Archaeological Notebook Canterbury 1949-51

ARCHAEOLOGICAL NOTEBOOK CANTERBURY 1949-51 By FRANK JENKINS JANUARY, 1949. The instaUation of a petrol tank on the south side of Gordon Road, not far from the east end of the old malt house, revealed an inhumation burial at a depth of 3 ft. below the surface in clean brickearth. Though no grave goods were associated, it is probable that it was of Roman date, for the site is within the area covered by an extensive cemetery of that period. The owner of the site informed the writer that some years ago he had found fragments of Roman pottery near the spot, but these are now lost. MARCH, 1949. The same burial ground was crossed during trenching for a gas main in Station Road East. The main finds were situated outside Messrs. Pickford's Furniture Repository, and consisted of fragmentary human remains lying at a depth of 3 ft. in yellow brickearth. Fragments of mid-lst- to 3rd-century Roman pottery also came from this spot. MAY, 1949. Site A. (Fig. 1, Nos. 18 to 20.) The Post Office Engineering Department made a deep excavation at the junction of Stour Street and High Street to accommodate an underground cable jointing chamber. The final depth reached was 9 ft. At the 8-ft. level a cambered spread of gravel and flints was encountered. A trial hole dug into this proved that it was more than 2 ft. 6 in. thick. It had the appearance of road metaUing, perhaps that flanking the forum said to exist thereabouts. The feature was sealed by a deposit of black water-logged soil. From this came a quantity of Flavian Samian ware representing Drag. Forms 18, 15, 27 and 37. Other finds were a small bronze brooch with hinged pin, the hob-naUed sole of a Roman boot, and an As of Titus (obv. T CAES IMP AVG F TRP COS V CENSOR ; rev. [VICTORIAE] NAVALIS (in field SC.) M&S 789A. A.D. 77-8). 63 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 •'/'».:// HI . ' d a k . J VLX's FIG. 1. (i) SEPTEMBER, 1949. Site B. (Fig. 3, Nos. 10 to 13.) Deep foundations for the extension of No. 59 Burgate Street, at the corner of Canterbury Lane, revealed a spread of rammed gravel and flints some 3 ft. thick and tailing off to a thin scatter at its edge, which was approximately in aUgnment with the lane. The build-up at the side of this feature produced pottery ranging from the mid-lst to late 2nd century. One Samian form 27 base, stamped OF VITA (Vitalis of La Graufesenque), and a form 31 base, stamped BORILLI OF (BoriUus of Lezoux), were recovered from this build-up. SmaU finds, other than pottery were a small stone spindle whorl, and a piece of worked antler, also a sestertius of Antoninus Pius (obv. iUegible, Head laur. r; rev. LIBERTAS PVBLICA COS II, in field SC), A.D. 131-61. I> Fia. 2. ( J) 64 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 At 12 ft. below the surface a wooden stake was found. It had been cut from a silver bhch sapling about twenty years old, but the growth rings were unsuitable for dating purposes. MARCH, 1950. Site C. (Fig. 3, Nos. 1 to 9.) In excavations for a block of flats between Nos. 16-24 Burgate Street, no Roman buddings were encountered, the area evidently remaining as an open space throughout the Roman period. A good test section was provided by a drain trench dug in the ceUar of No. 23 Burgate Street, at the extreme west end of the excavated area. It 3 7 rW7* r IQ.4&HB ) i> T 12 / tolPffqOfl) M^ 3 8 FIG. 3. (£) was dug from ceUar floor level, that is, 5 ft. 6 in. below street level, at right-angles to the street for a distance of 8 ft. 6 in., and continued for a further 7 ft. 6 in. as a tunnel beneath the modern roadway. The average depth was 5 ft., that is 10 ft. 6 in. below street level. This showed a stratification as foUows : 1. Natural brickearth. 2. Dark greyish brown loam mixed with minute fragments of carbonized vegetable matter, 10 in. thick. 3. Dumped sod, containing some Roman buUding debris, 1 ft. 9 in. in thiokness. 4. Dark greyish-brown soU equating with that found on the opposite side of the street, and 4th century in date. (No. 47 Burgate Street. Arch. Cant., LXIII, p. 92, key deposit No. 4.) As was to be expected, no trace was found of a Roman predecessor of the modern street. 65 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 Beneath the sites of Nos. 17 and 18 Burgate Street four stake-holes ranging from 1 ft. 6 in. to 3 ft. deep were seen in the surface of the natural brickearth, and at 10 ft. from ground level. A plan kindly drawn by Mr. W. G. Urry does not afford any clue to theh exact function. From the labourer's description, these holes suddenly appeared in the natural surface. In consequence they seem to be of early date, and were possibly Belgic, for traces of Belgic huts have been found on the south side of the street opposite this site. (Cf. Arch. Cant., LX, p. 75.) MAY, 1950. The clearance of a large bomb crater on the line of the City wall about 25 yd. north-east of the Riding Gate, prior to restoration, was closely observed by the writer in all its stages. The final depth reached was 6 ft. from datum, i.e. the present-day ground level outside the waU. Above this level the core of the waU was very badly constructed, the flints being sparsely set with mortar, almost ah-spaced in fact. Below this, however, it was very solidly buUt and compares favourably with the Roman work revealed by the excavations carried out by Mr. Sheppard Frere, just south of the Riding Gate, not many yards away from the present site (Journal of Roman Studies, XXXIX, p. I l l , fig. 25). Samples of the mortar have been taken at each 1-ft. level, and it is evident that those from the lower levels differ from those taken from higher up. It is hoped that an analysis of these samples and one other obtained by the writer from a spread of mortar dated late 2nd-early 3rd century A.D., at the foot of the Roman rampart on the site of No. 1 WatUng Street, just behind the present site, will prove without any reasonable doubt that the weU-constructed lower courses represent the original Roman work. NOVEMBER, 1950. Site D. (Fig. 1, Nos. 21 to 22.) During the extension of a basement area in front of No. 12 Station Road East, less than 10 ft. west of the human remains recorded in March, 1949, see above, workmen found three Roman inhumation burials with heads to the south-west. Though the work was well advanced by the time Mr. Maitland MuUer, of the Royal Museum, and the writer arrived at the site, enough evidence remained for a fahly accurate report to be made. The surrounding stratification showed a build-up of 4 ft. of domestic occupation material overlying the natural brickearth. From unspecified levels in this came two coins, Vespasian (M. & S. 758), and Faustina (probably the younger), respectively, a mid-lst-century type brooch, and pottery ranging from Belgic to 2nd century in date. 66 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 The graves were originaUy dug from a graveUy layer overlying this build-up. Two were relatively dated, the later probably containing a coin of Gratian (mint mark, Aries viz; in field OF I I , in exergue, CON*). From the side of the thhd grave the writer recovered a bone pin. The succeeding deposit consisted of a further 4 ft. of Roman build-up to within 2 ft. of the surface. FEBRUARY-MAY, 1951. Site E. (Fig. 1, Nos. 14 to 17, Fig. 2, Nos. 31 to 32.) No trace of any Roman buUdings was found over a wide area excavated for the extension of the Canterbury Gas Works plant,, on a site east of Church Lane, St. Mildreds, and about 40 ft. from the north-east angle of the church. The stratified deposits showed a normal build-up of about 5 ft. This consisted of mixed soil to a depth of 3 ft. containing pottery dating from the 13th-18th century. At this level the Roman deposits commenced. The first was black soU about 1 ft. thick, apparently a normal buhd-up as shown by the pottery which ranged from the 2nd-4th century in date. From near the surface of this deposit came three coins, one a Constantinopolis type, and the others of Tetricus I, and Victorinus respectively, the latter being a Virtus Aug type. One piece of decorated Samian Drag. Form 37 bears the erotic group Oswald type J (cf. Oswald, Figure types, pi. xc). At a depth of 4 ft. the soil became much cleaner, changing to a dark brown loam at 5 ft. From the former came some lst-century Roman pottery, including a smaU fragment of Samian ware, presumably from the base of a Drag. Form 29 bowl stamped LICIN[, probably Licinus a Claudian potter of La Graufesenque. From this same level elsewhere on the site came a small bronze brooch, Colchester type IV, probably dated to the third quarter of the 1st century A.D. At the lowest level, just above the natural brickearth, was a small quantity of native Belgic pottery representing native furrowed ware cooking pots, in character hardly later than the mid-lst century, and probably shghtly earlier than the Claudian invasion. Due to the presence of water at this level, and as the excavation was a commercial one, a number of pits seen outhned in the natural brickearth could not be examined, but they were evidently originaUy dug from a higher level. It is of interest to note that to-day the Roman levels are much water-logged, a condition which apparently did not prevail in Roman times. Thus the presence of occupation deposits at that level exhibiting no sign of contemporary marshy conditions, strongly suggests that here again is good evidence of a post-Roman submergence of the land surface, for the site is not many yards from the river. (For further remarks on this problem, see below.) 67 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 From close to the west waU of the Gas Works, in St. Mildred's Churchyard, Mr. W. F. Jenkins (Senr.), the Verger, found a coin of Constantine I (obv. CONSTANTINVS MAX AVG; rev. GLORIA EXERCITVS, two soldiers guarding two standards, in exergue, mint mark TRS.). MARCH-MAY, 1951. Site F. (Section Fig. 4. Fig. 5, Nos. 23 to 30.) The reconstruction of a collapsed revetting waU on the east bank of the River Stour at the rear of the G.P.O. premises in Stour Street, afforded a unique opportunity to study the archaeology of this branch of the river in circumstances, which if carried out as purely an archaelogical excavation, would have been an extremely costly and difficult undertaking. Briefly, the work involved was as foUows: First, steel piling, consisting of interlocking sheets, was driven into the riverbed paraUel to the bank for a length of 100 ft. Behind this effective barrier a foundation trench 3 ft. 6 in. wide- and 6 ft. 6 in. deep was dug. The final depth reached was 13 ft. 6 in. from present-day ground level and 6 ft. 6 in. below river level. Throughout the whole length of the trench the buUd-up of mixed soU (2) contained a quantity of 13th-century glazed pottery, mainly derived from jugs. That this deposit was fahly consolidated by the end of that century is shown by the presence of a rubbish pit which had been sunk into it for a depth of 5 ft. The filling of this contained a quantity of cut leather, mainly soles of shoes, and pottery of similar date. The surface of this build-up was the level from which the pit had been originaUy dug (1), and on it rested debris of simUar character. The maximum depth of the buhd-up was 6 ft., where it effectively sealed a hard compact bed of flints and gritty silt (3), needing a pick-axe to loosen it. From its sUty nature and homogeneity it evidently represents the bed of the 12th-century channel. The associated finds confirm this for on its surface rested a few objects which can be assigned to that period (A.D. 1200), namely, an iron flesh-hook (London Museum Catalogue, No. 7, pi. xxiv, No. 2 (C766), p. 125)'; a prick spur and a snaffle-bit. In this deposit a quantity of coarse pottery was found, dated by Mr. G. C. Dunning, F.S.A. as mid 12th-century, whue a trial hole to a further 2 ft. into it produced simUar material. I t is of importance to note the depth at which this medieval river bed hes in relation to present-day ground level, namely 13 ft. 6 in., and particularly with that of the lst-century Roman deposit which is known to lie at 8 ft. below the surface of Stour Street only a few yards north-east of the present site. Here the river bed hes a further 5 ft. 6 in. below that level. This in itself is significant when the origin of this branch of the river is considered. 68 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 GROUND LEVE1 &Mt&& RIVER LEVEL RIVERBED (MODERN) RIVERBED ^3 (EARLT0 FIG. 4 Site F. Section of river bank showing key deposits 1 and 2 = 13th-century 3 = 12th-century 25 26 27 _ _ _ . £ _ — / FIG. 5 Medieval objects from River Stour, Site F 69 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 The virtual absence of Roman debris throughout the excavated area at this level is remarkable, and can only be explained if we assume that at some time the river has changed its course, as no doubt it has. The natural scouring of this bank would be on the upstream end of the excavation, that is, on the outside curve of the river which begins to turn at that point. The consequent undermining of the bank would eventuaUy bring down the later deposits, thus effectively masking the earher ones which presumably he farther back. This could explain the absence of Roman debris, but is hardly what one would expect if the bank of a Roman waterway had been eroded. Another factor which does not appear to have occurred to others who have taken the Roman date of this branch of the River Stour without question, is the fact that in Roman times the land surface stood some 8 to 10 ft. higher in relation to sea-level than it does to-day. In other words a post-Roman submergence of the land surface has taken place. From this it foUows that throughout the Roman period the water level was that much lower, i.e. as much as 10 ft. In consequence, to have an effective waterway at that time the channel would have to be anything in the region of 15 ft. deep, and then could only hold about 5 ft. of water, which does not seem reasonable. To have a channel of this depth would have involved a great deal of unnecessary labour, for with the presence of the ford at Fordwich, and the Roman jetty at Sturry for. unloading ships, because of that natural barrier to traffic plying upstream to Canterbury, how would a navigable waterway at Canterbury be justified ? Further, why should we expect a river through the Roman town at this point ? The ground plan as we know it. would be extremely odd. In comparison with the defences on the south side of the town, its use in heu of a waU and bank seems unreasonable. There is also evidence that a pre-Norman chapel stood on top of the earlier West Gate, which immediately suggests that the gate was a masonry structure, and argues strongly in favour of its being Roman work. Here then is the crux of the matter. Did this branch of the River Stour exist in Roman times ? From the evidence which we have marshaUed together, the answer is overwhelmingly in the negative. We are thus left with the only alternative, and that is the stream came into existence in more recent times through natural causes. This occurred when the submergence of the land surface, inaugurated perhaps at the end of the Roman period, had reached the point where the water level had risen sufficiently to flood the whole of this area because of its lower lying situation, and in the course of time a channel formed through which the stream began to flow through the derelict Roman town. From information kindly provided by Mr. W. G. Urry, of the Chapter Library, it seems fahly clear that at the beginning of the 9th century 70 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-61 there were two branches of the Stour, one of which penetrated the walled area. (Cf. Bhch, Cart. Sax., Nos. 317 and 344.) With the evidence from this excavation before us, it is now of importance to reconsider the evidence obtained in the year 1935 (Arch. Cant., XLVIII, pp. 219 ff.). In that year, a few feet back from the river bank and just east of the present excavations, during the digging for the foundations of the Automatic Telephone Exchange, a massive structure consisting of large baulks of timber, evidently mortised together, were exposed, imbedded in black mud at a depth of 15 to 20 ft. below ground level. Around this structure were masses of reeds matted together as they had grown. In the present excavation this feature was absent. From our latest-evidence it would appear that this structure was not, as once thought, a Roman jetty, but something of more recent date. If it is not a medieval wharf it might perhaps be the first bridge erected in the late 13th century, to give access to the Grey Friars estabUshment oh the opposite bank. This hypothesis is given added weight by the fact that the timbers were sited approximately on the hne of an ancient right-of-way, a vestige of which stiU survives to-day as a narrow passage between the Telephone Exchange and the house known as No. 6 Stour Street, and stiU gives access to the present bridge which is Wth century in date. While these notes were in the press, supporting evidence has been forthcoming. In October, 1951, a large pit in the Post Office yard, 32 ft. east of the new river waU, showed that the mid-12th-century deposit lies at 7 ft. below ground level. Thus the difference in levels between this deposit on the river bank, and the contemporary river bed, is about 6 ft. During Easter, 1948, just north of the Blackfriars' Monastery similar conditions were found by Mr. John Wacher, who has kindly aUowed the use of his notes concerning the excavations which he carried out at that site. (Pubhoation pending.) JUNE, 1951. On the west corner of Station Road West, on land adjoining Mr. Marcel HaUett's garage, two inhumation burials were found, close together and side by side with the heads to the south-west. No trace of coffins was found nor any grave goods accompanied the burials. I t may weU be that they are outliers of the extensive Roman cemetery known to have been in this area. Judging from the contemporary ground level in relation to that of a rubbish pit which produced Roman lst-century pottery, the evidence is strongly in favour of this theory. A large fragment of a quern stone of Andernach lava was rescued from a simUar pit elsewhere in the excavated area. Mr. HaUett 71 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 informed the writer that other human remains had been found beneath his workshop nearer to the rahway. NOTES ON THE RESCUED POTTERY AND SMALL FINDS SITE C. Fig. 3. 1. T.S. Drag, form 45, late 2nd-early 3rd century. 2. T.S. Drag, form 37. Two ringed medalhon enclosing cupid, as at Alchester (cf. Antiquaries Journal, VII, p. 167, fig. 4, no. 17). Style of MOMMO, late Flavian. 3. Flask top, moulded mouth. Hard grey clay fired reddish brown. 4. Saucer-mouthed flagon. (Cf. Arch. Cant., LX, Fig. 8, No. 2.) Mid-2nd century. 5-7 inclusive. Flagon tops such as these are paraUeled with those found associated with others of type 4 above, in a deposit ante-dating the Roman rampart behind the city waU at No. 1 Watling Street, excavated by the writer. In consequence they cannot be later than the time of Septimius Severus, and a date about the mid-2nd century is indicated for the present examples. 8. Amphora neck with two handles attached, one stamped Q.P.PHRYXI. This stamp has also occurred at South Shields, Warrington, Rome, Xanten, Arentsburg and Strasbourg. Mr. M. H. CaUender, to whom the writer is indebted for the information concerning this stamp, states that there seems httle doubt that it originated in South Spain, and bases his dating on site distribution to the first half of the 2nd century A.D. 9. Stamp of herring-bone pattern on the rim of a buff ware mortarium. It is simUar to those which have occurred at SUchester, Newstead and BalmuUdy (cf. S. N. Miller, Roman Fort at BalmuUdy Report, pi. xl, B.l, type 32, p. 78). SITE B. Fig. 3. (N.B.—An asterisk denotes that the items were found WhUe the writer was actually on site, or the level was accurately indicated by the workmen.) *10. T.S. Drag, form 27, stamped OF VITA. Vitalis of La Graufesenque, period, Claudius-Domitian. *11. T.S. Drag, form 31, stamped BORILLI OF. BoriUus of Lezoux, period Trajan-Antonine. 12. T.S. Drag, form 37, probably Trajanic. 13. Stone spindle whorl. SITE E. Fig. 1. *14. Flagon top with three-ribbed handle, cf. Nos. 5, 6 and 7 above, hence mid-2nd century A.D. 72 ARCHAEOLOGICAL NOTEBOOK, CANTERBURY, 1949-51 *15. Flagon top with moulded mouth, dark grey clay, fired red with cream coating, probably 3rd century or later. *16. Mortarium, buff ware, weU gritted over rim, cf. Richborough type 349, dated there A.D. 80-120, a date which suits the present example. *17. Bronze brooch, Colchester type IV, dated thhd quarter of the 1st century A.D. *31-32. Furrowed ware cooking pots, native Belgic type. Both may weU be pre-Claudian. Another and larger version of No. 31 was found with these (not Ulustrated). SITE A. (With coin of Titus.) Fig. 1. *18. T.S. Drag, form 27, stamped OIIRO[NTI]. Frontinus of La Graufesenque, period Vespasian. *19. T.S. Drag, form 15/17, generaUy conforming to the pre-Flavian type. *20. Bronze brooch with hinged pin, complete and in an exceUent state of preservation when found. Colchester type XVIII, or Hod HiU type, pre-Flavian. SITE D. Fig. 1. 21. New Forest ware Cup, handle missing, but otherwise complete. Almost certainly from a grave with a coin of Gratian. End of 4th century A.D. 22. Bronze brooch. Colchester type VII. Sohd catchplate, spring pin, and tapering bow made from one piece of wire. Period Claudian. SITE F. Fig. 5. *23. Iron snaffle-bit, London Museum type D.III. c. 1200 A.D. *24. Iron prick-spur, LondonMuseum type BB(1)8, early 13th century. *25-26. Pottery spindle whorls, from Section 1, layer 2. *27. Stone spindle whorl, from Section 1, layer 2. *28. Chcular buckle, gUded, from rubbish pit, dated by the associated pottery, late 13th-early 14th century. *29-30. Typical examples of leather shoes, found in the abovementioned pit and the contemporary ground level. (Layer 1, Section, Fig. 4.) The reconstruction is based on the remains of two shoes obviously a matched pah. Many soles of the type Ulustrated were found, and these appear to be the commonest for only two broad-toed soles were with this group. AU had been discarded in a wellworn state and some of the uppers appeared to have been cut for the sake of the scrap leather. With the exception of the coin of Titus and brooch (No. 20), all finds are now deposited in the Royal Museum, Canterbury. 73