Turning to the 3rd century, an analogous situation can be observed in so far as Lezoux samian, normally ascribed dates within the 2nd century, appears frequently in 3rd century (and later) deposits. At Stonea Grange, for instance, Pit 10 by the temple site contained a group of samian believed to have been deposited in a single event and consisting of Lezoux items dated to the second half of the 2nd century; the feature though was dated to the early 3rd century by the excavators (Johns 1996; Jackson and Potter 1996).
Furthermore, samian vessels occur in some genuinely late Roman contexts at various site types. Portchester fort was constructed c. AD 280-90. From the excavations some 1200g of samian were recovered, with c. 42 vessels represented, amounting to c. 0.3% of all the pottery from the site. This material was mainly East Gaulish and, significantly, occurred in the absence of coarse wares that would be considered contemporary with it. In reporting this material Morris stated that: 'It is unlikely that this represents evidence of occupation in the late second and third centuries ... More probable is the explanation that these sherds represent old pottery brought to the site by the first occupants, possibly as treasured heirlooms or antiques' (1975, 277-8).
At the shore fort at Lympne, Kent, excavations in 1976-8 also yielded an assemblage of samian (Cunliffe 1980; Young 1980b), though it is possible that there was occupation at this site prior to the construction of the shore fort, c. AD 260/270. The samian in this case included pieces of Trajanic date, and more Central Gaulish than East Gaulish samian. (At Brancaster, also a site with a shore fort, the dating of the beginning of occupation in the area of the western vicus of c. AD 170 seems correct, with the coin evidence and the samian in agreement. Here samian is contemporary with activity/occupation in the later 2nd century).
Samian from late Roman era contexts at the rural site at Birdlip Quarry by Ermin Street suggests that here, as at Portchester, it may have been in use in the 4th century. At this site the excavators note that: 'In Period 2B [c. AD 330-380] large sherds [of Dressel 20 amphora] are virtually absent and small sherds have a similar distribution to other rubbish. It appears that all sherds from this period are residual which may imply that amphorae were no longer available or required. The contrast with the quantity of large sherds of samian in Area A in this period is of note but not easily explained unless it is assumed that some samian continued in use into the 4th century' (Mudd et al. 1999, 235). Consideration of the taphonomy of samian sherds will be important in future attempts to distinguish fragments of long-lived samian vessels from residual sherds within 3rd and 4th century deposits.
Groups of 3rd century date provide interesting data in this field. Some of the groups, despite their date, are numerically dominated by Lezoux vessels. The large group of samian (206 vessels) from Phase 7 at Bainesse, Catterick (Code Catte3), dated to c. AD 200-275 is dominated by Lezoux vessels, with East Gaulish samian accounting for less than 10% of the group. Doubtless a proportion of these Lezoux vessels will be residual in this phase, but it seems probable that a considerable number were still in use in the 3rd century. Smaller 3rd century groups from Heybridge, Elms Farm, Ceramic Phase 8 (Code HElms6) and South Shields (Code SShie1) are also dominated by Lezoux vessels. Yet this is not a universal trend: the small 3rd century groups from Bignor Period II to III (Code Bigno2), from Catterick Bridge, Phases 3-4 (Catte6), and the top filling of two pits at St Thomas' Street, Southwark (Swark11), all have more East Gaulish samian vessels than Central Gaulish items.
Colin Wallace has assembled a list of further instances of samian items in contexts that are later than the normal date range of the type (Wallace forthcoming).
© Internet Archaeology
URL: http://intarch.ac.uk/journal/issue17/1/5.8.3.html
Last updated: Mon Mar 7 2005