The excavations at Greyhound Yard in Dorchester, Dorset, produced a considerable quantity of BB1 that ranged in date from the early Roman period to the end of the Roman town sometime after c.AD 400 (Seager Smith and Davies 1993). The finds included a number of vessels that were classified in the Greyhound Yard type-series as round-bodied open bowls with everted rims and flat bases, often decorated with a narrow band of diagonal burnished lines around the girth (Seager Smith and Davies 1993, 233). This vessel was classified as a 'Type 18' bowl and it will be referred to as such here. It does, however, masquerade under other names. At Exeter it is a Type 21 (Holbrook and Bidwell 1991, 103) and it is described as a 'Class 2 necked bowl' at Bestwall Quarry (Lyne 2012, 209-11).
Typologically the vessel is a bowl (i.e. a vessel that is open-mouthed and usually wider than it is tall) but shows affinities with the standard late Roman everted jar form (i.e. a vessel that is taller than it is wide and with a closed-mouth) (Seager Smith and Davies 1993, Type 3). This causes a number of problems related to the identification of the vessel. When found in small fragments it is almost impossible to distinguish the standard jar form from the Type 18 bowl. One seemingly diagnostic attribute is the band of burnished diagonal lines. This is common on the Type 18 bowl but is not exclusive to these vessels. They can also be decorated with an obtuse lattice decoration, which is more typical of the jar form. Likewise the jar form can very occasionally be found decorated with just a band of diagonal lines.
In 2004 some 20 sites were thought to have produced definite or probable examples of Type 18 vessels (Gerrard 2004). The tally of sites has now increased and a review of the dating and distribution of the form might be considered apt (Table 1 and Figure 1).
The dating of this vessel type is entirely dependent on stratigraphic and artefactual associations. Consequently, the vessels tend to fall into one of three categories. The first category is vessels from sites or contexts that are considered to be either late Roman or 4th century. A good example might be the vessels from the old excavations at Dorchester Hospital and Colliton Park (Aitken and Aitken 1982, 112; Patrick-Greene 1993, 89 and 100). Here, Type 18 vessels were found in the rubble collapse deposits of late Roman buildings but no closer dating evidence is available. Similarly, an example from the High Ham villa in Somerset must post-date the construction of the villa and is thus '4th-century' but greater chronological precision is impossible (Leech 1977, 119). The second group of vessels come from excavations where the deposits that contained the Type 18 sherds have been phased to the 'late 4th century', 'AD 350-450' or the like by the excavators. There is some overlap between this group and the first group. The third and final group of Type 18 bowls are those from sites where precise associations with coins or other artefacts can be established. These are, by definition, of greatest interest to us in establishing a chronology for this BB1 form.
All examples of Type 18 vessels and possible Type 18 vessels known to the author are listed in Table 1. A selection of finds with the best dating evidence (i.e. vessels falling into groups 2 and 3 above) and/or vessels that were not included in the earlier paper (Gerrard 2004) are discussed below. They are reviewed on a site-by-site basis, starting with examples from production sites and progressing to examples from Dorchester and Dorset before moving on to finds from further afield.