A substantial portion of the Collections are made up of objects relating to the Later Iron Age and Romano-British periods; records attributed to these periods make up around a quarter of all archaeological records on the collections management database. Arguably, this is the most disparate group of material in the collections, as there is no clear geographical focus for the assemblages, nor any one significant assemblage to form a nucleus for the collections relating to this period. Many of the most exciting recent discoveries of Romano- British archaeology in Wiltshire have typically been the result of development in the Amesbury and Swindon areas, both outside of the Museum’s collecting area. More-so than other periods, the temporary pause on collecting archaeological archives between 2013 and 2022 also appears to have had a notable effect on the collections, with a number of potentially significant assemblages from as far back as the early 2000s awaiting deposition.
Despite this, probably the most significant individual object held in the Museum Collections outside of the Early Bronze Age dates to this period: the Marlborough Bucket (Figure 6.1). The bucket was discovered during gravel digging in St. Margaret’s Mead, Marlborough, in 1807, and eventually acquired by the Wiltshire Museum in 1878 (Cunnington 1887). The bucket was apparently discovered intact, and contained cremated human remains. It survived for just long enough that sketches could be made before it fell to pieces, the state in which it was received by the museum amongst a ‘box of broken urns and various fragments’ (Cunnington 1887: 224). Although initially enigmatic and dated to the Roman period based on surrounding discoveries, the date of the vessel is now firmly established as being Late Iron Age, and one of a number of funerary buckets found in Britain dating to this period (Stead 1971), and arguably one of the finest known. Yet, despite its huge archaeological and art-historical value, the last serious academic attention the vessel received was in the 1970s (Jope 2000), although this has recently changed (See Ellis 2021).
The most famous assemblage dating to this period may be that excavated at Nettleton Scrubb, initially by Priestly 1938-1954 and subsequently by William Wedlake over the course of the later 1950s and 1960s (Wedlake 1982). These excavations revealed a significant roadside settlement situated on the Fosse Way, and occupied throughout the Roman period. The site is best known for the long-lived shrine dedicated to Apollo identified on the site, with multiple phases of construction, but it is also notable for the considerable evidence for industrial production in the second half of the fourth century. This included only the second in situ pewter vessel workshop excavated in the country (Lee 2009). Unfortunately, the archive is divided between the Wiltshire Museum and Bristol Museum and Art Gallery, and the Museum holds only the material relating to Priestly’s earlier, smaller scale, excavations, yet this still equates to a moderately sized and varied collection, including c. 1,000 sherds of pottery, stone carvings and a copper alloy cockerell- shaped candle holder.
In addition to Nettleton, significant roadside settlements are also known at Cunetio, near Marlborough, and Verlucio, near Calne, both on the main east-west road connecting Bath and Silchester. Unfortunately, neither settlement is well understood. Verclucio remains the only unscheduled defended Romano-British roadside settlement in Britain (Linford et al. 2018), and remains unexcavated. The Museum holds a large assemblage of field-walked ceramics collected from within the enclosed area, as well as a large assemblage of metal- detected finds from an area immediately to the east of the defended area. Neither assemblage has been analysed or published in its entirety. Cunetio, another defended site, is particularly notable for the apparent preservation of the site, as well as for the later phase of its defensive circuit. This was constructed c. 360-380, and is of a scale and style that implies a central authority was involved in its construction (Corney 2001; Mattingly 2006: 333; Gerrard 2013: 43ff). The Museum holds the paper and material archive associated with relatively small scale, and largely unpublished, excavations ran by F.K. Annable and J.A. Clarke between 1957 and 1964 (WAM 1959; 1960; 1962; Annable 1966), including a relatively large ceramic assemblage. The archive deriving from more recent excavations within the town by Wessex Archaeology and Time Team (Wessex Archaeology 2011) will soon hopefully be deposited with the Museum. In addition to the excavated material, the Museum also holds a number of artefacts attributed to Cunetio from the prolific collector Jonathon Brookes, who had himself excavated a well within the town in 1912 (Brooke 1920).
The large assemblage of Romano-British material recovered during iron ore extraction at the Westbury Iron Works between 1877 and 1882 is evidence for a further significant, yet poorly understood, settlement occupied during this period (Cunnington and Goddard 1934: 175ff). The assemblage contains a large collection of ceramics, particularly early Roman, as well as iron tools, jewellery, and a collection of first or second century copper alloy vessels which probably derive from a single deposit comparable to the Kingston Deverill Hoard. Elsewhere in Westbury, excavations by amateur archaeologist Lt.-Col. Shaw between 1959-1964, discovered further evidence for Romano-British settlement at Wellhead Lane (Rogers and Roddham 1991). Shaw’s excavations were unsystematic and the extant material is extremely selective, yet a sizable assemblage remains.
The Museum also holds the archives derived form the excavations of a number of villas around the county, and in particular the archives from Castle Copse, Great Bedwyn (Hostetter and Howe 1997), Atworth (Erskine and Ellis 2008), and the villa or possible shrine at Box (Brakspear 1904; Hurst et al. 1987). Collectively, these represent a valuable collection of material, however, there are issues with all of the archives. In particular, no animal bone is extant form the excavations of Castle Copse, whereas the animal bone assemblage from Atworth had already become mixed by the early 1990s (Bourdillion 1991), limiting its usefulness. Similarly, with the exception of the Samian, the location of the 24,000 sherds of pottery excavated at Atworth (Timby in Erskine and Ellis 2008), one of the largest assemblages in the county, is unknown. Despite this, an interesting collection of small finds, ceramics, glass and structural stonework is extant from the three sites. The Museum also holds finds assemblages from Time Team’s limited excavations at Tockenham Villa, including sculpted stone spout from a probable bathhouse, and Swindon Archaeological Society’s unpublished excavations at Stanton Park Villa. Finally, the Museum also holds the extremely large assemblage from the excavations of Littlecote Roman Villa, 1978-91, however, the documentation associated with this assemblage is minimal, and a considerable project would be required to make the archive accessible.
The other material dating to this period largely derives from relatively small-scale excavations of rural settlements, with few notable exceptions. As with the above, the quality of the contextual information, and the completeness of the archives varies hugely from site to site. Perhaps two of the most notorious in this regard are Nan Kivell’s excavations of Stockton Earthworks (Nan Kivell 1926b) and Cold Kitchen Hill (Nan Kivell 1925; 1926). The former is classified as a as a nucleated agricultural settlement, or ‘village’, by the Rural Settlement of Roman Britain project (Allen et al. 2018), and Cold Kitchen Hill has already been mentioned for its Early Iron Age occupation, but in the Roman period, also became the site of a suspected shrine or temple. Large, but poorly documented finds assemblages survive from both sites, but unfortunately, the archaeological recording of these sites was below the standard even of the day, and little survives beyond sketches of notable finds, either published or in the Museum archives. In addition to the material excavated by Nan Kivell at Cold Kitchen Hill, the Museum also holds a number of collections related to the site, including the artefacts derived from an earlier excavation by Goddard (1894). A huge collection of brooches in particular attributed to Cold Kitchen Hill.
Unsurprisingly, the Museum also holds material deriving Maud and Ben Cunnington’s excavations of hill forts and similar monuments in the early 20th century which can be dated to this period. Late Iron Age and Roman period occupation was identified at Withy Copse, on the north side of Martinshell Hill (Cunnington 1909; 1910b), Casterley Camp (Cunnington and Cunnington 1913), and Knap Hill, although the evidence from the latter site is later than the former two (Cunnington 1911; Conah 1965). A further poorly understood assemblage derives from Shaw Mellor’s excavations of three large barrows in Colerne Park in the north west of the county (Shaw Mellor 1954). The barrows, whose exact location are not now known, were found to be of Roman date, with large quantities of material from the first to fourth century recorded. Shaw Mellor (1954: 338) speculated that the largest of the mounds may have served of the foundation for a building, but was ultimately unable to explain the site. The RSRB tentatively classifies the site as a shrine, although with little confidence (Allen et al. 2018). The extant assemblage comprises of ferrous and non-ferrous metalwork and a poorly- quantified ceramic assemblage.
Relatively more recent excavations include the substantial collection of material attributed to Fowler’s excavations of Overton Down XII, in 1966-8, a fourth to fifth century rural settlement at which a number of buildings were identified (Fowler 2000). A substantial portion of the c. 11,000 sherds recovered during the excavations are extant, as is a large collection of vessel glass and glass beads. Unfortunately, none of the animal bone from the site seems to have been retained. At Chapperton Down, an extensive nucleated settlement was partially excavated by Wessex Archaeology (Malim and Martin 2007). The site was constructed along the line of a major trackway between Bath and Salisbury, and occupied over the third and fourth centuries; the Museum holds a substantial finds assemblage from the site, and as would be expected from a modern excavation, this includes the animal bone, environmental samples and ferrous metal which has not consistently survived from earlier excavations. A small assemblage of both ceramics and animal bone also survives from Wessex Archaeology and Time Team’s excavations at Beach’s Barn, a rural settlement occupied throughout the Romano-British period, but preceded by Middle- Late Iron Age occupation (Wessex Archaeology 1995; Harding 2007). Roman settlement was identified at Black Furlong, Calne, where a late Roman corndrier/malting oven was excavated, along with a pair of fourth century graves (Phillips 2010). The site is suspected to have had a religious function based on the presence of a possible coin hoard, a number of unusual brooches and miniature socketed axeheads found previously on the site (Moorhead 2001; 2010). A short distance away, a potentially significant deposit of pewter and ceramic vessels was also discovered (WAM 1989; Partridge, in press). Further assemblages associated with rural settlement include those from excavations at Easton Grey, Bratton, near Westbury (Luckett 1981), and Cumberwell (Amadio et al 2011), as well as largely unpublished assemblages from Manor Farm, Allington, and Ashton Keyes (WANHM 1972: 173), all of which with very limited publication and discussion.
Finally, the Museum also holds assemblages derived from the excavations of the Minety and Brinkworth tile kilns, as well as the early Roman pottery kilns at Column Ride, associated with the Savernake ceramic industry.