Excavations at Gravel Walk, Canterbury, 1967

EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967* By J. H. WILLIA.MS, M.A. With A REPORT ON THE SA.MIAN BY MR. A. P. DETSICAS, M.A., F.S.A. and APPENDICES BY MR. R. A. G. O.ARSON, M.A., F.S.A., MISS D. CHARLESWORTH, M.A., F.S.A., MR. R. HARCOURT, M.SC., B.V.M.S., M.R.C.V.S., AND MR. D. M.ACKRETH, B.A. INTRODUCTION ExcAVATIONS commenced on 4th September, 1967, and continued for five weeks on a, temporary car park on the south corner of Rose Lane and Gravel Walk, Canterbury. The site (Fig. l}, measuring approximately 80 m. by 40 m., was to be developed as a multi-storey car park and any archaeological deposits would be damaged by piling. During the excavation the temporary car park remained in use and only a limited area was available at any one time for examination. It was, therefore, decided to trench as much of the site as. possible by machine, backfilling one trench before opening up another, in order to establish the character of the site. This in itself, however, was a slow process-with 3 ·5 m. of unstable archaeological deposits each trench had to be shored before investigation could proceed and removal of any spoil was difficult and time-consuming. Accordingly, only two long trenches, A and B (and small areas C and D), were excavated. This method of excavation is not recommended, for interpretation based solely on sections is unsatisfactory and results must be treated with caution. Little was known of the archaeological· potential of the site. Frere had postulated a Roman street running north-west to south-east across the site and forming part of a regular street grid,1 and sunken floor dwellings of probable fifth-century date had been found to the north-east of Gravel Walk in 1960 by Wacher.2 During the Angevin period the site appears to have been used as gardens.3 The excavations * The Department of the Environment has contributed to the cost of the publication of this report. 1 S.S. Frere, Rooian Canterbury, the City of Durovernum (1965), 8; reproduced in S. S. Frere, B1-itannia, i (1070), 84. 􀃄 S. S. Frere, The lih1d of Towns in Roman Britain, in J. S. Wa.cher (Ed.), The Oivitas Capitals of Roman Britain, (1966), 91. 8 W. Urry, Cante1·bury under the Angevin l(ings, (1967), 275; also map 26. Larger soele sheet 2. 119 J. H. WILLIAMS ET AL CANTERBURY 1967 A GRAVEL WALK - - - - -. - - - - - - - Scalo Metres o 10 ' F10. I. Location of Site 8 \ \ 20 30 •• 50 showed occupation on the site from soon after the Roman conquest. Although there was much pottery of 'Belgic' character, no group could be definitely assigned to the pre-conquest period, and it would seem that 'Belgic' wares continued into the second ha.If of the first century. Timber and stone buildings of first- and second-century date were revealed at both ends of the site but, although third- and fourth-century material was recovered, no structures could be assigned to that period. Furthermore, it is clear that no Roman street ran across the site along the projected line. No medieval structures were found, although several rubbish pits, predominantly of eleventh-century date, were recorded. The work was undertaken for the Canter.bury Excavation Committee and the Department of the Environment, and by courtesy of Canterbury City Council. Many persons from these bodies were most helpful, in particular, Mr. J. Boyle and Miss S. Butcher. I would like also to 120 TRENCHES A and C P25 P24 TRENCH A PI2 1 I. PI2 3 P2I 2 Collapsed Trench • --, ...... /􀀃 ' / '\ • ..-,A· f ' \ 􀀩 P20 -􀀪 '-( 􀀁 B ,, ,✓ -- A I-__ , l ; I I I l P4 P 3 P 2 : L ____________________________________ ---------------· 􀀴----------------􀀵 AREA 4 PIT 5 TRENCH C AREA 3 Traces of pre-stone timber buildings Scale Metres o FIG. 2. 2 3 4 5 10 [tau p. 120 TRENCH A East TRENCH B East KEY 􀁘􀁙LtEa rBtrho wn TRENCH C West East SURFACE MODERN DISTURBANCE 􀁚Dk Brown -Clay MODERN DISTURBANCE 2 3 4 s West SURFACE West 10 􀁛Eartf-\ Scale Metres o 􀁜Charcoal & 􀁝,i:,rs;....--:.:=:'........!;;;;;;;;;;;::;;l!!!!!!!!!!􀁞-:;:;;;;:;;;!!!!!!!!!!'!!!'!;-;;-;;;;;;;,a!!!'!!.11!!111!11;;::==;::;;;;;:::;:::;:::====::L􀁟-l­ 􀁠Burnt Daub 􀁡- Flo. 3. Sections, EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 thank all the supervisors and volunteers ·who assisted in the excavation. Many also have given advice on the preparation of the report and ruisisted myself and Mr. Detsicas with the pottery; in particular, Professor S. S. Frere, l\fiss M. Wilson, l\fr. B. R. Hartley, Dr. and Mrs. J. P. Wild, Mrs. J. Bird and Miss V. Rigby. Mrs. J. Brayne, Mr. M. Card and Mr. K. Connor a.F.· cit., no. 24. 67. Grey-fawn; grey; fawn. Smooth surface. 68. Grey; grey; grey. Very smooth surface. 69. Grey; grey; grey. Coarsish fabric. Smooth surface round neck. Cross-hatching on. shoulder and roughish body. Of. Frere, 1970, 01p. cit., fig. 9, no. 17. Dating: no. 60 perhaps appears somewhat later than the rest of the group being typically Flavian, but reeded-rim bowls are known at Verulamium from c. A.D. 60--see 99 below. The rest of the group dates to c. A.D. 60. Layer 36: Samian: S.G. Form 18, Claudian? Coarse wares: 70. Grey; grey; grey. Smooth but uneven surface. 71. Grey; grey; grey. Fairly fine fabric and smooth surface. Native imitation of Gallo-Belgic butt beaker. 72. Grey; grey; grey. Smooth but poor finish giving irregular profile. 73. Dark grey; grey; grey. Burnished neck with rough body. Of. Bushe-Fox, 1926, o:P· cit., no. 21; Frere, 1970, op. cit., fig. 9, no. 7. 74. Grey; grey; grey. Shell-filled. Smooth neck but rough body. Cf. Frere, 1954, op. cit., fig, 5, no. 24. 75. Grey; grey; grey. Smooth surface. Of. Frere, 1970, op. cit., fig. 9, nos. 4 and 16. 76. Grey; grey; grey. Smooth surface. 77. Grey; white; grey. Very fine terra nigra. Hawkes and Hull, 1947, o:P• cit., terra nigra Form 5A; cf. Frere, 1972, op. cit., no. 25. 78. Grey/fawn; grey; grey/fawn. Smooth surface. 79. Grey; grey; grey. Coarsish fabric but smooth surface. Of. Frere, 1970, o:P· cit., fig. 9, no. 10. Some fragments of painted wall-plaster were present. · · Dating: This group appears similar in date to Layer 38, c. A.D. 60. Some of the pieces are somewhat earlier. Layer 35: Coarse wares: 80. Grey; grey; 􀃆ey. Smooth. 81. Grey; grey; grey. Smooth. 82. Grey; grey; grey. Coarsish fabric but smooth, rather uneven surface. Dating: First century, A.D. 132 EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 -· 60 I ") 2 6 1 I 􀀁 I 6 2 1 \ \ ) 6 3 I \ ) 6 4 1 ( \ "=··I ./ 7 􀀇 66 ' 6 7 1 ) ,; 6 8 ! 7 J 69 Fro. 8. Coarse Ware, Trench B (Seo.le:¼). 133 J. H. WILLIAMS ET AL Layers 32-4: Coarse wares: 83. Fawn; grey; fawn. Smooth. 84. Grey; grey; grey. Smooth surface. 85. Grey; grey; grey. Dating: First century, A.D. Layer 3: Samian: S.G. Form 27, Neronian-Flavian. Layer 29: Coarse wares: 86. Grey-brown; grey; grey-brown. Smooth. Also part of a tegu"la. Layer 19: Samian: S.G. Form 36, Flavia.n. Dating: Late first-century, A.D. Well-Layer 40: Samian: (a) from bottom: Fig. 4, 3. S.G. Form 29. Three conjoining fragments, with brilliant glaze but poor relief owing to careless moulding. The upper zone consists of a scroll with a leaf used by LICINVS (R. Knorr, Topfer und Fabriken verzierter Terra-Sigillata dea ersten Jahrhunderts, (1919), Taf. 45, no. 28) who also used the squashed decorative detail at the upper left and lower right edges of this sherd (ibid., Taf. 45, no. 26; of. ibid., Taf. 56, no. 11, by MELVS, and Ta£. 58, no. 11 by MODESTVS). Elements of what remains of the lower zone's decorative scheme can be paralleled in the work of several contemporary potters, e.g. the inverted arcades, with rosettes at their junctions, were used by ACVTVS (R. Knorr,. Terra-SigillataGefasse des ersten Jahrhunderts mit Topfernamen, (1952), Taf. 1/E), CADMVS {ibid., Ta.f. 13/F) and SENICIO (ibid., Ta£. 56/M, N and Taf. 58/R); the serrated leaf to left was used by AQVITANVS (ibid., Taf. 4/D) and MVRR.ANVS (ibid., Ta.£. 44/B); the a.corn by DARIBITVS (ibid., Taf. 81/B). Sherds with similar decoration in their lower zones occur at Colchester in Claudian levels (Hawkes and Hull, 1947, op. cit., pl. XXIII, 10-13 and pl. XXIV, 1) and, though definite attribution to the style of an individual potter is impracticable, Mr. B. R. Hartley suggests a date of c. A.D. 40-55, on the basis of the rouletted central moulding. Fig. 4, 4. S.G. Form 29. In good condition, with remnants of a leafy winding-scroll, and a leaf used by PASSENVS (Knorr, 1919, op. cit., Taf. 62, no. 58) and GALICANVS (ibid., Taf. 33, no. 13). The triple bud on the left {ibid., Textbild 7, p. 15) was used by several S.G. potters as was the tendril binding (ibid., Textbild 6, p. 13). A pre-Flavian date for this sherd is very likely. Also: S.G. Form 27, Glaudian; S.G. Form 29, Claudian-pre-Flavian; S.G. Form 18 or 18/31, Flavian-Trajanic. (b) From remainder of well: S.G. Form 36, Flavian-Trajanic; O.G. Form Curle 15, first half of second century; S.G. Form 27, Flavian; S.G. Form 29, Flavian. Coarse wares: 87. Red; grey; red. 134 EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 􀀅 78 ' I ) 􀀊􀀋&;;;;=-------i-------=--l -- 77 ::,::::::---- 1 \ ( ) s2 1-I ------,1 --> / , 8 7 ,I C------,\ ' s 8 ! ( ( I 9 11 \ 9 2 4' .93 J ) ) \ 96 F10. 9. Conrae Ware, Trench B (Sonle: ¼). 135 J. H. WIUJAMS ET AL Traces of cream colour coating on exterior. Smooth. Cf. Jenkins, 1952, op. cit., fig. 4, 15. 88. Grey; grey; grey. Very fine fabric and smooth surface. 89. Purple-brown; grey; purple-brown. Fairly smooth with stabmarks on shoulder. 90. Grey; grey; grey. Very fine fabric with hard metallic finish. Cf. Williams, 1947, op. cit., fig. 7, no. 3; Frere, 1972, op. cit., no. 424; Frere, 1970, op. cit., fig. 10, nos. 12, 19 and 20. 91. Grey; grey; grey. Cf. Frere, 1972, op. cit., 447. 92. Grey-fawn; pink; grey. Very fine and hard fabric. Cf. Bushe-Fox, 1926, op. cit., nos. 76, 77; Frere, 1954, op. cit., fig. 8, no. 77. 93. Cream; cream; cream. Very fine. Top of butt beaker. 94. Grey; pink; grey. Very fine fabric highly burnished on top interior of rim. Hawkes and Hull, 1947, op. cit., Form 120. Miss V. Rigby comments-A fairly common post-conquest type, which occurs in a number of black fabrics of which this is the least fine and most commonly found in Britain. Claudian to early Flavian contexts with examples from Richborough, Wroxeter, Cirencester, Verulamium, Baldock, Camulodunum, etc. Frequently described a.a terra nigra, but it was not made by the potters who manufactured terra nigra and terra rubra cups and platters. Some of the thinner well-polished examples are stamped under the base, e.g. Gloucester, Fishbourne, Richborough, Wroxeter. 95. Grey; grey; grey. Slightly pimpled surface. Of. Hawkes and Hull, 1947, op. cit., Form 257 (p. 268, no. 8); Williams, 1947, op. cit., fig. 15, no. 9. 96. Grey-fawn; grey; grey. Coarsish fabric. Smooth but uneven surface. 97. Grey; grey; fawn-grey. Possibly burnished on neck but rougher below. Coarsish fabric. Scratch marks. 98. Grey/red; grey; grey. Rough surface and coarsish fabric. 99. Light grey; light grey; light grey. Sandy. Reeded bowls occur at Verulamium from c. A.D. 60 to the middle of the second century A.D. though apparently commonest c. A.D. 100. Cf. Frere, 1972, op. cit., no. 211, etc., also no. 659. 100. Grey; grey; grey. Coarsish fabric but smooth surface. Cf. BusheFox, 1932, op. cit., no. 276. 101. Grey; grey; grey. Sandy. Cf. Frere, 1954, op. cit., fig. 7, nos. 60, 61 and fig. 8, nos. 71, 72; also Frere, 1972, op. cit., no. 660. 102. Grey; grey; grey. Sandy. Cf. Frere, 1972, op. cit., no. 682. 103. Grey; grey; grey. Coarsish fabric. Smoothed on neck but combing on body. 104. White; white; white. Smooth fabric with 'knife-trimmed' surface. 136 EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 100 j 􀀫 05 􀀂? "' 1Ub I ) J ,01 I " 109 1 )' '-1101 7' 􀀉 \:: 112 j 7 \. 113 􀀬 ) 􀀄141 ;; J .,.. 11s I Flo. 10. Coarse Ware, Trench B (Scale:¼). 137 J. H. WILLIAMS ET AL 105. Pink; pink; pink. Sandy. 106. Grey; grey; grey. Sandy. 107. Dark blue-grey; pale blue-grey; dark blue-grey. Highly polished surfaces with hard fine grained dense paste. Hawkes and Hull, 1947, op. cit., Cup, Form 560. 108. Dark blue-grey; pale blue-grey; dark blue-grey. Of. 107 above. 109. Grey; grey; grey. Sandy. llO. Grey; grey; grey. Ill. Grey; grey; grey. Hawkes and Hull, 1947, op. cit., 222, nos. 26 and 27; Frere, 1970, op. cit., fig. 9, no. 19. ll2. Grey; grey; grey. Ha.rd burnished surface. Of. Bushe-Fox, 1932, op. cit., no. 235. 113. Grey; grey; grey. Hard burnished surface with cross-hatching. Of. Jenkins, 1950, op. cit., fig. ll, no. 28; M. R. Hull, Roman Colchester (1958), 29. 114. Pink; grey; pink. Smoothish but soft surface. Miss V. Rigby comments: Probably a fairly local product from kilns not necessarily imitating samian or Gallo-Belgic imports, but influenced by the demand for fairly shallow bowls, dishes or platters. Of. a range of decorated red and orange bowls with flat broad rims from north Lincolnshire and Yorkshire-Roxby, Brough, Malton, Rudston. Flavian-Antonine. 115. Grey; grey; grey. Sandy. Dating: Severa.I pieces of samia.n found at the well bottom suggest the well was dug, c. A.D. 60. Most of the rest of the material dates from the end of the first century to the middle of the second century and there is nothing necessarily later than c. A.D. 150. The well infilling probably dates to the first half of the second century. TRENCH B-SUMMARY The trench therefore showed: I. The earliest occupation o n the site was soon after the conquest. 2. There was probably a series of timber buildings in the late-first and early-second century. 3. A well out in the first century was probably filled by the middle of the second century. 4. There was little evidence for the later Romano-British and subsequent periods before pits covering the site in the medieval period. TRENOR D-TR:m EXOAV.A.TION (Fig. 11) Area D was opened at the end of the excavation to test the hypothesis of timber buildings in Trench B. Because of the time factor, the work was unsatisfactory and a detailed description of the results would be misleading; no pottery is illustrated as it was limited in quantity and not securely stratified. At least five phases of occupation were, 138 , I I I TRENCH t-- - - - I I I I I I I I I ' Unexcavated l I I f D PERIOD 2 CHALK FLOOR I I I I I I I I I Unexcavated PERIOD 1 SLOT L _____L J􀀏d._ ___ ; _____ N N, Scale a■M■e11:t=r=e=s=o-•••-=====2•••••3=====4•••••s Fm.ll. J. H. WILLIAMS ET AL however, recognized and these are summarized below in their excavation sequence. 1. Phase V. On removal of the overburden there was a series of clay deposits into which was cut a group of five post-holes arranged in no clear structural pattern. Among the clay deposits was a sestertius of Trajan (Coin list no. 8). 2. Phase IV. On removal of the clay levels two walls at right angles to each other (Layers 13 and 23) were defined by narrow gullies containing dark earth and packed with stones presumably to hold uprights rigid. To the north of 13 an orange clay level (14) underlay a grey occupation level (12b) which, in turn, was below a layer of burnt daub (12a), presumably the occupation and destruction levels associated with 13 and 23. 3. Phase ID. On removal of Layers 13, 14 and 23, a chalk floor (1) level covered with burnt daub was revealed. The south edge of the chalk formed a straight line parallel to the line of 13 and beyond its edge were possible post-holes of a wall. Additionally, twenty tiles c. 0 · 20 m. square were laid flat at the edge of and parallel with this line, possibly forming a doorway. The tiles could not have been original to this phase since they lay on top of the daub. 4. Phase II. At the north west end of the trench, there seemed to be evidence for a.n earlier beam slot. A trial hole revealed a slot (Layer 30) running a.t c. 30° to the later buildings. 5. Earlier occupation. Although there was earlier building, it was not possible to investigate it further. All the timber phases would appear to be of first- or early secondcentury date. THE Corns By R. A. G. Carson, M.A., F.S.A. l. Layer A 18. Constantine II. Rev. Gloria exercitu.s. Mint of Lugdunum. P. V. Hill and J.P. C. Kent, Late Roman Bronze Coinage, (1960), i, 187; .A..D. 330-335. 2. Unstratified, Trench B. Carausius. Rev. uncertain . .A..D. 286-293. 3. Unstratified, Trench B. Uncertain radiate ± .A..D. 270. 4. Unstratified, Trench B. Nuremberg token of Hans Schultes; sixteenth century. 5. Unstratified, Trench C. Hadrian. Rev. 008 III, Salu.s. H. Mattingly, E. A. Sydenham and C. H. V. Sutherland, The Roman Imperial Coinage, (1923), Hadrian, 975; .A..D. 134-8. 6. Trench C, Area 4, top clay level. Constantine period. Rev. Beata tranquillitas; c . .A..D. 320. 140 EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 7. Trench C, .Area 2, top clay level. Local imitation as of Claudius. Prototype, ibid., Claudius 66; c. A.D. 50. 8. Trench D, clay layer overlying Layer 12. Trajan. Rev. Bridge, ibid., Trajan 569; A.D. 103-111. THE GLASS By D. Charlesworth, 1\1.A., F.S.A. 1. Rim and part of neck of flask or bottle in blue-green glass. First to second century, A.D. Trench A, Layer 20. 2. Part of the side of a square bottle. c. A.D. 70-130. Trench C, Layer 6. 3. Three small fragments of blue glass with cut line. Probably first century, A.D. Trench 0, area 5. 4. Fragment of square bottle, blue-green, c . .A.D. 70-130. Trench D, Layer 13. THE SM.ALL FINDS (Fig. 12) J\, 􀀆 l®7 • Li C(:)@􀀋· . 􀀌--0 • @ 5 􀀍@ 0 ' . 􀀊k,,:jf:/#%®:t-.iD 􀀃rsz 􀀃,4 Fro. 12. Small Finds (Scale: ½), Tm: BROOOHES By D. F. l\1ackreth, B.A. 1. Colchester type. The spring is an integral part of the body of the brooch. The wings are plain and the bow appears to be plain and somewhat wide for its length. The details are obscured by considerable corrosion. The catch-plate is complete and seems to be unpierced. The spring starts immediately behind the head and coils out to the left. Four coils survive with the chord, which passes across the head of the bow. Only the stalk of the hook, which once held the chord down, is present. 141 J. H. WILLIAMS ET AL The Colchester type begins its life before the Roman conquest, but the present specimen belongs to the period after that event, and is probably late in the sequence, when it had begun to display features which belong to the succeeding derivative types. In the present case, little can be detected of any such characteristics, except the wide bow. The date range is possibly .A.D. 50-60. Trench D, unstratified. 2. Continental import. The spring arrangement is the same as in the preceding specimen, but survives complete. The wings are very short and on each is a small boss. The bow is divided into two parts of which the upper has a pronounced curve and a D section, while the lower part has a marked recurve. At the junction of the two parts is a small cross-moulding. The foot of the bow comes to a blunted point. Like so many of the brooches from the Continent in this period, the catch-plate is almost parallel with the bow. This type of brooch is to be found in the Rhineland forts belonging to the period before the conquest. However, it is clear that, by c . .A.D. 40, it was at the end of its life; the number coming into Britain is low and, although brooches of the general type reached Wroxeter (unpublished), the fioruit is effectively over by c . .A.D. 50. Trench B, well, Layer 40. 0TJIER SMALL FINDS 3. Bronze candlestick base, distorted. Trench C, Layer 6. 4. Possible part of bar of folding stool. Metal has high lead content and there are traces of bronze plating-such stools often had bronze decoration. Cf. J. Curle, Newstead, (1911), pl. LXIV; also J. Liversidge, Furniture in Roman Britain), (1953), 33. Trench B, well, Layer 40. 5. Grey, polished bone dice. Double ring-and-dot marking, fairly regularly spaced on sub-cube. Trench B, well, Layer 40. 6. Worked and smoothed bone, pointed at one end and with criss-cross decoration worn smooth in places. Grooves a.cross pointed end. Trench C, Pit 4. Tm: AmwL BONES By R. Harcourt, M.Sc., B.V.M.S., M.R.C.V.S. lNTRODUOTION Chronologically the material falls into two groups, Roma.noBritish and medieval, and has been so treated for study. The small size of the collection allows little more than an enumeration of the- species present. 142 EXCAVATIONS AT GRAVEL WALK, CANTERBURY, 1967 Description of the material I. Romano-British. The species prnseut were cattle (4), sheep (1), pig (2) and domestic fowl (2) represented by a total of fifty-eight fragmentary bones. (Bracketed numbers represent minimum numbers of individuals.) 2. Medieval. This group contained 109 specimens. Cattle: Five animals, indicated by this number of right metacarpals, were present and their measurements, as shown in the table, point to their being the typical small animals of the period. Sheep: Four animals were represented by skulls. Three of these were horned and one polled, a feature first seen in the Romano-British period, which was quite common by the medieval. Pig: A total of fifteen bones included three mandibles of different ages. Horse: This species provided only twelve bones from two animals, one adult and one half-grown. Dog: There was evidence of two animals, one adult but very small, c. 0 · 21 m., and the other immature but already standing some 0·66 m. at the shoulder. Birds: The four bones found were those of domestic fowl and goose. Measurements of Long Bones-medieval cattle: tl tlh pw msd Humerus 270 239 37 Radius 250 64 33 67 (2) Metacarpal 180 47 26 Tibia 305 88 37 Metatarsal 224 42 24 229 45 25 dw 69 64 55 49 52 58 58 Ht. 108 cm. 42" 120 cm. 47" 122 cm. 48" tl: total length. tlh: total length to head. pw: proximal width. msd: midshaft diameter. dw: distal width. Both extremities are measured across articular surfaces. Ht: height at shoulder. All dimensions a.re in millimetres unless stated otherwise. 143