Internet Archaeol. 17. Willis. Samian Lamps and Lighting

8.5 Samian lamps and lighting

8.5.1 The incidence of samian lamps

A small number of samian lamps in Britain were recorded in the course of work for the project. These are of two basic types. On the one hand are lamps with elaborately figured moulded relief appearing as a human head on the upper side, and on the other, conventional disc lamps, familiar from the form of the many none samian ceramic lamps in Britain. Lamps in samian ware are rare in Britain and elsewhere in the north-west provinces; Donald Bailey and Joanna Bird for instance, noted only items form, or likely to be from New Fresh Wharf in 1986 (Appendix 8.2). Hermet illustrates two conventional lamps, presumably from La Graufesenque (1934, Pl. 5, No. 23). All six samian lamps found in Britain though are, or are likely to be, Central Gaulish and of second century date. To this number may be added several apparent 'improvised' lamps fashioned from samian forms which are of particular interest.

The total of six samian lamps are recorded in Appendix 8.2 and come from three sites. Excavations at New Fresh Wharf (St Magnus House) 1974-8 produced a spout from a samian ware lamp in Central Gaulish fabric, although the type of lamp represented was uncertain. This work also yielded a group of plain disc lamps in 'black samian' otherwise known as Lezoux/Central Gaulish black-slipped ware, one stamped 'VIBIVS' (Appendix 6.2). Two samian plain disc lamps discovered c. 1880 are thought likely to have come from New Fresh Wharf as well. Both items are thought to be Central Gaulish and are stamped, one with the stamp 'VIBIVS', as with the black slipped lamp from 1974-8 (Appendix 8.2). One other samian lamp now in the Museum of London is apparently from London, though its find-spot is now unknown. This complete item is elaborately moulded, with the front view in the form of an African head, and with the side view designed to resemble the profile of a camel's head (cf. Oswald and Pryce 1920, Pl. 85 No. 7; Appendix 8.2). A similar lamp to this elaborate figured lamp was found at Alcester in 1976-7. This item is thought to be probably Antonine and displays similar moulded characteristics to the elaborate moulded lamp from London. The Alcester find came from a context of late 3rd to mid 4th century date just outside a substantial domestic style building (Appendix 8.2).

A ceramic lamp, not of samian, but stylistically similar to these items from London and Alcester, with the upper section fashioned to appear as a male head (a god?) in relief was recovered from grave 1 at Helshoven, north-west of Tongeren in eastern Belgium (Roosens and Lux 1974, 20, Fig. 10 No. 38). This rural grave is dated to the mid second century AD. In this case the lamp has a smooth underside/base with handle made to look like a palm leaf.

The final samian lamp recorded comes from Latimer villa, Buckinghamshire. The precise type represented is uncertain though the handle and wick end were represented; the exact context of the find has been lost (Appendix 8.2).

8.5.2 Lamps 'improvised' from samian vessels

Several cases of conversion, or potential conversion, of samian vessels into lamps can be documented.

The Fairy Knowe broch at Buchlyvie, central Scotland, was occupied and subsequently destroyed during the late 1st / early 2nd century AD (Main 1998). A wide variety of artefacts were recovered during excavations, including local/regional tradition pottery and a range of Roman pottery types. The Roman material from the broch dates to the period around the time of the first Roman military intervention in Scotland during the Flavian era and its occurrence may be directly or indirectly related to the presence of the Roman army in the vicinity of the site. Roman material arriving in this area of Scotland at this time will have been very different from what local people, such as those who presumably occupied the broch, were accustomed, in terms of form, colour, material composition, technology of manufacture and 'normative function'. Hence, how these items were understood and employed around the time of their receipt by the occupants of the broch is not certain and a matter for investigation. Nonetheless these items were used by local communities, in many cases being adapted to local requirements. Examination of the pottery from the broch shows that bases from three samian platters and one coarse ware flagon had been trimmed round at the junction of the vessel floor and the footring, with the trimmed edges then 'finished' by smoothing any rough fractures; these altered bases appear to have been adapted so that they could be inverted for use as shallow ash-tray type dishes (Willis 1998b, 321-5). This may represent pragmatic reuse of broken vessels or the intentional alteration of complete vessels.

At least one of the altered samian vessels shows signs of burning within the footring and the possibility that these vessels were adapted for use as open lamps therefore suggests itself. No ceramic lamps occur amongst the assemblage from this site, though two examples of the rare small dish Drag. 22/23 (Webster 1996, 36; Stanfield 1929, 124-5) are present and are approximately the size of an open lamp. It is understandable that some form of artificial lighting will have been necessary within the broch, illuminating its otherwise naturally gloomy interior components.

Jo Mills noted a vessel from Tort Hill East, a rural site adjacent to Ermine Street, Cambridgeshire, which had been trimmed in a manner similar to the vessels from Fairy Knowe (Mills 1998, 68-9). In this case though notches were also cut into the (inverted) footring, one of which was blackened suggesting the position of a wick. This vessel dates to the period c. AD 135-250 and seems a strong candidate as a base re-used as a lamp. Another example comes from Stockton West Moor, another rural site c. 2 miles north-east of York. This example was recovered during excavations in advance of pipeline works for Yorkshire Water and is not yet published (YORYM 1996.390, context: 1081/c138; Pearson 1997; cf. Pearson and Brinklow 1997; Roskams 1999, 59). The vessel in question is a Drag. 31R bowl in Lezoux fabric dating to c. AD 160-200. Again it was adapted for re-use via trimming, approximately at the junction of the vessel wall and floor, with these fractures then worn smooth. This base is now damaged but c. 80% is represented. Again, when inverted it forms a small dish-like receptacle and in three places the footring shows signs of burning; this seems to indicate that this item was used as an open lamp. Most remarkable, however, is an example of a samian Ritterling from 12 bowl from a rural inhumation burial at Whitcombe, Dorset (Simpson 1990, 79, Fig. 13). The burial dates to the late first century AD, and is of a young adult female. Grave goods included local 'Durotrigian' pottery, a decorated samian beaker and this samian bowl which seems to have been re-fashioned as a lamp (Appendix 9.1). The flanged rim of this vessel had been trimmed symmetrically to mirror the shape of a lamp with nozzle/spout like point and handle. The nozzle/point area shows burning, suggesting that a wick was maybe positioned there, as with 'purpose-made' open lamps (cf. Darling 1985, 87). Purpose-made ceramic lamps of the Roman era do not always display burning around the positions of their wicks or elsewhere (cf. Vernou-Magister 1991), though perhaps in past publications this aspect has not been consistently reported one way or the other.

The purpose/s to which these altered items were actually put is not certain, but their interpretation as lamps seems reasonable given the evidence. Lamps often occur amongst grave groups in southern Britain (cf. Philpott 1991; Eckardt 2000) and the occurrence of an altered samian vessel in the apparent guise of a lamp in a burial context at Whitcombe is therefore not entirely surprising.

Samian may have been preferentially chosen for making into lamps. Its selection may be relevant at several levels: for instance, since these improvised lamps were likely to be seen in social contexts, perhaps at meal times, maybe, it was appropriate that they were made from table ware; similarly the high-fired temperature used in samian manufacture and its gloss slip probably made samian much less porous to oils than was the case with other ceramic types. A range of samian forms have footrings which could effectively become vessel walls when the vessel is inverted; this contrasts with many other pottery types of the north-west provinces which were without footrings. At another level it may be thought no coincidence that the chosen medium is samian ware and the functional item produced is a ceramic lamp: both samian ware and ceramic lamps are types without an indigenous pedigree in the north-west provinces and are Mediterranean in genealogy. Hence within the north-west provinces they were perhaps symbolically 'highly charged' manifestations of Roman imperial culture.