EXCAVATIONS IN DOVER
By L. MURRAY THRBIPLAITD
A FEW war-damaged sites in the centre of Dover were stUl avaUable for
excavation and the results are given below. The excavations were
carried out on behalf of the Dover Excavation Committee.
SITE 6. MARKET SQUARE. (Figs. 2, 3)
In 1949 Mr. E. H. Bayly and a number of volunteers started work
on the west side of the Market Square, on the site of the east end of
St. Martin-le-Grand which had been later built over by the Carlton
Club and various shops. The area had been bombed during the
1939-45 war, and part of the debris was removed by bulldozer to assist
excavation. In 1950 a larger area was uncovered, and again, boys from
Dover CoUege and Dover Grammar School helped considerably.
The excavation showed the principal part of two rooms and parts of
others with evidence of four main periods of occupation. There are no
tesseUated pavements or concrete floors, but one floor is very carefuUy
constructed and there is evidence of an upper room with painted wall
plaster. It is possible that these are the lesser quarters of the fine
building1 found in 1881 on this site when the cellars of the Carlton Club
were excavated. In the first two periods the walls must have supported
a timber superstructure, but were eventually carried up in chalk
in a late second-century rebuUd. The complexity of building and
alterations to these rooms is shown by plans on Fig. 3.
An important factor in the early buUding on this site was the fall
away of the ground level to the east. In Period 1 (Fig. 3), the chalk
foundations of the east waU of Room A (Section A-B, Fig. 4) were
dug into the clay subsoU and the ground levelled up beyond it.
This made ground contains a considerable number of sherds including
Samian c. A.D. 130-140. The floor of Room A was carefully constructed:
first, with roughly squared tufa blocks (Plate I, a) laid on the original
turf, with larger chalk blocks buttressing the foundations of the walls
and the whole covered by a layer of clay, then a level of beach shingle2
topped with another layer of clay whose surface had been hardened and
reddened by fire (Section C-D, Fig. 4).
1 Arch. Journal, XXXVIII, 432, and Arch. Cant., XX, 120.
2 Still used as a floor make-up. v. Arch. Cant., LXD7, 1961, Fig. 4.
14
EXCAVATIONS IN DOVER
D O V E R . PLAN SHEWING WAR
DAMAGED AREAJ IN THE
CENTRE OFTOWN,AND
EXCAVATION.!
-J1TE
ITAKC
MARKET
SQUARE
ff
Kg. 1.
15
[Crown Copyright reserved
EXCAVATIONS IN DOVER
DOVER.1950.
MAIN PLAN OP SITE
AT MARKET SQUARE.
i x>
V/77//?77/////////////M WAMM
j 3 & ROOM.A
ROOMB. I
>UC'K£lt/ *£
WALL boor/ /
L'0CK?6
eZZZI=WALLJ OF PERIOD 2.
SCALE
IROOM.C.
Kg. 2.
DOVER 1950
MARKET JOUAR.E
PERIOD 1
PERIOD 1E23.PERI0D U,VMLLJ BUILT ON MADECROUND
• PERIOD Z
SCAltOFFtET
PERIODS
*"~ W I £LA? tMUDJUftFACC. H U
HAT>p«ltf • 2ATIR HjOCV.tD
•••PERIOD 5.WALU HEIGHTENED & REBUILT
'EST » PCRIODJA.DOOW BLOCKtD.CLAY PLATFORM IVAtB
• M . PERI0D4.P0JT DESTRUCTION REBUILD
E22U0LDWALLJ JTILL UTILISED
Fig. 3.
16
EXCAVATIONS IN DOVER
At roughly the same time (Period la), on the made ground to the
east, the walls of a room were built which must have abutted onto
Room A. Inside the room was a level of clean sea sand (not seen on
Section A-B) presumably acting as bedding to a stone slab floor which
had been removed. A hearth (Hearth 2) associated with the floor had
burnt a hollow in the east wall of Room A.
DOVER 1 J 5 0 MARKET SQUARE
SECTION C-D
L A T E R L E V E L L I NG
CTgjQQ^1, © FALLEN DIBRIJ © CHALK FLOOR. # ^ b - f r i j ^ f l &J
f
M
FC08TC
i m3 BD
H I N C LE
N O R T H
JOUTH
M B . W A L L OF PER.IOD2
W A L L . ! OF PERIOD 3
SECTION A-B V/W/ft • WALLOFPERIODL
L A T E R L E V E L L I N G O®
©
o Os D wm,. S\ff i o Cc2 F A L L E N DEBRIJ
O^CbJS © I o I RECENT FOOTINGS tHEARTH o © C L A V t U U D \J
CHALK 1 A L K " - 0 0
© OCCUPATlOmy
FOUND
Wl® ATIONJ WE.JT MADE G R O U N D S M
LDT
I I
f j a g ^ M . W A L L OF PERIOD IA : TUAJT
iiii.'i'-itj^'^i JCAU
OF FF£T
Eig. 4.
But these walls sank so badly towards the east on the made ground
(Section A-B) that they had been razed and new walls erected north and
east of them (Period 2). Puddled chalk floors were made over the razed
walls and up to the two new walls forming Rooms B and C. A cross
wall was built across Room A too, dividing it into A and D, and puddled
chalk floors also made in them.
In Period 3 a considerable reconstruction was carried out. All the
walls with the exception of that dividing Rooms B and 0 were built up
(the south wall of Rooms A and B a foot behind the earher wall) with
17
EXCAVATIONS IN DOVER
chalk blocks, replacing the presumed timber superstructure, and door
openings were left through the north wall of Rooms A, B and C. Trampled
clay floors were made over the whole area above the earher chalk
ones.
Some time later (Period 3a) the door openings were blocked up and
a small buttress built on the north outside the blocked door of Room A.
A gap or door, in the west wall of Room A too was roughly filled in so
that there was now no entrance to that room except from above. In
view of this, it was interesting to note two clay " platforms " measuring
about 3-4 feet square set at this time in the north-west corners of
Rooms A and B. They had more or less level tops but sloped down
towards the floor on the sides away from the walls (Section C-D, Fig. 4).
I t seems possible that they may have served as bases for wooden steps
or ladders.
After this the orderly succession of building and rebuilding was
interrupted by some disaster. In Period 36, Room A is filled with
debris probably coming from the room above, burnt timbers, broken
pieces of painted wall plaster, sherds and lumps of opus signinum, and
Rooms B and C are full, too, of fallen debris. The pottery in these
levels contain late second-century types with carinated pie dishes
predominating, and perhaps the destruction of the house could be
attributed to the last years of that century. Rather unusual, however,
are the numbers of short sections of " drain pipes " (they are discussed
more fully below) made from a dozen or so different moulds which are
found in the burnt debris of Room A.
About the middle of the third century A.D. there are signs of reoccupation
(Period 4). Although the top of the most northerly wall
had been nearly obhterated and in places was covered with mud, a wall,
only two courses of which now remain was built almost, but not quite, on
the line of it (Pig. 2 and Plate 16), and a puddled chalk floor put down
over the fallen debris. The other waUs appear to have been still standing
and to have been reused.
This re-occupation was abandoned in its turn, and the floor became
covered with faUen building material. Trampled into the top of the
rebuilt waU of Period 4 in two main concentrations (Pig. 2) and into
the top of the fallen debris adjacent, were some seventy odd coins,
belonging to the latter part of the third century A.D. It is difficult to
account for their rather curious distribution, and it is possible that later
when the whole area appears to have been levelled up, two small
hoards, one perhaps in a box (see below, p. 26) in the wall were partially
dislodged or broken up and were trodden into the adjoining ground
surface. Two similar coins found at the same level south of Room A
may well have been carried there in the mud on feet or sandals. The
levelling which contains fourth century A.D. material possibly repre-
18
EXCAVATIONS IN DOVER
sents the preparation of the site for the building of St. Martin-le-Grand.
A grave (Grave I) was dug into the top of one of the earher chalk
walls and was filled with the later levelling material—the filling containing
a coin of the House of Theodosius I (A.D. 388-395). It held
the extended skeleton of a middle-aged man with his head., pointing
west, pillowed on a hollowed chalk block. The top of the grave had
been sheared off by the bulldozer so that it was impossible to see from
what level it had been sunk.
I am indebted to Professor A. J. E. Cave, Department of Anatomy,
St. Bartholomew's Hospital Medical School, for the following notes on
this skeleton. " The cranial and more equivocally the pelvic characters
of this skeleton indicate with fair probability its male provenance.
From the condition of the cranial sutures and the degree of dental
crown wear the age of death . . . may be estimated as between 40 and
50 years. The estimated living stature was about 5 ft. 1\ in. The
bossed occipital region of the skull, the condition of the teeth, the
slight platymena and platycreumia of the lower limb bones and the
general characters of the skeleton suggest an assignment to the Romano-
British or early Saxon period. That is, the osteological features of
the skeleton can be matched by human material of such periods."
SITE 3. YEWDEN'S COURT (Fig. 5)
Another attempt was made in the autumn of 1952 to locate the wall
of the Saxon Shore fort. There was one area available crossed by the
presumed hne of the medieval town wall, and it seemed possible that
medieval builders might have made use of the massive structure of
fourth-century Roman defences, if they existed.
A first trench (Trench I) was dug roughly at right angles to a retaining
wall at the north of the court, below which was a drop of two to
three feet which was thought might represent the line of the medieval
town ditch. It was found, though, that this trench itself cut through
the filling of the medieval ditch which had been levelled in the early
nineteenth century. A second trench (Trench II) was therefore dug as
far east as possible and this revealed Roman and Medieval levels under
a recent cellar floor. Unluckily the medieval town wall was much farther
east than had been assumed and even the extreme eastern end of the
second trench just failed to find its face, which lay beyond the area of
the bombed site. The approximate line is shown on Fig. 1. It was,
however, shown clearly that Roman levels extended west beyond the
line of the medieval wall. A foundation of chalk blocks had been dug
into the original ground level with an early second century A.D. poppyhead
beaker in the filling overlying it. There were later levels : a floor
level, Level 2, containing mid-second century sherds and overlaid by
debris with early third century sherds, and another floor level, Level 4,
19
EXCAVATIONS IN DOVER
which was overlaid in its turn by debris containing a coin of Tetricus
(A.D. 270-3) and fourth century sherds. These Roman levels had been
cut into by the wide, shallow medieval ditch and the foundation trench,
packed with large flints, perhaps for the original medieval town wall
itself or a later patching.1 Above the builders' debris by the wall and
in the ditch are a number of later medieval levels, while above them is
a deep fill, as in the first trench on this site, of early nineteenth century
date.
D O V E R 1951. YEWDEN J COURT.
WEJT EAJT
CELLAR WALL
A CLAY FILL
(RECENT) ©
/
FOUNDATIONS
i :Js=H ILS CLAY FILL(RECENT) C6»
@12 UBBLE
5*1.11 IfcsJIW S UBBLE
\ \ \ \ \ 2N
• , \ w x \
3)FrITLLT NTG Srw^ - 0 \ V TRENCH OF \ \s w\v\ LT\E TAYi I MEWDAIELVLA L? TOWN / CHALKBL 3 0KS CONCRETE
^v TURF /
iiniiiiuiiiiiiiii
MED
I EVA NATURAL
TOWN' SOIL?
DITC ROMAN
NDATION ROMAN LEVELJ © ~ ©
MEDIEVAL] LEVELJ © ~@
fr LATER.
OF FEETV&&
Eig. 5
SITE 10. BELOW THE UNITARIAN CHURCH (Fig. 6)
In 1951-2 some trenches were dug, at the junction of Last Lane and
Adrian's Street, in a war-damaged area below the Unitarian Church,
where it was hoped to pick up a continuation of the Roman road found
on Site 2. Trench XXXI only showed a Roman levelling of reddish
clay, containing early second-century sherds, above the old ground
For Saxon sherds on this site, see below, p. 36.
20
NORTH-WEST
^OUTH-EAJT C O N C R E T E DOVER MAKE-UP
R E C E N T
SITE BELOW UNITARIAN
CHURCH.
TRENCHEJ
XXXoXXXlI
1551-2.
RECENT R E C E N T
ROOM
LOOR
CHALK WALL I
MEDIEVAL LEVELS
MEDIEVAL
LEVELS
DARK CLAV S ANDY- DEBRU
LIGHT C L A Y s ^ r ^ ^ ! . ^ ; ' < ^ ! ^ : " ' . ' V r : ' ' .
TRAM?Lt-i
NATURALfOIL?
(MAKE-UP
iii|!i:i|illli|liii (T\ ,
BROWN DEBRIS '.'lllMlllyllllllllllll11-
ROAD
SHXilKi)
ROMAN LEVELS©-®
Fig. 6.
[face p. 20
EXCAVATIONS IN DOVER
surface which here sloped down towards the east, but Trenches XXX
and XXXII showed a metalled and shingled surface leading steeply
down, south-east, towards the old estuary of the Dour. This surface
was laid partly on the old turf and partly on a chalk and clay ballast.
A late first-early second-century A.D. sherd was found in a trampled
level on the old ground surface, and a coin of Domitian in the make-up
under the metalled surface, but the road metalling contained abraided
Samian sherds of Antonine date. The later levels (4-7, Fig. 6) contain
late second-third century sherds and a number of broken pieces of
CLBR tiles, but the upper Roman levels in XXXII have been removed
by a medieval pit. The extent of the metalling was examined by
Mr. E. Bayly in another trench on this site, XXXIV, but it had been
deeply disturbed by later pits and it proved impossible to say whether
it represented a road or track or was part of a general hard on the slope
leading down to the estuary of the River Dour.
OTHER EXCAVATIONS
Trenches dug by Mr. E. Field1 between Adrian Street and Snargate
Street (Site 11, Fig. 7) in 1949 showed Roman walls and floors associated
with pottery of late first-early second century date. In Trench XX
west of this area the soil had been cut back and the cliff revetted in
recent times, for the concrete ground floors of the war-destroyed houses
lay directly on the natural soil.
In 1955 modern excavations cut back into the former area revealing
a fine section2 (Fig. 8). This shows on the east a cut through one of
Mr. Field's rooms with an opus signinum floor3 and plaster-faced tufa and
chalk block walls. There appears to have been a narrow passage or
corridor outside this room on the west, the other wall of which has been
destroyed by a recent sump. Beyond that a large room with plastered
floor and plaster-faced chalk wall contained a different filling from the
more orderly sequence farther east, for the floor was covered with a
layer of black ash while the room was filled completely with debris of
broken plaster, burnt material and tiles perhaps presupposing a local
or a general disaster as on Site 6 (p. 18). This occupation was built
on a hard artificial surface of puddled chalk over flints perhaps representing
earlier building or a rammed hard or quay at the sea edge (see
above). The section showed, too, the subsoil of the Dour valley overlying
the steeply sloping chalk of the western heights.
1 I am indebted to Mr. Field for this information. For the excavation of the
medieval Garde-robe on this site v. Arch. Cant, LXIX, 1956, 132.
2 Both Roman walls appear much wider than they actually are, as they are
cut through at an extreme angle.
3 Utilized by later builders as their cellar floor.
21
EXCAVATIONS IN DOVER
CONCLUSIONS
Although Dover suffered severely in the second world war, the
number of open spaces available for excavation in the old part of the
town were individually small and often cramped by high containing
walls, or cellars which in the main removed all but the deepest Roman
L> \
&ARDCROB!
O. fe \
>-
\
SM I 7 \ \
RS: N \
^
O 11
Fig. 7.
levels. Thus knowledge gained of the Roman town was unfortunately
limited. The known facts can be tabulated :
1. The occupation of the site between the Eastern and Western
Heights seems to have started in the late first-early second century A.D.
2. In the middle of the second century A.D. buildings with chalk
block walls, which must have supported a timber superstructure, and
associated floors'were found on Site 2 and on Site 6. A road going
roughly north-south, possibly the final stretch of the main road from
Canterbury to the coast, was also located on Site 2 at this date, and a
metalled area on Site 10, perhaps part of a hard along the foreshore (see
above also for Snargate Street Section) contained Samian sherds of the
Antonine period in the metalling.
3. In the late second century, chalk block walls replaced the
timber superstructure on Site 6. The north-south road on Site 2 at
this time was divided into three carriage ways by chalk gutters.
4. At some time at the end of the second century, Site 6 suffered
destruction, but whether this was local or general in Dover it is impos-
22
EXCAVATIONS IN DOVER
sible to say. There is possible evidence of a similar destruction in the
Snargate Street Section 21.
5. Rebuilding of a sort took place on Site 6, but by the end of the
third century the site appears to have been in a ruinous condition and
two small hoards of coins were secreted in the walls or wall tops.
6. There is evidence of considerable fourth-century activity in the
shape of sherds1 in later debris on aU sites, only in Level 5, Site 3 has
actual occupation been found and there has been no sign of Saxon Shore
fort defences.
7. Tiles bearing the stamp of the Channel Fleet (CLBR) have been
found in second century contexts and again in third but only as broken
fragments.
8. The Roman town, as it can be seen now, lay along the west bank
of the Dour estuary and stretched from St. Mary's Church in the north
to Snargate Street in the south and from Church Street in the east to
beyond the medieval defences at Yewden's Court on the west. So far
no defensive wall of any date has been found.
ACKNOWLEDGMENTS
My acknowledgments are due to the officers of the Dover Excavation
Committee, and to the owners who permitted excavations on their
property. I have been helped throughout by Mr. Edward Bayly of
Dover College and by his archseologically-minded schoolboys, and I am
grateful to Mr. E. Field for his notes on his excavations in Snargate.
Mr. Eric Birley, M.B.E., F.S.A., and Miss Grace Simpson have kindly
given notes on the Samian, and Mr. G. C. Dunning, F.S.A., on the
Saxon and medieval sherds. The late Mr. B. St. J. O'Neil, Chief
Inspector of Ancient Monuments for the Ministry of Works, made notes
on the coins and it was to his untiring interest that the Dover Excavation
Committee owed not only its inception but also a continuing fund
of enthusiasm and advice.
COINS
By the late B. H. ST. J. O'NEIL, F.S.A.
Site 6. Hoards
Philip II as Caesar (A.D. 244-246)
1. Obv. M.IUL. PHILLIPPVS CAES. Head radiate right.
Rev. PRINCIPI IWENT. Philip II standing left holding
globe and standards ; at foot a captive. M. & S. 219.
1 See also coin of A.D. 383-396 in Grave 1 (Site 6).
23
EXCAVATIONS IN DOVER
Gallienus. (Sole Reign, A.D. 260-268)
Obv. GALLIENUS AVG. Head radiate right.
Rev. [APOLLINI CONS AVG]. Centaur to right, drawing
bow.
Mint. Z (Rome). Antoninianus. M. & S. 163.
Obv. As last, head radiate right.
Rev. Illegible.
Mint. Illegible. Antoninianus.
Obv. As last.
Rev. MAR [TI PACIFERO]. Mars left with olive branch,
spear and shield.
Mint. A| (Rome). Antoninianus. M. & S. 236.
Obv. As last.
Rev. ? SALVS AVG. Salus feeding snake rising from altar
and holding cornucopiae.
Mint. Illegible. Antoninianus.
Obv. As last.
Rev. VBERTAS AVG. Ubertas standing left with purse and
cornucopiae.
Mint. |e (Rome). Antoninianus. M. & S. 287 var.
Obv. As last.
Rev. Illegible.
Claudius II (A.D. 268-270)
Obv. [IMP CLA] VD [IVS AVG]. Bust radiate (?) cuirassed
right.
Rev. VIRTUS AVG. Soldier standing left holding spear and
shield.
Mint. [Ill (Siscia).
Claudius II (posthumous) A.D. 270
Obv. DIVOCLAVDIO. Head radiate right.
Rev. CONSECRATIO. Altar aflame. Antoninianus. M. & S.
261/262.
Obv. [DIVO CLAUDIO]. Head [radiate] right.
Rev. [CONSECRATIO]. Altar [aflame].
Probus (A.D. 276-282)
Obv. IMP C PROB. Bust radiate.
Rev. Illegible. Antoninianus.
24
I W E i T D O V E R . . 1955, .SECTION BETWEEN ADRIAN .STREET AND
..SNARGATE STREET.
EA.ST.
RECEMT DEBR.U"
fcECENT CELLAR.
CHALK FACED ROMANS
WALL.PLAiTER FACED
INTERNALLY.": HAI.K,
RECENT
CELLAR
BOMB DEBRIJ
FLOOR .
177
„,°//CLA.Y
CHALK, FACED
ROMAN WALLMJU-IIDR
TUFA CORE \ 1 M
PLAJTERxX
FACEDS\ \\
INTERN ALL-. m ^ m
RECENT
SUMP RECENT
CELLAR
,;;-^;.<^-::!f*w ROMAN r-3«-: ^cfl-m
!fl,o
Previous
Previous
The Follies of Kingsgate
Next
Next