THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 



development at Paul's Dene (SU 14353225). The 

 finds were made some 450m SE of the Old Sarum 

 East Gate on Bishopdown hillside (Stone and Algar 

 1955). Despite these finds the paper concluded that 

 there was still no evidence of extensive occupation 

 in the Old Sarum area but suggested that the south- 

 western slopes of Bishopdown had attracted 

 settlement in the form of a few isolated dwellings 

 during the late 3rd and 4th centuries. The 1955 

 review was closely followed by an excavation in 1 957 

 inside Old Sarum, aimed primarily at elucidating 

 the continuity of the curtain wall around the outer 

 bailey, but as a by-product considerable quantities 

 of Roman period material were uncovered. The 

 excavation results indicated that occupation of Old 

 Sarum had begun early in the Roman period and 

 was of some consequence, lasting into at least the 

 3rd century AD (Rahtz and Musty 1958-60). In 

 the opinion of the excavators the evidence re- 

 opened the question of the exact location of 

 Sorviodunum. Also in 1957 a water pipeline trench 

 on Bishopdown revealed a number of Roman pits 

 but no structures (Musty 1959, 179). 



The pace of finds quickened during the 1960s 

 and 1970s when a number of excavations occurred 

 in Stratford-sub-Castle. A gas pipeline trench in 

 1969 revealed significant lst-4th century AD 

 occupation levels and building material scattered 

 over a distance of more than 200m (Algar 1970a). 

 In 1977, following earlier trial trenches across the 

 Roman road to Dorchester 1 (Musty 1958; Stratton 

 1965, 1966), part of a substantial building was 

 uncovered abutting the thoroughfare. 

 Unfortunately detailed results of the 1969 and 1977 

 excavations were never published, the latter due to 

 the untimely death of John Stratton the local 

 archaeologist in charge. One of the aims of the 

 present paper is to try and redress this omission by 

 providing a comprehensive summary of the results 

 (see Appendix). 



In 1983, as an outcome of a seminar on 

 archaeology and planning, held in the Salisbury 

 Museum in July 1982 for officers of the Salisbury 

 District Council, the Wiltshire Library and Museum 

 Service produced an excellent appraisal of the 

 information then available on the Wiltshire 

 Archaeological Sites and Monuments Records 

 (Borthwick and Chandler 1984). 2 The authors 

 pointed out that while the evidence for Roman 

 occupation in the Salisbury area was poor there had 

 been sufficient finds to suggest that Sorviodunum 

 could be located at any one of three sites. These 

 were Old Sarum; Bishopdown, (spreading SE from 



the East Gate of Old Sarum); or Stratford-sub- 

 Castle, where finds and land boundaries suggested 

 a possible Roman small town occupying 

 approximately 16ha. 



Finally almost 20 years later, in 2001, the 

 Wiltshire Archaeological and Natural History 

 Society has produced a very comprehensive and 

 valuable collection of papers in honour of Ken 

 Annable entitled Roman Wiltshire and After (Ellis 

 2001). A particularly apposite contribution comes 

 from Mark Corney who examines the nucleated 

 settlements across the county. Using evidence from 

 aerial photographs and excavations he concludes 

 that Sorviodunum was a roadside settlement, rather 

 than a town, with a core area extending for a 

 distance of approximately 500m to the north-east 

 of the River Avon giving a potential area of at least 

 lOha.The author felt there was the possibility that 

 Sorviodunum could have been an urban settlement 

 similar to those at Cunetio (Mildenhall) and 

 Durocornovium (Wanborough); however further 

 evidence was needed (Corney 2001, 5-38). 



Despite the publication of Roman Wiltshire and 

 After it is clear from current literature that the 

 majority of archaeologists are, in general, unaware 

 of the numerous finds. The principal aim of this 

 paper is to provide a sharper focus on the evidence 

 for Roman Sorviodunum, to try and establish its 

 location, extent, type of settlement, and how long 

 it was in existence. The need to carry out a synthesis 

 of all the archaeological data currently available is 

 widely recognized, particularly by the Wiltshire 

 County Archaeology Department who have 

 provided much help and encouragement in writing 

 this paper. 



THE EVIDENCE 



The archaeological evidence falls broadly into five 

 categories, viz. excavations; field walking; chance 

 finds; aerial photographs; and historical land 

 boundaries. The evidence in each group is described 

 below and analysed in more detail in the discussion 

 section. A chronological summary of the finds is 

 given in Table 1 . 



Excavations 



There are problems of definition when dealing with 

 'excavations', consequently it must be stressed that 

 the events and results described below include 



