The main obstacle for a research on imperial properties at the scale of the entire Roman empire is the current fragmentary knowledge of the sources. This is due to a deficit of general studies in the last century and to the continuous discovery of new epigraphic and papyrological sources. In order to solve this problem in the best possible way, the PATRIMONIVM project has created the Atlas patrimonii Caesaris (APC), a relational database where all relevant evidence for the study of the imperial properties is stored. Information on the more or less precise localisation of the imperial properties has been extracted from the evidence and represented in a cartographic way.

Before consulting the database, please make sure you have read the sections about the criteria we used for the reconstruction of the geography of imperial properties and on the way we displayed data on the map.


APC features five main components. Three are dedicated respectively to records for textual sources (literary, papyrological and epigraphic), places (document provenances, imperial estates, patrimonial districts, etc.), people (from peasants to procurators and freedmen in general). The remaining two contain the bibliographic references and the vocabulary (thesaurus) used in the database. The records of the different components are linked and cross-searchable according to a number of criteria. This powerful and versatile tool finally allows us to manage all the pieces of relevant information scattered over a massive number of publications.

APC is a large database with thousands of documents, places and people records and still remains under construction and development. While most of the sources for the reconstruction of the geography of imperial properties are already included, new documents are still added regularly, and modifications can be introduced into existing ones. In this first version, APC should be mainly seen as a supporting tool for the forthcoming publications on the history of imperial properties in the Roman world and not yet as an independent publication. In the long term, the database will not simply contain raw data: translations, historical commentaries and bibliographic references will be provided for every textual source, while descriptions attached to the different geographical entities will include topographic and archaeological informations. The prosopography will include all attested imperial slaves and freedmen as well as all persons working in connection with the imperial properties (administrations, tenants, contractors etc.). APC is just starting, and this living project still has a long way to go!

APC has been constructed following FAIR principles and is using the Exist-db environment. APC has been built on Linked Open Data principles and aims at a large interoperability with other online databases and services. APC is still under development, links to Trismegistos, Pleiades and other projects are still very incomplete.

APC also includes the PATRIMONIVM Editor, a innovative tool for creating and annotating Epidoc documents with an easy to use interface that significantly reduces the learning curve needed to produce digital editions of epigraphic texts. Read more about the technologies involved in the project here.

Planned features: data as XML, complex and full-text search tools, more interactive maps (including with timeline)

Criteria for the reconstruction of the geography of imperial properties

In order to reconstruct the geography of imperial properties in the Roman world, the PATRIMONIVM team has collected and reviewed the entire documentation for each province. The reviewing process – which has been conducted collectively and under the supervision of the PI – has allowed to establish which sources certainly or probably attest the existence of an imperial property. While differences for each regional context had to be taken into account, we tried to apply a consistent set of criteria to determine if a document can reliably be used.

Landed estates. Documentation concerning imperial estates is very diverse. Some documents, such as boundary stones, lease contracts, land registers, attestations of imperial coloni, attestations of administrators with precise titles, clearly prove the presence of properties of this kind. In many other cases, the reconstruction rests on indirect evidence. Funerary monuments of imperial slaves (even without specific administrative titles) in isolated, agrarian contexts have generally been taken as reliable sources. The presence of imperial slaves bearing titles such as vilicus, actor etc. has generally been used to localize imperial estates, when other interpretations (e.g. the administration of a customs station) could reasonably be excluded. Imperial slaves attested in provincial capitals were considered as pertaining to the provincial administration, apart from cases where specific titles indicated otherwise. Imperial freedmen without an administrative title connected with the patrimonium Caesaris where not considered as an attestation of imperial estates, but exceptions justified by the local context do exist. References in literary sources have been used only when they allowed to determine the location of a property with sufficient precision (attribution at least to a region or to a province). General references to the origin of the family of an emperor or of his wife, or to the fact that en emperor resided in a particular place for a certain period, have not been considered when not corroborated by other evidence. Information on production and other features has been added only if clearly attested in the sources or if it can be reasonably deduced by contextual elements.

Forests/pasturelands. This type of property cannot be easily distinguished from a (at least partly) cultivated landed estate, which could also be partly used for pasturage. We reserved this more specific type for land properties situated in mountain regions, where we can safely assume that cultivation played a minor role.

Mines. It is not easy to define what constitutes an imperial mine. Contrary to what can be found in older (and sometimes in recent) scholarship, the emperor did not own all mines of the empire. Public (that is of the populus Romanus), municipal and even private ownership continue to be attested in the first three centuries of the Empire. However, mines for which we have no information about ownership remain numerous. The inclusion or not of an ancient mine in APC is determined by two main criteria. Firstly, we have included all specific mines for which we have clear evidence of imperial ownership, either from literary texts (e.g. aes Livianum, aes Sallustianum) or from epigraphic attestations of dedicated administrators (mainly freedmen procurators). Secondly, we have included all mines laying within attested mining territories (e.g. metalla Ulpiana, ferrariae Pannoniarum) which were administered by equestrian procurators. The difference in the administration of this second group of mines suggests that not all sites belonging to these districts where in the ownership of the emperor (i.e. did not belong to the patrimonium Caesaris), but the issue remain debated. Another problem concerns the limits of these districts and the actual mines that were exploited. Information is very limited in this regard and we have generally decided to include all sites for which exploitation in Roman times is archaeologically attested and which can reasonably be attributed to one of these districts. This means that, for certain provinces where ancient mining is well attested (e.g. Britain), APC shows no data because we have no traces of specific imperial mines or of mining territories. ***Please note that due to the complexity of the identification of imperial mines, the review of documentation is still ongoing. Data displayed on the map should be considered as provisional.

Quarries. As for mines, imperial ownership of quarries was not general but was limited to a certain number of extraction sites, which produced prestigious decorative stones. For these quarries, we usually have evidence coming from quarry marks and funerary or honorary inscriptions mentioning slaves, freedmen and contractors. Certain quarries which were not owned by the emperor (e.g. Hymettos or Pentelikon), but for which we have evidence of imperial quarry marks, have also been included. For large quarrying districts (e.g. Karystos, Mons Claudianus), specific extraction sites, buildings and facilities are indicated and localized on the map.

Workshops and processing units. In the great majority of cases, imperial workshops producing amphorae, bricks, tiles etc. and processing units such as oil and wine presses are only indirectly attested and can be broadly localized in the region around the find spots of the relative evidence. When archaeological remains are known (e.g. Loron or Fazana), they are precisely located on the map. Bricks and tiles simply carrying the name of the emperor in the genitive case are included in the documentation, but are often not considered as proving the existence of imperial workshops, since the genitive may just indicate the ownership of the brick/tile and not of the workshop, therefore alluding to an imperial order for a specific construction. If the genitive precedes the name of the workshop, imperial ownership is more probable. ***Please note that due to the high number of stamped fragments of ceramics, brick and tiles, only a representative number of documents has been inserted for each workshop.

Animals. Outside Egypt, where imperial sheep, goats, pigs, and donkeys are often mentioned, attestation of imperial flocks are rare. When clearly mentioned, imperial flocks have been defined as production units, since they where used, for example, for producing wool and cheese. If flocks are not explicitly attested, but contextual elements indicate that a landed estate was used for pasturage/grazing, a reference to animals is added to the record of the relative property and document. When transport animals (donkeys, camels, horses) are mentioned, no specific production unit is added, but the relative document is marked with the corresponding keyword. Horse-breeding farms are considered landed estates producing horses. Otherwise, land qualified as pasturage in APC can also be taken as an indirect evidence to the presence of animals.

Patrimonial districts. Patrimonial districts (regiones, tractus, mining territories) and ousiai are indicated as circles to distinguish them from other points. Their record clearly indicate all imperial properties that where included in these administrative units.

Display of data on the map

The main block of geolocalized data present in the database is constituted by places of provenance of the documents, or places that are mentioned in the documentation. As provenance, we do not intend the find spot of a document, but the exact place or the area where it stood in antiquity (for epigraphic monuments) or where it was preserved (for papyri). If an epigraphic monuments have been reused or transported elsewhere and the original provenance cannot be reconstructed, a more generic provenance is given (an ancient city, a modern place, a province). The provenance of stamped amphoras, bricks, tiles, ingots or inscribed marble blocks coincides with the area where they have been transported or left in antiquity after their production. For these documents, we systematically indicate the workshop, mine or quarry where they were produced as their place of origin.

When the exact location of a place is not known, we preferred not to give it some arbitrary coordinates, which would mislead the user. In these cases, the position of an element of the map is given in relation to other localized places. For example, since the position of the four sectors (bracchia) of the Bacakale quarry of Dokimeion is unknown, they are simply indicated as « part of » Bacakale quarry. Their position on the map therefore coincides with the point representing the location of Bacakale quarry.

A second layer of places is formed by the imperial properties, called production units. When archaeological remains are present, exact coordinates are given on the map. However, since in most cases the attestation is only indirect, the position on the map is determined in relation to the place of provenance of the documents attesting the existence of that property. For example, a landed estate attested by an inscription mentioning imperial coloni is placed on the map exactly above the point indicating the provenance of that inscription, even if that point does not represent the position of the estate. This is a way to show that the information is unknown. In this case also, we preferred not to give arbitrary coordinates that would give the idea that the estate can indeed be located in a certain place. If an estate is mentioned by two or more documents (e.g. a series of boundary stones), the corresponding icon is automatically placed on the equidistant point between the known provenances of the relative documents. If we only known that a property was located in a certain district, or in a province, the corresponding icon is placed above the point representing that district or province. ***Please note that, sometimes, this automatic calculation places a production unit in unlikely spots, such as in the middle of the sea.

In certain cases, and particularly in certain parts of Egypt where documents are abundant, many landed estates can be indicated as being part of a nome, or in the vicinity of a certain village. The icons representing the estates therefore are piled above the point of the nome or village and only the one placed at the top is directly clickable. In these case, we suggest to click on the point representing the nome or village in order to have the full list of land plots attested.