Since 1996 the German Archaeological Institute and the American Academy in Rome have been conducting a joint urbanism project on the unexcavated parts of Rome's port city. The combined use of geophysical surveys of large areas, systematic analysis of aerial photographs, and selected stratigraphic sondages has not only complemented the previously known plan of the city but also brought much new information on the urbanistic development of previously unknown sectors. One of the most important results of the 2000 and 2001 seasons is the proof that a harbor basin existed just inside the ancient mouth of the Tiber. On the E side of that basin we investigated an unusual structure: a large terraced construction the vaulted substructure of which seems to have served in part as shipsheds and in part as storage and commercial space, and, above, a marble temple, oriented toward the mouth of the Tiber and surrounded by porticos.

In the Baths of Mithras at Ostia, a lead pipe from the public urinal carried fluids directly into a basement corridor which led to two small underground fullonicae (figs. 1-4). As must have happened in towns and cities all over the Roman world, this product of human excretion was flushed down the urinal to re-emerge as the quintessential industrial cleansing agent. The Roman fuller has achieved notoriety for his exploitation of urine for washing woollen cloth. In this paper, I intend first to attempt a definition of fulling, and to show that the process of identifying and reconstructing a fullonica requires us to think harder about Roman cleansing processes. I will argue that the topic of cleanliness is so culturally loaded that it is very difficult to reach a neutral account of fulling. Literary discourse on these processes and their agents offers us a set of contrasting responses, most notably in interpretations of urine. I will examine the ways in which the Romans played with some of these paradoxes in a world of limited chemistry. From this, I will suggest a topographical model of water and waste in which the fullonica was a significant unit, and examine how the proverbial smells it generated raise interesting archaeological questions about location and urban space. A final section addresses the social profile of fullers and the cultural stereotypes attached to this profession.

Chi avesse percorso la Sacra via tra gli ultimi decenni del I sec. a.C. e l'estate del 64 d.C. si sarebbe trovato in un profano e lussuoso quartiere commerciale: ori, argenti, perle e pietre preziose, aromi e spezie, primizie e cibi ricercati, strumenti musicali, libri, corone di fiori, cortigiane e prostitute — tutte le delizie della vita (come già ai Greci era piaciuto immaginarle) fornite dalla plebe urbana che qui gestiva il sistema di vendite al minuto più famoso e organizzato della città. Anche nelle altre zone intorno al Foro lo scenario non sarebbe stato diverso; nelle botteghe della piazza, lungo le strade circostanti e nei vicini edifici continuava il commercio di dispendiose mercanzie per una clientela di gusti ambiziosi e grandi disponibilità. La tradizione commerciale del luogo era molto antica e risaliva all'età dei re; stando alla tradizione, Tarquinio Prisco aveva per primo utilizzato la valle per attività pubbliche, facendo costruire appositi vani per artigianato e commercio (tabernae circa forum). Quando alla fine del VI sec. a.C. si impiantarono i quartieri della Sacra via, sul fronte degli isolati furono costruite una serie di botteghe tra le quali si aprivano gli ingressi alle domus. Nel periodo più remoto era lo smercio di prodotti alimentari, soprattutto le carni, l'attività primaria forse anche per la presenza di un mercato del bestiame collegato al Foro Boario (beccherie e beccai sono ricordati dal V sec. a.C). Fu negli ultimi decenni del IV sec. a.C. che un aspetto più conveniente (forensis dignitas) si sarebbe affermato per gli interventi di C. Maenius e per la trasformazione delle botteghe da lanienae in argentariae, da macellerie cioè in banchi di cambiavalute, usurai e banchieri, destinati a dominare la piazza fin agli inizi dell'Impero. I commerci più ordinari o le rivendite specializzate non dovettero scomparire del tutto almeno fino alla metà del II sec. a.C: alle tabernae argentariae si mescolavano i negozi di beccai e speziali, chiamati alla greca myropolae, i primi ricordati da Plauto e da Livio nel luogo della basilica Sempronia, i secondi soltanto da Plauto.

En los últimos decenios del s. I a.C. acaeció en Carthago Nova un intenso proceso de renovación urbanístico-arquitectónica, que probablemente encuentre su más remota génesis en los años posteriores a la fundación de la colonia y tras el viaje de inspección de César y Octaviano del año 45 a.C, momento en que el enclave asumió el nombre de lulia, pasando a denominarse colonia Urbs lulia Nova Carthago. El proceso afectó buena parte de la ciudad y encuentra sus más elocuentes expresiones a nivel arquitectónico en los nuevos proyectos y equipamientos monumentales alzados en sus sectores centro-occidentales, tales como el teatro, parte de los edificios del foro y, quizá, el anfiteatro. Sin embargo, el proyecto de monumentalización del centro urbano debió perdurar algunos decenios, prosiguiendo en las primeras décadas del s. I d.C. Buena muestra de ello son las estructuras, muy arrasadas, del edificio central de una composición arquitectónica de carácter monumental constatada en la calle Caballero; la excavación arqueológica ha posibilitado la recuperación casi íntegra de la planimetría del referido edificio que constaba de varias estancias (básicamente un vestíbulo y una sala interna) y debía estar emplazado al fondo de un pequeño atrio o patio, cuyos caracteres morfológicos y arquitectónicos son difíciles de precisar. El complejo se ubicaba en las inmediaciones de la plaza foral de la colonia, si bien su alineación respecto a ella y la presumible orientación hacia el NW de sus estructuras de acceso son indicios que no debía abrirse directamente a la plaza sino, más probablemente, a una presumible arteria de comunicación que desembocaría en ella de forma tangencial a su eje axial.

Demography has long been an essential ingredient of economic history. Students of the “ancient economy”, by contrast, have been late to give demography its due weight, and attempts to illustrate the potential relevance of population issues have been rare.1 This case-study of Roman Egypt aims to interpret empirical evidence of economic change with reference to demographic factors. I will argue that in the late 2nd c. A.D. a severe mortality crisis triggered price and wage shocks, and that during the following century the resultant population loss contributed to a decrease in the return on land and to a rise in the real wages of workers. I must stress at the outset that my model is deductive in so far as it predicts specific developments based on the internal logic of economic and demographic relationships as illustrated or corroborated by comparative evidence from other periods, and also in that it seeks to situate and explain disparate samples of empirical data within a preconceived unifying interpretative framework. In this it is my goal to provide the most economical and internally consistent explanation for the largest possible amount of the available data. No explanatory model can ever be “complete” or even “correct” to the extent that it would accommodate every single artifact of historical information, eliminate the need for complementary explanations, or fully disentangle the complexity of historical events; rather, it needs to be judged in terms of whether it exceeds (actual or potential) comparably comprehensive alternative models in its capacity to interpret and explain the evidence in a logically coherent and historically plausible fashion.

In the foregoing article, W. Scheidel builds on earlier work, most notably that of R. P. Duncan-Jones in JRA 9, to offer a model for the predicted effects of the Antonine plague and to argue that the model fits the evidence from Roman Egypt reasonably well within the limits of the quantity and quality of the latter. In his second footnote, he encourages critical response, suggesting that it “may either corroborate or undermine my interpretation.” The following pages are intended as a contribution to that discussion, but with lesser ambitions than either corroborating or undermining the model as a whole. They offer some of both, in fact, but more in the direction of undermining it.

There are three reasons for not claiming too much at this point and not offering any general conclusion (as I do not). The first is that I do not have any fixed views on the degree to which the plague was the prime mover behind the changes visible in late 2nd- and 3rd-c. Egypt. In the absence of any concerted attempt to formulate and test other hypotheses about the engines of social and economic change, it is hard to say if the degree of fit of evidence to model is impressive or not. The most obvious counter-candidate is the increased municipalization of Egypt during just this period, especially from A.D. 200 onward. It would be useful to generate a model of economic change from this force and see if it is equally capable of accounting for the evidence.

So-called magical gems constitute an especially rich body of material evidence for magic and religion in the Roman Empire. They differ from the ordinary run of gems in three respects: in their selection of iconographic types, normally divine images of one sort or another; by their use of magic words and occasionally longer texts, primarily in Greek script; and by their use of magic signs, usually called characteres. At least one of these three elements must be present for a gem to be identifiable as magical. These “Zaubergemmen” form the most easily distinguishable sub-group of the wider class of amuletic gems, that is, engraved stones of talismanic function. The majority of the iconographic schemes appearing on magical gems adhere closely to the classical Graeco-Roman and Egyptian traditions. Others, however, are unique to this class of gems: rare even on other magical objects, they are practically unknown outside this sphere in the whole variety of ancient art.

Divination was one of the most important features of the learned magical arts in the Imperial period. Not only do the Graeco-Egyptian ‘magical papyri’ contain an abundance of recipes which claim to enable the practitioner to know the future, but several ancient authors attest that divination was of special interest to occultists. Recent scholarship has indeed recognised the importance of divination in ritual-magical practice, but the relevant archaeological evidence has not been much discussed since the publication of the second volume of Th. Hopfner's Griechisch-ägyptischer Offenbarungszauber in 1924. The major new evidence here has been the Near-Eastern divination- and incantation-bowls. The present article, however, is concerned with the possible implications of a much older find, the divination kit from Pergamon, and its recently-discovered analogue from Apamea in Syria, for the study of specifically theurgic divination. The rôle of magical ritual within theurgy has received considerable attention in recent years, but the relevance of the divination kits has not hitherto been noticed. I shall argue that the physical instruments employed in theurgic divination help us to understand several features of theurgic practice. I shall also stress the possible contribution of magical gems in the same context, for in them we can recognise images and attributes of divine beings with whom magicians and theurgists identified themselves during their performances.

It is very welcome that A. Mastrocinque (above) has re-opened the dossier relating to the Pergamon Zaubergerät, first published by R. Wünsch almost a century ago. The discovery in 1977 of a very similar ‘triangle’ in the Maison du Cerf at Apamea seemed to confirm Wünsch's account of the kit as the equipment required to perform a type of divination very similar to that described in Hilarius' confession relating to the seance of A.D. 371. My remarks are not directly concerned with the possible theurgic background of the Pergamon kit, though I admit I am rather sceptical of it. I wish rather to take the opportunity to reconsider the kit as a whole, in the light of the find from Apamea.

The problem is easily outlined. Rhetorically, Wünsch's commentary leads up to a kind of revelation, the disclosure of the true sense of the kit as whole. In the last major section of his account, Wünsch discussed each object in turn, noting the use of similar objects in diverse magical contexts. For his interpretation the crucial apparatus was the inscribed disc, which has 24 fields in the three outer circles, that is, the number of letters in the Greek alphabet. That implied an alphabet-oracle, and it was then easy to point to Ammianus' already famous report. It is his story of the ‘wizard at work’ that caught the imagination of his readers, the story of the polished stones used as protective amulets, the ring hung from the nail over the circular disc, which was moved by the handle to create words or sentences from the signs inscribed on its surface. But if one looks closely at the disc, it is very difficult (indeed, in my view impossible) to credit that it could have served as an alphabet-oracle or anything similar. If so, does the disc belong to the triangular support at all? Can the other appliances be understood differently from the way Wünsch suggested? My argument is that we might read much of his own commentary as undercutting the final disclosure that depends so heavily on Hilarius, and that we should revert to his own initial conception of an ensemble, a group of instruments with a variety of ritual uses. Indeed, there are reasons for thinking that the individual items were not conceived as a group, but rather assembled over time from various sources as a collection. I incline to understand the ensemble as not so much a ‘kit’ as a rag-bag collection.

The Roman and Byzantine port of Butrint, situated on the SW coast of Albania directly opposite the island of Corfu, has been the focus of a major research project since 1994. The investigation of the site and its hinterland commenced with excavations within the walled town and a survey of sites and monuments in the region (Hodges et al 1997). Despite a brief hiatus caused by civil unrest in Albania in 1997, work continued with excavation and study seasons in 1998 and 1999. The results of the first five years of the project are due to be published shortly (Hodges, Bowden and Lako, forthcoming).

The second phase of the Butrint project, starting in 2000, has encompassed a wide variety of research aims. They have included extensive research on the archives of the Italian mission that conducted large-scale excavations between 1928 and 1942. Among other finds, this resulted in the discovery of the manuscript of L. Ugolini's Albania antica vol. 4, the hitherto unpublished results of the Italian excavation of the Hellenistic and Roman theatre (Gilkes, forthcoming). Other archival research has focussed on the records of the communist-period archaeological investigations, and has resulted in a much better understanding of the aims and results of these projects, which in some cases are almost wholly unpublished. Our project is also concerned with the controlled development of the site for tourism. This has resulted in the expansion of the UNESCO World Heritage site to 2900 ha and the creation of a National Park with the intent of protecting the archaeological and natural landscape around Butrint (Hodges and Martin 2000; Martin 2001). The present report is a synthesis of the first results of the major excavations of 2000 and 2001. While it is possible (and indeed likely) that interpretations may change as excavations continue, it was felt that the material was of sufficient interest to justify an interim statement.

Ancient Halmyris lies in the NW corner of the Dobrudja region in SE Romania. It lies c.2.5 km east of the village of Murighiol on a rocky promontory which is slightly higher than the surrounding marshes. This is at the E end of the Dunavat peninsula (known in antiquity as Extrema Scythiae Minoris: Jord., Get. 266) and it is bordered by the Danube delta on the north and east, Razelm lake on the south, and the Tulcea hills on the west (fig. 1). The site was occupied continuously from at least the mid-first millennium B.C. up to the 7th c. A.D. The local environment, flora and fauna were favourable to settlement until as a result of natural causes the Danube became almost inaccessible; from that point on, the settlement became vulnerable to human and other natural events and eventually it became deserted.

The site is known today as Bataraia or Cetatea. In the early 20th c. the locals still called it the Genoese stronghold (Geneviz-Kaleh). In antiquity it lay on the bank of the southern arm of the Danube called Peuce (now known as Sfantu Gheorghe). Today the southern arm of the Danube runs 2 km north of the site and it is connected to Lake Murighiol by the Periboina canal. Until 1983 there were two lakes, c.100 and c.200 m from the site, modern relics of the ancient course of the river. To the east lie the Dunavat hills and to the south is Dealul Cetatea (“fort hill”) (fig. 2).

During the last few decades most landscape archaeologists have noted the diffusion and the demographic importance of the rural landscapes of Archaic Etruscan communities and have tried to define their significance within Etruscan society in the same way as others have attempted to evaluate the political significance of the Greek rural landscape. Recent research on Italian landscapes has led to a great increase in the available data regarding the different paths of development for the various communities, allowing them to be outlined and compared.

The growing dichotomy between the studies of field archaeologists and historians or art-historians may appear to be a problem. Landscape studies in Italy have been dominated since the 1950s by an Anglocentric tradition of economic and environmental archaeology, with important work focusing on long-term phenomena. Historians and art-historians, on the other hand, have tried to define an interdisciplinary approach involving the use of several sources of evidence (art-historical, epigraphic, literary) and focusing on historical events and medium-or short-term phenomena. Yet field and historical archaeology are simply two sides of the same coin, and should be viewed as complementary rather than incompatible approaches to understanding the comolex evidence of the Dre-Roman cultures.

This is one of the rarest types of Domitian's coinage, struck somewhere between September 95 and September 96, as shown by the dating of the tribunician power. The issue of TR P XV was known to Cohen only from a piece in trade (Rollin), and to Mattingly in BMCRE only from Cohen. Since the publication of BMCRE vol. 2, an example has been acquired by the British Museum and a few others have appeared on the market. To this day there is no example of it in the extensive collection of the Bibliothèque Nationale, Paris. As to TR P XVI, no-one knew that the type had been struck up to that date until an example surfaced in the Mazzini collection, the only one ever to have been seen.

The type was obviously struck in very small quantities, given the few examples that survive. The issues of TR P XV and TR P XVI were struck from the same reverse die, which suggests that the first was probably produced late in the tribunician year, say late August to mid-September of 96. The TR P XVI issue was probably struck for less than a week, Domitian's dies imperii falling on 14 September and his assassination on the 18th. It is reasonable to suppose, too, that the rarity of the latest coinage reflects the murder: any stocks of Domitian's coin still held at the mint or in the Aerarium would have been reduced to raw metal and re-used for Nerva.

In 1904, a mosaic representing a hunt scene was found in House B in the Piazza della Vittoria in the centre of modern Palermo. The results of the excavations were published in 1921 by E. Gabriel. In 1931 H. Fuhrmann included an analysis and reconstruction of the mosaic in his book on the painter Philoxenos of Eretria. Work on both the context of the house and the mosaic has continued, confirming that House B was a richly-decorated Hellenistic house situated close to the E wall of the old town of Panormus. In this article I will re-assess Fuhrmann's reconstruction and subsequent interpretations. After describing the mosaic in detail, I will present and explain a new reconstruction.

The Hunt mosaic was found set centrally in a large room (r) to the north of the peristyle (figs. 1 and 6). The main picture of the mosaic (205 × 174 cm) is framed by two bands, the inner black and the outer white. An intricate floral border comes next, then a band of white tesserae set at a 45° angle, and an outer frame consisting of two parallel black bands with a white one running through the middle. An adjusting border of white tesserae, also set at a 45° angle, runs up to the base of the walls.

The future emperor Galerius was born in a small village in the province of Dacia Ripensis around A.D. 258. His mother's name was Romula. When he was confirmed as a member of the tetrarchy he undertook a building programme in his native village, transforming it into a monumental fortified palace. To honour his mother he gave it a new name, Felix Romuliana. In 311, before this ambitious programme was complete, he died and was buried and deified there.

Romuliana has been securely identified through epigraphic evidence with the impressive remains that exist near modern Gamzigrad (NE Serbia). This fortified palace was built in two phases: first came a fortification wall with 16 small rectangular or octagonal towers, dated in the first five years of the 4th c; then in c.305 this defensive system was partly demolished and a new, bigger fortification wall, strengthened by 20 huge, bastion-like towers (16-, 12- or 10-sided) was erected on the outside, encompassing and enlarging the original plan. This wall has been dated to 306-311. Inside the walls were erected two temples with altars, two palace complexes, a bath, a horreum, and auxiliary buildings. On a low ridge of hills east of the fortified palace two large consecration mounds and two mausolea were excavated in 1993. From the finds as well as the overall nature of the buildings at Romuliana, it was determined that this was where Galerius and his mother were buried and deified in the first years of the 4th c.

Lors des fouilles conduites par l'Ecole suisse en 1998 dans le secteur E/600 NW à Erétrie (Eu-bée), nous avons découvert un four présentant un aménagement peu courant. A double paroi, double bouche et banquette interne, sa configuration était pour le moins inhabituelle (fig. 1-2). En l'absence de toute trace de matériel de cuisson, seule la recherche de parallèles dans les différentes catégories de fours pouvait apporter des informations sur sa fonction initiale. Ma recherche s'est alors orientée principalement vers les fours céramiques et les fours à chaux. En effet, les autres structures analogues, destinées à la métallurgie, la verrerie ou à l'usage domestique, diffèrent sensiblement tant par la forme, la taille et le fonctionnement que par les déchets produits par la fonte du métal ou du verre.

A la différence des fours de potier, qui ont suscité un certain nombre de recherches, liées en partie à l'étude des provenances des céramiques locales, les fours à chaux restent peu étudiés. Ils ne produisent, il est vrai, aucune oeuvre d'art. En revanche, les chaufourniers utilisent parfois des marbres sculptés pour obtenir de la chaux.

Between 1993 and 2001 a British team led by S. Esmonde Geary, M. J. Jones and the author examined the Late-Roman defences of the ‘ville haute’ of Saint-Bertrand-de-Comminges (SW France). The project fell within the overall theme of studying the transition from the classical to the late antique/early mediaeval town, a principal objective of the international Trojet Collectif de Recherches’ at Saint-Bertrand. The primary aim of the British investigation was to document and analyse the construction of the Late-Roman defences and their subsequent development through a combination of architectural survey and excavation. During the nine seasons of fieldwork, the architectural remains of the entire wall circuit were analysed and 11 separate trenches excavated. The evidence obtained from these excavations dates the wall's construction to the early years of the 5th c.

The architectural survey included collating old photographs and unpublished excavation records; preparing a plan showing the surviving original and rebuilt stretches of the walls; making a general survey of the principal external and internal elevations, and recording the outline of all visible Roman facing and corework, vertical and horizontal breaks, offsets, tile courses, drains, re-used masonry and later building and repairs; making stone-by-stone drawings of the best surviving elevations and features; making a detailed analysis of the wall fabric, interpreting its building periods and phases of construction, and identifying changes in alignment of the defences, the presence of external towers, work-gang divisions, and so on. For ease of reference, the circuit was divided into 26 sectors on the basis of criteria such as change of alignment and state of preservation.

Le front oriental de l'enceinte intérieure de la Cité de Carcassonne (fig. 1) conserve plusieurs tours remontant à la fin de l'Antiquité. Celle située au sud de la porte Narbonnaise, accès principal au site, tire son nom de l'intégration de son étage, au Moyen Âge, à l'église Saint-Sernin. A cette époque, le mur de la tour regardant vers l'intérieur de la ville fut démoli, le sol intérieur surcreusé, et l'arrondi concave de la pièce transformé en l'extrémité orientale de l'espace cultuel. Un agrandissement de la fenêtre centrale fut autorisé par une lettre patente du roi Charles VII en l'année 1441. Après la désaffectation de l'église sous la Révolution, la nef, les collatéraux et le clocher ne furent pas conservés. Dans le cadre de la longue campagne de restauration de la Cité entreprise au milieu du XIXe s., les vestiges de la tour firent l'objet de travaux tardivement, vers 1889, sous la direction de Paul Boeswillwald chargé du chantier depuis le décès d'Eugène Viollet-le-Duc. Le restaurateur refit une partie des parements intérieurs, ajouta de nouveaux moellons disposés en arrachement à l'ouest, et reprit toute la limite extérieure du couvrement (fig. 2). Sous la voûte médiévale, on remarque encore deux fenêtres de la fin de l'Antiquité, celles qui sont mentionnées dans la lettre royale. Leur arc à profil demi-circulaire est formé de briques. A mi-hauteur de leurs montants, moellons et briques alternent régulièrement.