So dominant has Hadrian's sprawling villa near Tivoli been in our imagining of the rural retreats of Roman emperors that it is easy to forget that some 30 such sites lie within the orbit of Rome. Little was known of Villa Magna, 60km south-east of Rome, before the meticulous research described in the present volume was launched in 2006. It is one of those rare Roman sites where literary sources can throw direct and illuminating light on the archaeology. In the early 140s AD, the young Marcus Aurelius wrote three letters to his teacher, Fronto, mentioning that he was staying with the imperial court (it was the early years of Antoninus Pius’ reign) at the time of the autumn vintage in a villa near Anagni, and that they ‘dined in the pressing room’, where they happily heard the banter of the ‘rustics’ treading the grapes. Extraordinarily, the excavators at Villa Magna (the name is recorded on a marble inscription, reused in the later church on the site) found, through a stroke of luck, the precise location of this scene. An emperor dining ‘in’ a pressing room sounds improbable, but the location of a bath-suite and associated rooms sumptuously appointed in marble (including a large exedra on the central axis with a view across a room of sunken wine storage jars, or dolia, to a treading floor) matches perfectly with Marcus Aurelius’ description. There was no trace of dining couches (which may well have been movable furniture, not a permanent fixture), nor would the space have ever been interpreted as a dining area were it not for the literary passage; but the excavators’ conclusion must surely be right. It throws interesting light on the Emperor's ‘theatrical’ use of agricultural production as propaganda; and the presence of imported polychrome marbles from Chios and Sparta, used for paving even in the floor of the dolia room, provides a touch of luxury unique in Roman wineries.

Much else was learnt of Villa Magna in the course of this work, thanks to geophysical research, field survey and excavation. Overall, the Roman villa covers 6.2ha (less than 6 per cent of the size of Tivoli's 107ha): brick stamps show that it was first completed in its present form by the ever busy Hadrian. The main residential palace, joined by a road to the winery, lies some 250m to the north. Its outline is known from geophysics, but it was barely touched in the project described here. Under the Emperor Severus, in the early third century AD, infrastructure was improved with the addition of a showy fountain (nymphaeum), a large bathhouse, additional cisterns and a two-storey, barrack-like building, situated on the approach road on the eastern side of the complex. While quarters for a military guard are not out of place in imperial villas (e.g. Cercadilla in Spain, Mediana in Serbia), the poor amenities (earth floors) and the small finds (showing the presence of women, and with no hint of the military at all) suggest that this building is rare evidence for the slaves’ or attendants’ quarters on a great landed estate. The discovery of seven child burials under the floors, while attested at Roman villas in other regions (e.g. Hambleden in England), is unique for Italy; it too points to a civilian use for the building. The children all suffered from anaemic thalassaemia, suggesting the presence of families brought in from a coastal region: they were not locals.

How long Villa Magna enjoyed favour with successive Roman emperors is unknown (future work in the palace quarters might show how long it was maintained as an elite residence), but what the excavations did demonstrate very clearly was the continuing functioning of the place as a working estate and its gradual evolution over a longue durée. The baths by the winery went out of use towards the end of the second century, and by AD 250, at the latest, the winery itself had also been dismantled. A freestanding basilical hall of uncertain function was built in the open space in front of the palace, perhaps in the early fourth century AD. By the sixth century this had become a church; but a wide range of imported amphorae, carrying above all wine, was still reaching the site then, not only from southern Italy and Sicily, but also from the Greek islands of Samos and Kos, and from western and southern Turkey—unusually for a site so far from the coast. They imply a level of comfortable living still being enjoyed in early Byzantine times by at least some occupants of the estate.

No less significant a contribution has been made by the Villa Magna project to early medieval archaeology, also recorded in admirably full detail. Furthermore, surviving documentary evidence for the estate's management at this time supplements and helps contextualise the archaeological discoveries. A rare example of an early medieval stone-built villa was constructed on top of the Roman winery in the eighth or ninth century AD. The church was refurbished and upgraded with decorated chancel screens at the same time. By 976 (when a document records the donation of lands), an adjacent monastery had been founded, and part of its enclosure wall has been excavated. Between the eleventh and mid thirteenth centuries, the church was re-floored with a polychrome marble pavement of Cosmati type and a free-standing bell tower added, while the enclosure wall to the north had developed into the monastery cloister. The stone villa to the south had disappeared, to be replaced by a village of mud huts. All this is pioneering work: very few small-scale rural monasteries of early medieval date have been excavated anywhere in central or southern Italy. Equally pioneering is the full demographic, palaeopathological and isotope analysis of the extensive later medieval cemetery (some 500 individuals), which appeared in front of the church after the suppression of the monastery in 1297.

The amount of fresh information for the history of Villa Magna over the course of 1200 years that the project has unearthed is remarkable; superlative, too, is the level of detail provided, and within a very short time after the conclusion of the excavations. The volume is beautifully designed and printed (it even has that all too rare but indispensable item in any excavation report: a decent index). There are some disappointments (through no fault of the excavators): carbonised seeds for the Roman period were completely absent, for example, and the faunal samples for all periods were disappointingly small. One wonders, in fact, whether wine processing in the Roman and Byzantine periods may have been accompanied by the production of cereals and other crops, such as the Triticum aestivum (bread wheat) and the vetches abundantly attested in medieval levels; it is just by chance, perhaps, that the evidence for earlier periods has not survived (an absence of burnt levels?). Interpretation of the fourth-century AD aisled hall as a wine facility is particularly uncertain (one influenced by a single comparandum, at Passolombardo); could it have been an estate granary, for example, of the late Roman basilican type attested in the via Gabina 10 villa near Rome, and at Piazza Armerina and Gerace in Sicily? The very richness of the documentation presented here provides much food for thought, and a stimulant for further discussion. Elizabeth Fentress and her colleagues deserve nothing but praise for bringing their finely conceived project to such a triumphant conclusion.