SORVIODUNUM - A REVIEW OF THE ARCHAEOLOGICAL EVIDENCE 



19 



Two further trenches were dug in 1965. Trench 

 C was located in Fisherton Meadow on the west 

 side of the river Avon at SU 13303 160. The agger 

 was located almost immediately. At no point was it 

 more than 0.25mm below the present ground 

 surface. Its surface was of fine rolled flint above 

 river gravel interspersed with larger flints in two 

 layers. The average depth of the agger was 0.56mm 

 and the width 7.32m with a small drainage ditch 

 on each side. On each side at 12.8m from the centre 

 of the agger there were two larger side ditches. 

 Within these outer ditches was a series of chalk 

 floors which ran up to and part way under the agger. 

 The presence of the floors and the occupation 

 debris associated with them (samian and 'Belgic' 

 sherds, nails, and some indeterminate fragments 

 of iron), suggested that the road constructors had 

 lived alongside the road. One coin, an as of 

 Domitian minted in AD 84 and in fine condition, 

 was most probably lost in the last quarter of the 1st 

 century. 



Trench D was at SU 13403175 in the centre of 

 the meadow, which had been the site of the first 

 trench (A) . Here the height and width of the agger 

 and the separation of the side ditches corresponded 

 to those found in trench C. Again there were large 

 quantities of sherds but in contrast to the pottery 

 from trench C, the date range was from the 1 st to 

 the 4th centuries. Box tile, roofing tile, knapped 

 flint, greensand fragments, and plaster were also 

 found suggesting the presence of a substantial 

 building close to the road. 



A sectioning of the bank at SU 13503170 on 

 the line of the Roman road shown on the Ordnance 

 Survey maps revealed that it was an earthen bank 

 associated with three 1 8th century cottages; there 

 was also a stone conduit of similar date. Apart from 

 a single very abraded samian sherd all the pottery 

 was of the 18th century. Recent housing 

 development at Castle Keep Estate on part of the 

 meadow in which trenches A and D were located 

 revealed much evidence of settlement, chalk floors, 

 a cobbled area and a small oven. A length of Roman 

 road was also exposed. The finds included lst/2nd 

 century samian ware and coarse pottery (mainly 

 Belgic and Durotrigic derived wares, but including 

 a small quantity of 3rd century New Forest wares), 

 two very worn coins, probably of the late 1st century, 

 Purbeck stone and ceramic roofing tiles. 



There were numerous fragments of roof and 

 hypocaust tiles, and a small slab of polished Purbeck 

 marble. Samian ware included a number of early 

 forms of Flavian date and a retrograde stamp 



BELINICI M (Belliniccus of Lezoux, fl. Trajan- 

 Antonine) . Other pottery was probably of 1 st/early 

 2nd century; only two sherds of New Forest ware 

 were found. Subsequently further samian ware, 

 including a nearly complete platter of Dragendorf 

 Form 32 and the rare potter's stamp CRESIMI 

 (Cresimus of Montans, fl. AD 80-120), were 

 discovered. 



1969 NORTH SEA GAS PIPELINE 

 TRENCH 



In 1969 the meadow to the west of the Castle Keep 

 Estate was crossed by a North Sea Gas pipeline 

 trench and as a result the Romano-British 

 settlement bordering the Roman road was sectioned 

 from north to south over a distance of about 100m 

 (SU 13363 187- 13473 170). Along 65m were multi- 

 phased tiled and walled structures of timber and 

 flint, some with plastered walls and with at least 

 four superimposed floor levels. Building debris was 

 found scattered for over 200m. A large amount of 

 pottery (see list below) from the lst-4th centuries 

 AD was recovered including New Forest and 

 Oxfordshire wares, an amphora fragment and lead 

 glazed wares. Samian sherds included the forms 

 Drag. 18, 27, 37, 38, and 45 bat's head spout. A 

 dupondius of Nero AD 64-68 and a sestertius of 

 Commodus (Rome mint AD 187-188) were also 

 found. 



A pit, from the north end of the section, 

 approximately 23m north of the Roman road, 

 yielded a 3rd-4th century group (see list below). 

 Part of a tile stamped LHS was found on the spoil 

 heap. The sandy terracotta coloured fabric had a 

 smooth upper surface and underneath were the 

 marks of two areas of mortar: had the tile been 

 complete (260 x 410mm) there would probably 

 have been three. Part of another LHS tile was 

 recorded from the Outer Bailey of Old Sarum 

 (Rahtz and Musty 1960, 366; Darvill 1979, 328, 

 343). The Roman road c.6m in width had buildings 

 right up to its edge which had ultimately collapsed 

 on to it. The site is possibly delimited by ditches 

 approximately 6m in width on the north and south 

 but the water table was too high to be sure. 



List of pottery and other finds based on field 

 notes by Mrs V.G. Swan 



Unstratified Material Recovered from the Trench 

 Samian 



Dragondorf 45 bat's head spout and Drag. 38. Probably 

 late 2nd or early 3rd century. 



