16 



THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 



for supporting staff, camp followers and people 

 trading goods. Some of these individuals would have 

 come from the surrounding Iron Age settlements, 

 and others from further afield. Once established 

 the growth of a settlement would have depended 

 principally on economic forces (Wacher 1998, 102). 

 Other factors would have included the level of 

 military presence, construction of the London to 

 Dorchester road, and the amount of official activity 

 taking place in the area (Burnham and Wacher 

 1990). The apparent laying out of a street grid in 

 the settlement suggests at least some degree of 

 Romanization and deliberate planning which is 

 seldom seen at less important settlements 

 (Burnham 1987, 167). 



By the 2nd century many small and most large 

 towns had constructed an earthen and/or a walled 

 defence around part of the settlement (Millett 

 1 990). In the case of Sorviodunum the more easily 

 defendable area of Old Sarum would have been 

 invaluable during times of trouble or unrest. So far 

 no remains of an earth bank have been discovered 

 in Stratford-sub-Castle although there is some 

 evidence of ditches c.6m wide on the north and 

 south sides of the settlement (see Appendix below). 

 These are in addition to the two features identified 

 on aerial photographs as ditches by Corney (2001) 

 and mentioned above (Figure 3, d and e).The close 

 proximity of Old Sarum, Bishopdown, and 

 Stratford-sub-Castle to each other (Figure 1) 

 coupled with the fact that Old Sarum and Stratford- 

 sub-Castle both lie alongside the Silchester to 

 Dorchester Roman road leads to the conclusion that 

 all three 'sites' formed part of the Sorviodunum 

 settlement. 



The overall size of Sorviodunum could well have 

 changed during the course of the Roman period. 

 Most of the larger 'small' towns appear to have had 

 an area of between 4 and 14ha enclosed by a 

 defensive earth bank or wall (Todd 1970, 116; 

 Millett 1990, table 6. 4). The maximum area of land 

 available for settlement inside the Old Sarum 

 earthwork is 12ha. For the remainder of 

 Sorviodunum the evidence currently available 

 suggests Romano-British occupation in Stratford- 

 sub-Castle covered at least 1 6ha and may have been 

 as much as 25ha while that on Bishopdown was 

 8ha or more. These figures could well be an 

 underestimate of the actual size, but without further 

 excavations or geophysical investigations it is 

 impossible to be sure. Work elsewhere has shown 

 that the extent of extra-mural settlement can be 

 considerable. For example the 'small' Roman- 



British town of Durobrivae in Cambridgeshire 

 stretches for some 3km along Ermine Street when 

 extra-mural areas are taken into account (Mackreth 

 1995). Sorviodunum might have had additional 

 extra-mural suburbs, possibly on the land to the 

 NE of Old Sarum" or the areas covered by the 

 settlement site at Highfield 0.6km south (Stevens 

 1934) and the one in the vicinity of Moberly and 

 Netheravon Roads 1.3km SE (Stone and Algar 

 1955, 110). The latter two sites could well have 

 extended further than the areas already discovered. 

 At this stage it is not possible to determine if the 

 intra-mural area of the town was inside Old Sarum 

 or at Stratford-sub-Castle or whether it was at both. 



End of the Roman Period 



So far no evidence has been found to determine 

 the fate of Sorviodunum at the end of the Roman 

 period. Opinion is divided as to the likely scenario 

 in Britain after military withdrawal. Some experts 

 believe that there was almost complete collapse in 

 the face of Saxon incursions into Britain (Esmonde 

 Cleary 1989). Others postulate that the way of life 

 in existence at the end of the 4th century continued 

 with little change well into the Saxon period (Dark 

 2000). Corney (2001, 22) has pointed out the fact 

 that Sorviodunum is one of the few places in 

 Wiltshire which features in the historical Anglo- 

 Saxon record 12 implying there was still a sub-Roman 

 population and authority in the Salisbury area in 

 the mid 6th century AD. Excavations at Butterfield 

 Down, 10km to the north, indicate that the 'village' 

 was still functioning at the beginning of the 5 th 

 century with 9% of the coins coming from the 

 period AD 388-402 (Rawlings and Fitzpatrick 

 1996). The coin loss for the same period at 

 Sorviodunum was nearly 7% suggesting that 

 occupation also continued into the early 5th 

 century. 



WAY AHEAD 



The importance of towns, particularly those that 

 have until recently been classed as 'small', in Roman 

 Britain has been stressed on many occasions, most 

 recently in Brown (1995). Very few sites have been 

 extensively investigated even when they are still open 

 fields. One of the major problems is that vernacular 

 buildings in small towns often leave little trace of 

 floors or foundations. This is because the depth of 



