(cf. Fulford and Bird 1975, 179—81). It is possible that ‘Mayen’ ware
was exported to Britain in ships involved in delivering grain to the
Rhineland that had been collected as part of the annona militaris (corn
requisitions) in Britain (Ammianus Marcellinus, xviii, 2, 3), in much the
same way as BB1 from Dorset may have been transported to the northern
military zone of Britain through the Irish Sea (Fulford 1981, 202—3). The
commerce between Britain and the Rhine was intensified by the emperor Julian
in A.D. 359, when a large fleet was built to supply the imperial campaign in
the lower Rhine (Frere 1974, 390). Glass vessels may also have been exported
to Britain from the lower Rhineland (Price 1978, 75—7), but it is thought
that this trade may have declined in volume in the late fourth century (ibid.).
The high proportion of the total number of ‘Mayen’ vessels in
Britain that jars comprise has been used to support the hypothesis that
these vessels were imported laden with some commodity (Fulford and Bird
1975, 181). This is an interesting possibility, but it would be
necessary to compare vessel ratios in Britain with those in the area of
production, and recognise some differentiation between the two regions,
before this piece of evidence can be considered valid to the testing of the
hypothesis. The predominant role of the jars over other forms in Britain
bears a close resemblance to the patterns exhibited by ‘Portchester
"D" and late Roman shelly ware; while it is possible that these
were also imported for their contents, all three patterns may also be a
reflection of an especial esteem in which high-quality, often very
hard-wearing, jars were held. In Kent at least, local production of sandy
ware jars may have failed to fulfil the demand for such vessels in the
middle and late years of the fourth century, encouraging importation to fill
the void (see the following section). Slipped Nene Valley necked jars (q.v.
the preceding section; Howe et al. 1980, 9 and nos. 70, 75—7)
may also have taken a part in this trade.
Late Roman grog-tempered ware (4.IV.3) may
have
|
|
been the main local competitor with these imports in the late fourth century
throughout Kent. It evidently emerged in west Kent somewhat later than in
the east of the county, perhaps not becoming a significant element in
pottery assemblages until the second or third quarter of the fourth century.
Only one out of some 38 vessels represented at Bexley is in this ware, all
of these vessels conceivably being of fourth-century date. Five out of the
ten vessels published from one of the mid-fourth century pits at
Lullingstone are grog-tempered jars (Pollard 1987, Group XLV, Fabric 75;
q.v. Meates 1979, 56—7); several vessels are also represented
in the destruction and post-destruction deposits of the villa, and in the
occupation level south of the Temple-Mausoleum. A mid-fourth century
development of large-scale importation and/or local production (it is at
present not possible to differentiate between these two modes of
distribution) would be concomitant with the apparent rarity of the ware at
the Cobham Park and Maidstone villas, where occupation may have ended at
this time (Tester 1961, 97; the Maidstone evidence has been assessed by the
present author in consultation with the excavator, Mr D. Kelly). It would
also help explain the low proportion of grog-tempered ware in the Chalk ‘cellar’
infill accumulation (Appendix 5; the figure of 6 per cent is under half that
for this ware in the third-fourth century Rochester deposit, and one-quarter
that of the Springhead deposit of the same date). The layer containing most
of the pottery (layer 6), was sealed by a ‘nearly sterile’ layer (5)
prior to the deposition of building material (layer 4) c. A.D. 395
(Johnston 1972, 120). Deposition of the Rochester and Springhead material
may have continued into the fifth century (Redfern 1978, 54; the Springhead
coin list from the temenos area includes several late fourth- to
early fifth-century issues: Penn 1959, 1960, 1962; Pollard 1983a, fig. 64),
as did the destruction/post-destruction sequence at Lullingstone (Meates
1979, 41—2). The development of grog-tempered ware,
|