragoza
3, 1984, 165-192.
Lassére 1977 Lassére, J.-I., Ubique populus. Peuplement et mouvements
de population dans l'Afrique romaine de la chute de Cartha-
ge à la fin de la dynastie des Sévères (146 a. C. — 235 p. C.),
Paris 1977.
Laumonier 1956 Laumonier, A., Les figurines de terre cuite, (BAD XXI ΙΙ),
Paris 1956.
Leach 1997 Leach, E.W., Oecus on Ibycus: Investigating the Vocabula-
ry of the Roman house, in: Bon, S.E./Jones, R. (eds.), Se-
quence and Space in Pompeii, (Oxbow Monograph 77), Ox-
ford 1997.
208
Ledant 1995 Ledant, J, Carthage et l'Egypte, in: Actes IIl 1995, 41-50.
Lemerle 1935 Lemerle, P., Chronique des fouilles et découvertes archéo-
logiques dans l'Orient hellénique (1934), BCH LIX, 1935,
234-309.
Lemerle 1936 Lemerle, P., Chronique des fouilles et découvertes archéo-
logiques en Grèce en 1935, BCH LX, 1936, 452-489.
Le Roy 1960 Le Roy, Ch., TEP en 1959, BCH LXXXIV, 854-856.
Le Roy 1974 Le Roy, Ch., Deux notes détiennes, Π. La maison aux fron-
tons, BCH XCVIlI,1974, 388-396.
Le Roy 1985 Le Roy, Ch., Le tracée et le plan d?une villa hellénιstíque:
la Maison de Fourni à Délos, in: Dessin 1985, 167-173.
Lézine 1968 Lézine, Α., Carthage, Utique. Études d'architecture et d'ur-
banisme, Paris 1968.
Lézine 1971 Lézine, A., Utique. Note d'archéologie punique, AntAfr 5,
1971, 87-93.
LIMO Lexicon iconographicum mythologise classicae.
Ling 1991 Ling, R., Roman painting, Cambridge 1991.
Ling 1997 Ling, R., The Insula of the Menander at Pompeii. Volume L:
The Structures, Oxford 1997.
Lunas 1973 Llinas, Ch., Duas Ianuas à la Maison du lac, in: Études Dé-
tiennes 1973, 291-328.
Llorens i Rams et al. Llorens i Rams, J.-M./Pons i Brun, E./Toledo i Mur, A.,
1989 L'habitat fortifié du Ping Castellet à Lloret de Mar (Catalo-
gne), in: Habitat et structures domestiques en Méditerranée
orientale durant la Protohistoire, Pré-actes colloque interna-
tional, Arles- sur-Rhône, France, 19 au 21 octobre 1989, Aix-
en-Provence 1989,110-114.
Limas 1993 Lomas, K., Rome and the Western Greeks, 350 BC — AD
200. Conquest and acculturation in southern Italy, London-
New York 1993.
Lορéz Castro 1995 Lopéz Castro, J.L., Hispania piena. Los fenicios en la Hi-
spania romana (206 a.C— 96 d.C), Barcelona 1995.
Love 1972a Love, I.C., A Preliminary Report of the Excavations at Kni-
dos, 1970, AIA 76, 1972, 61-76.
Love 1972b Love, I.C., A Preliminary Report of the Excavations of
Knídos, 1971, AIA 76, 1972, 393-405.
Lund 1986 Lund, J., The Archaeological Activities of Christian Tuxen
Falbe in Carthage in 1838, in: Carthage VIII. Actes du Con-
grès international sur Carthage, Trois-Rivières, 10-13 octo-
bre 1984, troisième partie, (Cahiers des études anciennes
XVIIl), Università du Québec à Trois-Rivières 1986, 8-24.
Maluquer de Motes Maluquer de Motes, J./Huntingf οrd, E./ Martin,
et al. 1986 P./Rauret, A.M.a/Pallarés, R./del V. Vila, Μ.', Arquitectura
i urbanisme ibèrics a Catalunya, Barcelona 1986.
Mancinetti Santamaría Mancinettí Santamaría, G., Filostrato ai Ascalona, b an -
1983 chíere in Deli, in: Deli e l'Italia 1983, 79-89.
Mar/Ruiz de Arbulo Mar, R./Ruiz de Arbulo, J., El foro republicano de Empú-
1986 ries. Metrologia y composici~n, in: 6' Col.loqui Internacio-
nal d'Arqueología de Puigcerdà 1984, Puígcerdà 1986,
367-374.
Mar/Ruiz de Arbulo Mar, R./Ruiz de Arbulo, J., Sobre el ~gora de Emporion,
1988 ArchEspA 61, 1988, 39-60.
209
Mar/Ruiz de Arbulo Mar, R./Ruiz de Arbulo, J., Dos casas con inscripciones en
1989 griego en la Ne~polis ampuritana, in: Mosaicos Romanos
1989, 61-65.
Mar/Ruiz de Arbulo Mar, R./Ruiz de Arbulo, J., El foro de Ampurias y las tran-
1990 sformaciones augusteas de los fins de la Tarraconense, in:
Stadtbild und Ideologie 1990, 145-164.
Mar/Ruiz de Arbuló Mar, R./Ruiz de Arbulo, J., Ampurias romana. Historia, Ar-
1993 quitectura y Arqueologia, Sabadell 1993.
Marcadé 1951 Marcadé, J., Chronique des fouilles en 1950, BCH LXXV,
1951, 188-189.
Marcadé 1952 Marcadé, J., Α propos de statuettes hellénistiques en arago-
nite du Musée de Délos, BCH LXXVI, 1952, 596-624.
Marcadé 1953 Marcadé, J., Les trouvailles de la Maison dite de l'Hermès
à Délos, BCH LXXVII, 1953, 497-615.
Marcadé 1969 Marcadé, J., Au musée de Délos, (BEFAR 215), Paris 1969.
Marcadé 1973 Marcadé, J., Reliefs détiens, in: Études Détiennes 1973,
329-369.
Marcadé 1990 Marcade, J., Sur la sculpture héllenistique détienne, in: Ak-
ten des XIII. Internationalen Kongresses fur Klassische Ar
chäeologie, Berlin, 24. — 30. Juli 1988, Maiz am Rhein 1990,
145-150.
Marcet/Sanmartí, Marcet, R./Sanmartí, E., Empúries, Barcelona 1990 [1989].
E. 1990
Martin i Ortega 1997 Martin i Ortega, M.A., Ullastret, Museu d'Arqueologfa de
Catalunya 1997.
Mavrojannis 1995 Mavrojannis, T., L'aedicula dei Lares Compitales nel com-
pitum deg li Hermaístaí a Delo, BCH 119.1, 1995, 89-123.
Mayence/Leroux 1907 Mayence, F./Leroux, G., Remarques sur quelques statues
découvertes à Délos, BCH XXXI, 1907, 394-419.
Meals in a Social Context Nielsen, I./Nielsen Sigismund H. (eds.), Meals in a Social
1998 Context. Aspects of the Communal Meal in the Hellenistic
and Roman World, Aarhus 1998.
Merlin 1919 Merlin, Α., Statuettes et reliefs en terre cuite découverts
Carthage, BAParis 1919, 178-196.
Meyboom 1979 Meyboom, P.G.P., Α Mosaic Portrait at Delos?, BABesch
54, 1979, 111-115.
Meyer 1999 Meyer, K.E., Axial peristyle houses in the western empire,
AJA 12, 1999, 101-121.
Mezzolani 1996 Mezzolani, Α., Persistenza di tecniche ed ilizie puniche in
et à romana in alcuni centri della Sardegna, in: Khanoussi,
M./Ruggeri, P./Vismara, C. (eds.), Atti dell'XI convegno di
studio su L'Africa romana, Cartagine, 15-18 dicembre 1994,
Ozieri 1996, 993-1000.
Michalowski 1932 Michalowski, C., Les portraits hellénistiques et romains,
(EAD XIΙΙ), Paris 1932.
Millar 1983 Millar, F., The Phoenician cities: a case-study of hellenisa-
tíon, ProcCambrPhilSoc, new series 209, 1983, 55-71.
Morel 1969 Morel, J.-P., Kerkouane, Ville punique du Cap Bon. Re-
marques archéologiques et historiques, MEFRA LXXXI,
1969, 473-518.
Morel 1990 Morel, J.-P., Nouvelles données sur le commerce de Cart-
hage punique entre le VIΙ siècle et le 11e siècle avant J.-C.,
in: Carthage et son territoire 1990, 67-100.
210
Morricone 1970 Morricone, M.L., Mosaico and Pavimento, in: ΕEΑ Supple-
mento, Rome 1970, 504-531 and 601-605.
Morricone 1980 Morricone, M.L., Scutulata Pavimenta, (Studi e materiali
della ciuilai romana 9), Rome 1980.
Morricone latini 1967 Morricone latini, M.L. (ed.), Roma: Reg. X Palatium,
(Mosaici antichi in Italia, Regione I), Rome 1967.
Morricone latini 1971 Morricone latini, M.L., Mosaici antichi in Italia. Pavimen-
ti di signino repubblicani di Roma e dintorni, Rome 1971.
Morricone latini 1994 Morricone latini, Μ.L.; Scutulatum: precisazioni e rettifi-
che, ín: AISCOM 11994, 283-3 12.
Mosaicos Romanos 1989 Mosaicos Romanos. In memoriam M. Fernandez-Galiano.
Actas de la I Mesa Redonda Hispano francesa sobre Mosai-
cos Romanos, Madrid 1989.
Moscati 1988 Moscati, S., Fenicio o punito ~~cartaginese, RStFen XVI.1,
1988, 3-13.
Nappo 1997 Nappo, S., The urban transformation at Pompeii in the late
third and early second centuries BC, ín: Domestic Space
1997,91-120.
Nevett 1994 Nevett, L., Separation or seclusion? Towards an archaeolo-
gical approach to investigating women in the Greek house-
hold in the fifth to third centuries BC, in: Parker Pearson,
M./Richards, C. (eds.), Architecture and Order. Approaches
to Social Space, London-New York 1994, 98-112.
Nevett 1995 Nevett, L., Gender relations in the classical Greek house-
hold: the archaeological evidence, BSA 90, 1995, 363-381.
Nevett 1999 Nevett, . L.C. ; House and Society in the Ancient Greek
World, Cambridge 1999.
Nielsen 1994 Nielsen, I., Hellenistic Palaces, Aarhus 1994.
Nielsen 1998 Nielsen, I, Royal Banquets: The Development of Royal
Banquets and Banqueting Halls from Alexander to the Te -
trarchs, in: Meals in a Social Context 1998, 102-133.
Niemeyer 1987 Niemeyer, H.-G., Brefs Rapports Archéologiques. A llema-
gne, CEDAC 8, 1987, 8.
Niemeyer 1990 Niemeyer, H.-G., Karthago, Stadt der Phönizier am Mittel-
meer, Antike Welt 21.2, 1990, 89-105.
Niemeyer et al. 1993 Niemeyer, H.-G./Docter, R.F./Rindelaub, A., Mit Anhän-
gen von B. Bechtold, R.F. Docter, F.O. Hvídberg-Hansen,
Die Grabung under dem Decumanus Maximus von Kart-
hago. Vorbericht über die Kampagnen 1986-1991, RM
100, 1993, 201-244.
Niemeyer et al. 1995 Niemeyer, H.-G./Docter,.. R.F./Rindelaub, A., Die Gra-
bung under dem Decumanus Maximus von Karthago.
Zweiter Vorbericht, RM 102, 1995, 475-502.
Niemeyer et al. 1996 Niemeyer, H.G./Rindelaub, A./Schmidt, K. Karthago. Die
,
lieto Prieto/lolla 1985 lieto Prieto, F.J. /lolla, J.M.a, El yacimiento arqueol ~gico
submarin de meus-La Clota y su relation con Ampurias,
in: VI congreso international de Arqueología Submarina,
Cartagena 1982, Madrid 1985, 265-283.
Oliva Prat 1960 Oliva Prat, M., Actívidades del ServicioProvincial de Inve-
stigaciones Arqueol~gicas, Conserνaci~n y Catalogaci~n de
Monumentos de la Excma. Diputaci ~n de Germa, y de la
Delegacion de Excavaciones en 1960, Anales del Instítuto
de Estudios Gerundenses del patronato José M° Quadrado
XIV, 1960, 341-416.
Oliver Fox 1995 Oliver Fox, A.J., La presencia ρúnica en los asentamientos
ibéricos: una aproximaci~n a su problematica, in: Actes III
1995,282-296.
Olmos Romera 1989 Olmos Romera, R., Hedykoitos y Agathos Daimon. Ins-
cripciones en dos mosaicos tardohelenisticos de Ampurias,
in: Mosaicos Romanos 1989, 43-59.
Padrd/Sanmarti, E. Padr~ , J./Sanmarti, E., Serapis i Asclepi al m ~n hel.lenistic:
1993 el cas d'Emρúríes, in: Homenatge a Miquel Tarradell, Bar-
celona 1993, 611-628.
Ρadr~/Sanmarti, J. 1987 Padr~ , J./Sanmarti, J., L'ocu ρaci~~del territori per la polis
emporitana i la seva sign ~ficaci~~economica. Algunas hiρ~-
tesis, Fonaments 6, 1987, 23-26.
Palauf/iν~~1993a Palauf, Ll/Viν~ , D., Termes de la "Basilica" d'Emρúries,
in: Utilitzaci~~de 1'aigua 1993, 103-111.
Palaui/Viν~~1993b Palauf, Ll./Viv~ , D., Termes de les cases d'Em ρ~ríes, in:
Utilitzaci~~de l'aigua 1993, 118-1 24.
Papi 1995 Papi, E., I pavimenti dell e domus della pendice settentrio-
nale del Palatino (VI-II secolo a.C.), in: AISCOM 11 1995,
337-352.
Pariente 1989 Pariente, A., Chronique des fouilles en 1989. Delos. Ferme
aux jambages de granit, BCH CXIV.2, 1989, 812.
Paris 1884 Paris, P., Fouilles de Délos, BCH VIII, 1884, 473-496.
Pedroni 1991-1992 Pedroni, L., L'introduzione del signíno e del tessellato in
Sicilia, AccNapoliRend 58, 1991-1992, 649-662.
Pena 1985 Pena, M.-J., Le probleme de la supposée ville indigene a
coté d'Emporion. Nouvelles hypotheses, DialHistAnc 11,
1985,69-83.
Pena 1988 Pena, M.J., Hiρ~tesis noves sobre Empúríes a partir de l'a-
nàlisi de les fonts líteraries, Fonaments 7, 1988, 11-45.
Pena 1989 Pena, M.J., Ampurias — Les débuts de l'implantation ro-
maine, DialHistAnc 15.2, 1989, 219-248.
Pena 1991 Pena, M.J., Considerazioni sulla diffusione nel Mediterra-
neo occidentale dei bruciaprofumi a forma di testa femmi-
nile, in: Atti II 1991, 1109-1118.
Pena Gimeno 1988 Pena Gimeno, M.J., El problema del estatuto jurídico de
Emporia(e). Analisis de la documentation, in: Actas 1e'
Congreso Peninsular de Historia Antigua, Santiago de Com-
postela, 1-5 Julio 1986, I-III, Santiago de Compostela 1988,
455-466.
Pericot/Oliva Prat 1949 Pericot, L. /Oliva Prat, M., Actividades de la Comisaria
Provincial de Excavaciones Arqueol ~gicas en 1949, Anales
del Instituto de Estudios Gerundenses del patronato José M°
Quadrado IV, 1949, 318-322.
212
Pernice 1938 Pernice, E., Pavimente und figirliche Mosaiken, (Die helle-
nistische Kunst in Pompeji VI), Berlin 1938.
Pesando 1997 Pesando, F., Domus. Edilizia privata e società pompeiana fra
III e I secolo a. C., (Soprintendenza archeologica di Pompei.
Monografie 12), Pompeii-Rome 1997.
Pesce 1957 Pesce, G., Nora. Guida agli scavi, Bologna 1957.
Picard, C. 1951-1952 Picard, C., Vestiges d'un édifice punique à Carthage, Kart-
hago III, 1951-1952, 117-126.
Picard, G.-Ch. 1965 Picard, G.-Ch., Bulletin archéologique. Un art romain: la
mosaïque, REL 43, 1965, 506-523.
Picard 1985 Picard, G.-Ch., La recherche archéologique en Tunisie des
origins à l'indépendance, in Carthage V1 1985, 11-20.
Pittuu 1996 Pittau, M., Gli Etruschi e Cartagine: í documenti epigrafici,
in: Khanoussi, M./Ruggeri, P./Vismara, C. (eds.), Atti dell-
XI convegno di studio su L'Africa romana, Cartagine, 15-18
dicembre 1994, Ozieri 1996, 1657-1674.
Plana Mallart 1989 Plana Mallart, R., Le territoire d'Empúries: premiere phase
de l'implantation romaine, DialHistAnc 15.2, 1989,
249-281.
Plana Mallart 1994 Plana Mallart, R., La chora d'Emporion. Paysage et structu-
res agraires dans le nord-est catalan à la période pré-romaine,
(Centre de Recherches d'Histoire Ancienne 137. Espaces et
paysages 2/Annales Littéraires de l'Université de Besançon
544), Paris 19 94.
Plassart 1916 Plassart, A., Fouilles de Délos, BCH XL, 1916, 145-256.
Plassart 1973 Plassart, A., Un siècle de fouilles à Délos, in: Études Dé-
liennes 1973, 5-16.
Poccetti 1984 Poccetti, P., Romani e Italici a Delo. Spunti linguistici da
una pubblicazione recente, Athenaeum 62.3-4, 1984,
646-656.
Pour Sauver Carthage Ennabli, A. (ed.), Pour Sauver Carthage. Exploration et con-
1992 servation de la cité punique, romaine et byzantine, Paris-Tu-
nis 1992.
PPM 1990-1999 Pugliese Caratrilí, G. (ed.), Pompei. Pitture e Mosaici, I-X,
Rome 1990-1999.
Puig y Cadafalch 1908 Puig y Cadafalch, J., Les excavations d'Empúries. Estudi
de la topografia, Anuari del Institut d'Estudís Catalans
1908,150-194.
Puig y Cadafalch Puig y Cadafalch, J., La colonitzaci ~~greca. La colònia gre-
1915-1920 ga d'Empúries, Anuari del Institut d'Estudis Catalans VI,
1915-1920,694-712.
Puig ~~Cadafalch 1934 Puig i Cadafalch, J., L'arquitectura romana a Catalunya,
Barcelona 1934.
Queyrel 1988 Queyrel, F., Remarques sur la décoration sculptée de la
maison du Dionysos à Délos, BCH 112.1, 1988, 433-443.
Raeder 1988 Raeder, J., Vitruv, de architectura VI 7 (aedificia Graeco-
rum) und die hellenistische Wohnhaus — und Palastarchi-
tektur, Gymnasium 95, 1988, 316-368.
Rakob 1979 Rakob, F., Dossiers. Colloque de Stratigraphie. Allemagne,
CEDAC 2, 1979, 21-29.
Rakob 1981 Rakob, F., Deutsche Ausgrabungen in Karthago, in: 150
Jahre Deutsches Archäologisches Institut 1829-1979. Festve-
213
Santiago Alvarez 1994 Santiago Alvarez, R.A., Presencia ibérica en las inscripcio-
nes gríegas recientemente recuperadas en Ampurias y en
Pech labo, in: Iberos y Griegos 1994, 215-230.
Sant Marti d'Empúries Sant Marti d'Empúries. Una illa en el temps, Ampurias
1998 1998.
Santos 1998 Santos, M., Una aproximacíon al estudío de la arquítectura
domestica en la Ampurias tardorrepublicana, in: Mayer,
M./Nolla, J. Μ./Pardo, J. (eds.), De les estructures indigenes
a 1'οrganització provincial romana de la Hispània Citerior,
Homenatge a Josep Estrada i Garriga, Actes des les Jornades
Internacionals d?Arqueologia Romana, Granollers 1987,
(Itaca. Annexos 1), Barcelona 1998, 555-566.
Santos Retolaza 1991 Santos Retolaza, M., Distribución y evolución de la vivien-
de urb ana tardórrepublicana y altoimperial en Ampurias,
in: La casa urbana 1991, 19-34.
Saumagne 1924 Saumagne, Ch., Notes de topographie carthaginoise, la col-
line de Saint-Louis, BAParis 1924, 177-193.
Saumagne 1930-1931 Saumagne, Ch., Notes de topographie carthaginoise, BA-
Paris 1930-193 1, 64 1-65 9.
Saumagne 1934-1935 Saumagne, Ch., La `Maison du Ρaοn'à Carthage, BAParis
1934-1935, 52-57.
Schröder1996 Schröder, S.F., El `Asclepiο' de Ampurias: ?una estatua de
Agathodaimon del últimο cuarto del sigh o II a.C.?, ín: Mas-
só, J./Sada, P. (eds.), II Reunión sobre escultura romana en
Hispania, Tarragona 1996, 223-237.
Séveryns 1927 Séveryns, Α., Deux "graffiti de Délos, BCH LI, 1927,
234-238.
Seyrig 1929 Seyrig, H., La tride héliopolitaíne et les temples de Baal-
bek, Syria X, 1929, 314-356.
Siebert 1968 Siebert, G., Chronique des fouilles en 1967, BCH XCII,
1968, 1102-1116 and 1122-1123.
Siebert 1969 Siebert, G., Chronique des fouilles en 1968, BCH XCIII,
1969, 1031-1044.
Siebert 1971 Siebert, G., Sur la mosaïque de l'habitation V de l'ilot des
bijoux à Délos, BCH XCV, 1971, 147-165.
Siebert 1975 Siebert, G., '1 EF en 1974, BCH XCIX, 1975, 716-723.
Siebert 1976 Siebert, G., 1 EF en 1975, BCH C, 1976, 794-821.
Siebert 1978 Siebert, G., TEF en 1977, BCH CII, 1978, 848-852.
Siebert 1985 Siebert, G., Un programme d'architecture privée à Délos.
L'ilot des bijoux, in: Dessin 1985, 175-185.
Siebert 1987 Siebert, G., TEF en 1986, BCH CXI,1987,11, 629-642.
Siebert 1988 Siebert, G., TEF en 1987, BCH CXII, 1988, II, 755-767.
Siebert 1998 Siebert, G., Vie public et vie privée dans les maisons hellé-
nistiques de Delis, in: de Polignac, F./Schmitt Pantel, P.
(eds.), Entre public et privé en Grèce ancienne, Actes du
Colloque international organisé par le Centre Louis Gernet,
Paris 15-17 mars, 1995 (Ktèma 23), Strasbourg 1998,
171-178.
Siebert 1999 Siebert, G., Dédicaces déliennes et culture bilingue, in:
Khoury, R.F. (ed.), Urkunden und Urkundenformulare im
Klassischen Altertum und in den orientalischen Kulturen,
Heidelberg 1999, 95-101.
Siebert 2001 Siebert, G., L'Îlot des Bijoux. L'Îlot des Bronzes. La Maison
217
Designations of rooms marked by a solid circle (•) are the author's own.
Measurements marked with an asterisk (*) are the author's own.
As regards the number of supports (columns or pillars) those in the cor-
ners are counted twice.
1)
Cat. no.: Code in the present work. Designations given by the excavators.
Reference to plan(s).
Topography: Location in the city, quarter or insula.
Time of excavation:
State of preservation:
Dimensions: Length and width or total area.
N-S x E-W.
Building technique and material:
Date: Time of construction and abandonment or destruction.
Building history: Alterations and phases in general. For complicated cases,
references are given.
Orientation: The orientation of the house within the insula or city.
Ground plan: The shape of the ground plan. Number of rooms.
Entrance room: Identification based on location and size.
Atrium: Roofed room, identification based on the presence of an impluvium.
Courtyard: Open courtyard, may be provided with porticoes on one, two or
three sides.
Peristyle: Courtyard or garden with porticoes on all four sides.
Main room: Identification based on size, opening system, location, decora-
tion.
Delian houses: Reference to main room complex.
Other rooms: Identifiable rooms, alphabetically arranged.
Staircase: To upstairs rooms or terrace. Location.
Upper storey: Documented by fallen objects and/or staircase and/or other
indicators.
Water supply:
Drainage:
Wall decoration: Murals in situ of the First Style etc. Upper storey: Frag-
ments.
Pavements: Pavement types in situ. Upper storey: Fragments of pavement
types.
Finds: Sculpture, religious finds and epigraphical material (personal names).
These groups may overlap. Only finds from within the houses are included.
Miscellanenous:
Literature: Main references, chronologically arranged.
225
2)
Cat. no.: Code in the present work. Designations given by the excavators.
Reference to plan(s).
Topography: Location in the city or quarter.
Date: Time of construction and abandonment or destruction.
Description: The character of the remains.
Literature: Main references, chronologically arranged.
226
DELOS
General section:
Topography:
(Plan: Bruneau /Ducat. 1983, pls. I-IV and VI-V ΠΡ; Trümper 1998, fig. 1).
The Northern Quarter: North of the Sanctuary of Apollo.
The Stadium Quarter: North-eastern part of the island.
The Peribolos Street: Immediately, east of the Sanctuary of Apollo.
The Inopos Quarter: South-east of the Sanctuary of Apollo.
The House of the Masks' Quarter: South-east of the Sanctuary of Apollo.
The Theatre Quarter: South of the Sanctuary of Apollo.
The Area of the Sanctuary of Aphrodite: South of the Sanctuary of Apollo.
The Southern Zone: Southern part of the island
Time of excavation:
The Northern Quarter: Cf. the individual house.
The Stadium Quarter: 1912-1913.
The Peribolos Street: 1903-1904, 1906-1907.
The Inopos Quarter: Cf. the individual house.
The House of the Masks' Quarter: Cf: the individual house.
The Theatre Quarter: 1892-1894, 1904-1907.
The Area of the Sanctuary of Aphrodite: 1958.
The Southern Zone: Cf. the individual house.
State of preservation: Height of walls ranges from floor level to c. 4 m - 4.5 m.
Building technique and material: Outer and inner walls of roughly dressed blocks of
varying sizes bound with clay. More regular blocks may be used at the corners. Inter-
nal partition walls also built in pisé or with the lower part in stone and the upper part
in pisé.
Drainage: +
Wall decoration: First Style (a, b-c, e, f). Upper storey: Fragments (for example half-
columns and entablature).
Pavements: Opus tessellatum (a, b-c, e, f). Upper storey: Opus tessellatum. Opus ver-
miculatum.
Finds: -
Miscellaneous: -
Literature: Treheux 1949; Bruneau 1972, 127-130; Trümper 1998, 175-176.
Cat. no.: De113: House A west of the Building of the Poseidoniasts of Berytos
(Plan: Chamonard 1922-1924, pl. XXIX)
Topography: Southernmost house in the insula west of the Building of the Poseido-
niasts of Berytos.
Time of excavation: 1904. Partly excavated.
State of preservation: Cf. general section.
Dimensions: -
Building technique and material: Cf. general section.
Date: -
Building history:
Orientation: -
Ground p1a~ : -
Entrance room:
Atrium: -
Courtyard: -
Peristyle: -
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: -
Drainage: -
Wall decoration:
Pavements: -
Finds: -
Miscellaneous: -
Literature: Chamonard 1922-1924, 439.
Cat. no.: De114: House B west of the Building of the Poseidoniasts of Berytos
(Plan: Chamonard 1922-1924, pl. XXIX; Trümper 1998, 177, fig. 5; 329, fig. 60).
Topography: Insula west of the Building of the Poseidoniasts of Berytos, north of
De1N3.
Time of excavation: 1904.
State of preservation: Cf. general section.
Dimensions: 285 m2 .
Building history: Later phase: Alterations in the courtyard and south-eastern part.
Portico in front of the entrance. REF: Triimper 1998, 178.
Orientation: N-S/E-W.
Ground plan: Irregular. Nine rooms (without room i and j).
Entrance room: Room a (east).
Atrium:
Courtyard: Room b.
Peristyle:
Main room: Room f (and dining room), three rooms at the back (g, h, h'), north of
the courtyard (18).
Other rooms: Bathroom/kitchen (eastern part of c); dining room (e, f); latrine (east-
ern part of c).
Staircase: -
Upper storey:?
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (g).
Pavements: Opus tessellatum (a, b, e, f). Pebble mosaic (f, g).
Finds: -
Miscellaneous: Portico in front of the façade.
Literature: Chamonard 1922-1924, 440-441; Bruneau 1972, 146-151; Triimper 1998,
178.
Cat. no.: De115: House C west of the Building of the Poseidoniasts of Berytos
(Plan: Chamonard 1922-1924, pl. XXIX)
Topography: Insula west of the Building of the Poseidoniasts of Berytos, between
Del14 and DeΙN6.
Time of excavation: 1904. Partly excavated.
State of preservation: Cf. general section.
Dimensions: -
Building technique and material: Cf. general section.
Date: -
Building history: -
Orientation: -
Ground plan: -
Entrance room: Room a (east).
Atrium: -
Courtyard: -
Peristyle: -
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: -
Drainage: -
Wall decoration: -
Pavements: -
Finds:
Religious: Apotropaíon.
Miscellaneous: -
Literature: Chamonard 1922-1924, 441.
229
Cat. no.: De1N6: House D west 0f the Building of the Poseidoniasts of Berytos
(Plan: Chamonard 19224924, pl. XX ΙΧ; Trümper 1998, 177, fig. 5; 179, fig. 6).
Topography: Insula west of the Building of the Poseidoniasts of Berytos, north of
De1N4.
Time of excavation: 1904.
State of preservation: Cf. general section.
Dimensions: 209 m2.
Building technique and material: Cf. general section.
Date: -
Building history: Later phase: Shops (h, i) incorporated, peristyle? REF: Trümper
1998, 179-180.
Orientation: N-S.
Ground plan: Rectangular, minor irregularities. Nine rooms.
Entrance room: Room a (east). Entrance on the south side, admitting directly to the
peristyle.
Atrium:
Courtyard:
Peristyle: Room b (2 x 2 columns).
Main room: Room c, two rooms at the back (d, e), north of the peristyle (11).
Other rooms: Bathroom/kitchen (eastern part of g); latrine (h).
Staircase: -
Upper storey: -
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (b).
Pavements: Chip pavement (b).
Finds:
Religious: Lares Compitales: paintings on northern wall of room f, originally to the
right of the southern entrance, later inside the room. REF: Bulard 1926a, 70.
Miscellaneous: -
Literature: Chamonard 1922-1924, 441-442; Bruneau 1972, 151; Trümper 1998,
179 - 180.
Other rooms: Dining room (ΕΕ); kitchen (ED); latrine (EI); workshop? (EP, EQ).
Staircase: Internal, room EB. Internal and separate, room ΕI, closable.
Upper storey: +
Water supply: One cistern, two wells.
Drainage: +
Wall decoration: First Style ( ΕΒ', EE, EF).
Pavements: Opus tessellatum (ΕΕ). Chip pavement (ΕΒ, EB', EO). Upper storey:
Opus tessellatum.
Finds:
Sculpture, MA: 1 (base with inscription. REF: Siebert 1968, 1123, fig. 28).
Religious: Lares Compitales: Paintings on walls and altar; niche without paintings.
REF: Bulard 1926a, 64-69; Bruneau 1970, 591 and 599; Couílloud-Le Dinahet 1991,
111.
Miscellaneous: A column at the south-east corner of the insula.
Literature: Siebert 1968; Bruneau 1972, 151-152; Kreeb 1988, 125-126; Trümper
1998, 180-186; Siebert 2001, 55-73.
Literature: Siebert 1969; Trümper 1998, 186-187; Siebert 2001, 73-75 and 82-84.
Finds: -
Miscellaneous: -
Literature: Bruneau 1965; Bruneau 1966; Siebert 1969, 1039-1042; Bruneau 1972,
156; Siebert 1985; Trümper 1998, 190; Siebert 2001, 15-22.
Peristyle:
Main room: Room Q, room on one side (R), south of the courtyard (S5).
Other rooms: Bathroom/kitchen (N); latrine (0).
Staircase: Internal, room W.
Upper storey: +
Water supply: One cistern.
Drainage: +
Wall decoration: First Style (Q, U).
Pavements: Chip pavement (S). Upper storey: Chip pavement.
Finds:
Religious and epigraphical: Altar? with inscription. REF: Bruneau 1970, 223, 641
and pl. I.7.
Miscellaneous: -
Literature: Bruneau 1966; Siebert 1969, 1039-1042; Siebert 1985; Bruneau 1972,
156; Triimper 1998, 192-193; Siebert 2001, 31-35.
Building history: Room AK was originally placed at the back of room AL. REF:
Trümper 1998, 194-195; Siebert 2001, 39-43.
Orientation: N-S.
Ground plan: Irregular. Six/seven rooms.
Entrance room: _. Entrance on west side.
Atrium:
Courtyard: Room AI.
Peristyle:
Main room: Room AL (and dining room), one room at the front (AK), north of the
courtyard (16).
Other rooms: Dining room (AL); latrine? (AG'), apparently shared with De1115.
Staircase: -
Upper storey: -
Water supply:
Drainage: -
Wall decoration: First Style (AL).
Pavements: Opus tessellatum (AL). Opus vermiculatum (AL). Chip pavement (AI).
Finds: -
Miscellaneous: -
Literature: Bruneau/Siebert 1967; Bruneau/Siebert 1969; Siebert 1969, 1039-1042;
Siebert 1971; Bruneau 1972, 156-169; Meyboom 1979; Siebert 1985; Busing-Kolbe
1988; Kreeb 1988, 130-131; Trümper 1998, 194-195; Siebert 2001, 38-46.
1 Only general references ín the publication, not to individual rooms (Bruneau et al. 1970, 147).
238
Orientation: N-S/E-W.
Ground plan: Square, minor irregularities. Nineteen rooms.
Entrance room: Room A (south).
Atrium:
Courtyard:
Peristyle: Room I (4 x 4 columns). Two-storeyed.
Main room: Room N (and dining room), one room on either side (M, O), north of
the peristyle (12).
Other rooms: Bathroom/kitchen? (E); dining room (N, Q, R); latrine? (G).
Staircase: Internal and separate, room U.
Upper storey: +
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (A, B?, C, I, J, N, 0, P, Q, R, S, V). Figurative frize (N).
Upper storey: Fragments.
Pavements: Opus tessellatum (Q, R). Opus vermiculatum (Q). Chip pavement (Q).
Upper storey? 2 Opus tessellatum. Chip pavement
Finds:
Sculpture, BZ: 1.
Sculpture, ΜA: 6.
Sculpture, TC: 8.
Religious: Lares Compitales? Niche without paintings. -
Miscellaneous: -
Literature: Daux 1962, 963-966; Daux 1963, 869-872; Bruneau et al. 1970; Bruneau
1972, 172-174; Bruno 1985, 15-21; Harvard 1987, 153-155; Kreeb 1988, 132-142;
Trümper 1998, 202-205.
2 Cf. note 1.
239
Drainage: +
Wall decoration: First Style (AC, AD, AE, AF, AG, AH). Upper storey: Fragments
and figurative frieze.
Pavements: Opus tessellatum (ΑΕ). Opus vermiculatum (AE). Chip pavement (AC,
AF, AJ). Upper storey: Opus tessellatum. Opus vermiculatum.
Finds:
Sculpture, MA: 1.
Sculpture, TC: 4.
Religious: Lares Compitales? Niche without paintings.
Miscellaneous: -
Literature: Bruneau /Vatín 1964; Daux 1964; Bruneau et al. 1970; Bruneau 1972,
174-180; Harvard 1987, 155; Kreeb 1988, 142-145; Trümper 1998, 205-207.
Sculpture, ΜΑ: 1 .
Miscellaneous: -
Literature: Couve 1895, 492497; Chamonard 19224924, 411-416; Bruneau 1972,
182-183; Harward1987, 164; Kreeb 1988, 120-121; Trümper 1998, 211-212.
Miscellaneous: -
Literature: Plassart 1916, 228-232; Bruneau 1972, 195; Trümper 1998, 221-222.
Γ
246
Orientation: E-W.
Ground plan: Rectangular, minor irregularities. Twelve rooms.
Entrance room: Room a (south).
Atrium:
Courtyard: Room d.
Peristyle:
Main room: Room k, room on one side (1), east of the courtyard (E5).
Other rooms: Latrine (b).
Staircase: Internal, room c.
Upper storey: +
Water supply: One cistern.
Drainage: +
Wall decoration: Upper storey: Fragments and figurative frieze.
Pavements: Chip pavement (d). Upper storey: Opus tessellatum.
Finds: -
Míscellaneous: -
Literature: Plassart 1916, 247-255; Bruneau 1972, 194-195 and 197; Kreeb 1988,
174-177; Trümper 1998, 224.
Only general references, and not to specific houses or rooms (Bizard 1907, 488).
4 Cf. note 3.
249
Peristyle:
Main room: Room 14, two rooms at the back, north of the courtyard (11).
Other rooms: Latrine (X).
Staircase: -
Upper storey: +
Water supply: One well.
Drainage: +
Wall decoration: Upper storey: Fragments. 5
Pavements: -
Finds:
Sculpture, ΜΑ: 2.
Miscellaneous: -
Literature: Bizard 1907, 484-487; Bruneau 1972, 201; Kreeb 1988, 195-196; Trüm-
per 1998, 227-228.
Cf. note 3.
Cf. note 3.
250
Cf. note 3.
Cf. note 3.
251
9 Cf. note 3.
252
Topography: Insula west of the Sanctuary of the Syrian Gods, eastern part of the in-
sula.
Time of excavation: 1895, 1908.
State of preservation: Cf. general section.
Dimensions: 439 m2 .
Building technique and material: Cf. general section.
Date: -
Building history: The southern façade may have belonged to an earlier building. Lat-
er phase: Porticoes? REF: Trümper 1998, 241-242.
Orientation: E-W.
Ground plan: Irregular. Fifteen rooms.
Entrance room: Room h and k (both in the south).
Atrium: +
Courtyard: Room g, two porticoes (northern and eastern sides). Two-storeyed.
Peristyle:
Main room: Room d, room on one side (c), north of the courtyard (15) or room f,
east of the courtyard (E7).
Other rooms: Bathroom/kitchen (1); latrine (j).
Staircase: -
Upper storey: + (Capitals from an upper order, downpipe).
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (a, b, c, d, f).
Pavements: Chip pavement (g, j, k).
Finds:
Sculpture; ΜΑ: 2.
Sculpture, TC: 3.
Religious: Apotropaion.
Miscellaneous: -
Literature: Couve 1895, 506-509; Chamonard 1922-1924, 432-435; Bruneau 1972,
208; Kreeb 1988, 216-218; Trümper 1998, 241-242.
Upper storey: +
Water supply: One cistern.
Drainage: -
Wall decoration: First Style (a, f, g).
Pavements: Opus tessellatum (f). Chip pavement (f). Upper storey: Opus tessellatum.
Opus vermiculatum.
Finds:
Sculpture, MA: 2.
Religious: Lares Compitales? Altar without paintings.
Miscellaneous: Portico in front of the façade.
Literature: Chamonard 1922-1924, 435-437; Bruneau 1972, 209-217; Kreeb 1988,
218-221; Trümper 1998, 243.
Ground plan: -
Entrance room: -. Entrance on south side.
Atrium:
Courtyard:
Peristyle: (4 x 4 columns).
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage • -
Wall decoration: -
Pavements: Opus tessellatum (peristyle).
Finds:
Sculpture; MA: 1 (statue base with inscription, ID 1724).
Miscellaneous: -
Literature: Paris 1884, 486-496; Durrbach 1921, 212-214; Bruneau 1972, 226,
228-229;
Kreeb 1988, 223 -225.
Building history: First phase: Workshop and habitation. Second phase: Habitation:
Portico. REF: Trümper 1998, 252-253.
Orientation: E-W.
Ground plan: Irregular. Eight rooms.
Entrance room: + (north).
Atrium:
Courtyard: Room a, one portico (north side).
Peristyle:
Main room: Room e, one room at the back, east of the courtyard (E4).
Other rooms: Latrine.
Staircase: External, along the northern wall.
Upper storey: +
Water supply: +. Cistern in the south-west corner of the insula.
Drainage • +
Wall decoration: First Style (a, b, c, e).
Pavements: -
Finds:
Religious: Lares Compitales? Altar without paintings.
Miscellaneous: Portico along the façade.
Literature: Chamonard 1933 a, 166-168; Trümper 1998, 252-253.
Pavements: Chip pavement (c). Upper storey: Opus tessellatum (or from De1T6).
Finds:
Sculpture, ΜΑ: 7.
Miscellaneous: The house may have served as storage space for sculpture in its last
phase.
Literature: Mayence/Leroux 1907; Chamonard 1922-1924, 32; Bruneau 1972, 270;
Trümper 1998, 259-260.
Date: -
Building history: Later phase: Peristyle. REF: Trümper 1998, 261-262.
Orientation: E-W.
Ground plan: Irregular. Six rooms.
Entrance room: =. Entrance on the south side.
Atrium:
Courtyard:
Peristyle: Room e (2 x2).
Main room: Room b, one room at the back (d), west of the courtyard, (W1*).
Other rooms: -
Staircase: Internal, room e. External to the right of the entrance door.
Upper storey: +
Water supply:
Drainage: +
Wall decoration: -
Pavements: -
Finds:
Religious: Lares Compitales: Ground-floor unit: Paintings on walls and altar. Up-
stairs unit: Paintings on walls and altar. REF: Bulard 1926a, 88-93; Bruneau 1970,
594.
Miscellaneous: -
Literature: Chamonard 1922-1924, 32; Bruneau 1972, 270; Trümper 1998, 261 262.
-
Miscellaneous: -
Literature: Chamonard 1906, 575-576; Chamonard 1922-1924, 38; Trümper 1998,
266.
Pavements: -
Finds: -
Miscellaneous: -
Literature: Chamonard 1922-1924, 38; Trümper 1998, 267-268.
Main room: Room f (and dining room), one room at the back (h) and one room on
the side (g), north of the peristyle (13).
Other rooms: Bathroom/kitchen? (n/o); dining room (e, f); latrine (p).
Staircase: -
Upper storey: +
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (d, e, f).
Pavements: Opus tessellatum (e). Chip pavement (e, f, i, o, p). Mortar (central part of
peristyle). Upper storey: Opus tessellatum.
Finds:
Sculpture, MA: 2 (portraits, base with inscription, ID 1987).
Sculpture, TC: 1.
Miscellaneous: -
Literature: Chamonard 1922-1924, 39-41; Bruneau 1972, 270-273; Kreeb 1984, 323;
Kreeb 1985b; Harvard 1987, 175-176; Kreeb 1988, 282-284; Trümper 1998,
273-274.
Date: -
Building history: REF: Triimper 1998, 275-276.
Orientation: N-51E-W.
Ground plan: Irregular. Eleven rooms.
Entrance room: Room a (east). Room k (west).
Atrium:
Courtyard: Room f.
Peristyle:
Main room: Room d? g? j?
Other rooms: Latrine (in room a); shop (e).
Staircase: Internal and separate, room b.
Upper storey: +
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (a, g, k). Upper storey: Fragments.
Pavements: Chip pavement (f). Upper storey: Opus tessellatum. Mortar.
Finds: -
Miscellaneous: -
Literature: Chamonard 1922-1924, 42; Bruneau 1972, 271 and 274; Triimper 1998,
275-276.
Drainage: +
Wall decoration: First Style (House III S: a, d). Upper storey: House III S: Frag-
ments.
Pavements: Opus tessellatum (House III S: d). Chip pavement (House III R: c;
House III S: d). Upper storey: House III R: Opus tessellatum.
Finds:
Sculpture, MA: 3 (1 statue base for portrait?)
Miscellaneous: -
Literature: Chamonard 1922-1924, 48; Bruneau 1972, 282-284; Kreeb 1984,
320-322; Harvard 1987, 176; Kreeb 1988, 293-295; Trümper 1998, 281-283.
Peristyle:
Main room: -
Other rooms: Shop (51).
Staircase: Internal, room a.
Upper storey: +
Water supply: Two cisterns.
Drainage: -
Wall decoration: -
Pavements: -
Finds: -
ΜisceΠaneουs: -
Literature: Chamonard 1922-1924, 48-49.
Wall decoration: -
Pavements: -
Finds: -
Miscellaneous: -
Literature: Chamonard 1922-1924, 49.
Dimensions: 133 m2 .
Miscellaneous: -
Literature: Chamonard 1906, 577-581; Chamonard 1922-1924, 53; Bruneau 1972,
289; Kreeb 1988, 249; Trümper 1998, 291-292.
Peristyle:
Main room: Room g, one room at the back (i) and one room on the side (h), north of
the courtyard (13).
Other rooms: Latrine (b).
Staircase: Internal and separate, room a.
Upper storey: +
Water supply::. Numerous remains of lead pipes found in the courtyard.
Drainage: +
Wall decoration: First Style (g). Upper storey: Fragments.
Pavements: Upper storey: Opus tessellatum.
Finds:
Religious: Lares Compitales: Paintings on wall and altars; niche without paintings.
REF: Bulard 1926 α, 105 -107; Bruneau 1970, 595.
Miscellaneous: -
Literature: Chamonard 1922-1924, 55-56; Trümper 1998, 297.
Staircase: -
Upper storey:? (Downpipe)
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (d).
Pavements: Chip pavement (c).
Finds:
Sculpture, ΜΑ: 2 (bases).
Miscellaneous: The house may have functioned as storage space for sculpture in its
final phase.
Literature: Chamonard 1922-1924, 60-61; Bruneau 1972, 295; Kreeb 1988, 262;
Trümper 1998, 306-307.
Atrium:
Courtyard: Room d, porticoes on two/three sides.
Peristyle:
Main room: Room i north of the courtyard (17).
,
Date: -
Building history: REF: Trümper 1998, 313.
Orientation: -
Ground plan: Irregular. >_ seven rooms.
Entrance room: Room a (south).
Atrium:
Courtyard: Room b.
Peristyle:
Main room: Room c, two rooms at the back (d, e), east of the courtyard (El).
Other rooms: -
Staircase: -
Upper storey: -
Water supply: Two cisterns and one well or three wells.
Drainage: -
Wall decoration: -
Pavements: Chip pavement (a, b).
Finds:
Sculpture, MΑ:5
Miscellaneous: -
Literature: Chamonard 1922 1924, 66; Bruneau 1972, 301; Kreeb 1988, 272-274;
-
Main room: Room d, two rooms at the back (e, f), north of the courtyard (11).
Other rooms: Bathroom (c); latrine (b).
Staircase: Internal, room a.
Upper storey: +
Water supply: One cistern, one well.
Drainage: +
Wall decoration: First Style (d).
Pavements: -
Finds: -
Miscellaneous: -
Literature: Daux 1959; Marcadé 1969, 164-165 and 399 - 405; Trümper 1998,
314-315.
Orientation: N-S.
Ground plan: Rectangular. Six rooms.
Entrance room: + (south).
Atrium:
Courtyard: Room XIII.
Peristyle:
Main room: Room XIV, two rooms at the back (XV, XVI), north of the courtyard
(11).
Other rooms: Bathroom (XV ΙΙ).
Staircase: -
Upper storey: +
Water supply: One well.
Drainage: +
Wall decoration: -
Pavements: -
Finds:
Sculpture, ΜΑ: 1 (base, ID 2371).
Miscellaneous: -
Literature: Jardé 1906, 661-662; Bulard 1926a, 164; Trümper 1998, 316.
1962, 968-969; Bruneau 1970, 633; Bruneau 1972, 305-316; Ducrey/Ducrey 1973;
Marcadé 1973; Le Roy 1985; Triimper 1998, 317-318.
CARTHAGE
`HANNIBAL QUARTER'
General section:
Topography: Byrsa hill, south slope, sector Α of the French archaeological mission.
Five insulae (A-E) and five streets (I-V).
(Plan: Fig. 3; Lancel 1981, 177, figs. ha-1d; Lancel 1995, 149, fig. 75).
Time of excavation: 1947, 1952-1959 and 1974-1980 (French archaeological mission,
UNESCO).
State of preservation: Height of walls ranges from floor level up to three metres.
Building technique and material: Outer walls in dressed blocks of sandstone. Face of
outer walls provided with a packing of rubble coated with a waterproof layer. Parti-
tion walls between the houses often in opus africanum. Inner walls mostly in mud-
brick on stone socles. 10 Reused material in insulae B and D.
Date: Laid out at the very beginning of the 2nd century BC and destroyed in 146
BC.
Building history: First phase: The very beginning of the 2nd century BC. Second and
third phase: No absolute chronology. Fourth phase: No earlier than the end of the
first quarter of the 2nd century BC.
Cat. no.: CarHi: House Al
(Plan: Fig. 3).
Topography: Westernmost part of insula A.
Time of excavation: Cf. general section.
State of preservation: Cf. general section. The Roman structures have almost com-
pletely destroyed the Punic insula. The south and west corners are preserved, and
the south-eastern side towards street I is partly known. The north-western longitudi-
nal side is no longer visible. Room distribution is only partly visible in the south-
west.
Dimensions: C. 15 m x 10 m (EEC).
Building technique and material: Cf. general section
Date: The very beginning of the 2nd century BC.
Building history: -
Orientation: NW-SE.
Ground plan: Rectangular (EEC).
Entrance room: -
Atrium: -
Courtyard: -
Peristyle: -
Main room: -
Other rooms: -
Staircase: -
Upper storey: +
ο
Lancel et al. 1980, 19-20; Thníltier in Byrsa 11 1982, 61-64.
ιι
Lancel 1981, 176; Lancel in Byrsa 11 1982, 368-370 and 372; Lancel/Morel 1992, 49; Lancel
1995, 160-161.
288
Wall decoration; -
Pavements: -
Finds:
Sculpture, TC:
1. Fragment of an enthroned male statuette, head not preserved, raised right hand,
identified as Baal. FS: Coordinate Β-8. REF: Ferron/Pinard 1955, 59, N° 92, pl. LVI.
2. Fragment of headgear (plume or palm leaf?). FS: Coordinate Β-8. REF: Ferron/Pi-
nard 1955, 59, N° 93, pl. LVIII.
3. Fragment of volute. FS: Coordinate Β-8. REF: Ferron/Pinard 1955, 59, N° 95, pl.
LVIII.
4. Fragment of standing statuette, head not preserved. FS: Coordinate Β-8. REF:
Ferron/Pinard 1955, 60, N° 96, pl. LIX.
5. Woman's head. Coordinate B-7 or B-8. REF: Ferron/Pinard 1955, 60, N° 97, pl.
LIX.
Miscellaneous: -
Literature: Picard, C. 1951-1952; Ferron/Pinard 1955, 57-60.
Lancel/Thuillier in Byrsa I 1979, 193 - 195; Lancel 1981; Thuillier in Byrsa 11 1982, 79;
Lancel/Morel 1992.
Drainage: +
Wall decoration: -
Pavements: Opus fig/mum (a•). Mortar (d•, f•).
Finds: -
Miscellaneous: -
Literature: Ferr οn/Pinard 1960-1961, 97; Lancel in Byrsa I 1979, 76-77 and 81-94;
Lancel/Morel 1992.
Courtyard: Room H.
Peristyle:
Main room: Room I, south-east of the courtyard.
Other rooms: Bathroom (G); storerooms north of the courtyard towards street I.
Staircase: External, along the north-west side.
Upper storey: +
Water supply: One cistern.
Drainage: +
Wall decoration: -
Pavements: Opus tessellatum (G, I). Opus figlinum (G). Mortar (a•, E, F, H).
Finds: -
Miscellaneous: -
Literature: Ferron/Pínard 1960-1961, 96-97; Lancel/Thiiillier in Byrsa I 1979,
228-232; Lancel et al 1980, 16-17; Thuillier in Byrsa 11 1982, 61-84; Thuillier 1986;
Lancel/Morel 1992, 50.
Building history: Later phase: Room ß and y created by dividing one large area into
two smaller ones.
Orientation: NW-SE.
Ground plan: Rectangular. Three rooms.
Entrance room: Room a (south). Entrance to room ß.
Atrium:
Courtyard:
Peristyle:
Main room: -
Other rooms: -
Staircase: External, along the north-western side of the insula.
Upper storey: +
Water supply: One cistern.
Drainage: +
Wall decoration: -
Pavements: Fragments of opus figlinum (β). Mortar (a, β, y).
Finds: -
Míscellaneous: -
Literature: Lancel/Thuillier in Byrsa 1 1979, 232-235; Thuillier in Byrsa 11 1982,
74-80; Lancel/Morel 1992.
Small head of ram. FS: Courtyard. REF: Byrsa 11 1982, 130,10 Α.175.59 and 129,
fig. 163d.
Fragments of nude male statuette. FS: Room F. REF: Byrsa II 1982, 131, N°
Α.177.16 and 132, fig. 167e.
Woman's head. FS: Room F. REF: Byrsa 11 1982, 131, N° Α.177.17 and 132, fig.
167f.
Woman's head. FS: Room C, from upper storey. REF: Byrsa 11 1982, 139, N°
Α.179.58 and 138, figs. 174e and 174f; Lancel 1995, 349, fig. 222.
Foot from statuette (?). FS: Room C, from upper storey. REF: Byrsa 11 1982, 139,
N° 179.52 and 138, fig. 174c.
Religious: Fragment of censer in the shape of a Demeter head, TC. FS: Courtyard.
REF: Byrsa 11 1982, 130, N° Α.175.58 and 129, fig. 163c.
Miscellaneous: -
Literature: Lancel/Thuillier in Byrsa 1 1979, 67-75; Lance' 1981; Lance! in Byrsa II
1982, 105-141; Lance! et al. 1980, 17-18; Thnillier 1986.
'MAGO QUARTER'
General section:
Topography: Carthage-Hannibal, coastal plain. The excavation area is bordered by
Avenue de la République to the north, the coastline to the east, Rue Dag Hammersk-
jöld to the south and Rue Septime Sévère to the west. One insula, two possible insu-
lae and two streets, city wall.
(Plan: Fig. 4; Rakob 1989, between 182 and 183, fig. 7; Karthago 11991, 173, fig. 35;
Karthago 11991, plan 26-29)
Time of excavation: 1974-1980 (the German archaeological mission, UNESCO). All
houses partially excavated.
State of preservation: Generally very poor. Often dismantled right down to floor lev-
el
Building technique and materials: Outer walls and partition walls between the hous-
es of blocks of sandstone and El-Haouaría limestone bound with clay. 12 The outside
face of the outer walls provided with a packing of rubble bound with clay. Inner
walls: Foundations in sandstone bound with clay, the rest in opus africanum, usually
made of El-Haouaría sandstone. The large quantity of secondarily burnt mudbricks
found in the destruction layer provides evidence for walls of this material as well. 13
Date: The quarter was laid out around the middle or in the second half of the 5th
century BC and destroyed in 146 BC.
Building history. Three main phases: First phase: Mid-5th century BC — mid-3rd cen-
tury BC. Second phase: Second half of the 3rd century BC. Third phase C. 200-146
BC.
Cat. no.: CarMi: House I (Ια, lb, Ic)
(Plan: Fig. 4).
Topography: Western area of the insula.
Time of excavation: Cf. general section.
State of preservation: Cf. general section.
Dimensions: L House Ic: C. 15.85 m.
Building technique and material: Cf. general section.
'Z The El-Haouaria stone is both described as a sandy limestone and as a calcareous sandstone
(Karthago 11991,5: "Sandiger Kalkstein "; Karthago 11991,211: "E1-H.-Ka Ιksandstein").
Stanzi in Karthago I 1991, 211.
296
Date: Originally three separate houses: 6th century/first half of 5th century BC:
House Ic. Second half of 5th century BC: House Ia and lb.
Building history: Mid-3rd century BC — mid-2nd century BC: Amalgamation of the
three houses. No absolute chronology of the different phases exists. REF: Wiblé in
Karthago 11991, 133
Orientation: NE-SW.
Ground plan: Rectangular (EEC). Room P 1- Ρ 26. No evidence for the belonging of
room P 1, P 3-P 6, P 12, P 14-Ρ 15 and P 21-P 26 to the house in its last phase. >_
eleven rooms.
Entrance room: -
Atrium:
Courtyard: Two or three courtyards: Room P 9, porticus triplex: pillars on the eastern
(4), southern (3) and western sides (4) — the number of pillars are reconstructed.
Room P 18, porticus triplex: two deep porticoes (P 16 and P 19) and one narrow (P
17). Room P 1?
Peristyle:)
Main room: Room P 13?
Other rooms: Storeroom (P 25).
Staircase: -
Upper storey: -
Water supply: Eight cisterns, three wells.
Drainage: +
Wall decoration: -
Pavements: Opus figlinum (P 7). Chip pavement ( Ρ 9). Mortar (P 1, P 2, P 3, P 5, P
6, P7, Ρ 8, P9, Ρ ii, P13, Ρ 18, Ρ 21, Ρ 23, Ρ 24 -Ρ 24a -Ρ 24b, Ρ 26).
Finds:
Sculpture, TC:
1. A group of male and female idol, comedy scene? FS: Coordinate B/l, Punic de-
struction layer. REF: Kraus in Karthago 11991, 257, pl. 67.a.
Miscellaneous: -
Literature: Wiblé in Karthago 11991, 93-111.
Drainage: +
Wall decoration: -
Pavements: Opus tessellatum (P 59). Mortar (P 51, P 52, P 53, P 54, P 55, P 56, P 57,
Ρ 58).
Finds: -
Miscellaneοus: -
Literature: St an zi in Karthago I 1991, 9-11; Teschauer in Karthago I 1991, 136-151.
Atrium:
Courtyard: Two courtyards: Room P 72. Room P 75, porticus triplex: traces of square
bases on the northern (3), southern (3) and western (4) sides.
Main room: Room P 78, east of the courtyard.
Other rooms: Bathroom (P 79).
Staircase: -
Upper storey: -
Water supply: Four cisterns, three wells.
Drainage • -
Wall decoration: Fragments of the First Style not in situ. FS: House IVb, deposit.
REF: Laidlaw, A. in Karthago 11 1997, Nos. 5-7 and Nos. 28-30; 228.
Pavements: Opus tessellatum (P 75, P 80). Opus fig/mum (P 79). Mortar (P 72, P 73,
P 78).
Finds:
Sculpture, TC:
1. Left shoulder of enthroned male (Baal). FS: Coordinate G/2, filling beneath early
Roman pavement of room R 33. REF: Kraus in Karthago 1 1991, 257, pl. 67.b.
Miscellaneous: Fragments of Doric columns and capitals.
Literature: Teschauer in Karthago 1 1991, 152-159.
The number depends on the reconstruction of the boundary between Car Μ5 and CarM9.
299
Courtyard: -
Peristyle: -
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: Two cisterns.
Drainage: -
Wall decoration: -
Pavements: -
Finds: -
Miscellaneous: -
Literature: St an zt in Karthago I 1991, 34.
Atrium: -
Courtyard: -
Peristyle: -
Main room: -
Other rooms: Only workshops or shops? 15
Staircase: -
Upper storey: -
Water supply: No cistern/one cistern. 16
Drainage: -
Wall decoration: -
Pavements: -
Finds: ~
Miscellaneous: -
Literature: Teschauer in Karthago I 1991, 162 and 239.
Cat. no.: CarM10
(Plan: Rakob 1989, between 182 and 183, fig. 7)
Topography: West of Rue Septime Sévère.
Date: Cf. general section.
Description: Wall sections. Corridor. One cistern.
Pavements: Opus tessellatum. Mortar.
Literature: Rakob 1989, 182-189.
VARIOUS LOCATIONS
Cat. no.: Carl
Topography: Dar Saníat, between Dermech and Sidi-bou-Saïd, north-west of the.Ro-
man cistern complex.
Time of excavation: 1913
State of preservation: -
Dimensions: -
Building technique and material: Mudbrick walls.
Date: Period of the Punic Wars (3rd-2nd centuries BC). 17
Building history: -
Orientation: NE-SW.
Ground plan: Small rooms.
Entrance room: -
Atrium:
Courtyard: Two courtyards.
Peristyle:
Main room: -
Other rooms: Latrine.
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage • +
Wall decoration: - .
Pavements: Opus tessellatum (courtyard). Opus figlinum (both courtyards). Mortar.
Finds: -
Miscellaneous: Bathtub?
Literature: Renault 1911, 317; Renault 1912, 363-365 and 489-490; Cintas 1976,
124-125; Fantar 1985, 12 and 14.
17
Fantar 1985, 14. Renault suggests a dating in the 4th century BC (1912, 489).
3 03
State of preservation: -
Dimensions: C. 200 m2 .
Building technique and material: For example ashlar masonry, mudbrick, opus afri-
canum. 19
Date: Last quarter of the 4th century BC, 18 or middle/end of the 3rd century BC.
Building history: Later phase: Alterations in room 1, 2, 3.
Orientation: NE-SW.
Ground plan: Rooms with trapezoidal shapes. >_ twelve rooms.
Entrance room: -
Atrium:
Courtyard: +
Peristyle: -
Main room: Room 1.
Other rooms: Bathroom (2-3); cult room? (4); work room (9, 10).
Staircase: Internal, courtyard.
Upper storey: +
Water supply: Two cisterns.
Drainage: +
Wall decoration: Fragments of stucco not in situ, often moulded pieces and some-
times with traces of paint. FS: Destruction layer. REF: Chelbi 1984, 26.
18
Chelbi 1980, 38; Chelbi 1984, 29.
19 Lancel in Chelbi 1984, 33 (discussion).
305
Pavements: Opus tessellatum (2-3, 6, courtyard). Opus figiinum (2-3). Mortar (1, 4).
Upper storey: Mortar.
Finds:
Sculpture, TC:
Fragment of enthroned statuette (Baal). FS: Destruction layer. REF: Chelbi 1980,
35, fig. 14.
Standing female statuette. FS: Destruction layer. REF: Chelbi 1980, 35, fig. 14.
Religious: Censer. FS: Destruction layer. REF: Chelbi 1980, 35, fig. 14.
Miscellaneous: -
Literature: Chelbi 1980; Chelbi 1984; Chelbi 1992, 69.
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage: -
Wall decoration: -
Pavements: Opus sectile.
Finds: -
Míscellaneous:
Literature: Gauckler 1902, 383; Saumagne 1934-1935; Lance! in Byrsa Ι 1979, 35, n.
73; Lance! 1984, 48; Lance! 1990, 29-30; Lance! 1995, 148.
Courtyard: -
Peristyle: -
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: -
Drainage: -
Wall decoration: -
Pavements: Mortar with tessera design.
Finds: -
Miscellaneοus: -
Literature: Annabi, C. 1978, 19; Chelbi 1985, 81-82.
Byrsa 11 1982, 376; Morel in Byrsa 11 1982, 181-213; Lancel/Morél 1992, 60-64;
Lancel 1995, 161-162.
AMPURIAS
NEAPOLIS
General section:
Topography: South of Palaiapolis, immediately next to the sea.
(Plan: Fig. 9).
Time of excavation: 1908-1936. If the specific year is known, it is indicated in the in-
dividual sections on each house.
State of preservation: In general poor. Height of walls varies from floor level to c. 1.5
Μ.
Building technique and material: Socle in rubble masonry without mortar. Super-
structure in pisé.
Date: Abandoned at the end of the 1st century AD.
The order and names of streets, insulae and houses follow those given by Mar and
Ruiz de Arbulo. 20 The houses are first listed according to insulae and then numerical-
ly. Unless otherwise indicated, the dimensions of the houses are taken from Burés Vi-
laseca 1998.
Cat. no.: Ampli: House 18
(Plan: Fig. 9, Fig. 18).
Topography: North-western part of insu la A.
Time of excavation: 1917, 1934.
State of preservation: Cf. general section.
Dimensions: Exterior: C. 51* m 2 .
Building technique and material: Cf. general section.
Date: Built in the 2nd century BC. Material from the time of the abandonment.
Building history: -
Orientation: N-51E-W.
State of preservation: Cf. general section. The boundary to the north is not visible to-
day.
Dimensions: 477 m2.
Building technique and material: Cf. general section.
Date: Material from the time of the abandonment.
Building history: Later phase: Peristyle added.
Orientation: E-W.
Ground plan: Irregular. >_ ten rooms.
Entrance room: =. Entrance on south side.
Atrium: Room 1, Tuscan.
Courtyard:
Peristyle: Room 9 (6 x 6 columns) 21
Main room: Room 5, north of the atrium.
Other rooms: Dining room (6).
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage: +
Wall decoration: Fragments of red stucco not in situ. FS: cistern.
Pavements: Mortar (1, 7). Mortar with tessera design (6). Mortar with inset pieces
(impluvium).
Finds: -
Miscellaneous: -
Literature: Gandía 1916, 201; Puig y Cadafalch 1915-1920, 701; Puig i Cadafalch
1934, 250; Bahl Illana 1959, 18-19; Bald 1972α, 44; Campo/Ruiz de Arbulo
1986-1989, 155; Santos Retolaza 1991, 21; Mar/Ruiz de Arbulo 1993, 385-387 and
424-425; Burés Vilaseca 1998, 201-204; Santos 1998, 558-559.
21 According to Balil, the courtyard was provided with columns on three sides only (1972 α, 44).
314
Wall decoration: -
Pavements: Mortar/mortar with inset pieces (1).
Finds: -
Miscellaneous: Garden situated south and west of the house.
Literature: Ping y Cadafalch 1915-1920, 700-701; Puig i Cadafalch 1934, 249; Alma-
gro 1949, 74-75 (Corte II); Bald Plana 1959, 18; Bali] 1972 α, 44; Mar/Ruiz de Arbulo
1993, 354.
Wall decoration: Unknown decorative system (3). REF: Gandia 1918, 305.
Pavements: Mortar (impluvium). Mortar with tessera design (2, 3). Mortar with inset
pieces (atrium).
Finds: -
Miscellaneous: Garden with two silos.
Literature: Ping y Cadafalch 1915-1920, 701; Ping i Cadafalch 1934, 249-250; Bald
Illana 1959, 18; Balil 1972α, 43; Santos Retolaza 1991, 21; Mar/Ruiz de Arbulo 1993,
366; Burés Vilaseca 1998, 204-207.
Courtyard: Room 2.
Peristyle:
Main room: -
Other rooms: -
Staircase:-
Upper storey: ?
Water supply: One cistern.
Drainage • +
Wall decoration: -
Pavements: -
Finds: -
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1993, 403; Burés Vilaseca 1998, 207-209.
Atrium:
Courtyard: -
Peristyle:
Main room: Room 1, western part.
Other rooms: Shop? (29).
Staircase: -
Upper storey: -
Water supply: -
Draínage: -
Wall decoration: -
Pavements: Mortar (1).
Finds:
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1993, 356.
Pavements: -
Finds: -
Míscellaneous: -
Literature: Mar/Ruiz de Arbulo 1993, 356-357.
Date: -
Building history: -
Orientation: N-S/E-W.
Ground plan: L-shaped. Six rooms.
Entrance room: Room 1 (west).
Atrium: -
Courtyard: +
Peristyle:
Main room: Room 2-3, south of room 1.
Other rooms: -
Staircase: -
Upper storey: -
Water supply: -
Draínage: -
Wall decoration: -
Pavements: Mortar with tessera design (2-3).
Finds: -
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1993, 403 and 406.
Atrium:
Courtyard: Rooi 1.
Peristyle:
Main room: Room 2, west of the courtyard.
Other rooms: Kitchen (3).
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage: +
Wall decoration: Fragments of mouldings not in situ (2). REF: Gandia 1935, 103.
Pavements: Mortar with tessera design (2).
Finds: -
Miscellaneous: -
Literature: Campo/Ruiz de Arbulo 1986-1989, 155; Mar/Ruiz de Arbulo 1993, 357
and 432; Burés Vilaseca 1998, 220-222.
Wall decoration: -
Pavements: -
Finds: -
Miscellaneous: -
Literature: Puig y Cadafalch 1915-1920, 703; Mar/Ruiz de Arbulo 1993, 406-407.
Cat. no.: AmρN22: House 52 (Casa del mosaico con inscripci ~n Hedykoitos)
(Plan: Fig. 9, Fig. 20).
Topography: Southernmost house in insula O.
Time if excavation: 1919, 1924.
State of preservation: Cf. general section.
Dimensions: -
Building technique and material: Cf. general section.
Date: -
Building history: -
Orientation: N-S.
Ground plan: Probably with a garden. >_ six rooms. It is not certain if room 7, 8 and
9 belonged to the house.
Entrance room: =. Entrance on east side.
Atrium: -
Courtyard: -
Peristyle:
Main room: Room 4, north of room 1.
Other rooms: Dining room (2).
Staircase: -
Upper storey: -
Water supply: -
Drainage: -
Wall decoration: -
Pavements: Opus tessellatum (4). Mortar (5). Mortar with tessera design (2).
Finds: -
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1989, 62-63; limos Rimera 1989, 45-52; Mar/Ruiz
de Arbulo 1993, 358-359.
Pavements: -
Finds: -
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1993, 369-370.
Wall decoration: -
Pavements: Mortar with tessera design (2).
Finds: -
Miscellaneous: Garden or terrace north of the courtyard.
Literature: Mar/Ruin de Arbulo 1993, 372-373 and 445-446; Burés Vilaseca 1998,
234-236.
Cat. no.: Amρ127: House 80 (Casa del mosaico con inscrí ρci~n Xaire Agathos Dai-
mon)
(Plan: Fig. 9, Fig. 20).
Topography: Middle part of insula Ν.
Time of excavation: 1935.
State of preservation: Cf. general section. The northern and eastern limits have dis-
appeared.
Dimensions: -
Building technique and material: Cf. general section.
Date: -
Building history: -
Orientatíon: -
Ground plan: >_ four rooms.
Entrance room: Room 1 (south).
Atrium: -
Courtyard: -
Peristyle:
Main room: -
Other rooms: Cult room/dining room (3); shop? (81).
325
Staircase:
Upper storey: -
Water supply: -
Drainage: -
Wall decoration: -
Pavements: Mortar (4). Mortar with tessera design (3).
Finds: -
Miscellaneous: -
Literature: Mar/Ruiz de Arbulo 1989, 63-64; limos Romera 1989, 52-57; Mar/Ruiz
de Arbulo 1993, 374.
Atrium: -
Courtyard: -
Peristyle:
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: Two cisterns.
Drainage • +
Wall decoration: -
Pavements: -
Finds: -
Miscellaneous: Two silos in room 8.
Literature: Mar/Ruiz de Arbulo 1993, 407 and 452-453; Burés Vilaseca 1998,
242-244.
Date: -
Building history: -
Orientation: -
Ground plan: . ? three rooms.
Entrance room: =. Entrance on south side.
Atrium: Room 1 , tetrastyle.
Courtyard:
Peristyle:
Main room: -
Other rooms: -
Staircase: -
Upper storey: -
Water supply: One cistern, one well.
Drainage: +
Wall decoration: -
Pavements: Mortar (1). Mortar with tessera design (impluvium).
Finds: -
Miscellaneous: -
Literature: Bald 1972a, 45; Mar/Ruiz de Arbulo 1993, 375; Burés Vilaseca 1998,
244-246.
Cat. no.: AmρN32: House 87
(Plan: Fig. 9, Fig. 18) .
Topography: North-western corner of insula B.
Time of excavation: 1920.
State of preservation: Cf. general section.
Dimensions: 84 r2.
Building technique and material: Cf. general section.
Date: Material from the time of the abandonment.
Building history: -
Orientation: N-S.
Ground plan: Rectangular. Five rooms.
Entrance room: + (west).
Atrium:
Courtyard: Room 1.
Peristyle:
Main room: Room 2, north of the courtyard, interior dimensions: C. 4* m x 6* m.
Other rooms: -
Staircase: -
Upper storey: -
Water supply: One cistern.
Drainage: +
Wall decoration: -
Pavements: Mortar (4).
Finds: -
Miscellaneous: -
Literature: Campo/Ruiz de Arbulo 1986-1989, 156; Mar/Ruiz de Arbulo 1993, 360
and 442-445; Burés Vilaseca 1998, 248-250.
RΟMAΝ CITY
Cat. no.: AmpRi: Roman house 1 (Casa Villanueve)
(Plan: Fig. 10, Fig. 12).
Topography: North-eastern part of the city, north of Am ρR2 and AmpR3.
Time of excavation: 1923-1924, 1936, 1942-1947.
State of preservation: C. 0.25 m-1.3 m. Eastern part up to e. 2.5 m.
Dimensions: 3492 m2.
Building technique and material: First phase: Socle in opus incertum reinforced at the
corners with ashlars 22 Superstructure in pisé. Second phase: Reuse of earlier materi-
al. Third phase: Socle in opus certum/vittatum. Superstructure in pisé. Fourth phase:
Socle in opus certum /vittatum and reused material.
Date: Beginning of the 1st century BC. Abandoned in the second third of the 2nd
century AD.
Building history: First phase: First quarter of the 1st century BC: Two zones: Habita-
tion on the western higher level and service sector on the eastern lower level.
Second phase: No later than the second half of the 1st century BC: The entire insula
occupied. Peristyle with garden and cryptoporticus added to the south. Rooms on
the northern side of the peristyle. The rooms south of the atrium remodelled.
Third phase: Later part of the first half of the 1st century AD: Extensions towards
the north and beyond the city wall (demolished). Courtyard with three porticoes and
various rooms. Atrium.
Fourth phase: End of the 1st century AD: Rooms south of the peristyle. The garden
divided down the middle by low walls.
Orientation: N-S (originally E-W).
Ground plan: Rectangular. > fifty rooms.
Entrance room: Room 30 (west).
Atrium: Room 30a•, Corinthian, hexastyle (originally tetrastyle?). Room 18, Tuscan?
Courtyard: Room 12, three porticoes (northern, southern and western sides). Room
50.
Peristyle: Room 46.
Main room: Room 37, tablinum, east of the atrium, one room on either side of the ta-
blinum opening onto the atrium (36, 38).
Other rooms: Dining room (5, 42, 43, 53•).
Staircase: -
Upper storey: Probably above the northern part of the house.
Water supply: Nine cisterns.
Drainage: +
Wall decoration: Esquema 1 (15, 26, 47). Unknown decorative system (10, 12, 17,
18, 42). Panel A not in situ (north-western zone of the house).
Pavements: Opus tessellatum (49, 51 and rooms situated in northern part of the
house, for instance 17, 18). Opus vermiculatum (for example emblema from room
north of the peristyle). Opus sectile (rooms situated in northern part of the house).
Mortar (12, 13, 24, 30, 34, 35). Mortar with tessera design (36, 38). Mortar with inset
pieces (impluvium, 37).
Finds:
Sculpture, MA:
Portrait of a woman. FS: Room on the south side of the peristyle.
Two herms, heads not preserved. FS: Roman house 1.
22
Opus incertum at Ampurias differs from the Italian version, for example the mortar is visible in
the surface of the wall (Mar/Ruiz de Arbulo 1990, 147, note 5).
330
Miscellaneous: Additional garden zone east of the northern part of the house.
Literature: Puíg í Cadafalch 1934, 252; Bald 1972b, 90-101; lieto Prieto 1977; lieto
Prieto 1979-1980; Gimeno Pascual 1991, I, 127-131, 235-237, 400-404 and 443; II,
886-888; Santos Retolaza 1991, 22-31; Mar/Ruiz de Arbulo 1993, 242, 390-391 and
395-397; Burés Vílaseca 1998, 325-334; Santos 1998, 559-565.
Building technique and material: First phase: Socle in opus incertum. Superstructure
in pisé.
Second phase: Socle in opus incertum. Superstructure in pisé. Third phase: Socle in
opus incertum. Superstructure in pisé. Fourth phase: Socle in opus certum/vittatum.
Mudbrick also used for superstructure.
Date: Beginning of the 1st century BC. Abandoned in the second half of the 2nd cen-
tury AD.
Building history: First phase: First quarter of the 1st century BC.
Second phase: Second third of the 1st century BC: Extension to the east beyond the
city wall: Peristyle with garden and adjacent rooms, including room 15 from ΑmρR2.
Further to the east room 42-49.
Third phase: 1st century AD: Extension to the north: Baths (room 30-33), peristyle
with garden and room 34-36 to the north.
Fourth phase: End of the 1st century AD or beginning of the 2nd century AD: Room
46 and alteration of the eastern facade. Garden zone with porticoes to the east.
Orientation: E-W.
Ground plan: Irregular.
Entrance room: Room 50• (west).
Atrium: Room 51•, Tuscan. Two alae (18, 24).
Courtyard:)
Peristyle: Room 35a•, with garden. Room 40a•, with garden.
Main room: Room 52•, tablinum east of the atrium.
Other rooms: Bath (30-32); dining room (15, 16, 21, 25); kitchen (33).
Staircase: -
Upper storey: -
Water supply: Three cisterns.
Drainage: +
Wall decoration: Esquema 1 (40a•). Esquema 3 (16, 17, 26, 37). Decoration resem-
bling the Third and Fourth Styles (41). Unknown decorative system (40a•, 47) 2 4
Pavements: Opus tessellatum (15, 40, 43). Opus tessellatum with inset pieces (implu-
vium). Opus sectile (35). Mortar (atrium, tablinum zone, 16, 25, 41). Mortar with tes-
sera design (15, 17, 19, 20?, 21, 26, 31-32). Mortar with inset pieces (30, 31).
Finds:
Religious: Altar, paintings. FS: Peristyle garden (room 40a•).
Miscellaneous: -
Literature: Balil 1972b, 101-104; lieto Prieto 1977; lieto Prieto 1979-1980; Gamer
1989, 28-29 and 232-233; de Chazelles 1990; Santos Retolaza 1991, 31-33; Mar/Ruiz
de Arbulo 1993, 239-240 and 391-394 (the house is erroneously called 2A); Palauí/
Viν~~1993b; Carrion Masgrau/Santos Retolaza 1995; Burés Vilaseca 1998, 336-340.
Z" Other fragments have been found within the area of ΑmρR2 and ΑmρR3.
CONCORDANCE BETWEEN
CAT. NOS. AND HOUSES
Trümper 1998 Kreeb 1988
Delos, Northern Quarter
De1N1 = House north of the French School Cat. no. 1
De112 = House east of the French School
De113 = House A west of the Building of the PB*
De114 = House B west of the Building of the PB* Cat. no. 2
De1N5 = House C west of the Building of the PB*
De1N6 = House D west of the Building of the PB* Cat. no. 3
De117 = House I-II, Îlot des bronzes Cat. no. 4-5 Cat. no. 4
De118 = House III, Îlot des bronzes Cat. no. 6 Cat. no. 4
De119 = House IV, Îlot des bronzes Cat. no. 7 Cat. no. 4
DelN10 = House VI, tot des bronzes Cat. no. 8 Cat. no. 4
De1N11 = House VII, Îlot des bronzes Cat. no. 9 Cat. no. 4
De1N12 = House I, Îlot des bijoux Cat. no. 10 Cat. no. 5
De1113 = House II, Îlot des bijoux Cat. no. 11 Cat. no. 5a
De1114 = House III, Îlot des bijoux Cat. no. 12 Cat. no. 5
Dell 15 = House IV, tot des bijoux Cat. no. 13 Cat. no. 5
Dell 16 = House V, tot des bijoux Cat. no. 14 Cat. no. 5b
De1N17 = House VI, Îlot des bijoux Cat. no. 15 Cat. no. 5
Dell 18 = House VΙΙ, tot des bijoux Cat. no. 16 Cat. no. 5
Dell 19 = Maison du Diadumène Cat. no. 17 Cat. no. 7
De1N20 = Maison aux frontons Cat. no. 20 Cat. no. 6a
De1121 = Maison des comédiens Cat. no. 18 Cat. no. 6b
De1122 = Maison des tritons Cat. no. 19 Cat. no. 6c
De1123 = Maison des sceaux Cat. no. 21 Cat. no. 3
De1124 = Maison de l'épée
De1125 = Maison de la colline Cat. no. 22 Cat. no. 2
De1N26 = House west of Maison de la colline
De1N27 = Maison du lac Cat. no. 23 Cat. no. 9
De1N28 = House west of Maison du lac
De1129 = Maison de Skardhana Cat. no. 24
* = Poseidoniasts of Berytos
Car6 = Juno hill, at the foot of the east slope, near the Carthage-Hannibal station, Terrain des. Héri-
tiers Grosjean
Car7 = Beneath Maison du Paon
Cara = Carthage-Dermech, Terrain Clariond
Car9 = Between the Salammbo and Carthage-Byrsa station, on the Avenue Habib Bourguiba, Ter-
rain Ghali
Car10 = Le Kram, Avenue Habib Bourguiba, opposite the municipal stadium
Carll = Byrsa hill, south slope, sector B of the French archaeological mission
Ampurias, Neapolis
Ampli = House 18
ΑmρΝ2 = House 19
ΑmρΝ3 = House 20 (Casa del ngulo)
ΑmpΝ4 = House 23
ΑmpΝ5 = House 1 (Domus de las ínscripciones)
AmpΝ6 = House 5 (Casa G)
ΑmρΝ7 = House 7 (Domus H)
ΑmρΝ8 = House 9 (Casa de la escalera)
ΑmρΝ9 = House 10 (Casa de los cuatro departamentos)
Ampl10 = House 28 (Casa de la herrería)
Amphi = House 32
ΑmρΝl2 = House 33
Αmp113 = House 34 (Domus del atrio tetr~stilo)
AmpΝ14 = House 35
Ampl15 = House 41 (Casa de la puerta gemínada)
ΑmpΝ16 = House 42 (Casa de la conducci ~n de inforas)
ΑmpΝ17 = House 43 (Casa de la esquina del agora)
ΑmρΝ18 = House 50
ΑmρΝ19 = House 55 (Casa sobre el dep ~sito de espades)
ΑmρΝ20 = House 56 (Casa sobre el homo gríego)
AmρΝ21 = House 57 (Casa del homo)
AmpΝ22 = House 52 (Casa del mosaico con ínscripci~n Hedykoitos)
ΑmρΝ23 = House 70
ΑmρΝ24 = House 74
Ampi" = House 75 (Casa de la cisterna en L)
ΑmρΝ26 = House 76 (Casa de la cisterna -eremitorio)
ΑmρΝ27 = House 80 (Casa del mosaico con inscrιpci~n Xaíre Agathos Daimon)
ΑmρΝ28 = House 82
ΑmρΝ29 = House 83 (Casa de los silos de ladríllo)
ΑmpΝ30 = House 84
337
~~
~υ
CJ a κκ κκ π. κ π. π. κκκκ
Scu lpture
κ κ κ κ κκκ κκ κ
Pavemen t
0
~~
,
types
~~ ~ ~~
~~ ~~ ~~~~ ~~ ~~ ~~~~~~ Ó β υ ~~~~~~~~ ÓÓ Ó ~~
decoration
Wa ll
~~ ~~ ~~
-i---1- +
κ κ κκ n. κκκκκκκκκκκ κκ κ
~α,
~ κ κ κκ κ κ κκ κ κ κκκ κ κ
C)
Οι
. (-,
κ κ κκ κκκκ κ κκκ κ κ κ
Upper
storey
κ n. κ κ κπ. κ κ κ κ κ κ κπ. κ κ κ κ κκ κ π.
π.
Other rooms
π. ~
Q) 0 Cl
~
b '2Q)
3
Γα -- ;
Q)
;
~~ cC -ti
π . --~ .
Q) Q)
~ ω 02 ~ o ~ n ~~ ó~°~ o ,o »o „n n
Ζ Ζ Ζ ΖΖ ~ ~ Ζ ~ Ζ ΖΖ ΖΖΖ ~ ΖΖ ΖΖ zZ Ζ
Main
room
ΖΖ
κ κ κκ κκκκκκκκκκκκκκκκ κκ κκ π.
Circulation
ο ο ο n.
0 c â ~. ~ 0
~ι ~ ~υ p. no- Μ
space
υα υ ()Q υΡΗ u
Entrance
room
κ κ κκ
κ κκκκ κκκ κκκκ κκκ κ κ κ π.
ο
Φ
.--~ η Σ' ' - r- ii lι' N- η Vr Φ 1"- 'Ο 1" lι' N- η η Λ
ι
Λ Φ O Ι » '-1 "-) (Η "-Η Μ ~1 CC"-) Σ' Ο η- Vr Μ.-) .-) η 00' N-
2 Ν (Η «-•~«-•- «-•-- «-•-- «- Ν '-1 000' 1"' lt Ν M M Ν
R,e ~ 2 ~
C L Ν
b7) CC bp ε , CC CC y , CC ~, 4 δρ ε, c~C CC
ü ν ~~ν v V ú ~~ú ú v v v V ~~ v V~ v V O ω
~"
.~..
0 α 000'
M NN NN N N
Cat. no.
ΦΠ .-
Zz Ζ z Ζ Z Ζ Z Ζ%. ~~
~ %. %. %a%.%,%, ΧΖΖΖΖΖ
- - - - Ζ %.Ζ Z Ζ z
-
Q: GJ QJ G% CJ EJ GJ L Ú N L W L W L W N Ú N d~ L W WLÚ N W N
Α ΑΑΑΑΑΑ ΑΑ ΑΑΑ Α Α ΑΑΑ Α ΑΑ Α ΑΑ ΑΑΑΑ ΑΑ
342
~~
~ν
Vg) λ κ κ π. κ π. κ π. κκ π.
Scu lptu re
κκ κ κ κκ κ κκ κκ κ κ κ κκ
ο
ο
Pavem en t
~~ ~ ~~0 ~~ ~~ ~~
types
000 ~~0µ
Ó ~~ ~ 0Φ-, ~ ~~Ó ,~~xυ" ~ ~~~~ Ó Ó ~~
decoration
Wa ll
~~
+~~ ~~~~ ~~ ~~
+
κ κκ κκ κ κ κκ κ κκκκ κκ κ κκ κκ
~ g) κ κ κ κ κ κ κ κ
g)
,4
V κκ π.κ κ κ κ κ κκ
Upper
storey
.- -
r ~
~~ ~ zz ~ ~~~~ z z ~~ ~ z ~ zz z zz
κκκκ z XXX κ κ κ κκ κ Ζ κκ κ κκ
Circulation
ο ~
~~ ~+ ~~ ~~
s
space
s
υρ.,υυ ~ υ ~, υ υ υυυυυυυ υυυαα υα
d'
Entrance
κ
room
κ κ
κκ κ κκκ κκκκκκκ n. - κ κ κκ κ ' κκ
c
Vr O.-- ιΠ ι Οιη Vr . Ν
Ο
^
Ν V
~1 ~\ ν1 V' '0
V1 ' Ν - '--i V' ~ Ο ~ ~ Ο Μ Ν η O V
~ g) Γ--ι ~--~ Μ Ν Ν Ν Ν -ι
ΓΜ ~ Ι- Ι- '½ <1' .--ι ~~
~ ~ ~ c~α
Ground-
~ ~ ~ g) ~ ~
δΑ ~,., bi) δΑ ~ ~ bi) bi) bi) bi) bi)
g) tg)
ρlαη
ηΓΛηΓΛ αζ, .-
α. -ι F-~F-ι F -1 f•~-ι ι~ F~-ι G Γ. -
- -/.. Γ.
@ L Φ Φ . . Φ 71 .Φ 71 ~) 71 71 ν - L W 71 ~1 Ν ΦΦ Φ L Φ
ΑΑΑΑ ΑΑΑΑ Α:ΑΑΑΑΑΑΑ ΑΑ Α' ΑΑ ΑΑΑΑΑ
343
~~
~
sV χ
Λ. κ κ Λ. κ
ν
λ
~
+α
0.
\ι
η κκ κκ κ DC
Pavement
υ G µ
ς
types
decoration
Wall
+
κ κ κ κκ κ κκκκκκχκ κκκχ κ
~
ν
~~ κκ κκ κ κκ κ κ κ χκκκ κ
~λ
.h
V κκ κκ κ κκ κκ κ κ κκ
Λ.
Upper
storey
,-q
η
κκκκ Χ ''. κκ XXX XXX ~ κκκκ κκ
Other rooms
3τ
44
Λ. ι.. Λ. ~
^
~
i ~~ ~~
_$ ._$,~ á -]
^ "'i . -ι C ~ .- V]
. ~~ ς.
λ
~
y •-
η-ι ^ •-
εF
η Γ^η
4 W 4 W
y- ι
ι4 C~ ~
I~
η
-
y η .--ι ~
Γ -ι -
η •- ι ^
η η
4 ~ F-ι Fy W !~ 4 4 (Λ F-ι Lι
~Π M
Γ -ι ^ y y η
/ti (η W^ ~ 1η
I ~ ^ .--ι η - ι
η
f-ι L-ι 4 •--~
\
κκκκ XX κκχκκκκκχχκκ κκκκκκ κz
ο
ο _
Circulation
Do ,o _ οο οο ο
Μ CL ~ η η O ιιη
η ~ Rι L1ι
space
κ κ χ
χ κ κ κκ κκκ κκκκκκκ κκκκκ κκκ κ
Ο
~ 'D ~
Ο
η~ Ν
ο ΟΟ 1 V' οΟ Φ V
ι εΥ O η.--ι Μ .-ι Ο Ó ~ η Μη Vι O
Ν •--) •--4 Ν .D Ν .--ι .-ι
~ ΜΟ Μ Ν ~ If'
Μ •--ι Μ Μ •--ι . --ι .D Ο. -ι εΥ (V Μ •--ι + •--4 .--ι N (V
44 44 44 4444 4Ο 44 444444444444 ca cE 44 44 n 44
η
~ λ µ λ - ~ '-
QQ QQ ., QQ QQ QQ QQ QQ + bAbAbAbA DA W bAbA W bA c3 bA QQ
tω
ρlαη
00
ι. ι. í.
υ000 ι ι ι .ι ι ι. ~ υ Ú τ .ι . ~. ι ι. ι. ιι t) O ) 0
υ ω ω ω ω ι . ιι . .µ C Γ ι
ι
. ~ .~ ~4 .. ...ι . ,~. ..~
. . ,~.. .~ 4 4 4 4 •.
~ ι... ..ιι-.ι .~
-ι • .-ι ~ ιλ.
.ι.. .. - ι
υ .~ .~ 4 . .
~
~
~~ ~
1y
~υ 0
Ua κ η• .• κκ κ π. χχ η• κ
~
~ti
>~
Scu lpture
Ο
0
ι.
ο0
χ κ κ κ χ κ χ κκκ κκ >~
ο
[
Ο
Pavemen t
~~ ~~ ~
µ
~
~~ ~
types
~~ ~~S νΝ
0
3 .~
decoration
L
II ν
8 40
Wa ll
W
+ + + ν Ε'ι
κκκκκ κκκκκ κκ κχκκ κκ χ Ο Ν
• 1η
0
bo ν
~
.S v
~ κκ κκ κκη. κ κχκκ χ κ XXX
~
• ~ο
00 "
~
~ 0
.r~ ο
3 •~
~~
~~ κ κκ x κκ κκχχ n. xx
νο
'-
Upper
~~
storey
~ Μ - ú4)0κ
Ο .-
Λ. .χ -2 O Λ. ~ ‚-ι. ‚-ι. hη M1 .
‚_ι.
;
~. Y
η
Υy
~γ
\ 00
\y
~ ~~.~ \ µ
Ν -2
@ @@
,.Ο ..Οa ti ~~ ~ti.~ ,.Ο ..Ο ΓΟ ~~ ^~~
"0
Ii 2
~
~~ ο
2
ZW
κκ
ZZZZ2ZZ
κκκχ κκκκκκ χχ κκκκκ
~ ZZZ W f-τλΓ1λΖΖ W ΖΖΖΓ~λ
κ χχ κ
κ~
Q-4-' -'
i-ι ι
ο ν
Entrance Circulation
Ι
space
cV
QQ υαυυυυυυυυυαω ~ υ ~ υ ~ Q, υ υυυ ~~ '
b.
$
ο ο.
room
κ κ χκ κ νι 0.
κ κκ κκκκη.κκκ κκ XXXXX XXXX
~c ω
-X ο
Ν ρ 00)
r' 'D
r' '
Γ-. .--ι 00 '0 ι'~~ 1Λ --1 ∆\ η∆\ 0' Ν•—
η V' Μ V' 1ΛΛ Vr ~
ηι Ο ηΟ
00 η' .--ι
r- Ο -Ω
ην Ν ηη CY Μ fV ~\
~
λ r+ η .- .--1 ~ Φ.--ι .--ι .-ι η 7 .-. η .--ι .-ι η .-ι Φ .--ι .-ι ~~
Ο
40 II
'4 '4 '4 '4
rt ~'α
1 ~~ ‚4 ι'
Ground-
"
cα cα α α ~ É ~ ` ~ c'a" ca ι~
α ια
ε. ~7 II ~
~ ~~~~~ ~cv. v
µ λ
bA bA bA bA δΑ bp bp ~Ν bA cC OA
v C! W v v ν~ ν Ο
plan
µC/ ν. µ µνΟ µ
ν ν ν ν ν µv µv ~
. .~ ..~. ιΟ. ...~. .~ .~ Ο"
ΙΙ.
Ι.µ
ν N
Ν
φ ~-
So uthern
-
~
Zon e
ν
Α
Sanctuary of
Area of the
Ap hrodite
C)
Α Α
Ο[t ~ Π V' Μ Η Μ Ιπ 00 0\
Masks Quarter Quarter
N O N Μ εt O Μ εt Γ` Ν εt If' Γ`
Theatre
1f' 'D
L L
ΑΑ
Quarter
-
Inopos
ν
υ ω
Α Α
Peribolos
f' 'f'
Street
0. ~ι
~~~~
ΑΑ
,-ι Ν
Quarter
Stadium
F F
νι υ7
~~ ~~
ΑΑ
Northern
Ο
zΝ
Quarter
-ι
zzzzzzzzzzz
00 ~\ ' .~ -1
~Ν ~Μ
-ι ΝΓ
η -ι ' Η φ
~~ υυυωωωωωυυυ
Α ΑΑΑΑΑΑΑΑΑΑΑ
Ε
~~ ~~
α Ο Ν
C) Ο
-y .--
µ
λ Ο
346
To tal
.~
η
0'
Sou thern
Ν
η
Zone
~1
Α
Sanctuary of
Area of the
Ap hrodite
νΓ
Ν.~ Ν Ν Μ 7 7 V' V• Λ.--ι Ν.--~ d' Μ
Ε Ε Ε Ε Ε Ε Ε Ε Ε- Ε Ε Ε Ε Ε Ε
ω ω ω ω ω ω '~~ω ω ω ω ω_ ω
~~ ω
ΑΑΑΑΑΑΑΑΑΑΑΑΑΑ Α
House of the
v, .- εη
~~ ~~ ~~
Α Α Α
Quarter
Inopos
Α ~
ν ~~
Α Α
Peribolos
-
Street
0.
~~
Α
Quarter
Stadium
ΕΕ Ε Ε
η η η η
~~~~~~ ~~
ΑΑ Α Α
Northern
Ν. -~-ι
Quarter
ΝΝ Ν N
Ζ ΖΖΖ Ζ. _Ζ Ζ..Ζ.g)
Λ1 ~1 GJ ~1 ω ω 'ω CJ g) -..
A AA AAA Α Α ΑΑ
O
O
Ο
O
ο
O
O O OO
O O O O
347
~
To tal
~
Southern
~
ν~Α
Zone
Sanctuary of
Area of the
Ap hrodite
~~
ωω
ΑΑ
ν>~
~
λ
Ο. -ι N ιΛ et η V' 00 O ΝΜ7υ1 ~,V' Λ00 ON ΜεΥη V' Λ0001ΟΜυ1
Masks Quarter Quarter
Theatre
~
^
ω
Α
~. ~.
Quarter
Inopos
^
~ `~
~a ~a ~~
AAA
~Λ
~
Η
RS
Peribolos
Ψ ~~
υ
~~ι ~ι i~ l i~ I ~
Street
~~
ωω î αω υ
~ ΑΑΑΑΝ
3
ο
ν
,-~
Ψ
Ψ
..ιΜ Λ
Quarter
Stadium
λ
Ψ
.`~.
h~~ Ε
C! Λ1 ~
~~
ΑΑΑ
Ψ ~
υ
~ b
ι.
ΙΙ, ~
Northern
Q ~ O .- -~ 7 υ1 V' I~ OO O M~h l~
Quarter
~~
Cλ ~
.~~ ÿ
ι.
σο
space
Ψ
,:ο
F
~
U
348 Total
ο
~
\ Μ
0Ο - 'c Ν
Sou thern
Ν
~
Zone
41
Α
Sanctuary of
Area of the
Ap hrodite
X χ χ χ - cχ' ι
n. d' Μ ~ ' ...~.......`-'
Ν Ν -
Masks ' Quarter Quarter
-- -- Ν
Theatre
Ν"
ir' r ‚· •
FFFF
g)g) ~α iα iα iα ωωωωωω ~ iα ~~
ΑΑΑΑ ΑΑΑΑΑΑΑΑΑΑΑΑΑ
House of the
ΧΧ
- -
13 13
η 1313
Α Α ΑΑ
~η
x
~
‚
Quarter
Μ~
Inopos
~~ ~~ Α Α
~Α
ν
χ
~.
Peribolos
-
Street
ν
~
(Ν Ν
FF
Quarter
Stadium
F
νι νι νι
~~ ~~~~
Α ΑΑ
c c
x x
Χt-«'
~
Northern
Quarter
N~V~ Ν NN
zzzzzz~
g)g)g)g)g)g)g)
zzzzzz
~~~~~~~
Circulation
, -o
gg Ο gg Ο gg Ο η~
λ •~
Ν
λ λ• ~ λ•~~ λ
space
ION ~~0ι ON ÿ ~
UU rι Uc Υ ' 0.
349
η
To tal
~ φ
---
N Μ Η.Μ C,
+ Vr Ν
Southern
Ν
Co
Zon e
41
Α
Sanctuary of
Area of the
Ap hrodite
~
ν ν
AA
η.
Ν Μ Λ Μ I Λ Vr Ο 00 Ν Μ yr 4, 11± Vr Λ V1 Ο
Masks Quarter Quarter
Theatre
.--~
(‚(~~ - 1•
FFFFE-FFFFE-~FFFF Ε-- E~E-i
ωωωωωωωυωωωωωω ' ω ~~~~î ~~ω ~~~~
ΑΑΑΑΑΑΑΑΑΑΑΑΑΑ ΑΑΑ GAA AA
House of the
~~ ~~
A Α
η.
Quarter
Inopos
~~~~ ν
AA Cλ
Peribolos
V'
Street
(J
4
n.
Quarter
Stadium
ΕΝ F ~~
CoCo Co
ω ~~ ~ι
AA Α
ο' Ν
Northern
Λ Μ ' Λ ΟΟ Λ ν^ Ο
Quarter
z (NV' z
~
z z z
350
Main room Northern Stadium Peribolos Inopos House of the Theatre Area of the Southern To tal
comp lex Quarter Quarter Street Quarter Masks Quarter Quarter Sanctuary of Zone
Ap hrodite
/
\
cc
\
/
οο
\\
°—
~ \\\
~ ~~
\ §
// /
~η ~
Ε--
°2®
/
æ'
\
/
\ \
‚'-
7
\
\
~~
~~
/α m
ηΒΒΒΒΒΒωω
/~ :
/
~~
« Ν -
\ \ \
.
\
~~
~/ /
Ii: main room north of the circulation space, two rooms at the rear.
11*: main room north of the circulation space, originally two rooms at the rear, but in the last
phase one room or three rooms.
12: main room north of the circulatín space, a room on either side.
13 • main room north of the circulation space, two rooms on one side or one room at the back and
one on the side.
14: main room north of the circulation space, a room at the rear.
15• main room north of the circulation space, a room on one side.
main room north of the circulation space, a room at the front.
main room north of the circulation space, without additional room(s).
18• main room north of the circulation space, three rooms at the rear.
lx• main room north of the circulation space, unknown arrangement.
El: main room east of the circulation space, two rooms at the rear.
B2: main room east of the circulation space, a room on either side.
main room east of the circulation space, a room at the rear.
main room east of the circulation space, a room on one side.
main room east of the circulation space, a room at the front.
main room east of the circulation space, without additional room(s).
Si: main room south of the circulation space, two rooms at the rear.
main room south of the circulation space, a room on one side.
main room south of the circulation space, a room at the front.
main room south of the circulation space, without additional room(s).
Sx: main room south of the circulation space, unknown arrangement.
Wl: main room west of the circulation space, two rooms at the rear.
Wl main room west of the circulation space, originally two rooms at the rear, but in the last
phase one room or three rooms.
main room west of the circulation space, a room at the back.
main room west of the circulation space, a room on one side.
W7: main room west of the circulation space, without additional room(s).
352
Dell! x off-centre x
x x off-centre x
De114 x central, wide x
x x al. central x
De117 x central, wide x
~ Cd
x x central, wide x
De1122 x sl. off-centre x
Cτ1
DelIS x x
,F . ._... .-. ,.,, . -•αa ry (D
x x al. central x
De1T21 x off-centre x
De1T23 χ 3 doors x
n
De1T24 x 2 doors x
φλ
De1T25 x ? central x
De1T31 x central, wide x
x sl. off-centre x
x sl. off-centre x
De1T45 x x al. central x x
x off-centre
De1S2 x central x
z
x central, wide' x
•v
Key to Table 5:
al = almost; sl. = slightly.
κ ι . χ κ χ ~. κ χ χ χ κ κ χ ~. κ χ κ κ κ κ . κκ χ
in front of
the façade
Portico
κ
Staircase
external
o
ο κ κ
Staircase
separate
internal
XXX κ χ
Staircase
internal
κ XX κ κ XXX ~. κκ
Pavement
us
κ κ κ κ x χ •~. κκ χ ~• κ χκ
decoration
Wa ll
χ κχ
us
κ κο~ κ κ κ
Pavem en t
κ κ κ χ κ κ κ x κ x χ κ Χ κ κ χ κ χ
81
decoration
Wa ll
+
κ κ χ χ '• κ χ χ κ χ κ χ κ χ κ χ χ κ κ χ κ χ
χ κ κ κ χ κ χ κ χ κ χ κ χ χ κ κ κ κ κ
Zí1 frοnt Of
the façade
Portico
κ κ χ κ κ κ κ κ χ κ
Staircase
ex ternal
κ κ κ x κ κ κ
Staircase
separate
internal
n. χ x κ
Staircase
internal
x κ κ x x κ η. κ κ κκ
Pavement
us
κ χχ κ χ κκ x κ κ κκ η. κη. η.
decoration
Wa ll
+
κ κκ
us
κ χ κ κ χ κ χ χ κ χ κκ κ
Pavemen t
κκχ κχχ
gf
χ κ κκ κ χκχχχ κ x χ
decoration
Wa ll
~~ ~~
κ ~ κκ
gf
κ κ κ κ χ κ χ κ χ κ χ κ χ κ
~~ .τ η ~ v~ n.ς
Cat. no.
ι ~
ςι η 1η ς ~~ ςM~ ς
G) ~) ~) L N N C) W W W L C/ WNWU W N U N U L W E) W L W C)
Α ΑΑΑ. ΑΑ ΑΑΑΑΑΑΑΑΑΑ Α Α ΑΑ Α -ΑΑΑΑΑΑΑΑΑΑΑ
Upper
storey
355
κ κ
Staircase
separate
internal
κ κ
Staircase
internal
κ κκ ' -Y
Pavem ent
κ κ
us
κκκ κκ κ κ κ
decoration
Wa ll
~~
us
κκκ χκ κ κχ κ κκ k ΝΧ ΧκΧ
Pavement
gf gf
κκ κ κ κ κ κκ κκκ κ
decoration
Wa ll
κ κκ κ χκ κ χ κ κ χ κ χ κ
Table 7. Delos: water supply and cistern types (fir key, see below the table)
Cat. no. Cistern Cistern Cistern Cistern Well
Type I Type II Type III Shape unknown
De1N1 2 1
De1N2
De1N3
De1N4 1 1
De1N5
De1N6 1 1
De1N7 1 2
De1N8 1
De1N9
De1N10 1
De1N11 1
De1N12 1 1
De1N13 1 1
De1N14 1
De1N15
De1N16
De1N17 1
De1N18
De1N19 1 1
De1N20
De1N21 1 1
De1N22 1 1
De1N23 1 1
De1N24
De1N25 1
De1N26
De1N27 2 3
De1N28
De1N29 1 1
De1ST1
De1ST2
De1ST3 1
De1ST4 1
De1ST5
De1ST6 ? ?
De1ST7 1
De1P1 1
DelP2 1
De1P3
De1P4
De1P5 1
De1P6
De1P7
De1P8
De1P9 1 1
Delll 1
De112 1
359
De1T38 1 ? ?
De1T39
De1Τ40 1
De1T41 1 1
De1T42 1 1
De1Τ43 1
De1T44 1 1
De1Τ45 1
De1T46 1 1
De1T47 1
De1T48' 2/0 1/3
De1T49
De1T50
DelAl 1 1
De1Α2 1 1
De1S1 1
De1S2
Total 52 2 1 ? >_ 61
Key to Table 7:
Type I: Quadrangular, vertical walls.
Type II: Circular, vertical walls.
Type III: Irregular.
Ν ~
Sou thern
'.
~
Zon e
ν
Α
Sanctuary of
Area of the
Ap hrodite
Ν ν
A Α
Quarter
Quarter
Inopos
~~
~~
AA
Periholos
η
Street
α~
ν
Α
~
Quarter
Stadium
~~
Α
Η
~~
z "z
Northern
~~
Quarter
zzzz zz " z
~ χ ~~~~~~ ~~ χ ~~
Ν 0η Ν '~
0 rt
. . .
ó x -ι Ν
εµ
361
us
Pavemen t
χ ~
~
~ υ υ Ο =ι_
gf tessera design gf types
O O ι- ~, ι+O n. Ο ~~
Ó Ó Ο υ Ο Ó ÓÓÓ ΟÓÓ O Ó ÓÓÓ
Mortar
Mortar
gf
Chip pavemen t
κ κ
gf v ermiculatum gf
Op us
κ κκ
tessellatum
Opus
κ κ κ κ κκκ κ κ
Ο ,- ,-ι
ι Ν M CU1V r r-.. Q0 If' Ο CO C
, ι ~ι .η ν ' η Ν CO α' ,-ι .-ι Γ-ι ,-ι Γ-ι ' . Ν Ν (‚Ι Ν Ν Ν Ν ('Ι rv F F F F
Cat. no.
ΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖΖ ~ ηηη
ω ω ω ω ω ω ω (J'
ω ω ω ω ω ω ~~
ω ω ω ω ω ~~ ωωωωωωω CII ω ω
ΑΑΑΑΑΑ Α ΑΑΑΑΑΑΑΑ Α ΑΑΑΑΑ Α ΑΑ Α Α Α ΑΑΑΑ Α Α
363
us
Pavement
~~
ο
gf tessera design gf types
~~ ~~
000 000 00 ~~ ~~ ~~
Mortar
Mortar
Chip paveme nt
κ
tessellatum
Opus
κ κ κκκ κ κ
ο
Ι
Α ο '- N Μ ~Υ V1 V~ Λ οο ο' Ο ti N Μ V' Ii' V• Λ 00 0 ' 0.-ι N M eh u' 'D ΟΟ 0\ O .-4 N
Cat. no.
α; -.-ι -- - - - . -i. -i rι N N N N N ('‚Ι N N N N MMN1 u' u' u' u' u' u' u' M
FFFFFHFFFFHFFFFFFFFFFFFFFFFFFFFFF
WWW Ε! L~4 Ú N L L W W ~~~
W W L W W W L U W W W i7 01~0
CJ W W W L
ΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑΑ
365
Table 10. Delos: sculptural finds (for key, see below the table)
Cat. no. Bronze Marble Terracotta Po rtrait
De1N1
De1N2
De1N3
De1N4
DellVS
De1N6
De1N7 x (1)
De1N8
De1N9
De1N10
De1N11 x (3) x (1)
De1N12
De1113 x (fragments) x (fragments)
De1N14
De1N15
De1N16
De1N17
De11V18
De1119° x (5) x (4)
De1N20
De1121 x (1) x (6) x (8)
De1122 x (1) x (4)
De1123 x (1) x (2) x (number ?) x (2)
De1124 x (1)
De1125 x (1)
De1N26
De1127 x (5) x (1)
De1N28
De1N29
De1ST1
De1ST2 x (1) base
De1ST3
De1ST4
De1ST5
De1ST6
De1ST7
De1P1* x (12)
De1P2* x (6)
DelP3 x (2)
De1P4
De1P5 x ( 2)
De1P6
De1P7
De1P8 x (1)
De1P9 x (1)
Delíl x (1)
DelI2 x ( ξ 30) x (3)
Dell3 x (2) x (3)
De1I4 x (2)
DeIIS
DelI6 x (1) base
De1M1
De1M2
De113 x (fragments)
De114 χ (4) x (8) x (1)
De1M5
De1M6
De1T1
Cat. no. Ex terior Exterior Ex terior Exterior Ex terior Exteior Exterior Ex terior Exterior
wall, left wall, right wall, third n iche, left niche, right n iche, third altar, left altar, right altar, third
s
ofen trance 0/en trance location of entrance of en trance location of en trance of en trance location
ω
i,
κ
κ
κ
XX
.D Λ,--ι Γ-ι
g)
XX
"
κ
#
1,
N
XX
,- ι Ν Ν Ν
-0<
g)
κ
X
κ
~
i<
Χ
Ν rv ro
o
XXX
XXX
ΖΖΖΖΖΖΖΖΖΖΖ
V• Λ Λ
ν
~
κ
^
Ο
κ κ
ix
0g)
b
ω
κ
Χ
κ κκ
κ κ
g)
κ
φ Ν(V Μ ~t
ΑΑΑΑΑΑΑΑΑΑΑ ~αΑΑΑΑΑΑ
κ
κ
g)
Ζιηνινινινιιη
κ
κ
'κ
~~.
t
1,
X
κ
XX
XX
κ
κ κκ
κ
κ
Α ΑΑ ΑΑΑΑΑΑ
Χ
~~
ω ω ω ω ω ω ω ω ω ω ~ ω ω υ ω ω ω ω ω ω ω ~~âι âι âι ~~
367
369
κ κ κκ κ
~~
Ε
Pavemen t
0 ~
~~ ο ~~ o ~ ~~
EE ~~~~~ E ~ E ~ E ~~~~~~ ~~ ~~~~ E ~~ Ε ~~~~~~
Wa ll
κ κ κ κ
~
~ κ κ κ
Cistern
3
Other
3 ι'-- ~
gg ~
~ ~ ÿ -o . ι-Q ,o ε-χ
ο
Circulation
0
~~
~~ ~~
αο α ~
ΜΜ Μ Ν
• -+ } }M1. rt.M1. + +
υυυ ~ νυυυυ υυ υ
Entrance
• •
ο ο
ο ο
.
ι^'^ • • ο ο
• • •
ο If'
ν̀ Ο
"O εr
•
ν
'-q ~
'—q η r rr MM ~~ .-η-ι Vr ~ ~ ΛΙ ~ Γ -ι
η
Ground-plan
•
' Ο ^d
Λ) Π
Α. Α. ,
~
χ ~••• • • • • ••~•
ΥΥΥ "° ΥΥΥ Q ΥΥ °° ΥΥΟ Qω ΟΥ Υ ΥΥαΥ
0 νν ι~ ~. ~αο~ ω
ννω ωωαινν0 ~. 0ω αιω αι
α ~~ τ~ ι~
~
Ο -r Ν
-
~~
-
Cat. no.
~c*C b '-'~C Κ
~ cΕ - ι0 Cl ;K --Κ c*3 ~
* ctl ~ Κ ~ -Κ ~ Κ Κ R ç- *
ιa 'C c~a 'C 'C
υυυυνυυ ~ υυυ ~ υυ υ υυ ~ υυνυ ~ υυ
Scu lpture 371
0
µ Ε
o
Pavem ent
~~
~
ο ο ο
types
o 00 Ε Ε
Ε
decoration
Wa ll
κκ κ
We ll
κ κ κ κ κ κ κ
Cistern
κ κ κκκκ κκ κ κκκκκκκκ κκ κ
Upper
storey
κ
0
33
Other
rooms
qq .qq
χ_ ;ο ;~ο χ ~
Main
room
κ
Circulation
space
υ υ υ υυ
C5
Entrance
room
κκ - Χ
ο ~
ο η
Ν
ς
Ground-p lan
Cat. no.
η 0Ο Γ+ M V v ' V' 1 » 00
•-ι 11' "0 η 00 Cλ '- ι -ι rι rι rι ti ~-ι.a.•-ι ~• - ι NNNNNNNNN
c~α ro ι~α m c~α κ *cα c~a c~a c~α c~a c~α c~a iα ro b c~α rs c~a c~a m
ro ro υυυυ
c~α c~a c~α c~a c~e
υ υυυυυυυ υυυυυυυυυυυυυυυυ
373
Car16 1
Car17 6 2
Car18 1
Car19 3'
Car20 6
Car21
Car22
Car23 x
Car24 1
Car25
Car26 1
38
Car27
Car28
Total > 80 4 >8 14
The number depends on how the boundary is drawn between CarΜ5 and CarM9.
2 Cf. note 1.
3 "...en forme de bouteille" (Fantar 1985, 12).
Table 15. Cartha ge: `Mago Quarter', Quarter of Decumanus Maximus and Cardo X: cisterns: location,
dimensions, date
Cat. no. Room Cistern Length Width Depth Capacity Date
(Type) (Type)
This room m ay belong to CarM5 or CarM9 depending on how the boundary between the house
units is reconstructed.
2 Cf. note 1.
~
Q)
κ
κ
υ
χ
χ
Μ
Fragment of cornice
λ",
κ
χ
Q
υ
b
Μ
Fragment of cornice
µ
κ
κ
ιΤι
χ
; ;
υ
b
Q
Μ
Fragment of cornice
,
xi
υ
κ
b
Q
Μ
Fragment ofdentilfrieze Ferron/Pínard1960- 1961, no. 3 13, pl. XLVIΙ
µ ,
κ
χκ χκ
b
υ
x
υ υ υ υ υ
Q
Two fragments ofdentilfrieze, traces o
~~ ~~ ~~ ~~ ~~
~~ ~~ ~~ ~~ ~~
µ
pigment κ
ὶri
χκ
υ
~
Q
υ
Fragment with traces of pigment
~~
~~
Fragments with traces of pigment and relief
377
378
design gf types us
o
~
Mortar tessera Pavemen t
o ~ ~ o
~.
Υ
s
κ κ ~~ κ
~-.οtι _
Μ
s
ΝΙ Ι ,< 1 555 ~
x
,-
~
κκ κ κ κκκκκκκκκκκκ κκ κ κκ κ κκκκ
`5υ
ν
s ~~~
α, Υ
ΟΝ κ
pavem en t gf
Chip
κ
.-ευ
Q' ~
κκ κ κ κκκ κ κ κ κ κκ
tessellatum gf verm iculatum gf
Opus
Υ
~~ ε4
s s sss ~~ ^~~
,
Opu s
κ κ κ ~. κ κκ κ κ
Γ, N
οΟλ '- Ηι-Ν-ι,-M-ι ~ NM V' 'ιπ V' ςι ~ι
Cat. no.
N 'ι' O Λ r" Q
§%
gf gf design gf types us
ω
pavemen t gf sectile
±4
tessellatum gf verm ículatum gf
~
\\
~ ωχ
ω
* *
II' 'D
3
00 D'
χ
χ
«
ω
lfg 'D-
00
ω
\\ \\\\\\\\\\\\ } }3222} } \
379
381
λ α ~ ι Ι 1 ,"~~ λ λ 11 χχχχχχχχχχ
W W 03 @ W W Φ Οι W ~0
~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ ~~ν υ υ υ υ υ υ υ υ υ
χΦ χΦ Φχ . χΦ χΦ χΦ χΦ χΦ χΦ χΦ Φχ
χ χχχχχχχχ χ
House
Φ uΦ
u υ υ υ υ u u u u υ u υΦ υΦ υΦυΦ υ
ΦυΦυ
ΦυΦυ
α
—
-op
~ - '-
ss ss 4s -5s '-
ss
Hy M M M M M M M M M M M "~ "p "~ ' n~ ' Λ~ ' ~
Find-spot
Ο! λ λ λ λ ι7 α α λ λ λ
µ ~~~~~~~~~~~~Α, R, α A,a, ¢,a,A,n,¢,
; Lλ Cλ QQ Q Ä Γλ Ä ÄQQ y µY YµY Y Y
~~ α α α α α ααα α
'r ' s- ô
φφφφ CO O'
Reference
~~ η ς ς ς ς ς ς ς ς ς ς
ο
η ~~ ~~ ~~ ~~ Ó Ó Ó Ó Ó ~~
s-
η Φ Φ cd Ν c~d λ̀ b Ν c~
α c~α b c~α c`~~ c~α λ̀ c~ö cα
µ c~d c~α Ν Κ
~ S Sι, 5 $ .5
ι-~-ι ~ ς Α ς ~ ~ ~ ~ ~ ς \α ~ ~ ς 0 ς Q 00000
tl ΟΟ ΟΟΟOOΟ 0 0 Ο O O O O Ο 0 0 0 0 0
a. s~.ι ~~ µ
τ ~~
51)51)
iµ. iµ
. 1. 4 4 s. i . s. ! ι. ι-ι 4 S~-ι II)L 4 4
s-s SS s
~~ ~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
µ µ µµ
µ~
µ µ µ µ
-~~ v
~ ν
µ ~
~ @ µ ν ~
'ω-
ν " ° ~~b
µ ~ φ
Object
~ ~ 2'- '
~~"~
Q s- 55 Q)
,_ς W µ χ µ µµ
~~ g
4's--s
~~ ~~~~
χ χ 4. 55 τ -ti t
'- t
OC ο~η b οη χιΛ ‚ ' 3' b ~~ ~~ α " ~α ο3
Q 0 ~~ '
ΦΦ 00 ΦΦΦφφΦφ 00 φΦ 00 Φ
χχχχχχχχχχχχχχχ
ει ει ει ει ει ει ει ει ει ~ ει ει ει ει ει
υυυυυυυυ
r
υυυυυυυ
τ. _4
τ. ε. ε. ε, 1 ε. ε. ε. τι τ. τ. ε. ε, S.
Ο 0 0 0 OO Ο 0 0 Ο Ο Ο 0 Ο 0
r ηιΛ. ι1ΛΡ ΛΡ
η.I ιIΛΡ ~ ~M
ιιΛ
g- ΛΡ Ν t— t-' t--' ι--
ι-'ι ηη-
ιτ µ µ
η.I εF~ µ ηyy ιη~-ι- 1 Fι ~- ιτ ιιτ
ει ει ει ει ει ει ει -τ. χ
ει χ
ει χ
ει sι τι χ
τι χ
ει
Φ Φ Φ Φ Φ Φ Φ Φ Φ Φ Φ c6 Φ c6 Φ
ζ.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.) ς.)
Η Η Η Η Η :
b -0 b b ~
. cd cςα Κ 0
cα ~ Η
O O O O O O O 0 O O O O O O o Η Η - Η Η
w Φ Ri - Φ Φ α Φ Φ Φ Φ w w w π$ α F. Η Η F.
Η
Η ς'-
.Si Ji -ς -ς ,- ς .-C, ,-4" .-ς '-O .rti ~. -ς '-Cι ~--i
Find-spot
w ti υ Ψ Ú) ν Ν \
fy
• τ • i ~ ' ~. sι i, ~, i, • . • ~ι • ~~
• i,+ z' 0 0 0 ~ Ó ~
υ υ ν υ υ υ υ υ υ υ υ υ υ υ υ-
eι eι αι υ
Η
αι αι Ο ν αι υ αι ν αι αι αι ς Γ. 0 Γ, ~ Γ, ω ~
0 0 0 0 0
• ON
0 0 ~~0 0 0 0 0 0 0 ON ON ΟΝ ON
Γ. •O Γ . ' O Γ. ~ Γ. .-ι <α
µ
-+ NC w w Υ ςςιι ηη Υ 4) 0
.Φ ~~ Ý ÿC Ν ..`α.~ .". ,.Φ+ ~ .-Φ+ ~~ ~~ Ν Γ'-ι ν Γ-.
Φ Φ Φ Φ w w w ~<à w
- ι '.ς Γq 'ς ' ς . -ς
Φ Φ Φ Φ Φ Φ
ς '.ς _λ'-ι '.ς
Ν µ
‚4 Ν µ Ν
Ο
ν µ Ν ~ ~~W
λ'" ' ς. L., » n
Γ.' 4-' Γ. n 4-
' Γ4'
λ .Γ.
ιn H m ~ vι H vι υ Q
ι-λ vι
m
Η Η Η Η Η Η
~
H
-- l ιΗ
— ι Η~ _ H
H _
_ H H '
~ ~ ~ ! ~ ~ ~ ~~ ^ ~
w Γ, Γ, Γ, ΓJ ' , ~
~ ~
Γ, Γ, Γ~, ...
!ÿ C~ jÿ C ~ jÿ C ÿC~ j C~ ~ ~G
~~j ÿ C~ jχ Φ Γ0
j λ ~~ ~ λ H~ j λ [~ÿ~ ~j υ -0 Ο
η
ÝY
λ
~γ Γηγ Γηχµ
λ ~ ~ !µ[ηµχ
~! Ν
γ ηγ ηµ_ γ γ γ γ {η ηγ γ, 7 V 7
ηµµ
1'Y µ χµ µ
1'Y ~1
, ~~ jµ ~
'_0
YY YY YY YY YY l'Y ~ ~ YY YY ✓Y YY ~ ~
ααα ααααα α αααααα ~, ~, ob ob ob ~
'-ς '-ς '-ς ' .-)
ó ~~ ~ ‚g;
Ν d' ιΠ O Ó N M V η Vr Λ Ν
η Vr Λ Φ
.Ν N N Ν
04-04-04- O Ο O O O O O Γλ
υ1 ι1 ,.ÿ --', .-~.
~
σ' ι
Π
η η Ν Γ, '~-- ι Γ•
Reference
0 0 0 0 0 0 O Ó ° Ο 0 0 0 0 Ó N .
~~ ~ . ~ Ο - Ν
Vr Vc Vr Vr Vr νΓ Vr V
Γ v̀
Πι ' ~ Vr Vr ~ Vr VJ O O O Ο Ο Ν
ι
ιΠ °~ Π ιΠ Π
~ Πι Πι Π
ι ιΠ ιΠ ιΠ ιΠ Πι ιΠ Λ 1~ Ο
Ο
'.0 0 '0.0 -0 "0'.0 -~~ '~~-0 -0 '0 -0 ΓΟ ,
ς
w w w
Φ Φ c~~ π4 w Φ
^
cα
~ '
Φ Φ Φ Φ
O
nt Φ,
C Cí
c0 Φ . Φ Φ
Γ!
Γ λ
ΓΡ '
λ
ΓΡ
Π ' '
λ
ΓΡ
η '
.~,
Λ " η " η " " ~
O C
η '
ΓΡ, λ 4'
λ λ
Γ' ΓΡ
λ
'
v1 ,~~
ηη ηη η Λη η
η η η ηη I η ηη
0 ς 0 0 O A O A A Q Q ς ς A ς- H H -H H H H
0 0 0 0 0 0 0 0 0 0 0 0 0 0 o α α α α α α
λ λ λ ~ λ λ
Ψ Ν υ Ν Ψ ν Ψ Ν Ο υ Ν Ν Ν υ Ψ λ λ λ ?ι . λ λ
ωωωωωωωωωωωωωωωω ω ω ω ω ~
µ µ
~. ÿ
. y
~ ~. µ
Object
~ Γ~ ~ µ
- ~ µ
~ - λ.
µ. ~ ~ -~ µ
-00 w
~ τ3 cï ~~ b
" 4 ,
~~ -
ωι ~ cα C ~
~~. ' -
~ οη ai ,~
Ο
εΦ.
w ια O ο εΦ, "d
c Ow ~~ Κ ~~ ~~ ~ " . O . ~
w w Ο
Vr Λ Φ σ . Ο '- N Μ
ci - c- w'w' ' Μ - V - - -
Table 20. Carthage. `Mago Quarter'. Sculpturalfinds of unc provenance, various materials
~ρ
~.
~
~
υ
Fragment of thigh, stone, stuccoed Kraus in Karthago I 199 1, 256, Taf. 65.h, 65.i Coordinate J/4, Roman fill to Punic cistern beneath
room P 100
Fragment ofleg, stone, stuccoed Kraus in Karthago 1199 1, 25 6, Taf. 65.g Coordinate JK/8, Punic destruction layer
,
Fragment of mask, Pan tc Kraus in Karthago 1 199 1, 25 6, Taf. 66. d Eastern part ofCardo XVIII
Fragment ofdrap ery, tc Kraus in Karthago 1199 1, 257, Taf. 67.h Excavation beneathCardo XVIIΙ, stratum p 2
Fragment of relief, frieze ofuraeiand Kraus in Karthago I199 1, 257, Taf. 67.g Punic destruction layer
winged solar disc, tc
Table 21. Carthage. Car la. Sculpturalfinds of unc ovenance, various materials
383
384
Scu lpture
.0 •0
~
Pavemen t
~~
α ~
0 ~ 0
0
~
0 0
~ ~
~~ ~ ~ ~~ ~ ~ ~~ ~ ~
IWa ll
κ κ
~ κ
I Cistern
κ κ κ κκ κ κκ κ κ κκ κ
Tipp er
η .
Γ4ΓΛ
Other
~
Η 3
~ ~~
° κ κ κ κ κκκ κ κ κκ
Circulation
ο
ο
X Χ
~~ υ ~ Fι QV υυ ¢ QVV υ ~ υ ~~
Entrance
κκ κκ κκκ κ κ κ
~~
~~ Μ ~
.- - ι ~ ~ Ν.-
η -ι ~~ Ν ~
Ν•- -ι ~ ;~~~ ~~ Ν ~~ ~~ΛI .-My .-
~ -i
Groun d-p lan
'-o ι.
~ Κ ~ ~ ~ Ψ at
Η
c~3 W cö ~bp Η
cC ~ ~bA ~
bA ~~-+ bA, ~: .-~~ bA
~" µΨ ~" Ψ ~' µµΝ µ ΨΨ NN Ψ N
ι. ι. .
µε.Ψ
Ε5
' ς ' 1: i Q Q Q Q ~i Q Qi Qi ~i Q Qi Q Qi Qi Qi Qi Qi Q~i QQQ
Scu lpture 385
° ~~ ~~
Pavement
α
~~ ~~ ~~ ~ ¢ ο •
~
00
Wa ll
κ κκ
~
m
~ κ κ
Cistern
κ κ κ κκ κκ κ
Upper
(. κ κ
Other
κ
~
z zκ ~~
κ κ
ûα ûα
κκ
z
(.
Circulation
0
~χ ~° ο
Porticoes on three sides and a well on the fourth side.
Entrance
jas: main features of the houses
κ κ κκ κκ κ
Μ
~ ~
Φ Ó
η
ti ~ íN
Λ 00
N
00 ~'
Μ ΛI Φ ~ ΛΙΜ VΝ
Ground-plan
κ m
ςα
c°µ ου „ 0000
~ α υ υ υ υ ) υ
ΜR
Cat. no.
~. ~. ~ , ~ , ~ , ~~~
ξ
¢ Q ¢¢¢¢¢ ¢¢ ¢
387
Ampl1
Amρl2 1
Amρl3 1
Amρl4
Ampl5 1
AmρΝ~~
AmρΝ7 1
AmpΝ8 1
ΑmpΝ9 .
AmpliO
Amph 1
AmρΝΙ2
AmpΝ13 1
AmpΝ14
AmpΝ15
AmpΝ16 1
AmpΝ17 1
AmpΝ18
AmpΝ19 1
AmρΝ20
AmρΝ21 1 1
Amρ122
AmρΝ23
Amρ124
Amρ125 ΙΙ
AmρΝ26 1
AmρΝ27
ΑmρΝ28 1
Amρ129 2
AmρΝ30
Amρl31 1 1
Amρ132 1
Amρl33
Amρ134 1 1
AmpRi 5 1 3
AmpR2 1
AmpR3 1 1 1
AmρR4 1
Total 6 1 1 23 1 3
Table 25. Ampurias: cisterns: location, dimensions, date (for key, see below the table)
Cat. no. Room Cistern Length Width Depth Capacity Date
(Type) (Type)
Ampli
ΑmρΝ2 3 (C ?) 3( IV) 3.66 m 0.92 m 3.8 m 12.1 m3
ΑmρΝ3 4 (C) 38 (IV) 3.2 m 1.7 m 0.2 m
(pres.)
ΑmρΝ4
ΑmρΝS 1(A)-5 (mr) 16 ( TV ) 5.1 m 1.14 m 4.3 m 21.9 m3
ΑmρΝ6
ΑmρΝ7 1 (A) 15 (I) 4.5 m 1.69 m 2.75 m 20.5 m3 Older th an
present house
ΑmρΝ8 2 (C) 14 (IV) 4.84 m 1.14 m 3.04 m 15.6 m3
ΑmρΝ9 .
AmpliO
Amphi
ΑmρΝ12
~mpN13 2 ( Α )-3 17 (IV ) 5.02 m 1.1 m 3.3 m 17.3 m3
ΑmρΝ14
Ampi"
ΑmρΝ16 1 (C) 20 (IV) 3.6 m 0.9 m 1.9 m 5.8 m3 2 "α
(pres.)
ΑmρΝ17 1 (C) 19 (IV) 4.16 m 1.16 m 1.96 m 8.8 m3
ΑmρΝ18
ΑmρΝ19 1 (C) 21 ( IV ) 2.98 m 1.08 m 3m 8.9 m3
ΑmρΝ20
ΑmρΝ21 2 (P) 22 (IV) 4.5 m 1.1 m 1.56 m 9.4 m3
ΑmρΝ22
ΑmρΝ23
ΑmρΝ24
~mρΝ25 1 (C) 25 ( IV) 14.4 m 1.6 m 3.3 m 22.4 m3
ΑmρΝ26 1-2 (mr) 24 ( IV) 4.66 m 1.66 m 2.76 m 18.4 m3
ΑmρΝ27
~mpN28 1 (C ?) 30 ( IV) 3.24 m 1.12 m 3.74 m 12.5 m3
~mpN29 1 28 ( IV) c. 5 m 1.5 m
1 29 ( IV) c. 5 m 1.5 m
~mpN30
ΑmρΝ31 1(A) 40 (1V) 4.2 m 1.1 m 3.1# m 13.5 m3
ΑmρΝ32 1 (C)- 34 ( LV) 8.1 m 1.06 m 3.2# m 26.7m3
2 (mr)
ΑmρΝ33
ΑmρΝ34 1(P) 9 ( IV ) 4m 1m 2.08 m 8.1 m3
389
ο ο ο
υ υ υ
+ + +
κ κ κ κ κ χ κ κ
q
ο
.~
η,
~
Mortar
+
κ κ κ κ κ κ κ
gf sectile gf
Op us
verm iculatum gf pavemen t
Op us Chip
inset pieces gf
tesse llatum gf tessellatum
Op us Opus
-o
κ κ
Table 28. Terms used for pavement types: Delos, Carthage, Ampurias
Delos Carthage Ampurias
Mosaic pavements Opus tessellatum Figlinum tessellatum Opus scutulatum
Opus vermicualtum Opus figlinum Opus tessellatum
Pavements de galets nu- Opus segmentatum Opus vermicualtum
lés Opus tessellatúm Scutulata
? Lithostroton
Opus scutulatum
Opus segmentatum
Scutulatum
LIST OF FIGURES