A New Roman Site Discovered at Rochester

A NEW ROMAN SITE DISCOVERED AT ROCHESTER BRIAN PHILP This site (TQ 7452 6218) lies behind shops and properties which flank the angle formed by Victoria Street and Eastgate (Fig. 1). In 1990 this was derelict land awaiting a possible development known as the Havisham Centre. The planning consent required an evaluation excavation and this was carried out in March 1990. As this produced important new evidence, an area excavation was also undertaken by KARU in the following month. Significantly, the site lies well outside the Roman town of Durobrivae, about 300m south of its East Gate. It probably lies close to the junction of two Roman roads leaving Rochester, one (Margary Route lB) to Dover and the other (Route 13) towards Maidstone and the Weald. The underlying soil is mostly an orange brickearth and the ground level elevation is 5-6m OD. THE EXCAVATION The initial evaluation consisted of a 47m long trench (A/B) down the main SW-NE axis of the site. A 6m spur trench (C) was cut on its south side (Fig. 1). Few features were found in the south-west half of Trench A, but concentrated features were found in the north-east half. It was here that the original trench was quadrupled in width to occupy most of the available space (Fig. 2). A total of about 200sq.m were thus excavated, to a depth of l -2m. The work identified 14 7 individual contexts, 103 features and a total of 7,739 finds. The features include four Victorian pits and brick tanks, 12 postmedieval pits and 87 Roman features, the latter including postholes, gullies and pits. (The basic data on all of these is shown in Appendix 1). The site was backfilled on completion as the development failed to go ahead. 203 London • Canterbury IO 1G ,o » «>... ------=--.,,,,,,-, 􀀄 O a 􀀅 M 􀀆 BRIAN PHILP Om 10 50 Oft 100 Fig. 1 Site Location. and plan of excavated area. The Roman Features Of the 87 features which can reasonably be regarded as Roman, nine relate to a probable East Post-line, of which four posts can be identified [F39, F 31, F30 and F 24 ]. Four more features seem to relate to the West Post-line, of which two posts can be identified [F83 and F9 l] and two others may have been removed by later pits [F61 and 204 __ J 0m Rochester. Havisham Centre 1990 Plan of features in Main Area 5 Fig. 2 Plan of Features in Main Area. 20 BRIAN PHILP F70]. These two lines of post-holes contained over 100 potsherds, mostly of late first- or second-century date. They appear to be late in the Roman sequence on the site (Catalogue No. 1). Nine Roman features appear as shallow, linear gullies and cross the centre of the site. Of these, F76 and F32 seem to join at a right angle. So, too, may F80 and F25; the others are randomly arranged [F6, F21, F36, F47 and F48J. As regards dating, about 60 potsherds were recovered from the four associated gullies and the latest are dated to the second half of the second century (Nos 2-3). One gully [F32] was cut through by two of the post-holes of the East line. The random gullies produced another 160 potsherds (Nos 4-8) and mostly seem to date from the end of the first century. Thirty of the Roman features have been identified as post-holes and these are mostly 18-45cm in diameter and 10-26cm in depth. Several of the larger ones could in fact be small pits. Most of the post-holes seem randomly placed though it is possible to suggest short patterns. Seven fall on the line of two parallel gullies and may relate to a rectangular structure. Four of these [F64, F89, F90 and F85] are positioned along the south gully [F76]; three more [F86, F87 and F88] fall along the north gully [F80]. These help reinforce the view that the gullies outline a timber-framed structure. Another 34 Roman features are regarded as pits and these range in size from about 0.40-3.00m. Six have a length less than 60cm, 13 are up to l.00m, nine up to 2.00m and six have a length greater than 2.00m. Of these 18 are 0.40-l.75m in depth (Nos 9-11, 43-44, 46-47, 51). In terms of distribution a large cluster of 13 pits occupy the southwest corner of the site. Others are generally scattered about, though two large ones [F49 and F98] occupy the south-east corner and two more large pits [F92 and F93] occupy the north end. These are immediately adjacent to each other and one (F92J contained over 300 potsherds, including samian, mostly dating to the second half of the second century. The other [F93] contained over 50 potsherds, dating to the middle of the second century. The fill of each had contracted and the overlying layers had slumped into them, including Chalk Deposit G. It is clear that some pits cut earlier pits, but their precise relationship across the whole site is difficult to determine due to the limited extent of some of the horizontal stratification. Even so, some of the pits occur at different levels and a wide date-range within the Roman period is indicated. The Roman Horizontal Features and Related Stratigraphy (Figs 3-8; Sections l-16) A large part of the centre of the site was found to be covered by a sequence of laid formations and dumped soils. Nowhere did these 206 Nortn-Eaat Sect. 1 Chalk(G) N Key 􀀂] Loams 􀀁 Clay 􀀂 Chalk Sect. 2 l!􀀅•I Flint Mortar 􀀃 (;) Oyster Chalk wash Om I l Fig. 3 Sections I and 2. Dep. A B C D E F G/H I J K L M BRIAN PHIL'F TABLE 1. ROMAN HORIZONTAL STRATIGRAPHY (* MIN. DIMENSION) Compos. Length Width Depth Context Illust.