Time Team - Brancaster Roman Fort

making sense of heritage
Brancaster Roman Fort, (Branodunum),

Norfolk
Archaeological Evaluation and Assessment of Results
Ref: 85209.01
September 2014
BRANCASTER ROMAN FORT

(BRANODUNUM), NORFOLK
Prepared for:
Videotext Communications Ltd

11 St Andrews Crescent
CARDIFF
CF10 3DB
by
Wessex Archaeology
Portway House
Old Sarum Park
SALISBURY
Wiltshire
SP4 6EB
Report reference: 82509.01
Scheduled Ancient Monument number (old county style) NF 208

(The National Heritage List for England entry 1003983)
September 2014
Wessex Archaeology Limited 2014 all rights reserved

Registered Charity in England and Wales, No. 287786. In Scotland, Scottish Charity No. SC042630
Brancaster Roman Fort, Norfolk

DISCLAIMER
THE MATERIAL CONTAINED IN THIS REPORT WAS DESIGNED AS AN INTEGRAL PART OF A REPORT TO AN INDIVIDUAL
CLIENT AND WAS PREPARED SOLELY FOR THE BENEFIT OF THAT CLIENT. THE MATERIAL CONTAINED IN THIS REPORT
DOES NOT NECESSARILY STAND ON ITS OWN AND IS NOT INTENDED TO NOR SHOULD IT BE RELIED UPON BY ANY
THIRD PARTY. TO THE FULLEST EXTENT PERMITTED BY LAW WESSEX ARCHAEOLOGY WILL NOT BE LIABLE BY
REASON OF BREACH OF CONTRACT NEGLIGENCE OR OTHERWISE FOR ANY LOSS OR DAMAGE (WHETHER DIRECT
INDIRECT OR CONSEQUENTIAL) OCCASIONED TO ANY PERSON ACTING OR OMITTING TO ACT OR REFRAINING FROM
ACTING IN RELIANCE UPON THE MATERIAL CONTAINED IN THIS REPORT ARISING FROM OR CONNECTED WITH ANY
ERROR OR OMISSION IN THE MATERIAL CONTAINED IN THE REPORT. LOSS OR DAMAGE AS REFERRED TO ABOVE
SHALL BE DEEMED TO INCLUDE, BUT IS NOT LIMITED TO, ANY LOSS OF PROFITS OR ANTICIPATED PROFITS DAMAGE
TO REPUTATION OR GOODWILL LOSS OF BUSINESS OR ANTICIPATED BUSINESS DAMAGES COSTS EXPENSES
INCURRED OR PAYABLE TO ANY THIRD PARTY (IN ALL CASES WHETHER DIRECT INDIRECT OR CONSEQUENTIAL) OR
ANY OTHER DIRECT INDIRECT OR CONSEQUENTIAL LOSS OR DAMAGE
QUALITY ASSURANCE
85209
-
SITE CODE
PLANNING APPLICATION REF.
VERSION
STATUS*
PREPARED
BY
NB
APPROVED
BY
ACCESSION CODE
2012.240
CLIENT CODE
578209, 344020
NGR
APPROVERS
SIGNATURE
LNM
DATE
FILE
02/09/14
X:\PROJECTS\85209\POST
EX\REPORT\85209_BRANCASTER ROMAN
FORT_REPORT V1.DOC
* I= Internal Draft E= External Draft F= Final
WA Project No. 85209
ii



Contents
Summary ............................................................................................................ vi
Acknowledgements ............................................................................................ vii
1
INTRODUCTION .................................................................................................1
1.1 Project Background .....................................................................................1
1.2 The Site, location and geology ....................................................................1
1.3 Archaeological Background and Previous Archaeological Work .................2
AIMS AND OBJECTIVES ...................................................................................3
METHODOLOGY ................................................................................................3
3.1 Geophysical Survey.....................................................................................3
3.2 Evaluation Trenches ....................................................................................4
3.3 Copyright .....................................................................................................4
RESULTS ............................................................................................................5
4.1 Introduction ..................................................................................................5
4.2 Geophysical Results ....................................................................................5
4.3 Conclusions .................................................................................................8
4.4 Evaluation Trenches ....................................................................................8
FINDS ................................................................................................................14
5.1 Introduction ................................................................................................14
5.2 Pottery .......................................................................................................15
5.3 Ceramic Building Material..........................................................................20
5.4 Mortar, opus signinum and wall plaster .....................................................21
5.5 Stone and Worked Flint .............................................................................22
5.6 Glass .........................................................................................................22
5.7 Metalworking debris...................................................................................22
5.8 Coins .........................................................................................................22
5.9 Metalwork ..................................................................................................24
5.10 Worked Bone .............................................................................................26
5.11 Human Bone..............................................................................................26
5.12 Animal Bone ..............................................................................................27
5.13 Marine Shell...............................................................................................29
PALAEO-ENVIRONMENTAL SUMMARY ........................................................30

6.1 Introduction ................................................................................................30
6.2 Charred and mineralised plant remains.....................................................31
6.3 Wood Charcoal ..........................................................................................32
6.4 Land and aquatic molluscs and marine shells ...........................................32
6.5 Small animal and fish bones......................................................................33
6.6 Foraminifera ..............................................................................................33
DISCUSSION.....................................................................................................33
7.1 Introduction ................................................................................................33
7.2 Evidence for an earlier fort (Trench 5 and Trench 3).................................34
7.3 The eastern vicus (Trench 3).....................................................................34
7.4 The main fort .............................................................................................35
iii

POTENTIAL AND FURTHER RECOMMENDATIONS .....................................37

8.1 Potential.....................................................................................................37
8.2 Proposals...................................................................................................39
ARCHIVE ...........................................................................................................39
10
REFERENCES ..................................................................................................41
10.1 Bibliography ...............................................................................................41
10.2 Online resources .......................................................................................46
APPENDIX 1: TRENCH SUMMARIES.......................................................................47

APPENDIX 2: SUPPLEMENTARY FINDS INFORMATION ......................................54
Table 1: Finds totals by material type and by trench (number / weight in
grammes) ..................................................................................................54
Table 2: Pottery assemblage by trench .............................................................55
Table 3: Pottery quantification by fabric type .....................................................56
Table 4: Pottery fabric concordance for greywares ...........................................56
Table 5: Pottery vessel forms by fabric ..............................................................58
Table 6: Trench 1 pottery fabrics .......................................................................59
Table 7: Trench 1 pottery forms by fabric ..........................................................60
Table 8: Trench 2 pottery fabrics .......................................................................61
Table 9: Trench 2 pottery forms by fabric ..........................................................61
Table 10: Trench 3 pottery fabrics .....................................................................62
Table 11: Trench 3 pottery forms by fabric ........................................................62
Table 12: Trench 4 pottery forms by fabric ........................................................63
Table 15: Quantification of retained CBM by type and by context (fragment
count) ........................................................................................................65
Table 16: Coin list ..............................................................................................66
Graph 1: All coins from the site ..........................................................................68
Graph 2: Probable hoard from layer 125 ...........................................................69
Graph 3: Coins from Brancaster (without the coins from layer 125) ..................69
Table 17: Marine shell by context ......................................................................70
Table 18: Oyster shell analysed in more detail ..................................................71
Table 19: Comparative size of analysed oyster shell.........................................72
Graphs 4-8 showing size of analysed oyster shells by context..........................72
APPENDIX 3: PALAEOENVIRONMENTAL RESULTS ............................................75
Table 20: Assessment of the charred plant remains and charcoal ....................75
Table 21: Land and aquatic molluscs and marine shell assessment .................76
iv

Figures
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Location of Site, trenches and geophysical survey areas including

features identified from aerial photography
Magnetometer survey results
GPR survey results
Trench 1, survey and sections
Section 1: East-facing section of sondage at north end of trench and
possible surface 134
Section 2: West-facing section of wall 128, robber cut 127 and
adjacent deposits
Trench 1, plates
Plate 1 : Southern end of Trench 1, view from the south-east
Plate 2 : Central area of Trench 1, view from the south
Plate 3 : West-facing section adjacent to wall 130, oblique view from
the south-west
Plate 4: Lorica squamata in situ adjacent to wall, fragment after
cleaning, x-ray plate of fragments
Plate 5:
Northern end of Trench 1, view from the north
Trench 2 survey, section and plates
Section 3: East-facing section features 208, 223 and wall 205
Plate 6: Northern end of Trench 2, view from the north
Plate 7: East-facing section features 208, 223 and wall 205, oblique
view from the north-east
Plate 8:
Southern end of Trench 2, view from the south
Trench 3, survey, section and plates
Section 4: West-facing section through ditches 313 and 305
Plate 9: Trench 3, view from the east
Plate 10: East facing section of 312
Section 5: East-facing section of wall 403 and associated rampart,
robber cut 408 and ditch 406
Plate 12: Southern end of Trench 4, view from the south-west
Section 6: West-facing section through ditches 503 and 506
Plate 13: Trench 5, view from the south-west
Postulated layout of the Brancaster complex, based on excavation,
geophysical survey and cropmark evidence
Front Cover Working shot, Trench 1

Back Cover (clockwise from top left) Working shot Trench 2; Investigating wall 205;
Working shot Trench 1; Working shot Trench 3


Summary
Wessex Archaeology was commissioned by Videotext Communications Ltd to

undertake a programme of archaeological recording and post-excavation work on an
archaeological evaluation undertaken by Channel 4s Time Team at Brancaster
Roman Fort (Branudonum), Brancaster, Norfolk (NGR 578209, 344020). An
evaluation consisting of five trenches, magnetometer survey and Ground Penetrating
Radar (GPR) survey sought to characterise both the chronology and character of the
main fort and also the eastern vicus. A trench was also positioned on a possible
earlier fort identified from cropmark evidence to the north of the main fort. The
fieldwork was carried out between 710 August 2012.
Trenches 1, 2 and 4 lay within the fort. These indicated activity in the fort from the
2nd into the 4th century AD, with some evidence for robbing, abandonment and
possible later activity in the 4th century. The geophysical survey, particularly the GPR
survey, enabled the layout of the fort and a number of structures within in it to be
identified, and indicated the generally good preservation of the below ground
remains.
Trench 3 was situated within the eastern vicus. This revealed considerable truncation
by modern ploughing. A number of features were identified, dating to the 2nd and 3rd
century AD, overlain by a metalled road surface of 3rd or 4th century AD date.
Trench 5 was positioned to the north of the fort, targeting double ditched features
identified by aerial photography. An earlier date for this structure could not be
determined as the Romano-British pottery recovered from these ditches could not be
tightly dated, though there was some evidence that the features may have been
deliberately backfilled.
Despite limited size, these investigations have considerably augmented existing
knowledge of this nationally significant monument and have the potential to
contribute to our understanding of the Saxon Shore network. Only limited further
analysis is proposed, but a short summary article of the results of the evaluation,
incorporating the analytical results, will be prepared for submission to the Norfolk
Archaeological Journal.
vi


Acknowledgements
This programme of post-excavation and assessment work was commissioned and
funded by Videotext Communications Ltd, and Wessex Archaeology would like to
thank the staff at Videotext, and in particular Sin Price (Series Editor), Chris
Rushton (Director), Val Croft (Head of Production), Dan Wheatley (Production Coordinator) and Maddy Gerry (Researcher) for their considerable help during the
recording and post-excavation work.
The geophysical survey was undertaken by John Gater, Jimmy Adcock, Emma
Wood, Graeme Attwood and Rachel Brown. The excavation strategy was devised by
Francis Pryor. The on-site recording was co-ordinated by Naomi Brennan, and onsite finds processing was carried out by Matt Kendall, both of Wessex Archaeology.
The excavations were undertaken by Time Teams retained archaeologists, Tracey
Smith, Phil Harding, Rob Hedge, Ian Powlesland, Matt Williams, Raksha Dave and
Cassie Newland assisted by Rob Brown, Sarah Leppard, Simon Greenslade,
Charlotte Mecklenburgh, Tom Jamieson and John Ames. The metal detector survey
was carried out by Kevin Elfeet and Mark Nicolson.
The archive was collated and all post-excavation assessment and analysis
undertaken by Wessex Archaeology. This report was written and compiled by Naomi
Brennan with specialist reports prepared by Rob Perrin (pottery), Kayt Marter Brown
(CBM), Nicholas Cooke (coins), Jacqueline McKinley (human bone), Lorrain Higbee
(animal bone), Lorraine Mepham (all other finds), Kevin Hayward (geological
identifications), Chris J. Stevens (environmental charred and mineralised material)
and Sarah Wyles (Marine shell; environmental molluscs). The illustrations were
prepared by Rob Goller. The post-excavation project was managed on behalf of
Wessex Archaeology by Lorraine Mepham.
Wessex Archaeology would like to thank Will Fletcher (English Heritage), Angus
Wainright (National Trust) and David Gurney (Norfolk County Council) for their advice
and input during the evaluation process. We would also like to thank Philippa Walton,
Mark Corney, Alice Lyons and Naomi Sewpaul for their specialist input during the
course of the investigation. Finally thanks are extended to the owners, Jeremy
Thompson and the National Trust, for allowing access to the Site for geophysical
survey and archaeological evaluation.
vii


INTRODUCTION
1.1
Project Background
1.1.1
Wessex Archaeology was commissioned by Videotext Communications Ltd

to undertake a programme of archaeological recording and post-excavation
work on an archaeological evaluation undertaken by Channel 4s Time
Team at the site of Brancaster Roman Fort (Branodunum), Brancaster,
Norfolk, National Grid Reference (NGR) 578209, 344020 (hereafter the
Site) (Figure 1).
1.1.2
This report documents the results of archaeological survey and evaluation

undertaken by Time Team, and presents an assessment of the results of
these works.
1.2
The Site, location and geology
1.2.1
The Site lies within the parish of Brancaster, some 11km to the north-east of
Hunstanton and 5km to the north-west of Burnham Market, and is situated
less than 2.5km from the present edge of the Norfolk coast. The saltmarsh,
sand dunes and associated littoral zone just to the north of the Site forms
part of the North Norfolk Coast Site of Special Scientific Interest (SSSI)
(citation number 1001342).
1.2.2
Brancaster Roman Fort is a Scheduled Ancient Monument (old county

number NF 208, The National Heritage List for England entry 1003983). The
Scheduled Area is around 27ha in extent and covers not only the field where
the fort is known to situated but the fields to the north and east of this, as
well as the northern portion of the fields to the south of A149. This was in
order to encompass the spread of the vicus and other associated features.
The housing estate to the west, though forming part of the vicus, is not
included as this was subject to excavation in the 1970s before development.
1.2.3
The area considered as part of this evaluation consisted of the main field
where the fort lies (under the ownership of the National Trust), the field
directly to the north of this and the field immediately to the east (under
private ownership).
1.2.4
The fort area is still visible as a distinct raised plateau at a height of

approximately 15m aOD. On each side the ground slopes steeply down and
into the hollows, indicating the surrounding defensive ditch. The northern
field is slightly lower at a height of 7-9m aOD and slopes gently towards the
coast. The eastern field is generally flat and lies at around 13m aOD. All the
areas are currently under grass, though the eastern field has been ploughed
within the last few years. Though the fort area has been ploughed during the
19th and 20th centuries, in 1984 the National Trust acquired the site and
were therefore able to remove it from cultivation and further damage (Flack
and Gregory 1988, 164).

1.2.5
The bedrock is listed as White Chalk Subgroup while the superficial geology
is the Ringstead Sand and Gravel Member; a spur of Head (clay, silt, sand
and gravel) crosses the eastern field to a point just by the bend in Green
Common Lane to the south (BGS 1:50,000 mapping).
1.3
Archaeological Background and Previous Archaeological Work
1.3.1
Brancaster forms one of a line of Roman forts that was constructed on the
south and east coasts of England, generally known as the Saxon Shore.
The Notitia Dignitatum, thought to reflect the situation in the western empire
in the late 4th century AD, lists nine forts under the command of the comes
litoris Saxonici (Count of the Saxon Shore), although the physical remains of
ten or possibly twelve remain (Cunliffe 1977, 1). Brancaster has long been
identified from the list of these installations given in the Notitia Dignitatum
with Branodunum, where a regiment of Equites Dalmat, or Dalmatian
Horse was garrisoned. Its form, with rounded corners, internal banks and no
bastions, suggests a 2nd or early 3rd century date, making it one of the
earlier forts to be constructed and probably of a similar date to Relculver
(Regulbium) in Kent (ibid., 3). Though the Dalmatian cavalry are unlikely to
have been stationed here before the late 3rd century, there is some
evidence to suggest that Cohors I Aquitanorum, a Gaulish infantry regiment,
was originally stationed at the fort (Hassall 1977, 9). The position of a fort at
Brancaster and its likely date suggest that it was constructed to guard the
approach of the Wash from pirate attacks (Cunliffe 1968, 261), though more
recent interpretation has seen the Saxon Shore forts supporting inland
garrisons and facilitating the movement of people and resources through the
territory and the wider empire (Bidwell 1997, 42-43).
1.3.2
In 1846 the Reverend Lee Warner examined Brancaster with the hope of
elucidating the plan of the fort. His investigations were successful in locating
remnants of the masonry wall at the north-east corner as well as locating the
base of corner tower, apparently contemporaneous with the main wall. He
also notes the removal of stonework from the foundation walls some 50
years previously in order to prepare ground for ploughing and to provide
stone for the construction of a nearby barn (Lee Warner 1851, 12).
1.3.3
The first systematic excavations of the fort were in 1935, when a number of
trenches were excavated within the western part of the fort on the north,
south and western defensives (St. Joseph 1936). This was able to establish
confidently the size and shape of the fort as well as to locate sections of the
wall, ditch and rampart. The rampart was found to be internal and
contemporaneous with the wall which it directly abutted. In several places
the wall was found to have been complete removed and in places where it
was present, much of the facing stone had been removed. Other trenches
explored the north-west corner and the west entrance, confirming the
presence of a corner tower as well as the west gateway and road, though
here the remains had been heavily disturbed by ploughing. One trench was
extended to explore the interior of the fort and here found two phases of
structures and associated occupation separated by a layer of refuse and
debris. The later structure was of fairly crude construction; finds from this
layer suggest it is late 4th century AD.

1.3.4
Some brief unpublished notes suggest the presence of Roman structures in

a field to the west of Straithe House (Rotham 1960), potentially in the
eastern vicus settlement.
1.3.5
In 1974 and 1977 excavations took place to the west of the fort to record the
western vicus settlement prior to the construction of a new housing estate
(Hinchliffe and Sparey Green 1985). This encountered a certain amount of
truncation due to ploughing which was thought to have removed some of the
structural remains, but a series of ditches and pits was located. The activity
would seem to indicate settlement in the late 2nd and through the 3rd
century AD, with some 4th century activity. The alignment of the settlement
differs from the fort and it has been suggested that the vicus pre-dates the
fort; the establishment of such a settlement, therefore, within the rural
hinterland may indicate an earlier military presence.
1.3.6
In 1985 three trenches were excavated in the western part of the fort, two
across the defensive ditch and one at right angles along its edge (Flack and
Gregory 1988). Though the trenches were not able to excavate a full profile
across the ditch, in both cases they located a gully at the base of the
western (outer) edge. It was hoped that the trench at right angles to the ditch
might locate remains of the road, but nothing was found to have survived.
Intriguingly the ditch does not appear to be directly parallel to the fort and
runs directly in front of the west gate, posing questions about the connection
between the fort and the western vicus.
1.3.7
As well as the evidence for Romano-British activity, Neolithic and Mesoltihc

flints were discovered during the 1985 excavations (Flack and Gregory
1988, 169).
AIMS AND OBJECTIVES
2.1.1
A project design for the work was compiled (Videotext Communications

2012), providing full details of the research aims and methods. A brief
summary is provided here.
2.1.2
The aim of the project was to characterise the nature and date of the Site
and place it within its historical, geographical and archaeological context.
Three research aims were outlined in the project design:
x
Research Aim 1: What is the character of the archaeology

represented by cropmarks at both the main fort site and the eastern
vicus? Do the differing alignments suggest the presence of an earlier
fort?
Research Aim 2: What is the chronological sequence of fort

construction at Brancaster? Is there a 2nd century AD fort at the site?
Research Aim 3: Does any evidence survive for shore-side

development at Brancaster?
METHODOLOGY
3.1
Geophysical Survey
3.1.1
Prior to the excavation of evaluation trenches, a geophysical survey was

carried out across the Site using a combination of resistance and magnetic

survey. The survey grid was tied in to the Ordnance Survey grid using a
Trimble real time differential GPS system.
3.2
Evaluation Trenches
3.2.1
Five trenches of varying sizes were excavated, their locations determined in

order to investigate and to clarify geophysical anomalies and address
specific research objectives (Figure 1).
3.2.2
The trenches were excavated using a combination of machine and hand

digging. All machine trenches were excavated under constant
archaeological supervision and the machine was only used to remove
modern topsoil or ploughsoil. When machine excavation had ceased all
trenches were cleaned by hand and archaeological deposits investigated.
3.2.3
At various stages during excavation the deposits were scanned by a metal

detector and signals marked in order to facilitate investigation. The
excavated up-cast was scanned by metal detector.
3.2.4
All archaeological deposits were recorded using Wessex Archaeologys pro

forma record sheets with a unique numbering system for individual contexts.
Trenches were located using a Trimble Real Time Differential GPS survey
system. All archaeological features and deposits were planned at a scale of
1:20 with sections drawn at 1:10. All principal strata and features were
related to the Ordnance Survey datum.
3.2.5
A full photographic record of the investigations and individual features was

maintained, using digital images. The photographic record illustrated both
the detail and general context of the archaeology revealed and the Site as a
whole. Digital images have been subjected to a managed quality control and
curation process which has embedded appropriate metadata within the
image and ensures the long term accessibility of the image set.
3.2.6
At the completion of the work, all trenches were reinstated using the
excavated soil.
3.2.7
A unique Site code 85209 was agreed prior to the commencement of works.
The work was carried out between 710 August 2012. The archive and all
artefacts were subsequently transported to the offices of Wessex
Archaeology in Salisbury where they were processed and assessed for this
report.
3.3
Copyright
3.3.1
This report may contain material that is non-Wessex Archaeology copyright

(e.g. Ordnance Survey, British Geological Survey, Crown Copyright), or the
intellectual property of third parties, which we are able to provide for limited
reproduction under the terms of our own copyright licences, but for which
copyright itself is non-transferrable by Wessex Archaeology. You are
reminded that you remain bound by the conditions of the Copyright, Designs
and Patents Act 1988 with regard to multiple copying and electronic
dissemination of the report.

RESULTS
4.1
Introduction
4.1.1
Details of individual excavated contexts and features, the full geophysical

report (GSB 2014), the summary of the landscape and earthwork survey and
details of artefactual and environmental assessments, are retained in the
archive. Summaries of the excavated sequences can be found in Appendix
1.
4.2
Geophysical Results
4.2.1
Geophysical survey was carried out over a total area of 84.5 hectares using
magnetometer survey with a 0.8 hectare area within the fort interior
subjected to detailed Ground Penetrating Radar (GPR) survey. Conditions
for the survey were good. It should be noted that depths referred to in the
GPR data are only an approximation. The following discussion and
accompanying data is taken from the report compiled by GSB (2014).
4.2.2
4.2.3
Magnetometer survey (Figure 2)

Earlier unpublished work at Brancaster by English Heritage in 1973 and
1975 (Bartlett 1973; Bartlett 1975) provided a glimpse of the potential that
could be achieved by carrying out magnetic survey at the site. Ditches, pits
and field systems, following a rectilinear pattern, were detected immediately
to the west of the fort, on the site of the present housing estate. It is perhaps
surprising. therefore, that prior to this investigation, no further geophysical
survey had been carried out, apart from an evaluation project to the east of
the vicus (GSB 2006).
The fort
The main elements of the fort are clearly visible in the magnetic data. The
defences, comprising banks and ditches, stand out in the east and west but
unfortunately are obscured by modern field boundaries in the north and
south. Gateways [1, 2, 3 & 4] are visible at the cardinal points, some more
clearly than others, and the courses of the internal roads are discernible but
only the east-west route is conspicuous. Barrack blocks are apparent in the
south-west quadrant but much more clearly in the south-east [5] where the
lines of buildings are visible, with negative magnetic responses
corresponding to the wall foundations. Similarly many of the rooms and the
courtyard which make up the principia [6] show as negative anomalies (due
to the lack of magnetic material compared to the immediate surroundings). It
is interesting to note the strong positive anomalies which presumably relate
to magnetic deposits which have built up inside the individual rooms; in
some cases these are likely to represent the sources of heat for the
hypocaust system. The results correlate extremely well with the radar
findings [Figure 3, B and D]. In the northern half of the fort the complexity of
the magnetic responses is even greater, but buildings visible in the radar are
not as clear magnetically. For example, the granary buildings walls are at
best poorly defined but more often absent in the data; this is likely to be a
consequence of the lack of magnetically enhanced deposits which are
present in and around the barracks and principia. This interpretation could
explain why many of the small buildings, including probable workshops,
elsewhere in the fort are visible. The magnetic data from the three rooms at
[7] match extremely well with the radar [Figure 3, J & I]; the clarity is such
that the magnetic data even indicate the line of the flue into the hypocaust

room [I]. Other anomalies which have been highlighted are those at [8 & 9]
which could be large pits or areas of intensive burning and the responses
[10] which correspond with the enigmatic radar results [Figure 3, G]. A line
of four ferrous-like anomalies [11] is perplexing; it is uncertain whether they
relate to the fort or to much more recent features.
4.2.4
4.2.5
4.2.6
4.2.7
The vicus
An area in the field to the east of the fort shows a complex of responses very
similar to the earlier English Heritage surveys. The line of the east-west road
which runs through the fort is apparent, although it veers slightly southwards
and there appear to be later features cutting through. The rectilinear pattern
of ditches and presumed tracks indicates a formal layout to the majority of
the vicus mapped by the survey. Overlapping and intercutting anomalies
[e.g. 12] suggest multi-phased activity. The data suggest a lack of pits
throughout which is perhaps surprising given the nature of the settlement.
The general results concur well with aerial photographs; clearly the vicus
extends well beyond the area which was surveyed.
Area to the north
The density and complexity of the anomalies in this area is much less than
compared with the vicus field. In the south-western extension of Area 1 there
is a similar pattern of anomalies [13] as in Area 3 and in the English Heritage
survey in the housing estate to the south. The diminished magnetic
response may be due to a phenomena referred to as a habitation effect
(Gaffney and Gater 2003) whereby the strength of magnetic responses
decreases away from the core of activity. However, in this instance it is
possible that a post-Roman deposition of alluvium is resulting in a weaker
anomaly strength. In the northern half of the area there appears to be a
separate double-ditched enclosure [14] with internal divisions which seems
to have no direct association with the fort, apart from the fact it follows a very
similar alignment. At [15] there is an unusual curving response which is
difficult to interpret. In this context it could be of archaeological interest but
the nature of the anomaly suggests a natural (alluvial) origin is perhaps
more likely; hence the uncertain interpretation category.
GPR survey (Figure 3)
After the success of the magnetic survey, there were high hopes for the
results of the GPR survey. Despite only covering a relatively small
percentage of the whole site, they exceeded expectations revealing a wealth
of detail indicating numerous structures, construction details and multiphased elements of the fort. As the MIRA system and processing software
were on loan for the Time Team project, there was only a limited amount of
processing and interpretation that could be carried out to produce this report.
It therefore gives just a basic overview of what was found, highlighting some
key features of note; it should provide an excellent basis for any further
investigations in the future.
The survey area was chosen based on crop-marks which appeared to show
the principia. This building dominates the southern third of the data, with the
walls and large rooms clearly visible and spanning a total of approximately
51m by 38m. Other features include: a possible monumental feature [A]
within the central 20m-sided courtyard, plus two more similar features
immediately to the north-east and north-west [D]; a grid of very small
reflectors [B], presumably the pilae within an extensive hypocaust system;

potential structures [C] immediately outside the building complex; and one
room (approximately 8m by 5m) with significantly greater depth extent
(0.45m 2.2+m) than the others (which, at the northern end of the principia,
peter out at around 1.5m below ground level).
4.2.8
North of the principia is a large range of buildings one of which [E] has the
hallmarks of a large granary, 7m by at least 21m, with a central division or
drain, floor pillars for air circulation and buttressed walls. On the north side
of this, a large rectangular space [F] is around 18m north-south and could be
as much as 40m east-west based on the magnetic data. What is unusual
about this structure is the oval response that contracts towards the centre of
the space with increasing depth, to reveal a smaller inner rectangle [G]
approximately 10m by at least 18m and which extends down beyond 2.2m
below ground level. The oval shape could be the effect of tip-lines within
demolition material filling this feature rather than a response to a physical
structure. There is also a strong magnetic response [Figure 2, 10]
coincident with the inner rectangle.
4.2.9
Further north again, beyond a range of buildings adjoining [F], is another

large structure [H] with no obvious internal divisions although this does not
preclude their existence. It has a small porch-like extension on the north side
but it is difficult to tell whether the structure is cruciform as the south side is
masked somewhat by an area of increased response, perhaps demolition
material.
4.2.10
On the eastern side of the survey area is a three-cell building although

whether they are all contemporaneous is arguable. The southern-most room
has a cross-flue hypocaust system [I] which, with depth, fades to reveal a
semi-circular feature beneath it (from around 1.5m below ground level) that
seems to extend off the southern side of the central room [J] and has a starshaped response at its centre.
4.2.11
South of the three-celled structure is a large building [K] containing a

number of rooms, three of which [L, M, N] appear to have intact floor
surfaces at around 0.3m, 0.4m and 0.7m below ground level, respectively.
This building is interesting as it lies on a slightly altered alignment from the
majority of the other structures; most are aligned with the principia and thus
the Saxon Shore Fort phase of defences, whereas [K] is in the same
orientation as the vicus to the east. One explanation could be that this is a
relatively early mansio that was built on the line of a pre-existing Roman
road running through the original 2nd century fort, which was then
subsumed within the larger fort construction, whose focus in the landscape,
and thus orientation, was somewhat different.
4.2.12
Down the centre of the survey block runs one of the principal thoroughfares,
leading to the principia and along this can be seen a series of narrow linear
anomalies [e.g. O], presumably drains. Some of these can be seen to
branch off [P] towards the buildings down each side.
4.2.13
There are numerous other linear anomalies and zones of response that are
undoubtedly further structural elements but which are less clear in the data,
due to variation in either preservation, overburden or construction, for which
interpretation is accordingly more ambiguous. Examples include the
rectilinear features up against the northern boundary [Q]; zones of increased

response [R], possibly indicating metalling; the linear responses [S] and [T]
which have markedly different orientations to the other features identified but
which could be more drains or some such.
4.3
Conclusions
4.3.1
The magnetometer survey was able to provide a detailed plan of the vicus to
the east of the fort and other archaeological features to the north as well as
complementing the radar results within the interior of the fort. Use of the
GPR was focused within the fort interior and was able to achieve incredibly
detailed results, identifying the principia, a possible mansio, and a granary
as well as hypocaust systems, drainage routes, intact floor surfaces, pillar
bases and buttresses.
4.4
Evaluation Trenches
4.4.1
Five trenches were excavated as part of this investigation. Trenches 1 and 2

both lay within the interior of the fort, targeted on buildings of varied
alignment. Trench 3 was positioned in the eastern field over the eastern
vicus settlement. Trench 4 was situated over the northern fortifications of the
fort and Trench 5 lay in the northern field over a possible earlier defensive
structure. The size and shape of the trenches varied to account for the
potential targets that they were sited on and the archaeology subsequently
uncovered. Any substantial remains were left in situ. Those trenches in the
main and eastern field lay at heights of between 10-15.5m aoD while Trench
5 lay at around 8m aOD.
4.4.2
The trenches within the fort interior saw the removal of between 0.25-0.50m
of overlying topsoil while Trench 3 saw 0.58m of overlying topsoil. Only in
Trench 5, which lay to the north of the fort was a thin (0.15m) subsoil seen
beneath 0.40m of topsoil. Where encountered the natural geology was sand
with gravel inclusions.
4.4.3
Trench 1 (Figures 4 and 5)

Trench 1 was positioned over what had been identified as the principia or
headquarters of the fort. The trench was situated over the central part of the
southern range of rooms. In general, though largely relating to abandonment
and later activities, the pottery suggests activity from the 2nd into the 4th
century.
4.4.4
The earliest stratigraphic features identified were two sections of east-west

aligned wall, 128 thought to be the external south wall and 130 an internal
wall. Both walls lay over 1m below the present ground surface and
continued below the limit of excavation.
4.4.5
Wall 128, which lay in the southern part of the trench, was at least 0.90m
wide and constructed from stone bonded with a dark yellow mortar. Sand
deposits 142 and 143 lay to the south of the wall, separated by clay deposit
141. As no construction cut was visible and the wall depth continued into
143 these layers must be re-deposited and either levelling or possible
construction cut deposits. The origin of 141 is curious but could potentially
provide a firmer footing during the construction of the wall.
4.4.6
Above re-deposited sand layer 142 was a thin layer of stone chippings (136)
which may relate to the construction or potentially the demolition of wall 128

(Figure 4, Section 2; Figure 5, Plate 1). Overlying this was 122, which may
represent general build up during the life of the fort and contained 2nd
century AD pottery. Above this was a distinctive tile-rich deposit (121) that
could indicate collapse of the roof or, given the fragmentation of the tiles,
material discarded during dismantling of the roof. This layer was equivalent
to 115, identified slightly further to the south. Covering tile deposit 121 was
120, a spread of demolition material; this is equivalent to 114 which overlay
115 to the south.
4.4.7
Cutting through demolition debris 120 was robber cut 127, which had
removed the greater proportion of wall 128 (Figure 4, Section 2; Figure 5,
Plate 1). The homogeneous nature of the single fill 126 suggests it was
deliberately backfilled and 4th century AD pottery was recovered from this
deposit. Backfill deposit 126 lay beneath 107, which is equivalent to 105 and
106 seen elsewhere in the trench.
4.4.8
The interior wall 130, located in the central area of the trench, was similarly
east west aligned and its stone block construction is largely obscured by a
pale yellow mortar (Figure 5, Plates 2 and 3). With a width of only 0.48m it
is narrower than 128 but this is to be expected for an internal division.
Though its position suggests it divides two rooms, the deposits to the north
and south of the wall differ considerably.
4.4.9
Against the southern face of wall 130 at the limit of excavation a portion of
lorica squamata (Object Number [ON] 83) was recovered (Figure 5, Plate
4). This artefact (assigned context 129) was situated on a possible pinkbrown mortar surface 132, though only a small area was seen along the wall
face. When ON83 was lifted, a sand layer (139) could be seen beneath 132,
though as the wall could be seen still to continue, this layer is thought to be
re-deposited. Above 132 but not extending fully to the wall face was another
mortar deposit of pale-yellow white mortar (131), a possible later surface.
4.4.10
Above mortar 131 and extending fully to abut wall 130 was occupation
deposit 125. Within this layer were a number of pieces of articulated animal
bone representing cuts of meat (Animal Bone Group [ABG] 82) as well as a
probable scattered coin hoard deposited in the mid 4th century AD. An
environmental sample (ES 1) taken from this context found low numbers of
cereal seeds, but a number of plant seeds mainly found in wasteland and
arable margins were present, supporting the idea that this context relates to
the final or post-abandonment phase of the fort. This deposit was overlain by
124, a layer of alternating lenses of mortar and dark silts which must surely
post-date the decommissioning of the fort. Over this, and capping the
remnants of wall 130, was a spread of mortar (123) (Figure 5, Plates 2 and
3).
4.4.11
Mortar 123 also covered the layer of built up material to the north of wall 130
(138). Unlike 124 to the south, 138 was a fairly clean sandy deposit with
occasional fragments of stone. It was stratigraphically above a sand deposit
(140) which, for the same reasons as 139 to the south, must have been redeposited. These two layers are at a similar height in relation to wall 130 and
may therefore be equivalent deposits.
4.4.12
Overlying 123 and directly above wall 130, though at a slightly divergent
alignment was a defined area of pink-red mortar or crushed tile (137)

(Figure 5, Plate 2). It was seen mostly clearly in the western part of trench
though it may continue on the eastern edge of the sondage. Set virtually
upright at the south-east corner of this possible structure were two or three
mortared tiles. What is unclear is whether this is a deliberate structural
element in situ or whether it is a structural element, possibly an arch, which
has maintained a degree of structural integrity despite falling out of position.
Covering this and 123 was 106, an artefact-rich deposit which appears to
post-date the main demolition of the building, equivalent to deposits 105 and
107 elsewhere. Pottery from this context spans the period from the 2nd to
the 4th century AD, while the environmental sample (ES 3) contained
evidence for grass and heather and utilisation of hazelnut and sloe in
addition to some wheat, barley and spelt charred remains.
4.4.13
At the far northern end of the trench and constructed on layer 106 was a
very rough series of large blocks 112. Although only one course high it does
appear to be represent a possible plinth, of likely post-Romano-British date
(Figure 5, Plate 5).
4.4.14
A sondage dug immediately to the south of 112 located a potential rough

surface 134 at the limit of excavation (Figure 4, Section 1; Figure 5, Plate
5). It lacked coherent structure but did contain a number of flat stones and
appeared to be overlain by a layer of trampled, charcoal-rich material (133).
Over this was a tile- and mortar-rich deposit 119 similar to 115/121 and
which is also likely to derive from dismantling or demolition of the roof.
4.4.15
Indications of possible post- Romano-British activity can be seen in levelling

deposit 118 which was above 119, onto which layer 117 and then possible
mortar surface 116 appear to have been laid. This layer of mortar is sealed
beneath 106.
4.4.16
Another later phase of use may be indicated by 111. Although it may have
been merely a spread of rubble, the sharply defined limits of its extent could
indicate a rough surface. Other discrete demolition deposits were seen in
the central part of the trench where rubble-rich 109 was partially overlain by
similar deposit 113.
4.4.17
Rubble layers 111, 113 and possible plinth 112 all lie beneath 102, a deposit
rich in domestic debris and datable late Romano-British artefacts, but which
must surely post-date the abandonment of the fort. It is equivalent to 104
and 108 and was seen through the full area of the trench. Such a wide
extent suggests that this material has been deliberately spread, probably by
agricultural activity. Deposits 104 and 108 directly overay layers 105 and
107 which are equivalent to 106.
4.4.18
A single feature was observed cutting 102. Feature 135, which lay beneath
the modern topsoil 101 was not fully excavated and was difficult to
distinguish in plan but may be a small pit. It contained a single fill 110. No
dating evidence was recovered from this feature.
4.4.19
Trench 2 (Figure 6)
Cropmarks clearly indicated another building within the fort interior just to the
north-east of the principia. It had been suggested that as this appeared to be
on a different alignment to the fort it may have belonged to a different phase
of activity. Trench 2 was situated on the southern edge of this structure and
10

also on the northern edge of another building identified from cropmark

evidence, fronting the via principalis.
4.4.20
It is believed that undisturbed natural geology 211 was only encountered at

a depth of 0.8m+ below the current ground surface and that the sand
deposits into which the majority of the archaeological features were cut, or
on which they were situated (209 and 210), are in fact made ground
deposits, probably to level the ground prior to construction. Artefactual
material was recovered from both 209 and 210 (which were identical
deposits, originally numbered separately as they were encountered during
excavation). The presence of a clearly diagnostic sherd of Middle Saxon
pottery (Ipswich ware) in deposit 209 is problematic, but the sandy and
friable nature of the layer and overlying deposits meant that considerable
bioturbation was evident, as shown by the dark root lines. The rest of the
pottery recovered from the layer suggests a 2nd to 3rd century AD date, in
keeping with the postulated date of establishment of the fort. Interestingly
the presence of human neonatal bones in deposit 210 suggests that this
initial occupation included a female presence. At the interface of 210 and
211, a thin layer of darker material (221) was seen which may represent an
area of trample. A single sherd of pottery recovered from this layer is
probably of Iron Age date.
4.4.21
The base of a small chalk rubble wall (205) ran on an eastwest alignment
(Figure 6, Section 3, Plates 6 and 7). This was relatively narrow with no
real foundation, suggesting it is more probably an internal division than an
external or load-bearing wall. Comparison with the interpretation of the
cropmarks (Hincliffe and Sparey Green 1985, fig. 2) suggest that in fact this
wall is the northern wall of a long building situated along the eastwest road,
rather than part of the differently aligned building. To the north of this were
remnants of surfacing (Figure 6, Plate 6). Surface 204, which directly
abutted wall 205, contained several layers of make-up, the lower portion
composed of mortar with chalk and gravel inclusions overlain by a sandier
layer with frequent gravel. In places, patches of a final mortar surfacing were
observed. Further north, the surface was damaged and truncated by
additional areas of surfacing which were seen at the northern end of the
trench, here numbered 203. In this northern part of the trench, a disturbed
interface layer 202 was recorded beneath the topsoil and above surface
203.
4.4.22
Despite the cropmark evidence, no indication of another wall was found in

the northern part of the trench though the metalled surface 203 suggests a
yard in this area. It is possible that further masonry remains lie beneath the
limits of excavation or that the area of disturbance seen within 203
represents the removal of a wall. The GPR results suggest that the main
structure of the building lay slightly further to the north-west (Figure 3).
4.4.23
Cutting through 204 was a small east west aligned gully 208 filled with a
single fill of topsoil-derived material (207). This feature, cutting through the
Romano-British deposits, is likely to be much later in date, and the finds
within it residual, although they are still of largely 2nd and 3rd century date.
4.4.24
To the south of the wall 205 was a rubble spread (206), thought to represent
demolition debris; it was composed of chalk and flint and may represent
11

remnants of wall 205. Removal of 206 in the western part of the trench
showed that it overlay a number of features (Figure 6, Plate 8).
4.4.25
Three discrete areas of mortar (212, 213 and 214) appear to have been
post-pads for a north-westsouth-east aligned structure. Further post-pads
may line beyond the limit of excavation to the west and beneath 206 to the
south-east. A fourth feature which may also relate to this group was a
possible posthole 223. Directly alongside the wall 205, its relationship to this
feature was unclear, nor was it clear whether large fragments of stone and
chalk rubble within its fill (224) were remnants of post-packing or collapsed
material from the wall.
4.4.26
Two small pits were found at the southern end of the trench (215 and 219).
Pit 219 was sub-oval in shape with a single fill (220) and a concentration of
charcoal flecks near the base. Pit 215 was slightly more irregular in shape
with what appeared to be a posthole incorporated into its southern end.
There were some indications of in situ heating along the cut margins with a
red discolouration 218. At the base of the main part of the pit but not within
the posthole was a more clayey deposit (217) which may be a deliberate
lining. The final fill 216 was a more general mixed deposit which also filled
the posthole void. Indications are that pits 215 and 219, as well as the group
of post-pad 212, 213 and 214 represent an area of occupation but not one
necessarily associated with the military use of the fort. However the pottery
recovered still falls within the 2nd to 3rd century AD period seen elsewhere
in the trench.
4.4.27
Trench 3 (Figure 7)
Trench 3 was situated outside the fort, within the eastern vicus settlement. It
was located on the south-western edge of a probable crossroads identified
from the cropmark evidence. Overlying the archaeology was a deep former
ploughsoil, indicating that the archaeology had potentially been truncated
since the cropmarks were identified. Indications from local residents are that
the field has been deep ploughed within recent memory. The only 2nd
century AD coin recovered from the Site was found unstratified in this
Trench (ON 29).
4.4.28
The earliest feature encountered was a north-eastsouth-west aligned ditch,

numbered 312 and 313 in the slots that explored the northern and southern
edges of this feature respectively (Figure 7, Plates 9 and 10). The feature
was not fully excavated but was 2.3m wide and over 1.2m deep. Its size
suggests a large enclosure ditch. The lowest fill excavated was 307/314, a
distinctive banded sandy deposit (Figure 7, Section 4 and Plate 10),
thought to be the result of a sequence of depositional events, probably the
result of water action. Above this was a deep but fairly homogeneous fill
(303/304/315). The uniform nature of this deposit and the presence of a
number of large, unabraded sherds of pottery suggest that this was
deliberately backfilled. The pottery from both the initial deposits (307/314)
and the later backfilling (303/304/315) all fall broadly within the 2nd and 3rd
century AD, but the presence of a sherd of Late Saxon Thetford ware from
the lower fill 307 should be noted; this may be intrusive.
4.4.29
In the eastern part of the trench, feature 310 was partially exposed in one of
the excavation slots. As it was not fully exposed in plan its exact nature is
unclear, but it would appear to have been either an eastern ditch terminus or
12

an elongated pit. It contained a single secondary fill 311. No relationship

between 310 and enclosure ditch 312/313 could be determined.
4.4.30
Feature 310 lay beneath and was truncated by curvilinear feature 308. This
shallow but relatively wide feature also cut through enclosure ditch 312/313.
It contained a single secondary fill (309) incorporating occasional fragments
of animal bone and 2nd to 3rd century pottery.
4.4.31
Few traces of the road surface itself remained though some patches of
metalling were identified (302 and 316), particularly where they had settled
into the top of earlier features. Pottery recovered from the road metalling 302
and 316 suggests 3rd and 4th century activity. The southern roadside ditch
305 was also identified, which was found to have a steeper nearside edge
(Figure 7, Section 4). Some of the gravel metalling from the road had been
eroded and was incorporated into the upper part of ditch fill 306. Ditch 305
cut ditch 308.
4.4.32
Trench 4 (Figure 8)
Trench 4 was located over the northern (seaward) fort defences on the
western edge of the entranceway.
4.4.33
At the northern end of the trench was a defensive wall (403), constructed
from micaceous stone and flint nodules. It was a substantial structure, 2.5m
in width (Figure 8, Plate 11). As observed on previous excavations, the
rampart (413, 414), constructed from re-deposited natural sand, was internal
and directly abutted the wall. Even allowing for a certain amount of spread
along the southern edge, the width appears to be over 5m. Dividing the very
similar upper (413) and lower (414) deposits was a thin discontinuous lens
of charcoal, 415. Environmental evidence obtained from a sample (ES 2)
suggests that this charcoal lens represents an occupation layer, which could
therefore indicate a second phase of construction, perhaps to increase the
height of the rampart (Figure 8, Section 5).
4.4.34
Defensive ditch 406 lay to the north of wall 403 (Figure 8, Section 5). No
direct relationship between the two features could be established but they
are likely to have been contemporaneous. Indeed, the lowest ditch fill
encountered (420) appeared to be mortar debris deriving from the
construction of the wall. Neither the full depth nor profile of the ditch 406 was
exposed. Both the upper deposit (404) and deposit 409 below are likely
either to post-date the defensive life of the feature or to have occurred very
late within the sequence, although pottery recovered falls within a date
range of 2nd to 3rd century AD.
4.4.35
Robber cut 408 ran along the upper part of the wall 403 and down its
northern face (Figure 8, Section 5). The cut was filled with several deposits
(402, 405, 407 and 418); both 405 and 407 are likely to have been
deliberate backfill events while 418, which contained a large number of
stone chippings, may represent reclamation debris.
4.4.36
Overlapping the southern edge of the upper portion of rampart bank 413
was a mixed deposit (419) likely to reflect demolition or abandonment
(Figure 8, Plate 12). This was cut on its southern edge by robbing event
411. This comprised several robbing episodes, and had cut through what
was presumably originally one area of surfacing (412). Its position, just
13

within the rampart and adjacent to the road, suggests it could have been
flooring or foundation within a guardhouse - indications of a building are
shown here on both the magnetometer and GPR survey (Figures 2 and 3).
The date and exact purpose of robbing event 411 is unclear.
4.4.37
Exposed at the base of cut 411 was a possible makeup deposit (416)
containing a large fragment of a 2nd or 3rd century Rhenish mortarium, and
beneath this in turn was another mortar layer (417). This could indicate an
earlier structural phase beneath the structure represented by 412.
4.4.38
The presence of redeposited neonatal bones in the topsoil and an adult tibia
in the upper robber cut backfill 402 could indicate nearby burials.
4.4.39
Trench 5 (Figure 9)
Trench 5 was situated in the northern field across a double ditched feature
identified from cropmark evidence, thought potentially to be the remnants of
an earlier fort.
4.4.40
Only in this area was a thin, underdeveloped subsoil identified (502),

beneath the modern topsoil. As in the other excavation areas, the natural
(510) consisted of sand with some coarse flint gravel.
4.4.41
The two ditches (503 and 506) were found to have very similar profiles
(Figure 9, Section 6), though the outer ditch (503) was slightly more
substantial. The main fills of both ditches were very similar and
homogeneous, necessitating a slightly arbitrary division between the upper
and lower deposits in each case as the interface between them was
extremely diffuse. The exception was 509, the lowest fill of 506 which was
derived from the collapse of the southern edge. The pottery recovered
indicates largely 2nd and perhaps some 3rd century activity. A single
neonatal bone was also recovered from fill 505 (ditch 503).
4.4.42
Such dark homogeneous deposits could be reflective of a long period of

gradual infilling and sediment accumulation or equally they could indicate a
single period of deliberate backfilling. Whichever is the case is likely to
indicate whether there was a hiatus between this defensive earthwork and
the fort to the south. Once the main fort was established it seems unlikely
that another defensive structure in close proximity would be tolerated; either
enough time must have passed for the earlier fortifications to have naturally
infilled, which is perhaps unlikely given the timescale, or if there is continuity
of occupation the ditches must have been deliberately backfilled.
FINDS
5.1
Introduction
5.1.1
Finds were recovered from all five of the trenches excavated, although
quantities from Trench 5 were relatively low. The assemblage is
predominantly of Romano-British date, and relates to the construction and
use of the fort complex; there are also a few prehistoric items (worked flint,
pottery sherd), and a small quantity of post-Roman (or probable postRoman) material (pottery, ceramic building material, gaming die, metal
objects).
14

5.1.2
The whole assemblage has been quantified by material type within each
context; finds totals by material type, subdivided by trench, are presented in
Table 1 (Appendix 2). As part of this assessment stage, all finds have been
at least visually scanned, and preliminary identifications and spot dates
recorded. On this information is based an assessment of the potential of the
finds assemblage for further research.
5.1.3
The assemblage is described and discussed by material type below.
5.2
Pottery
5.2.1
The pottery was divided into fabric groups and quantified by number of
sherds, weight and rim estimated vessel equivalent (EVE) per fabric. As an
additional measure, vessels identifiable to form (mostly rim and base
sherds) were recorded for each context by fabric. The pottery data was
entered onto an Excel spreadsheet. The total assemblage comprises some
653 sherds, weighing almost 15 kilos and with a rim EVE of 17.7 (Appendix
2, Table 2). This is almost exclusively of Romano-British date, with one late
prehistoric and three post-Roman sherds.
5.2.2
The pottery is compared to the substantial pottery assemblage (260kg)

recovered from excavations at Brancaster in 1974 and 1977 and
subsequently published (Andrews 1985).
5.2.3
5.2.4
Prehistoric pottery
One small, undiagnostic body sherd in a coarse shelly fabric from layer 221
is not particularly chronologically distinctive, but is tentatively dated as Iron
Age.
Romano-British Fabrics
Just over 50% of the Romano-British pottery comprises various reduced
grey wares, with regionally-traded wares from Dorset, the Thames estuary,
the Lower Nene Valley, Oxfordshire and Hertfordshire, and continental
imports from France, Spain and the Aegean accounting for around another
third. Table 3 (Appendix 2) shows the fabric proportions.
5.2.5
The various reduced grey wares are all quartz sand-gritted. Mica is
prominent in some of the wares with one sherd being highly micaceous.
Some of the vessels have burnished surfaces and the most noticeable of
these has a black, highly burnished, almost polished surface; it is possible
that the black colour may be in fact due to a slip. The regionally-traded
wares are BB1 and BB2, Lower Nene Valley colour-coated and cream wares
(LNVCC, LNVCW) and Oxfordshire colour-coated, parchment and whiteslipped wares (OXCC, OXPA, OXWS). While there are definite examples of
BB1 and BB2 in the assemblage, there are some similar fabrics which may
be of more local origin. The continental imports comprise samian ware
(CGS), Rhenish ware and amphora from France, Spain and the Aegean.
The only other ware which occurs in appreciable amounts is shell-gritted
ware and there are a few sherds in flint-gritted, mica-dusted, buff and other
oxidised wares.
5.2.6
All of the fabrics are present in the 1974 and 1977 assemblage. Table 4
(Appendix 2) provides a fabric concordance for the reduced grey wares.
15

5.2.7
5.2.8
The 1974 and 1977 assemblage contains some fabrics not present in the
2012 pottery, comprising Lower Nene valley grey ware, Colchester colourcoated ware and Dales shell-gritted ware. Andrews notes that fabric RW10
represents a number of fabrics (Andrews 1985, 92-3) and that fabrics
RW11-20 comprise grey wares with no outstanding characteristics which
were, however, consistently distinguished and sorted (ibid., 93-4). RW11
was difficult to distinguish from BB2 and some of RW10 from BB1.
Romano-British Vessel Forms
The minimum number of identifiable vessels, as represented by mainly rims
and bases, totalled 171, of which five are amphora. Some of the 166 others
may be the same vessels, and there are a large number of body sherds
which could be from these or different vessels. Table 5 (Appendix 2) shows
the occurrence of forms by fabric.
5.2.9
The assemblage has a wide range of vessel forms. Jars occur in the most
fabrics but many fabrics also include bowls and/or dishes in their vessel
range. Some wares are more specialised, as with the LNVCW, OXWS and
Lower Rhineland mortaria, the Rhenish beaker and the shell-gritted jars. The
LNVCW mortaria comprise bead and grooved and reeded flange types while
the OXWS mortaria are mainly Young (1977) type M22. The Lower
Rhineland mortarium is an extra large vessel, which can probably be
attributed to the workshop of Verecundus at Soller; an edge of a stamp is
just visible on this vessel. The OXCC and CGS vessel range is mainly
confined to bowls and dishes, while the latter ware also has a cup and a
mortarium. Many of the OXCC bowls are Young forms C75 or C77, together
with a C50 and a C52. The CGS occurs as forms 18/31 or 31, 18/31R or
31R, 33, 37, 38 and 45.
5.2.10
The BB1 and BB2 vessels are mainly the more widely-traded bowl and/or
dish types, the former ware occurring as flanged bowls with intersecting arc
decoration and the latter ware comprising rounded rim types. The LNVCC
vessel range is one of the widest with flanged bowls, plain-rim dishes,
beakers and wide-mouthed jars or bowls all well represented, together with
a flagon and two lids, one flanged and the other the upper part of a Castor
box.
5.2.11
The vessels in the distinctive highly burnished (or slipped?) black-surfaced

ware have some interesting characteristics. Many of the bowls and dishes
are curved-sided and have one or more external grooves below the rim. A
few dishes have internal grooves and one dish has a thickened rim boss
with two grooves cut on its upper surface. Other bowls have either a low
flange or a small, high flange and some have facetted burnishing akin to that
found on BB1 vessels. Many body sherds, probably from jars, have
unburnished bands down the vessel wall which are decorated with lattice,
wavy lines and, occasionally, stabbing. One or two of the jars have frilled
rims. A number of body sherds in other grey wares have horizontal bands of
closely-set combed wavy lines and the grey, Rustic ware sherds all have the
linear form of rustication.
5.2.12
These forms all occur in the 1974 and 1977 assemblage (Andrews 1985,
figs 53-66). As examples, fig. 63, types 150-1 are black-surfaced ware bowls
and dishes, linear rustication occurs on fig. 56, type 100 jars, fig. 53, types
50.1 and 50.3, and fig. 55, type 94 are examples of the LNVCC funnel-
16

necked folded beakers, fig. 54, type 61 are similar Castor box lids and fig.
63, types 145 and 147.1-4 are BB1 or BB1 type vessels. The 1974/1977
assemblage even has another Aegean hollow foot amphora (ibid., 84, no.
75, 104, not illustrated) but contains many forms not present in that from the
2012 excavations; this is not surprising, given its size. It does not include,
however, any of the dark grey burnished vessels, probably jars, which have
unburnished bands down the vessel wall decorated with lattice, wavy lines
and, occasionally, stabbing, though these all occur individually on some of
the 1974/1977 vessels.
5.2.13
Sources
It is likely that most of the reduced grey wares will have been locally
produced. There are no kilns known around Brancaster itself, but there were
large production centres at Brampton, Pentney and Shouldham and others
at Sheringham, Snettisham, Lyng and Witton (Swan 1984, map 15). These
are all between 30 and 50 kilometres from Brancaster but military sites such
as Brancaster had more elaborate and far-reaching supply routes. Some of
the grey ware may have originated from the kilns at Horningsea near
Cambridge and kilns in the East Midlands which were part of the East
Midlands burnished ware production. The sherds in reddish-yellow ware
vary with some being highly burnished and others having red- or brownpainted horizontal bands or a white slip. Some are probably from the Lower
Nene valley kilns, but other sherds may be from Much Hadham in
Hertfordshire. It is likely that most of the buff sherds are also from the Lower
Nene valley, but some could be from the Oxfordshire kilns.
5.2.14
The vessels in BB2 will have originated from kilns in along the Thames
estuary and those in BB1 from south Dorset BB1 but, in both cases, some
may, in fact, be good locally-produced imitations, as the fabrics are not
always readily distinguishable from some of the other reduced grey wares.
There is far more certainty with the colour-coated wares and cream wares
from the Lower Nene Valley and the Oxfordshire kilns; the buff colour-coated
sherd is probably from the former. The samian ware is from Central Gaul,
the Rhenish ware from the Trier region and the amphora from southern
Spain, southern France, together with the handle of an unusual hollow-foot
amphora, probably of Aegean origin (Peacock and Williams 1986, 193-5,
class 47). The shell-gritted ware may have been produced in the Harrold
areas of Bedfordshire.
5.2.15
A programme of heavy mineral analysis was carried out on some of the

1974 and 1977 pottery (Andrews 1985, 82, 88, 90, 92). This indicated that
fabric RW1 has links with the Shouldham kilns, though it is noted that the
glacial drift deposits forming the clays used there are widespread in Norfolk,
Indeed, Lyons (2004) states that large quantities of pottery from kilns at
Snettisham occur at Brancaster. Shouldham, Pentney and Blackborough
End, Middleton (Gurney 1990) are all part of the same Nar Valley/West
Norfolk pottery industry, so the any one or a combination of these could
have been the source for the Brancaster material. The analysis also
suggests that fabrics RW2 and RW are products of the Brampton kilns while
kilns in the Homersfield/Wattisfield area are thought to be the source for the
micaceous wares.
17

5.2.16
Date
The assemblage contains a lot of 4th century AD pottery, especially the BB1
flanged bowls, the LNVCC flanged bowls, plain-rim dishes, beakers and
wide-mouthed jars or bowls, the OXCC bowls and dishes and some of the
shell-gritted ware jars. Some of the LNVCC beakers and the mortaria in
LNVCW and OXWS are types which originated in the later 3rd century but
continued into the 4th century. More definite 3rd century vessels are the BB2
bowls and dishes and a LNVCC Castor box lid. The CGS, Rhenish ware,
mica-dusted ware and grey Rustic ware attest 2nd century activity. The
earliest pottery, possibly of pre-Roman date, comprises the sherds in flintgritted ware. The kilns at Shouldham, Pentney and Brampton were in
operation from the 2nd to 4th centuries.
5.2.17
The trenches were positioned to hopefully provide additional dating evidence

for specific areas of the Brancaster site. The pottery from Trench 1, located
over the central part of the southern range of rooms of the principia or
headquarters building, is predominantly 4th century in date. That from
Trench 2, in the main fort area, appears earlier in date than Trench 1, with
an overall range of mid 2nd to 3rd centuries. The date range of the Trench 3
pottery, from within the east vicus area, appears to be 2nd to 4th centuries,
while that from the fort defensive ditch and rampart (Trench 4) and the
double ditch feature thought to potentially be the remnants of an earlier fort
Trench 5) both have a mid 2nd to 4th century date range, though the latter
may start in the later 2nd century.
5.2.18
The chronology postulated by Hinchliffe in the report on the 1974 and 1977
excavations (Hinchliffe 1985, 180-1) was that a fort was established in the
late 2nd century and an adjacent settlement was soon established. This fort
was replaced by a larger one in the second quarter of the 3rd century with
occupation continuing, based on coin evidence, into the 5th century. The
settlement itself appeared to have been abandoned by the 4th century. The
dating evidence provided by the 2012 excavation pottery broadly supports
this chronology.
5.2.19
5.2.20
Assemblage characteristics
The pottery is generally in good, unabraded, condition with a number of
large sherds, complete bases and vessel profiles. One jar rim has a pierced
hole in the neck just below the rim and some other jar sherds have internal
limescale accretion. The approximate 50:50 ratio of reduced grey wares
and other wares which were probably locally produced to regionally-traded
and imported continental wares would be unusual on most rural Norfolk
sites, but reflects the military nature of the main occupation, with its
attendant civilian vicus. This character is borne out in the wide range of
vessel types within the various fabrics, which includes unusual vessels like
the Aegean hollow-foot amphora.
Post-Roman pottery
The three post-Roman sherds comprise the rim and spout from a middle
Saxon Ipswich ware spouted pitcher (c. AD 720850) from 209; the rim from
a late Saxon Thetford ware jar (10th to mid 12th century) from 307; and a
small body sherd in modern (19th/20th century) refined whiteware from
Trench 3 topsoil.
18

5.2.21
5.2.22
5.2.23
The Trenches
Trench 1
Trench 1 was located over the central part of the southern range of rooms of
the principia or headquarters building. The contexts containing pottery
excavated comprise post abandonment layers, a post stone-robbing deposit,
occupation debris/refuse and a finds retrieval cleaning interface. The Trench
1 contexts produced a large percentage of the total site assemblage,
ranging from 40-55% (Appendix 2, Table 2). Table 6 (Appendix 2) shows
the fabric proportions. Most of the fabrics represented in the overall site
assemblage occur in the Trench 1 contexts and in similar overall
proportions. There is, however, a higher percentage of Oxfordshire wares
and the most noticeable absentees are BB2, Lower Rhineland and Rhenish.
Table 7 (Appendix 2) shows the vessel forms per fabric from Trench 1. The
contexts again have most of the forms occurring in the overall site
assemblage and, in some cases, contain all the forms in certain fabrics; the
form ratio is also similar. The LNVCC vessels include flanged bowls, plainrimmed dishes a narrow-mouthed jar and a jar sherd with bosses (cf Corder
1961, fig. 24, 5). The LNVCW mortarium has a bead and grooved flange (cf
Howe et al. 1980, fig. 8, 102), while two of the OXWS mortaria are Young
(1977) type M22. The OXCC comprises Young (1977) forms C50, C52, C75
and C77.
Trench 2
Trench 2 investigated the main fort area. The contexts contain around a fifth
of the total site assemblage (Appendix 2, Table 2) and Table 8 (Appendix
2) shows the fabric proportions. The range of reduced grey wares is smaller
than Trench 1 and there are no Oxfordshire wares present. The only
Rhenish ware from the Site occurs in a Trench 2 context.
5.2.24
There are far less forms from Trench 2 and the range occurring is also more
limited (Table 9, Appendix 2). The possible BB2 dishes have plain and flattopped rims while the jar has lattice decoration. The LNVCC dish has a plain
rim and the lid is flanged (cf Perrin 1999, fig. 62, 214). The CGS cup is form
33.
5.2.25
The presence of a sherd of Middle Saxon Ipswich ware from sand deposit
209 can be noted.
5.2.26
5.2.27
Trench 3
Trench 3 is located within the east vicus area. The features comprise
various ditches, including a road ditch and contain between 11% and 16% of
the pottery from the excavations. The reduced grey wares are the most
numerous fabric types and there are more BB2 and CGS than in Trenches 1
and 2, with Lower Nene Valley and Oxfordshire wares being correspondingly
lower (Appendix 2, Table 10).
The form range is greater than that in Trench 2 (Appendix 2, Table 11). The
amphorae are from southern Spain and southern France and the LNVCW
mortarium is a hammer-head type (cf Hartley and Perrin, 1999, fig. 77, M20).
The LNVCC includes an imitation samian ware form 38 and the OXCC a
beaded-rim bowl. The possible BB1 bowl has a flanged rim, facetted
burnishing and intersecting arc decoration.
19

5.2.28
5.2.29
5.2.30
5.2.31
A single sherd of Late Saxon Thetford ware was recovered from the lowest
excavated fill in the large enclosure ditch 312/313.
Trench 4
Trench 4 was located to investigate the fort defensive ditch and rampart.
The assemblage size is similar to that in Trench 2, also comprising around a
fifth of the overall excavation total (Appendix 2, Table 2). All bar eight
sherds come from the topsoil. The proportion of reduced grey wares is lower
than the previous three trenches with the amount of other fabrics, especially
BB2 and LNVCC, correspondingly higher. Weighty sherds also boost the
amphora percentage (Appendix 2, Table 12).
The number of vessels is a little higher than Trench 2 (Appendix 2, Table
13). The contexts contain five of the six LNVCC beakers from the
excavations. Two of these are tall funnel neck types (cf Howe et al. 1980,
fig. 4, 43) and another is a folded beaker with rouletted bands (ibid., fig. 4,
42). The LNVCW mortarium is a bead and grooved flange type (ibid., fig. 8,
102) and the CGS form 45 mortarium is the only samian ware mortarium
from the excavations. The BB2 vessels both have rounded rims. One of the
amphorae is of significant interest, comprising the handle of a hollow-foot
amphora, thought to have been manufactured in the Aegean and occurring
in Britain in late 3rd to early 4th century contexts (Peacock and Williams
1986, 193-5, Class 47).The other amphorae are from southern France.
Trench 5
Trench 5 was situated across a double ditch feature which is thought to
potentially be the remnants of an earlier fort. The assemblage from the
contexts is quite small, accounting for a maximum of 5% of the pottery from
the excavations (Appendix 2, Table 2) and the various reduced wares
comprise over three-quarters of the total (Appendix 2, Table 14). The
contexts contain the only possible pre-Roman pottery, flint-gritted pottery.
5.2.32
Only four vessels occur in the Trench 5 assemblage, but these include a
micaceous grey ware dish with a plain rim and the lid of a LNVCC Castor
box (cf Perrin 1999, fig. 62, 210).
5.3
Ceramic Building Material
5.3.1
The complete CBM assemblage recovered from the Site amounted to 627
fragments. An on-site selection strategy was adopted, in which undiagnostic
flat fragments were quantified (count and weight) and then discarded. An
assemblage totalling 433 fragments was retained for further processing and
assessment. This assemblage is almost entirely of Romano-British date;
only three post-Roman fragments were recorded (medieval roof tile
fragments from ditches 312 and 308, and an early post-medieval brick from
ditch 312).
5.3.2
The retained assemblage has been quantified by type (imbrex, tegula, etc)
within each context, with thicknesses and other selected dimensions (e.g.
tegula flange height) recorded, as well as the presence of features such as
combing on box flue tiles, paw prints and finger-smeared signatures. For
tegulae, upper cutaway type, where present, was classified using Warrys
typology (2006). The fabric types were noted but not recorded in detail as
the majority of the assemblage comprised fragments in non-distinctive hard-
20

fired, slightly sandy fabrics firing orange-red. A very few fragments differed
significantly from this norm.
5.3.3
Table 15 (Appendix 2) gives the breakdown of CBM types by context. Roof

tiles (tegulae and imbrex) formed the major component of the assemblage
(21% and 24% respectively by count). A very small quantity of box flue tiles
(tubuli) from hypocaust heating systems was recovered. A significant
proportion of the retained assemblage, however, comprised flat fragments
lacking diagnostic features on which to assign them to specific tile or brick
types; these were divided into those less than 30mm in thickness, and those
of a greater thickness; the former are likely to represent further examples of
tegulae, imbrices and box flue tiles, while the latter probably derive from
bricks of various forms. Pieces lacking any dimensions or distinguishing
features were classified as unidentifiable fragments.
5.3.4
No complete dimensions were noted amongst the assemblage. Within the

tegulae fragments it was apparent that thickness, as well as flange width
and height, varied. Flange height is generally considered to be roughly twice
the tile thickness in this instance it ranged from 30 to 50mm; flange profile
was generally squared, occasionally with a single finger smeared groove
along the flange top. Fifteen cutaways were observed, both upper and lower
examples; those complete enough to be identified to type comprise three of
Warry type C5 and single fragments of a C4 and D16 (Warry 2006). Within
Warrys classification of cutaways type C is assigned a date range of 140
260 AD and type D is considered to date 240 AD onwards.
5.3.5
There were two imbrex fragments with finger smear marks along the length
of the upper surface, which appear to be deliberate markings. Signature
marks were relatively uncommon; four were recorded on flat tile fragments
and two marks on tegulae. All comprised two or more finger smears in a
curved design. One paw print was recorded, on a flat tile from trench 5.
Occurring alongside this print were two joining fragments of a tile which had
been tapered along one edge. A second tile (from Trench 1) had possibly
been re-worked into an S shape. Most box flue fragments carry some form
of keying for mortar in the form of linear (often cross-hatched) combing.
5.3.6
CBM was recovered from a total of 37 contexts across all five excavated
trenches. Most contexts are described as occupation debris/abandonment
layers. By far the majority of the assemblage was recovered from Trench 1,
not unexpected, given the location of the trench over the principia. The small
number of tegula cutaways that can be identified to the mid 2ndearly 3rd
century AD may lend further weight, albeit somewhat tenuous given the
small numbers involved, to the possibility of a 2nd century structure on the
Site.
5.4
Mortar, opus signinum and wall plaster
5.4.1
Building material is also represented by small quantities of mortar, opus

signinum and wall plaster, recovered mainly from Trenches 1 and 2 in the
fort interior, and largely from demolition or post-abandonment contexts; none
was found in situ. The wall plaster is all monochrome white in colour.
21

5.5
Stone and Worked Flint
5.5.1
Of the four pieces of stone recovered, two represent building material, one is
a portable object, and the fourth is of uncertain origin.
5.5.2
The building material comprises one fragment of hard chalk incorporated in

road metalling 302, and one small fragment from a roof or floor tile in a
micaceous fine sandstone from ditch 313. The chalk could have come from
Upper Cretaceous chalk quarries to the south of the Site, or from the shores
of the Wash to the west (Allen and Fulford 1999), while the micaceous
sandstone derives from Triassic or carboniferous sandstone, from
Nottinghamshire or the Pennines.
5.5.3
The potable object is a crudely made rectangular die, measuring 25 x 20 x

15mm, made of Upper Cretaceous chalk, possibly a re-used tessera, and
found in layer 102. The die does not follow the conventional marking, with
opposing faces totalling 7, but is marked with rough point indentations,
apparently far more randomly: 3/3, 5/?6 (damaged face), and 9/incised
cross. No parallels for this marking have been found. Although dice made
from reused tesserae are recorded in Romano-British contexts elsewhere,
for example Dorchester, Dorset (Woodward 1993, fig. 104, 26), this object is
very crude for a Romano-British die, and is more likely to be medieval in
date (I. Finkel pers. comm.; I. Riddler pers. comm.).
5.5.4
The fourth object is a small fragment from Trench 2 topsoil with one flat,
smooth surface, but with no incontrovertible signs of working.
5.5.5
There is also a single prehistoric flint waste flake, recovered as a residual

find from ditch 503.
5.6
Glass
5.6.1
Of the five pieces of glass recovered, two are of Romano-British date. The
more diagnostic of these is a small bowl rim in pale greenish glass from
layer 102. An undiagnostic small body fragment in pale blue glass came
from the topsoil in Trench 4.
5.6.2
Two fragments are certainly post-Roman. These comprise a very heavily

degraded and oxidised vessel fragment, possibly of late medieval or early
post-medieval date; and a post-medieval window glass fragment, both from
Trench 4 topsoil.
5.6.3
The fifth fragment, in clear glass, from Trench 4 topsoil, is completely

undiagnostic, and could well be of post-medieval date.
5.7
Metalworking debris
5.7.1
A very small quantity of slag was recovered, from Trenches 1 and 3, deriving
from iron smithing. This includes a possible hearth bottom from ditch 312.
Quantities are insufficient to postulate on-site metalworking, although this
would not be unexpected.
5.8
Coins
5.8.1
Seventy-nine Roman coins were recovered from the excavations at

Brancaster. All of these are Roman in date, the vast majority comprising
22

antoniniani and nummi of the late 3rd and 4th centuries AD (see Appendix
2, Table 16). All of the coins are small denomination copper alloy issues. In
general these are in fair condition, although a few show evidence for postdepositional of corrosion, whilst many also show signs of pre-depositional
wear. Despite this, the majority could be identified to period. Only 14 of the
79 coins could not be closely dated, although the size and shape of the flans
suggest that these date to the 3rd or 4th centuries AD.
5.8.2
Sixty-five of the coins from the Site could be dated to period, using the 21
periods described by Reece (1991) for the analysis of coin assemblages.
The breakdown of these coins by period can be seen in Appendix 2, Graph
1.
5.8.3
The earliest coin from the site is a worn dupondius of the Emperor Caracalla
(AD 180192). A further eight coins are radiate copies of late 3rd century
antoniniani, the majority probably minted between AD 270 and 296. These
radiate copies were copies of official coinage, possibly struck to
compensate for gaps in supply of coinage to Britain and to supply sufficient
small change for the provinces needs. It is unclear whether these copies
were officially sanctioned, if at all, but they are not uncommon as site finds,
and seem to have circulated in the same fashion as officially struck coins.
5.8.4
The majority of the coins from the site date to the first half of the 4th century
AD, with a sizeable peak of coin loss in period 17. Once again, more than
half of these coins (23 of the 43) were copies or probable copies. Much
smaller numbers of period 18 and 19 coins were present, although the six
coins of period 21 indicate that coin use continued on the site into the late
4th century, and perhaps into the 5th.
5.8.5
5.8.6
Coins from layer 125

Some 42 of the coins from the site were recovered from a single deposit
layer 125. This was a final use/disuse deposit recorded in one of the small
rooms within the principia building. Only a small portion of this deposit, so
the recovery of so many coins is worthy of comment. Looking at these coins
by period (see Appendix 2, Graph 2), it is clear that the vast majority of
those that could be dated (some 30 coins in all) date to the AD 330s and
340s (period 17). This clearly suggests that the deposit represents a
scattered hoard. The latest coin from this deposit is an official Fel Temp
Reparatio Fallen Horseman issue of period 18, probably minted between
AD 350 and 355. This suggests that the hoard was probably deposited in the
AD 350s. Some caution should be exercised here, as only a portion of the
hoard was recovered much of the deposit from which they were recovered
was left in situ.
Other coins from the site
Inevitably, the presence of a hoard within a site assemblage skews the
pattern of normal coin loss. This can clearly be seen by looking at the
remaining coins from the site by period (Appendix 2, Graph 3). With the
hoard removed, it is clear that the assemblage is no longer so dominated by
period 17 coins. The peaks of coin loss in the late 3rd century (period 14)
and period 17 are common in late Roman assemblages, and reflect the
vagaries of the supply of coinage to Britain from the Imperial mints, most of
which were based on the continent.
23

5.8.7
The single late 2nd century coin from the site may well have been in
circulation for some considerable period of time before its loss. Earlier
excavations and fieldwork on the site (Hinchliffe and Sparey Green 1985)
recovered a number of coins spanning the four centuries of Roman rule.
Work on the combined assemblage from the earlier work suggested that
coin loss on the site began in the middle of the 2nd century AD (Sparey
Green and Gregory 1985, 191), and continued until the end of the 4th
century AD, although with a marked decline in coin use after period 17 (from
AD 348 onwards). There were also apparent differences between the
patterns of coin loss from within the fort to that recorded outside the fort,
although the sample sizes were not equal, with significantly fewer coins
recorded from within the fort. To a certain extent, the overall pattern of coin
loss from the current evaluation matches that from the earlier excavations,
with the only significant change being the number of Theodosian (period 21)
coins recovered from the evaluation. Only one of these was recorded from
the previous excavations, from the settlement outside the camp itself. All six
of those recorded from the recent work were recovered from unstratified
spoil from Trenches 1 and 2, within the fort. They clearly indicate that there
was activity in the fort at the end of the 4th century, and probably into the
5th. Beyond this, the small size of the assemblage (13 coins) recovered from
Trench 3, the only trench outside the defences, makes any meaningful
discussion of the distribution of the coins recovered impossible.
5.9
Metalwork
5.9.1
Apart from coins, the metalwork includes objects of copper alloy, lead and
lead alloy, and iron.
5.9.2
All metal items apart from the coins have been X-radiographed as a basic
record, and also to aid identification. The iron objects in particular are badly
corroded.
5.9.3
Copper alloy
The copper alloy objects fall into four functional groups: personal items,
writing implements, military items, and miscellaneous objects. There are
also 14 very small fragments of copper alloy sheet, from various contexts,
some with rivets in situ, but of unknown function.
5.9.4
Amongst the personal items are a brooch pin from layer 122; part of an
armlet made from two twisted strands (layer 106); a finger ring formed from
a narrow plain strip bent into a circle (layer 102); and a pair of toilet
implements held on a suspension ring (layer 108), comprising a doublelooped object and a shank, probably from an ear scoop, but with the end
missing. A shank from Trench 1 topsoil may belong to another toilet
implement; it seems too thick for a hairpin. All these objects are of well
documented Romano-British types.
5.9.5
An almost complete pin (tip missing) from layer 102, however, is of more
ambiguous date. The pin has a faceted cuboid head: the type is known from
Romano-British contexts, for example in Colchester (Crummy 1983, fig. 29),
but also from the mid Saxon period, for example in Hamwic (Southampton)
(Hinton 1996, fig. 9).
24

5.9.6
Two personal items are certainly of post-Roman date. These comprise a

small, D-shaped buckle, lacking the pin, and a modern button (both from
Trench 3 topsoil).
5.9.7
One stylus was recovered, from cleaning layer 103 in Trench 1. The shank
is bent, and the pointed end is missing; the opposite end is expanded and
subrectangular.
5.9.8
Perhaps of most interest from the Site, however, is a section of lorica

squamata (scale armour), found on mortar surface 132, next to wall 130
(Figure 5, Plate 4). This was made from small subrectangular scales
attached to a fabric backing. It is typically seen on depictions of standard
bearers, centurions, cavalry troops and even auxiliary infantry, as well as
regular legionaries. The X-radiograph clearly shows groups of overlapping
scales (measuring 16mm x 12mm) fastened together with wire; each scale
has three paired perforations for attachment, one at each side (for
attachment to each other) and one at the top (for attachment to the backing).
The lower corners of each scale are rounded.
5.9.9
A small stud with a solid domed head (diameter 10mm, length 15mm) from
Trench 1 topsoil is presumed to be of Romano-British date, but is not
particularly chronologically distinctive.
5.9.10
Lead and Lead Alloy

Much of the lead comprises waste pieces, or small undiagnostic fragments
(16 objects). Identifiable objects include two weights. One of these, from
cleaning layer 103, is ovoid, with a transverse perforation through one end; it
weighs 36.6g. The second, from Trench 2 topsoil, is bun-shaped, and
weighs 162g (although including some iron corrosion, possibly from a
suspension loop). Three possible pot mends were recovered, all from topsoil
contexts.
5.9.11
The most diagnostic object is a pewter platter from layer 102. This has been
damaged, and is in four pieces, with some distortion to the rim. The platter is
almost certainly of late Romano-British date datable pewter finds in Britain
are nearly all from later 3rd of 4th century contexts, and reflect the revival of
Roman interest in the Cornish tin mines around the middle of the 3rd
century, leading to extensive pewter production in Britain (Barker and
Hatcher 1974, 9). Production seems to have been widespread: moulds have
been found in several locations, including East Anglia, although there is very
little regional variation in forms.
5.9.12
A possible post-medieval impacted musket shot came from Trench 5 topsoil.
5.9.13
5.9.14
Iron
The majority of the ironwork consists of nails of various sizes (99 examples).
Other structural items comprise three strip fittings. Neither the nails nor the
structural items are chronologically distinctive; most if not all are assumed to
be Romano-British, although a high proportion (approximately two-thirds)
came from topsoil contexts.
Other identifiable objects fall into four functional categories: personal items;
knives and other tools; household equipment; writing implements. Personal
25

items are limited to three hobnails from Trench 3 topsoil, probably from
footwear.
5.9.15
There are three knives. The most complete came from layer 102, and falls
into Mannings type 24, the back of the blade and tang having an S-shaped
profile. This knife type was originally an Iron Age form, and continued to be
used into the early Romano-British period (Manning 1985, fig. 29, 118-9). A
second knife came from the topsoil in Trench 1, and appears to belong to
Mannings type 16, the tang lying on the midline of the blade; it seems to
have been a long-lived form (ibid., fig. 28, 116). The third knife, from 315,
survives as a triangular blade only, lacking the tang: Mannings type 12,
another long-lived type, would accommodate this example (ibid., fig. 28,
114).
5.9.16
A tapering bar from Trench 1 topsoil seems likely to be a tool of some kind,
perhaps a chisel or punch. This is not a chronologically distinctive tool type,
and cannot be definitively dated as Romano-British.
5.9.17
A small key was found in Trench 4 topsoil. This is a lever-lock key, the most
advanced key type used during the Romano-British period, operating on the
same principles as the modern lever-lock. This example has a rectangular
handle with circular bow, and a simple bit.
5.9.18
Two styli were identified (demolition debris 206, Trench 4 topsoil), as well as
the possible shanks of two others (both from layer 102). The example from
Trench 4 is of Mannings type 1, tapering to a point at one end and flattened
into a small eraser at the other; this is a common type (Manning 1985, 85,
fig. 24). The stylus from 206 belongs to Mannings type 3, with eraser and
point clearly separated from the stem, the eraser having concave sides
(ibid., 85, fig. 24). The two shanks cannot be attributed to specific forms, and
are not definitively identified due to lack of distinctive features.
5.10
Worked Bone
5.10.1
A single object of worked bone was recovered, from layer 102. This
comprises a strip with concave sides and ends (150 x 23mm). The object is
neatly made, with the upper surface and edges polished; it is of uncertain
function, although it could have been used as an inlay.
5.11
Human Bone
5.11.1
Redeposited human bone was recovered from five contexts in three

trenches, two within the confines of the fort (Trenches 2 and 4) and one from
the potential early fort to the north (Trench 5). The bone was subject to a
rapid scan to establish the minimum number of individuals (MNI) and other
demographic data.
5.11.2
The bone is generally in good condition. Complete or near complete skeletal

elements are represented with limited fragmentation.
5.11.3
Parts of a minimum of two neonates (0-2 weeks) were recovered (Trenches

2 and 4), and an adult (> 18 yr) of indeterminate sex (Trench 5).
5.11.4
The neonatal remains were recovered from the topsoil (201 and 401), madeground (210) laid to facilitate construction of the fort, and the fill of ditch 506
26

believed to be part of the defences associated with the earlier fort.The adult
tibia was recovered from the fill (402) of a robber trench cut through the
backfill of the defensive ditch 406.
5.11.5
The exclusion of neonates from cemeteries and the recovery of their

remains in association with domestic buildings and properties is a common
feature in the Roman period (Philpott 1991, 97-102; Mays 1993; Scott 1999,
115; Struck 1993). The factors affecting both the death and place of burial of
such young individuals may be numerous and have been subject to both
general and site-specific discussion elsewhere (Mays 1993; McKinley 2011;
Philpott 1991, 101; Riddle 1997, 85-6; Scott 1999, 30-32, 70 and 115- 118).
As the original place of burial of the neonates at Brancaster is unknown
limited comment can be made regarding their deposition. However, the
condition of the bone and presence of numerous skeletal elements from
different areas (upper and lower limb and axial skeleton) suggests the
remains have not moved far from their original place of deposition. The
presence of the adult tibia is less easily accounted for in the absence of
precise dating. It may have derived from a grave within the vicinity of the fort
displaced by its construction or, following a break-down in
organisation/abandonment of the fort, parts of the area may have served a
temporary mortuary function (potentially sub-Roman in date), burials made
there being disturbed and redeposited before the ditch silted up.
5.12
Animal Bone
5.12.1
A total of 1,322 fragments (or 22.714kg) of animal bone was recovered from
the site during the normal course of excavation. Once conjoins are taken
into account this figure falls to 1,149 fragments. The majority of the bone
comes from Roman layers in Trenches 1 to 5, very little was recovered from
cut features.
5.12.2
5.12.3
5.12.4
Methods of assessment
The following information was recorded where applicable: species, skeletal
element, preservation condition, fusion and tooth ageing data, butchery
marks, metrical data, gnawing, burning, surface condition, pathology and
non-metric traits. This information was directly recorded into a relational
database (in MS Access) and cross-referenced with relevant contextual
information.
Preservation condition
Bone preservation is on the whole quite good and only a small number of
fragments show signs of weathering and abrasion. This suggests that soil
conditions are favourable for the preservation of bone and that contexts
containing bone have not been significantly disturbed and redeposited. The
number of fragments displaying signs of gnawing is also relatively low (c.
4%), which suggests that the majority of bones were rapidly buried out of the
reach of scavenging carnivores. Indeed some of the bones were still in
articulation when uncovered, for example the pork joint from occupation
deposit 125 (animal bone group 82).
The assemblage
Bone was recovered from 41 separate contexts, mostly layers resulting from
demolition and levelling. The following species have been identified and are
listed in order of their relative abundance: cattle (40%), sheep/goat (33%),
27

pig (12%), bird (includes domestic fowl, duck and corvid), horse, dog, deer,
cat and edible crab. A few intrusive rodent bones were also recovered.
5.12.5
Species frequencies are similar to those reported from earlier excavations at

the fort (Jones et al. 1985, 135 and 137), and this reflects the importance of
cattle to the Romano-British economy and the dietary preferences of the
solders stationed at the fort (see King 1984, 198; 1991, 17; Dobney 2001,
36-7; Davies 1971).
5.12.6
All parts of the beef, mutton and pork carcass are represented in the
assemblage and this suggests that the fort was supplied with livestock on
the hoof probably from the immediate rural hinterland (see for example
(Davies 1971, 127; Thomas and Stallibrass 2008, 9; Thomas 2008, 38 and
44-5). Cattle bones are more extensively butchered than the bones of other
species, more so than might be considered necessary to reduce a large
carcass into manageable portions. This could indicate that cattle bones were
reused in stews or to flavour soups, the type of dishes that are relatively
quick and easy to prepare for large groups, and which stretch available food
resources.
5.12.7
A distinct pattern of butchery marks was observed on a number of cattle

scapulae. This takes the form of trimming around the glenoid cavity, removal
of spine and nick marks on the margo thoracic border. A few of the scapulae
also have hook damage to the blade. This type of butchery is typically
Roman and has been recorded on cattle scapulae from a large number of
sites in Britain. The marks result from the preparation of shoulders of beef
for curing, most probably by using the technique of cold-smoking (i.e.
immersion in brine; see Dobney et al. 1996, 24-7), a process that preserves
meat for long-term storage and is thought to have evolved in response to
military food requirements (Grant 1987; Maltby 1989). Cattle scapulae
recovered from previous excavations at the fort also show this distinct
pattern of marks (see Jones 1985, 130; Jones et al. 1985, 144).
5.12.8
Additional sources of meat include domestic fowl and duck, and based upon
the butchery evidence, it would seem that horse was also eaten. There is
limited evidence for the exploitation of marine resources in the form of both
fish and crab.
5.12.9
Two red deer bones were recovered; these include a lower molar and a
piece of antler. The latter is a single tine, which shows signs of use wear on
the very tip. It is unclear what type of implement this piece might have
broken-off from, antler tools such as picks and rakes are known from the
prehistoric period in Britain, however these tools are rather rudimentary in
comparison to the wooden and metal digging tools available during the
Romano-British period.
5.12.10 The assemblage also includes a small number of dog and cat bones, and
based on the demographics it would seem that a breeding population was
present. These animals are like to have had a semi-feral existence.
5.12.11 Also present are a number of crow and raven bones, these species might
have been attracted to the site by the opportunity to scavenge on midden
material.
28

Conclusions
5.12.12 Assessment of the animal bone assemblage from Brancaster Fort indicates
that the military diet was primarily based on the consumption of beef, and
too a lesser degree mutton and pork. These animals appear to have been
supplied on the hoof, although there is also some indication that cured
shoulders of beef were available. Domestic poultry, fish and crab provided
some dietary variety, and horsemeat also eaten but perhaps only in times of
severe hardship.
5.13
Marine Shell
5.13.1
The marine shell assemblage consisted of 519 shells, representing 329

minimum number of individuals. These were retrieved from 31 deposits of
Romano-British date in five trenches. Where deposits were encountered
which contained a large proportion of marine shell, a representative sample
was collected of sufficient size for analysis.
5.13.2
All the shell has been recorded by species and by context, with the oyster
shell being sub-divided into left and right measurable and unmeasurable
valves. The results can be seen in Table 17 (Appendix 2).
5.13.3
The predominant species of the assemblage was oyster (Ostrea edulis),

forming 90% of the minimum number of individuals. Mussels (Mytilus edulis)
represented 5% and whelks (Buccinum undatum) 3.6% of the minimum
number of individuals. The remaining 1.4% of the assemblage was
comprised of cockles (Cerastoderma edule) and periwinkles (Littorina
littorea).
5.13.4
Although the marine shell was retrieved from five trenches on the site, 36%
of the assemblage was recovered from Trench 1, 29% from Trench 4, 19%
from Trench 2, 15% from Trench 3 and only 0.6% from Trench 5. No
indication of areas of preparation or consumption could be discerned, as
there were no significant differences between the spread of the 203 right
oyster valves and the 273 left oyster valves in the five trenches.
5.13.5
The oyster shell was analysed in more depth from five deposits of RomanoBritish date, two from Trench 1 (103) and (108), one from Trench 2 (202),
one from Trench 3 (304) and one from Trench 4 (401). The oysters were
subdivided into measurable and unmeasurable left and right valves. 70% of
the shells from the selected deposits were measurable. The measurable
valves were then measured and examined, both for traces of infestation and
physical characteristics. A summary of these results can be seen in Table
18 (Appendix 2).
5.13.6
The analysed oyster shells were generally large, with the majority of the
shells having a maximum width and length of between 60 and 100 mm, as
can be seen in Table 19 of comparative shell sizes and the graphs of shell
size distributions (Appendix 2). The shells were generally slightly elongated,
indicative of softer substrates. Although the shells were generally large, over
a quarter of the shells (26.7%) were misshapen and 10% of the shells had
oysters attached. This may be an indication of competition for space in a
less well managed oyster bed as high levels of both irregularity of shape and
clumping of shells are indicative of natural oyster beds where there is
competition for space. The general lack of other small marine shells together
29

with paucity of small oysters may be indicative of a managed natural oyster

bed being fished with a dredge net of a fixed size, rather than a specially laid
oyster bed. It may also be an indication of some form of selection before
they were brought to site.
5.13.7
Traces of infestation, mainly only in small amounts, was seen on 59% of the
analysed shells. This was mainly traces caused by the polychaetic worm
Polydora ciliata, observed on 50.7%, with some evidence of the boring
sponge Cliona celata on 11.5% of the shells. A small number of shells (4%)
showing traces left by barnacles were recorded from Trenches 2, 3 and 4
and there were calcareous tubes, made by marine worms of the Serpulidae
family such as Pomatoceros triqueter and Hydroides norvegica, on two
shells from Trench 4. Polydora ciliata is widespread and is most prevalent
on hard, sandy or clay grounds particularly in warm shallow water, while the
boring form of Cliona celata is also widespread on a variety of coasts.
5.13.8
The shells were generally in fairly good condition with 26% being worn and
7.8% flaky. Taken in conjunction with the fact 70% of the shells were
measurable, this may be indicative of a relatively fast rate of deposition of
the shells.
5.13.9
There were notches and traces of opening on 34% of the shells. There was
a square hole in the centre of a large right valve from Trench 2 (202), and a
smaller right valve from Trench 4 (topsoil 401), which were not the result of
predators. Such complete holes in the centre of the shells have been
observed elsewhere from Roman contexts, such as Westhampnett, West
Sussex (Wyles 2008) and Tolpuddle, Dorset (Winder 1999), and are thought
to be possibly a result of fork tines or even the deliberate perforation of large
shells for use as temporary roof tiles. It is possible that some shells were
used as temporary labels on barrels or even as decoration/advertising on
oyster stalls. Georgius Agricola refers to sheds usually named from some
animal or other thing which is pictured on a tablet nailed to it (Agricola 1556,
modern translation 1950). It may be that oyster shells themselves were used
in a similar way to distinguish between different barrels or buildings, using a
pictorial or representational form of labelling rather than writing for what
would have been largely a non-literate society.
Conclusion
5.13.10 The marine shell assemblage only represents an augmentation and variety
of the basic diet rather than forming a significant part of the diet.
5.13.11 It is likely that the shells came from a managed natural oyster bed on a soft
substrate on the East coast.
6
PALAEO-ENVIRONMENTAL SUMMARY
6.1
Introduction
6.1.1
Three bulk samples were taken: from a possible occupation layer (125)
associated with Romano-British coins and possibly located within the strongroom in the headquarters building; from the late Romano-British or postRoman demolition layer (106) sealing many of the structures in Trench 1;
and from a charcoal lens (415) within the defensive rampart excavated in
30

Trench 4. The samples were processed for the recovery and assessment of
charred plant remains and charcoal.
6.1.2
The bulk samples were processed by standard flotation methods; the flot
retained on a 0.5 mm mesh, residues fractionated into 5.6 mm, 2mm and
1mm fractions and dried. The coarse fractions (>5.6 mm) were sorted,
weighed and discarded. Flots were scanned under a x10 x40 stereobinocular microscope and the preservation and nature of the charred plant
and wood charcoal remains recorded in Table 20, Appendix 3. Preliminary
identifications of dominant or important taxa are noted below, following the
nomenclature of Stace (1997) for wild plants, and traditional nomenclature,
as provided by Zohary and Hopf (2000, tables 3 and 5), for cereals.
6.1.3
The flots varied in size with low numbers of roots and modern seeds that
may indicative of stratigraphic movement and the possibility of
contamination by later intrusive elements. Charred material comprised
varying degrees of preservation.
6.2
Charred and mineralised plant remains
6.2.1
Charred remains of cereals were quite low in the sample from the
occupation layer 125, but occasional grains and glumes of hulled wheat
(Triticum dicoccum/spelta) were present along with charred seeds of hedge
parsley (Torilis sp.), annual meadow grass (Poa sp.), vetch/wild pea (Vicia
Lathyrus sp.), buttercup (Ranunculus sp.) and ribwort plantain (Plantago
lanceolata). There were also some charred stems, and thorns of
sloe/hawthorn (Prunus spinosa/Crataegus monogyna).
6.2.2
This same sample also produced a large number of mineralised seeds,

mainly of more wasteland and arable species, rather than cultigens or
domesticated species. The seeds present included corncockle (Agrostemma
githago), dock (Rumex sp.), hedge-parsley, knotgrass (Polygonum
aviculare), wild mignonette (Reseda lutea), knapweed (Centaurea sp.),
perennial rye grass (Lolium sp.) and brome grass (Bromus sp.). More
unusual was a probable seed of Spergularia cf. marina, a species of sandy
and muddy coastal areas. Seeds of orache (Atriplex sp.) were also quite
common and can be found in similar coastal situations, or within wasteland
and arable areas. Also present were a few seeds of forget-me-not (Myosotis
sp.), given the general ecologies of the other species in the assemblage and
the modern distribution of species of Myosotis, early forget-me-not (Myosotis
ramosissima) a plant of dune slacks, sandy wastelands, arable fields and
calcareous soils, would seem the most probable candidate.
6.2.3
The later sample that seals this earlier deposit from demolition layer 106 had
stems of heather (Ericaceae), along with a few grains of hulled wheat and
several more of barley (Hordeum vulgare). A few glume bases including
some of spelt (Triticum spelta) were also seen in the sample. A small range
of charred seeds, predominately of larger seeded species including black
bindweed (Fallopia convolvulus), fumitory (Fumaria sp.), cleavers (Galium
aparine), oats (Avena sp.) and brome grass (Bromus sp.) were also present.
Smaller seeds included a few of sedges (Carex sp.) and heath grass
(Danthonia sp.). Also present were several fragments of hazelnut (Corylus
avellana) shell and a fragment of a stone of sloe (Prunus spinosa).
31

6.2.4
The sample from layer 415 contained several grains of barley and many
charred capsules of runch (Raphanus raphanistrum). The sample also
contained a few other weed seeds although these were few in number and
included those already recorded from the other samples, including cleavers,
black bindweed, dock, goosefoot and orache.
6.2.5
The samples indicate the presence and potentially the processing of cereals,
although glume bases, which were rarer than grain, might be expected in
some quantity if larger amounts of processing waste were present. It is
possible that such processing was carried out outside the fort, or, as noted
at South Shields and potentially also at Bremetenacum, Ribchester (van der
Veen 1992; Huntley 2000; Fuller and Stevens 2009), that grain arrived and
was stored in such forts after dehusking. The dominance of seeds or
capsules of larger seeded species would certainly suggest that crops arrived
at the fort in a relatively processed state.
6.2.6
The range of species present are not generally indicative of the cultivation of
any particular types of soil and, as such, could all come from locally grown
crops. In particular runch can become very dominant on sandier, usually
non-calcareous soils, which formed on the Pleistocene deposits just to the
south of Brancaster.
6.2.7
The charred remains of hazelnut and sloe are in general keeping with the
probable collection and use of local wild food resources. The remains of
heather, and stems, including seeds of heath (Danthonia decumbens) and
probably sedge, are all likely to come from heathland vegetation that lies to
the south of Brancaster, again growing upon the Pleistocene deposits that
can still today be associated with heathland to the south of the site. Such
remains are common on sites in the north where they have been associated
with the use of heathland turves for fuel (see Hall and Huntley 2007).
6.2.8
The mineralised seeds from the occupation layer 125 might result from cess
although no cess type material was seen. The high presence of fish bones
and fish bones in general, can introduce phosphates and together with the
calcareous nature of these deposits, can explain the resulting calcium
phosphate and mineralisation. None of the identified seeds present relate to
edible species and all are more typical of disturbed, wasteland settlement
soils with some coastal elements.
6.3
Wood Charcoal
6.3.1
Wood charcoal was noted from the flots of the bulk samples and is recorded
in Table 20 (Appendix 3). Generally very little wood charcoal was
recovered within the samples, although small round wood was identified
from layer 125. The absence of wood charcoal may be a result of
preservation, although equally it might reflect the regular use of other
material for fuel; e.g. possibility turves and heather.
6.4
Land and aquatic molluscs and marine shells
6.4.1
The flots of the bulk samples were rapidly assessed by scanning under a
x10 x40 stereo-binocular microscope to provide some information about
shell preservation and species representation. The numbers of shells and
the presence of taxonomic groups were quantified (Table 21, Appendix 3).
Nomenclature is according to Anderson (2005) and habitat preferences
32

according to Kerney (1999). The presence of these shells may aid in broadly
characterising the nature of the wider landscape.
6.4.2
The mollusc assemblages observed within layers 125 and 106 in Trench 1
were indicative of the presence of a number of different habitats in the
vicinity of the site. The local environment may have been one of an area of
open grassland as shown by the general range of open country species,
intermediate species and shade-loving species in the assemblages. A
variety of other habitats in the vicinity is indicated by the mollusc
assemblage from layer 125. These include more marshy areas and freshwater flooding, as shown by the presence of Succinea/Oxyloma sp. and
Anisus leucostoma, and brackish environments, such as estuaries or salt
marshes, as indicated by the occurrence of Myosotella myosotis and
Hydrobia spp.. The possible identification of Cochlicella acuta is noteworthy.
This species in Britain is almost exclusively maritime, inhabiting cliffs, sand
hills and waste ground and rarely straying more than a mile or two inland
(Kerney 1999, 186 citing Aubertin et al. 1931). Although this species on the
South and West coasts of Britain, it has been recorded in Norfolk at Wellsnext-the-sea.
6.4.3
The small assemblage from charcoal layer 415 is indicative of an open

environment.
6.4.4
A few marine shell fragments were also observed in the samples from
Trench 1. These included fragments of periwinkles, oysters, mussels and
limpets.
6.5
Small animal and fish bones
6.5.1
During the processing of bulk soil samples for the recovery of charred plant
remains and charcoals, small animal bones were noted, and their presence
recorded in the flots (Table 20, Appendix 3). These included those of
birds/small mammals, anurans (frogs, toads)/fish. The sample from the
strong-room (125) contained a number of small and small-medium mammal
bones including a rodent jaw bone. Fish bone was also present which
included a vertebra of eel (Anguilla anguilla), and a number of smaller fish
vertebra, and other bones, including otoliths.
6.5.2
The other two samples had less of such material with only a few mammal
bones available from demolition layer 106.
6.6
Foraminifera
6.6.1
Tests of foraminifera, including Quinqueloculina and Elphidium sp. were

recorded within the sample from occupation layer 125. These remains were
not seen within the remaining samples.
DISCUSSION
7.1
Introduction
7.1.1
Though small scale this evaluation has effectively highlighted the well
preserved remains and structures within Brancaster fort through both
geophysical survey and intrusive investigations. The eastern vicus was also
investigated and though many of the cropmark features were identifiable
from the magnetometer survey, excavation showed some truncation of the
33

remains. Such truncation was also noted during the 1974 and 1977
excavations of the western vicus (Hinchliffe 1985, 176).
7.2
Evidence for an earlier fort (Trench 5 and Trench 3)
7.2.1
Trench 5 was positioned to target the double-ditched enclosure thought to

represent an earlier fort. The location of the enclosure would be consistent
with establishing an early position to guard the river navigation, and the
shifting course of the river and mudflats may have necessitated its
relocation. Generally little occupational debris was recovered from Trench 5
and, even given its position on the defensive margins, when contrasted with
Trench 4 it suggests less intensive occupation - it may have been more a
camp than a fort. No evidence was found for any upstanding earthwork
defences.
7.2.2
No clear difference in dating could be determined through the pottery,

though it seems doubtful that this structure was established very long before
work began at the main fort. Most probably this fort would have been
abandoned once the main fort was established, though it may have served
an auxiliary function. Evidence from Trench 5 suggests that the ditches may
have been deliberately backfilled once they went out of use.
7.2.3
A large enclosure ditch (312/313) was excavated in Trench 3. The position

of this ditch in conjunction with the magnetometer survey and cropmark
evidence suggests the position of a large rectangular enclosure on a similar
alignment to the early fort. The magnetometer survey indicates that this
may have been another double-ditched feature - could this also be another
earlier fort? The presence of such a feature, pre-dating the vicus, could
explain the general alignment of the vicus, which is at odds to the main fort.
7.3
The eastern vicus (Trench 3)
7.3.1
The 1977 excavations indicated that the western vicus was established in
the 2nd century AD, with continued occupation throughout the 3rd century
(Hinchliffe 1985, 178). While the 1974 excavation areas slightly to the west
suggested the presence of stock enclosures dating to the late 3rd and 4th
centuries AD (Sparey Green 1985, 13).
7.3.2
The evidence from Trench 3 seems to indicate that both settlements were
contemporaneous, with largely 2nd and 3rd century material recovered from
the trench. The exception was the road surface itself, stratigraphically later,
and from which pottery suggests a 3rd to 4th century date. It also overlay
some of the earlier features. As this represents the earliest surviving
metalling phase of the road, this seems to suggest that the road network
within the vicus may not have been laid out or formalised until this later
period. This would imply continued and perhaps more significant settlement
into the 4th century AD. On current evidence, this would have coincided with
the abandonment of the western vicus and the decline and possible
abandonment of the fort, suggesting the emergence of the eastern vicus as
an entirely civilian settlement.
7.3.3
While the vast majority of the features seen in the magnetometer survey and
identified through cropmark evidence follow a broad west-north-west eastsouth-east alignment, some possible features can be seen on a divergent
orientation, most obviously, a north-west south-east aligned linear feature
34

to the east of Trench 3. However, without excavation it is impossible to tell

whether this represents earlier or later activity, though a break in continuity
is suggested.
7.4
The main fort
7.4.1
Roman forts generally conform to a playing card shape with gates in the
centre of the four sides (portae principales) (Bidwell 1997, 28). A main road
(via principalis) forms the main axis of the fort with the headquarters
(principia) fronts this with another extending to the front gate from this (via
praetorian) and another behind (via decumana). Other standard buildings
within the interior are barrack blocks, the commanders house (praetorium)
and granaries, the latter two usually situated either side of the principia.
7.4.2
Some variation from this typical layout design can be seen at Brancaster.
This may reflect the date of construction - some changes in layout are seen
from the early 3rd century (Bidwell 1997) - or differences in function. At
Reculver, thought to have been contemporaneous with Brancaster, the
barracks also appear to lie to the east of the principia, while to the west lay
granaries (Philp 2005); unfortunately the northern half of Reculver has been
reclaimed by coastal erosion, so the parallels cannot be further pursued, but
in the north-eastern part of the fort lay a bath house and officers quarters
(ibid.).
7.4.3
Perhaps most anomalous feature of Brancaster is the west-north-west

east-south-east aligned building in the north-eastern are of the fort. This
shares a common orientation with the postulated earlier forts to the north
and west and its walls appear to extend substantially deeper than the
buildings fronting the via principalis to the south. As there is evidence in
Trench 2 that the ground may have been built up before the construction of
the more conventionally west east orientated building, this difference in
construction depth may represent a change in the ground surface rather
than different foundation depths. There is a suggestion that this complex of
rooms may be a mansio, or official accommodation. The presence of a
hypocaust system in the complex to the north suggests a detached bath
house though this may not be contemporary.
7.4.4
A large barrack block was identified from the geophysical survey to the east
of the principia, and a granary in the north-western part of the fort. Further
buildings are suggested to the east of the principia, though these are less
distinct and within the northern part of the fort.
7.4.5
An unusual oval footprint underlain by a rectangular structure and within a

wider rectangular structure was identified during the GPR survey. The report
(GSB 2014) speculates that the oval response, which contracts with depth,
could be demolition period. The possibility of a raked structure such as
seating should perhaps also be considered, although the location and size
would be unusual for an amphitheatre. A building showing an oval within a
rectangular footprint is known from Chester legionary fortress (Deva Victrix)
where is has been interpreted as a temple.
7.4.6
A number of possible structures are visible in the northern and western parts
of the fort along the interior of the ramparts, a situation which typically
reflects later development and, where investigated by St. Joseph, proved to
be late 4th century in date (Edwards and Green 1977, 25). The evidence
35

from the magnetometer survey (Figure 2) suggests more industrial activity

may have been concentrated within these areas.
7.4.7
What is not clear from the available evidence is the location of praetorium.
While the nature of the structures to the west of the principia is uncertain
there are no responses in the geophysical survey that suggest the kind of
stone built structure you would expect. The postulated mansio could also
equally be the praetorium, though the different orientation does suggest
either an earlier or later date than the main fort layout. It is even possible
that the building changed function as the fort developed with the earlier
mansio being commandeered for military use.
7.4.8
Earlier excavations confirmed the position of the corner towers and the
present investigations have confirmed the earlier findings of a wide stonebuilt wall with an internal rampart behind and at least one external ditch
(Trench 4). Analysis of the stone recovered from the excavations confirms
the evidence from earlier studies (Allen and Fulford 1999; Allen et al. 2001)
that the construction materials from the fort were sourced relatively locally
and probably utilising the coastal access.
7.4.9
The charcoal layer with the rampart (415) may correspond to that noted
during the 1935 excavations, where the removal of the facing stones of the
defensive wall was also noted (St Joseph 1936, 447). Significant amounts of
charcoal were also noted by the Reverend Lee Warner during his
excavations of the north-east corner tower (Lee Warner 1851, 14-16) and
this could perhaps have resulted from a widespread fire amongst the
defences.
7.4.10
The earliest stratigraphic deposits within Trench 1 were not reached during
excavation, but the pottery suggests activity from the 2nd into the 4th
century AD. Evidence of robbing and infilling of the rooms of the principia
appears to date to the 4th century.
7.4.11
The central room of the office block of the principia can be seen on the GPR
survey to extend to a considerable depth below the surrounding rooms, with
responses still be received at 2.2m below ground level. This basement room
was in all likelihood the location of the fort strong-room. A similar feature
was uncovered during the excavations at Reculver, where it was thought to
have been overlain by the garrison shrine (sacellum) (Philp 2005, 43-46).
7.4.12
The position of wall 205 indicates that it was mostly probably the north wall
of a long building fronting the via principalis. However, excavation showed
that this was not a substantial structure. Such a shallow foundation may
have formed the support for a timber superstructure rather than a stone
building. There were no clear indications of the nature or purpose of this
building, though a storeroom or stable is perhaps most likely. Dating
evidence from Trench 2 indicates activity within the overall range of mid 2nd
to 3rd centuries AD.
7.4.13
In both Trenches 1 and 2 there was evidence of later, seemingly more

rudimentary structures, but it was not clear whether this presented military or
civilian occupation. This activity could not be clearly dated but seems to
have post-dated some demolition of structures within the fort.
36

7.4.14
The pottery is largely derived from the local region though there are some
continental imports (Rhineland and central Gaul). Rather than being a point
of entry for foreign imports, however, the fort is more likely to have been
supplied from internal trade routes. The prevalence of locally derived Nene
Valley and Oxford wares is similar to the pattern seen within the western
vicus excavations, though here Colchester-derived wares were also seen
(see Andrews 1985). Interestingly the finds assemblage suggests the
presence of women on the site with neonatal bones recovered from three
contexts (Trenches 2, 4 and 5). The excavations at Reculver found at least
ten infant burials within the fort (Philp 2005, 225).
7.4.15
Very little higher status indications were recovered from the investigation
with the finds of opus signinum and wall plaster largely coming from Trench
1. Three of the five styli recovered also came from Trench 1 and this, along
with the evidence for more luxurious decorations, is perhaps a reflection of
its function. However it must be noted that all these items were retrieved
from the later demolition and abandonment deposits.
7.4.16
Evidence for diet is limited, but does demonstrate that meat and cereals
were, perhaps unsurprisingly, being supplemented by fish and seafood.
POTENTIAL AND FURTHER RECOMMENDATIONS
8.1
Potential
8.1.1
This investigation represents a small-scale evaluation of a large and

complex site, although the use of geophysical survey has enabled some
wider conclusions to be drawn. As a Scheduled Monument the fort is of
national significance and the results here, albeit limited, will inform research
about the wider network of Saxon Shore forts.
8.1.2
8.1.3
Finds
The finds assemblage from Brancaster is of moderate size, dominated by
pottery, ceramic building material and animal bone, with a smaller quantity of
metalwork, including coins. Other material types are represented in
negligible quantities. The assemblage is similarly limited in its range of
object types. Domestic and structural refuse is overwhelmingly predominant,
but there is otherwise little functional evidence (several styli, but no grain
processing or textile-working equipment, for example), and personal items
are noticeably scarce. Given the clear evidence for substantial buildings on
the Site, the scarcity of higher status building materials, such as painted
wall plaster, and anything more than a scattering of opus signinum, is
perhaps surprising. There was no evidence for tessellated pavements. The
small amount of stone building material recovered indicates both local and
regional sources of supply, and this is also likely to be the case for the
ceramic building material. Longer-distance contacts are demonstrated by the
pottery assemblage, but continental and Mediterranean imports are more
likely to have arrived at the fort via internal trade networks rather than
directly from overseas trade.
Despite the relatively large number of identifiable animal bone fragments in
the assemblage, the amount of specific information (e.g. age, biometry and
butchery) available for further study is quite limited and unlikely to provide
the level of detail needed to significantly enhance our understanding of
37

military food supply networks, butchery practices or the dietary preferences

of solders stationed at Brancaster Fort.
8.1.4
Certain items and groups of items amongst the finds are, however, of
intrinsic interest, such as the fragment of lorica squamata, and a probable
dispersed coin hoard in Trench 1. The coin assemblage as a whole provides
important new information on coin use and coin loss within Brancaster and
its environs. It also has the potential to inform discussion on the longevity of
activity within the fort and, through discussion of the likely hoard and its
context, on the chronology of the principia in particular. This information is
sufficiently important to merit publication as part of a wider report on the
excavations themselves. The hoard itself, whilst only partially complete, is
also worthy of publication in its own right.
8.1.5
The small group of post-Roman material, including Ipswich ware pottery and
(probably) the chalk die, is of interest, but its potential is limited by the small
quantities involved, and the probable residual provenance.
Environmental samples
8.1.6
8.1.7
8.1.8
8.1.9
8.1.10

The charred plant remains have the potential to demonstrate the range of
crops brought into the fort, as well as information on their processing and
from the weed seeds potentially crop-husbandry practices. Additionally the
charred material has also the potential to examine the use of local heathland
resources for fuel. The small number of remains and small range present
however make such potential very limited.
The mineralised remains have the potential to provide information on the
local vegetation growing on the site during the deposition of the occupation
material within layer 125. As most of the seeds have been identified to
species, and the range of species only appears to include plants associated
with disturbed and nitrogen rich soils in coastal areas, further potential from
full analysis is likely to be limited.
Wood charcoal
Wood charcoal can inform on the range of species collected as fuel, as well
as providing evidence for aspects of woodland management and
composition. However, given the low amount of wood charcoal there is little
further potential.
Land and aquatic molluscs
Further analysis of the mollusc assemblages from the Trench 1 deposits
would not assist in determining to nature of the local landscape to a greater
extent due to the mixed nature of the deposits. The assemblage from Trench
4 is too small for any analysis.
The small animal and fish bones from context 125 can provide information
on the broader diet beyond the usual range of larger domestic and wild
animals, whose bones are recovered through hand excavation.
38

8.2
8.2.1
Proposals
Finds
All the finds have already been recorded to an appropriate archive level, and
no further analysis or reporting is proposed on any of the material types.
Data gathered as part of the assessment stage, and assessment reports as
presented here, will be incorporated as appropriate in the proposed
publication report.
Environmental samples
8.2.2

It is proposed to analyse the plant remains from all three samples.
8.2.3
All identifiable charred and mineralised plant macrofossils will be extracted

from the 2 and 1mm residues together with the flot. Identification will be
undertaken using stereo incident light microscopy at magnifications of up to
x40 using a Leica MS5 microscope, following the nomenclature of Stace
(1997) for wild plants, and traditional nomenclature, as provided by Zohary
and Hopf (2000, Tables 3, page 28 and 5, page 65), for cereals. and with
reference to modern reference collections where appropriate, quantified and
the results tabulated.
8.2.4
The samples proposed for analysis are indicated with a P in the analysis
column in Table 20, Appendix 3.
8.2.5
Wood charcoal
No further work is proposed.
8.2.6
Land and aquatics molluscs

8.2.7

8.2.8
Publication
It is recommended that short article summarising the results of these
investigations and the further environmental analysis undertaken be
submitted to the Norfolk Archaeological Journal. An report of approximately
3000 words is proposed, with 3-4 accompanying figures, and finds and
environmental data tabulated as appropriate.
ARCHIVE
9.1.1
It is recommended that the project archive resulting from the excavation be

deposited with Norfolk Museums and Archaeology Service. The Museum
has issued an Event Number (ENF129426) and an accession number for
the project (NWHCM : 2012.240), but is not currently accepting archives due
to lack of storage space. Deposition of any finds with the Museum will only
be carried out with the full agreement of the landowner.
9.1.2
The complete site archive, which will include paper records, photographic
records, graphics, artefacts, ecofacts and digital data, will be prepared
following the standard conditions for the acceptance of excavated
39

archaeological material by Norfolk Museums and Archaeology Service, and

in general following nationally recommended guidelines (SMA 1995; IfA
2009; Brown 2011; ADS 2013).
9.1.3
9.1.4
9.1.5
9.1.6
All archive elements will be marked with the Event Number and accession
code, and a full index will be prepared. An OASIS online record
http://ads.ahds.ac.uk/projects/oasis/ will be initiated and key fields
completed on Details, Location and Creators Forms. All appropriate parts of
the OASIS online form will be completed for submission to the HER. This will
include an uploaded .pdf version of the entire report.
Copyright
The full copyright of the written/illustrative archive relating to the Site will be
retained by Wessex Archaeology Ltd under the Copyright, Designs and
Patents Act 1988 with all rights reserved. The recipient museum, however,
will be granted an exclusive licence for the use of the archive for educational
purposes, including academic research, providing that such use shall be
non-profitmaking, and conforms with the Copyright and Related Rights
Regulations 2003.
This report may contain material that is non-Wessex Archaeology copyright
(e.g. Ordnance Survey, British Geological Survey, Crown Copyright etc.), or
the intellectual property of third parties, which we are able to provide for
limited reproduction under the terms of our own copyright licences, but for
which copyright itself is non-transferrable by Wessex Archaeology. You are
reminded that you remain bound by the conditions of the Copyright, Designs
and Patents Act 1988 with regard to multiple copying and electronic
dissemination of the report
Security copy
In line with current best practice (Brown 2011), on completion of the project
a security copy of the paper records will be prepared, in the form of a PDF/A
file; PDF/A is an ISO-standardised version of the Portable Document Format
(PDF) designed for the digital preservation of electronic documents through
omission of features ill-suited to long-term archiving.
40

10
REFERENCES
10.1
Bibliography
ADS, 2013. Caring for Digital Data in Archaeology: a guide to good practice,
Archaeology Data Service & Digital Antiquity Guides to Good Practice
Agricola, Georgius, 1950 [1556]. De Re Metallica translated by Herbert
Hoover and Lou Henry Hoover, Dover
Allen, J R L and Fulford, M G, 1999. Fort building and military supply along
Britains Eastern Channel and North Sea Coasts: The Later Second
and Third Centuries, Britannia 30, 163-84
Allen, J R L, Fulford, M G and Pearson, A F, 2001. Branodunum on the
Saxon shore (North Norfolk): a local origin for the building material,
Britannia 32, 271-5
Anderson, R, 2005. An annotated list of the non-marine Mollusca of Britain
and Ireland, Journal of Conchology 38, 607-37
Andrews, G, 1985. The Coarse Wares in Hinchliffe and Sparey Green
1985, 23, 82-98, 100-117, 123-125
Aubertin, D, Ellis, A E and Robson, G C, 1931. The natural history and
variation of the pointed snail, Cochlicella acuta (Mll.), Proc. Zool. Soc.
67 (1930), 1027-55
Barker, T C and Hatcher, J, 1974. A History of British Pewter
Bartlett, A D H, 1973. Geophysics: Brancaster, Norfolk, English Heritage
AM Lab Report (Old Series), ref. 1605
Bartlett, A D H, 1975. Geophysics: Brancaster, Norfolk, English Heritage
AM Lab Report (Old Series), ref. 1989
Bidwell, P, 1997. Roman Forts in Britain, London: English Heritage/ B T
Batsford
Brown, D H, 2011. Archaeological archives; a guide to best practice in
creation, compilation, transfer and curation, Archaeological Archives
Forum (revised edition)
Corder, P (ed.), 1961. The Town and Villa at Great Casterton, Rutland.
Third report for the years 1954-8
Crummy, N, 1983. The Roman small finds from excavations in Colchester
1971-9, Colchester Archaeol. Rep. 2
Cunliffe, B, 1968. The British Fleet, in B. Cunliffe (ed.) Fifth Report on the
Excavation of the Roman Fort at Richborough, Kent, Report of the
Research Committee of the Society if Antiquaries of London 23, 25571
41

Cunliffe, B, 1977. The Saxon Shore some problems and misconceptions,

in D. E. Johnston (ed.) The Saxon Shore, Counc Brit Archaeol Res
Rep 18, 1-6
Davies, R W, 1971. The Roman military diet, Britannia 2, 122-42
Dobney, K, Jacques, D and Irving, B, 1996. Of Butchery and Breeds:
Report on the Vertebrate Remains from Various Sites in the City of
Lincoln. Lincoln Archaeol Studies 5
Dobney, K, 2001. A place at the table: the role of vertebrate
zooarchaeology within a Roman research agenda for Britain, in S
James and M Millet (eds.), Britons and Romans: advancing an
archaeological agenda. Counc Brit Archaeol Res Rep 125, 36-45
Edwards, D A and Green, C J S, 1977. The Saxon Shore fort and
settlement at Brancaster, Norfolk, in D E Johnston (ed.) The Saxon
Shore, Counc Brit Archaeol Res Rep 18, 21-9
Flack, S and Gregory, T, 1988. Excavations at Brancaster, 1985, Norfolk
Archaeology 40, 164-71
Fuller, D Q and Stevens, C J, 2009. Agriculture and the development of
complex societies: An archaeobotanical agenda, in Fairbairn, A. and
Weiss, E. (ed.) From Foragers to Farmer: Papers in honour of Gordon
Hillman, Oxford: Oxbow Books, 36-57
Gaffney, C and Gater, J, 2003. Revealing the Buried Past: Geophysics for
Archaeologists, Stroud: The History Press
Grant, A, 1987. Some observations on butchery in England from the Iron
Age to the medieval period, Anthropozoologica, Premier Numro
Spcial, 53-58
GSB, 2006. Brancaster: Geophysical Survey, unpublished report, ref. 64
GSB, 2014. Geophysical Survey Report G1252: Brancaster Roman Town,
Norfolk, unpublished report, ref. 2012/52
Gurney, D, 1990. A Romano-British Pottery kiln at Blackborough End,
Middleton Norfolk Archaeology 41 (1990-3), 83-92
Hall, A R and Huntley, J P, 2007. A review of the evidence for macrofossil
plant remains from archaeological deposits in Northern England.
Research Department Report Series no. 87-2007, Portsmouth: English
Heritage
Hartley, K F and Perrin, J R, 1999. Mortaria from Excavations by E
Greenfield at Water Newton, Billing Brook and Chesterton 1956-58, in
Perrin 1999, 129-35
42

Hassall, M W C, 1977. The historical background and military units of the

Saxon Shore, in D E Johnston (ed.) The Saxon Shore, Counc Brit
Archaeol Res Rep 18, 7-10
Hinchliffe, J C, 1985. Discussion, in Hinchliffe and Sparey Green 1985,
176-81
Hinchliffe, J and Sparey Green, C, 1985. Excavations at Brancaster 1974
and 1977, East Anglian Archaeology Report No. 23, Norfolk
Archaeological Unit
Hinton, D A, 1996. The gold, silver and other non-ferrous alloy objects from
Hamwic, and the non-ferrous metalworking evidence, Southampton
Finds Vol. 2
Howe, M D, Perrin, J R and Mackreth, D F, 1980. Roman Pottery from the
Nene Valley: A Guide. Peterborough City Museum Occasional Paper
2.
Huntley, J P, 2000. The plant remains, In K. Buxton & C. Howard-Davis
(eds.) Bremetenacum. Excavations at Roman Ribchester 1980, 19891990, Lancaster: Lancaster Imprint Series 9, 349-59
IfA, 2009. Standard and Guidance for the creation, compilation, transfer and
deposition of archaeological archives, Institute for Archaeologists
Jones, G, 1985. The animal bones from the 1974 excavations, in Hinchliffe
and Sparey Green 1985, 129-31
Jones, R, Langley, P and Wall, S, 1985, The animal bones from the 1977
excavations, in Hinchliffe and Sparey Green 1985, 132-74
Kerney, M P, 1999. Atlas of the Land and Freshwater Molluscs of Britain
and Ireland, Colchester: Harley Books
King, A, 1984. Animal bones and the dietary identity of military and civilian
groups in Roman Britain, Germany and Gaul, in T F C Blagg and A
King (eds), Military and civilian in Roman Britain: cultural relationships
in a frontier province. Oxford: Brit Archaeol Rep 136, 187-218
King, A, 1991. Food production and consumption meat, in R F J Jones
(ed), Roman Britain: recent trends. Dept. Archaeol. and Prehist.,
University of Sheffield: J. R. Collis Publications, 15-20
Lee Warner, J, 1851. Notices of the Original Structure of the Roman
Fortifications at Brancaster, (the ancient Branodunum), Norfolk, Proc
Archaeol Institute, Norwich, 9-16
Lyons, A L, 2004. Romano-British Industrial Activity at Snettisham, Norfolk,
East Anglian Archaeol Occas Paper 18
Maltby, M, 1989. Urban-rural variations in the butchery of cattle in RomanoBritish Hampshire, in D Serjeantson and T Waldron (eds), Diet and
Crafts in Towns. Oxford: Brit. Archaeol. Rep. 199, 75-106
43

Manning, W H, 1985. Catalogue of the Romano-British Iron Tools, Fittings

and Weapons in the British Museum. London: British Museum
Mays, S, 1993. Infanticide in Roman Britain, Antiquity 67, 883-8
McKinley, J I, 2011. Human Bone in C Barnett, J I McKinley, E Stafford, J
Grimm and C J Stevens, Settling the Ebbsfleet Valley: High Speed I
Excavations at Springhead and Northfleet, Kent. The Late Iron Age,
Roman, Saxon and Medieval Landscape. Vol. 3: Late Iron Age to
Roman Human Remains and Environmental Reports. OxfordWessex
Archaeology, 1-14
Peacock, D P S and Williams, D F, 1986. Amphorae and the Roman
Economy. Longman Archaeology Series
Perrin, J R, 1999. Roman Pottery from Excavations at and near to the
Roman Small Town of Durobrivae, Water Newton, Cambridgeshire,
1956-58. Journal of Roman Pottery Studies 8
Philp, B, 2005. The Excavation of the Roman Fort at Reculver, Kent, Kent
Monograph Series 10, Dover: Kent Archaeological Rescue Unit
Philpott, R, 1991. Burial Practices in Roman Britain. Oxford: Brit Archaeol
Rep 219
Reece, R, 1991. Roman Coins from 140 Sites in Britain, Cotswold Studies 4
Riddle, J M, 1997. Eves Herbs. A history of contraception and abortion in
the west. Harvard University Press
Rotham, A S, 1960. Notes of a trial excavation at Brancaster (789443) on a
Roman site in the filed to the west of Straithe House, unpublished
notes
Scott, E, 1999. The archaeology of infancy and infant death. Oxford: Brit
Archaeol Rep Int Series 819
SMA, 1995. Towards an Accessible Archaeological Archive, Society of
Museum Archaeologists
Sparey Green, C, 1985, The 1974 Excavations, in Hinchliffe and Sparey
Green 1985, 4-14

Sparey Green, C and Gregory, T, 1985. Appendix 4. Surface Finds, in
Hinchcliffe ad Sparey Green 1985, 190-3
St. Joseph, J K S, 1936. The Roman Fort at Brancaster, Antiq J 16, 444-60
Stace, C, 1997. New Flora of the British Isles, Cambridge: Cambridge
University Press (2nd ed.)
44

Stallibrass, S and Thomas, R, 2008. Feeding the Roman Army: the

archaeology of production and supply in NW Europe. Oxford: Oxbow
Books
Struck, M, 1993. Kinderbestattungen in romano-britischen Siedlungen-der
archologische Befund, in M. Struck (ed), Rmerzeitliche Grber als
Quellen zu Religion, Bevlkerungsstruktur und Sozialgeschichte.
ArchologischeSchriften des Instituts fr Vor-und Frhgeschichte det
Johannes Gutenberg-Universitt, Mainz 313-18
Swan, V G, 1984. The Pottery Kilns of Roman Britain. Royal Commission on
Historical Monuments Supplementary Series 5
Thomas, R, 2008. Supply-chain networks and the Roman invasion of
Britain: a case study from Alchester, Oxfordshire, in Stallibrass and
Thomas 2008, 31-51
Videotext Communications, 2013. Proposed Archaeological Evaluation:
Brancaster Roman Fort, Norfolk, unpublished project design
Walker, K, 1990. Guidelines for the preparation of excavation archives for
Long-term Storage, UKIC Archaeology Section
Warry, P, 2006. Tegulae manufacture, typology and use in Roman Britain.
Oxford: Brit Archaeol Rep 417
Winder, J M, 1999. Oyster shell, in C M Hearne and V Birbeck, A35
Tolpuddle to Puddletown Bypass DBFO, Dorset, 1996-8 Wessex
Archaeology Report No. 15, 202
Woodward, P J, 1993. Counters and dice, in P J Woodward, S M Davies
and A H Graham, Excavations at Greyhound Yard, Dorchester, 19814, Dorset Archaeol Natur Hist Soc Monogr 12, 190-4
Wyles, S F, 2008. Oysters, in A P Fitzpatrick, A B Powell and M J Allen,
Archaeological Excavations on the Route of the A27 Westhampnett
Bypass West Sussex, 1992 Volume 1: Late Upper Palaeolithic-AngloSaxon, Wessex Archaeology Report 21, 225-7
Young, C J, 1977. Oxfordshire Roman Pottery. Oxford: Brit Archaeol Rep
43
Zohary, D and Hopf, M, 2000. Domestication of plants in the Old World: the
origin and spread of cultivated plants in West Asia, Europe, and the
Nile Valley, Oxford: Clarendon Press (3rd ed.)
van der Veen, M, 1992. Crop husbandry regimes; An archaeobotanical
study of farming in northern England 1000 B.C. - A.D. 500, Sheffield
Archaeol Monogr 3, University of Sheffield, J R Collis Publications,
Department of Archaeology and Prehistory
45

10.2
Online resources
Geological Survey of Great Britain 1:50,000 mapping available at:
http://www.bgs.ac.uk/data/services/digmap50wms.html
Scheduled monument information available at:
http://list.english-heritage.org.uk/
46

APPENDIX 1: TRENCH SUMMARIES

bgl = below ground level
Machine excavated
TRENCH 1
Type:
Dimensions: 20.20x1.80m
Max. depth: 1.16m
Ground level: 14.65-15.56m aOD
Context Description
Depth (m)
101
Modern topsoil. Dark brown sandy loam. <1% flint, sub-angular 0.00-0.25
Topsoil
sub-rounded, <1-3cm. Loose and friable. Bioturbated. Homogeneous. bgl
Under grass. Overlies 110.
102
Equivalent to 104 and 108, northern area of trench. Occupation 0.34 deep
Layer
debris likely post-abandonment. Very dark grey-brown sandy silt
loam. 5% stone, gravel and chalk, sub-angular sub-rounded, <16cm. Frequent oyster shells, animal bone and CBM. Slightly mixed.
Fairly compact. Overlies 111, 112 and 113.
103
Number assigned for finds retrieval, cleaning over 104, base of 101
104
Equivalent to 102 and 108, south end of trench. Occupation debris 0.34 deep
Layer
likely post-abandonment. Very dark grey-brown sandy silt loam. 5%
stone, gravel and chalk, sub-angular sub-rounded, <1-6cm.
Frequent oyster shells, animal bone and CBM. Slightly mixed. Fairly
compact. Overlies 105.
105
Equivalent to 106 and 107, northern end of trench. Very dark brown 0.30 deep
Layer
sandy loam. 1% stone/gravel, sub-angular sub-rounded, <1-4cm.
Frequent oyster shell, occasional animal bone and CBM. Rare
charcoal and chalk flecks. Fairly friable but moderately compact.
Overlies 114.
106
Equivalent to 105 and 107, southern end of trench. Very dark brown 0.33 deep
Layer
Environmental sample 3. Overlies 116 and 137.
107
Equivalent to 105 and 106, central area of trench. Very dark brown 0.29 deep
Layer
Overlies 126.
108
Equivalent to 102 and 104, central area of trench. Occupation debris 0.43 deep
Layer
likely post-abandonment. Very dark grey-brown sandy silt loam. 5%
stone, gravel and chalk, sub-angular sub-rounded, <1-6cm.
Frequent oyster shells, animal bone and CBM. Slightly mixed. Fairly
109
Discrete dump of demolition material. Very dark grey-brown sandy silt 0.31 deep
Layer
loam. 5% stone, sub-angular, 2-8cm. Rare oyster shell. Slightly
mixed. Moderately compact. Overlies 106.
110
Deliberate backfill of possible pit 135. Dark grey-brown sandy silt 0.26+ deep
Deposit
loam. 2% chalk fragments and oyster shell, rare CBM. Largely
unexcavated. Overlies 135.
111
Possible rough surface. Dark grey-brown sandy loam. 60% stone and ?Surface
chalk, occasional mortar and CBM. Compact. Unexcavated. Overlies
106.
112
Possible wall remnant or structural feature composed of large chalk 0.12 high
Structure
blocks. No bedding material or core. 0.85x0.40m. Only single course
remaining. Overlies 106.
113
Demolition rubble spread in central part of trench, potentially 0.25 deep
Layer
associated with later robbing. Mid grey-brown sandy loam. 60%
stone, sub-angular, 2-10cm. Occasional CBM fragments. Overlies
109.
47

114
Layer
115
Layer
116
?Surface
117
Layer
118
Layer
119
Layer
120
Layer
121
Layer
122
Layer
123
Layer
124
Layer
125
Layer
126
Deposit
127
Cut
128
Structure
129
Layer
130
Structure
131
Surface
132
Layer
133
Layer
134
?Surface
135
Cut
Equivalent to 120. Demolition debris. Mid orange brown sandy loam.

15% stone, sub-angular, <1-3cm. Occasional tile fragments and
mortar fragments. Fairly compact. Overlies 115.
Equivalent to 121. Tile debris, possible roof collapse or reclamation.
Mid orange-brown silty sand. Abundant tile fragments. Fairly
homogenous. Moderately compact. Overlies 122.
Possible surface. Mid yellow-pink mortar with small gravel inclusions.
Compact. Overlies 117.
Very dark grey-brown sandy loam. Rare charcoal flecks. Loose and
friable. Overlies 118.
Possible levelling layer. Mid yellow-brown sandy silt loam. <1%
stone/gravel, rounded, <1-2cm. Fairly homogeneous. Overlies 119.
Tile debris, possible roof collapse or reclamation. Mid yellow-brown
sandy mortar. Abundant tile fragments. Fairly homogenous.
Moderately compact. Overlies 133.
Equivalent to 114. Demolition debris. Mid orange brown sandy loam.
15% stone, sub-angular, <1-3cm. Occasional tile fragments and
mortar fragments. Fairly compact. Overlies 121.
Equivalent to 115. Tile debris, possible roof collapse or reclamation.
Mid orange-brown silty sand. Abundant tile fragments. Fairly
Mid orange-brown sandy loam. 2% stone/gravel, sub-angular subrounded, <1-3cm. Occasional small CBM fragments. Moderately
homogeneous. Fairly compact. Overlies 136.
Mortar spread. Mid yellow. Moderately homogeneous and compact.
Overlies 124.
Possible deliberate backfill relating to building demolition. Mixed
lenses of pale yellow-white mortar and very dark-brown sandy silt
loam. Occasional CBM fragments within mortar. Fairly compact.
Overlies 125.
Occupation deposit. Dark yellow-brown sandy silt loam. <1%
stone/gravel, sub-angular sub-rounded, <1-3cm. Rare mortar
flecks. Contains ABG 82. Environmental sample 1. Overlies 131.
Deliberate backfill of robber cut 127. Dark yellow-brown sandy silt
loam. 5% stone, sub-angular, <1-6cm. Rare CBM fragments. Fairly
Robber cut which has removed upper portion of 128, filled with
126. East west aligned. Steep, straight side, flat base. Full
width not exposed 0.90m+ wide. Cuts 120.
East west aligned wall, southern exterior wall. Face material
squared stone blocks mostly obscured by dark yellow sandy mortar.
Full surviving height not exposed. At least 0.90m wide.
Material around lorica squamata (ON83) collected for find retrieval.
Overlies 132.
East west aligned wall, interior wall. Face material squared stone
blocks mostly obscured by pale yellow sandy mortar. Full surviving
height not exposed. 0.48m wide.
Mortar surface. Pale yellow-white mortar. Does not fully extend to
wall face 130. Only partially excavated. Overlies 129.
Exposed along edge of wall, only removed where ON83 lifted. Pale
pink-brown mortar, possible surface. Overlies 139.
Trample deposit. Very dark grey-brown silt. Frequent charcoal flecks.
Laminations visible. Compact. Unexcavated except to establish
textural class and depth. Overlies 134.
Possible rough surface. Mid yellow-brown sandy silt loam. Includes
several large (8-15cm) flat stones. Unexcavated.
Seen mostly in section, largely unexcavated. Difficult to see in
plan. Possible pit filled with 110. Cuts 102.
48
0.08 deep
0.21 deep
0.05 deep
0.06 deep
0.09 deep
0.14 deep
0.10 deep
0.22 deep
0.20 deep
0.10 deep
0.38 deep
0.10 deep
0.65 deep
0.65 deep
0.28+ high
0.54+ high
0.02 deep
0.02 deep
0.01 deep
0.26+ deep

136
Layer
137
Structure
138
Layer
139
Layer
140
Layer
141
Layer
142
Layer
143
Layer
Pale orange sandy silt loam. Abundant small angular stone chips.
Possible construction or reclamation debris. Abuts wall 128. Overlies
142.
Possible structural element but post-dates robbing. Area of pink-red
crushed CBM or mortar includes some near vertical tiles mortared
together. Unexcavated. Overlies 123.
Build up to north of wall 130. Mid orange-brown sand. 10%
stone/chalk, sub-angular sub-rounded, <1-10cm. Slightly mixed
deposit. Fairly compact. Overlies 140.
Apparently re-deposited natural as wall depth continues. Mid orange
sand. Only very small area seen beneath 132 where ON83 removed.
Unexcavated. May be similar to 140. Abuts 130.
Apparently re-deposited natural as wall depth continues. Mid redorange sand. 2% flint/gravel, sub-angular sub-rounded, <1-6cm.
Fairly homogeneous and compact. Unexcavated. May be similar to
139. Abuts 130.
Defined area or lens of mid red clay apparently abutting wall 128.
Overlies 143.
Apparently re-deposited natural as wall depth continues. Mid redorange sand. <1% flint/gravel, sub-angular sub-rounded, <1-3cm.
Fairly homogeneous and compact. Very similar to 143. Overlies 141.
Apparently re-deposited natural as wall depth continues. Mid redorange sand. <1% flint/gravel, sub-angular sub-rounded, <1-3cm.
Fairly homogeneous and compact. Very similar to 142. Abuts 128.
0.06 deep
0.10 high
0.20 deep
0.07 deep
0.33 deep
0.28+ deep
Machine excavated
TRENCH 2
Type:
Max. depth: 0.98m
Context Description
Depth (m)
201
Modern topsoil. Dark brown sandy loam. 1% flint/stone, sub-angular 0.00-0.35
Topsoil
sub-rounded, <1-4cm. Loose and friable. Homogenous. Bioturbated bgl
Under grass. Overlies 202, 206 and 207
202
Disturbed interface/ layer between topsoil 201 and surface 203. Mid 0.12 deep
Layer
grey-brown sandy loam. 2% stone, sub-rounded, <1-3cm. Occasional
sand lenses. Overlies 203.
203
Same as 204. 40% flint gravel and chalk flecks, sub-angular, <1-2cm Surface
within pink-grey mortar form base. Finer mid orange sandy ?mortar
with 50% flint gravel , sub-angular, <1-2cm form upper layer.
Compact. Left in situ. Overlies 209.
204
Same as 203. 40% flint gravel and chalk flecks, sub-angular, <1-2cm 0.12 deep
Surface
within pink-grey mortar form base. Finer mid orange sandy ?mortar
with 50% flint gravel , sub-angular, <1-2cm form upper layer. Some
traces of mortar, possibly opus signinum on upper surface. Compact.
Abuts 205.
205
East west aligned chalk wall. Chalk sub-angular facing blocks, 8- 0.12 high
Structure
12cm, chalk rubble core. Only 1 course remaining, no discernable
foundation. Pale pink grey mortar. 0.50m wide.
206
Demolition debris. Mid brown sandy silt loam. 40% stone and flint 0.15 deep
Layer
rubble, angular sub-rounded, <1-3cm, 8-34cm. Slightly mixed.
Fairly compact. Some bioturbation. Overlies 205, 212, 213, 214, 216,
220 and 224.
207
Secondary fill of gully 208. Dark brown silty sand. 1% flint, sub- 0.40 deep
Deposit
rounded, <1-2cm. Occasional chalk and mortar fragments. Fairly
homogeneous. Moderately compact. Some bioturbation. Overlies
208.
208
East west aligned gully, filled with 207. Straight, near vertical 0.40 deep
Cut
sides, concave base. 0.36m wide. Cuts 204.
209
Same as 210. Made ground, re-deposited natural. Pale to mid brown 0.53+ deep
Layer
49

210
Layer
211
212
Natural
Structure
213
Structure
214
Structure
215
Cut
216
Deposit
217
Deposit
218
Deposit
219
Cut
220
Deposit
221
Layer
222
Layer
223
Cut
224
Deposit
sandy silt loam. <1% stone, rounded sub-rounded, <1cm.

Homogeneous. Compact. Some bioturbation. Overlies 211.
Same as 209. Made ground, re-deposited natural. Pale to mid brown
sandy silt loam. <1% stone, rounded sub-rounded, <1cm.
Homogeneous. Compact. Some bioturbation. Overlies 211.
Natural geology. Mid yellow-brown sand. No inclusions. Compact.
Possible post-pad, sub-rectangular. 0.35x0.24m. Composed of pale
white-grey mortar. Possible group with 213, 214 and possibly 223.
Possible post-pad, sub-rectangular. 0.36x0.25m. Composed of pale
white-grey mortar and tile fragment. Possible group with 212, 214
and possibly 223.
Possible post-pad, sub-rectangular. 0.15+x0.20m. Composed of pale
white-grey mortar. Possible group with 212, 213 and possibly 223.
Small sub-oval pit fill with 216, 217 and 218. Includes possible
posthole in southern end. Steep straight sides, generally very
slightly concave base except for concave area at southern end.
1.08m long, 0.50m wide. Cuts 210.
Secondary fill or possible deliberate backfill. Dark brown sandy loam.
2% flint, sub-rounded sub-angular, <1-4cm, 5-18cm.Occasinal
charcoal flecks. Very mixed deposit, mid orange-brown and mid red
mottling. Some bioturbation. Infills area of possible posthole. Overlies
217.
Possible deliberate deposit, possible clay lining. Mid red-brown clay.
No inclusions. Very occasional mid brown mottles due to bioturbation
but otherwise homogeneous. Compact. Overlies 218.
Possible in-situ heating. Mid to pale red sand. Homogenous. Some
bioturbation. Compact. Overlies 215.
Small sub-oval pit filled with 220. Moderate, concave sides,
concave base. 0.65m long, 0.45m wide. Cuts 210.
Secondary fill of pit 219. Dark brown-grey sandy silt loam. <1% stone,
sub-rounded, <1cm. Occasional charcoal flecks concentrated at base
of deposit. Very slightly mixed. Some bioturbation. Overlies 219.
Possible trample layer beneath 210. Mid grey sand. No inclusions.
Overlies 211.
Thin layer seen in south-west part of trench. Mid orange-brown sand.
<1% stone, sub-angular sub-rounded, <1-2cm. Slightly mixed.
Fairly compact. Overlies 210.
Possible group with 212, 213 and 214. Possible posthole filled
with 223. Not seen in plan. 0.40m wide. Cuts 222. Adjacent to
wall 205 but relationship unclear.
Secondary fill of possible posthole 223. Dark grey brown sandy loam.
10% stone and chalk rubble, 6-20cm either post-packing or
collapse from wall 205. Fairly homogeneous. Overlies 223.
0.34 deep
0.80+ bgl
-
0.42 deep
0.42 deep
0.14 deep
0.02 deep
0.15 deep
0.15 deep
0.02 deep
0.12 deep
0.40 deep
0.40 deep
Machine excavated
TRENCH 3
Type:
Max. depth: 0.58m
Context Description
Depth (m)
301
Modern topsoil/ former ploughsoil. Dark black-brown sandy loam. 2% 0.00-0.58
Topsoil
flint, sub-angular sub-rounded, <1-5cm. Fairly loose and friable. bgl
Homogeneous. Bioturbated. Under grass. Overlies 302, 306 and 316.
302
Remnants of road metalling, very patchy. Has survived within the top 0.10 deep
Layer
of some of the earlier features. Same as 316. Dark grey-brown silty
sand. 40% flint/ gravel, sub-angular sub-rounded, <1-4cm.
Compact. Fairly homogeneous. Overlies 309.
303
Possible deliberate backfill of ditch 312. Same as 304. Dark brown 0.66 deep
Deposit
silty sand. 2% flint/gravel, sub-angular sub-rounded, <1-4cm. Fairly
50

304
Deposit
305
Cut
306
Deposit
307
Deposit
308
Cut
309
Deposit
310
Cut
311
Deposit
312
Cut
313
Cut
314
Deposit
315
Deposit
316
Surface
317
Natural
307. Same as 315 within 313.

Possible deliberate backfill of ditch 312. Same as 303. Dark brown
silty sand. 2% flint/gravel, sub-angular sub-rounded, <1-4cm. Fairly
307. Same as 315 within 313.
Roadside ditch, east west aligned filled with 306. Straight,
sides, steep north edge, moderate south edge. Flat base. 1.08m
wide. Cuts 309.
Secondary fill of ditch 305. Dark brown sand. 20% flint/gravel, subangular angular, <1-8cm. Sand lens at base of deposit along
interface with cut. Some eroded metalling from road 302/316 within
top 0.12 of deposit. Compact. Overlies 305.
Secondary fill of ditch 312. Mixed/ layered deposit of mid to dark
brown lenses and pale yellow-white lenses of sand. No inclusions.
Compact. Banding suggests probably water action. Not fully
excavated. Lowest fill encountered within 312. Same as 314 within
313.
North south aligned curvilinear ditch. Filled with 309.
Moderate, concave sides, concave base. 1.2-2.0m wide. Cuts 303
and 311.
Secondary fill of ditch 308. Pale brown sand. 2% flint/gravel subangular sub-rounded, <1-4cm. Fairly homogeneous. Moderately
Possible pit or ditch terminus, filled with 311. East west
aligned. Concave, moderate sides, concave base. Full extent not
seen in plan. Cuts 317.
Secondary fill of ditch 310. Pale brown sand. 2% flint/gravel subangular sub-rounded, <1-3cm. Fairly homogeneous. Moderately
Large north-east south-west aligned ditch filled with 303, 304
and 307. Northern edge. Same as 313. Steep, straight sides. Not
fully excavated. 2.3m wide. Cuts 317.
Large north-east south-west aligned ditch filled with 314 and
315. Southern edge. Same as 312. Steep, straight sides. Not fully
excavated. 2.3m wide. Cuts 317.
Secondary fill of ditch 313. Mixed/ layered deposit of mid to dark
brown lenses and pale yellow-white lenses of sand. No inclusions.
Compact. Banding suggests probably water action. Not fully
excavated. Lowest fill encountered within 313. Same as 307 within
312.
Possible deliberate backfill of ditch 313. Dark brown silty sand. 2%
flint/gravel,
sub-angular
sub-rounded,
<1-3cm.
Fairly
314. Same as 303/304 within 312.
Remnants of road metalling, very patchy. Has survived within the top
of some of the earlier features. Same as 302. Dark grey-brown silty
sand. 40% flint/ gravel, sub-angular sub-rounded, <1-4cm.
Compact. Fairly homogeneous. Overlies 315.
Natural geology. Mid orange sand and gravel. 10% gravel, subangular sub-rounded, <1-3cm. Slightly mixed, some pale yellow
orange patches. Compact.
0.66 deep
0.44 deep
0.44 deep
0.50+ deep
0.30 deep
0.30 deep
0.30 deep
0.30 deep
1.20+ deep
0.95+ deep
0.14+ deep
0.66 deep
0.10 deep
0.44+ bgl
Machine excavated
TRENCH 4
Type:
Max. depth: 1.40m
Context Description
Depth (m)
401
Modern topsoil. Dark brown sandy loam. 2% flint, sub-angular sub- 0.00-0.50
Topsoil.
rounded, <1-4cm. Loose and friable. Homogeneous. Bioturbated. bgl
51

402
Deposit
403
Structure
404
Deposit
405
Deposit
406
Cut
407
Deposit
408
Cut
409
Deposit
410
Deposit
411
Cut
412
Surface
413
Layer
414
Layer
415
Layer
416
Layer
417
Layer
418
Deposit
419
Layer
420
Deposit
Under grass. Overlies 402 and 410.

Demolition debris, upper fill of robber cut 408. Mid orange silty sand.
15% flint and stone, sub-angular sub-rounded, <1-8cm. Slightly
mixed. Fairly compact. Overlies 407.
Northern exterior defensive wall of fort. East west aligned. 2.5m
wide. Face material pale white micaceous stone and flint nodules
within pale yellow lime mortar. Lower portion slightly stepped out.
Mortar widely slobbered so coursing pattern not apparent. Left in situ.
Secondary fill of ditch 406. Dark grey-brown silty sand. 25% stone,
sub-angular, <1-10cm. Occasional charcoal flecks. Slightly mixed.
Fairly compact. Overlies 409.
Secondary fill or possible deliberate backfill of robber cut 408. Mid
orange-brown sandy silt loam. 2% flint, sub-rounded, <1-2cm. Slightly
mixed, some bioturbation. Fairly compact. Overlies 408.
Northern defensive ditch, east west aligned, filled with 404, 409
and 420. Not fully excavated so profile not seen.
Secondary fill or possible deliberate backfill of robber cut 408. Dark
grey-brown silty sand. 5% stone an flint, sub-angular sub-rounded,
<1-4cm. Slightly mixed. Fairly compact. Overlies 418.
Robber cut associated with wall 403, situated over top of wall
and down northern face. East west aligned, filled with 402, 405,
407 and 418. Straight steep sides, stepped base. 3.8m wide. Cuts
404 and 413.
Secondary fill of ditch 406. Dark brown silty sand. 5% flint and stone,
sub-angular rounded, <1-3cm. Occasional mortar and oyster shell.
Very slightly mixed. Fairly compact. Not fully excavated. Overlies
420.
Secondary fill of robber trench 411. Dark brown sandy silt loam. 2%
chalk and gravel, sub-angular sub-rounded, <1cm. Topsoil/
ploughsoil derived material. Fairly homogeneous. Moderately
Series of robber trenches in southern end of trench. Thought to
be from either early modern robbing or investigations. Filled
with 410. Cuts 412 and 419.
Mortar surface or wall foundation. Shape in plan highly truncated. Mid
yellow-grey sandy mortar. 10% chalk and lint, sub-rounded subangular, 1-5cm. Compact. Overlies 416.
Rampart bank. East-west aligned. Mid orange sand. 2% flint/ gravel,
sub-angular, <1-4cm. Fairly homogenous. Compact. Some
bioturbation. Overlies 415.
Rampart bank. East-west aligned. Mid orange sand. 2% flint/ gravel,
sub-angular, <1-8cm. Fairly homogenous. Compact. Some
bioturbation. Abuts 403.
Thin discontinuous charcoal rich lens. Dark grey-black sandy loam.
Frequent charcoal. Environmental sample 2. Overlies 414.
Possible makeup. Mid yellow-brown sandy silt. 5% stone and chalk,
sub-angular sub-rounded, <1-4cm. Frequent mortar. Overlies 417.
Possible surface or bedding layer. Mid brown sandy silt with pale
pink-orange mortar. 2% flint and chalk, sub-rounded, <1-4cm. Heavily
truncated. Compact. Slightly mixed. Left in situ.
Deliberate backfill of robber cut 408. Likely reclamation debris, east
facing section only. Mid grey sandy silt loam. 60% sub-angular stone
chippings, <1-4cm. Fairly compact. Slightly mixed. Overlies 405.
Possible abandonment, demolition debris. Mid grey-brown sandy
loam. 1% flint, sub-angular, <1-8cm, Frequent oyster shell, rare stone
and mortar flecks. Mixed with frequent mid orange-brown mottles.
Fairly compact. Overlies 413. Largely unexcavated.
Possible lower fill of ditch 406, lowest deposit encountered but could
52
0.35 deep
0.75+ high
0.50 deep
0.31 deep
0.88+ deep
0.30 deep
0.62 deep
0.28+ deep
0.48+ deep
0.48+ deep
0.10 deep
0.52 deep
0.17+ deep
0.03 deep
0.13 deep
-
0.11 deep

be on the edge of the ditch edge rather than within ditch itself. Mid
orange brown sand and mortar. Debris from wall construction.
Unexcavated.
Machine excavated
TRENCH 5
Type:
Max. depth: 1.36m
Context Description
Depth (m)
501
Modern topsoil. Dark grey-brown sandy loam. 2% flint, sub-angular 0.00-0.40
Topsoil
sub-rounded, <1-5cm. Loose and friable. Bioturbated. Homogeneous. bgl
Under grass. Overlies 502.
502
Thin, underdeveloped subsoil. Mid orange-brown sand. 5% flint, Sub- 0.38-0.52
Subsoil
angular sub-rounded, <1-6cm. Fairly loose and friable. Slightly bgl
mixed. Bioturbated. Overlies 510.
503
East west aligned probable double defensive ditch, filled with 0.96 deep
Cut
504 and 505. Straight, moderate sides, concave base. 2.56m
wide. Slightly diffuse in plan and section. To south of parallel
ditch 506. Cuts 510.
504
Secondary fill or possible deliberate backfill of ditch 503. Dark brown 0.56 deep
Deposit
sandy loam. 2% flint, sub-angular sub-rounded, <1-4cm. Slightly
arbitrary division with 505, slightly more orange but very diffuse,
gradual change. Overlies 503.
505
Secondary fill or possible deliberate backfill of ditch 503. Dark brown 0.40 deep
Deposit
sandy loam. 2% flint, sub-angular sub-rounded, <1-4cm. Slightly
arbitrary division with 504, slightly darker but very diffuse, gradual
change. Overlies 504
506
East west aligned probable double defensive ditch, filled with 0.75 deep
Cut
507, 508 and 509. Straight, moderate sides, concave base. 2.46m
wide. Slightly diffuse in plan and section. To north of parallel
ditch 503. Cuts 510.
507
Secondary fill or possible deliberate backfill. Mid brown sandy loam. 0.40 deep
Deposit
2% flint, sub-angular sub-rounded, <1-6cm. Occasional diffuse dark
brown mottles. Slightly arbitrary division with 508 as very diffuse,
gradual change. Overlies 509.
508
Secondary fill or possible deliberate backfill. Dark brown sandy loam/ 0.35 deep
Deposit
sand. 2% flint, sub-angular sub-rounded, <1-6cm. Rare diffuse mid
brown mottles. Bioturbated. Slightly arbitrary division with 507as very
diffuse, gradual change. Overlies 507.
509
Derived from collapse of south edge. Mid orange-brown sand. 5% 0.25 deep
Deposit
flint, sub-angular sub-rounded, <1-8cm. Very mixed bioturbated.
Overlies 506.
510
Natural sand. Mid orange. 5% sub-angular gravel, <1-6cm. Paler 0.30+ bgl
Natural
yellow patches and mottles. Compact.
53
Material
Pottery
Prehistoric
Romano-British
Post-Roman
Ceramic Building Material
Fired Clay
Opus Signinum
Wall Plaster
Mortar
Stone
Worked Flint
Glass
Slag
Metalwork (no. objects)
Coins
Copper Alloy
Lead
Lead Alloy
Iron
Worked Bone
Human Bone
Animal Bone
Marine Shell
Tr 1
263/6577
263/6577
342/63,957
4/360
3/109
7/2237
1/8
1/3
3/173
125
57
22
6
2
38
1/12
735/11484
182/7477
Tr 2
131/2685
1/3
129/2629
1/53
113/20,032
8/738
1/151
1/13
1/121
1/20
28
7
8
13
7/16
149/2106
103/4038
Tr 3
93/2420
91/2386
2/34
54/5749
2/527
2/1110
41
13
5
23
313/7532
81/3343
Tr 4
134/2968
134/2968
104/18,175
1/15
1/323
4/7
47
2
6
39
9/162
117/1433
165/5285
Table 1: Finds totals by material type and by trench (number / weight in grammes)
APPENDIX 2: SUPPLEMENTARY FINDS INFORMATION
Tr 5
33/362
33/362
14/3611
1/4
3
1
2
1/2
10/171
2/42
Total
654/15,012
1/3
650/14,922
3/87
627/111,444
9/753
5/511
4/122
8/2560
4/656
1/4
5/10
6/1303
244
79
27
21
2
115
1/12
17/180
1324/22,726
533/20,185
54


Table 2: Pottery assemblage by trench
Trench
Trench 1
Trench 2
Trench 3
Trench 4
Trench 5
Total
No.
Sherds
263
131
93
134
33
651
%
40.2
20.0
14.2
20.5
5.1
Wt (g)
6577
2685
2420
2968
362
15,012
%
43.8
17.9
16.1
19.8
2.4
EVE
984
301
193
262
30
1770
%
55.6
17
10.9
14.8
1.7
55

Table 3: Pottery quantification by fabric type

Fabric
Grey
Grey, oxidised core/edges
Grey, burnished
Grey, burnished oxidised surface
Grey, micaceous
Grey, micaceous, burnished
Grey, micaceous, oxidised core edges
Grey, micaceous, burnished dark grey/black
Grey, micaceous, burnished dark grey, oxidised core edges
Grey, highly micaceous, burnished dark grey
Misc grey
Dark grey
Dark grey, oxidised core
Dark grey, oxidised core edges
Dark grey, micaceous, oxidised core edges
Dark grey, micaceous, oxidised surfaces
Grey, Rustic
BB1
BB2
LNVCC
LNVCW
OXCC
OXPA?
OXWH?
OXWS
Lower Rhineland
Rhenish
CGS
Amphora
Buff
Mica dusted?
Buff CC
Reddish brown
Reddish-yellow
Shell
Flint gritted
Middle Saxon Ipswich ware
Late Saxon Thetford ware
Modern refined whiteware
Total
No.
sherds
78
9
39
2
5
20
6
87
3
1
3
54
3
7
3
5
18
11
37
90
15
24
1
1
3
3
1
22
18
7
3
1
3
17
49
2
1
1
1
654
%
10.4
6
3.1
13.4
8.1
1.1
2.8
1.7
5.7
13.8
2.3
3.7
3.4
2.8
2.6
7.5
Wt (g)
1251
137
480
10
88
830
200
2045
129
40
114
940
18
259
48
244
226
442
626
1278
735
820
8
4
146
363
2
393
829
278
22
3
244
464
1204
5
53
32
2
15,012
%
8.4
3.4
5.8
1.4
14.3
6.3
1.8
1.7
1.6
3
4.2
8.6
4.9
5.5
6.1
35
38
19
300
21
2
2.1
1.1
17
1.2
22
163
1.2
9.2
11
7
2.7
4.5
16
1
2.4
29
1.6
2.6
5.6
1.9
79
4.5
1.7
3.1
8.1
10
57
126
3.2
7.1
1770
1974/1977 fabric
RW2
RW13
RW2
RW2
%
8.5
47
80
288
16
127
4
Table 4: Pottery fabric concordance for greywares

2012 Fabrics
Grey
Grey, oxidised core
Grey, burnished
EVE
150
33
108
56
7.2

Grey, micaceous
Misc grey
Dark grey
RW4
RW4
RW4
RW1
RW5
RW5
RW10
RW10
RW8
RW14
RW1
RW1
57
Fabric
Grey
Grey, burnished
Grey, micaceous
Misc grey
Dark grey
Grey, Rustic
BB1
BB2
LNVCC
LNVCW
OXCC
OXPCH?
OXWS
Lower Rhineland
Rhenish
CGS
Amphora
Reddish brown
Reddish-yellow
Shell
Total
Table 5: Pottery vessel forms by fabric
2
3
7
67
1
8
2
1
1
1
1
2
3
J
8
4
4
2
4
7
2
3
J/B
31
22
1
1
2
4
15
B/D
2
D
1
1
1
7
BKR
BKR/J
1
J/F
F
1
3
1
L/J
Total
13
4
9
3
5
28
2
3
1
1
9
2
1
1
1
4
7
25
3
13
1
3
1
1
10
5
2
6
7
171
L/D
58


Table 6: Trench 1 Romano-British pottery fabrics

Fabric
Grey
Grey, burnished
Grey, micaceous
Grey, micaceous, burnished dark grey
Grey, micaceous, burnished dark grey, oxidised core
edges
Dark grey
Grey, Rustic
BB1
LNVCC
LNVCW
OXCC
OXPCH?
OXWH?
OXWS
CGS
Reddish brown
Reddish-yellow
Shell
Total
No.
Sherds
47
8
10
2
4
14
6
53
2
1
2
3
5
1
1
30
2
23
1
1
3
4
1
12
27
263
%
17.9
3
3.8
1.5
5.3
2.3
20.2
1.1
1.9
11.4
8.7
1.1
1.5
4.6
10.3
Wt
(g)
738
125
172
10
54
732
200
1046
74
40
40
48
244
20
20
581
150
815
8
4
146
63
20
364
863
6577
%
11.2
1.9
2.6
EVE
110
33
64
%
11.2
3.3
6.5
11.1
3
15.9
25
38
19
195
2.5
3.9
1.9
19.8
1.1
11
1.1
22
7
2.2
8.8
2.3
12.4
153
9
122
4
15.6
2.2
1
29
5
2.9
5.5
13.1
36
102
984
3.7
10.4
3.7
59
12.4

Table 7: Trench 1 Romano-British pottery forms by fabric

Trench 1
Grey
Grey, burnished
Grey, micaceous, oxidised
core edges
dark grey
dark grey, oxidised core
edges
Grey, highly micaceous,
burnished dark grey
Dark grey
Dark grey, micaceous,
oxidised core edges
Dark grey, micaceous,
oxidised surfaces
Grey, Rustic
BB1?
LNVCC
LNVCW
OXCC
OXPCH?
OXWS
CGS
Reddish-yellow
Shell
Total
J
5
4
1
J/B
D
1
B/D
1
BKR
BKR/J
1
J/F
L/D
3
5
Total
8
4
2
3
6
19
1
1
1
1
1
1
1
1
1
13
1
12
1
3
4
5
5
93
1
4
1
1
1
1
1
2
22
12
1
3
3
5
42
60


Fabric
Grey
Grey, burnished
Grey, micaceous, burnished dark grey, oxidised core
edges
Misc grey
Dark grey
Grey, Rustic
BB1
BB2
LNVCC
LNVCW
Lower Rhineland
Rhenish
CGS
Buff
Buff CC
Reddish brown
Shell
Total
No.
Sherds
10
4
11
1
1
28
10
3
11
33
7
2
1
2
2
1
1
2
130
%
7.7
3.1
8.5
21.6
7.7
2.3
8.5
25.4
5.4
1.5
1.5
1.5
1.5
Wt
(g)
161
52
501
55
14
492
124
46
223
412
101
67
2
28
130
3
200
21
2632
%
6.1
2
19
EVE
12
67
4
22.3
2.1
10
3.3
18.7
4.7
1.7
8.5
15.7
3.8
2.5
106
35.2
47
45
15.6
15
14
4.7
4.9
7.6
301

Trench 2
Grey, burnished
Dark grey
Grey, Rustic
BB2?
LNVCC
CGS
Rhenish
Reddish-brown
Total
J
2
1
5
1
1
BKR
2
1
1
1
1
1
11
Total
2
4
5
1
3
2
1
1
1
20
61


Fabric
Grey
Grey, burnished
Grey, micaceous
Dark grey
Grey, Rustic
BB1
BB2
LNVCC
LNVCW
OXCC
CGS
Amphora
Buff
Shell
Total
No.
Sherds
2
5
1
6
6
20
5
2
4
11
5
1
1
15
5
1
1
91
Wt
(g)
26
58
34
98
105
336
233
21
288
279
95
356
5
223
149
73
7
2386
%
2.2
5.5
1.1
6.6
6.6
22
5.5
2.2
4.4
12.1
5.5
1.1
1.1
16.5
5.5
1.1
1.1
%
1.1
2.2
1.4
4.1
4.4
14.1
9.8
EVE
12
10
6.2
5.2
19
35
9.8
18.1
12.1
11.7
4
14.9
27
12
13
14
6.2
6.7
5
60
2.6
31.1
9.3
6.2
3.1
193

Trench 3
Grey
Grey, burnished
Dark grey
Grey, Rustic
BB1?
BB2
LNVCC
LNVCW
OXCC
CGS
Amphora
Total
J/B
1
1
B/D
F
1
L/J
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
2
Total
1
2
2
2
3
1
1
1
2
2
1
1
4
2
25
62


Fabric
Grey
Grey, burnished
Dark grey
Grey, Rustic
BB1
BB2
LNVCC
LNVCW
Lower Rhineland
CGS
Amphora
Buff
Mica dusted?
Reddish brown
Reddish-yellow
Shell
Total
No
Sherds
19
7
14
2
1
4
3
15
23
1
1
1
13
4
2
1
5
18
134
%
14.2
5.2
10.4
1.5
3
2.2
11.2
17.2
9.7
3
1.5
3.7
13.4
Wt
(g)
326
88
334
64
23
51
88
124
200
102
296
79
680
75
9
24
100
305
2968
%
11
3
11.3
2.2
1.7
3
4.2
6.7
3.4
10
2.7
22.9
2.5
3.4
10.3
EVE
40
20
14
%
15.3
7.6
5.3
11
4.2
20
21
74
7
7.6
8
28.2
2.7
10
21
24
262
3.8
8
9.2

Trench 4
Grey
Grey, burnished
Grey, Rustic
BB1?
BB2
LNVCC
LNVCW
Amphora
CGS
Reddish brown
Reddish yellow
Shell
Total
J
3
J/B
D
1
1
1
1
B/D
1
2
2
BKR
1
2
1
5
1
1
1
1
2
9
Total
4
3
2
1
1
2
2
6
1
3
1
1
1
2
30
63


Fabric
Grey, burnished
Dark grey
Misc grey
Grey, Rustic
LNVCC
Mica dusted?
Shell
Flint gritted
Total
No.
Sherds
1
13
3
1
4
1
2
1
3
1
1
2
33
%
3
39.4
9.1
3
12.1
3
6.1
3
9.1
3
3
6.1
Wt
(g)
12
110
59
3
23
3
100
10
16
13
8
5
362
%
3.3
30.4
16.3
EVE
16.7
22
73.3
10
6.4
27.6
2.8
4.4
3.6
2.2
1.4
30
64

Table 15: Quantification of retained CBM by type and by context (fragment

count)
Context
101
102
103
104
105
106
108
113
115
119
121
122
124
126
201
202
207
210
213
216
220
222
301
303
304
306
307
309
315
401
404
405
409
410
501
502
505
TR2 u/s
TR4 u/s
TOTAL
Tegula
15
2
5
4
1
4
3
5
5
Imbrex
1
3
23
5
1
1
4
Box
Flue
6
8
9
Flat Tile
(<30mm)
6
4
3
2
3
11
3
2
4
2
Brick
(>30mm)
Fragment
PostRB
Total
1
36
53
13
3
13
24
13
11
17
16
2
16
7
7
21
1
4
15
3
2
4
2
2
6
6
7
5
7
32
3
35
5
18
1
1
12
1
5
433
12
21
4
7
6
2
2
4
2
3
9
2
1
1
2
9
3
1
3
6
4
1
1
1
1
1
2
2
1
1
2
6
2
2
5
1
1
2
90
14
2
1
2
1
1
3
15
9
2
3
13
2
3
1
5
2
94
1
115
1
2
4
1
11
1
7
105
1
1
2
1
2
7
4
4
1
19
65

Table 16: Coin list

Object Context Type
Issuer / type
Issue
Date
AD 337
Reference
101
Cu Alloy Nummus
101
Cu Alloy Nummus
102
Cu Alloy Nummus
10
201
Cu Alloy Nummus
11
201
Cu Alloy Nummus
15
101
Cu Alloy Nummus
Constantine I/Quadriga. Commemorative issue,

Trier
House of Theodosius/Victory l dragging captive.
Salus Reipublicae type
House of Constantine/Fel Temp Reparatio (Fallen
Horseman) Copy
Eugenius/Winged victory l with wreath. Victoria
Auggg. ?Copy
House of Constantine/Fel Temp Reparatio (Fallen
Horseman) ?Copy
House of Theodosius/Salus Reipublicae type
As LRBC II, 796
Corroded C3-C4 coin. Dated by size alone
AD 388 402
AD 353 360
AD 392 394
AD 353 360
AD 388 402
AD 270 296
AD 330 345
C3 - C4
16
105
Cu Alloy Antoninianus
Radiate copy/Pax Aug reverse. Copy
22
201
Cu Alloy Nummus
Urbs Roma/Wolf and Twins. ?Copy
23
201
24
201
Cu Alloy
Antoninianus/nummus
Cu Alloy Nummus
Corroded C4 coin. Dated by size alone
C4
26
301
Cu Alloy Nummus
Constantius II/ Fel Temp Reparatio (Fallen

Horseman) Copy
Commodus/uncertain reverse
AD 350 360
AD 180
192
AD 332
Copy as LRBC II,

25
/
29
301
Cu Alloy Dupondius
31
301
Cu Alloy Nummus
32
301
Cu Alloy Nummus
Constantine II/2 soldiers, 2 standards. Gloria

Exercitus type. Trier mint
Constantinopolis/Victory on prow. Trier mint
AD 330
LRBC I, 52
36
301
Cu Alloy Nummus
Constantinopolis/Victory on prow. ?Copy
AD 330 345
AD 270 296
AD 347
348
AD 270 296
AD 327328
AD 335
? Copy as LRBC I,
52
/
41
401
Radiate copy/Illegible reverse. Copy
42
301
Cu Alloy Nummus
44
101
Constans/2 facing victories with wreaths.

Victoriaeddauggqnn type. Trier mint
Radiate copy/Illegible reverse. Copy
45
301
Cu Alloy Nummus
49
301
Cu Alloy Nummus
50
301
Cu Alloy Nummus
51
101
Cu Alloy Nummus
57
301
Cu Alloy Nummus
58
301
Cu Alloy Nummus
Constans/2 facing victories with wreaths.

Victoriaeddauggqnn type. Trier mint
House of Valentinian/Winged victory l with wreath.
Securitas Reipublicae type
Constantine II/2 soldiers, 2 standards. Gloria
Exercitus type. Trier mint
Urbs Roma/Wolf and Twins. Copy
AD 348
LRBC I, 163
As LRBC II, 82
AD 364
378
AD
330/331
AD 330 345
C4
59
201
Cu Alloy Nummus
70
101
C3 - C4
101
Cu Alloy
Antoninianus/nummus
71
Radiate copy/Uncertain reverse. Copy
72
101
Cu Alloy Nummus
73
101
Cu Alloy Nummus
75
401
84
125
Cu Alloy
Antoninianus/nummus
Cu Alloy Nummus

Salus Reipublicae type. ?Copy
Valens/ Winged victory l with wreath. Securitas
Reipublicae type
AD 270 296
AD 388 402
AD 364 378
C3 - C4
House of Constantine/2 soldiers, 1 standard.

Gloria Exercitus type. Copy
AD 335 345
Copy as LRBC I,
87
Constantine I/Camp Gate Providentiaeaugg type.

Trier mint
Urbs Roma/Wolf and Twins. Trier mint
LRBC I, 114
Copy as LRBC II,

25
?copy as LRBC II,
171
?copy as LRBC II,
196
?copy as LRBC II,
796
/
?copy as LRBC I,
51
/
LRBC I, 63
LRBC I, 149
/
LRBC I, 38
LRBC I, 85
LRBC I, 56
Copy as LRBC I,
51
/
?Copy as LRBC II,

796
As LRBC I, 340
/
66

85
125
Cu Alloy Nummus
AD 330 335
AD 330 345
As LRBC I, 48
Cu Alloy Nummus
Constantine I/2 soldiers, 2 standards. Gloria

Exercitus type.
House of Constantine/Winged victory on prow.
Constantinopolis reverse. Mule. Almost certainly a
copy
86
125
Cu Alloy Nummus
87
125
C4
88
125
Cu Alloy Nummus
Urbs Roma/Wolf and Twins. Copy
89
125
Copy as LRBC I,
51
/
90
125
Cu Alloy Nummus
91
125
Cu Alloy Nummus
92
125
Cu Alloy Nummus
House of Constantine/2 soldiers, 2 standards.

Gloria Exercitus type. Lyons mint
Urbs Roma/Wolf and Twins. Lyons mint
AD 330 345
AD 270 296
AD 330 345
AD 330
AD 330
LRBC I, 184
93
125
Cu Alloy Nummus
AD 339
LRBC I, 131
94
125
95
125
Cu Alloy Nummus
96
125
Cu Alloy Nummus
97
125
Cu Alloy Nummus
98
125
Cu Alloy Nummus
99
125
Cu Alloy Nummus
100
125
101
125
Cu Alloy Nummus
108
301
Cu Alloy Nummus
110
101
111
Constans/2 soldiers, 1 standard. Gloria Exercitus

type. Lyons mint
Copy as LRBC I,
48
As LRBC I, 180
Cu Alloy Nummus
AD 270
296
AD 335Gloria Exercitus type.
345
Constantinopolis/Victory on prow. Copy
AD 330
345
AD 335
Gloria Exercitus type.
345
House of Constantine/2 soldiers, 2 standards.
AD 330 Gloria Exercitus type. ?Copy
345
Constantius II/2 soldiers, 1 standard. Gloria
AD 335 Exercitus type.
341
AD 270 296
Constantinopolis type/2 soldiers, 1 standard. Gloria AD 335 Exercitus type. Mule. Almost certainly a copy
345
AD 330 345
C4
101
Cu Alloy Nummus
116
101
Cu Alloy Nummus
Copy as LRBC I,
52
As LRBC II, 796
117
101
Cu Alloy Nummus
119
301
Cu Alloy Nummus

Salus Reipublicae type
Arcadius/ Winged victory l with wreath. Victoria
Auggg. ?Copy
Constantinopolis/Victory on prow. ?Copy
144
201
Radiate Copy/uncertain reverse
154
125
Cu Alloy Nummus
155
125
Cu Alloy Nummus
156
125
Cu Alloy Nummus
157
125
Cu Alloy Nummus
Constantinopolis/Victory on prow. Mint Mark: PTR

Trier mint, Copy
Constantinopolis/Victory on prow. Mint Mark: SLG
Lyons mint
AD 330 345
AD 388 402
AD 388 402
AD 330 345
AD 270 296
AD 330 345
AD 330 345
AD 331
C4
158
125
Cu Alloy Nummus
159
125
Cu Alloy Nummus
Urbs Roma/Wolf and Twins.
Copy as LRBC I,
52
As LRBC I, 51
160
125
Cu Alloy Nummus
161
125
Cu Alloy Nummus
Constantine II/2 soldiers, 1 standard. Gloria

Exercitus type.
AD 330 345
AD 330 335
AD 330 335
C3 - C4
162
125
Cu Alloy Nummus
C3 - C4
As LRBC I, 87
Copy as LRBC I,
52
As LRBC I, 87
?Copy as LRBC I,
48
As LRBC I, 89
/
/
Copy as LRBC I,
52
/
?Copy as LRBC II,

164
Copy as LRBC I,
52
/
Copy as LRBC I,
52
Copy of LRBC I,
52
LRBC I, 191
As LRBC I, 226
/
67

163
125
Cu Alloy Nummus
164
125
Cu Alloy Nummus
165
125
Cu Alloy Nummus
166
125
Cu Alloy Nummus
167
125
Cu Alloy Nummus
168
125
Cu Alloy Nummus
169
125
Cu Alloy Nummus
170
125
Cu Alloy Nummus
171
125
Cu Alloy Nummus
172
125
Cu Alloy Nummus
173
125
Cu Alloy Nummus

Gloria Exercitus type.
House of Constantine/ Soldier spearing a fallen
horseman. Fel Temp Reparatio type
Urbs Roma/Wolf and Twins. ? Copy
174
125
Cu Alloy Nummus
175
125
Cu Alloy Nummus
176
125
177
125
Cu Alloy
Antoninianus/nummus
Cu Alloy Nummus
AD 335 345
AD 348 360
AD 330 345
Constantinopolis/Victory on prow. ? Copy
AD 330 345
AD 330 345
AD 330 345
Urbs Roma/Wolf and Twins.
AD 330 335
Urbs Roma/Wolf and Twins. Mint Mark LG. Lyons AD 330 mint
335
C4
As LRBC I, 87

Gloria Exercitus type. ? Copy
AD 335 345
AD 335 345
C4
?Copy as LRBC I,
87
Copy as LRBC I,
87
/

AD 335 345
C3 - C4
Copy as LRBC I,
87
/
C3 - C4
As LRBC II, 25
?Copy as LRBC I,
51
?Copy as LRBC I,
52
Copy as LRBC I,
52
Copy as LRBC I,
52
As LRBC I, 51
As LRBC I, 184
/
Graph 1: All coins from the site
Coins from Brancaster (Time Team) by period (after Reece 1991)

50
45
40
35
No of cois
30
Copies
official coins
25
20
15
10
5
C
4
C
2
-C
4
3
C
-u
to
A
-4 D4
1
1
3 to
-5 5
4
4 4 to
-6
68
9
5
- 9 to
9
6 6t 6
-1 o
1 11
7 7t 7
-1 o
3 13
8 8t 8
-1 o
61 16
1
9
- 1 to
10 80 180
t
-1 o
11 93 192
- 2 to
12 22 222
- 2 to
2
13 38 38
- 2 to
2
14 6 0 6 0
- 2 to
15 7 5 275
- 2 to
2
16 96 96
- 3 to
3
17 17 17
- 3 to
18 30 3 3 0
- 3 to
3
19 4 8 48
- 3 to
2 0 6 4 36 4
- 3 to
21 7 8 378
- 3 to
8 38
H
ou C 8 to 8
se on
4
of sta 02
C nti
o n ne
st
an I
ti
C ne
1
-C
2
C
1
-C
3
Period
68
-u
to
A
-4 D4
1
1
3 to
-5 5
4 4 to 4
-6
6
5 9 to 8
6 96 t 96
-1 o
1
7 17 1 7
- 1 to
1
3
8 8 38
- 1 to
1
9 61 61
to
10 1 80 18
- 1 to 0
11 93 19
to 2
12 222 22
- 2 to 2
1 3 3 8 23
to 8
1 4 26 0 26
to 0
15 27 5 2 7
- 2 to 5
16 9 6 2 9
to 6
17 317 31
to 7
18 330 33
- 3 to 0
19 48 34
to 8
2 0 3 6 4 36
- 3 to 4
21 7 8 37
- 3 to 8
3
H
ou C 88 88
se on to
s
of ta 402
C nt
on i n
st e I
an
t
C ine
1
C C2
1
-C
3
C C2
3
-C
Po
4
st
m C
ed 4
ie
M val
od
er
n
No of coins
-u
2
to
A
-4 D4
1
1
3 to
-5 5
4 4 to 4
-6
6
5 9 to 8
6 96 t 96
-1 o
1
7 17 t 17
-1 o
1
3
8 8 38
- 1 to
1
9 61 61
to
1 0 1 8 0 18
- 1 to 0
11 93 19
to 2
12 222 22
- 2 to 2
1 3 3 8 23
to 8
1 4 2 6 0 26
to 0
1 5 27 5 27
- 2 to 5
16 9 6 29
to 6
17 317 31
to 7
18 330 33
- 3 to 0
19 48 34
to 8
2 0 3 6 4 36
- 3 to 4
21 7 8 37
- 3 to 8
3
H
ou C 88 88
se on to
s
of ta 402
C nt
on i n
st e I
an
t
C ine
1
C C2
1
-C
3
C C2
3
-C
Po
4
st
m C
ed 4
ie
M val
od
er
n
No of coins

Graph 2: Probable hoard from layer 125
Coins recovered from layer 125
35
30
25
20
15
Copies
official coins
10
Period (after Reece 1991)
Graph 3: Coins from Brancaster (without the coins from layer 125)
Coins from Brancaster (without the probable hoard in layer 125)
14
12
10
8
Copies
official coins
Period (after Reece 1991)
69

Table 17: Marine shell by context

Trench
Context
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
3
3
3
3
4
4
4
4
4
5
Total
103
104
105
106
106 <3>
108
109
113
121
122
125
125 <1>
126
201
202
206
210
216
301
302
303
304
306
307
314
315
316
401
402
405
409
410
504
OY
LV
18
1
0
13
2
19
OY
UMLV
8
0
1
0
5
3
OY
RV
15
0
0
1
4
19
OY
UMRV
2
0
0
0
1
7
OY
MNI
26
1
1
13
7
26
13
17
1
1
4
1
13
22
9
0
1
0
2
0
13
4
5
0
2
0
52
0
3
1
1
2
203
1
0
0
0
2
9
0
0
1
0
5
0
4
0
1
0
0
1
24
1
0
0
0
0
70
0
0
8
0
8
14
3
2
3
2
0
2
10
5
8
1
0
2
36
0
1
1
4
0
157
0
0
0
0
1
8
0
0
0
0
2
0
7
0
0
0
1
0
14
0
0
0
0
0
46
2
1
8
1
15
31
9
2
3
2
7
2
17
5
8
1
2
2
76
1
3
1
4
2
296
Whelk
Periwinkle
Cockle
MNI
1
1
2
1
Mussel
MNI
2
1
1
1
1
1
1
1
1
3
12
17
KEY: OY = oyster, LV = left valve, RV = right valve, UM = unmeasurable, MNI = minimum number of
individuals
70
Total
28
1
2
18
7
27
3
19
1
2
1
8
1
15
31
10
3
3
2
7
4
17
5
9
1
3
2
88
1
3
1
4
2
329
RV
15
19
14
10
35
93
42.9
Context
103
108
202
304
401
Total
%
18
19
22
13
52
124
57.1
LV
18
26
17
15
34
110
50.7
Polydora
Ciliata
5
5
4
4
7
25
11.5
Cliona
celata
2
2
0.9
Calcareous
tubes
Table 18: Oyster shell analysed in more detail
3
3
3
9
4.1
Barnacles
3
2
1
6
12
5.5
Thin
5
1
3
5
14
6.5
Thick
1
3
2
3
2
11
5.1
Chambered
4
3
5
11
23
10.6
Chalky
dep.
5
10
12
7
23
57
26.3
Worn
6
1
3
7
17
7.8
Flaky
4
6
2
3
7
22
10.1
8
8
10
10
22
58
26.7
Irreg
shape
15
15
12
5
27
74
34.1
Notches
Oysters
attached
71

21
29
22
15
41
128
59.0
Shell
infested

Table 19: Comparative size of analysed oyster shell

Context
103
103
103
103
108
108
108
108
202
202
202
202
304
304
304
304
401
401
401
401
Valve
LV width
LV length
RV width
RV length
LV width
LV length
RV width
RV length
LV width
LV length
RV width
RV length
LV width
LV length
RV width
RV length
LV width
LV length
RV width
RV length
Number
18
18
15
15
19
19
19
19
22
22
14
14
13
13
10
10
52
52
35
35
Mean
81.05
79.7
81.5
76.7
81.9
78.7
79.8
72.5
80.1
75.5
72.1
67.3
88.8
90.3
80.8
78.9
81.5
78.9
76.6
71.4
Min
64
50
70
62
55
56
62
58
60
57
50
45
75
70
60
58
55
55
58
50
Max
105
90
92
90
110
113
100
88
105
118
90
90
120
135
110
115
102
108
120
100
Graphs 4-8 showing size of analysed oyster shells by context
103
5
4.5
4
3.5
Left
Right
2.5
2
1.5
1
0.5
0
6569
7074
7579
8084
size
8589
9094
9599
Left
100- 105104 109
Number of shells
110
72

108
5
4.5
4
3.5
3
2.5
2
Left
Right
1.5
1
0.5
0
50-54 55-60
60-64 65-69
70-74
75-79
80-84
85-89
90-94
95-99
100104
Right
Left
Number of shells
105109
110
202
4
3.5
3
Left
Right
2.5
2
1.5
1
0.5
0
50-54 55-60
60-64 65-69
70-74
75-79
Size
80-84
85-89
90-94
95-99
100104
Right
Left
Number of shells
105109
110
73

304
3.5
3
2.5
2
Left
Right
1.5
1
0.5
0
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
100104
Size
105109
Right
Left
Number of shells
110115
115 >
401
Left
Right
0
55-60
60-64
65-69
70-74
75-79
80-84
Size
85-89
Right
90-94
95-99
Left
100-104
105-109
Number of shells
110
74
Context
Layer
125
Sample
10
Ltrs
Vol.
106
415
10
15
125
60
Flot
(ml)
% roots
Grain
Chaff
A*
Other
Charcoal
>4/2mm
12x barley, Raphanus raphanistrum ++,

Galium aparine, Fallopia convolvulus, 1/1ml
Rumex sp. Chenopodium sp. Atriplex,
Heather stems x4-5, some hulled

wheat, grains and glumes,
several
grains of hulled barley x5, Fallopia
convolvulus, Avena sp., Fumaria sp. 2/3ml
Carex sp., Bromus sp., Galium aparine,
Prunus sp. Corylus avellana. Danthonia
sp.
Charcoal includes small round wood.

Hulled wheat, glume bases, Torilis, Poa
Vicia, Ranunculus Plantago lanceolata,
Charred stems. thorns of Prunus
spinosa/Crataegus
monogyna.
Mineralised seeds of Spergularia 6/8ml
marina, Myosotis sp., Lolium sp.,
Agrostemma, Rumex, Daucus, Torilis
type,
Bromus.
Reseda,
Orache,
Aethusa,
Polygonum,
Centaurea,
Urtica/Aphanes fly larvae,
Comments
Charred Plant Remains
Flot
Moll-t (C)
Smb (C)
Moll-f (B)
Moll-t (A)
Molf (A)
Mollt (A)
Min (A
Eel (C)
Smb (A)
Other
Analysis
75
Key: A*** = exceptional, A** = 100+, A* = 30-99, A = >10, B = 9-5, C = <5; Smb = small animal bones, Moll-t = terrestrial molluscs, Moll-f = aquatic molluscs;
Analysis: P = plant
Layer
Trench 4 Romano-British
Demolition layer
Trench 1 Late Romano-British-Post-Roman
Occupation
Strong-room
Trench 1 Romano-British
Feature
Samples
Table 20: Assessment of the charred plant remains and charcoal
APPENDIX 3: PALAEOENVIRONMENTAL RESULTS


Table 21: Land and aquatic molluscs and marine shell assessment
Site Phase
RB
LRB-Post RB
RB
Feature type
layer
layer
layer
Context no.
125
106
415
Sample no.
Volume (L)
10
10
Pupilla muscorum
Vertigo spp.
Helicella itala
Vallonia costata
Trench
Open country species
Vallonia excentrica
cf. Cochlicella acuta
Intro. Helicellids
Trochulus hispidus
Cochlicopa spp.
Cepaea spp
Oxychilus cellarius
Aegopinella nitidula
Clausilia bidentata
Myosotella myosotis
Anisus leucostoma
Hydrobia spp.
100+
100+
Oyster
Frags
Frags
Mussel
Frags
Limpet
Frags
Intermediate species
Shade-loving species
Marsh-loving species
Succinea/Oxyloma sp.
Aquatic species
Burrowing species
Cecilioides acicula
Approx totals
Marine shells
Periwinkles
Key: A = >10, B = 9-5, C = <5; + = fragment present
76
Trench 4
S
H
J
I
F
G
M
N
T
Trench 2
344000
B
A
D
C
Trench 1
50 m
A419
Geophysics data reproduced by permission of GSB Prospection LTD

Digital data reproduced from Ordnance Survey data Crown Copyright (year) All rights reserved. Reference Number: AL100018665
Contains Ordnance Survey data Crown Copyright and database right 2014.
This material is for client report only Wessex Archaeology. No unauthorised reproduction.
Date:
Scale:
Path:
GPR survey results
19/08/2014
1:1000 @ A4
Revision Number:
Illustrator:
0
RG
X:\PROJECTS\85209\Graphics_Office\Rep figs\assessment\2014_08_15\85209_master.dwg
Figure 3
Section 1
112
S
133
101
102
Section 1
14.32 mOD
112
134
106
116
117
118
119
133
134
106
West-facing section of wall 128, robber cut 127 and adjacent deposits
111
106
106
140
Position of lorica (ON 83)
137
130
Wall
132
Surface
131
Ceramic building material

(section only)
123
113
107
Section 2
Section
2
N
128
104
143
104/108
14.75 mOD
107
141
114
142
121
102/104/
108
120
115
122
126
136
142
136
141
114
Wall 128
127 (Robber cut)
115
East-facing section of sondage at north end of trench and possible surface 134
105
1m
5m
Scale for plan
Scale for sections
This material for client report only Wessex Archaeology. No unauthorised reproduction
Date:
Scale:
Path:
Trench 1 Survey and sections
19/08/2014
Plan @ 1:80 section @ 1:40
Revision Number:
Illustrator:
0
RG
X:\PROJECTS\85209\Graphics_Office\Rep figs\assessment\2014_08_15\85209_trench01.ai
Figure 4
Trench 1, plates
Illustrator:
Revision Number:
RG
X:\PROJECTS\85209\Graphics_Office\Rep figs\assessment\2014_08_15\85209_trench01plates.ai
NTS
Path:
19/08/2014
Date:
Scale:
Plate 2: Central area of Trench 1 , view from the south
Plate 4: Lorica squamata in situ adjacent to wall, fragment after cleaning, x-ray plate of fragments
Plate 1: Southern end of Trench 1, view from the south-east
Figure 5
Plate 3: West-facing section adjacent to wall 130, obli ue view from the south-west
203
202
203
208
207

209
204
Section 3
205
211
214
213
206
210
Natural
feature
Plate 7: East-facing section features 208, 223 and wall 205, obli ue view
from the north-east
212
Wall/structure
220
Surface
215
219
Ceramic building material

(section only)
Section 3
S
N
11.53 mOD
201
206
205
222
204
224
204
202
207
210
209
221
211
223
208
East-facing section features 208, 223 and wall 205

0
2.5 m
Plate 8: Southern end of Trench 2, view from the south
Scale for plan and section
Date:
Scale:
Path:
Trench 2, survey, sections and plates
19/08/2014
Plan
section @ 1:40
Revision Number:
Illustrator:
0
RG
Figure 6
305
313
Section 4
306
Road metaling 302/316
312
315
316
309
308
313
305
West-facing section through ditches 313 and 305
314
Section 4
308
310
309
303
306
316
Surface
Scale for section
317
Sand lens
Scale for plan
1m
1m
12.98 mOD
Tree-throw
Illustrator:
Revision Number:
RG
Figure 7
Plan @ 1:40, section @ 1:20

Path:
21/08/14
Date:
Scale:
Plate 10: East-facing section of 312
Plate 9: Trench 3, view from the east
Section 5
S
418
10.35 mOD
401
402
413
415
402
403
414
405
409
404
408
East-facing section of wall 403 and associated rampart, robber cut 408 and ditch 406
1m
0
404
Scale for section
409
Wall
405
420
407
Section 5
403
414
413
ixed interface between 413
419
419
410
416
412
410
Area of robbing 411

412
416
417
1m
410
412
Scale for plan

410
Plate 12: Southern end of Trench 4, view from the

south-west
Date:
Scale:
Path:
21/08/2014
Revision Number:
Illustrator:
0
RG
Figure 8
Section 6
506
507
505
Plate 13: Trench 5, view from the south-west
503
Section 6
N
S
8.27 mOD
501
502
505
508
509
507
504
503
506
West-facing section through ditches 503 and 506

1m
0
Scale for plan and section
Date:
Scale:
Path:
Trench 5, survey, section and plate
21/08/2014
Revision Number:
Illustrator:
0
RG
Figure 9
Western
Vicus
?Barracks
Principia
A419
Trench 1
Via Principalis
Granary
Trench 4
Barracks
Trench 2
?Mansio
?Bath house
Digital data reproduced from Ordnance Survey data Crown Copyright (year) All rights reserved. Reference Number: AL100018665
Contains Ordnance Survey data Crown Copyright and database right 2014.
Geophysics data reproduced by permission of GSB Prospection LTD
Trench 5
?Early Fort
Postulated layout of Brancaster complex based on excavation, geophysical survey and cropmark evidence
Postulated layout
Cropmark
378000
RG
Figure 10
X:\PROJECTS\85209\Graphics_Office\Rep figs\assessment\2014_08_15\85209_master.dwg
Path:
Illustrator:
1:2000 @ A3
Revision Number:
18/08/2014
100 m
344000
Date:
Trench 3
Eastern
Vicus
Scale:
?Possible Camp
High water mark
Wessex Archaeology td registered office Portway ouse, Old Sarum Par , Salisbury, Wiltshire SP4 6E
Tel: 01722 326867 Fax: 01722 337562 [email protected]
www.wessexarch.co.u
Wessex Archaeology td is a company limited by guarantee registered in England, company number 1712772. It is also a Charity registered in England and Wales,
number 287786 and in Scotland, Scottish Charity number SC042630. Our registered office is at Portway ouse, Old Sarum Par , Salisbury, Wiltshire SP4 6E .

Time Team - Brancaster Roman Fort

Uploaded by

Copyright:

Available Formats

Time Team - Brancaster Roman Fort

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Time Team - Brancaster Roman Fort

Uploaded by

Copyright:

Available Formats

making sense of heritage

Brancaster Roman Fort, (Branodunum),

BRANCASTER ROMAN FORT

Videotext Communications Ltd

Report reference: 82509.01

Scheduled Ancient Monument number (old county style) NF 208

Wessex Archaeology Limited 2014 all rights reserved

Brancaster Roman Fort, Norfolk

* I= Internal Draft E= External Draft F= Final

WA Project No. 85209

Brancaster Roman Fort, Norfolk

BRANCASTER ROMAN FORT

Archaeological Evaluation and Assessment of Results

AIMS AND OBJECTIVES ...................................................................................3

PALAEO-ENVIRONMENTAL SUMMARY ........................................................30

WA Project No. 85209

Brancaster Roman Fort, Norfolk

POTENTIAL AND FURTHER RECOMMENDATIONS .....................................37

APPENDIX 1: TRENCH SUMMARIES.......................................................................47

WA Project No. 85209

Brancaster Roman Fort, Norfolk

Location of Site, trenches and geophysical survey areas including

Front Cover Working shot, Trench 1

WA Project No. 85209

Brancaster Roman Fort, Norfolk

BRANCASTER ROMAN FORT

Wessex Archaeology was commissioned by Videotext Communications Ltd to

WA Project No. 85209

Brancaster Roman Fort, Norfolk

BRANCASTER ROMAN FORT

WA Project No. 85209

Brancaster Roman Fort, Norfolk

BRANCASTER ROMAN FORT

Wessex Archaeology was commissioned by Videotext Communications Ltd

This report documents the results of archaeological survey and evaluation

The Site, location and geology

Brancaster Roman Fort is a Scheduled Ancient Monument (old county

The fort area is still visible as a distinct raised plateau at a height of

WA Project No. 85209

Brancaster Roman Fort, Norfolk

Archaeological Background and Previous Archaeological Work

WA Project No. 85209

Brancaster Roman Fort, Norfolk

Some brief unpublished notes suggest the presence of Roman structures in

As well as the evidence for Romano-British activity, Neolithic and Mesoltihc

AIMS AND OBJECTIVES

A project design for the work was compiled (Videotext Communications

Research Aim 1: What is the character of the archaeology

Research Aim 2: What is the chronological sequence of fort

Research Aim 3: Does any evidence survive for shore-side

Prior to the excavation of evaluation trenches, a geophysical survey was

WA Project No. 85209

Brancaster Roman Fort, Norfolk

Five trenches of varying sizes were excavated, their locations determined in

The trenches were excavated using a combination of machine and hand

At various stages during excavation the deposits were scanned by a metal

All archaeological deposits were recorded using Wessex Archaeologys pro

A full photographic record of the investigations and individual features was

This report may contain material that is non-Wessex Archaeology copyright