THE ROMAN FORD AT IDEN GREEN,
BENENDEN
CECILY LEBON
INTRODUCTION
Iden Green ford and subsequent bridges appear to have crossed a
tributary of the Rother's Hexden Channel at N.G.R. TQ 50133225,
and close to a boundary stone inscribed with the names of three
Hundreds, Cranbrook, Rolvenden and Selbrittenden, now somewhat
worn and damaged.
Following its course from north to south, one can see the ancient
road marked by traces of iron slag at intervals in the east bank of a
hollow way which runs southwards from Seven Acre Shaw to Stream
Farm, with the present-day footpath parallel to it on higher ground at
the edge of field 103 (Fig. 1). On plot 102, when barns were recently
rebuilt, their foundations and drains went down into burnt earth of
undisclosed depth immediately west of the presumed line of the road.
Containing hardly any iron slag, the feature may possibly indicate a
brick-making site for the building of the early eighteenth-century
Stream Farmhouse and for the adaptation of the older, timberframed
farmhouse (on Plot 125) which was soon afterwards converted
into three labourers' cottages. On the 1769 mapping by
Andrews, Dury and Herbert (Fig. 2) the old farmhouse is named
Pruiny Hall. It is now again converted into a single dwelling which
appears near the western edge of the location map (Fig. 1). The
confluence of several potentially useful streams and a network of
tracks converging in the vicinity suggest a history of early clearings
and the pursuit of economic activities such as stone-quarrying,
brick-making and iron-working, besides the agriculture and stockrearing
which persist.
In a nodal position, as seen on Fig. 1, a conspicuous feature is the
small, roughly pentagonal, enclosure where the excavations were
undertaken in close proximity to the boundary stone. The main
69
C. LEBON
• ^ j ^ ^ ^ ^ p , » , J i j i.
Seven A^re Shaw * ' . O
ft-.Q
LXXI. 13
>Wood
Slr*arn
Farm
a c A C.6-IJ4 o G «
niiiiooil
•'<•" nfe
.ft' \Og&w/iWifoi ' i „ ._
itttf-rViWlltmir
tp$!-' /idOlcrjiruMf
Fig. 2. Andrews, Dury and Herbert Mapping, 2 in. to 1 mile, 1769; from sheets
reproduced by H.H. Margary, 1975.
71
C. LEBON
stream flows in a deep course bending to border two sides of the plot
which is now overhung with trees. Other features are two old farm
gates with a disused cart track between them crossing the northern
portion of the plot; and a timber bridge constructed a few years ago
and replacing an earlier footbridge, to allow cattle to pass to and from
the old lane (118) and the field adjoining the lane (119). Stream
banks are fenced only on the far side of the water course. Although
open to the enclosure, they are steep except at one spot where it is
easy to descend to the stream by a trodden slope. Much of the area on
both sides is often waterlogged, and during the 1980 and 1981 seasons
was frequently invaded by cattle.
In the south-western quarter of Fig. 1, No. 118b is a lane which on
the ground looks like an old carriageway, bordered with trees and
displaying outcrops of Tunbridge Wells sandstone in its banks and
similar, but detached, stones here and there underfoot. Strangely, its
stony surface does not reach to the stream side opposite the site
enclosure, but it peters out as the lane opens into the marshy strip
alongside the south bank. At its other end this lane leads into
Coldharbour Lane which crosses the south of the map as a continuation
of Chapel Lane.
A more obvious and relevant route for our purpose of discovering
the point where the Roman road crossed the stream is the old, deeply
sunken and tree-lined lane, 118, leading up to Iden Green from the
present and former bridges on the south-east side of the site
enclosure. It is called Field Farm Lane, and is not now a public
footpath. Drainage from field 119 has created a seasonal tributary
which flows down the lower part of the lane; and, usually, a great deal
of mud impedes the way. However, further up on the gentle climb the
road is clearly founded on, or has cut down into, natural sandstone
which also outcrops in its banks. After passing Field Farm (115), it
goes on to form a cross-roads at the junction of Coldharbour Lane
and Chapel Lane, and another where it crosses Mill Street obliquely,
having formed one side of the triangular shape of Iden Green hamlet
(Fig. 2). The next mile is known as Standen Street and consists of a
long descent converging towards the Hexden Channel and in the
direction of Newenden which had a wharf and market, granted to
Christ Church, Canterbury, before the Norman Conquest.
Prior to the coming of Benenden's turnpike roads, for which a trust
was set up in 1768/69, the way south from Hemsted Manor in the
parish of Benenden was via the ford (or a bridge replacing it) and
across Iden Green where there was a choice of Mill Street southwards
or Standen Street continuing south-eastwards. (In Fig. 2 Standen
Street is named Saunding Street). This was surely the route taken by
Queen Elizabeth I when she rode with a large retinue from Hemsted
72
THE ROMAN FORD AT IDEN GREEN
to Northiam and Rye on her Progress in the summer of 1573. In the
first quarter of the twentieth century, before Hemsted was sold to
Benenden School, estate and domestic servants living in Iden Green
made regular use of Field Farm Lane in going to and from their work.
In the 1960s local people still frequented it for leisure walking and
children for playing in and exploring. One Iden Green boy found
there a Roman coin, said to have been dated by the British Museum
to A.D. 208. The lane has probably been in continuous use from
Roman to recent times.
On the southern margin of Fig. 1 the Ordnance Survey marks
'Roman road (course of)'. This shows conformity with the track
adhered to by Margary on a course which takes it for granted that the
road did not veer just before the stream crossing, but plunged straight
across, continuing due south 'up the spine of the spur',1 etc. It is
strange that he made no comment on the obvious and currently used
Field Farm Lane, although elsewhere he looked carefully at possible
tracks branching towards harbour sites. He may have relied too much
on the impressive photograph by Crawford (Plate I) which he
published as Plate XV in his Roman Ways in the Weald. This
photograph has been calculated to date to 1935 on the basis of the
present age of one of the boys who appear in the scene. Crawford,
Margary and subsequent commentators, as far as I know, accepted
without question the supposition that the array of massive slabs then
protruding at high level along the north bank and recorded in the
photograph represented in situ the remains of an original north-south
ford, the stream having, since the ford went out of use, cut down its
bed to a much lower level. This assumption is questionable on a
number of grounds:
(i) The line of the Roman road credited by Margary for a mile
or two south of the ford site rests on rather slender clues;
e.g. 'agger by hedge'.2 Furthermore, two bits of negative
evidence may be adduced. Doubts were raised when no trace
of ancient road remains was found during replacement of the
old hundredal bridge near Watermill House, Mill Street, in
the 1970s; nor did any appear when a gas-main trench was
dug across the alleged line in the north of Sandhurst parish,
(ii) From a viewpoint on the excavation site looking south across
the stream, the supposed course seems topographically
improbable owing to the marshy state of the opposite landing
ground and, 650 m. beyond, to a steep bank where the spur
1 I.D. Margary. Roman Ways in the Weald (1948), 220.
2 Margary, op.cit., 224.
73
C. LEBON
PLATE I
M..: v £l.,J
Mr-
W * *:j"**^»
«4a»»<*i—*
Part of Ford Paving, photographed by Crawford in the 1930s. Reproduced 1984.
rises as a formidable barrier. Indeed, test holes were dug at
intervals on the flat 'landing' ground beyond the south bank
of the stream and no hard core was encountered,
(iii) It is important to consider where the camera was pitched in
1935 and how restricted the view is (Plate I). A longer shot
might have revealed where, and in what direction, the ford
stones had crossed the stream. As it is, with the forward slabs
cut out, the picture leaves uncertain which way this streamside
part of the paving was intended to make the passage
across. Although not shown in the photograph, it is close to
the foreground that the stream makes a big right-hand bend,
compelled by the force of a tributary entering opposite and
running out of the lower stretch of Field Farm Lane, fed by
field drains. Unfortunately, on the projecting bank at this
bend, already subject to much erosion by water, a tall alder
has been allowed to grow up and its roots have disrupted and
dislodged most of the ford stones seen in the right foreground
of the photograph,
(iv) If the row of stones seen in the north bank towards the left
side of the photograph were in fact an undisturbed part of
74
THE ROMAN FORD AT IDEN GREEN
the ancient ford, it is conceivable that they were aligned
along the course rather than across it and that the stream has
eaten away the side of a causeway below the bend as well as
through it further upstream. Alternatively, there may have
been originally, or at a subsequent repositioning of these
stones, an intention to use the place beside the ford for
industrial or domestic purposes requiring hard-standing
close to the water; for example, for brick-making, laundering,
watering animals and other farm work.
Crawford who first identified the features he found on this site as a
Roman ford, seems to have carried out an excavation on the north
side of the stream and found a pavement of roughly-squared stone
blocks.3 The paved area measured 5.72 m. east to west along the
north bank and 3.89 m. from north to south. He exposed a wooden
post lying horizontally under the paving and 51 cm. below its top
surface. He must, therefore, have taken up some of the stones, but
presumably replaced them in back-filling. However, by 1980 most of
the paving-stones visible were lying in the stream. It is not unlikely
that others had been furtively removed to serve as doorsteps and in
gateways. A Benenden resident recalls by name a farmer who carted
away at least one load of paving-stones to break up for road- and
yard-making. Thus, the array of large slabs had vanished before the
1980-83 excavations. A half-sized slab was lying on the ground
beyond the bridge at the entry to Field Farm Lane; and another was
observed on the near side at the foot of the alder tree. This latter may
be reckoned as the sole survivor in situ of those in the right
foreground of Crawford's photograph. One or two smaller, rounder
stones are firmly clasped in the roots of the alder.
Margary described the paving stones as sandstone, instead of
paludina limestone, locally known as Bethersden marble. His accuracy
in following the Roman road immediately south of Iden Green
ford is now called into question; yet, to do so is not to forget how
faithfully he examined the great network of roads for clues, which he
clearly defined, to the benefit of future field workers.
THE EXCAVATIONS, 1980-1983
Towards the end of 1979, it was recognised that the Iden Green ford
remains had greatly deteriorated over the previous thirty or forty
years. Trees had grown up to maturity, with thick roots penetrating
where there had been an intact pavement, and many of the paving
3 Information from O.S. site card in National Monuments Records.
75
C. LEBON
stones had slipped into the water-course or disappeared entirely from
the scene. The stream banks had been severely cut back to provide
sufficient emplacements for a new bridge of heavy timbers; and cattle
passing over the bridge were churning up the ground on either side.
In 1980, after permission had been obtained from the landowner
and tenant of the scheduled site, and with the consent and encouragement
of H.M. Inspectorate of Ancient Monuments, brief excavations
were undertaken during two years with the help of Mr. Peter Rowe,
then headmaster of Cranbrook School, and a few of his pupils. In
1982 and 1983 some members of the archaeological group associated
with Tenterden Museum assisted on several occasions.
At the outset only two parallel trenches, outlined on the plan (Fig.
3) as Areas I and II, were opened. The hard layer in Area I was
found, where intact, to be buried only 5-15 cm. It consisted mainly of
angular stones, typically 3-6 cm. in dimensions, mixed with pebbles,
a little slag, fragments of roofing-tiles, bottle- and sheet-glass,
crockery, clay pipes, brass curtain-rings and bits of bicycle: in short,
domestic and builders' rubbish, dating to no further back than the
eighteenth century.
The layer had evidently been cut into and turned over during some
previous excavation in the central and western portion of Area I
along its northern side near the boundary stone, although it seemed
unimpaired closer to the stream bank and the western fence. At the
west end a little patch was scraped right up to the brink in order to
investigate a comparatively small paving stone sticking in the bank
with its flat top about 10 cm. below the excavated surface. After
probing to discover its extent, this rare stone was left in situ and not
further exposed for fear of spoiling what might prove in future to be
important evidence of ford structure.
Area II also contained a top layer of road stones with the same
range of relatively modern rubbish, but its metalling was continuous
along the trench for nearly 4 m. It was spread directly on a lower
layer of similar metalling resting on clay at the west end and on a
rust-coloured, hard substance at the east end. The whole area
became waterlogged before it could be recorded properly. By
probing, however, we found another hard layer beneath the west end
clay and rightly suspected it might be continuous with the third hard
surface found further east, although there without clay overlying it.
In an attempt to drain Area II, we cut a narrow trench joining its east
end with that of an extension to Area I where we had discovered
hard, reddish slag, horse-shoes of various sizes and pieces of leather
boot. These finds might not be ancient, but they were lying lower
than the top metalling.
The ground surface and the upper road levels both sloped down-
76
THE ROMAN FORD AT IDEN GREEN
Roman Road and Ford Site,
Iden Green
\
^
I ^ ^ ^ Car l Track */^
/Gale ^- ^ ^
/ ^^^ '^
\ < ^ ' \i;\iv
\ v 1
\ Boundary Slone ^ '
Fence \ •. - ^ ' ^ "'X
\ .^"' ' ."-"' ^1
\ £ •-•"' ^ r -k < ^
\ ^^^K\Al 3/
\ SUca"1
^ \ vA-ly
• ' 0 1 2
I
L ho
J
Gale
rd
/ Fence
/ » • >
E l
"ay"
o
V \ Stream v. T
Held (-arm Lane
.Tributary ' / V
s
N
f
\
^
•\»4l)
Fig. 3. Roman Ford Site: Plan of Excavations, 1980-83.
77
C. LEBON
wards in Areas I and II. Was this feature a part of the camber of a
north-south aligned road, or did it simply show that the postmedieval
road was heading downhill to a ford or bridge situated, as
today, upstream and east of the hitherto assumed place for the
Roman crossing?
Complying with a request from the Inspectorate of Ancient
Monuments, we 'rescued' six of the largest Bethersden marble slabs
out of the stream and buried them in Area II in 1981." As to the
original use of these much displaced paving stones, it is tentatively
suggested that they may have served as pedestrian causeways on the
upstream side of the road-like part of the ford used by carts, riders
and animals. Such a construction would strengthen the crossing in
times of spate.
After the first two seasons, the site was freed from the passage of
cattle. It became a priority to open up a wide area near the bridge,
Area V on the plan (Fig. 3), in order to see more of the spread of iron
ore and slag, and to obtain dating evidence. The layer has been
represented on the plan by stippling which does not appear in Area I
because there we refrained from digging below the upper metalling,
noting that it had already been much cut into.
The road was undoubtedly based on a natural layer of iron which,
although of geological origin, was apparently supplemented by slag.
The bed seems in most places to be composed of iron nodules and
slag more or less cemented by iron pan, presumably deposited from
iron salts carried in the water which flowed over a wide area for
centuries, and still does at times.
Topping up with loose, small particles of road material and
domestic rubbish could never have been a successful method for
durable road repairs near the ford, since ordinary, small stones and
rubble, as well as light articles dropped, would be carried into the
water, particularly at a low part of the bank just downstream of the
modern bridge.
The evidence has already pointed to an early ford having led into
Field Farm Lane. Standing down in the stream and examining the
banks carefully (Plate II), one could see an outcropping band of iron,
10-20 cm. thick, running horizontally half-way up the north-west
bank from where a modern drain-pipe projected under the bridge to
its obliteration among the exposed roots of the alder tree, a distance
of about 2.40 m. (unfortunately cut at both ends). The opposite bank
showed a corresponding, but thinner, band, both under the wooden
4 The stones are all about 16.5 cm. thick. The largest has a maximum length of 87
cm. and a width of 66 cm.
78
THE ROMAN FORD AT IDEN GREEN
PLATE II
••:-
(Photo. N.R. Aldridge)
Roman Hard-core, seen outcropping in Stream Bank.
bridge and where the almost vertically cut bank turns away up the
tributary on the threshold of Field Farm Lane.
A paving-stone lying under the bridge may well have been part of
the causeway. Another lies on the surface where Field Farm Lane
begins. It is probable that the ford was a structure about 3 m. wide
and that its length was greater than a span which could easily be
bridged, supposing the stream to have been wider than today and no
less liable to flooding.
Near the north-east corner of Area V silty clay replaced the iron
surface and gave it an edge pointing to where the drain-pipe emerged
under the bridge. So here is only a recent drainage trench and not the
edge to the original road which is more likely to be found intact
further back from the ford.
CONCLUSIONS
Map and field work on the course of the Roman road or roads
through Iden Green have been combined with excavations to investi-
79
C. LEBON
gate afresh the site and structure of the Roman ford in the valley
between Hemsted (Benenden School) and the Iden Green ridge.
The course south of the ford can hardly have been up a steep spur
where no track remains, as claimed by Margary. An old lane to the
south-west of the ford seems more likely to be medieval or Tudor, for
it shows no similarity of metalling, nor continuity with the Roman
levels north of the stream. But the sunken lane to the south-east of
the ford vicinity does contain iron nodules and slag similar to that
composing the earliest road surface found north of the stream. I now
believe that Field Farm Lane dates from Roman times. It was part of
a route which passed through Iden Green hamlet and continued via
Standen ('Saunding') Street towards a wharf on the Hexden, probably
above the site of Hope Mill where the marshes begin (Fig. 2).
The position of a road junction to link Iden Green with Bodiam, on
the main Rother, remains uncertain and perhaps less important,
considering that iron and timber for export would tend to be taken to
the nearest navigable river. Thus, long overland haulage would be
avoided wherever possible, especially by the well-organised Classis
Britannica which was active in the eastern Kentish Weald as well as in
Sussex.
Although evidence does not warrant the easy assumption that the
Classis Britannica constructed the Iden Green ford, certainly byproducts
of local iron-works were used there." The ford was built up
on foundation materials similar to those of the roads it connected.
There is a conspicuous out-cropping of a stratum of iron substances
under, and immediately downstream of, the present-day bridge.
This, I suggest, was the basic ford material which carried some of the
massive paving-slabs now widely dispersed. The valuable 1935 photograph
showing a pavement fails to provide sufficient evidence for
deciding the direction of the crossing. It is quite consistent with
postulating a passage south-east into Field Farm Lane, as today. In
fact, the slope of the pavement appears to descend that way, i.e.
towards the camera. Today, this is the general trend of the surface
and of the layer which formed the foundation of the Roman road.
The excavation was too late to find the pavement intact.
ACKNOWLEDGEMENTS
I am grateful to the Honorary Editor, Mr. A.P. Detsicas, who drew
the excavation plan from my poorly executed version; and to Mr.
5 On Little Farningham farm in Cranbrook, some fifty CLBR stamped tiles were
found in association with Roman iron-working two miles north of the ford site. (Arch.
Cant., lxxii (1958), lx-lxii.)
80
THE ROMAN FORD AT IDEN GREEN
N.R. Aldridge for photography and other work on the site; also, to
Mr. H.H. Margary, Mr. H.W. Edwards, executor of O.G.S. Crawford,
and the Director-General of the Ordnance Survey for permitting
me to use already published material among the illustrations to
this article.
81
