More information : This camp has been re-assessed in connection with RCHME's survey and
publication of Roman Camps in England. The following descriptive
account is taken from the published text.
Previously recorded under NAR Number SE 79 SE 45; now assigned unique
identity.
The cluster of military earthworks at Cawthorn consist of four major
elements. A camp, C, of unusual polygonal design, is partly overlain
by a slightly later fort, D, which is probably datable to the late
1st century (NAR SE 79 SE 45; Jones, 1975, 140-1 (1a). To the E of
the camp are two structures which have often been classified as
camps; on balance, however, the more westerly of the two is best
regarded as a fort, A, which was subsequently provided with an annexe
on its E side, thus forming a much larger defended area, B.
The earthworks were excavated between 1923 and 1929 (Simpson 1926
(1b); Richmond 1926 (1c); 1929 (1d); 1932 (1e)). The identifying
letters A-D usually ascribed to the earthworks are retained here, but
most of the highly speculative functions and relationships put
forward by Richmond have now been discarded. Certainly there seems
to be no overriding reason to consider the sites as practice works.
The few finds suggested that occupation may not have continued later
than c AD 120.
SE 7852 9012 (FCE). Only 75 m to the E of camp C (SE 79 SE 66) are
the impressive earthworks of a fort, A. Since internal and external
claviculae survive at its gates, thus conforming broadly to the rules
of contemporary castramentation, it has often been loosely described
as a camp. However, it is markedly dissimilar from the other camps
in England and it seems more reasonable to classify it as a fort to
which, subsequently, an E annexe, was attatched. The scale and the
method of construction of the defences seem to make this
reclassification more appropriate. The interior of the fort and of
its E annexe were still obscured by bracken, piles of timber and
brash at the time of the survey by RCHME. No detailed description or
analysis will be provided here (NAR SE 79 SE 45), but it is
appropriate to underline the features that may illuminate the
dividing line between the two monument types.
On its W side, on to which the three gates of camp C seem to face,
the defences are 14 m across overall and consist of a rampart, an
outer ditch, and a slight counterscarp bank only 0.2 m high. The
rampart is 1.7 m high internally and 2.8 m above the bottom of the
ditch which is 1.6 m deep. The position chosen, on the crest of the
ridge, is similar to that of fort D. The true summit is just to the
NW of the centre of the site and so most of the interior slopes
gently to the S. Excavation revealed that the rampart consisted of
material upcast from the ditch, with a thick capping of turf which,
the excavator suggested, was added in the second phase when the
annexe was constructed on the E. However, it is difficult to be sure
quite what was revealed in the 1920s: the sections drawn (Richmond
1932, 25 (see auth 1e)) do not corespond to the description in the
text. At one stage or another there seems to have been extensive use
of timber in the construction of the defences. There was a series of
vertical posts on the lip of the ditch 10ft (3.0 m) apart with, some
6ft (1.8 m) farther back, a continuous trench for a pallisade, cut
into the upcast material. Between the gates on the S and E, deep
holes for stout posts were found along the rear of the rampart, 1.8 m
behind the palisade and 1.5 m apart. Such extensive use of timber
makes it clear that this was a permanent or semi-permanent fort and
not a camp.
Along the E side of the fort, the top of the rampart is markedly
broader than elsewhere and its height has been reduced to 1.0 m
internally and 1.0 m to 1.7 m externally. This change was presumably
madfe when the annexe was added. The security provided by the ditch
was not seriously compromised for it is still 1.0 m deep. It appears
unlikely that the intention was to slight the defences on this side.
This reinforces the impression that the additional defended area to
the E was intended as an annexe to fort A and not to form one larger
integrated entity (Richmond's B).
The annexe, much more like a camp in character, is laid out on an
irregular subrectangular plan. This is especially marked on the N
where the defences were set along the natural crest; they had to be
realigned at the gate on that side in order to take account of a re-
entrant valley which provides access down the escarpment. The bank
stands up to 0.7 m high internally and was found to be constructed of
turf; the ditch was 8ft 9in (2.7 m) wide and 3ft 9in (1.1 m) deep
(Richmond 1932, 52 (see auth 1e)). A counterscarp bank, now no more
than 0.4 m high, was provided throughout. These defences are
altogether slighter than those of the fort. The two gates, on the N
and S, are protected by internal and external claviculae and are
offset to the W, confirming that the internal layout was focused on
fort A. No gate was provided in the rear, E, rampart.
The single gates in the W and S sides of fort A are each defended by
internal and external claviculae, none of which was provided with a
ditch; the N gate has an internal clavicula only. On the W, the gate
has been damaged by the later hollow-way. In each case the internal
clavicula, at only 0.6 m to 0.9 m in height, is appreciably lower
than the adjacent ramparts. The report of the excavations (Richmond
1932, 22-30 (see auth 1e)) confirms the impression that these
claviculae were additions to the defences. In the original design the
ditch of the fort was 15ft (4.6 m) wide and 7ft (2.1m) deep, and was
driven across the line of the gateways. The latter, which were
originally inturned and rectilinear, were defended by traverses on
the E, W and S; the ditches of these were soon back-filled. The W
example was unusual in that it was keyhole-shaped on plan, and it
was suggested that this and the other traverses were never completed.
No traverse was found at the N gate; either it had been eroded away
or the steep natural slopes rendered such a provision superfluous.
In a subsequent phase, when the claviculae replaced the traverses,
only an internal clavicula was provided at the N gate. The
excavations uncovered a single E gate, defended by a traverse, only
about 40 m from the SE angle. No clavicula was subsequently provided
here before the gate was demolished (Richmond 1932, 30, pls 7 and 20
(see auth 1e)).
The E and W sides of the fort are parallel but the E defences are
slightly longer than the other three sides. This plan, combined with
the positions of the gates, must have affected the regularity of the
internal arrangements. These were briefly studied by Richmond (1932,
29-30(see auth 1e)) who was of the opinion that only some of the many
pits and ovens identified belonged to the first phase of occupation.
Subsequently, at the time that the defences were extended to the E,
turf structures were laid out which he interpreted as wind breaks
rather than the lower courses of buildings. These now stand up to
0.6 m high as banks up to 2.5 m broad; some of the areas enclosed
appear to be as little as 2m across internally. A single line,
extending N to S within the fort as a discontinuous bank, suggests
some regularity of planning. The layout of banks in the E annexe is
rather clearer in places, especially in the SE corner. These were
interpreted by the excavators as tented barrocks. Despite this,
there is insufficient clarity to define a conventional road-plan and
it is not clear which way the fort originally faced. Some of the
turf structures seem to be laid across internal lines of access to
the N and S gates of the fort, and are thus not likely to be strictly
contemporary. A mound close to the centre of the fort, now standing
0.9 m high and measuring about 9 m across, was identified by Richmond
as the tribunal, a dais for the commanding officer. This was
constructed by adding a turf platform to the N side of an earlier
mound. The latter seems to have been a prehistoric barrow with a
central pit which had been disturbed by earlier excavators (Richmond
1932, 61-3 (see auth 1e)). Aerial photography by RCHME in 1993
revealed that the barrow was surrounded by a circular ditch more than
20 m in diameter (see SE 79 SE 55). Further investigation is
required, for the date, associations and functions of all the
features within the fort and its annexe are unclear. Other
structures, apparently not wholly dissimilar, survive outside the W
and S side of the defences.
A low bank, nowhere more than 0.5 m high, defines two sides of a sub-
rectangular enclosure which takes in most of the NE corner of fort A.
Its layout suggests that it may be later in date than the other
internal features but this cannot be demonstrated without excavation.
Its function is unknown (see SE 79 SE 64 ). Full information is
included in the NMR Archive. (1)
An air photographic evaluation (2-3) was undertaken by the EH Aerial Survey section, in conjunction with the Metric Survey section, as part of a wider research project investigating Cawthorn Camps from 1998-2002. In addition to the air photographic work, this research has included geophysical survey, topographic survey of Fort A and Annexe B, and two seasons of excavation. (4-5)
Photogrammetric survey using specially commissioned, large scale air photographs has enabled the production of a detailed plan of the earthwork remains at a scale of 1:500 and an accuracy of 10cm or below. Rectification of this plan with other photographs for the site has enabled further interpretation of the earthworks. In particular, use of photographs taken in 1925 of the excavations undertaken by Simpson, Kirk and Richmond in the 1920s (sources 1b-1e), has enabled the positive identification of many extant earthworks on site as remains of the 1920s trenches and spoil heaps. These excavations were particularly extensive in Fort A and Annexe B. Similarly, features attributable to World War II activity have been identified from photographs dating to 1945 and 1946, including ordnance craters in Fort A and Annexe B. New earthwork features, some as little as 10cm in height, have been recorded in the interior of the forts, camp and annexe. Within both Fort A, Annexe B and Camp C numerous embanked features are evident; many of these features are tentatively thought to be of one general phase and possibly contemporary with the main defences. This theory is based on observations on form, alignment and condition of features. The limited excavated evidence, described below, indicates a Roman date for these. (3)
However, the date of many of the interior earthworks is yet to be proven. Richmond (source 1e) considered them to be Roman in date, military in nature and contemporary with the main defensive earthworks. More recent research (6-7) has proposed a post-Roman date for some of the features, in particular, the system of streets and enclosures in the south-east of Annexe B, also certain of the pits excavated by Richmond which he termed as 'officers' dugouts'; these latter features are now considered to represent possible sunken featured buildings of early medieval date. The air photographic interpretation has identified further depressions across the site, but particularly in Fort A and Annexe B which is where all of Richmond's 'dugouts' were located. These now require further investigation as potential sunken featured buildings.
The excavations undertaken in 1999-2000 by Dr P.Wilson (EH) comprised ten trenches, four of which examined Fort A. Three trenches investigated the defences of Fort A on its north, west and east sides; the trench on the east side also included the re-excavation of one of Richmond's 'dugouts' located on the top of the east rampart, which is now re-interpreted as a probable early medieval sunken featured building as suspected. The excavation of the east rampart has called into question some of Richmond's findings, in particular elements of timbering which he recorded as part of the rampart structure. The only evidence found for a timber structure was the palisade visible as a trench running along the top of the rampart. The eastern rampart appears to have been of one phase, however, there was evidence to corroborate Richmond's conclusion that there were two phases for Fort A. The trenching of the north and west ramparts of the fort produced evidence for two phases of construction for both ramparts. The earlier rampart was of dumped construction associated with palisade trench, with a capping of clay representing the later phase. The remains of four ovens were also discovered - three were cut into the rear of the northern rampart and one was found behind the west rampart. Three of the 1999-2000 trenches investigated the internal embanked structures in Fort A and Annexe B, the one in Fort A was located across the lines of east-west banks delineating one side of the north-east 'enclosure', see SE 79 SE 64. These trenches produced evidence of streets and multi-phase buildings. One of the trenches in Annexe B produced an archaeo-magnetic date of late 1st and early 2nd century AD date. For the other two trenches, the limited finds evidence, comprising pottery and glass melon beads, also indicate a Roman date. (8-9)
|