Here's what I've found out so far (please let me know of any errors!):
The site report for the Roman Tilery site at St. Oswald’s Priory, Gloucester (Heighway and Parker 1982: 25-77), indicates the presence of local wheel-made BBI pottery in hard, sandy fabric from sub-Roman levels (ibid. 46). Further examples of later local BBI were sought. Local production of Roman pottery has been noted with increasing frequency at the end of the Roman state in
Various centres have produced examples of BBI in local Grey Ware fabric, including the Thames Estuary region and a separate fabric noted in the South of Britain (Tyres 1996: 182). Distribution of BBI had reached the South, West, and lower
It has been established that the Oxford Ware industry had developed a tradition of imitation BBI (Young 1977: 205) to include the ‘dog-bowl’ form similar to that from Crickley. Later versions included the type R53, which bears a similarity to the form of the Crickley bowl, and has been dated to AD 240-400+ (ibid. 221, fig. 82). However, the Crickley example bears traces of a second beading on the rim, of which the most comparable example is from Birdoswold, dated to c. AD 350-400 (Gillam 1968: 71, no. 321), although this form is deeper. This latter form gained popularity during the later Roman period, with all kiln finds extant from the end of the 4th century or later, with local production possibly limited to the end of the Roman period (Young 1977: 206). Kiln sites include Cowley, St. Luke’s Road, Allen’s Pit, Nuffield Orthopaedic Centre, Churchill and Foxcombe Hill, Dorchester and Garsington, Sandford and Shakenoak, Oxfordshire, the latter datable to c. AD350-420 and AD350-400 (ibid. 247, 248, 252, 352).
Further examples of BBI from late 4th – early 5th century levels are not unknown, the most notable examples being located at Poundbury,
[i] Pers. comm. Phil Greatorix,
[ii] Pers. comm. Alan Vince; J. Timby (1986: 63). Non-local grey coarse ware, wheel made micaceous. C3 introduction into
[iii] Pers. comm. Ron Firman.