Discoveries at Reculver, 1955-57

DISCOVERIES AT RECULVER 1955-57 By BRIAN J. PHILP THE recovery of material disturbed by sea erosion west of the fort was continued by the writer in 1955.1 The extension of the sea defences in the area by some 700 feet in September, 1957, however, terminated the need for further rescue work. Thanks are due to Mr. Brian KeweU for his assistance with the excavations on the foreshore during 1957 and also to the staff of the GuUdhaU Museum, London, for their help concerning the finds. WELLS Five more weUs were discovered during the period bringing the total now recorded to twelve. Well H. Situated 520 feet west of the fort and 17 feet from the chff face. OriginaUy 15 feet in depth of which only the lower 4 feet remained intact. It was rectangular in shape measuring 50 by 40 inches at beach level narrowing to 34 by 32 inches at the bottom. A shaUow clay lining puddled the shghtly cupped bottom and a few inches of the sides. The vegetable fiUing characteristic of the other weUs was present and contained the usual mass of bones, wood, tile and pottery. Samian forms 31(2), 33, 37, 45(2) and a few coarse ware rims of late-second/early-third century date were the earhest datable finds from the weU. Parts of five third century vessels and a fourth century mortarium suggest the weU to have remained open for some considerable period. Well J. About 1,000 feet from the fort and some 55 feet from the chff. The position of this well was pointed out to the writer after it had been completely washed away. It had been rectangular and not more than 12 inches deep when found. It was from this weU that our member, Mr. A. W. Jan, recovered the " Black Samian" vessel decorated with the cut-glass technique, dated late second or early-third century.2 Well K. 1,120 feet west of the fort and just 2 feet from the chff. It is the most westerly of the weUs discovered and is the deepest owing to the rise of ground in that direction. Of its origmal 20 feet (23 feet to present ground level) only the lower 88 inches remained 1 Detailed report on findings for 1952-54 in Arch. Cant., LXXI (1967), 167. 2 Illustrated in Journal of Soman Studies, XLVII (1957), PI. XV, No. 3, and now in the British Museum. 160 DISCOVERIES AT RECULVER, 1955-57 intact. At this depth it was rectangular measuring 58 by 42 inches ; 40 inches lower it had narrowed to 52 by 33 inches and was reduced to 30 by 18 inches at the bottom. At between 18 and 36 inches the sides of the shaft were badly blackened where rubbish in the weU had been burnt. Water had been drawn from this well by means of counter-weights. By this system the weU-rope, weighted by heavy stones, was attached to a framework at the bottom of the weU. When not in use the weights would draw in slack rope equivalent to the weU's depth (here 20 feet). The other end of the rope (no doubt with a metal hook attachment), would then be drawn to the top and secured. When functioning, a vessel would be hooked on to this end and lowered by pulling against the weights which would rise. The vessel filled upon reaching waterlevel whereupon the rope would be released, the weights drop, and the water drawn to the surface without further effort. The weights and frame were placed at the bottom of the shaft in order to avoid any possible collision with the moving water-vessel. The reconstruction which has been conjectured (Fig. 1) is based upon the interpretation of the available evidence, as hsted below. 1. 40 inches from the bottom of the weU and covering just half of it, was the impression (half an inch deep) of a rectangular framework plugged into the shaft at its corners. In aU probabihty this had been of wood though no vestige of such remained. 2. Shghtly above the frame were cut, actuaUy behind the surface of the Sandstone, two short channels several inches long and half an inch in diameter acting as rope-guides. 3. Four weights, two of flint and two of Sandstone, with holes drUled through them and showing signs of rope wear were found in the weU. 4. A two-handled vessel (Fig. 2, No. 1) one of those used for drawing water, was found intact at the bottom of the weU. On one handle was fixed an earthenware ring, probably the survivor of a pair, from which the pot had been suspended. Each handle, including the ring on the one side, was embraced by an irregular mass of oxidised iron. This indicated that some form of iron-wire had joined the handles; this had broken whilst the vessel was being drawn up the shaft leaving it beyond recovery. A smaU piece of the pot's rim had been broken off in antiquity and signs of wear indicated that the gap thus left was used as a spout. The weight of the pot and its contents when fuU (capacity 3£ pints), would have totaUed. in all 6 pounds. The stone weights were certainly in excess of 12 pounds and could with the aid of simple puUeys have raised considerably more. Other finds from the weU were few; pieces of lava quern-stone, a little pottery and the skeleton of a smaU dog. 161 DISCOVERIES AT RECULVER, 1955-57 A pie-dish rim of Antonine date and another of the early third century, the Antonine pinched-neck flagon (Fig. 2, No. 2), and the second century two-handled water-drawing vessel indicate the well to have been in use at the end of the second century. GROONb LEVEL UWIUWWUIVWUWUWI' to«.U'**AAH4W'*'wjwt ^ > COWEaVREO BEAM AT WELL- HEAD \ fi I J t ft I THANET I 1 *-\ WELL R0P SAND 0 \- V / WATER'LEMEL. / I I FRAME ROPE GU\t>EyS . CORNER PLUS 20 Fig. 1. Well K showing (conjectured) counter-weight method of drawing water. Well L. Situated only 480 feet from the fort and 18 feet from the cliff face. Rectangular, measuring 53 by 36 inches with a depth of 30 inches. The tacky vegetable filling was again present and a hning of clay (9 inches deep) lined the bottom and the sides to a thickness of 1 inch. Although the weU had previously been excavated, by persons unknown, several potsherds were recovered. Samian (Form 31 (?)), a fragment of early Rhenish beaker and part of a rouletted beaker aU date to the end of the second century. 162 DISCOVERIES AT RECULVER, 1955-57 Well M. This weU was the nearest to the fort of the whole group being a mere 420 feet from the west waU. It was located against the face of the cliff and of its original depth of 15 feet, only 56 inches of filhng remained intact. Unlike the others it had been cut circular with Vo i 2 X / L \. 3 y,;:/tfi'W. !'< Fig. 2. Coarse pottery from Well K (Nos. 1 and 2) and Well M (No. 3). (i). a diameter of 43 inches. Vegetable sUt and clay lining were absent. The bottom 35 inches of the filling contained burnt matter, debris, pottery, bone and a smaU bronze ring. A complete red-coated flask (Fig. 2, No. 3) had been dropped into the weU and rested near the 163 DISCOVERIES AT RECULVER, 1955-57 bottom. Unfortunately it had been smashed in antiquity by a weUaimed Sandstone block, found embedded in the pieces. Samian, forms 31,43, and a base (31 ?) and two coarse ware rims date this lower filling to the second half of the second century; a date also apphcable to the smashed flask. The debris in the bottom of the weU was sealed by various layers of sand probably coUapse from the sides of the shaft. In this sand at a depth of 4 | feet was an iron buckle and the skeleton of a young dog which had very hkely faUen into the weU. A coin with an obverse of Gordianus I I I (A.D. 238-44) but minted about A.D. 250 (hybrid type), was found at the same level suggesting that the weU had long been in disuse by that date. When the erosion had first uncovered the well a skeleton of a woman complete with shale and bronze ornaments had been revealed in the upper part of the well shaft.1 Such an unorthodox resting place excludes the likelihood of it being a burial and even if the woman had faUen in the weU her body would have been recovered for more customary disposal. It can only be concluded therefore that the woman had been killed and her body thrown into the disused shaft. The coin of c. A.D. 250 found at a lower level proves this event to have taken place after that date. The possibUity that the woman was a victim of Saxon raiders must therefore be considered. RUBBISH PITS AND HEARTHS Four more shaUow pits were revealed in the cliff and examined. One (Pit 6), produced pottery mainly of third and fourth century date and also contained a worn first century native rim, clearly a stray from the Belgic settlement area. Other pits (Nos. 7 and 8) contained pottery of Antonine date and pieces of imported lava-stone. A fourth pit (Pit 9) contained pottery of the late third or early fourth century. Three shaUow hearths were noted in the area but only one, used for smelting, contained pottery. This hearth (No. 2), measured 41 inches in length and 12 inches in depth, contained numerous flakes and fused masses of bronze and iron. A damaged bronze finger-ring was recovered from the ash and vessel-glass, Samian (Form 31) and a pie-dish rim had been thrown in after the fire was out. The pottery indicates the hearth to have been used at the end of the second century. During the construction of the sea-waU several feet of top-soU on the chff edge were mechanically removed. No trace of any structure was then revealed along the entire length of the new waU but several pits and hearths had been cut through. The Road into the Fort. Local topography indicates that the road 1 Details of the discovery are at present held by the finder, Mr. P. A. Clayton. 164 DISCOVERIES AT RECULVER, 1955-57 approach to the fort was from the west. A map of Reculver dated 1685 shows a road " The Kings highe Way," as terminating at a gap in the centre of the west waU of the fort and there can be httle doubt that here at least it foUowed the course of its Roman predecessor. The first indication of the Roman road outside the fort was revealed in 1953. At about between 850 and 950 feet from the fort there extended a pebble and gravel band some 12-16 inches below the surface. This was later destroyed as the cliff feU but in 1958 the metaUing was stUl visible at between 1,120 and 1,164 feet from the fort. The stretch first seen must have run almost paraUel to the cliff face but the second, owing to further erosion, entered the chff obhquely at an angle of about 30 degrees. The fact that this metaUing sealed the smelting hearth indicates that it was not put down before the end of the second century. An extended line through the two sections strikes the fort waU north of the central west gate and at an angle to it. This indicates that the Roman road left the fort, crossed the ditches by means of the causeway, and continued straight for another 100 feet or more before bearing some 15/25 degrees to the south. From there it crossed several fields to HiUborough and then via Sturry to Canterbury, a distance of about nine mUes. The Anchorage. An extension of the marsh-levels south of the fort has often been regarded as the site of the Roman harbour. The storms of early 1953 breached the sea-waU and inundated the marshes, the flood water foUowing the course of the Roman channel. The area south of the fort was also flooded and it may be concluded therefore that it was simUarly covered in Roman times. Such an expanse of water would have provided a safe anchorage for a number of ships, a fact which no doubt determined the site of the coastal fort. THE POTTERY Well K. 1. (Fig. 2, No. 1.) Two-handled vessel of sandy white ware ; so-caUed. " Honey p o t " type. Flattened rim and external median groove. An earthenware ring, no doubt the survivor of a pair, encircles one of the handles. A mass of oxidised iron on each handle showed where they had been joined by an iron wire-like handle. With this addition the pot could be used in the weU for drawing water. The rings on the handles are a rare feature and only two other examples have been noted, one in the British Museum and the other at Chesters (Northumberland). " Honey pots " are normaUy more bulbous than the present vessel and those known mostly date from the first century. However, the 165 DISCOVERIES AT RECULVER, 1955-57 type had a long life and here the flattened profile is consistent with a second century dating. 2. (Fig. 2, No. 2.) Pinched-neck flagon of hard white ware with two-ribbed handle. These vessels date to the first and second centuries and are not particularly common. At Richborough they are dated A.D. 50-80; South Carlton (Lines.) A.D. 140-180; Verulamium Had./ Antonine and at Canterbury to the second half of the second century.1 The present example is therefore hkely to be of Antonine date. Well M. 3. (Fig. 2, No. 3.) Bulbous flagon of hard grey ware with red shp. Belgic type with three distinct cordons on the neck. Five white painted bands are partly superimposed on the cordons. The' vessel's profile is common to many of the burial vessels from the Ospringe cemetery, there generaUy dated to the second half of the second century. 1 For details of this type see note by F. Jenkins, F.S.A., in the 1956 volume of the Antiquaries Journal, 48. 166