An Early Iron Age Site at Borden

AN EARLY IRON AGE SITE AT BORDEN By F. H. WOBSEOLD, F.S.A. IN 1874 a new house was erected for Mr. Francis Prentis of Milfon near the farm caUed Hart's DeUght, a smaU hamlet on the western confines of TunstaU parish where it adjoins the parish of Borden, close to Sittingbourne. His new home he called " CunobeUne " to mark a find that had been made on the site of the garden of two gold coins of that British King's reign (5 B.C. to A.D. 40-43), together with an aureus of Augustus struck 18 B.O. (see Arch. Cant., XXVII (1905), p. Ixxi). A further gold coin belonging to this treasure was recovered in the same place in 1943. Illustrations of two of the pieces appeared in Arch. Cant., IX (1874), p. 299. At the time of the discovery of the coins, Canon Scott Robertson suggested that they had been secreted by someone from a nearby Celtic Camp. The Canon thought this might have been at TunstaU ; and he gathered a place-name clue therefrom. If such a camp existed at TunstaU aU traces appeared to have been removed by intensive cultivation (see George Payne, Collectanea Cantiana (1893), p. 9). Knowing the configuration of the countryside thereabouts intimately I always felt that there was a far more likely site at Borden. The knoll on which Borden church is built stands at 199 feet O.D. and would be ideal for such a hill-fort, commanding as it does extensive views from the N.W. round to the E. Moreover Milton Creek, up which this last incursion of the Belgse no doubt entered these parts, is in the middle distance. It was not, however, until in 1939 Messrs. WiUs and Packham, Ltd. of Sittingbourne acquired the five acre field immediately adjoining the north side of the viUage Playstol for the removal of brickearth, that my idea of a hiU-fort existing here could be put to the test. Nearly two-thirds of the field had been cleared before any trace of occupation was revealed, except that a few worked flint implements and " pot-boilers " of indeterminate age had been found. Then early in 1943 as the excavators neared the Playstol (see Plan Fig. 1) a V-shaped ditch at Al was cut through (for section see Fig. 2, Al). From this ditch were recovered portions of a Belgic cordoned bowl (see Fig. 3, No. 8). It was the herald of good things to come. From then onwards till the ground was worked out in 1945 the ditches and pits as shown in Fig. 2 were uncovered. They contained hundreds of sherds belonging to the period immediately before the Roman occupation ; while the after presence of the Romans here is attested by four cremation burials (shown as" black spots on the right-hand section of Plan, Fig. 1). The 148 AN EARLY IRON AGE SITE AT BORDEN V?, •••:.••. :c? Sketch PIMI of E&rly Iron Ai[eSil"e. BORDEN.N'Sinln^bourne.KENT. t l t l l l l l l l l I I I 1 sw\ V Ftapole (jlrSen./ O.D. I9S. \ re? The PLystol <**!• FIG. 1. IIIISPWSJIS • •"••-' •••Sealtiiiidifelfool-. •'-'•'.' 'B.g-.'V ••••->,• •:••.:,•/:••••••.••••• HUMUS &&yv>-A-:•>;••:•;.•:••.!. w. | i11l . J,i.iA«r\B»Wrf> ««0?nwswS3!3Ql!S2P^ '•C:^::-::. •:••:. itrmSS^i^SSSP^^r^ ;- • \?T-f*.* -t...'..l-!-'.. —•'.'.".'.'.' ••"'• •' PIG. 2. SECTIONS OE DITCHES AND RUBBISH PITS As numbered on Plan. Dotted areas indicate Brick Earth. 149 AN EARLY IRON AGE SITE AT BORDEN burial over-riding ditch B is indicated in section B2 on Fig. 2. From it were obtained fragments of a much damaged urn, a platter and carinated vase (drawn in Fig. 3 Nos. 13,14 and 15). Materials from the other three burials were too badly damaged to yield the shapes of the pottery. The disconnected nature of the ditches A and B should be noted, but such lack of connection is not at all unusual on Belgic sites. For instance, at Cassington, Oxon, there was a short length only (Ant. Journal, XV (1935), p. 33) and Major AUen in his air-photographs has revealed a number of sites with straight or irregular lengths of ditches in the Thames VaUey area. At Worthy Down unconnected lengths of ditch were also found (see Proc. Hants E.G., X, 178-195). The smaU cross in the stipple of the lower haU of the excavated ground (Fig. 1) marks a flint knapping site. Perhaps the flint implements referred to earher originated here. Many partiaUy-trimmed flint nodules, an anvil-stone, punch, several round hammer-stones, cores and various finished implements, including two large pointed flint handpicks came from this spot. Whilst these remains might all be attributed to NeoUthic times, it is quite possible they were the work of Early Iron Age folk; and it may be recaUed that a good series similar to these was excavated in conjunction with Early Iron Age pottery at the hiUfort at Loughton, Essex, by Mr. Hazzledine Warren (see Essex Naturalist, XXII, pp. 127-136). Two portions of a saddle-quern and a grinding-stone were found close to our flint work. It should be stated that the rectangular space indicated by broken hnes on either side of ditch B (see Plan, Fig. 1) is too evenly deUneated. It represents a spread of broken ware. Between 1943-1945 the last patch of brickearth on the western edge of the ground paraUel with Wise's Lane which leads from Borden down to the London Road, the Old Watling Street (see Plan, Fig. 1), was removed. Signs of occupation soon became manifest here, although in the uppermost layers the ground showed signs of considerable disturbance, medieval glass and pottery sherds being intermixed with Early Iron Age material. Disturbance was speciaUy marked immediately in the rear of the recently destroyed Vesey House. . Below this foUowed a section of pure brick-earth, and a weU-marked V-shaped ditch was encountered (Plan, Fig. 1, C and Section, Fig. 2, No. C4) running N.W. and S.E., that is in the direction of the ditoh B in the other part of the field, but not quite in the same aUgnment. The ditch C yielded an exceUent selection of Belgio sherds, as did a smaU shaUow and roughly chcular pit adjacent (see Plan D, Fig. 1 and Section D5, Fig. 2). Later a further extensive dump of occupation material was worked through between the pit just mentioned and the Playstol N.W. gate (see Plan E, Fig. 1 and Section E6, Fig. 2). 150 AN EARLY IRON AGE SITE AT BORDEN ^ M immmmw/#mm S i ^ V J-*--&slS X i3 ( ( ..' 9 r 10 sit 12 FIG. 3. COARSE POTTERY EROM BORDEN. 151 AN EARLY IRON AGE SITE AT BORDEN This spot is chiefly to be remembered as yielding fragments of curious circular blocks of burnt clay; when pieced together these resembled two large shaUow reels. Photographs of them (Plate 1) have been submitted to various antiquarian coUeagues but none have been able to throw any real hght on the use to which the objects may have been put. One is inclined to surmise that they are of the nature of the " tournette " which Professor Gordon Childe in his Bronze Age, p. 50, mentions as being in use among the Egyptians of Old Kingdom times. The tournette was a disc, pivoted so that it could be revolved at wiU by the feet during the shaping of the pot. This piece of apparatus friends of mine who went through the 1939-45 War say is stiU in use in Egypt and in Cyrenaica as weU. It may be added that in the fabrication of pottery the Maya Indians of Yucatan, Central America, use similar blocks, placing them on the ground between theh knees so that they can revolve them slowly in either direction by using theh feet; this leaves the hands free for shaping the vessel. A graphic photograph of a Maya Indian so employed is shown in the Geographical Magazine, VIII, No. 4 (February, 1939), p. 257 ; it bears this caption, " Pottery made without a wheel recalls the ancient Maya's lack of that invention. With his hands he shapes the vessel while his feet turn the block on which it rests." Our blocks may weU be a forerunner of the true potter's wheel which the Celt introduced here from the Continent. The animal remains from our site were not extensive. Dr. J. WiUrid Jackson, F.S.A., kindly identified those sent to him as " bones and teeth of the Ox, the canines of Pig, tooth of Red Deer and Horse, and the bones of a Dog." He remarks with reference to the last named that both bones and mandible are larger than examples from Colchester (Belgic), that the mandible agrees with some from Salmonsbury (100 B.o. to A.D. 50), and that the metacarpal of the Ox agrees with one also from Salmonsbury, but that the Borden coUection as a whole is too smaU for the purpose of comparison with others. The oyster, whelk, cockle and mussel also occurred in the workings, chiefly in ditch C. No iron work was encountered and the only suggestion of metal was a green stain on a fragment of bone, which may have betokened oxidized bronze. POTTERY. By far the larger proportion of the Borden sherds are of the coarse hand-made type, plain rimmed, and of a corky fabric with a considerable grit content. The surface is duU black and rather soapy to the touch and it gives a " tar-mac " appearance to the vessels. Although plainly of Iron Age " A " derivation, they are without a doubt contemporary with the recognized later wheelmade " 0 " ware, a seleotion from which, together with three of the hand-made pieces, has been drawn in Figs. 3 and 4. The coUeotion as .a whole resembles very nearly 162 *' CLAY " REELS " FROM BORDEN ( x i). AN EARLY IRON AGE SITE AT BORDEN the pottery described by J. B. Ward Perkins from an Early Iron Age Site at Crayford, Kent (see Proc. Prehist. Soe, 1938, pp. 151-168), and only shghtly less closely that recovered by Cotton and Richardson from a Belgic Cremation Site at Stone, Kent (Proc. Prehist. Soe, 1941, pp. 134-141). From my pieces the Swarling pedestal cordoned type of vase is absent, although there are two foot-ring bases from ditch A which may weU belong to this style. The complete profile of only one vessel was recovered from Borden (Fig. 3, No. 4 from pit D). With respect to the coarse hand-made derived " A " ware, various parts of rims are drawn in Fig. 4, Nos. 6-11. These originated from ditch B and from the adjoining spread. Belonging to the same group (Fig. 3, No. 1) is shown a reconstructed " flower pot " shaped vessel with " comb " decoration ; and (No. 2) immediately below is a " sauce-pan " shaped pot exhibiting faint vertical lines rather widely spread. No. 11 is a smaU bowl also of this " A " type with one finger groove below its black " tar-mac " neck, set on a matt-surfaced body. Coming to the " C " ware, the three wheel-made cordoned bowls depicted (Fig. 3, Nos. 3, 4, 8) are aU of the typical late Belgic hard grey paste. No. 3 from ditch A has two cordons above a zone of horizontal lines which are used as body decoration, and there is an (inferred) foot ring in the figure, added to conform with that of a similar bowl from Stone. No. 4 has but one cordon ; it has a flat base and is decorated on the body with a wide faint trelhs pattern. This came from pit D as did No. 8. The taU wheel-made finger-grooved vase, No. 6 was from ditch B. It is shghtly gritted and burnt to a light briek-red and is almost identical in character with one from Crayford (see Proc. Prehist. Soe, 1938, p. 160, Fig. 6, No. 1). The barrel-shaped cordoned wheel-made vessel No. 7, also from ditch B, closely resembles one from Dumpton Gap near Broadstahs, Kent (see Arch. Cant., XXX, (1914), pp. 309-312) from which the (inferred) beadrim and base of ours are copied. No. 12 is a taU wheel-made jar from pit D, paste very duU red and gritted, with one shght cordon on its short neck and bearing on the shoulder a very irregular chevron pattern. The base could not be determined, but if it were like that of an urn shown in Proc. Prehist. Soe, 1942, p. 137, Fig. 2, No. 7, the two vessels would strongly resemble one another ; the date assigned by Cotton and Richardson to the vessel they figure is late first century B.o. Nos. 9 •and 10 are two bead-rim bowls. The former is decorated with rather broad horizontal lines on the shoulder and has similar vertical lines in zones on the body, The paste is a soft blue-grey clay. The other vessel is made of heavily gritted dark grey paste; its body is ornamented with a series of four enchchng lightly drawn lines horizontaUy spaced four inches apart. Both these vessels are wheel-made and they came from ditch B. Although sherds for both rim and base of No. 5 were 153 AN EARLY IRON AGE SITE AT BORDEN missing, measurement of those belonging to the body revealed a shape not unlike an omphalos bowl found at Canewdon (see Proc. Prehist. Soe, 1938, p. 164, Fig. 10, No. 1), which has shoulder decoration almost identical with that of ours. This style is to be found with Belgic pottery of the first century A.D. and is designated " South Eastern B." The great food vessel from spread E (Fig. 4, No. 1) is the most perfect of three of this kind from Borden. Its paste is of a grey earthy t V . XMWV .i VU/AMA 5 J \ / f 2 4 e. 7. 8. g. to. ' w a • M = » ™ = M B e = 3 Cm*. WOOOLAtiP FIG. 4. COARSE POTTERY FROM BORDEN. texture, not heavily gritted ; it has a shght reddish outside shp to rim and neck, a clean grey inside surface and it is probably wheel-made.- The decoration is arresting. There is one neck cordon, and on the shoulder a band of horizontal combing upon which is superimposed a series of " horse-shoe " loops ; below, there are rough vertical combings. J. P. T. Burchell in his " Final account of Investigations carried on at Lower Halstow, Kent" (Proc. Prehist. Soe, 1925-27, Plate II, No. 2, p. 293) shows a decorated sherd with combing almost identical with 154 AN EARLY IRON AGE SITE AT BORDEN that of our crock. The pot fragments from our site drawn in Fig. 4, Nos. 2,3,4,5 give further styles of ornament. No. 5 with the concentric ring and roulette design may weU be from another " South-Eastern B " type of pot. The dating of our Borden pottery caUs for Uttle comment. It approximates very closely to that from Crayford, upon which J. B. Ward Perkins after careful study pronounces " Some at least of it belongs to the closing years of the pre-Roman period . . . others such as a corrugated vessel [like our Fig. 3 , No. 6] seem to belong to a somewhat earher phase. The majority of these vessels are hardly sufficiently determinate for further precision within the compass of the Belgic period, but aU could weU belong to the last seventy-five years before the Roman conquest " (see " An Early Iron Age Site at Crayford, Kent," Proc. Prehist. Soe, 1938, p. 152). Into this period the Hart's Delight treasure conveniently fits, for it is likely that Borden Camp was overrun by the Claudian troops in A.D. 43 or at any rate soon after their landing, so causing the flight of the occupants, one of whom smuggled out his golden hoard to hide it close by, and failed to return to recover it. The occupation of the Roman conquerors is patent in the crematory pottery revealed (Figs. 3, Nos. 13, 14, 15). My thanks must be tendered to Messrs. Wills & Packham for aUowing me the run of theh ground during the removal of the brick-earth, and to Professor Christopher Hawkes for his kind advice and help in the production of this report. I am also especiaUy indebted to Messrs. J. White and G. Woodland, to Mr. White for the time and care exercised in assembling and measuring the sherds, and to Mr. Woodland for the skfiiul scale-drawings he has made of the material so arranged. [The Author and Editor are also grateful to the Rev. Dr. S. Graham Brade-Birks for assistance in the preparation of this paper.] 155