r

172

MAESBURY
MAESBURY CASTLE,
CASTLE, SOMERSET
SOMERSET

.' .'
-;" \'J

"','

"','

Maesbury Castle, Somerset

(O.S. 6 in. to 1 mile, Somerset 41 S.E.

N.G.R. ST 6147.

I
!,

* Mr. H. Freke of the Department of Geography has drawn the map from
which this figure is made.

\\\

~\ \" \\\
\'\,'

"

"'~" fJ~~':~::;\~... ..:-,\.\

....' ~ ' ,'.

By
E. K. TRATMAN, O.B.E., M.D.S., F.S.A.
F.S.A.
This is one of the most often mentioned contour hill forts of Somerset.
Presumably it dates from the Early Iron Age. It is mentioned in several
places by Allcroft (1908). Bothamley (19II) and Burrow (1924) have both
given short accounts of it as a bivallate work and both mention a detached
outwork to the south-east very faintly marked. Gray (1908) has given the
fullest account. He could find no record of any objects having certainly been
found within the camp, though he records finds made in the vicinity. He
places on record the alleged discovery of a polished flint axe near the water
on the north-west inside the camp and correctly notes that its discovery,
if authentic, could have no bearing on the date of the construction of the
the
camp. Gray (1908, p. 81) also consulted Boyd Dawkins about finds inside the
camp and the latter confirmed the view that none had been made. Boyd
Dawkins (1906) considered that the main rampart contained a dry stone wall
similar to that found at Worlebury.
The object of the present paper is to give a revised account of the camp
and to give some tentative conclusions about the site based on the facts
noted. In several ways the present author disagrees with Gray.
The camp is essentially a bivallate contour fort making use of the steep
steep
slope of the hill on the south and of the gentler slopes elsewhere. The camp
lies at the western end of a long ridge of high gr~d at over 900 ft. above
O.D. It is ovoid in form and the steepest natura1
natural slope is on the south. To
To
the east the ground is fairly level for some way ~efore it rises again. On the
north and west the slopes are gentle, while the area enclosed within the inner
bank is fairly level from east to west but has a genrte
gende slope down from north
to south. The area enclosed within the inner ram~r
is about 7 acres out
ram~ris
of the total of over 15 acres within the fortifications, the limit of which is taken
to be the outer lip of the outer ditch. About 400 yards to the north is the
Rom~n
Roman Road from Charterhouse on Mendip running approximately south
southeast towards Old Sarum. About 2 miles to the east is Beacon Hill with its

,'\

",
",,\,\\
,\'

.',

Figure 38.*)
38.*)

. \\"",
: . ' ,,'"" "\ .
~ ~ .:. . .' : ' '0:. " "..",
"

,'"

"',~

~
,' ...' "
. ~.,,:::::,,,,,,,,, <,,,,,
::~ ..."' "' ...'

~~

\~~'

\~.

~;

...\.

~ .... 2 ::..~.... "",'

"".. .
'" ~ ....

'::::~.:,

:::-::~ ....

" ...~

~::~:~

mUff
-----

- -

;~~~ ~ ~
:.~::= ~ :.
----_....

_- -

~~\\~ ~ ~

\~~~r\

\;~;l\..\
' .......... , ..- 't'
~~~~-:. ~

-:..-=.-:.:.";.
.............. ,

-:;'-:;.-:;.-:-:.

~ ,II'
,,-,

-:~:.

-:)\\&:

.tl"

S~-:~-:;'j:~
7<0 __ ~:"

.. -

............. ,
~:. ~\i;;-.. ; ~
..~ ; . !'\~

-:,..

l
I
I

~
I

considerable group of b
date, and the rest are
crosses the west end of 1
between Dinder on th,
from south-west to nOI!
It has been observed:

I
I

tiET

MAESBURY CASTLE, SOMERSET

173

~erset

R.

ST 6147.

Figure 38.*)

).S., F.S.A.

Ibtour hill forts of Somerset.

;. It is mentioned in several
MI Burrow (1924) have both
W.both mention a detached
;.=Gray (1908) has given the
bbjects having certainly been
lis made in the vicinity. He
~
~ed flint axe near the water
~ notes that its discovery,
~ of the construction of the
~s about finds inside the
~ne had been made. Boyd
1st contained a dry stone wall
L
M

(revised account of the camp

~Ute site based on the facts
l!t.
p'ees with Gray.
ifort making use of the steep
;40pes elsewhere. The camp
pound at over 900 ft. above
slope is on the south. To
~fore it rises again. On the
~ enclosed within the inner
~tle slope down from north
p.mpart is about 7 acres out
. . the limit of which is taken
yards to the north is the
iuming approximately south
~ east is Beacon Hill with its

nu

100

IIPbY

has drawn the map from

considerable group of barrows, some of which are known to be of Bronze Age

date, and the rest are presumably of the same period. The Fosse Way
crosses the west end of Beacon Hill from north to south. The parish boundary
between Dinder on the west and Croscombe on the east crosses the camp
from south-west to north-east. The interior of the camp is under cultivation.
It has been observed under varying conditions at different seasons under

174

MAESBURY CASTLE, SOMERSET

to drain off this w8I

Without this cuttial
probably be water-:i
The camp is thus:.
within the precinctl!

different crops but no crop markings have been observed at any time and
nothing appears in the air photographs. Perhaps the soil is too thick to
produce crop markings.
Gray considered those parts of the outer ditch that he recorded, on the
south along the north edge of Castle Hill Wood and along the western part
of the north side of the camp, to be modern and to have been made when
the boundary hedge was planted. Neither he nor other writers mention a
quarry ditch on the inner side of the inner rampart and there is certainly no
superficial evidence of one there now, 1959.

THE INTEWj
(

The width of _
north the portion C
west of the bound8
effects of cultivatiol
been used to better;
Croscombe parish ~
banks along eithed
ditch, that is fonnin
ditch. The banks 81
of 10 ft., sometim~
south-east gap but I
On the north, ~
on this interspace. ,~
the bank is about 3':
of about 15 ft .,
along the outer e~
as far as A. But her
but do become high
the flat interspace. J
detached mound at 1
part of the defencelo
On the south Ii
the outer edge of til
vertical drop of abo
trace of a bank aloDl
till at G, where the J
is almost certainly.
track to enter. ~
low flattened counte
and there has evide

THE INNER RAMPART

The spoil from the inner ditch seems to have been taken almost entirely
artifi
uphill to form the inner rampart and the sides of the hill to have been artificially steepened, presumably to improve the defence and to provide more
material for the bank. For most of the perimeter the bank stands about 5 ft.
above the interior of the camp but at the eastern end the bank height increases
to 10 ft. above the interior. Here the top is 20 ft. wide, and fla! with its
sides sloping very steeply to the base, which is estimated to be 35 ft. wide.
There are two gaps in the rampart, one to the south-east and the other to
the west. At the south-eastern gap traces of a stone core to the bank can be
seen on the north side of the gap. From the south side of the western gap
and extending east to the parish boundary there is a well-marked berm
about 6 ft. wide between the outer edge of the bank and the lip of the steep
descent to the ditch, K-K'. Traces of this berm can be seen in other parts.
The steepness of the sides of the bank together with the berm must indicate
a wall structure within the bank. This wall is probably about 20 ft. wide
and is probably still standing to a height of up to 6 ft. at the east end. If so
it would compare well with the dry stone wall of the bank at Little
Solsbury Hill camp (Dowden, 1957).

.wb!

THE INNER DITCH

This has a flat bottom and is not of the usual U form. Some of the
flattening may be due to cultivation. Its base now is about 15ft. wide and
lies about 6-8 ft. below the outer lip. The outer side of the ditch is remarkremark
ably steep. On the south of the camp the present bottom of the ditch is
estimated to be 18ft.
18 ft. below the top of the inner bank, and all round the
slope up is very steep. There is a wide gap at the south-east opposite the
gap in the inner bank and a smaller gap at the west end where the ends of
the ditch on either side of the gap there are splayed. On the north side
of the south-east gap the ditch usually holds water in the wetter months.
Along the north side of the camp from A-B (Fig. 38), the ditch holds water
all the year round.. In comparatively recent times an attempt has been made

From F-G thi

about 12ft. wide. J

':

l_-

7_=...5_-

',,-
..
/ __
\"--

MAESBURY CASTLE, SOMERSET

175

"

:observed at any time and

~ the soil is too thick to
.. that he recorded, on the
!lid along the western part
~to have been made when
~ other writers mention a
~ and there is certainly no

'1'
f,

taken almost entirely

the hill to have been artifi

~ce and to provide more
the bank stands about 5 ft.

ill the bank height increases

Jt. wide, and flat with its

~ated to be 35' ft. wide.

both-east and the other to
core to the bank can be
side of the western gap
~ is a well-marked berm
.. and the lip of the steep
~ be seen in other parts.
ith the berm must indicate
>robably about 20 ft. wide
6 ft. at the east end. If so
iaU of the bank at Little

idl

IIa1 U form. Some of the

.. is about 15ft. wide and
ide of the ditch is remark
Ilt bottom of the ditch is
r batik, and all round the
lie south-east opposite the
cat end where the ends of
Iayed. On the north side
Iter in the wetter months.
'38), the ditch holds water
an attempt has been made

to drain off this water by cutting through the outer bank approximately at B.
Without this cutting the whole length of the ditch along the north side would
probably be water-filled to a depth of a foot or more above the present bottom.
The camp is thus one of the very few that have a copious water supply
within the precincts.
THE INTERSPACE BETWEEN THE INNER AND OUTER
DITCHES
The width of this space varies somewhat between 30 and 40 ft. On the
north the portion east of the parish boundary is very much flatter than that
west of the boundary. The difference would seem to be mainly due to the
effects of cultivation. The land in the parish of Croscombe has, in the past,
been used to better advantage than the land in Dinder parish. In the part in
Croscombe parish there are very low, never more than one foot high, wide
banks along either side of the interspace; along the outer edge of the inner
ditch, that is forming a counterscarp bank, and on the inner edge of the outer
ditch. The banks are about 10 ft. wide and there is a level area between them
of 10 ft., sometimes more. These very low banks can be traced across the
south-east gap but not across the west gap.
On the north, at C, in Dinder parish there are two well developed banks
on this interspace. On the outer edge, that is the inner lip of the outer ditch,
the bank is about 3 ft. high and 15ft. wide at its base. Then there is a space
IS ft., which is level, to the less well developed counterscarp bank
of about 15
along the outer edge of the inner ditch. This form of twin banks continues
as far as A. But here the banks do not fuse as they approach the western gap
but do become higher and steeper sided reaching l! height of over 4 ft. above
the flat interspace. Here shallow quarrying at E confuses the picture and the
detached mound at D appears to be spoil from this quarry and not an outlying
part of the defences.
On the south side of the western gap from F-G the modern hedge marks
the outer edge of the interspace between the two ditches. There is a nearly
vertical drop of about 8 ft. to the outer ditch. Along this stretch there is no
trace of a bank along the outer side of the interspace, which gradually narrows
till at G, where the modern track enters, it is only 15
IS ft. wide. This narrowing
is almost certainly largely due to the cutting that has been made to allow the
track to enter. Along the inner edge of the interspace there are traces of a
low flattened counterscarp bank of the inner ditch. These traces are very faint
all,
and there has evidently never been much of a bank here at allTHE OUTER DITCH
From F-G this is seen as a very shallow ditch, a few inches deep and
about 12 ft. wide. At the modern track the ditch is obliterated. For the rest

176

MAESBURY CASTLE, SOMERSET

of the perimeter it is continuous, save at G and G' where it has been filled in
in modern times, right round to the north side of the western gap, where its
end has been destroyed by the quarry E. It is also interrupted by the bank of
the Outwork (see below). Its depth and definition vary considerably. Its
width averages about 15ft.
IS ft. between present-day lips, and its present bottom
is, in places, 6 ft. below the platform of the interspace between the ditches.
The depth is variable. There is no real evidence for a counterscarp bank
on the outer side except for a short stretch on the north immediately west of
the parish boundary. The ditch is only just discernible from L-L'.
The modern hedge line roughly follows the centre line of the ditch with
a tendency to lie towards the inner side. Gray, as noted above, considered
this ditch to be modern but there seems no doubt that it is ancient though
altered, when the hedge was made, by re-excavating the outer side of the
ditch and throwing the earth inwards to form the hedge bank within the
hollow of the ditch. The double profile so formed is distinctive. The outer
side of the ditch is steep.
THE OUTWORK
Bothamley and Burrow refer to a detached work to the south-east. This
could not certainly be identified in 1959. On the other hand, there is, beginning
at H and running for 30 yards to I, a much ploughed down bank about 1I ft.
high by 10 ft. wide, and outside it is the corresponding shallow ditch. The
bank blocks the main outer ditch, which has been turned outwards to be
continuous with the ditch of the outwork. Beyond I the outwork cannot be
traced and the manner in which it ends abruptly indicates that it was never
built beyond this point.
THE SOUTH-EAST AND WEST GAPS
Gray (1908) considers the south-east gap to be the origirial
origi~al entrance, having
been widened in modern times,
times, and the west gap to be recent. Bothamley
and Burrow thought the reverse, and so does the present author. The south
southeast gap is wide and the central 30 ft. appear to be more ancient than the rest,
and perhaps connected with the faintly marked sunken way coming up from
the south-east. On the other hand the cut ends of the inner bank oppose each
other with no indication of inturning or overlapping. The low
low banks on the
platform of the interspace between the two ditches and the outer ditch itself
are all continuous across the gap. There seems to be little doubt that the
whole of the south-east gap is modern. Only excavation can prove the validity
of this conclusion, but it does seem unlikely that a form of straight through
entrance would be made at the most easily accessible place and where the
inner bank has been heightened, presumably to increase its defensive
value.

Thewest~

fusion producet~
inner bank. N~
angles to the preli
continues past ~
been much ni.~
bank south Of~
about 20 yard&,
part is shown '.
cannot now be
thrown into the ill
side the west ga~~
Its form is
entrance. A simi
Combe. Here
'-~".'i

unusuit

th&.
~

.l1

The inner bIi

width of the ~
thereon of spoil
twin banks sud
second bank, by~
this plan, for r~
be a sti1l1ater addl
completed.
The inner bai
Sandstone qu~
suitable for such J
those of Little Sol
(Falconer and ~
inner bank at Mal
the floor of the ia
both ditches,
seem to indicate d
Richardson 1957.;
copious water ~
the very existence I
site for the fortre8l
east. The absence
would be in keepiJ
rampart in its IDII
feeble countel'1lcaJ

ii

an&

1
)
MAESBURY CASTLE, SOMERSET

P'e it has been filled in

The western gap presents an entirely different picture in spite of the con
confusion produced by modern alterations, which include the gap near A in the
inner bank. North of the gap the inner bank is running nearly west at right
angles to the present end of the inner bank on the south side of the gap. It
continues past the end of the bank on the south for about 20 yards. It has
been much mutilated. Further, Bothamley (1911) describes how the inner
bank south of the gap turned abruptly outwards, west, and continued for
about 20 yards, thus matching the bank north of the gap. This out-turned
part is shown on his plan but not on that of Burrow. This out-turned part
cannot now be traced. Presumably it has been levelled and the material
out
thrown into the inner ditch. This would account for the large flat area outside the west gap. There seems to be no doubt that this is the original entrance.
Its form is unusual and appears to be the reverse of the more usual inturned
entrance. A similar form of entrance occurs at the small camp of Tap's
Combe. Here the entrance was funnel shaped (Thorburn, 1926).

~tem

gap, where its

!fiupted by the bank of
flPY considerably. Its
pad its present bottom
c.between the ditches.
~;la counterscarp bank
~ immediately west of
tfrom L-L'.
sli.ne of the ditch with
ICed above, considered
~ it is ancient though
e outer side of the
lI'dge bank within the
'.llistinctive. The outer

r
the south-east. This

~, there is, beginning

fawn bank about I ft.

shallow ditch. The
Ifmed outwards to be
IlIIe outwork cannot be
~ that it was never
~

.'

i GAPS

!igUial entrance, having

lie recent. Bothamley

ill author.

The south
ancient than the rest,
~way coming up from
,mer bank oppose each
. e low banks on the
the outer ditch itself
Jllittle doubt that the
Lan prove the validity
~ of straight through
place and where the
~e its defensive
I

I.

DISCUSSION
The inner bank and ditch seem to form a unit and to be complete. The
width of the flat interspace between the ditches coupled with the placing
thereon of spoil from both ditches in relatively small amounts so as to form
twin banks suggests a plan for strengthening the defences by adding a large
second bank, by deepening the inner ditch and cutting the outer ditch, but
this plan, for reasons unknown, was never completed. The outwork seems to
be a still later addition possibly to form a cattle kraal and that too was never
completed.
The inner bank almost certainly contains a dry stone wall. The Old Red
Sandstone quarried from the side of the hill and from the ditches would be
suitable for such a structure and its postulated dimensions are similar to
those of Little Solsbury Hill Camp, which is of Early Iron Age "A" date
(Falconer and Adams 1935, and Dowden 1957). But this interpretation of the
inner bank at Maesbury is based on only the slenderest data. The flatness of
the floor of the inner ditch together with the steepness of the outer sides of
both ditches, and the flatness of the interspace between the ditches would
seem to indicate that Maesbury Castle was a Belgic stronghold (Wheeler and
Richardson 1957, pp. 8-14), and therefore very late. The existence of a
copious water supply within the camp points in the same direction, and
the very existence of this supply may well have determined the selection of the
ground to the
site for the fortress rather than the higher and less well watered ground
east. The absence of crop markings suggests an absence of pits and this again
1943, p.
p. 58).
58). The
The main
main
would be in keeping with a Belgic date (e.g., Wheeler 1943,
rampart in its massive strength, the flat
flat interspace between the
the ditches, the
feeble counterscarp
counterscarp bank
bank of
of the
the inner ditch
ditch and
and the similarly
similarly feeble
feeble bank
bank of
of

178

MAESBURY CASTLE, SOMERSET

the outer ditch all present similarities to those which have been described as
the Fecamp series (Wheeler and Richardson, 1957). The entrance at the west
end does not in its present form show the typical Fecamp type of massively
in-turned entrance but instead an out-turned one. All the same, the general
conclusion is that Maesbury is possibly a Belgic stronghold, constructed by
them and possibly by the Dumnoni and is quite a fair example of the Fecamp
series. This would put the date of construction in the later part of the
1st century B.C. and possibly as late as the beginning of the 1st century A.D.
It is tempting to consider it as being in process of alteration at the time of the
Roman conquest of Britain.
Belgic pottery is amongst that found at Kingsdown Camp, Mells (Gray,
1930). This camp is about 6 miles east of Maesbury. It is a very late and
feeble work. The evidence from the excavations points to it being probably
entirely of post-Conquest date and there is no evidence for dating the little
Belgic pottery found any earlier. The other camps nearby are Newbury,
Wadbury and Tedbury. Only the last has any massive defences which, for
part of the west side, consist of twin large banks, probably containing dry
stone walls, with a very wide ditch between. It is too overgrown to be
examined properly. It is associated with two streams. These features could,
very doubtfully, be interpreted as indicating a Belgic origin. On the other
hand trial excavations at Burledge Camp (ApSimon, 1957), which is about
6 miles north-west of Maesbury, yielded no Belgic pottery. So perhaps
Maesbury, if it is Belgic, marks the western limit of pre-Conquest Belgic
penetration of the area.

Aue

A Recons;.,

<

t
Prof.
.,,~

In studYioa
from Flint Ja~
by its affinities
excavations of

1
,.

REFERENCES

Proc. = Proceedings, University of Bristol Spelaological Society

ALLCROFT, A. H., 1908, "Earthwork of England".
1911, "Croscombe-Maesbury Castle", V.C.H. Vol. II, 491-492.
BOTHAMLEY, C. H., 19II,
BoYD DAWKINS, W., 1906, "Early Man in Somerset", V.C.H. Vol. I, 203.
BURROW, E. J., 1924, "Ancient Earthworks and Camps of Somerset", 80.
DOWDEN, W. A., 1957, "Little Solsbury Hill Camp", Proc., Vol. 8, 18-29.
FALCONER, J. P. E., and ADAMS, S. B., 1935, "Recent Finds at Little Solisbury Hill
Camp, Near Bath", Proc., Vol. 4, 183-221.
GRAY, H. ST. GEORGE, 1908, "Maesbury Castle", Proc. Som. Arch. Soc., Vol. 53,
Pt. ii, 73-81.
ArchtEol., Vol. 80,
- - 1930, "Excavations at Kingsdown Camp, Mells, Somerset", Archaol.,
59-9 8 .
59-98.
THORBURN, M., 1926, "Field Work", Proc., Vol. 2, 279-280.
WHEELER, SIR MORTIMER R. E., and RICHARDSON, K. M., 1957, "Hill-Forts of
Northern France", Rprt. Res. Comm. Soc. Antiq. Lond., No. XIX. Especially
pp. 8- 1 4.
WHEELER, MORTIMER R. E., 1943, "Maiden Castle, Dorset", Rprt. Res. Comm. Soc.
Antiq. Lond., No. XII. Especially p. 58.

Fig.

3~

Dr. W. N. E~
British Musem
remains in the