The raw material of the majority of gems in the Campbell Bonner Magical Gems Database has not been analyzed: the archaeometry of magical gems remains a task for the future. Since the CBd-team rarely has the opportunity to study gems first hand, and has practically no resources to carry out archaeometrical investigations, we are left with what information is provided by museums in possession of the gems or described in the available literature. These descriptions tend to follow various terminologies, which is further complicated by inconsistencies in translation. Thus, for the sole purpose of search result optimization in the database, we have decided to create a slightly simplified system, which is inevitably artificial and arbitrary, but manages to give usable results. (Full-text references naturally preserve the original information.)

Some guidelines for the system
- Mineralogical classifications are avoided (e.g. agate and jasper are not grouped under chalcedony).
- In cases where the material of a gem is uncertain, the name of the material is followed by a question mark, e.g. jasper (?). These uncertain cases are also included in the search result given for the general query of the material, so 'jasper (?)' gems will appear both in the list for ‘jasper (colour unspecified)’ and in ‘jasper (all colours)’.
- The colours of gems are indicated within the material field, in additive lists. Rather than listing all possible hues, we aim at creating groups which cover approximate colour ranges, for example ‘jasper, green to black’ will include all jaspers of green(ish) / grey(ish) / brown(ish) / black(ish) colour. The lightness or darkness of a colour is not differentiated. This way, it is possible to search for jaspers of a certain colour (e.g. 'jasper, yellow'), and for all the jaspers stored in the database (‘jasper, all colours’).
- Stones which always tend to appear in the same colour, eg. amethyst (purple) or citrine (yellow), have their colours given in brackets for the sole purpose of facilitating a colour search.
- With banded stones and gems with inclusions, both characteristic colours are usually given (e.g. ‘agate, brown and white’; ‘jasper, green and red’).

Notes on peculiar cases
- Heliotrope, Nilkiesel, Nicolo, Sardonyx. Heliotrope appears in the drop-down search list, and can be queried, but it will not appear on the datasheet itself, where one will see ‘jasper, green and red’ instead. Simiarly, Nilkiesel and nicolo – although both can be queried from the drop-down search list – will appear on the datasheets as ‘jasper, yellow’ and ‘onyx’, respectively. The same applies for sardonyx, which is also invisibly linked to the onyx group.
- Bloodstone. The designation ‘bloodstone’ is avoided whenever the gem can be identified as ‘haematite’. In these cases, too, the gem preserves the tag ‘bloodstone’, but this will not show up on the datasheet, where the gem will be designated as 'haematite'.
- Glass vs glass paste. While a distinction exists between glass and glass paste to refer to an ancient and a modern category, it has so far proved impossible to follow this classification consistently in the database. Therefore, CBd uses ‘glass’ to designate both categories and to facilitate searching for a stone substitute, be it ancient or modern. Stone substitutes will appear in the search results for the respective stones (e.g. ’glass, imitating amethyst’ for ’amethyst’).

This system, and the list in the ‘material’ field is in statu nascendi – we would be very glad to receive comments on how to make it better.