From
http://www.academia.edu/820052/Biggins_J._A_and_Taylor_D._J._A._2007_The_Roman_Fort_at_Castlesteads_Cumbria_A_Geophysical_Survey_of_the_Vicus_Trans._C._and_W._Soc._VII_15-30:
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
1
The Roman Fort at Castlesteads
A geophysical Survey of the Vicus
by
J. Alan Biggins and David J. A. Taylor
February 2007
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
2
The Roman Fort at Castlesteads, Cumbria: a Geophysical Survey of the Vicus
by
J. Alan Biggins and David J. A. Taylor
1. Introduction
The Roman fort at Castlesteads (Camboglanna), is positioned between the Wall forts of
Birdoswald and Stanwix, and sited on an escarpment above the Cam Beck. The fort is
unusual in that, although it is classified as a Hadrianic Wall fort, it is not built on the line of
the Wall but is positioned some 400 metres to the south. The easier line, which the Wall now
follows, was probably taken in preference to the direct line between Milecastles 56 and 58,
which would pass through the fort. One explanation why this route may have been taken was
to avoid constructing the Turf Wall across the Cam Beck and up a steep escarpment to the
east.
Details about the fort itself are meagre as it was largely destroyed in 1791 when the
present house was built, and walled gardens laid out over the site of the fort, although
Hutchinson (1794) in his County History of Cumberland recorded some details of the fort.
Castlesteads is thought to have been occupied during the Hadrianic period by a cohors
quingenaria peditata (Breeze and Dobson 2000, 54).
An excavation by Richmond and Hodgson (1934, 159-165) attempted to establish if
any trace of the fort remained after its destruction. Some evidence of gates towards the east
and west of the fort, together with the curtain wall was identified with a 4.8m wide ditch and a
berm 3m wide, although little was seen within the intra-mural area itself. Significantly, it was
estimated that some 100 feet (c. 30 m) of the fort had fallen into the Cam Beck. Due to the
siting of the fort, with the north gate overlooking the steep escarpment above Cam Beck, it is
likely that the porta praetoria faced east instead of the more usual northern aspect, facing the
main gate. This would mean that the fort was similar in general layout to the east-facing fort
of Housesteads.
The excavators found evidence of a rampart and ditch from an earlier turf and timber
fort set on a different alignment to the later stone fort (ibid., 163-5). Trenching at the south
east angle revealed the remains of a turf rampart, at least 3m wide, resting on flagging and
stones set in clay to the rear of the Stone Wall. This rampart base is probably the remains of
the earlier turf and timber fort. An east facing scarp, 0.4m high, parallel to the east rampart
was discovered during a survey by the Royal Commission on the Historical Monuments of
England in 1991. Richmond and Hodgson also pointed out that the later stone fort was not set
out centrally within the diversion of the Vallum to the south but asymmetrical to it; a point
made earlier by Collingwood (1922, 202). Collingswood’s explanation for this irregularity
was that the fort could have had an annex built on to the east. Richmond and Hodgson
suggest that this inconsistent construction was possibly due to an earlier fort being set further
to the east with a north-east orientation, which would have then been symmetrical with the
Vallum diversion. This is a certainly a possible explanation and reflects a similar relationship
between the Vallum and the stone fort at Birdoswald (Wilmott 1997, 44-5).
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
3
An external bath-house to the north-east of the fort was located and partly excavated in
1740, although the exact position and plan is not known. The Vallum survives as a buried
feature throughout this section with no remains visible above ground. Its route has been
confirmed by Haverfield who cut trenches in 1898, 1901 and 1902 to determine its course.
The vicus, which is usually associated with Roman forts, is not now visible, except
in places where it is seen as very shallow platforms. However, a letter from Richard Goodman
writing to Samuel Gale in 1727 mentions traces of an extensive settlement on the slope
towards the south east of the fort. He noted the existence of foundations of walls and streets,
which were being removed to construct new buildings and to allow the land to be ploughed.
A temple outside the Castlesteads fort to the south-east, which had ‘fallen in through age’ was
repaired by a centurion, Gaius Julius Cupitianus, and dedicated to ‘the Mother goddesses of
all nations’ ([Deabu]s| [Mat]ribu[s]| omnium | gentium | templum | olim uetus | tate
conlab|sum G(aius) Iul(ius) Cu/pitianus | (centurion) | p(rae)p(ositus) restituit); (RIB 1988).
The high quality of the geophysical survey data almost certainly reflects the absence of
recent sustained deep ploughing over the site. It is possible that the apparent absence of subsurface
features to the south of the survey may indicate greater destruction by ploughing than
those to the north; this could be reflected in the level of occupational debris present on the
surface of that part of the site. Conversely, taphonomic processes downslope may have
preserved deeper levels by colluviation. Many large stones can be seen on or just below the
surface over the area of the settlement. Also, extensive surface collections of pottery, brick
and tile were seen in stubble after the field had been laid out to arable cultivation in 2001.
2. Methodology
On this site, magnetic survey was the preferred method, which was carried out in three phases
in 1999, 2000 and 2001 during periods of variable weather. The site of the fort is set within
woodland and the area surveyed (17.9ha in total), is located to the south of the field boundary
dividing the woodland from open pasture. Geoscan FM36 fluxgate gradiometers were used to
carry out a magnetometry survey employing 1.0 m parallel traverses with 0.5 m sample
intervals. The 30 metre grids were set out approximately parallel with the south-east curtain
wall of the fort using a Leica TC 403L EDM, and these grids and other relevant mapping
features were recorded.
The data was processed using Geoplot software (Geoscan Instruments) and the data
presented as a grey scale plot superimposed upon The first edition Ordnance Survey base map
of 1856 (figure 1). This figure shows the relationship of the vicus with a less developed
landscape. One point of note shows that the surveyors of the day indicated the location of the
Vallum as some 500m north of its currently known position, which is probably a point worthy
of further investigation. The magnetic grey scale plot was used to produce an anomaly plan,
which, as is usual, produced a palimpsest of superimposed subterranean responses (figure 3).
A number of hypothetical features, including the fort and bathhouse have been incorporated
on the map, which give both a sense of scale and place the survey in a spatial context. In
order to clarify the very complex central portion of the survey area around the vicus an
additional larger scale anomaly plan has been included (figure 4). Neither can generally be
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
4
depicted as an interpretative plan, because of the complexity of superimposed features.
However, the outline of the individual fields and their associated road and track systems were
plotted as a CAD image, from the primary geophysical survey data (figure 5). Metric units of
length and the field areas (in m²) were determined from this data.
3. Interpretation
A double-ditched major road (1), up to 11m in width, approaches the fort and vicus from the
south-east. This road continues as a single-ditched road towards the junction in the northwest.
The road can be seen as a raised linear feature, approximately 500-600mm in height
above the surrounding ground level. At its south-eastern end the agger is double ditched, with
an intervening strip or pathway some 2.5 to 3.0m in width. A ditched enclosure is sited to the
north of the road (2), which it is suggested could be a cremation cemetery (maximum
dimensions; 64m by 49m; 0.23ha). This interpretation was made based upon the morphology
of the magnetic anomalies, some of which are circular, whilst another resembles a circular
ditch within a 10m square stone surround or kerb. The siting of a cemetery is typical, being
located by the side of a major road leading to a deliberate constriction, adjacent to which a
circular positive anomaly may indicate a roadside well or shrine (3).
This road curves to the north at its junction with a second major road (4) running
approximately north-south and extending beyond the limit of the survey. At the mid-point in
the survey it passes through the vicus where it diverges forming several minor roads. The
magnetic anomalies indicate that the roads to the north of the easterly road may not be
metalled. The road leading south from the Vallum crossing (5), continuing in a straight line to
the outer ditch, is on a different alignment to the major road leading south-west (4) away
from the fort. This latter road is on a direct alignment with the Stanegate fort of Old Church,
Brampton. One of these roads (5) crosses the Vallum (6) and leads to the fort. Several roads
branch from the major road (4) most of which run in south-easterly direction south of the
vicus.
The line of the Vallum (6), together with its crossing, is clearly defined to the southwest
of the survey area, although the angle in the change of direction is not as great as that
previously published (Daniels 1978, 227). The strength of the magnetic response suggests
that the Vallum was left open for a considerable period, perhaps for the greater part of the life
of the fort. These strong positive readings indicate the presence of anthropogenically modified
organic deposits resulting from the gradual infilling of the ditch, possibly creating anaerobic
conditions, and by implication, the potential preservation of organic material. Some evidence
of buildings can be seen within the Vallum enclosure, although the limit of the survey denies
interpretation. Similar buildings have been identified in this position at the fort of Halton
Chesters (Taylor et al, 2000).
Close to the point where the two roads meet, some 40m from the Vallum, is a strong dipolar
anomaly (7); this is thought to be of later origin due to its location and morphology and could
indicate the site of a kiln. This feature is close to the position where the road crosses a
substantial ditch (8), which was intersected at an angle to the south of the Vallum. The ditch
returns to the north-west and north-east where it closely follows the line of the Vallum and
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
5
can be seen to cut several features relating to field boundaries. It is probable that all the roads
crossing this ditch have been cut with the exception of the Vallum crossing. A building,
possibly of stone, which may be associated with a channel or wall, has been built at the
conjunction of three of the ditches to the south-west (9).
The purpose of the ditch outside the Vallum can only be conjectural. Its line however,
does mirror that of the Vallum and it is possibly an earlier cut of the ditch that was later
abandoned. It is significant that the line of the road from the Vallum crossing to the south
deliberately skirts its southern lip, suggesting an earlier date. Alternatively, the ditch may
have enclosed an annex to the fort, as previously suggested by Collingwood (1922, 22); it is
noteworthy that the space between the two ditches does enclose several substantial buildings,
all probably built of stone. A group of four substantial stone buildings (10) can be seen to the
south-west of the road leading from the Vallum crossing. The larger of these buildings is c. 8
m square and is subdivided.
Henry MacLaughlan’s map of the fort, designated by him as Petriana, clearly indicates
that there is a watercourse some 120 m to the east of the fort running north-west to south-east.
The watercourse is shown to cease on the line of the present northern field boundary, and
would appear to be culverted from that point. This watercourse was detected as a slight linear
anomaly by the magnetometry survey, but can be seen to clearly divide the fort and major
settlement to the west from the Romano-British design of field system to the east (15).
Implicitly, it may indicate that it might well be culverted with a stone channel, moreover it
may still be functional, i.e. water-logging and associated anaerobic conditions are not evident
along its route. However, the path of the watercourse can still be observed as a substantial
surface depression, running roughly parallel to the remains of a relict field boundary, which is
also indicated on MacLaughlan’s map as a tree line leading to a permanent spring.
Several ‘lanes’ have been identified, which can be seen to run between the blocks of
property within the settlement to the south of the Vallum, leading to the open land to the rear.
One such ‘lane’ (11) to the east of the settlement passes through a right angle, around a site
boundary, before proceeding in a north easterly direction possibly passing to the south of the
field systems to the east of the watercourse. A lesser road (12), some 20m to the south and
running east crosses the watercourse leading directly to a series of small enclosures to the
south of the field system. A small building of uncertain function can be seen in one of these
fields.
An area free from buildings can be seen to the south of the outer ditch (13), which is
bounded by a ditch to the north and buildings to the other three sides. The main road from the
west enters this open space towards the north-west, with additional roads leading off to the
south and south-east. It is possible that open space was created to form a small market place
or some similar focus for the community. A number of small circular and square (c. 2m)
strongly positive anomalies are contained within this area. It is entirely possible that they will
indicate the location of wells. A narrow linear negative anomaly (possibly of stone) can be
seen to traverse this area from the direction of Castlesteads house and continue towards the
stream bed. This, perhaps arbitrarily, is not indicated as it may be a relatively modern ceramic
pipe, which appears to cut features identified as Roman.
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
6
The road leading to the north-east (4) can clearly be identified as far as the occupation
road running approximately north-west to south-east. Beyond the occupation road line, its
course has been ploughed out and only traces remain of its sinuous route. The lack of
formality of this length of road, in that it does not reflect the perceived Roman preference for
a direct, straight line as seen to the south, could suggest a pre-Roman inception. It is probable
that a branch leading off this road to the south of the occupation road (14) could connect with
the road leading from the east gate, the possible porta praetoria.
Located to the south of the fort is the vicus with roads and buildings set at right angles
to the main thoroughfares. In the northern sector is a field system of more than one phase, set
out to either side of the major road (4) running north-east. Significantly, the alignment of the
former boundaries is similar, implying a change in land ownership or tenure rather than
modifications in crop or husbandry management. These fields continue to the east of the
occupation road, but have largely been ploughed out. The broad impression is that the
secondary field boundaries north of the major road enclose larger plots than during the
primary phase. Almost all of the fields have entrances in a corner; this would imply that they
were all used for animal husbandry at some time, even if it was just for manuring or security.
Understandably, it is easier to herd animals through a field boundary, which is emplaced in the
corner (Pryor 1998, 101, 103, 121). In other words, if the entrance to an enclosure is central,
the assumption is that it is not intended to be used for stock.
Several possible roundhouses can be provisionally identified over the site (16), which
do not appear to have defensive enclosures, but these could be masked by later activity. One
close to the west bank of the watercourse, associated with the primary field system, is cut by a
later boundary. Suggestions of further roundhouses can be seen close to this. Several pits
can be seen over the survey area, although it is possible that those lying adjacent to the
watercourse could be wells (17); the percolation of the water through the ground was thought
to have offered some form of natural filtration (Johnson 1983; Hodge 1995, 71). In total,
there could be as many as 16 wells, most of them giving a magnetic response some 2-3m in
size and are either square or rectangular, although the excavated size could be smaller (see
Johnson 1983, 205, fig. 156). Wells could also be associated with individual buildings or
plots, but in that instance, they could equally be interpreted as refuse or latrine pits.
The boundaries to the settlement to the south-east of the vicus are complex. However,
many of the boundaries appear to be arranged on a similar alignment to those to the east. It is
clear that many of the southerly field boundaries may have been affected by ploughing, and
their southern limit is not now known, although additional survey to the south-east may be
productive in some areas. To the south of the settlement, however, the field boundaries are
aligned to the major road systems (1 & 4). These can be seen to overlie an earlier field system
similar to that found in the northern sector of the settlement. The fields generally are larger
and many contain buildings; some of the larger ones almost certainly used for livestock.
A curved linear feature to the west of the road on the south-west of the survey area
could suggest the line of a ditch to an earlier fort (18). However, if the major road from the
south-east (1) were centred on its length it would suggest a fort with its longest dimension of
c. 220m, which would make it the largest fort on the Wall. This is an area where further
investigation is required in order to validate any interpretation. What should be noted is that
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
7
there is no valid or obvious reason, such as the topography, which dictates that a major road to
the fort should be offset by some 350m.
At the edge of the survey area to the north, evidence can be seen of the Vallum (5) and
a crossing (19), seen as a mix of positive and negative anomalies. This mixture of anomalies
represents both the ditch and the upcast from it. There is some slight evidence, in the form of
a diffuse negative linear area, north of the Vallum crossing of a possible area of metalling,
possibly indicating a road surface.
4. Discussion
This survey has given some insight into the arrangement of the vicus to a fort about which less
is probably known than any other Wall fort. The survey shows that the vicus is set out to the
west of a watercourse and that this feature dictated the eastern extent of most of the
settlement. Romano-British field systems were apparently largely unaffected by building and
could reflect the importance of an existing source of supply to the Roman army. The annona,
an important component of the military diet was almost certainly levied on the frontier. For
example, depopulation on the Eastern Frontier at Dobrudja on the Pontic Shore, meant that
after pacification, the military was forced to draft in farming families from Roman Thrace to
resolve their problem with wheat supply (Williams 1996, 62). The vicus itself is made up of a
loose group of buildings, which has spread out from the Vallum crossing. Its full extent is not
known, as buildings could be present within the woodland to the west of the survey area.
The contrasting character of the fields to either side of the watercourse (15) can clearly
be seen on figure 5, which shows the simplified layout of the field boundaries, road and ditch
systems in bold line, whilst omitting other complicating features. It is significant that the
ditch system connected with the fields, particularly in the east, have a large span and depth
(implied from the strength of the magnetic response). The fact that the ditches follow the
contours indicates that they were not used for drainage purposes alone, and it would seem that
their main purpose would have been to prevent cattle and other animals entering the cultivated
areas or hayfields, or alternatively to protect them. A very similar system was seen within the
field system at Maryport (Biggins and Taylor 2004b; compare figs. 5.2 & 5.6, 105 & 110),
where gentle drainage along the contours probably prevented erosion.
The character of the enclosures to either side of the watercourse differs to quite a
marked degree. Those to the east are made up of a series of small linked enclosures, which
are typical of Romano-British field systems and entrances to these fields from the road are
evident. Those to the west, overlying an earlier phase of much smaller enclosures, are larger
in size. Some of these latter enclosures probably contain buildings, whilst it is suggested that
others could be for the containment of stock. Some indication of a primary field system was
seen in the angle where the road from the southeast joins the road running northeast. The
roads are quite clearly set out either side of the main road entering the fort and are
contemporary with the vicus. It would seem that a field was set out to each side of the roads
entering the fort, being accessible from these roads. The field boundaries running from the
road are set at a right angle to the inner boundary. The fields behind these would appear to
have been accessed from the roads running out from the vicus itself.
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
8
Further examination of this field system to the east of the watercourse shows that the
earlier fields were considerably smaller than the later phase. It is also clear that the minor
road running north-west (14) is later as it cuts earlier field boundaries, whilst it is almost
certain the main north-easterly road (4) continued to be used concurrently. The size of the
fields falls mainly within the approximate areas of 600 -700m². This is about half of an actus
quadratus (c. 1260m²), and although difficult to prove at this remove in time, may have been a
standard allocation, even if they were farmed or managed by an indigenous population. A tidy
Roman mind might have been responsible for basic survey and subsequent allocation. It
should not be forgotten that a governor of Britain, Sextus Julius Frontinus (c. 74-78), who
later wrote about survey, could have had a particular interest in it, and was possibly even a
land commissioner (Dilke 1980, 41; Campbell 2000, xxvii-xxxi). It is also significant that
Frontinus stressed the importance of straight lines and right angles in land measurement,
features which are evident in the southern field boundaries (Campbell 2000, 13).
The earlier field system to the west of the watercourse can be seen to extend under the
vicus and to the southern edge of the survey area. It is probable that it extended up to the
limits of the classified road, although the archaeological evidence has been either destroyed by
ploughing, or in an ideal world covered by colluvium. The difference in character to those to
the east implies either a different use or different tenants, who were possibly members of the
indigenous population. The plot divisions to the east of the road leading south (4) suggest the
character of the later burgage plots of the medieval period and retention of an existing
successful agricultural regime. In many cases, buildings can be seen within their boundaries.
The regularity of many of the field boundaries suggests that the boundaries were set out to a
predetermined layout and are not the result of arbitrary division.
Although it is apparent that the fields have been laid out in a regular pattern, it is
possibly subjective to try to relate the field dimensions to a unit of measure. Some evidence
of the use of the pes Drusianus in military buildings on the Wall and elsewhere has been
identified (Taylor 2000, 41-42). Due to the small difference between the dimensions of a pes
Monetalis and the pes Drusianus (296 mm and 333mm; Duncan Jones 1980, 85-98)
comparison can be subjective and special pleading could be suggested for his hypothesis.
This is due to the possible inaccuracy of any measuring device and the ability of the surveyor
when setting out the boundaries. However, rather than concentrating upon units of length, the
issue of area should perhaps be considered. The area of many of these plots falls within three
approximate ranges; 650m², 1250m² and 2500m² (respectively circa 0.5, 1 and 2 actus
quadratus). The semi-actus quadratus may even have been an accepted sub-division of land
allocations, where good agricultural land was at a premium. It is also significant that many of
the fields fall within the areas 1222m2 and 1276m2, and several of the larger fields are almost
double the size of these stated areas.
It could be noteworthy that the width of the fields A and B measure approximately 150
and 500 pes Drusianus respectively (see figure 4). The size of the plots east of the
watercourse, in many cases, falls between 500 - 600m². One factor, which cannot be
determined, and was proscribed by Roman access law, was the width of access roads or paths
to a neighbour’s property. This law, the Lex Mamilia, attributed to Julius Caesar, guarantees a
width of a minimum of five feet leading to a neighbour’s property (Dilke 1971, 104-5).
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
9
Presumably, this allocation was in force even or especially on frontier settlements and was
removed from the tenant’s or occupier’s land allocation. The problem today is in determining
which plot holder (and hence the field or enclosure) was the responsibility of which
landowner or tenant. It follows that an individual’s land allocation may appear smaller and not
fall within the accepted surveyor’s range, either in terms of length or area. However, it is
generally accepted that each piece of land allows the boundary to stretch to a width of two and
a half feet (Campbell 2000, 23-5).
The late prehistoric period saw an expansion of agricultural settlement into the
uplands, and extensive clearing of woodland has been dated to the first decade of the first
century AD (Woodside and Crow 1999, 30). Evidence of cultivation and field systems can be
seen beneath most of the forts per lineam valli, and Roman field systems have been identified
at Housesteads (ibid., 33). It is apparent that much of the land on the line of the Wall was
being cultivated prior to its construction (Bidwell and Watson 1996, 8-17; Bennett 1998, 19).
The Romans would have found on their arrival well-established arable cultivation almost
certainly growing higher yielding spelt wheat (predominant in the North; van der Veen and
Palmer 1997, 163 -182) and barley (Huntley 1999, 49-64). Both wheat for men and barley for
animals were staples which could be requisitioned from the civilian population under a system
known as frumentum emptum (Rickman 1971 271), not necessarily at disadvantageous terms.
When the forts on the Stanegate frontier, between Carlisle and Corbridge, were built it is
possible that new Romano-British settlements were established attracted by the presence of
the Roman army. Conversely, those settlements displaced by the Roman army, during the
construction of the defensive works associated with the Hadrianic frontier, would have had to
be resettled in a location dictated by the army. A further layer of dependency on the forts
could have been created, with these settlements providing foodstuffs for the vici and the forts.
Examples of forts in a close juxtaposition with field systems are known at Old
Carlisle and on the Antonine wall at Bar Hill and Carriden (Keppie et al 1995, 602-6).
However, some authors propose an alternative view that small fields in association with forts
(e.g. Carriden) may represent some form of allotment by the Roman rather than the
embodiment of an indigenous agricultural system (Armit 2005, 62). Field systems, of
unknown magnitude, have been identified to the west of the fort at Carvoran following an
unpublished geophysical survey by the authors (Breeze 2006, 280-281). Ditched enclosures to
allotments and fields have been identified to the south-east of the fort at Brough-on-Noe
(Dearne 1993, 155).
A major road (1) joins that running south some 275m from the road leading to the
Vallum crossing. This road is wider than any other leading to the fort and bends towards the
River Irthing, where it is likely that an unrecognised bridge was constructed. It is probable
that this road joined the Stanegate, whilst the road running to the south headed towards
Brampton Old Church. The presence of this road from the south-east is surprising as it is not
laid out on the axis of the known or any postulated earlier fort. The breadth of the major roads
entering the settlement is wider and better defined than is known at other Wall forts such as
Birdoswald, Carvoran and Housesteads and the associated site of Maryport (Biggins and
Taylor 2004a, b & c).
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
10
The road crossing the outer ditch from the fort changes direction at the edge of an open
space, which has tentatively been identified as a possible a market place. There are few
apparent buildings in this area to the south of the outer ditch, and its importance is enhanced,
as it is the place where at least four roads converge. The absence of buildings is supported by
the presence of clear undisturbed secondary field boundaries within the space. No market
place is known to any vicus, in Britain, although their existence has been inferred (Sommer
1984, 39-40). A similar open space has been seen on the road leading west from the fort at
Birdoswald (Biggins and Taylor 2004, 176). In this case, the road widens out to form an
elliptical space close to the edge of the settlement.
A speculative model for the pre-Roman and Romano-British occupation of the site at
Castlesteads can tentatively be put forward. Millet (1994, 44) has hypothesised that Rome
dealt with the British tribes by a combination of threats, promises and military action. Their
reward for co-operation with the Roman army could have been independence within
prescribed limits. Seen in the context of the Stanegate frontier, the Romans would have
secured the obedience of the indigenous settlements to the north. These communities may not
have been dispossessed of their land, but have been obliged to supply the Roman army
certainly with food and possibly in the form of conscripted labour or troops (Breeze 1984,
277). Additionally or alternately taxation may have been levied, such an example is that of the
Frisii, who were taxed in the form of commodities such as ox-hides (Tacitus, Histories, V,
25).
The prehistoric settlement at Castlesteads was built in a strategic position and probably
also occupied the low hill to the north-east of the fort. The change in the field patterns,
possibly prior to the building of the Hadrianic fort and settlement, may suggest a change in the
management of the agricultural pattern brought about by the involvement of the Romany
army. It is known that the Roman army was allocated land around fortresses to provide
essential materials. At the Legionary fortress of Xanten in Germany the area the territorium
was not less than 3,400 ha. or 8,500 acres (Petrikovits 1960, 63). Estimates of British
territorium coloniae, for much larger settlements are implied at Colchester (1,100ha), York
(900ha), and Lincoln (900 ha) (Rodwell 1975, 98). Whilst at Chester-le-Street, the only
known example of a frontier auxiliary military territorium is noted (RIB 1049; dated to AD
216), but gives no indication of the size. It has been suggested that a revision of the taxation
system for the benefit of the army was emplaced in the third century or even earlier, the
annona militaris (corn levy), which continued until the end of Roman Britain (Rickman 1971
278-283).
Whether the apparent change of pattern, seen in the creation of smaller fields, but
based upon the original Iron Age model, could have reflected the growing of different crops to
feed the army is a speculative hypothesis. Who controlled, or even farmed the fields close to a
fort is also uncertain. When the decision was taken to construct the Wall, the site chosen for
the fort at Castlesteads in view of its strategic position was founded on what was probably a
Romano-British settlement. The gentle, southern facing slope with fertile soil would have
provided an ideal settlement environment. Whether an earlier Stanegate fort was already in
situ is largely conjectural; this hypothesis is supported by evidence of a major road system
being offset from logical access to a ‘later’ fort. It is possible that a conscious decision was
also taken, by the pragmatic Roman military, in laying out the settlement, to retain part of the
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
11
existing field systems for agricultural use, either by the soldiers themselves or the indigenous
population.
5. Conclusion
The size of the settlement shown is small and there could probably be less than 60 buildings
shown on the magnetometry image. The possibility arises that the settlement may extend
much further to the west and follow the line of the Vallum, outside the limits of the survey.
The present known extent is some 750m along the largest axis of the survey. The Vallum
may not have been left open away from the fort for any extensive period, and further survey to
the north and west of Castlesteads House could clarify this very important point. It can be
strongly argued that at Castlesteads the Roman fort and vicus was founded on a Romano-
British settlement, and that the field system continued to be maintained and modified.
This evidence of the field systems identified on the Wall would tend to discount the
proposal put forward by Sommer (1984, 38) that agricultural activity in the military vici of the
Highland Zone can be decried and only small scale vegetable growing agriculture was
demonstrated. The probability is that extensive field systems will prove to have been
established at all the Wall forts, but this evidence is becoming rapidly degraded by modern
agriculture and building programmes.
The size of the identified vicus is much smaller than those detected elsewhere and
associated with the Wall at Birdoswald, Housesteads, Chesters, Vindolanda, Halton Chesters
and Corbridge, yet the field systems are considerable in extent. It is manifest that the scale of
any field systems and other features around the forts indicated with larger vici will have been
much greater in extent. At the present time there is very little archaeological research being
carried out on the Wall and no agreed inclusive research agenda (November 2006). This is
despite the Wall being a World Heritage Site with tourism being actively encouraged. The
full extent of the archaeological deposits around the forts listed above unknown and it is
almost certain to extend beyond those limits presently assumed.
Acknowledgements
Grateful thanks must be given to the main landowner Major Johnson and the tenant Mr Ian
Douglas together with the adjoining landowner’s Mr and Mrs Sutcliffe for allowing the survey
to take place. Thanks must also be given to Mr Andrew Brown, of Smiths Gore Chartered
Surveyors, for his help with the arranging access. The survey was carried out with the benefit
of generous grants from the Cumberland and Westmorland Antiquarian and Archaeological
Society and the Royal Archaeological Institute. Thanks are also due to June Biggins, Bob
Coxon, Alex Greene, Ben Johnson, Rupert Lotherington, Janet Mears, Claire Nesbitt and Julia
Robinson who assisted with the survey.
JAB would like to dedicate this report to the theatre and ward staff at Ward 36, RVI,
Newcastle, and in particular, Mr D. Karat, consultant surgeon, without their skilled surgery
and care JAB would not have survived to contribute to this paper.
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
12
Bibliography
Armit, I., 2005, Celtic Scotland, Batsford, London
Bennett, J., 1998, “The Roman Frontier from Wallsend to Rudchester Burn Reviewed”,
Archaeologia Aeliana, ser. 5, 26, 17-37
Bidwell, P. T. and Watson, M., 1996, “Excavations on Hadrian’s Wall at Denton, Newcastle
upon Tyne, 1986-89”, Archaeologia Aeliana, ser. 5, 24, 1-56
Biggins, J. A. and Taylor, D. J. A.. 2004a, “Geophysical Survey of the vicus at Birdoswald
Roman fort, Cumbria”, Britannia 35, 159-178.
Biggins, J. A. and Taylor, D. J .A., 2004b, “The Roman fort and vicus at Maryport:
geophysical survey, 2000-2004”, in R. J. A. Wilson and I. Caruana (eds.), Romans on
the Solway, Maryport.
Biggins, J. A. and Taylor, D. J. A., 2004c, “A geophysical survey of Housesteads
Roman fort”, Archaeologia Aeliana, 23, 51-59.
Breeze, D, 1984, ‘Demand and Supply on the Roman Frontier’, in R. Miket and C.
Burgess (eds.), Between and Beyond the Walls: Essays on the History and Prehistory of
North Britain in Honour of George Jobey, John Donald Publishers Ltd. Edinburgh, 264-
286.
Breeze, D J, 2000, Hadrian’s Wall, Harmondsworth
Campbell, B., 2000, The Writings of the Roman Land Surveyors, Society for the Promotion of
Roman Studies, Journal of Roman Studies Monograph no. 9, London
Collingwood, R. G., 1922, “Castlesteads”, CW2, 22, 198-233
Daniels, C. M., 1978, J. Collingwood-Bruce’s Handbook to
the Roman Wall,(13th ed..) Newcastle upon Tyne.
Dearne, M. J., 1993, Navio – the fort and vicus at Brough-on-Noe, Derbyshire, BAR Brit. Ser.
234, Oxford
Dilke, O.A.W., 1971, The Roman Land Surveyors: An Introduction to the Agrimensores,
David and Charles, London
Duncan-Jones, R. P., 1980, “Length-Units in Roman Town Planning: The Pes Monetalis and
the Pes Drusianus”, Britannia, 11, 127-133.
Heslop, D., 1987, The excavation of an Iron Age settlement at Thorpe Thewles,
Cleveland, 1980-2, CBA Excavation Report 65
Hodge, T. A., 1991, Roman Aqueducts and water Supply, London
Huntley, J. P., 1999, “Environmental Evidence from Hadrian’s Wall”, in P. Bidwell (ed.),
Hadrian’s Wall 1989-1999, 49-64
Hutchinson, W., 1794, The History of the County of Cumberland, Carlisle
Keppie, L. J. F., Bailey, G. B., Dunwell, A. J., McBrien, J. H. and Speller, K., 1995 “Some
excavations on the line of the Antonine Wall 1985-93”, PSAS, 125
Johnson, A., 1983, Roman Forts, London
Mann, J.C., 2002, “The Settlement of Veterans Discharged from Auxiliary Units
Stationed in Britain”, Britannia, 33, 183 – 188
Millett, M., 1994, The Romanization of Britain, Cambridge
Petrikovits, H. von, 1960, Das Römishe Rheinland: Archäeologishe Forshungen Seit
1945, Köln.
Pryor, F., 1998, Farmers in Prehistoric Britain, Tempus, Stroud.
Richmond, I. A. and Hodgson, K, S, 1934, “Excavations at Castlesteads”, CW2, 34, 159-165
Geophysical Survey of the Vicus & Vallum at Castlesteads Roman Fort, Brampton, Cumbria
13
Rickman, G., 1971, Roman Granaries and Store Buildings. Cambridge.
Rodwell, W., 1975, Milestones, Civic Territories and the Antonine Itinerary, Britannia 6, 76 -
101.
Sommer, S., 1984, The Military Vici in Roman Britain, BAR Brit. ser. 129, Oxford
Tacitus, Histories.
Taylor, D. J. A., 2000, The Forts on Hadrian’s Wall, BAR Brit. ser. 305, Oxford
Taylor, D. J.A., Robinson, J, and Biggins, J. A., 2000, “A Report on a Geophysical Survey of
the Fort and Vicus at Halton Chesters”, Archaeologia Aeliana, ser. 5, 28, 35-46
van der Veen, M. and Palmer, C., 1997, “Environmental Factors and the Yield
Potential of Ancient Wheat Crops”, Journal of Archaeological Science, 24, 163 -182
Williams, D., 1996, The Reach of Rome, St. Martin’s Press, New York.
Wilmott, T., 1997, Birdoswald, English Heritage Archaeological Report 14, London
Woodside, R, and Crow, J., 1999, Hadrian’s Wall, An Historic Landscape, National Trust
List of Figures [NOTE: Problems with the PDF formatting prevented me from getting the last Figure to post, and the text between the Figures also had to be sacrificed. For the missing images and information, please see the Academia.edu page listed at the head of this article. My apologies for any inconvenience.]
