sequence covering all vessels, allocating number-blocks to form groupings
(e.g. Gillam 1957; 1970; Cunliffe 1971; Fulford 1975b), sometimes with short
parts of the sequence left open for the insertion of future types (e.g.
Hawkes and Hull 1947; Hull 1958). The inflexibility of these methods is a
clear disadvantage. The pottery reports from most major sites have opted for
less synthesised systems, illustrating a representation of the assemblage as
a whole or by contexts with some repetition of forms (e.g. Bushe-Fox 1913;
1914; 1916; 1926; 1928; 1932; 1949; Cunliffe 1968; Frere 1972; Orton 1977b;
Whiting et al. 1931).
|
|
The type-series developed by researchers studying a particular
industry have used a combination of major divisions of fabric and continuous
numerical sequences (e.g. Fulford 1975a; Young 1977a) or a single numerical
sequence covering all fabrics (e.g. Lyne and Jefferies
1979). The present author has adopted the latter approach for the
illustrations of the major wares, known or suspected to have been produced
in the study region, and of other vessels of particular significance, that
accompany this study. The illustrations have been selected as
|