Database structure

Technical specifications

The online version of Trismegistos works with MySQL and PHP, but for data entry a set of relational Filemaker databases are used, currently in a mixed Filemaker 12-13 environment.

The core of the database is Trismegistos Texts, where each individual record is identified by a unique, stable Trismegistos number (for criteria what constitutes a record, click here). This number is used to link to most of the other constituent databases of Trismegistos, and allows for easy implementation of one-to-many and many-to-one relationships.

The Texts database and its satellites

Around Trismegistos Texts, there are two related tables for dates and publications, as well as four main related sets of databases: Trismegistos Collections, Trismegistos Archives, Trismegistos Places, and Trismegistos People.

Collections and Archives

The first two of these have the simplest structure with just two tables: in each case a database of Collections (COLL) and of Archives (ARCH) is connected with the Texts (TEX) database via intermediate databases, one of collection affiliations for texts (COLLREF) and one for archival affiliations for texts (ARCHREF).

COLLREF contains inventory numbers (current and sometimes obsolete) or just general locations of ancients texts in collections, both public and private.

ARCHREF contains a relation between an ancient text and a collection of texts in antiquity, either certain, presumed, or erroneous.


Places is somewhat more complicated, because the database with Places (GEO) is connected with two separate intermediate databases, one for attestations of ancient places in the texts (GEOREF), and the other for indications of provenance of the texts themselves (GEOTEX).


The most complex structure is that of People, where there is only a single intermediate database with attestations of people (and names) in texts (REF), but this is linked with two other databases, one of people (PER), which contains individuals, and one of names (NAM), which contains names but is in fact only indirectly linked to the attestions through intermediary databases of name variants (NAMVAR) and declined name variants (NAMVARCASE).