In 2004 it was tentatively suggested that the Type 18 bowl could be dated to between AD 350 and perhaps AD 450 (Gerrard 2004, 71). Over the last decade the excavation of new sites and the re-evaluation of archives and publications indicate that the chronology of this vessel can be put on a firmer footing. The first observation to make is that the Type 18 bowl has not been identified in any of the extensive and detailed excavations along Hadrian's Wall (Bidwell and Croom 2010). This would suggest that the form entered production after exports to the frontier region had ceased. The date at which this occurred is currently considered to be c.AD 360 (Bidwell and Croom 2010, 34). This late date receives further support from the dated occurrences of the form in southern Britain. Of the 33 examples listed in Table 1, just over half (17) can only be assigned to the 4th century or later on stratigraphic grounds alone. Of this relatively poorly dated sub-sample, ten were considered to have been recovered from late 4th- or early 5th-century deposits or phases. The absence of the Type 18 bowl in late 3rd- or early 4th-century contexts is noticeable.
In sixteen cases associations and stratigraphic relationships between Type 18 bowls and coins, metalwork and other items can be identified with some certainty. At four sites Type 18 vessels are associated with coins of the House of Valentinian (AD 364-378) but eleven occurrences have associations with later coins of the House of Theodosius (AD 388-402). Very late 4th- or 5th-century belt fittings occur at sites alongside these late coins and Type 18 vessels in seven cases. At one site a Type 18 vessel occurs in association with early Saxon pottery.
Walton and Moorhead (this issue) as well as others have recently argued that the use of low-value Roman coins may have continued in some regions until the AD 430s. If this was the case, then coins struck between AD 364 and 402 must have formed a significant component of the money in circulation. Certainly in the West Country issues of the House of Valentinian (AD 364-378) formed the most important influx of coinage in the region after c.AD350 (Walton and Moorhead this issue). The association of Type 18 vessels almost exclusively with this late coinage, the absence of the vessels in the northern frontier zone and the recent discovery of a Type 18 vessel in an Alice Holt fabric alongside a fresh early Saxon sherd strengthens the case for these vessels being used and deposited during the early 5th century. The presence of other elements of the normal late Roman BB1 assemblage (such as the Type 2/3 jars, Type 20 dishes and Type 25 bowls) suggests that BB1 pottery continued to be useful in the West Country, at least, up until and after the traditional end of Roman Britain. The significance of this vessel as an indicator of the 'latest Roman', or 'earliest post-Roman' activity should not be underestimated. In time, it is hoped that more deposits containing this vessel will be identified and that the other artefacts, ecofacts and stratigraphic relationships associated with these vessels will allow us to characterise the early 5th century in a way that has lain for too long beyond our grasp.