EXCAVATIONS IN 1999 ON LAND ADJACENT TO WAYSIDE FARM, DEVIZES 



207 



belt fitting at Bishops Canning (Guest er al. 1997) 

 and 11km west of an early 5th-century hoard of 

 gold, silver and bronze coins at Stanchester near 

 Pewsey (Paul Robinson in CBA Wessex News, 

 October 2001, 18). 



There appear to be two types of deposits present 

 on the site; first, those which on the basis of the 

 material evidence and by association are or may be 

 of religious or funerary character (e.g. the midden, 

 pit F4225, the burials, the north to south flanking 

 ditches), and secondly; those which may be linked 

 to the above deposits but are more likely to represent 

 evidence for settlement-based activity (the ovens, 

 ditches and the remaining cut features). 



Deposits of religious or funerary 

 character 



The type and quality of some of the artefacts 

 recovered from these deposits taken as a group are 

 more characteristic of ritual or religious beliefs and 

 could be considered as votive offerings rather than 

 the typical range of material recovered from 

 settlement sites of this date. Specific objects would 

 in particular include the curse tablet (Plate 6) and 

 the garment collar (Figure 14.02). Many of the 

 contributors to this report have highlighted aspects 

 of the finds assemblages which may be atypical of 

 the range or composition of collections from 

 unequivocally domestic sites. In addition to the 

 coins, the high proportion of iron objects to nails 

 and the presence of items such as an iron spoon, 

 stylus and hipposandal indicate a site of some status 

 and possibly specialised (ritual) function. In this 

 context, the high proportion of pottery fine wares 

 and animal bone butchery waste, including skulls, 

 may also be indicators of non-domestic activity. 

 Wayside Farm may lie close to a site or building 

 with a more specialised function such as a temple 

 or a shrine, as it is not unusual for such places to 

 have associated pit or midden deposits. However, 

 although no structure of this date was identified at 

 Wayside Farm, evidence for late Roman buildings 

 can be notoriously elusive despite abundant finds 

 (Mudd er al. 1999, 148). It cannot be discounted, 

 therefore, that a building of this date may once have 

 been present within the excavated area which did 

 not survive in the archaeological record. This, 

 however, is considered unlikely, with a specialised 

 Roman building immediately to the north of Area 

 2 seeming more probable. 



The midden deposit at Wayside Farm appears 

 to have been demarcated by the north to south 

 aligned ditches F4261, F4288 and F4254, the upper 



fills of each containing midden soils. The majority 

 of the midden deposit was identified within the 

 excavation area, with just a small area seemingly 

 continuing beyond the northern excavation limit. 

 The midden covered an area of c. 1 150m : , with an 

 average thickness of 0.15m; therefore in terms of 

 volume c. 170m 3 of material was present. On the 

 basis of the homogeneous nature of the deposit, 

 the probability that material from the midden was 

 incorporated within the upper fills of ditches F4261, 

 F4288 and F4254 and pit F4225 (see below) and 

 the presence of conjoining pottery from the pit fill 

 and the midden, it is likely that the midden was 

 formed as a result of dumping or spreading over a 

 relatively short space of time; the latest coin from 

 the deposit dates to between AD388-402.This also 

 appears to have been the case at a possible late 

 Romano-British midden site at Colerne Park 

 (Mellor 1954, 337), where the deposit is also 

 described as homogenous. 



Late Roman midden deposits are rare but not 

 unknown in Wiltshire and the surrounding regions 

 either as settlement 'rubbish heaps' or as 

 repositories for votives and other material. Evidence 

 for the former type of deposit has been identified 

 recently during excavations at Weaver's Bridge, near 

 Cricklade (Mudd er al. 1999), where the quantity 

 and quality of the artefacts recovered was far lower. 



Middens as religious deposits are thought to 

 be present and partly excavated at Cold Kitchen 

 Hill, near Maiden Bradley (Goddard 1893, Nan 

 Kivell 1925), and at Colerne Park (Mellor 1954). 

 The midden at Cold Kitchen Hill was located 

 adjacent to a probable Bronze Age burial mound, 

 but significant quantities of Romano-British 

 artefacts were recovered which are comparable in 

 terms of type, quality and date to the Wayside Farm 

 assemblage, although the longevity of that site was 

 far greater. Further afield, there are also cases of 

 Bronze Age burial mounds being used as 

 repositories for votives and other material during 

 the Roman period, for example at the Roman 

 temple on Brean Down, Somerset (ApSimon 

 1 966). The Colerne Park investigation was far more 

 systematic than that at Cold Kitchen Hill and the 

 material recovered was again similar to that from 

 Wayside Farm. As with the current site, no in situ 

 evidence for an associated structure has so far been 

 identified at either Cold Kitchen Hill or Colerne 

 Park. 



Pit F4225 was partly filled with midden material 

 and is a likely component of the probable religious 

 site. Pits within and associated with Romano-British 



