Dover: Stembrook and St. Martin-le-Grand, 1956

DOVER: STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 By PHTLIP A. RAHTZ SUMMARY THIS report describes the structures and objects found during mechanical excavations on two sites in Dover, Stembrook and St. Martin's. Indications of Roman buUdings were seen on the latter site, whUe at Stembrook were remains probably of a. quay and jetty, associated in both cases with the former estuary of the River Dour. INTRODUCTION Towards the end of 1955, two large scale building operations began at Dover, in the angle between Church Street and Stembrook Street (Stembrook), and on the west side of the market square (St. Martin's). The structures and objects found were reported in the local press and were examined by Messrs. F. Jenkins and Paul Woodfield, the latter, together with the Borough Surveyor, being responsible for the plan of the medieval walls at St. Martin's. The writer visited the site in February,. 1956, on behalf of the Ministry of Works, and observed further excavation on both sides. The present report is an amalgam of the information obtained by all the parties involved.1 It would appear that the St. Martin's site lies on the boundary between the aUuvium of the Dour VaUey and the brickearth exposed on the ground rising from it on either side. The Stembrook site lies entirely on the aUuvium (information from Mr. S. C. A. Holmes, Geological Survey and Museum). PREVIOUS EXPLORATION Finds from Dover up to that date were summarized by Amos and Wheeler in 1929.2 In the 1939-45 war several sites were cleared by 1 Acknowledgements are due to the contractors and staff of the excavations, who were unfailingly helpful; to the Borough Surveyor, who prepared a plan of the structures at St. Martin's ; to Mr. D. Philpott, the Quantity Surveyor on the Stembrook site, who made measured drawings of the quay and jetty ; to Mr. C. Morecraft, the Clerk of Works ; to Frank Jenkins for his notes on the coarse pottery and to L. G. Harris for drawing i t ; to Miss S. Butcher and Miss D. Charlesworth for notes on the Stembrook samian, and Brian Hartley for notes on that from St. Martin's; to Eric Higgs for his note on the flints; to Paul Woodfield for the appendix on the St. Martin's tomb and for several helpful suggestions ; and to L. Biek, of the Ancient Monuments Laboratory, for teohnioal advice (Appendix D and footnotes " L.B."). a Arch. Journ., Vol. LXXXVI, 2nd series, Vol. XXXVI, pp. 47-58, hereafter referred to as " Amos & Wheeler, 1929." See also V.C.H., I l l , 42 ; J.R.S., XXXVI (1946), 146 ; XXXVIII (1948), 98 f. ; XLI (1951), 140 ; XLII (1962), 102 f. (material colleoted by Miss M. V. Taylor for Professor Haverfleld). I l l DOVER : STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 enemy action, and were explored by the Dover Excavation Committee1 ; trenches were dug at Stembrook in 1945-47, and on St. Martin's in 1950. On the Stembrook site the previous trenches were not extended far enough eastwards to locate the wooden pUes of the " quay ", nor the chalk revetment behind ; nor were the same depths encountered. No finds are recorded from this area previously, except the massive waU2 from the south end of Church Street. The St. Martin's site has received attention frequently ; a large Roman building, with hypocausts and a tesseUated pavement, was noted here in 1881,3 and parts of probably the same buUding were located in trenching in 1950.* The Roman levels on this site were overlaid by the massive foundations of the Norman Church of St. Martin-le-Grand5 which have been identified in the present excavation, and those of 1950.6 STEMBROOK (Figs. 1 and 2) The excavations shown on the plan (Fig. 1) were for stanchions and a boUer house for a new block of buUdings erected by Dover Borough CouncU. The holes were dug during the autumn and winter of 1955-56. Only parts of 9-15, dug in February, 1956, were seen by the writer, and most of the measurements were taken by Mr. D. V. PhUpott.7 The site is low-lying ground (about 20 ft. above O.D.) on the west side of the present course of the River Dour, below the Market Square. Some 200 yards to the south-east a large timber and shingle foundation was discovered in 1857, interpreted by Wheeler as a quay, causeway or slipway.8 The relationship of this structure to the Roman channel of the Dour is uncertain. Two timber and chalk structures have been located at Stembrook at a distance of some 60 ft. from each other. " A " was a structure of pUes and planking on the edge of a wedge-shaped mass or platform of chalk blocks—this is interpreted as part of a quayside. " B " was a smaUer isolated chalk platform surmounted by timber staging and planking, interpreted as a jetty. " A "—the quayside—was seen mainly in the boUer house excavation. The section (Fig. 2) shows that the basal very compact gravel (Layer E) slopes graduaUy to east and south, from 4-5 ft. above O.D. near Church Street to 2»4 ft. below O.D. in Hole 5 near Castle Street. The lower part of this slope was made up to a level surface by a platform of chalk 1 Arch. Cant., LXIV (1951), hereafter referred to as Dover, 1945-47. a Amos and Wheeler, 1929, p. 50, No. 9. 3 Amos and Wheeler, 1929, p. 49, No. 5. * Arch. Cant., LXXI (1957), pps. 14 ff. 6 Arch. Cant., IV, pp. 1-26. • Also Amos and Wheeler, No. 10. 7 Details of the stratification of these holes are in the Ministry's records. 8 Amos and Wheeler, 1929, No. 10, p. 52. 112 PLATE I m Dover, St. Martins : General View of pit. PLATE II Dover, St. Martins : Lowor Layers in Section D. \fare p. 112 PLATK III LENGTH HOUSE %S°<>°9 "a o&va , •* %•'<$ DEPTHS INDICATED ARE IN rCET, AND REPRESENT THE DEPTH OP NATURAL GRAVEL RELATED TO ORDNANCE DATUM. B B B C A S T L E STREET SCALE =1 FEET P L A N of Excavations and Structures L.V.P PA.R. Eie. 1. DOVER : STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 blocks, averaging 1 ft. 6 in. by 9 in. by 9 in., roughly coursed, without mortar. These were held in place mainly by their own weight, and the edges secured by pUes.1 One was possibly a re-used ship's timber (PI. ILT and Appendix D). Only a few were found in situ and their original spacing is uncertain. Planking was nailed to their heads apparently as a further support for the top of the chalk platform. In the boUer house area the chalk platform was continuous, with a weU defined eastern edge, but in Hole 18, although a pUe was seen on the original ahgnment at its south-east corner, there was no firm coursing of chalk blocks ; there were three further piles near to the south-west corner, with a curved plank naUed to two of them. Furthermore, there is clearly a sharp drop from the boUer house area towards Hole 17 of nearly 4 ft. It seems hkely, therefore, that the quayside turned westwards in the vicinity of Hole 18, though further chalk blocks are recorded from Hole 9. The surface of the chalk platform was level with the top of the piles at 6 ft. above O.D., and may have been the original quay surface, though it is possible that planking was laid on the chalk, as on structure " B " , and perhaps further timbering on the edge. " B "—the jetty—was 60 ft. N.E. of " A ", and consists of a platform of chalk blocks 6 ft. 9 in. from east to west. On this was laid a framework of timbers enclosing an area of planking. No pUes were found holding this chalk in place, though two, not in situ, were seen in Hole 1. A further plank, shown on the plan, was found further to the north, loose in the filling. It wUl be seen from the section (Fig. 2) that the timber framework was secured by means of a half-joint on the plank, resting on an off-set formed by a double timber on the edge. Only the north, east and west edges of the jetty were found ; it cannot have extended far to the south or it would have been seen in Hole 4. It may have turned eastwards between Holes 1 and 4, or may have been an isolated " island ". STBATIPICATION (see Sections Fig. 2) Layer A is the most recent, from houses recently demohshed. Layer B is probably post-Roman sUt and debris ; no pottery was recorded as coming from this layer, and the dating of its accumulation is uncertain ; in the hand-cut trenches of 1945-47, a similar layer was mostly medieval. It continued without any apparent change down to the level of the quayside, blacker and lower in the vicinity of the jetty. 1 Nine piles and seven other timbers ; piles 4-7 feet in length with an average section of 6 inches, usually squared ; squared tapered ends culminatmg often in a battered end with " shoulder ", where driven into basal gravel. (See Eig. 10 and PI. III.) 114 DOVER 1956 STEMBROOK W.NW 20.16 PRESENT SURFACE A.) BRICK RUBBLC AND CELLAR roUNDATIONS DARK LOOSC CLAYEY PEATY SOIL, GRAVEL PATCHES Or CHALK RUBBLE 31) BLUISH CLAYEY SILT AND BRUSHWOOD CHALK BLOCKS O CHALK BLOCKS AND DARK SILT PLANKING 1 O « O 0) at/rr ri./wr GRAVEL AND SILT NATURAL ORDNANCE DATUM YELLOW SAND AND HEAVY GRAVEL (NATURAL) SECTIONS through Structures A and B '©?'.' Q TIMBER Q . • (Q . Q • • ^ . > m ; j \ ':"«?•; smm?i \ ' ?.:°.---.:_@>:.-.°.v.-.;..% _jt° V a t \ " \ \ EAST icy *•*•*.•. Vs.' ** - 'F12v • xV * ; \ > £ ^ i i g s f <«i® •*.*•".- - -' —; : BRICK EARTH ® CHALKY CLAY ond FLINTS U N E X C A V A T E D SECTION A - NORTH SIDE of PIT SCALE ^ | . _2f FEET FIG, 4. DOVER : STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 E Puddled clay and chah\, merging with G at base=B(S). F Buff orange chalky clay. G Black-brown soU and charcoal (F12 Roman floor)=B(S). H Buff-brown clay, less clayey to East=B(J). J Compact buff and grey clay, charcoal-flecked. J l Buff clay. J2 ChaUc and clay J3 Charcoal-flecked grey and grey-buff clay (as J). J4 Puddled chalk. K. Buff clay and chaUs: rubble. Kl Grey-buff clay. K2 ChaUcy gravel. K3 Dirty brown and buff clay, charcoal-flecked. L Buff clay, With orange straw at base. M Chalky clay. Ml Dark brown chalky clay. N Buff-brown chaUcy charcoal-flecked clay. Nl Orange earth and buff chaUcy clay. N2 Buff-yellow chaUcy clay=B(L). N2 Buff-yeUow chalky clay=B(L). 0 Brown clay with some chalk and flint. 01 Charcoal-flecked brown clay. 02 Grey-buff clay, charcoal-flecked, some chalk. 03 Brown clay, charcoal-flecked, chalk flint and gravel. 04 Very dark brown clayey soil, frs. oyster and Roman brick. Pr\ Pmu ddleid. cnh aUc J T i_i I\ =TR>o man af loors inFn9 .=TD\/(TJ\) . Q Clay, chaUc and pebbles J v R Brown soil (pre-Roman soU)*=D(0). 5 Dark grey-brown soil with charcoal and many pot-boilers=F.ll (prehistoric occupation level)*. T Buff-grey soU (old soil line under S)*. U Brown sticky soU merging into T. V Chocolate-brown fibrous sticky peat* with some flint and scraps of Roman tile. W Orange-brown brick earth (natural).* X Chalky clay and flints (natural). Section B. The make-up continues into Seotion B, but only Layer L can be directly equated with Layer N.2 in Section A. Layers M-P cannot be thus equated, but should be of similar date (Flavian samian 1 in Layer 0). The massive chaUc foundation F.2 probably cuts these and layers J and K above them, though K yielded ?Antonine samian 5 1 For layers R, S, T, V, W see footnotes on pp. 119 and 124. 121 DOVER : STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 and may abut on F.2, which is itself dated to Trajanic-Hadrianic times or later by samian (Nos. 20, 21, 22) In Layers W and X below it. It may indeed be much later: of the three coarse sherds sealed beneath it in layers W and X, two are late first-early second, but the third (Fig. 9, No. 9) might be much later. WaU F.l, however, hmits the continuation of the second century occupation level F.12 seen in Section A, and should itself be of that date or earher. Layer G appears to represent the destruction of F.2, and itself contained Hadrianic samian (No. 2) ; a similar layer (H), above F.2, contained Hadrianic-Antonine and Antonine samian (4, 25, 26). Layer E is presumably a floor-level laid on all these destruction layers, and may be medieval. On the right of the section no clear edge was seen to F.2 and R may be similar to G. Section B—East side of Pit Key to layers in drawn Section (Fig. 5) A Concrete and modern disturbance. B Dirty gravel—modern make-up. C Dirty soil and gravel—modern sewer trench filling. D Dark soU with some flint and chalk gravel—late Roman, medieval and later debris=A(B)-(D). E Compact yeUow clay—? medieval floor level. F Buff-orange clay -with few chalk blocks—post-Roman destruction of Roman structure? G Chalk rubble and dirty clay —1 destruction of structure F.2. H Chalk blocks in buff-orange clay=C(A). I Buff-orange clay with httle chalk. J Brown orange clayey soU=A(H). K Charcoal and black soU. L Chalky buff-orange clay=A(N2). M Dark grey-buff clayey chahky soU. N YeUow gravel. 0 Brown clayey soU with some flint. P Puddled chalk—some flint. Q Chahk blocks in dirty buff clay—F.2 structure. R SmaUer chaUc rubble. S Laminated chaUi: and gravel layers (==Roman floor F.12)=A(E) and (G). T ChaUc blocks=waU F.l. V Dirty flint gravel. W Dark brown clayey soil, charcoal-flecked. X Flint and chalk gravel, ochreous-stained, with dirty clay and oyster-sheU. 122 DOVER 1956 ST. MARTINS NORTH MODERN SURFACE: »> II) I SOUTH BRICK • . © • ' : OLD BANK BASEMENT ' WALL M \ L ' \ \ N v \ v \ vconcPete and red brick \ \ \ \ \ \ \ \ FOOTING WEFT; UNEXCAVATED SECTION B-EAST SIDE of PIT 5 O SCALE I FEET PA.R FIG. 5. DOVER : STEMBROOK AND ST. MARTIN-LE-GRAND, 1956 Section C. The west slope of the estuary is again seen here ; the lowest layer N1 was not seen elsewhere, though equating in depth with Layer V in section A. Layer M is apparently a destruction layer of a timber and daub structure of the first century A.D., destroyed probably before the end of the century (samian 11-17 and coarse sherds of similar date). It may be from the timber superstructure of waU F.3 ; Layer E, wliich should be the destruction layer of this waU, contained only Flavian samian (Nos. 18 and 19). Layer D, filling the space to the east of F.3, is of second century or later date (samian 8-10, Hadrian- Antonine and Antonine) and confirms a second century or later date for F.2, which cuts Layer D, part of wliich could be seen in the interstices of the chalk blocks of F.2. The make-up Layer C may be simUar to that in Section A, and should be of late first-second century date, as also should the waU F.6 which cuts it at the S.W. end of the section. F.4 and F.5 are presumably part of St. Martin's Church, and each shows evidence of a rebuild ; the old bank waU is inserted alongside F.4. Section C—South-east side of Pit Key to layers in drawn Section (Fig. 6) A Heavy chalk rubble and buff clay, sheU in top+B(H). B Dirty clay (part of L(D) to right)=(packing of F2). C Buff chalky clay=D(C). D Dark brown clayey soil, charcoal, burnt clay, red-painted plaster, fr. glass bottle. F Puddled or trodden chaUi. G Charcoal-flecked buff and brown clay=D(B). H Chalk rubble. M Dark brown clayey soil, much oak charcoal and burnt red clay. N Fine clean grey sUt. Section D had a very ragged curving face. The buried soil (Layer O)2 lies uniformly on the brick-earth. F.7 and F.8 were apparently aU part of the same waU, whose N.E. face was destroyed. It was later than the lowest Roman layer (J), but appeared to hmit the laminated floors F.9 on its S.W. side. H, E, F are perhaps destruction layers of an earher buUding. F.10 is on the ahgnment of the church waUs, and cut floors F.9. Layer D may be the waU-trench 1 Highly calcareous very fine sand ; no evidence of definite bedding ; could be either natural or man-laid, but more likely the former, possibly due to solifluxion ; probably not estuarine. (L.B.) 2 " brickearth " (very fine sand or silt loam) with some " fresh " angular flint; similar to layer R in Section A. Such a grade of basic material is very commonly deposited by wind, but that has not been proven here. (L.B.) 124 DOVER 1956 ST. MARTINS OI NE. PRESENT SURFACE • REBUILD SECTION OBSCURED vxet^n F5 . " f )