Canterbury Excavations September-October 1944

( 64 ) CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. REPORT BY MRS. AUDREY WILLIAMS, E.S.A. IN the war-devastated areas of Canterbury open ceUars and unoccupied ground floors afford a unique opportunity for acquiring by excavation information about the city in earUer times. In 1944 Major F. W. TomUnson, F.S.A., Honorary Secretary of the newly-formed Canterbury Excavation Committee, organized sufficient voluntary labour to begin this work. The supervisor was lent by the Ancient Monuments Department of the Ministry of Works. The first area investigated lay at the junction of St. George's Street and Burgate Lane. Here the cellar west of Burgate Lane (Fig. 1, and Plate I) was fuUy trenched and across the ceUarless ground east of the lane a section was cut at a right angle to the line of the medieval town waU (Fig. 1, II). / N 8M49.4 ^ v SITE I I T E I i.«4 recr 100 100 FIG. 1. SITE I. In the ceUar undisturbed ground lay 4 to 8 inches below the floor and 7 feet 9 inches to 8 feet 1 inch lower than the present road level. It was for most part a bright yeUow clay, but to the west the clay PLATE I. THE EXCAVATIONS ON ST. GEORGE'S STREET, CANTERBURY, with Major F. W. Tomlinson, Hon. Secretary of the Canterbury Excavations Committee. RI-7 I RO - ROMAN PITS < CULLY Ml-20 I MTI-2 - MEDIEVAL PITS I TRENCHE5 EXCAVATE O TO SURFACE OF PITff) DEPTHS OF PITS TAKEN FROM NATURAL SURFACE HEARTH KSW 4 POST-HOLES ( • ) AT DIFFERENT LEVELS IN FIUINO OF PITf - NATURAL „ O GKAVEL-, Q GROUND FLOOR OF •, ' • •• ,'.1/ .M?"_ M O D E R N B U I L D I N G 7 . 4 4 F T ABOVE CELLAR FLOOR NATURAL CLAY FLOOR OF CELLAR 0-MT.U MT2 R5 -I'T 4 8"vx FIQ. 2. SITU I. PLAN CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 65 >.;.j 1;*J: SV..-..!"-"?-:;-.-: •< v. 66 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER • 1944. changed to a breccia-Uke mass of smaU flints naturaUy cemented together. Into this old surface had been dug a series of pits and a few smaU trenches (Pig. 2). Seven of these pits (Rl-7) were of Roman origin, producing nothing of later date. They were roughly round or oval (Plate II, 1), measuring at the Up 6 feet or more across. Their depth varied from less than 3 feet to 7 feet at most. Some had been fiUed in and then cut into by subsequent pit-digging (Nos. 1-4). Their filling consisted chiefly of soft black soil containing household rubbish—potsherds, indeterminate scraps of bronze and iron, a few coins and quantities of animal bones (pig, ox and sheep or goat) and oyster sheUs. The homogeneous nature of the fLUing indicated that each pit was quickly fiUed. There were, however, as the sections (Fig. 3) show, occasional deposits of clean yeUow clay. Probably this clean material, obtainable from the cutting of fresh pits, was thrown in for sanitary reasons. The pottery and coins (see hst below, and Pig. 5) point to these rubbish pits having been in use from the middle of the second.century into the fourth century. Some of the sherds were abraded and had obviously been lying about for a long time before reaching the midden. WhUe the fresher fragments give a rough chronology for the pits, the possibility of a considerable time-lag must be admitted ; the pits may be rather later than even the latest sherds found in them. Pit Rl is of early-middle second century date, with a third century pit (R2) impinging. The two pits (R3 and 4) interlocking with R2 must both belong to the fourth century since a mortarium of late type occurred in R3, though the later pit (R4) had late third century coins and no distinctively fourth century material. Pit R5 dates to the late thirdfourth century, R6 to the late third century and R7, on the evidence of a coin of Oonstantine I, to the middle of the fourth century at earhest. Two short lengths of guUy (RG) were uncovered (Plate I I , 2). They were identical in character and presumably belonged to the same feature, but the presence of a shop floor at street level prevented fuller examination. Funnel-shaped in section (Fig. 3 BB1 and DT)1), the guUy was 3 feet wide at the top, 10 inches wide at the bottom and 2 feet 2-4 inches deep. In the absence of packing material or carbonized wood it must be taken as a drain of some kind rather than a paUsade trench. At either end it was cut into by medieval pits, but on t ae west, at least, it obviously never ran much farther than at present; o" iherwise it would have reappeared between pits M16 and 18. The filhng of the gully resembled that of the pits. Its few scraps of Roman pottery were not datable, but a radiate (? Tetricus I) coin in worn condition showed that it was open as late as the late third-fourth century. The use of this ground as rubbish dump persists in medieval times. OnSitel twenty medieval pits were found, on SiteIlyet another. This last must belong entirely to the later period, in that it began very Uttle CANTERBURY EXCAVATIONS ". 'SEPTEMBER-OCTOBER 1944. 67 below present street level, but some of the pits in the ceUar may well have been dug in Roman times and struck .again by medieval diggers startmg work at a higher level. This is borne out by the proportion of medieval pits to Roman in this hmited space. In turn, the medieval pits had been decapitated by the excavations for the modern cellars. A trial cutting (Pig. 2, IV) was made through the ground floor north of the ceUar in the hope of establishing later levels. Unfortunately building operations had disturbed the upper part of the section but medieval pit filling was encountered nearly 3 feet above that in the cellar itself. GeneraUy bigger than the Roman examples, the medieval pits varied greatly in shape and size. Some were wide-mouthed, bag-shaped holes, others cylindrical shafts with vertical or undercut sides (Fig. 3). The deepest (M9) dug in clay, was 10 feet deep, another in the hard gravel ran down over 9 feet to underlying chalk (Plate III, 1). New pits frequently overlapped older ones. By removing the fining to section faces showing the relationship of successive pits it was possible to arrive at the chronological sequence indicated by their numbering on the plan. The filUng was again black soU but looser in texture and, owing to an admixture of chalk and mortar, lighter in colour than the Roman fining. Again intermittent deposits of clean clay occurred presumably for the same reason as in Roman times. From the filling came fragments of pottery, animal bones (ox, young pig and two retriever-like dogs) and innumerable oyster shells. This filling suggests that these pits, Uke their Roman predecessors, were dug expressly to receive the rubbish they now contain. If they had been intended as weUs or water-receptacles they would surely have been deeper with probably traces of a hning of timber or, in the graveUy sub-soil, clay might have been expected. Had they been cess-pits certainly a different type of filhng would have been found. Two of the medieval cavities appeared to be trenches rather than pits ; MT1 was older than pit M10 which partly overlay i t ; MT2 just cut into the edge of a Roman pit (R7). They both had parallel, almost perpendicular, sides and square ends. MT1 was 3 feet 3 inches deep and contained black soU and flint nodules; MT2 with a depth of 18 inches was fined with a mixture of pieces of chalk and black soil. How far to the north these trenches ran could not be ascertained. Minor features of interest were a hearth at a high level in one pit (M10) and a pair of post-holes at a lower level in another (M13). The hearth, used when the lowest 3 feet of the pit had been filled up, was built up of a layer of the flint breccia described above (3 feet by 18 inches by 4 to 6 inches deep) covered with a mass of powdered charcoal and reddened stones" and soil. The post-holes were dug in two of the patches of yeUow-brown clay which were fairly thickly distributed in the lower part of the pit M13. The holes were uniform, roughly 68 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 6 inches square and 12 inches deep. They had held timbers which tilted shghtly to the north. What purpose they served, sited as they were at a depth of 3 feet 6 inches in a large pit, it is difficult to guess. If they had feUows these lay to the west: there was none to the east. n // P L AN a i CLmAY BANK v u l . mm I '//')•• IU \ MOPE*** F/LL//M/C• // P I T / W E L L VA . / i. *i hi SECTION B P I T WALL v LAfTtLELIMfQi . •" -Tt lACK W W MOKTAP. CLAY BANK FlOO A Of nuesis t CtUAR-p (.-•SHELL! •^I^TYj-^Cl^AY SLACK SOIL FLINT sxrcc'A \ . + luoicATsy POTS HE FIDS f£ET 10 IS FIG. 4. SITE II. The pottery from these pits wiU be discussed later, along with that from the Site II pit. It aU belongs to the last quarter of the thirteenth century and as most of the sherds are in good condition, with some vessels discarded praoticaUy complete, that date must apply fairly closely to the pits themselves. One exception must be made : Pit M7 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 69 produced a soUtary object and that of a later date, a fragment from the base of a fifteenth-sixteenth century glass tumbler. This hole was abnormally shaUow and seemed at sight to be outside the series. ' It would appear to be merely the bottom of a later pit dug from a higher level than the majority of the pits in this area. SITE n . The cutting east of Burgate Lane (Fig. 4) was made with the purpose of testing the Roman defences. It ran, at a right angle to the town wall, for 21 feet across ground undisturbed by ceUarage. Excavation was complicated by the appearance of a comparatively modern well at 2 feet below the surface. The filling of this shaft—dirty soU, chant and mortar—was removed to a depth of 9 feet. The only notable find, and that of no dating significance, was a London farthing of Richard II (1379-90) in worn condition. The north face of the cutting, however, gave a satisfactory section (Fig. 4, ABO) showing three definite features—a bank of clay, a rubbish pit and the face of the medieval wall (Plate I I I , 2). The clay of the bank was bright yeUow-brown in colour, perfectly clean and evidently derived from virgin soil elsewhere. Beneath it ran a dark grey streak denoting the old turf line below which lay undisturbed clay, and the flint breccia already described for the west end of Site I. The level of this old surface ranged from 43 • 8 to 44 • 8 feet O.D. This tallies with the undisturbed level at just over 44 feet O.D. in the cellar to the west, and may be taken as general in this part of the town in Roman times. The bank can be accepted as the usual bank against the inner face of a Roman town waU, as for instance at Caerwent and Verulamium. Its surviving height was 6 feet and the amount of its total width exposed in the cutting.was 14 feet 6 inches. Its inner limit lay beneath Burgate Lane ; its outer face had been cut into, at the point examined, by a medieval pit. This pit caUs for speoial comment because of its bearing on the date of the waU. It first showed at 2 feet beneath the surface, below modern build-up. It was funnel-shaped in section with an inward sloping side from the mouth to a depth of 3 feet and then a vertical or slightly undercut line to the total depth of 11 feet. Although its north and south margins lay beyond the Umits of the cutting, and time would not allow of their being exposed, and although on the east waU building had destroyed its profile, enough of the pit was avaUable to show that it had been approximately oval, measuring about 12 feet by 9 feet at the top. East to west the narrower part (the shaft) was 4 feet 6 inches across. In the filling of this pit four distinct deposits could be seen : (i) Just over 3 feet of stiff black soil with household debris covered the bottom of the shaft. For this and subsequent deposits the section showed a 8 70 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. sagging tip-line which became less pronounced in the upper part of the pit. (U) Above came a layer (12 to 18 inches thick) of dirty brownblack clay, (in) Above again was looser black soU with an outstanding quantity of oyster sheUs admixed, (iv) The final layer, filling the top of the shaft and the wider mouth, consisted of stiffish black soil, pieces of chalk, a few sheUs and fragments of pottery. The first three of the layers above described clearly antedated the waU. The waU foundations sUghtly oversaUed the bottom five feet of pit filling and the trench cut to receive the foundations was visible as a narrow slot between the waU face and the upper part of the shelly black deposit (in). There was no sign of this waU trench in the final filling (iv) which must have been thrown in agamst the faced waU when it was freestanding and very soon after it was erected, so soon that the filling had penetrated the stiU open wall-slot. It wiU be seen below that the pottery from this post-waU filling is of the same date as that from the pre-waU deposits. The waU face stood 4 feet high on foundations 18 inches deep. It was of roughly coursed Kentish ragstone, with foundations of rubble (chiefly flint), and clay. Neither below the waU nor immediately in front of it was there any sign of Roman work. Without danger to the modern waU standing above the cracked medieval masonry it was not possible to undercut farther than 18 inches in search of the Roman wall. If it exists here its inner face must he well in front of that of its successor. The present cutting estabhshes two facts which indicate that its continuation beyond the extant waU is most desirable. First, since the inner face of the Roman waU lay somewhere in the untested thickness (4 feet) of the medieval waU its outer face should he in the cellar outside the present waU. Secondly, the accepted Roman undisturbed level is shghtly lower than the floor of this ceUar so that even if the wall proper has been reduced its foundations, sunk into the old surface, should survive. The pottery (Figs. 6, 7 and 8) from the pits in the cellar and from the pit against the waU belongs to the same period, the late thirteenth century. Further, no distinction can be made between ware from the pre-waU and the post-waU fiUinas of the " waU " pit. The ware is generaUy weU fired. It has a fine sandy texture, though some few pieces have a coarser grit or shell admixture. Its colour varies from red and buff to grey and black. Many of the cooking-pots are blackened with smoke. Glazed sherds rarely occurred and decoration is hmited to thumb-marks and finger-prints and stabbing. Cooking-pots predominate. The most popular form (Fig. 6, 1-2 and 6-9) here, as on subsequently excavated Canterbury sites, has a long neck and a simple or sUghtly flanged rim, beveUed either externally or internaUy. This type was not, apparently, found at Tyler Hill1 1 Arch. Cant., LV, pp. 57-64. PLATE II, I. SITE I. ROMAN P I T (CUTTIX'G II, RJ), FROM THE EAST. f'r.ATK I I . :.'. S I T K I . ROMAX" GULLY (CLTTlXG II. HG),FROM T H E KA.ST. /•' PLATE III, 1. SITE I. MEDIEVAL PIT (CUTTING III, M18), FROM THE SOUTH. PLATE III, Sin. II. MEDIEVAL WALL-FACE WITH SECTION .SHOWING PI'I' AND CLAY HANK. CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 71 the pottery site near Canterbury from which much of our pottery may, on grounds of form and fabric, derive. Nor are paraUels easy to find farther afield. The form may be a local one of limited distribution. The second common cooking-pot (Pig. 8, 21 and 23) with globular body and sharply outbent rim has analogies at Tyler HiU. Jugs are represented by fragments only, but at least two (Pig. 6, 6 and Fig. 8, 14) belong to the taU slender type with thumbed base, general in the thirteenth century.1 The jug with a short tubular spout (Fig. 7, 2) is a survival of a twelfth century form, as in a group dated c. 1294 at Bungay Castle, Suffolk.2 There are few dishes and bowls. Flanged bowls such as Fig. 7, 1 and Fig. 8, 13 appear at Tyler HiU, but the latter, with its straphandles, is unusual and closer to an example from Maidstone.3 It will be noticed that this bowl and the cooking-pot with elbow handles (Fig. 8, 24), in imitation of a metal cauldron with angular handles, are the only innovations in the post-wall series. They do not, however, affect the date (1275-1300) already put forward in that they occur elsewhere in thirteenth century groups, the bowl at Maidstone, as has been mentioned, and the elbow-handled pot at Tyler Hill and at Leicester, though it is more usual in the foUowing century.4 None of the pottery from the upper filhng of the " wall" pit need be later than 1300. The implication as to the date of the waU is obvious. As aU the sherds in both pre- and post-waU fillings are fairly fresh it follows that the pit was dug and two-thirds infiUed round about 1300 and, as has been already argued, the waU was built and the filling of the pit completed without loss of time. An alternative suggestion must be mentioned but only to be dismissed. The late thirteenth century filling of the upper part of the pit might have been removed at a later date, the waU erected and the same filling thrown back. It is scarcely credible that the filling should have been replaced unadulterated with objects of later date or, a more serious objection, without scrapings of the soU on which it had rested when displaced. Here that soil would necessarily have been the clay or topsoil of the bank, both easily recognizable by texture and colour when mixed with the black filhng. THE FINDS. Site I. From the Boman Pits (Fig. 5, 1-13). PIT Rl. Coarse Ware. 1. Pie-dish with bead-rim and beveUed base, good buff-grey ware, smoothed surface decorated oblique tooled lines (Fig. 5, 1). 1 London Mus. Cat., fig. 69, 1, 4 and 5. 3 Suffolk Inst. Arch. Proc, XXII, p. 334. * Arch. Cant., LV, fig. 3, 4. * London Mus. Cat., p. 224. 72 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 2. Dish with heavy bead-rim and shghtly beveUed base, smooth black ware (Fig. 5, 2). 3. Poppy-head beaker, thin buff-black ware, decorated vertical panels of smaU studs, cf. Richborough Report I, 54 ; Antonine, but other examples " not far removed from A.D. 100 " (Fig. 5, 3). P IT R2. Samian. 1. Drag. 38, with flange unusuaUy grooved on lower edge. Probably Hadrianic. Coarse Ware. 1. Butt beaker, thin grey ware with metalhc glaze, decorated rouletted bands, cf. Richborough Report II, 142 ; third century (Fig. 5, 4). P I T R3. Samian. 1. Drag. 37, style of IVLLINVS of Lezoux. Hadrian- Antonine (A.D. 130-40). Coarse Ware. 1. Jar with short concave ridged neck and heavy bead rim, grey-black ware (Fig. 5, 5). 2. Dish with bead rim and beveUed base, smooth buff-grey ware decorated obhque tooled lines (Fig. 5, 6). 3. Mortarium, rim approaching hammer-head type, pinkish buff ware. cf. Richborough Report I, 100 ; fourth century (Fig. 5, 7). 4. Fragment from double-handled storage jar, smooth cream ware (Fig- 5, 8). Also part of a whetstone, triangular in section, of a sandy palaeozoic siltstone, probably from a drift deposit. Length 4 in. PIT R4. Coins. 1. Tetrious I (A.D. 270-3). Antoninianus. M. & S.1110-12. 2. Claudius I I (posthumous). Antoninianus. M. & S. 261-2. Samian. 1. Drag. 18-31. Probably Hadrianic. Coarse Ware. 1. Dish with obhque sides and beveUed base, burnished black ware (Fig. 5, 9). P IT R5. Samian. 1. Drag. 45, with unusual decoration of shaUow corrugations round base of waU. cf. Oswald and Price, Terra Sigillata, PI. LXXIV, 2. Late second century (Fig. 5, 10). Coarse Ware. 1. Flagon with moulded neck-ring, cream ware. cf. Richborough Report I I , 164. Late third-fourth century (Fig. 5, 11). PIT R6. Coin. 1. Galhenus. A.D. 260-8. Antoninianus. Samian. 1. Drag. 37. Rheinzabern. ware. Antoninine. Also half a clay weight, circular on plan, flattened oval in section with cylindrical perforation. Diam. 5-5 in. Depth at centre, 2-2 in. 1 M. & S. =Mattingly and Sydenham, Boman Imperial Coinage. CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 73 \ X mmaui. ULtliiliJII imfSBm aw FIG. 5. SITE I. ROMAN POTTERY (£) 74 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. P IT R7. Cain. 1. Constantino I. A.D. 330-7. Antoninianus. Samian. 1. Drag. 45. Second half of second century. Coarse Ware. 1. SmaU carinated bowl, thin hard grey-black ware, decorated crude incised lattice on neck (Fig. 5, 12). 2. Dish, straight-sided, smooth buff-grey ware (Fig. 5, 13). GULLY. Coin. 1. ? Tetricus I (A.D. 270-3). Antoninianus. Samian. 1. Drag. 31. Indeterminate fragment. Coarse Ware. Indeterminate fragments. Of the Roman sherds found in the medieval pits none was of outstanding interest and they are therefore not here recorded. The foUowing coins were found in the later pits. P IT Ml. House of Valentinian I (A.D. 364-83). 3 AE. PIT M6. Tetricus I (A.D. 370-3). Antoninianus (barbarous radiate). PIT Mil. Postumus (A.D. 258-67). Sestertius. M. & S. 167. PIT M12. Constantino I (A.D. 330-7). 3 AE. TRP (Trier). Constantinopolis type. PIT M17. Constantine I. As above. From the Medieval Pits (Fig. 6, 1-11). PIT Ml. *, 1. Cooking-pot with long everted neck, thickened rim beveUed internally ; hard sandy grey ware, smoke-blackened (Fig. 6, 1). 2. Cooking-pot of simUar type, but rim everted and flat-topped; ware simUar (Fig. 6, 2). 3. Socket for wooden handle of a skUlet, buff ware with black coating, outer face brushed with twigs, cf. Rye,1 PI. XI, 4, and Ashstead, Surrey,2 Fig. 5, 25 (Fig. 6, 3). PIT M4. 1. Upper part of jug, rim bevelled internaUy, neck ridged ; sandy grey ware with red core and patches of yeUow-brown glaze externally (Fig. 6, 4). 2. Rim fragment of cooking-pot similar to Ml, 1 ; sandy buff ware. '* SuBBex Arch. Coll., LXXIV, p. 69. 3 Surrey Arch. Coll., XLVII, p. 63. CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 75 1 s ~y 7 J FIG. 6. Snaa I. MEDIEVAL POTTERY (J). 76 CANTERBURY EVACUATIONS : SEPTEMBER-OCTOBER 1944. 3. Sherd of simUar ware with pinched band ornament and patches of green glaze on inner face. P IT M5. 1. Thumbed base of jug of taU slender type, reddish buff ware, cf. Tyler Hill,1 Fig. 1, 1 (Fig. 6, 5). P IT M7. 1. Fragment from truncated conical glass tumbler with frilled ring round base. Fifteenth-sixteenth century, cf. an example, in Guildford Museum, made at Chiddingfold. P IT M14. 1. Cooking-pot of same general type as Ml, 1 but shoulder more angular and rim simple with external bevel; sandy buff ware, smokeblackened (Fig. 6, 6). PIT M16. 1. Rim fragment of smaU cooking-pot of same general type as Ml, 1 but rim flanged internaUy and decorated with fingerprints externaUy ; red sandy ware (Fig. 6, 7). 2. Cooking-pot of same general type but neck sUghtly concave, outbent rim bevelled externaUy; sandy buff-red ware, partly smokeblackened (Pig. 6, 8). 3. Cooking-pot of same type, rim beveUed externally ; black sandy ware (Fig. 6, 9): PIT M17. 1. Rim of cooking-pot with concave neck and simple clubbed rim ; sandy yeUowish-buff ware (Fig. 6, 10). 2. Rim of storage-jar (diam. 16 inches at rim), concave neck, thickened rim grooved along outer crest, pronounced mternal bevel and shght convex moulding below ; sandy yeUowish-buff ware (Fig. 6, 11). Site II. From the pre-wall filling of the pit (Fig. 7, 1-12). 1. Bowl with hammer-head rim decorated with finger-prints on outer edge ; red sandy ware. cf. Fig. 6, 7. 2. Jug with short tubular spout; red sandy ware streaked black, cf. Bungay Castle 13.2 3. Jug with shght internal flange; red-buff ware with reddish patchy glaze. Probably a jug of squat type. 4. Cooking-pot with shghtly concave neck and flat-topped rim flanged internaUy; hard sandy grey-black ware with coarse white " backing." 5. Cooking-pot with short shghtly everted neck and hammer-head rim ; hard sandy red-buff ware with grey core, partly smoke-blackened. 1 Arch. Cant., LV, p. 59. 2 Suffolk Inst. Arch. Proc, XXII, p. 337. CANTERBURY EXCAVATIONS ". SEPTEMBER-OCTOBER 1944. 77 r ^ ^ 0 0 0 2 2 5 ^ f 3 r 7 7 7 7 3 1 7 FIG. 7. SITE II. MEDIEVAL POTTERY (J) 78 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 6. Cooking-pot with longer neck of same type and shorter hammerhead rim ; sandy reddish-buff ware, partiahy blackened. 7. Cooking-pot with concave neck, vertical rim beaded internally, grooved externally ; hard sandy red ware. 8. SmaU cooking-pot, similar to 7 ; fine sandy grey black ware. 9. Cooking-pot simUar to 7 and 8 but rim section more square ; fine sandy buff-grey ware. 10. Cooking-pot with short sharply everted neck and heavy bead ricu with internal bevel and concavity below; hard sandy buff ware. 11. Cooking-pot, straight neck, with two tooled lines, beaded rim with internal ledge for Ud ; coarse reddish ware with large particles of sheU or burnt flint admixed, termed " backing ". 12. Cooking-pot with recurved rim thickened internaUy with bevel and ledge. From the post-wall filling of the pit (Fig. 8, 13-24). 13. Dish with flanged rim and a pair of vertical strap handles, rim and handles irregularly stabbed ; red ware with black coated outer face. 14. Part of jug of tall slender type, edge heavily thumbed ; good red sandy ware with bands of yeUow shp and green glaze on upper part of body. 15. Fragment from jug, neck cordoned, rim sUghtly bevelled internaUy ; sandy buff ware. 16. Neck of jug with flat everted top, handle with three grooves ; buff ware with grey core. 17. SimUar jug ; sandy grey ware. 18. Cooking-pot, concave neck and thickened fiat-topped rim ; coarse red-grey ware with buff inner face, speckled white backing. 19. Cooking-pot; shghtly concave neck and flanged rim with square section and shght internal thickening; hard sandy buff-grey ware with white backing. 20. Cooking-pot, very shghtly concave neck, everted clubbed rim.; hard red-buff ware, partiaUy blackened. 21. Cooking-pot of squat form, short thipkened neck, everted rim with internal bevel, decorated slats through bevel to outside of neck and vertical lines, finger prints on body ; fine sandy black ware. 22. Cooking-pot, coated neck, rim bevelled externally and beaded internaUy ; black-coated red-buff ware with traces of brownish glaze on both faces. 23. Cooking-pot sharply everted rim with short external bevel and pronounced hoUow internal bevel with ledge for Ud, decorated slats • through internal bevel, narrow cordon round body; sandy red ware, cf. Tyler Sill, 10.1 1 Arch. Cant., LV, Fig. 2. •CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944. 79 S V "A ^ J 7 FIG. 8. SITE II. MEDIEVAL POTTERY (i). 80 CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944.- 24. Cooking-pot with sharply everted rim beaded internally and elbow handles irregularly stabbed through ; grey ware with red inner face. cf. Tyler Hill, 16.1 MARTYRS FIELD ROMANO-BRITISH CEMETERY. . While the above work was in progress trial trenches were also cut at Martyrs Fields (Fig. 9) on the Une of a proposed extension of Oxford Road. It was supposed that part of the known inhumation cemetery2 / MAHTYKS MEMOJl/AL y y S" / o^s y y oy / / y y fy y / / oly y *}' S V1 X X x x x 77*/Ai TKCNCHiJ A/W SHOWN J7IPPL*£, -y\ ricr too FIG. 9. might here be encountered. Undisturbed yeUow clay, similar to that of Site I above, lay at 2 feet 9 inches to 4 feet below the surface. No burials were found. Acknowledgments. Help from Mr. Martyr Smith who surveyed the sites and Mr. Mann who took levels is gratefuUy acknowledged. In preparing this report I have been indebted to Dr. Felix Oswald, Mr. B. H. St. J. O'Neil, Mr. D. B. Harden, Mr. L. F. Cowley and Mr. F. H. Edmunds for their identifications of Samian ware, coins, glass, animal bones and geological 1 Ibid., fig. 2. 2 V.C.H. Kent, PI. XII, Cemetery No. 4. CANTERBURY EXCAVATIONS : SEPTEMBER-OCTOBER 1944 81 material. I am also grateful to Mr. G. C. Dunning for discussing the medieval pottery. A.W. The Canterbury Excavation Committee also acknowledges with thanks the kind permission to dig on their property given by the foUowing : J. G. Craik, Esq., Mrs. Bates, Frank Mason, Esq., Mrs. M. Pettit, and F. S. Barr, Esq. (Martin's Ltd.) ; the loan of tools by the O.C. 12th I.T.C.; the use of a shed for storage lent by the Canterbury Co-operative Society; and much practical help by H. M. Enderby, Esq., O.B.E., City Surveyor. F. W. TOMLINSON, Major, Hon. Secretary.