albeit in small numbers. They are more frequent than cavetto-rim jars, which
comprised a very small proportion of BB2 in east Kent throughout the second
to fourth century, perhaps for reasons discussed above (4.III.3). Dog-dishes
are second to plain pie-dishes in quantitative importance. Taking all forms
together, BB2 comprised up to one-quarter of third-century Canterbury
assemblages (up to half of the coarse wares), quite possibly with similar
quantities in use in both early and later years.
The third coarse ware frequently encountered in early
third-century deposits, ‘Native Coarse Ware’ (q.v. 4.III.3),
continued to be used throughout the century, probably declining in
importance in the latter half (to under 5 per cent of total assemblages).
The forms — mostly jars, but occasionally including bowls (e.g. Jenkins
1952, no. 34) and dog-dishes — exhibit no discernible typological
development. They occur at Wingham and possibly Reculver (Philp 1957, nos.
22—3, 29 have profiles and scoring suggestive of this ware) in mid- to
late third-century levels, as well as at Canterbury. Other, unstratified,
examples in east Kent could also be of later third-century date (cf. Fig.
45).
BB1 is a regular occurrence in Canterbury deposits of the mid-
to late third century, generally comprising up to 9 per cent of total
assemblages and 20 per cent of coarse ware assemblages (by vessel rim
equivalents). Vessels are usually bead-and-flange dishes (Gillam 1970, Type
228) and decorated dog-dishes (ibid., Type 329), or more rarely jars
with rims of wider diameter than the body (the ‘oversailing’ rim, ibid.,
Types 147 and 148; e.g. Jenkins 1952, no. 21). The jars, and the
bead-and-flange dish, are dated to the late third century onwards in the
northern military zone (ibid.), and the dog-dish (Gillam Type 329)
from the late second century onwards. There is no reason to suppose that
these forms reached the south-east any earlier than they did the north
(Williams 1977); in consequence, it may be supposed
|
|
that the major period of importation to east Kent began in the last third of
the third century. Williams has demonstrated that some BB1 from Dorset
certainly reached Canterbury (ibid.), including two bead-and-flange
dishes that may be dateable to the late third century (Williams 1947, fig.
8, nos. 4 and 9; although the excavator dates their contexts to the fourth
century, the coins and other finds do not rule out a third-century date for
the fills, and this is certainly more likely on ceramic grounds). On the
basis of visual inspection under a hand-lens, it is possible to ascribe most
BB1 from third-century contexts to a Dorset source. Canterbury, and other
settlements throughout Kent including Wingham and Reculver (Philp 1957, no.
7, and unpublished vessels from the foreshore), were evidently sought out by
the Dorset distributors in a general intensification of trade with the
south-east of Britain (Figs. 35 and 36 here) subsequent to the reduction in
the scale of the northern region’s consumption (Williams 1977, 204—5).
The coincidence of this trade with the development of the Saxon Shore
defensive system in the English Channel, Straits of Dover and North Sea is
of interest, particularly in view of the presence of oolite stone on several
sites in Kent (Williams 1972, fig. 9 and Appendix 3). More research into the
dating and sources of this stone is required before it is possible to
provide good evidence to support the attractive hypothesis that the Dorset
BB1 was imported as a by-product of the transportation of Portland oolite
stone. Purbeck Marble may also have been shipped up the Channel in the late
Roman period (ibid., citing Hull 1958, 96). This kind of commerce
might accord with Fulford’s hypothesis (1981, 202—3) that Dorset BB1 was
distributed over long-distances as a ‘makeweight’ in cargoes of more
important commodities. The imposition of a system of taxation in kind on the
south-east as well as in the north and west in the third century (cf.
Hopkins 1980, 116 ff.) may have occasioned a restructuring of the tax
collection system that resulted in an
|