Note 1 [go back]. The use of the terms samian and terra sigillata has generated discussion in the past (cf. Oswald and Pryce 1920, 3-4) and over recent years. The term samian ware or just samian is normally applied in Britain, while on the continent terra sigillata has been preferred, particularly in Germany, meaning 'earth with little figures' (Brunsting 1972). Tony King has advocated the use of 'samian' suggesting that this is likely to approximate to the contemporary Latin name for this pottery, meaning red or black glossy pottery (King 1980).

Note 2 [go back]. The phrase 'dataset' is mainly employed in this text as an umbrella term, refering to all the various elements of collated samian information collected for the Project.

Note 3 [go back]. Ideally an alternative option would be the use of EVEs (Estimated Vessel Equivalents) data; that is to replace the 'number of vessels per samian form' type of data with the EVE measure per type per group. The EVE method would be preferable for it is a statistically robust approach for establishing the relative frequency of pottery types (eg. Orton 1989; Orton et al. 1993; Tyers 1993). This sort of data is now being recorded for samian in routine processing and the number of reports employing this method of quantification that have been published has grown, particularly very recently. However, this kind of record is almost invariably published in the site samian/pottery reports as an aggregate, in summary form per assemblage or phase (cf. Tables 27 and 28); that is, not in a manner which makes constructing a Database of Samian Values by EVE, as an alternative to the present Spreadsheets, possible. In order to build a dataset based upon EVE data what would be required are simply the rim equivalent amounts recorded per samian form type per group. The EVE data used for Tables 50-52 is of this type but derives from data collected first hand by the author relating to forthcoming reports. See Section 8.2.7.

The formal structure of a database via EVEs would be identical to that of the existing Values Spreadsheet. Quantities of samian types would be expressed as values. Hopefully this will cease to be presented as an aggregate in the future, but in proper detail documenting EVE amounts per form type per group. As the increase in electronic publication allows for the relatively easy placement of detailed records into the public domain more information of this type will become available.