Beruflich Dokumente
Kultur Dokumente
Doctor of Philosophy
at the University of Leicester
by
Alfonso Burgers BA
(Cape T o w n ) , MS
(Purdue)
October 1997
Dissertation Publishing
UMI U095284
Published by ProQuest LLC 2013. Copyright in the Dissertation held by the Author.
Microform Edition ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
ABSTRACT
TABLE OF CONTENTS
VOLUME 1
Page
Table of Contents
List of Figures
vi
List of Tables
viii
Acknowledgements
xii
Chapter 1.
The Background: Historical content of water supply
1. Introduction
2. Objectives
3. Database
4. Water supply
4.1.
Chronology.
4.2.
Roman Britain.
6. Social contexts
10
6.1.
The Romanizationdebate.
10
6.2.
11
6.3.
12
6.4.
Religious aspects.
12
7. Conclusion
15
Chapter 2.
Methodology, database, floods and Roman engineering in
ancient texts
16
1. General
16
16
3. Data collection
16
4. Database recording
18
23
5.1.
Aqueducts.
24
5.2.
Roman baths.
25
5.3.
Wells.
25
5.4.
Drains.
26
5.5.
Pipes.
26
5.6.
Tanks.
28
5.7.
Springs.
28
5.8.
Sewers.
28
6. Typology of sites
29
6.1. Forts.
30
31
34
34
35
37
38
39
41
9. Floods
43
46
48
Chapter 3.
Water supplies: 1) Aqueducts, springs and dams
52
52
2. Springs
53
3. Dams
54
4. Roman surveying
55
57
58
5.1.1. Leats.
58
59
59
5.2. Tunnels.
61
62
63
64
64
7. Site types
66
67
8. Aqueducts
68
8.1. General.
68
69
71
9. Inscriptions
72
73
74
76
ii
76
76
82
86
88
92
93
12.
94
Conclusion
Chapter 4.
Water supplies: 2) Rainwatercatchment and wells
95
1. General
95
2. Rainwater catchment
95
96
2.1.1. Forts.
96
97
3. Wells
99
3.1. General.
99
3.2.
3.3.
wells.
101
103
4. Steining of wells
104
4.1.
107
4.2.
109
115
5. Stability of wells
116
118
119
121
7.1.
Forts.
121
7.2.
122
7.3.
125
7.4.
Civitas capitals.
128
7.5.
Small towns.
13 0
7.6.
Settlements.
131
7.7.
Villas.
131
132
8.1.
132
8.2.
133
8.3.
137
9. Conclusion
138
iii
Chapter 5.
Other water-related features: Baths, drains and sewers
141
1. General
141
2. Roman Baths
143
2.1.
Types of baths.
144
2.2.
147
149
152
152
158
169
171
3.1. Terminology.
171
171
3.2. Drains.
17 6
4. Conclusion
180
Chapter 6.
The distribution of water-related features in Roman Britain
182
1. General
182
186
189
189
3. Towns
195
3.1.
Coloniae.
197
3.2.
Municipia.
206
3.3.
Civitas capitals.
211
4. Small
towns
5.
213
Settlements
213
6. Villas
219
7.
221
7.1.
at military sites.
221
7.2.
at villas.
223
7.3.
224
8.
Aqueducts
227
9.
Baths
229
10. Wells
230
iv
231
232
234
14. Conclusion
23 8
Chapter 7.
Some social, economic, organizational and construction aspects of
water-related features
255
1. Introduction
255
2. Romanization
255
258
260
study
264
264
267
268
272
273
277
281
5. Digging of wells
283
285
288
292
8. Conclusion
293
Chapter 8.
Conclusions and Recommendations
294
1. Conclusions
294
2. Recommendations
299
VOLUME 2
Appendices
Appendix 1: Data from database
3 02
317
474
Appendix 4: Abbreviations
490
Appendix 5: Bibliography
497
(Last p a g e
vi
537)
LIST OF FIGURES
Chapter 3.
Water supplies: 1) Aqueducts, springs and dams
52
Fig.
56
Fig.
60
Fig.
64
Fig.
70
Fig.
71
Fig.
72
Fig.
77
Fig.
77
79
Fig.
85
Fig.
89
Fig.
93
Chapter 4.
Water supplies: 2) Rainwater catchment and wells
95
99
107
Well 1 at villa
Fig. 4.4b:
Well 1 at villa
timber lining
110
112
113
Fig. 4.4c:
Well 2 at villa
Fig. 4.4d:
Well 2 at villa
timber lining
114
115
Fig.
124
Fig.
127
Fig.
128
Chapter 5.
Other water-related features: Baths, drains and sewers
141
148
vii
(b) Bearsden,
Fig
(c) Bothwellhaugh
(c) Vindolanda,
154
(a) Red House,
(d) Benwell
155
Fig
156
Fig
159
Fig
160
Fig
161
Fig
163
Fig
165
Fig
166
Fig
168
Fig
172
Fig
174
Fig
175
Fig
5.15: Open channel drain and lead pipe from Caerleon fortress 177
Chapter 6.
The distribution of water-related features in Roman Britain
182
Fig
240
Fig
241
Fig
242
Fig
243
Fig
244
Fig
245
Fig
246
Fig
247
Fig
248
Fig
249
Fig
250
Fig
251
Fig
252
Fig
253
Fig
254
Fig
192
Fig
184
Fig
198
Fig
203
207
area, London
2 08
235
Chapter 7.
Some social, economic, organizational and construction aspects of
water-related features
255
Fig.
259
Fig.
260
260
Fig.
263
Fig.
264
Fig.
27 0
Fig.
278
Fig.
280
ix
LIST OF TABLES
Chapter 2.
Methodology, database, floods and Roman engineering in
ancient texts
16
17
17
33
40
Chapter 3.
Water supplies: 1) Aqueducts, springs and dams
52
67
83
84
Chapter 4.
Water supplies: 2) Rainwater catchment and wells
95
96
100
101
109
129
131
Chapter 5.
Other water-related features: Baths, drains and sewers
141
150
169
Chapter 6.
The distribution of water-related features in Roman Britain
182
182
184
185
186
213
217
219
221
223
224
236
Chapter 7.
Some social, economic, organizational and construction aspects of
water-related features
255
267
274
276
284
xi
ACKNOWLEDGEMENTS
Many persons have contributed to my academic studies during the past
decade since my retirement. In particular I am indebted to Prof. John
Atkinson, Dean of the Faculty of Arts, University of Cape Town, and Dr
Margaret
Mezzabotta
of
the
Classics
Department,
UCT,
who
first
my
supervisor,
I owe
special
debt
of
thanks
for
his
complexity
experience.
To
of
Dr
archaeology,
Lin
Foxall
subject
I also
owe
new
to
special
my
academic
thank
you
for
intricacies
whose help I would not have been able to produce this thesis.
without
I also
thank all the staff of the School of Archaeological Studies and the
postgraduate students for their assistance whenever needed.
The University of Leicester offered me a unique opportunity to study
as a mature student,
Sciences
Research
Council
of
South
Africa
the
award
of
the
Inez Burgers,
fifty years,
BA.LLB,
and artist,
last
and support during the last four years. To her I dedicate this thesis.
CHAPTER 1.
THE BACKGROUND: HISTORICAL CONTEXT OF WATER SUPPLY
1. INTRODUCTION
With the introduction of aqueducts to Rome in the late 4th century BC
the
Romans
water
for
was
The
Roman
had
army
developed
introduced
desire
into
the
to
new
have
unpolluted
province
of
Roman
Britain
organized water supply at their fortresses and forts. Where towns and
new settlements developed under Roman rule, organized water supplies
were also introduced.
have
been
found
in
Amongst
Britain
to
the
of structures
Romano-British
that
period,
following the
settlement
but
engineering
civilization.
improvements
is
involved,
The
improvement
on
nature,
especially
constitutesa
of
when water
characteristic
of
the
to
their
homes
or
to
from the
specially
ability
of man
constructed
public
to bring
bathing
facilities.
2. OBJECTIVES
In this thesis I shall bring together much of the available evidence
from the archaeological
record
on water-related
features and discuss the possible impact they had on British society.
The
thesis
drainage,
is
also
and their
about
the
mechanics
of
water
distribution at Romano-British
provision
sites.
and
It also
seeks to address the social and economic impact of water supply and
its use. In the discussion I shall look at:
1. the sites where water-related structures have been found;
2. the type of water-related structures;
3. the inter-dependence of these structures upon each other;
4. the distributions of different water-related features at
different categories of sites.
Arising out of this information some questions are relevant, such as:
(a) how reliable are the data?
(b) how representative are the data for all the Romano-British
sites known from the archaeological record?
(c) are current interpretations of some of the Romano-British
aqueducts and wells acceptable?
(d) how to reconcile older and more recent excavation reports
on water-related
structures
such
as
aqueducts,
baths
and
wells?
(e) what is the relevance of water supply to a community?
This
thesis
structure
will
not
associated
give
with
detailed
water
descriptions
supply,
rather
of
each
site
it
will
or
examine
fortlets,
civitates,
small
and
chartered
towns,
towns
settlements
(municipia
and
and
villas.
coloniae),
There
are
in chapter
2.
The
categories
of water-related
structures
record,
Three months
project.
The hope is to have the database published in a gazetteer
provide
easy
reference
to
the
data
set
on
the
format to
water-related
3. DATABASE
The database
the
available
water-related
sites
have
been
processed.
The
Site
types
settlements
with
the
category
of
problems
in Chapter
to
instance,
and villas.
problems
seems
for
those
be
of
how
Scholarship
some
sites.
2.
sites,
I
shall
For
reasonable
to
categorize
small
is generally not
mainly
discuss
fortresses,
because
some
forts
agreement. Large
towns
in agreement
of
definitional
aspects
and
towns,
of
these
fortlets,
there
such
as
the
four
seem to present
problems
of
either.
Since
John Wacher's
seminal book The Towns of Roman Britain appeared in 1974, there has
been
an
intense
debate
on
how
to
define
settlements,
towns,
and
definitions.1
The
authors
analyse
their
development,
morphology,
in
the first five chapters of their book on 'small towns', discuss the
above aspects of towns, but have little to report on water supply or
wells.
As
an example,
Neatham
less
than
eleven wells, but despite the fact that there were so many wells for
the single community,
there
1 Brown (ed.) 1995. It is striking h ow the nineteen authors of articles agree on the
lack of a definition, but that they turn up with almost as many definitions as there
are authors.
2 . Data types :
2a. Aqueducts, for instance,
materials and the archaeological record is not always clear what was
the form of the aqueduct. Uncertainty often exists whether a stone
channel
was,
in
fact,
the aqueduct
or whether
it
only
carried
most
commonly
excavated
area
of
site,
providing
obvious
at
and it is
ritual
have been excavated they have often been recorded primarily for the
small finds found in them.
2d. Drains and sewers are often confused with each other.
2e. Tanks
are
sometimes
only
inferred
from
base
that
has
been
identified.
These
are problems
assembled.
which
The definitions
revealed
themselves
of categories
as
the
database
of both sites
was
and water-
4. WATER SUPPLY
The different types of water-related features will be presented in a
series of chapters: aqueducts and springs (Chapter 3), wells and rain
water catchment (Chapter 4), baths, drains and sewers (Chapter 5).
4.1 Chronology.
Water
supplies
from
antiquity
came
in
variety
of
forms.
The
Man-made
wells
instance,
wells
was
also
an
early
form
of
water
supply,
as
(c.
3000
BC) ,
or
the
religious
wells
from
the
Indus
for
from
from as
early as the later Bronze Age.4 The palace-temples of the Bronze Age
such as at Knossos,
In
Romans
(Winslow
1963,
171-6)
dating
from
the
4th
century
BC
through into the 4th century AD (Hodge 1992, 92). Many of these Greek
and
Roman
countries
remains
where
have
they
become
are
national
situated.
Yet,
monuments
modern
in
the
scholarship
modern
gives
aqueducts
incorporated
large
bridge
structures
as
found
elsewhere.
Water is a prime social need and the search for water supplies must
have been of great importance to ancient communities, both during the
prehistoric
period
and
later.
Food
could be
obtained
from
remote
sources, but in antiquity water was obtained nearly always from local
sources such as a nearby stream, a spring, or from a purposely sunk
well
from the Bronze Age of spring-heads being specially adapted for ease
of obtaining water and these would appear to have often developed as
religious shrines.
2 Genesis 21.15, 26.15, 21.18-21. Abraham had wells dug in the Neg ev desert and so did
his son Isaac, who had reopened the wells which were destroyed by the Philistines.
3 Mackay 1935,
40-2,
4 Bromehead 1942,
85.
183-96.
5 Mathioulakis 1966,
6 Castleden 1994,
55-8,
32.
23.
7 Report p repared for the open day of the Clwyd Powys Archaeological Trust,
in
conjunction with other interested groups held during the summer of 1995. In a personal
letter to me Dr. Caroline Earwood, the leader of the excavation team has indicated
likely that
clay-lined
to collect
rainfall
have
(1989,
but that
121).
Such
it was most
supply,
he
says,
would
been
seems
to
be
the
most
feasible
explanation
from
studies
of
origins.
Gloucestershire
the
Neolithic
to
are
at
Crickley
Hill
(Dixon
1994),
Montgomeryshire
Neolithic
Examples
Bronze
Age;
Breiddin
Hillfort
(Musson
1991),
Late
Bronze
Age,
and
Danebury
(Cunliffe
1995,
91)8,
(Musson
England.
There
is
growing
perception
that
far
from
welcomed
neighbours.
the
The Romans
invader
also
for
protection
against
aggressive
to the
roads,
bridges,
(especially
that the findings will not be completed for a few years as they are havi ng extensive
dendrochronology tests done on wood recovered from the site. Letter to A Burgers from
Dr Caroline Earwood, dated 5 February 1996. Another example is the Budsene spring on
the Danish island of Moen w hich was enclosed in a hol low alder wood (Brondsted 1958,
2.202). The figure shows the tree trunk which was inserted into the springhead, and
also the relics which were deposited by worshippers of the spring goddess.
8 Dating from pot tery indicators were from before the 4th century BC, but it is
thought that ha bitation might have been from the Neolithic period, though
the main
activity dated to the 4th c BC and later.
supposed
'benefits'
improvements
water
Britons,
the
Yet
provision
attention
it
that
at
the
Romans
brought
Romano-British
warrants.
Richmond
sites
(19 68,
to
the
has
2.87)
not
said
received
the
that
the
"in
because
from about
British
towns,
of
the
bath
was
probably
one
of
the
most
significant
The
non
Suddenly
water
culture
to
be
virtually
shock.
confronted
on-the-tap
with
must
this
have
that
foreign
been
concept
something
the provision
of
of
of
a
running
water for domestic use could have had on such prehistoric societies
has been observed time and time again during modern colonization.
Whether
or
supplies,
not
individual
Britons
water
supply was
no doubt
had
access
viewed
as
themseves
a symbol
to
of
such
Roman
power. Both colonial powers, Britain and France, having occupied many
African and Far Eastern countries for economic and political reasons
during
the
late
19th
and
early
20th
centuries,
made
attempts
to
Iron
Age
nucleated
water supplies.
communities
assimilated
new
cultural
along
built
main
or
roads
and
encouraged
military
the
communication
construction
of
routes.10 The
houses,
streets,
9 I was involved with a project through my firm working for a mu nicipality in the Cape
Province of South Africa,
drilling for water in a remote rural area in order to
prov ide drinking water to the African community. When the scheme was completed and
communal water poi nts laid on in the area, celebrations lasted for a week to thank the
Inkosis for this wonderful gift.
10 Frances Condron provides an extensive bibliography in her PhD thesis,
development of settlements and towns in Britain after the Roman invasion.
1996,
on
the
at
Lincoln
(Bidwell 1980,
(Wacher
1995,
138,
and
my
Fig.3.9),
and
at
in this
thesis,
related structures,
their
settlements
into
towns
and
villas
with
these
it
built
with
timber,
century
most
of
those
other
settlementsand
villas.
The
indigenous
British
It can
had
living
impressed
and
villas
communities. However,
water
shows
supplies
support
large
there are
still many sites which do not show any remains of these facilities.
Can one conjecture from the data available that all or most of the
forts, towns, settlements and villas actually had these water-related
structures, but that they simply have not been found,
or that they
have been destroyed during the passage of time? From the excavation
reports of many sites I gain the impression that many more excavated
sites than are indicated in my database,
related
features.
materials,
Comparing
smallfinds)
where
descriptions
no
in fact,
of
sites
water-related
features
have
been
explicitly recorded with sites where they have, it seems likely that
many of the former group could also have had aqueducts,
wells and
baths.
11 W acher 1995,
2nd e d . , (1974);
Brown
(ed), 1995;
Smith 1987.
Britain,
aqueducts
were
often
of
the
leat
type,
specially
Chesters,
aqueducts
of
Winchester
the
and
stone channel
Wroxeter.
type
served
(Dorset),
At
Lanchester
the
fort,
three
constructed
along contours of the land between the source and fort. Elsewhere
wooden pipes were used. At Chester (Hanson 1970, 185) and at Lincoln
(Lindum) (Thompson 1954)
the water supplies. At York the suggestion is that it may have been
supplied with a lead pipe encased in concrete
Both
wells
and
shallow
tanks
also
were
(Hanson 1970,
important
water
192).
supplies
during Romano-British times and they were probably the most important
forms of water supply for most communities.
I think that much of the evidence pertaining to the ancient water
supplies in Britain has become lost through a combination of lack of
interest, natural decay and the activities of industrial and property
development
during
and
after
the
industrial
revolution.
of
wells
and
aqueducts
were
much
more
common
towns,
than
present
settlements and
villas.
the
problem
of
who
paid
for
the
amenities
which
were
site
types,
their distributions
related
structures.
The tables
in
the
categories
and distribution
distribution
of
site
of water-
figures
show
that
types
across
the
excavation
reports,
are
often
susceptible
to
new
the
the
fortsuggestthat it must
have had a running water supply. Why is there not even the smallest
physical
evidence
Similarly,
the
to
indicate
aqueducts
at
that
both
there
Chester
was
such
and
Lincoln
interpretations
as
to the
an
aqueduct?
leave
many
functions
of certain
deep
Why
did
communities)
some
settlements
or
villas
(usually
10?
13 wells,
At the small
not
large
such as the
14
discussthese questions,
but
6.
SOCIAL CONTEXTS
so-called
complex.
romanization
of
conquered
peoples
by
the
is
Romans
successful
in
Italy
itself
or
even parts
It may have
of
Gaul,
but
but will
consider
in Chapter 7 Burnham's
10
comment
on
the
121).
Roman
rule
architecture,
basilicas,
new
the
ways
of
provision
construction,
of
major
changes
buildings
in
such
domestic
as
fora,
The perception
but did it
romanize them? Even before towns or villas were built many Britons
would have come into contact with the new way the Romans built their
forts and with the amenities they provided for themselves.
I shall
structures,
2nd
supply administration,
in Britain.12 It
century
and
the
3rd
and
other
amenities.
would
appear
century
that
people
during
of
the
Britain
Places
like
Dorchester,
Leicester,
Lincoln,
London, Silchester, Wroxeter, York, and many other British towns grew
in population,
these
centres
temples,
fora,
How successful
able
to
and with
embarked
on
that
extravagent
accompanied
building
sustain
development
the wealth
this
this
extravagance? Were
progress?
How
long
projects
tribal
these problems
(in chapter 7)
as
such
as
it
authorities
last?
growth,
the
did
this
factor?
they affected
Did
urban
extent was
I shall
look at
the water
related
set about
local tribes
(in Chapters
2 & 7).
the
This process
Caesar,
with
increasing
trade
between
Britain
and
of
Gaul
relates
to Rome:
Vitruvious
Water Distribution in Ancient
and
Frontinus.
Also,
Rome, The Evidence of
13
Julius Caesar invaded Britain during 55 and 54 BC in two separate expeditions. A
number of scholars have written on the subject, such as Bruun (1981) , Ward-Perkins
(1970, 1-19), Mac Mullen (1974), Duncan-Jones (1974 and 1990), and others.
11
in Roman Britain.
costs of structures
case studies,
order
of
cost
limitations
However
for
such
as we have
structures.
to make many
of
through two
Such
studies
assumptions
have
about
certain
conditions
or
the
army,
and
the
quoted
for services
or
estimated
such as
costs
of
is
maintenance.
generally
This
directly
applies
now as
proportional
it must
to
have done
their
regular
in antiquity.
construction,
additional
funding
or
after
would
be
period
required.14
of
negligence,
From
the
and
that
archaeological
in
drinking,
antiquity
had
far
wider
implications
than
its
use
for
Religious
especially
at
spring
sources.
In
Celtic
religion
the
(Jones and
14 Gaius Plinius Caecilius Secundus, Pliny the Younger, AD 727-113?: Pliny, Letters
and P a n e g y r i c u s , in two volumes, tr. by Betty Radice, Loeb Classical Library, 1972,
(1969), letters 37, 38, 39, 40,41, 42, 61,62, 90,91, 98 & 99. All these letters refer
to some technical prob l em related to water or construction and the finances for the
projects. It would appear in the case of the prov ince of Bithynia Trajan h ad given
Pliny the authority to allocate funds for spending on construction proje cts without
re ference to a higher authority.
12
Mattingly 1990,
of
but,
Often
true to
rivers
and
wells
were
associated
with
Celtic
religious
functions and became focal points for their cult practice and ritual
long before the coming of Rome. The pre-Roman Celtic Britons appear
to have
adopted
the
concept
of
the
sacred well
similar
to
their
the well
also would have served the purpose of providing the community with
water and some of the early wells were probably dug by these early
Britons.
Rivers
were
important
because
they
were
associated
in
Celtic
tradition with fertility and with deities such as the divine mothers
and sacred bulls.
functioned
in
(foretelling
1967,
the
the
20). There
role
of
future), have
was
war-goddesses
wide
and
association
connection with
prognostication
with
fertility
water
which
(Ross
could
be
rivers
so named,
its
1967,
nature
20). Ross
(Ross
of the
to the
Ross reports other cult object finds found in rivers and at springs.
Some
finds in
rivers
are
images
swords
indicates
and
that
Celtic
one
has
metal
to
be
of
deities
which
probably
were
Although
careful
the
in
Jane
Webster
interpretation
of
13
22-47,
the
the Celtic Clota name the "washer, the strongly flowing one"(45), the
Severn
Brent
from
the Braint
"Brigantia"
from Anglesey
river goddess
and
the
(278-9) . The
itself"
(Ross
1967,
22).
shrines associated
with aquatic cults in the region of the Severn estuary and adjacent
areas (p.22).
There is other direct evidence for the cult of wells, pools and lakes
in the British
isles.
for the
Strabo,
quoting Posidonus,
reported
"that
lake at
reports that Celtic metalwork was found in the small lake Llyn Cerrig
Bach,
in Anglesey,
and that
the manner
in which
it was
deposited
(1958) also
in rivers,
the most
likely reason for their presence here is...of the placing of precious
objects
in water
(Ross
1967,
24).
On
the
it
would
seem
that
but
new
of
the
original
interpretations is necessary.
It would seem that the Celtic religion was highly developed amongst
Britons
at
the
time
of
the
Roman
conquest.
The
Romans,
in
the
14
7. CONCLUSION
In the chapters that are to follow I shall look closer at the aspects
that
have
been
presented
above.
outline
the
methodology
of
and
questions
archaeological
data.
that
A
have
simplified
emerged
from
presentation
my
of
study
the
of
the
database
section reviewing
the analysis
that
15
has
been presented
CHAPTER 2.
METHODOLOGY, DATABASE, ROMAN HYDRAULIC ENGINEERING
IN ANCIENT TEXTS
1. GENERAL
After
trying
several
databases,
Microsoft
ACCESS
version
2.0
was
adopted for creating a database and the other type tables to record
the information on water-related structures from the archaeological
record of Roman Britain.
is
material
no
on
comprehensive
water-related
gazetteer
structures
summarizing
for
Roman
the
Britain
available
for
easy
remains
Roman water
might
have
type
been
structures,
overlooked.
but
inevitably some
primary motive
for
publications
the
research
presented here has been the need to bring together information spread
out in many publications. Also important in the study is the inter
relationship of the various features, and their significance in the
romanization process, and in their social and religious contexts. The
data can never be complete as new excavated sites are continually
reported
in
the
literature
and
vast number
of
sites
remains
unpublished.
3. DATA COLLECTION
The most
Roman
important
Britain
is
previous work
the
unpublished
on water
thesis2 of
Julie
Hanson,
in
which
types,
the types
of
details about the sites. Her primary purpose was to look at the known
1 At the present time there is no comprehensive summary for the wate r sup plies of the
empire or even for the various provinces, though one may note the w ork of Gsell (1902)
and Birebent
(1962), both for Algeria in North Africa.
Research is currently in
pr ogress to summarize information on water supplies for the M ed iterranean area. The
Germans have done some work in collation of water supplies through the FrontinusGesellschaft in three volumes Die Wasserversorging antiker Stadte edited by Garbrecht
et al, 1986, 1987, 1988. For this reason I have produced a gazetteer type da tabase of
in formation on water supply features at Romano-British sites.
2 Hanson 1970,
PhD Thesis,
16
Faculty of Arts.
investigations
requires updating.
aqueducts,
but
conjectured
that
in the
and inevitably
light
her discussion
of
recent
of many
sites
because
such
of
other
sites
may
circumstantial
have
had
an
evidence
aqueduct-type
of
she
water
supply.
Stephens discussed sites with civic and military aqueducts in his two
1985 papers,3 but he did not go into the sort of detail that Hanson
covered. Both authors in a few instances make comments on technical
detail that require some re-interpretation, which I shall attempt to
do in the sections of this thesis where it is appropriate.
The
features
discussed
by
Hanson
and
Stephens
are
summarized
in
76
-9
-2
65
** Corbridge i) Flavian,
ii)
Severan,
iii)
4th century
BA+
WP++
SP
TA
Military: (certain/conjectured)
29/13
13/-
22/2
6/1
8/-
12/6
7/-
13/-
6/2
1/-
Total:
41/19
20/-
35/2
12/3
9/-
the
pipes
shown
by
collar
to the forts.
remains
within
forts
were
conjectured
55
Civil:
31
Total:
86
3 Stephens 1985a,
4 Hanson 1970,
197-207;
1985b,
216-36.
1985b,
216-36.
358-74.
5 Stephens 1985a,
197-207;
17
to
indicate
For comparison,
(see Chapter 6, Table 6.1). Neither of these two authors dealt with
the wider range of archaeological sites with water-related features
that have been reported on during the past 200 years.
I have collated information on water-related structures for 7 types
of
sites.
Military
sites
are
divided
into
fortresses,
forts
and
supply,
either
of
wells
or
aqueducts
or
a water
source
baths,
such as found
could have been provided with water from wells, or some other source,
by
soldiers
or
slaves
carrying
water
to
the
baths.
Baths
are
typology of
and the
4. DATABASE RECORDING
The
procedure
used
for
compiling
the
database
was
to
record
information obtained from the libraries for each site on index cards
with
as
many
bibliographic
references
as
possible,
and
then
to
types
of
sites
are
given
in the
records
of
the
'site
type'
6 The database table was originally headed by fifteen fields of w hich the last eight
represents the actual typological features data. The 7th field was the 'Reference'
memo field, but the large size of each memo record made it impractical to handle
w ithin a database, and was transfered as a table to Appendix 2.
18
The initial two fields provide the site name and category of site
(for
simplicity
colonia,
defined
municipium,
civitas
nine
types:
capital,
fortress,
small
fort,
town,
fortlet,
settlement
and
villa). Field three gives a Roman name for the site name where it is
known. The fourth and fifth fields give the location of each site in
the
International
'northing'
Grid
format
in
coordinates based on
terms
100km
of
the
(100,000m)
'easting'
grids.
The
and
sixth
field provides the location for each recorded site using the Ordnance
Survey National Grid system for Britain, with the two-letter notation
for each 100 square kilometre grid area.
in
the
northings
International
Grid
format
based
on
eastings
and
The database
is presented
in this
(Appendix 3);
to
include
the
dates
of
the
various
structures
in
the
of the
the
number
of
fields
and
would
have
made
printing
of
the
references
of
Appendix
2.
The
typology
of
water-related
7 Areas such as Wentlooge Level (Gwent) show Roman expertise in this type of hydraulic
engineering. On Wentlooge Level, the Romans used many drains to lower the water table
in order to reclaim land which was affected not only by surface flooding, but also by
erosive attack from usually violent sea waves. At Rornney Great Wharf alone more than
19
features
for
information and to
clarify
aspects
always are given at the end of a reference record where they apply.
It is also to be noted that many of the references in the database
and the bibliography are dated in the last and previous centuries.
These
authors
give
descriptions
of
structures
which
they
have
actually seen or had first hand knowledge of and many of which may
have disappeared
references
since.
are the
only
ones
that
features
Icouldtrace.
the
older
However,
early
archaeological reports often did not give sufficient details for the
precise location of sites or for features such as wells or aqueducts
at
particular
sites.
Nevertheless
the
reports
can
provide
visual
for the
references
are included
endeavoured
to
make
Romano-British
in the
these
as
period
table
of
complete
for example,
(Appendix
Appendix
as
1) .
2, and
possible.
Many
I have
Where
the
they are indicated in the field as W1+. Where the specific number is
given it would be indicated.
The
inclusion of structures
form of water
is
supply by
implication. Large public baths would most likely have required some
form of constant running water supply such as an aqueduct. Baths in
some forts, villas or homes may have been small enough for them to
have been filled by water carriers. Many of the sites listed in the
database show only a bath or a drain or piping, without also showing
a water supply. It is to be inferred that such sites had some form of
a water supply source. A number of these may have had running water
supplies of which the structures are now no longer visible,
TIBRS:
TIRCGL:
(Oxford).
Britannia Septent ri o na li s.
C o nd at e- G levum-Londinium-Lutetia.
20
or they
land
may
have
been
filled
from water
in a
tank
or
cistern,
or
water
items such as
every
such
tile
would
indicate
the
presence
of
bath.
of a bath
have
not
investigations.
been
This
found
could
or
recorded
equally
apply
during
to
archaeological
other
water-related
30
unpublished
years
of
rescue
archaeology,
remain
and
this
(Zienkiewicz, et al,
period
sites
which
has
evidence
of
water-related
information about
other Romano-British
are many
more
which
have
shown
none.
This
seems
to
apply
excavations,
because
they
had
been
completely
destroyed
had
or of course,
water-related
structures.
is
led
to
genuinely not
the
tentative
be
remaining
reasonable
sites must
quantitatively
or
indicator
that
qualitatively
many,
if
to water
different
not
supplies
from
21
most,
the
of
that
the
were
traditional
What are the chances that the sites that had no record of having had
water structures, did in fact have them? I am conscious of my lack of
archaeological
excavation
experience
of
sites.
and
of
However
the
problems
find
involved
myself
asking
during
whether
One
constant
factor must
have been:
could
they
subsist at those sites, and for this the two prime requirements were
food and water.
Defence was
also
important,
livelihoods
and
commercial
but under
'new'
prosperity
Roman
rule
became
the
motivating
forces.
Since water was vital to existence it can be assumed that they gave
it some priority in their planning.
works
on Roman Britain
it is difficult
to
find
the word
seems
'water'
to indicate
or
community.
When
reports
do
mention
well,
for
have been in use. That the well in itself was of importance to that
site in its own right is usually ignored. This, it can be argued, may
account for the fact that many sites have been reported on and no
water-related structures recorded. It may also be because interest in
water or the technical aspects of what made a site viable as a place
to
live at was
outside
the ken or
interest
of
the
investigating
to
underestimate
abundance of water
the
significance
of
water
leading
supply.
An
site.
22
5.
records
of
eight
water-related
tanks,
features:
sites
aqueducts,
and
baths,
1.
An aqueduct
conveys water
from a
Baths,
floors
from underneath and walls built with flue pipes to heat the rooms,
and almost
a water-bearing
layer
(aquifer)
to convey rainwater,
and
waste
Tanks
were
receptacles
in
which
water
was
tanks,
but
these
stored,
usually
It is likely that
are
less
visible
archaeologically.
7. Springs were sources of water which issued from the ground fed by
an aquifer in the form of a perched water table, usually along the
slopes of hills or mountains.
8. Sewers were stone-built structures,
and also on
23
of
5.1 Aqueducts
These consisted of the following types:
la. Leats: open ditches cut into soil or rock.
lb. Stone channels: usually lined with clay or lime-based mortar to
render them impermeable to seepage; on more elaborate aqueducts the
channel
cleaning purposes,
popular
enough
for
sometimes
they are covered with ashlar slabs. Another issue concerns whether
a stone channel was open, as at Lanchester, or covered.
lc.
fit
into an enlarged
adjacent pipe,
opening
at
the
thick
iron collar.
end
of
These
an
iron
collars are often the only remains which indicates the line of such
wooden pipes. Roger Wilson discusses the wide use of wooden pipes
as aqueducts, and in particular refers to them also being used in
inverted siphons (1996, 22).
Id. Lead pipes: were either
(see Chapter
3,
Fig. 3.2);
system where
inverted
round,
triangular,ovalor
siphons
were
pear shaped
used in an aqueduct
necessary,
or
in
intra-site
distribution systems.
le. Ceramic pipes: were made from terracotta
(earthenware)
and were
either cased in concrete, or uncased. They were hollow, conicallyshaped, with a small diameter end which could fit into the large
end of the adjacent pipe. The sizes varied considerably, depending
on the expected amount of
water to be delivered.
At Lincoln the
Ceramic distribution
pipes found at Wroxeter were about 7 0mm at the large end and 3 540mm at the small end. Pipe lengths varied from about a half metre
(as at Chester) , to one metre
Stone pipes:
made
from bored-out
stone
were
used
in
several
places in the eastern part of the Empire, but are rarely attested
in Britain.
In the
database
I have
used
the
classification
aqueduct
and
the
not
always
possible.
Aqueducts
24
are
more
fully
discussed
in
heating,
which
provided
graded
temperatures
to
generally
two
heated rooms (lb above) and the larger baths also included a dry-heat
sweating room
the
first
two hot
rooms there were pools fed with hot water from tanks associated with
the heating system (Nielsen 1990, 14-8). There were also many other
rooms used for dressing, scraping, massaging and other activities in
the larger bath complexes. The public baths of towns usually had a
basilica attached to them, where exercise activities were available,
and where food and other commercial products could be obtained, which
at the same time provided a social community centre associated with
the baths. There were also private baths which were run as commercial
enterprises,
particularly
in places
like
Rome,
Pompeii
and
Ostia
private
baths
in
Britain.
Baths
varied
tremendously
in
size
across Britain and the large ones usually were architecturally very
elaborate structures. At fortresses, baths were usually large as they
would have been used by large numbers of soldiers (Zienkiewicz 1986).
A
number
of
towns
also
had
large
baths
such
as
Leicester
and
1991,
8-11).
In Chapter
6 I
Some other
BA)
in the database
discussed
their
their
sizes
table,
classification
but
in the
and where
Baths
text
(Chapter
appropriate,
have been
5)
have
referred
to
comprehensively
5.3 Wells
These were an important source of water for all site types. They were
dug into both soil and soft
fractured rock,
and,
where necessary,
1983,
25
is widely
spread
lm to 2.5m,
and
also were made in square format, particularly when they were steined
(lined)
with
wooden
planks.
Some
wells
were
steined
with
used
generally
feature
except
(30m) (App.2)
have
in
not
a
been
few
studied
cases,
as
such
a
as
special
the
construction
Wilsford
(40m and
shaft
51m) (App.2)
collapse,
as at London,
Lancaster and
in the database.
5.4 Drains
Drains varied tremendously typologically, having been constructed as
open soil type ditches, pipes of various kinds and channels in stone,
some open and others covered.10 Many excavation reports show drains
of several kinds for some town sites. At forts they generally seem to
be timber
or stone-lined.
allowed to deteriorate,
drains were
often
(App.2).
5.5 Pipes
These
were
used
in
aqueduct
systems,
water
distribution
systems
26
a ter m
in modern
were used more often than either lead or earthenware pipes, whether
as
aqueducts
(as
at
Cirencester,
Caistor-by-Norwich,
London,
Silchester,
Carpow
and
Verulamium
and
Wroxeter)
(Hanson
1970, 419). Wood was cheaper and more accessible than either lead or
materials
Where
as
(Wilson
Caerwent shows, they were prone to burst under pressure (Hanson 1970,
85) . There seem to have been no standard wooden pipe diameters or
lengths, these depending probably on the boring equipment and lengths
of trees available to a pipemaker. Iron collars used at the joints of
wooden pipes were found at many sites but surprisingly were not used
much at military sites as at Birrens, Brough-on-Humber, Fendoch,
Pen
Llystyn and South Shields, although they were used at the legionary
fortresses
of
Caerleon
and
Carpow
(Hanson
197 0,
421) . Pipes
are
cheaper
wooden
pipes
rather
than
lead
or
earthenware
aqueducts
in
Roman
pipes
The
use
of
earthenware
pipes
as
Britain
is
recorded only for Chester and Lincoln, whereas they were used widely
in distribution
systems
at many
sites.
When
they were
used
as
over
their
lengths
for different
sites
Their diameters
from about
30-50mm
27
pipes
to
Chester
fortress
and
probably
also
to
the
5. Tanks
Tanks were widely used in all categories of sites to store water.
Many
excavation
reports
have
recorded
tanks,
some
of
which
were
filled with rainwater run off from roofs and others were placed where
it was most convenient to fill them,
modest
size
(capacity about
1.5m3 to 3m3).
in buildings and of
In private homes
they
usually were below floor level in a convenient place where they can
receive water from rooftops.
even in the
5.7 Springs
These
were
the
preferred
source
of
water
for
aqueducts,
though
It must have
At Winchester
it is reported that
several
springs
springs
along
the route of
but that a
'culvert'
(Hanson 1970,
the aqueducts
179).
was used to
Although there
fort, the northern aqueduct was extended further west, where two dams
were built across a stream to provide the source of water.
Springs
5.8 Sewers
Sewers
are
some
of
the
best
preserved
structures
from
the
Roman
28
were
used
in
Roman
times
for
the
removal of
foul
and
1993,
high
46,
(Whitwell 1976,
slab-covered
138)
1-55),
and
structures
arched
They were
such as at
structures
such
waste,
Lincoln
as
at
York
distribution
of
incompleteness of
the
water-related
features, and
shows
the
one would expect to find at sites. For instance at a site that has a
bath one would expect to have some form of water supply. If the bath
is large the expectation would be that it was serviced by a running
water
supply,
such as
at
Caerleon
or Wroxeter.
However
it
often
happens that a town may have a public bath, but no water supply of
any kind has
as yet
been
found,
as
at
High Wycombe.
I shall
be
6. TYPOLOGY OF SITES
The
term
'site'
in modern
archaeology,
it
is
suggested,
must
be
"descriptive
features occurs"
region,
label
(ibid.
or the route
for
place
where...
artifacts
and/or
of an aqueduct
as a site,
but
usually
such
11 There is no doubt some logic behind the reasoning why the term 'site' as a notion
should not be used, but if qualified with a name that gives the location of a supposed
site it provides a practical means of referring to a locality. The w or d site comes
under the category of generic words like 'love', 'war' or 'object', each of which have
bee n
analyzed
etymologically
and
philosophically,
without
specific
accept able
definitions. Yet they are read, in spite of Duneell's statement that "site, as an
ar chaeological concept, has no role to play in the d i s c i p l i n e ... In spite of the
technical problems its abandonment will cause, the concept of archaeological sites
should be discarded" (1992, 36-7, quoted at Greene 1995, 53).
29
6.1. Forts.
In the database tables I have used three categories of forts namely
fortresses
(legionary),
forts
(auxiliary/cavalry),
and
fortlets.
All three
its active
existence.
I have
not
included military
camps as a category, because they did not have permanent status, and
did not generally have permanent water supply features,
water most
1.
Fortresses
were
of
two
kinds:
legionary
(c.
20-25ha)
and
of
legionary
and
auxiliary
troops.
Wilson
lists
ten
(1980, 92-3),
but the military disposition of fortresses and forts was very fluid
through the conquest period (c.AD 43-68), the Flavian (AD 70-96), the
Trajanic-Hadrianic
63).
Often
both
(AD 97-138)
fortresses
and
forts
were
deployed
(c.AD 142-
during
these
periods to serve particular military needs, so that they may not all
have been occupied at the same time. Jones and Mattingly
(1990, 88-
101)
4.32)
show
in a
series
of maps
(Map 4.23,
4.24,
4.31,
the
all the known military sites for the period from AD 43 to AD 168. In
the
database
have
listed
nine
legionary
and
four
vexillation
of deployment.
In Table
6.2 where
I have
given
the
summarized information for all site and feature types, I have classed
all military sites together as forts, for the purpose of analysis.
12
However,
some
anomalous,
semi-permanent
structures
were
erected
during
the
construction of major sites, as with the officer's compound at Inchtuthil. (Pitts and
St Joseph 1985, 215).
30
numbers at some forts were increased to about double the former unit
size (Breeze 1983, 15). Their duties were in general to keep lines of
communication open and preserve the peace
in conquered
territory.
During the 3rd century the north of Britain was relatively peaceful
(Welsby
1982,
8),
but
in
the
south-east
of
Britain
there
were
the
construction
of
the
so-called
forts
of
the
Saxon
Shore
(Johnson 1979). These forts differed in some respects from the type
of forts described above, but I have not listed them separately. They
have been listed as ordinary auxiliary/cavalry forts. Holt is listed
as a fort, but was actually a supply base to other forts of special
products such as pottery tiles and ceramic pipes.
3. Fortlets were usually small forts (generally less than 0.5ha) and
manned with a detachment of about a centuria. They were often used as
outposts
bridge,
or a road
installation
(Breeze 1983,
such as
43). Milecastles
for a period
and,
when military
and some
circumstances
large
urban
centres
or
major
towns,
each
serving
specific
functions within the province. The three initial coloniae were built
for the specific purpose of housing large populations of discharged
soldiers,
as
at Colchester,
Gloucester
and
Lincoln.
York,
on
the
other hand, was raised to that status in the early 3rd century during
Severus' reign (Salway 1993, 391-2). The water-related structures for
coloniae were
usually
inception
their
of
an
integral
development.
part
The
of
their
coloniae
planning
would
at
have
the
been
established with the approval of the Emperor and the colony would
have been governed by a council known as an ordo with a constitution
modelled on that of republican Rome. Officers would have been elected
by
the
council
as
executive
magistrates,
who
would
have
been
31
chartered
chartered towns
towns
(the
coloniae
Distinctions
and
are drawn
municipia)
and
non
the latter and 'small towns'. Verulamium is the only town in Roman
Britain for which there is evidence that it was granted a charter as
a municipium (Wacher 1993, 18). The status of Roman London seems to
be uncertain; however, I have designated it a chartered town
in that
and their
families.
of
Lincoln and
later York.
All
four
initially
started as
fortresses.
2. Municipia were
often pre-existing
towns
that
were
promoted
to
but
they
the
did
coloniae
not
have
(Frere
the
1967,
full
200).
administrative powers
Verulamium was
vested
a municipium
in
(probably
from the AD 50s) and London may have achieved that status soon after
the Boudician revolt in AD 60 (Frere 1967, 93-4). However, London is
a problem with
regards
to
its
status.
Morris
argues
that it
was
its probable
status
from its
more
than
that
mere
settlement
or
ordinary
town,
13 Morris 1982, suggests that Leicester was also a municipium, based on The diploma
of M. Ulpius Novantico (CIL x v i , 160), a Coritanian/Corieltauvian auxilliary soldier,
gives his origo as Ratis, not Coritanus, implying that Leicester had risen from the
status of civitas to that of municipium by AD 106. For the contrary v i e w see Frere
1978, p p . 2 3 5 - 6 (Morris 1982, 71, n.31 p . 354). Frere's v iew is that since Novantico
was already a Roman citizen as a result of a special grant in the field...", it has
no bearing on the status of Leicester. This is an historical opinion, but one cannot
but wonder why so few towns in Britain had municipium status.
32
of
Canterbury,
"their
other
evidence
of
intense
status,
particularly
romanization",
such
as
but there
self-governing
function
on
the
powers
model
of
of
the
Roman
chartered
towns
with
towns,
an
but
ordo
they
who
did
elected
17
tribal
areas.
These
tribal
areas
were
the
artificial
Millett lists 16 civitas capitals (1990, 106, Table 5.1; 154-6, Table
6.5), which is three less than the number of tribal areas which he
shows in Fig.16,
Degeangli
in Wales,
nor
Tribe
Atrebates
Trinovantes
Dobunni
Cantiaci
Dumnonii
Brigantes
Regini or Regni
Corieltauvi
Belgae
Iceni
Silures
Catuvellauni
Cornovii
Durotriges
Carvettii
Demetae
Parisi
33
Modern town
Silchester
Chelmsford (or
Colchester)
Cirencester
Canterbury
Exeter
Aldborough
Chichester
Leicester
Winchester
Caistor-by-Norwich
Caerwent
St. Albans
Wroxeter
Dorchester (and
Ilchester)
Carlisle
Carmarthen
Brough-on-Humber
Civitates
perigrinae
coloniae but
magistrates
without
serving
had
similar
quaestors
in
constitution
and
local
servi
government
could
charter.
positions
members
some
wealthy
that
of
Augustales. The
to
the
elected
acquire
Roman
the
community
to
and
wealth.
They would have been responsible for the collection of taxes for the
state, but also for the planning of
raising the funding for buildings.
become
generous
benefactors
to
and
city
development
projects.
By
the
(Wacher 1993,
(Salway
1993,
391).
11
civitas
capitals have been directly attested and 3 further sites are also
suggested
administered
The
suggestion
is
but
opinion
seems
to differ on this
(Wacher
1995,
207). The formation of the civitates is complex (Haselgrove 1984, 3143), but generally was based on prior Roman experience in Gaul.
In
'tribal'
boundaries
(Haselgrove
1989,
34;
Birley
1988
11,
24ff).
in my database
for the
6.3.1.
Small
towns
were
'settlements'
that
developed
the
34
is
contentious
(1995a,
7-17)
with
little
gives
agreement
detailed
amongst
analysis
of
scholars.
settlements
Burnham
and
what
list of 97 small
towns
(1986,
186-7,
small
towns
Burnham's
(1995a,
approach
7-17).
Frances
(Burnham 1995a,
7-17)
Condron
has
synthesized
to the classification
(Condron 1996,
of
57-8).
(1986) Burnham listed some 97 sites, but in later work he has cited
in
texts
variously
52
sites
(1987,
187),
and
60
and Millett
(1990,
154-6)
lists
117
(1988)
(1975,
'small
3)
and
54
list 78
towns'
but
16 civitates
town'.
unless
they
Other
nucleated
sites
I have
called
'settlements'
civitates or municipia.
6.3.2.
The term
settlement,
as
it will
be
used
in
this
thesis,
economic
interests
such as pottery
farmers,
and
'settlements' dating from the Iron Age developed during the RomanoBritish period into villas or towns,
Keynes.
Ultimately,
35
Even
the name 'town' which has become synonymous with settlements in Roman
Britain as 'fortified places with a civilian character'
(Wacher 1993,
(1986,
187).
Hingley
(1989,
2-3)
acknowledges
the
historical
Condron
(1995,
103)
seems
to
have
town'
here
archaeologists
description
to
of
is
the
the
taken
as
identified
settlements
provided
label
rather
themselves".
sensible
modern
sites,
than
This
applied
by
a meaningful
seems
to
be
lived.
It is a
for
water-related
features.
The
existence
of
such
is not paid to the parameters used to categorize a villa site. This may of course be
due to problems with the limited data of remains found on the ground, but it is not
always the
only problem. The problem of categorizing villas applies equally well to
other site
types. She quotes Millett's approach to the
term 'villa'
(1990,
91-2),
which I will in general accept.
15 Hingley 1989, 2-3, cites Burnham & Johnson 1979; M Jones and D Miles 1979; Reece
1982 and Cunliffe 1984 who is critical of the historical approach. Since 1989 a nu mber
of scholars have
reconsidered the basic approach and assumptions
to settlement
classification.
16 Burnham 1993, 99-110; Condron
authors of
the articles mention
definitions of small towns and
necessity discussed by them.
1995,
103-18; Rodwell
& Rowley (eds.),
1975. The
the disagreement over
details of some aspects of
by implication also on settlements, wh i ch are of
36
to be general agreement
there
on what
this
constituted a villa,
category,
which
I will
have
not
distinguished
between
different
types
of
To
quote
one
recorded as a villa,
suggest
that
the
example,
the
have
been
'a bigger
of
this
example
is
enterprise
than
Lydney
site might
although
thesis
to
investigate
but
whether
the
sites
could
be
differently classified.
Usually the excavation reports refer only to the site comprising the
structures
within
the
villa's
built-up
area.
Actually
the
villa
form the villa estate (Hingley 1989, 100-9, 121-3). When it comes to
the other occupied areas,
characteristics
features
stone.
associated
of Roman
with
Hingley refers
province
as
ranging
type buildings,
Roman
construction
that
or
is,
being
the
built
'extensive
with
over the
communities
to
village-type
linear
and Hingley
Non-villa
(1989,
settlements
100-9,
are often
121-3)
indicates
some relationship between them and the villas. On the other hand many
non-villa settlements were not associated with villas or other nearby
communities and Hingley
(1989,
100-9,
127-8)
could have been the homes of absent landlords. He also suggests that
there may have been families living in simple settlements but also
had some contacts with more sophisticated urban communities
1989,
24-25).
between
It
different
seems
kinds
to me
of
that
estates
37
the
in
exact
these
form
rural
of
(Hingley
dependency
areas
is
not
clearly established.
(1989,
133-44)
also
refers
to
more
isolated
non-villa
traditionally
inherited
from
generation
to
generation
from
communities,
it
seems
unnecessarily
in
these
sites
is
whether
they
possessed
water-related
features and how their presence affected the status of the so-called
settlement. It is therefore not possible to adopt Hingley1s criteria
for what constitutes a settlement for the purposes of this thesis. I
have simply grouped all the sites which are not forts, major towns,
small towns or villas as settlements in the records of the database.
The
dispersed
minor
dwellings
that
dominate
the
Romano-British
academic
priorities
more
focussed
on
upper
levels
of
the
settlement hierarchy.
that
the
settlements
like
that
at
Alice
Holt,
Hants.,
and
Cantley, West Yorkshire (see App.2), owe their existence to the very
extensive
potteries
that
developed
during
the
late
2nd
and
3rd
sites of Romandate,many
of which continued on
period.
Dolaucothi
(App.2)
and
dependent on aqueducts
(App.2),
were both
Roman iron
mining and iron working sites but only at Lydney (App.2) have I been
able to trace an aqueduct. Beauport Park
38
(App.2)
is referred to in
of
had
what
two
water
wells
and
source
was
it may
classed
have
as
had.
Walton-le-Dale
settlement,and
is
sites I formed
the opinion
reading
several
of
conclusion
These
that
sites
references
for
no water-related
are
referred
to
in
those
features
the
sites
were
database
cameto
actually
as
closer
the
reported.
'falsus'
sites,
because it is likely that other people may form the same impression
as I did that water-related features have been found. That is not to
say
that
they
did not
these
features,
but
merely to put on record the fact that they have not yet been found contrary to the impression one can form.
Many reports of sites incorrectly imply that they did have waterrelated structures, especially baths. For instance in describing the
features of a site with a bath, usually the bath had an hypocaust,
drains, special kinds of tiles and tibuli, mosaics and often painted
walls.
A number of sites are described in the literature referring to such
type of finds, and the expectation would be that the sites may have
had baths, but none have been found. Similarly from descriptions of
certain sites the expectation is that they may have had other waterrelated features, but again none have been found as yet.
17 Britannia 10,
1979,
139-56;
18 Britannia 19,
1988,
217-74.
19,
1988,
217-74.
39
Site type:
Ardoch
fort (large)
Limestone Corner
fort
Neath
fort
Whickham
fort
Coldharbour
settlement
Brantingham
villa
Frampton
villa
Kingscote
villa
fort
Bowness-on-Solway
fort
Lidgate
villa
found
yet.
The
following
three
sites
(not
included
in
the
database), would seem to have all the requirements to have had waterrelated
'large
structures.
civil
Brancaster
settlement'
water-related features.
in Norfolk,
(23ha),
(EAA
23,
Bowness-on-Solway
fort
1985)
(refs),
(3.1ha)
has
no
with
reported
the terminal
fort
on Hadrian's Wall, a largish fort (c. 2.77ha) and with a large civil
settlement, has no reported water-related structures. In the light of
evidence for elaborate baths and water supplies to some of the other
forts associated with Hadrian's Wall it is surprising that this site
has not yielded any evidence of their existence. Lidgate in Suffolk,
a large winged corridor type villa with 20 or more rooms,
has no
further
examples,
Kingscote(App.2)
in
Gloucestershire,
is
county where they were usually well provided with a bath and water
supply system.
Frampton
(App.2),
been
part
of
the
a villa in Dorset
is another site
complex,
but
none
have
been
features
found.
to
This
444ff;
EAA 23, 1985;
JRS 67, 1977, 157-8. Bowness-on1971, 224-5;
Britannia 3, 1972, 330-1; 5, 1974, 258.
40
more reported villas and many more settlements of various kinds. Some
may have been too poor to afford such luxuries, but there seems to be
too large a number without those features. They may of course not
have been found yet or have been irretrievably destroyed.
therefore,
In general
would be expected but none have been found, stresses the problem of
some
archaeological
data.
There
are
some
detailed
reports
on
aqueducts and wells, but very few in comparison with the number of
sites that have been excavated and which would obviously have had
some sort of a water supply. The reports of the Colchester
(App.2)
excavations pay particular attention to water supply and other waterrelated structures.20 For many
sites
it would appear
that
amateur
interested in aqueducts
or
wells.
Historically,
archaeological
past.
These
archaeological
investigations
(Aubrey, Camden,
antiquarians
excavations
and
had
no
indeed
started as a result
Leland,
Stukeley)
particular
their
had in
background
initial
of
interest
in
was
and
antiquarians
by
the
second
half
of
the
in Europe,
'scientific'
nineteenth
century
had developed
stratification..."
excavation of Corbridge
as
one of
(1995,
the
62).
He
sites where
of
quotes
the
"aims
the
the
and
(1995,
69-76).
Leonard Woolley
of
today
(Greene
1995, 70). The remains of the aqueduct at Corbridge are still visible
within the confines of the excavated site, but its course outside the
20 Crummy 1984,
Report N o . 3; 1992,
Report N o . 6.
41
main site is not well reported. From my study of the reports on the
remains of water-related features it seems that there was a limited
interest in these features during early archaeological work and it
was more by chance observation that features like leats were reported
when they were still visible during the 17th to early 20th centuries.
Most of the physical evidence of leat-type features seem to have been
irretrievably lost as a consequence of intensive ploughing and is not
easily identified even on aerial photographs. The exceptions are the
leats at Dorchester, Great Chesters, possibly Winchester and the leat
at Wroxeter
leats
are
only
along
certain
sections,
the
rest
being
supplies and
seems
to
be
an
anomaly
and
can
be
attributed
to
several
different factors.
1. wells have often not been found because of the limited area of
excavation at many of the sites;
2. wells may have been covered with so much overburden that ordinary
physical excavation techniques have not uncovered them;
3.
excavators
working
before
archaeology
became
strict
science
during this century have not been interested in wells unless small
finds were found in them;
4. wells have collapsed or been irretrievably destroyed;
5. where leats were the water supply system these have most
likely
where
water
wooden,
lead
supply
or
consisted
earthenware
of
pipes,
water-mains
they
have
in
the
been
forms
robbed
of
out,
romanized types
of
sites,
such as
42
farmsteads,
round houses,
need
to
be
published
and
this
also
limits
the
available
literary
evidence
is
available,
which
complements
the
not
features,
ancient
is
the
poor
evidence
from
only
for water-related
literary
sources.
Recently valuable evidence has become available from the many writing
tablets found at Vindolanda,
(Bowman 1994a,
likely to provide new evidence about what was built in Britain during
the Roman period. Four inscriptions
supply
have
limited
knowledge
of
flood
control
measures
in
antiquity.
flooding of
the
lower
city
(Livy 35.9.2;
food crisis
in Rome,
resulting
in Augustus
54.17).
appointing
Suetonius mentions
(Tac. Hist.
1.86;
43
flooding by the
considerable
damage
due
to
flooding of
the Anio
(Ep.
and Tiber
8.17).
rivers,
Further major
floods are mentioned during Marcus Aurelius' reign (AD 217) and again in
AD 374 (Sha. M. Ant. Aurelius 8; Dio 79.25; Ep. de Caes. 32.3; Ammianus
29.6.17-18). Olivia Robinson discusses the administration and officers
who were appointed by different emperors to deal with the flood problems
in Rome (1992, 3, 85-9).
In modern
times
agencies
concerned
with
water
hydrology
generally
in order
to
impound
nor the
been
intended
for
flood
control
structures,
and
others
as
44
both
for water
supply
and
flood
The
the Clyde,
the
along
the
Rhine and
Danube
must
have
been
banks
of
the
river Foss.
Richardson
(1959,
56),
who
excavated
the
the
Hungate area, shows the levels of silting of more than a metre (Fig.3)
that occurred during the Roman period into the Medieval period as a
result of
flooding
to have occurred
during the late 5th to early 6th centuries on the Hatfield Moors
and
also in the Humber area duringthe Dark Ages (Ramm 1971, 183). No doubt
similar floods took place elswhere in Britain during Roman times, which
probably resulted in erosion and instability of the embankments of leat
aqueducts, but we have no record of specific instances of such damage.
No study of flood damage to structures seems to have been carried out
for the Romano-British period. What, may be asked, was the impact on
Roman water supply systems and how did the communities and the Roman
authorities
deal
with
the
potentially
devastating
effects
of
such
repair the flood damage at the wharf in York and this could have been
the situation at many places in Britain. Usually flood damage is of such
proportions that for the period under discussion it would have been
easier
to
facilities
leats were
abandoned. It is likely
rather
severely
than
repair
damaged
they
the damage.
were most
When
likely
evidence
indicates
that
some
water
supply
systems
as
result
of
flooding,
which
could
have
triggered
Negligence in
45
of
indicating
regular
practice. For a city like Rome with specific departments responsible for
such work it would have been a normal practice, but in Roman Britain the
local town administrations may not have had specific maintenance units
to do regular maintenance. However, we do not know what the maintenance
practice was in Roman Britain for any of the urban centres.
reports,
'Gentleman's
particular
information
or
available
only
libraries,
or
generally
magazines
Magazine'),
counties
is
and
(such
books,
some
in
of
the
the county
as
which
or even parishes.
in
recorded
the
are
Much
and parish
now
often
of
country's
in
this
discontinued
specific
to
literature
is
specialized
records and
journals,
copyright
the
National
recent
years
aqueducts
some
specialized
in particular
books
have become
on
water
available,
supplies
of which
(1992)
and
the most
for Roman
aqueducts and Dora Crouch (1993) for Greek aqueducts. Neither pretend
to be exhaustive,
work for some time. Other books on water supply have been written by
technical people,
(1971,
1976),
such as Robins
usually
supplies
from
(1991),
Birebent
(1946), Bromehead
engineers who
antiquity
(1962),
to
the
Gsell
covered
present.
(1902)
Scholars
and Landels
such
as
(1978),
Bruun
describe
Deman
has
(1934) .
Much
specialist
literature
recently
become
and
lead pipes,
but
only
in passing
from
provide
the
fact
that
drinking water,
the
nor
Roman
aqueducts
to promote
hygiene.
on the actual
not
Nearly
built
all
to
Roman
46
individual houses, and some cities (such as London) got through their
entire history without having had an aqueduct at all. In most, when
the aqueduct
arrived,
it came belatedly
and
only
as
result
of
the
existed without
one
for
(p.5). I
truth in it and
3. His treatmentof
the
wide
detail
subject
is
ranging,
providing
much
on
of
the
the
Empire.
He
has
also
provided
some
information
on
the
recent
information on aqueducts
and water
supply
in the
English language
overview of the
existing
literature he
stresses many
with
the
famous
areas
In
his
where
there
an
review of the
are major
du Gard bridge
and
the very
long
is
particularly
significant,
because
believe
that
also
it
seems
to me
that
information
has
become
(1993)
in
It was
in these type of
formations
that people
looked
for
of water
supplies was
47
planned
and
constructed
by
the
Greeks, and how they were managed. However she is not too clear on
some technical detail.
Long before
engineers
seriously
of
early water
Vitruvius
amplifications by Faventinus
and Frontinus
(AD 90s),
and Palladius
(c.
with
5th c.),
(Plommer 1973, 1-2). One of the problems with the studies by scholars
from the different
disciplines
is
that
classicists,
historians
and
surveying,
architects.
The
modern
engineering,
study
of
has
ancient
come
from
technology
and
classicists22,
specifically
ancient
Roman
historians,
knowledge we know nothing, except for the remains of how they applied
it. Aqueducts are one such manifestation,
which includes
the simple
1988;
Isaac
1958;
Smith
1971,
1976,
48
1990,
Blake
1991;
1959;
Landels
Sprague
1978;
1978;
and
White
many
surveying manuals,
the Corpus
certain
land
survey projects,
but
they
did
not
describe
how
century.
planning
healthier
of
abundant
developed
out
supply
which
water
of
the
became
existing
supplies
(Smith,
1976b,
93).
The
example
of
the
extensive aqueduct system of Rome became a model for the supply of clean
water for modern aqueducts over long distances from the cities. Thus
interest in the technology of ancient water supplies was brought into
focus, particularly Roman hydraulic technology and the applications of
the technologies that preceded it.
We do not
Vitruvius,
engineering practices.
it
clear
that
the
architect/engineer
produced
plan
even
24 Vitruvius,
25 Vitruvius
comments on what was expected from architects and
engineers while
practising their professions. Most technical training was obtained while men served in
the army, he
himself being army trained. He gave his
views on what
the educational
background
of
architects
and engineers
should
be.
He gives
criteria
for
an
architect/engineer's
liberal
educational
background
in
order
to
understand
the
technical, social and spiritual implications of his profession (1.1.3) and the need to
appreciate the social issues in practicing his profession and his responsibility as an
advisor and designer to his client (6, Pref.6). He emphasizes the need to rely on the
experience of others (7, Pref.6) and gives his reasons for producing a manual on
architecture and engineering techniques (7, Pref.18).
49
expertise successfully,
in
spite
of
the
flow rate,
size of conduit
and friction
for
the
operation and
by
which
is
the
unnecessary
force
for
of
the
the
air
system
may
be
relaxed"
to work.
(8.6.5-6),
These misconceptions
which
by
those
authors have been misinterpreted from as far back as the 16th century.
Modern engineering
scholars
have
tried to make
some
sense
of what
Vitruvius was trying to say in terms of the physics involved, both with
regards
to
the
standard
siphon
and
the
incorrectly
named
liquid
flowing
an
inverted
siphon.
The
physics
and
mathematics
of
but
in
U-tube
pressure system in the aqueduct has not yet been proved, although it has
been suggested.
There is circumstantial
to the
50
in Rome
showed
lack
on the construction of
of understanding
of the
part
to
results",
of calcined
"kind
found
in
of
the
lime used as a
powder which,
by nature,
neighbourhood
of
Baiae
produces wonderful
and
Mount
Vesuvius
made with
lime admixed with pozzolana was stronger than concrete that had only
clean
quality
of the
Pantheon (Smith 1976a, 49), the dome for the Bath of Mercury in Baiae
(McKay 1978, 48-9) , and vaulted roofs for many structures, and the cores
of masonry dams.
Other aspects of ancient and Roman construction are similarly poorly
understood.
NAF
article on the
Smith,
a civil
"Problems
engineer,
as
it relates
to
He suggests
that,
since
the
parameters would have been important to consider. How these were arrived
at would be most interesting to know.
There
ammount
following
three
chapters,
followed
distribution in Britain.
51
in
Chapter
by
their
CHAPTER 3.
WATER SUPPLIES:
for the
aqueducts they built all over the Empire. The sources of water supply
used by them consisted of:
1. water from springs, streams and dams; transfered in aqueducts in
the form of various kinds of stone channels, leats, wooden and
earthenware pipes;
2. rainwater,
British period that carried the channel of an aqueduct over deep valleys
as are found in so many other parts of the Empire.
Many
settlements
and
towns
in the
Empire
had
only wells
as water
large number of
they were
not
able
to yield
sufficient
quantities
of water,
source of water for aqueducts and dams were often built in the vicinity
52
(see
Section 3).
2. SPRINGS
Vitruvius,
source if springs were not obviously visible. The engineer had to lie
flat on the ground and looking at the horizon would see in the distance,
if the circumstances were right, where vapours rose from the ground,
indicating a source of underground water. If he then dug there he would
encounter a water source.1 This is a misinterpretation of a natural
phenomenon, because the apparent vapour so observed is merely the mirage
of heat radiation creating turbulant air flow upwards
ground. He does qualify this procedure with advice on the type of trees
and plant growth which could also indicate that there were underground
streams in the vicinity. Vitruvius made comments on springs and the
quality of water derived from them, which were a mixture of practical
advice and often
truth.
His
views
many
technological
aspects,
no physical
often
unsound,
of
water
From great
supplies
for
antiquity
humans
which
such springs
and
animals.
flows by
is relatively
have
In
been
time,
the
cults
5-11); the
villa of Chedworth had a spring water source and so did the palatial
53
3. DAMS
Often dams were built below springs,
or across
streams
to ensure a
dams
as
their
supply
source,
notably
the
dams
at
Alcantrilla
Cornalvo (Merida), an earth dam. These dams, built across streams, had
fairly large capacities: the Proserpina with a capacity of 10 million m3
of water. This is a large dam for its low height of 12m. Hodge lists 13
Roman dams built
during
the
2nd and
3rd
centuries
AD
(1992,
82) .
Schnitter (1987) records many dams from antiquity which served as water
supplies to cities, of which a number were built by the Romans.
Roman dam building probably
represents
the acme
of
their
empirical
but
with a down stream slope of 1:5, is reported as still being in use today
(Hodge 1992, 89). The dams at the source of the Lanchester aqueduct are
said to have been almost intact during the early 19th century
(Steer
over a small stream that had a spring as its source. About 0.25km higher
up
the
slope
there
was
another
spring
54
at
which
dam
(B)
was
constructed, which discharged into the lower dam. It is not known how
the system worked between the two dams, but it is possible that the
supply from the lower spring into its dam did not supply sufficient
water,
upper dam.
It has been suggested that the Saughy Rigg Washpool was the source for
the Great Chesters aqueduct (App.2), but it is not certain (Mackay 1990,
285)2. I saw no evidence of a likely place at this position where the
aqueduct could have obtained water from the river or evidence of dam
remains when I visited the site, nor could I see any indication of the
source on the relevant aerial photographs held at RCHM(E) archives at
Swindon. Putnam has suggested that the Dorchester aqueduct may have had
as its source of supply a 4m high dam, which was built across the
Steppes Bottom stream near Littlewood,
interim report). There may have been other Roman dams in Britain which
have now completely disappeared.
4. ROMAN SURVEYING
In my judgement, one of the most significant feats of Roman surveying
and construction is that of the aqueduct from Ucetia (Uzes), the source,
to Nimes and that beautiful aqueduct bridge structure, the Pont du Gard.
From the surveying point of view, a standard of levelling was achieved
that would do a modern surveyor proud, providing a route through rough
and mountainous country with a fall of 17 metres in 51km, that is, a
fall of one third of a metre (1 foot) in 1 km (Hauck 1988, 78-84). This
quality and skill of surveying, done with the crudest of instruments,
can only be appreciated if one has tried to do surveying over similar
distances.
of Gaul
(Nimes,
51km;
precision
surveying
had
special
of
significance
for
the Dorchester,
This
Roman
Great
skill
in
Britain
as
Chesters
and
surveying
skills
were
important
aspects
of
Roman
technology
55
controlled by a
mensor (Dilke 1971, 51) . Most of the ancient writings on surveying were
based on accounts by Vitruvius, Hero and the Corpus Agrimensorum, which
described surveying for agricultural and land purposes,
engineering, for which little research has been done
61) . The Roman surveyor
of
the
1st
century
the
dioptra
had
probably
improved
(Smith
1990, 61).
Knots f o r ^
Establishing
Grade
Notched Stick
adiusted at Right
Angie to Plumb Line
Level Line /
_ol S ight
\/j
- Cord
ENLARGED
D E T A IL .1
W e ig h t--!
Fig. 3.1: Levelling with an A-frame level, ( Butler 1933, 73, Fig. 5).
However, Roman surveyors were not infallible. A prime example of 'How
Not to Build an Aqueduct' discussed by Nicholas Horsfall, is the Saldae
aqueduct in Algeria which was intended to bring water from the spring of
El Anseur,
21km away.
A Roman engineer,
original plan, and carried out the initial survey. But it took a long
time, something like 15 years, from the start of the project (c AD 137)
until water eventually flowed. When construction had progressed for some
short distance, the contractors realised they had a problem constructing
the aqueduct over the valley of El Hanai'at. Datus' services were again
called for and the aqueduct was placed on arches along the valley. But
progress was slow and he was required elsewhere. He was recalled again
56
four years later, because it turned out that the two units who were
constructing the aqueduct worked
from
"It was apparent that the digging had strayed from the
line, so much so that the upper tunnel (source end) turned right, to the
South, and likewise the lower tunnel turned North,
two ends
were out
of
to
Datus
had
which was
of
it was
originally
many
pipes
and
channels
which
operated
in similar
limestone
environments, which could imply that water did not in fact flow in the
pipe for any length of time. There are other uncertainties about this
aqueduct, in particular its source of water and the delivery of water
from the source to the town (see section 9.3, p.87).
Roman
approach
to
water
supply.
Aqueducts
that
were
used
Their
None
of these
wealth
for
the
communities where they were situated. However all three were indicators
of wealth
of the
community or of
do
not
appear
to
be
aqueducts
constructed
specifically
for
57
water
in Britain
Wentlooge in Wales,
and at
logged areas. Some aqueducts were constructed for industrial use such as
those at Dolaucothi where three leat aqueducts were constructed for the
purpose of hydraulic mining
scale,
at
(Jones,
of Cant ley
(Doncaster)
two very
short
also have
served
some
economic
bridge structures,
that
with
some
effort,
other
sources.
Although
there
were
facility for comfort and pleasure, and these, in part at least, depended
on the construction of aqueducts to make running water available.
In Britain there are none of the magnificent arched bridge remains
carrying the aqueduct conduits to urban centres. The Romans utilized
various forms of aqueduct conduit that are also found in other parts of
the Empire. Where there were no problems with valley crossings which
could not be effectively circumnavigated by
following
contours,
the
preferred method was to use simple leat channels, ditches dug into soil
and rock forming the water conduit.
Other conduit
types are
listed
below.
It would seem
the general
58
leat at Dorchester
it traversed rock,
as at
Dorchester. The second kind are channels made from dressed stone cut for
the purpose of forming channels with two vertical sides and a floor.
These channels could either be sealed with a lime-based mortar to render
them impermeable to leakage, or be the carrier of some form of pipe.
Remains
types
are
found at
several
sites
in
coal mining
1938).
(Steer
Vitruvius advises
that
such a
channelled
were
below ground
level,
firstly,
for
protection
against
pollution, and secondly, to ensure that the supply could not easily be
cut off by an enemy.
a) wooden,
b)
ceramic (terracotta),
c)
In Britain,
carried in a channel of stone, but not invariably so. The iron collars
associated with wooden pipes vary in size from different sites. More
than 20 collars
were found
at
Wroxeter
pipe lengths of
the order
of
between
) , whereas at Silchester
found at
1896(2), 422-4).
some
forts
Although
such as
at
Birrens, Fendoch, Pen Llysten and South Shields, no iron collars have
been traced (Hanson 1970, 421). Instead two pipes were fitted together
and the join then "solidly packed with clay" to prevent leakage (Arch.
J. 125, 1968, 125) . Collars were found at the legionary fortress sites
of Caerleon and Carpow.
59
b) cross-section of
Roman lead pipes and
soldered seam
\ h
*-v_
(a)
d) methods of jointing
pipes to each other
(b)
(c)
Fig. 3.2: Construction of lead pipes and making of joints (Hodge 1992,
309 Fig. 215, 312 Figs. 216 & 217, 314 Fig. 219).
60
b)
is that of the
An earthenware pipe
is also reported
for the
diameter
(external) , 18m long lead pipe was found, "which seemed to have carried
water across the bridge and down the centre of the widened road" (Wacher
1995, 175; Britannia 21, 1990, 325) . This is the largest known Roman
lead pipe found in Britain. Lead piping was extensively used elsewhere
in inverted siphons, and for this use the thickness of the pipe has been
reported as being from 19 to 25mm. Several techniques of forming lead
pipes were developed and an elaborate technique of soldering the joint
was one method of sealing the pipe, this being the main procedure used
for inverted siphon pipes. Pipes were also formed on a circular mandrel,
then bending the two edges of lengths of pipe over on themselves and
forcing the edges tightly against each other making a lap joint. This
type of joint would not have been able to take the internal pressure of
inverted siphons. Other joints were also made for specific purposes.
Sketches of lead pipe making and forming of joints are shown above (Fig.
3.2, p.60).
d) Stone piping were used mainly in the eastern Empire such as at Patara
and Aspendos
been extremely costly to produce in terms of both time and money. It was
not used in Britain as far as I can ascertain.
5.2. Tunnels.
Tunneling was not used in Britain as it was in the rest of the Empire.
Tunnels were constructed through both soils and rock, of which the K61n
aqueduct is an example. There is the famous tunnel of Eupalinos on Samos
island (Rihll & Tucker 1995, 403-31), and the tunnel section described
by
engineer who
left
an
inscription
describing
61
the
(Horsfall
1987) . The aqueduct from Vers to Cahors in France had a short length
constructed as an arched channel in a soil tunnel (de Garros, 1989) .
remains
of aqueducts
structures and arcades, and they are some of the most magnificent stone
constructions produced by the Romans. Long arcades with elaborate arches
over land and as bridges over rivers and roads are found all over the
Empire, some of the most conspicuous being those at Aspendos, Carthage,
Kbln,
Pont
du Gard,
Segovia,
and
Rome's
own
contribution
of
many
covered,
or ashlar
slabs.
Fig.3)
for
substructure
piers
which
most
probably
would
At
have
carried
Beckfoot,
Joseph
Fig.l) . From the 1:25,000 OS map of the area it would appear that the
aqueduct would have had to be carried on a raised structure to enter the
fort as Collingwood suggests, but the evidence is inconclusive. In a new
survey of the Great Chesters aqueduct by the RCHM(E) , it is suggested
that at the crossing of the valley at Benks Bridge, which is about 6m
below the course of the aqueduct, a bridge would have been necessary to
carry the aqueduct across the valley and the river. It is likely to have
been of wood construction, but its length and actual height is not known
and no surface remains survive (Britannia 21, 1990, 288) . The aqueduct
originally
serving
the
fortress
and
later
the
town at
Exeter,
was
322).
suggested which would have required bridge structures, but this would
need special investigation to confirm it.
62
have
the
knowledge
to
calculate
exactly
the
flow
in pipes
or
They
are
likely
to
have
relied
on
their
experience
and
been
encased
in concrete.
The
Lincoln
aqueduct
must
have
because the
upward
end
slope
of
evidence
for the
source
of the
aqueduct has not been found to confirm this. Suggestions have been made
by Thompson, Wacher and Lewis for possible sources, but so far no source
has been identified. Since the earthenware pipe encased in concrete has
been traced near the Roaring Meg, the postulated piers at the Roaring
Meg could have supported a low bridge structure as proposed by Wacher
(see Fig. 3.7b) carrying the pipes southwards towards the town. However,
such an interpretation is dependent on identifying
source.
explained.
Figure
intake
siphon the
3 Hodge
mentions
aqueduct
depth of
have been
viscosity,
and others.
sides
63
had
meet the bottom section of the conduit will bear the greatest thrust
and is usually anchored.
source
Ah
castellum
inverted siphon
anchor block
low bridge
structure near
The
source,
wherever that was, would have had to be higher by a distance Ah, for
water to flow in the pipe. Neither of these two heights are known.
6. TECHNICAL ASPECTS OF AQUEDUCT CONSTRUCTION
Adam
(1974)
discusses
in some detail
so,
for example,
the well-known
instability,
or cracking,
is
with consequent
64
The non-
and
this
is
likely
to have
applied
to
all
the
other
suggesting
of
instability.
Modern
soil
mechanics
have
shown
how
the
Roman
structures
must
have
failed
because
of
poor
foundation
theory
of
stability
analysis
that
can
be
used
to
assess
the
are
measured
in
situ
or
in
laboratory.
Avery
quotes
inoperative
during
their
lifetimes
due
to
some
construction,
structural
or
subsequent
embankments,
or
due
to
erosion
during
flooding.
The
regular
of
sophisticated repair-conscious
long
linear
features
were
age.
possibly
The
total
maintenance
prohibitive.
This
is
probably one of the reasons why a number of the aqueducts of Rome and
elsewhere fell into disuse.
65
where the
been
blockages
responsible
and
other
for
them
breaches
becoming
that
might
inoperative,
have
impeded
resulting
their
in
proper
aqueducts
used
on many
sites
(as
at
Cirencester,
Caistor-by-
which could crack if allowed to dry out and burst if under high internal
pressure.
The
seals
at
joints
between
individual
pipes
could
have
deteriorated over a period of use and would have needed repairs. The
joins of stone channels would have been potential weak points subject to
leaks, because the sealing mortar would have been very sensitive to even
slight lateral and vertical movements in the individual channel blocks.
It is reported that several sites had stone channels that carried wooden
pipes, which may be the result of experience with leaking joints of the
original supply channels. Hence the need to support wooden pipelines for
their joints not to be disturbed. I would suggest that on certain sites
the stone channels were the original water conduits,
but because of
problems with the integrity of the joints of the channel blocks to hold
water, it was belatedly decided to use wooden pipes and support them on
the channels. It could not have been cost effective from the beginning
to build a channel and use a wooden pipe to carry the water.
7. SITE TYPES
Owens observes that "An adequate supply of water was one of the major
factors in deciding the location of a new city, and many cities were so
situated as to take advantage of naturally occurring supplies of water
from springs, wells and even rivers"
(Owens 1992,
158).
It is likely
that many urban centers did not start with an aqueduct as the initial
source of water supply.
66
Aqueducts have been used to deliver water to all the listed site types
in Roman Britain as shown in Table 3.1.
Table 3.1: Aqueducts at site types.
Site type
Aqueducts
Percent
forts/fortresses
66
48
coloniae
municipia
<1
civitas capitals
11
small towns
settlements
20
15
villas
24
18
Total sites
135
6.1
(forts),
6.2
(civitas capitals),
6.3
(small towns),
6.4
the
number of known Roman settlements increases over the time span. However,
the number of sites where the water supply features have been reported
comprises a small percentage.
The rescue context of many excavations is a further factor affecting the
details of the archaeological
features
reuse.
Clearly,
the
information
obtained by me
for
any
specific site may not be complete, so that the presence of one type of
structure and the absence of another may not necessarily reflect the
true archaeological potential of the site. Hence, during the analysis of
site data these shortcomings in the data set and therefore the biases
they introduce must be borne in mind.
In Table 3.1
it is shown that
the number of
aqueducts listed in the database, showing that the army had a preference
for aqueduct water supplies compared to any other supply. The reason for
this may have been that there were usually a large number of troops
67
would have been used daily, and it would have been preferable to have a
running water supply than having to have it carried to the site or drawn
from wells. It is not clear why many of the other forts also did not
have aqueducts,
excluding
if for no other
8. AQUEDUCTS
8.1 General.
There must be a number of factors which relegated aqueduct construction
in Roman Britain to the expedient based mainly on leats, simple channels
at ground level, or pipe conduits. Comparing the lengths of aqueducts
(many over 30km
was easier to
Topography
but cost
significant
may
in
have been
the
decision
factor,
to build
the
may
cheaper
find in Britain.
have
type
been
more
structures.
structures
and
long
arcades
did
not
arise.
Leats
could
be
63
that is, channels cut into the soil or into rock, were major
undertakings. Even their relatively lower costs would have been con
sidered carefully, because towns had to generate their own revenue for
public facilities, whereas the province financed its own expenditure on
capital public works.
8.2. Leat aqueducts.
Examples
of
leats
are Bowes,
Dolaucothi,
Dorchester, Great Chesters, Haconby, Haltonchester, Hardknott, Tomen-YMur, Winchester and Wroxeter. Dolaucothi is the only known gold mine in
Britain,
Romans,
worked
that
was
served
by
three
and
further
aqueducts
of
developed by
leat
the
construction5.
the ground.
Where
they were
cut
into
soil,
leats
have
been
(Fig.
7.3), and portions of the Dolaucothi aqueducts, that is, the Cothi (A),
Annell
(B) , Gwenlais
(D)
(Fig. 3.4)
(Jones and
5 Lewis
Burnham
1962,
41-7.
71-84,
pls.I-V;
6 This may not be entirely correct, as apparently even if a site has been extensively
ploughed, if the light and the time of the year when the growth of the crops is just
right, a cropmark may show up and reveal a structure that is invisible at the surface.
69
kilometres
Fig. 3.4: Routes of the Dolaucothi aqueducts (after Jones and Mattingly
1990, 182, Map 6.3).
Only minor excavations have been carried out on the Great Chesters
aqueduct (Fig. 3.5), but it has been surveyed twice, the first time in
1850, and again in 1988 by the RCHM(E)
sections have completely disappeared,
lengths of the
aqueduct have survived to show its general route. It has been suggested
that the source of the aqueduct may have been a dam at Saughy Rigg
washpool (A), of which there is now no evidence. At Benks Bridge (B) it
has been suggested there was a bridge structure carrying a pipe conduit
across
the
stream,
but
no
remains
have
been
found.
70
The
aqueduct
Fig. 3.5: Route of Great Chesters aqueduct (after Mackay 1990, Fig. 6).
Present surveys of these aqueducts usually consist only of a line on a
poorly contoured map, making it difficult to visualize their spatial
impact on the landscape.
For instance,
it so that
can be
illustrated.
Although some work has been carried out on the British leat aqueducts,
there
is
no
published
complete routes.
information
showing
cross-sections
of
their
leats can be plotted in a format which will show them in the context of
the local topography. These two aqueducts
such as at Birdowswald,
Birrens,
Catterick,
Glenlochar,
High
71
"when the
channels were apparently as 'visible as the day they were made'" (Reed &
Austin 1976, 214) .7 By 1938 it was difficult for KA Steer (1938, 210-34)
to locate the route during a geometric survey of the channels
(Fig.
3.4). Reed and Austin state that destruction was due to "Three different
forms of land-use - agriculture, new housing and opencast coal-mining (and these)
can be
archaeological
sites".
Natural
erosion
by
the
elements
threaten
has
also
contributed to this process, but not to the extent and at the rate at
which human
agencies
destruction
was
have
opencast
caused damage.
coal-mining,
prime
since
cause
"as
of
wanton
result
of
deforestation during the last war and the succeeding two decades, some
previously
protected
archaeological
remains,
such
as
the
dams
and
aqueducts of the Roman fort, were destroyed" (Reed & Austin 1976, 213)8.
DETAILED SURVEY:
lanchester area
ROMAN AOufOUCI SYSTEM
Aqueduct
Fig. 3.6: Lanchester fort aqueducts (Reed & Austin 1976, 215, Fig. 37).
9. INSCRIPTIONS
There are no ancient literary reference or epigraphic information for
any of the known leats in Britain,
found, one about the repairs to an aqueduct, and three refers to the
provision of new aqueducts, all four referring to forts.
7
The article discusses the early discovery of
its subsequent deliberate destruction.
the
three
aqueducts
to the
fort and
8 A recent case of continued destruction of archaeological sites has been the three
medieval bridges at Hemmington on the Trent, (Current Archaeology 12(8), 1994, 316-
21) .
72
..| ...NL...
of
the
inscription
based
on
morphological
grounds,
mentioned by Stephens (1985b, 228-3 0), suggests that there may have been
more than one aqueduct channel; one supplied the fort and the other the
extramural baths.
that
'... of the troopers of the Cavalry Regiment ... Antoniniana ... domainland ... brought in a water supply, and erected a bath-building from
groundlevel under the charge of ... emperor's propraetorian legate, in
the consulship of Sabinus for the second time and of Anullinus.'
9 Collingwood & Wright 1965, (RIB) Inscriptions for aqueducts have been identified as
follows: Caernarfon - RIB 430; Chester-le-street - RIB 1049; Chesters - RIB 1463 ;
South Shields - RIB 1060.
73
It would seem that the work was done during the reign of the Emperor
Caracalla,
for
which
comparable
inscription
was
found
at
High
Rochester (RIB 1279). This aqueduct is known only from the inscription,
as there is no archaeological evidence for it. The fort has a sewer and
the baths mentioned in the inscription also suggest that there was a
running water supply. Limited excavation or loss of the remains of the
aqueduct account for it not having been found.
coh(ortis)
Gallo(rum)
inlduxit curante Mario Valeriano I leg(ato) eius pr(o) pr(aetore)
'The
Emperor
Caesar
Marcus
Aurelius
Severus
Alexander
Pius
Felix
of
Gauls,
under
the
charge
of
Marius
Valerianus,
his
propraetorian legate.'
This is one of the most complete inscriptions in Britain relating to the
provision of a water supply amenity. It indicates that there may have
been two aqueducts at South Shields, the original aqueduct that went out
of commission and the new one provided by the emperor Alexander Severus,
some time during AD 222-35.
/ alae II Astur(um)
| sub Vlp(io)
Marcello
/ leg(ato)
favoured
date,
the
earlier
whereas
174). Haverfield
Birley
(1988,
174)
(1897,
179)
assigns
the
it is
pointed out that the ala II Asturum was a 3rd century garrison of
74
Chesters, and since the posting of garrisons to Hadrian's Wall was rare
at the end of the 2nd century,
1, 1965, 472). However Birley also points out that a certain Q. Baienus
Blassianus, a prefect, "had commanded an auxiliary unit in Britain, in
his case the cohors II Asturum, which was his first commission." He is
reported in a papyrus to have been in Egypt in AD 168, and was active
from c.AD 140-168 (1988, 51), so he is unlikely to have been with the
cohors II Asturum in Chesters in AD 217. To reconcile these two persons'
dates it would seem that the inscription refers to the earlier governor.
This inscription is also complete, brief and concise. Birley has pointed
out that during the 2nd and 3rd decades of the 3rd century there were
many examples of the provision of this kind of amenity
(Collingwood &
and 1912
(Collingwood &
Hodge
(1990,
346-8)
gives
some
of
these
discharge
statistics for a number of aqueducts over the Empire, though none for
British aqueducts. These statistics should be treated simply as orders
of
magnitude
calculating
because
flow
in
of
the
channels
many
or
variables
pipes.
In
that
are
Roman
involved
Britain,
in
some
calculations have been done for the Dorchester aqueduct and those at
Dolaucothi.
However
there
is
very
little
hydrographic
data
of
in their
descriptions.
Jones et al.,
per day. The Cothi leat discharged into a tank with a capacity of about
14,300m3 (14.3 million litres). Water was used from this and other tanks
on the site for hushing of the ore to be processed for its gold. In the
RCHM(E)
10 The formulae on which they based their calculations are simplified versions of
hydrological calculations, but provides adequate orders of flow quantities. Hodge also
gives these formulae in his appendix (349-55), explaining in some detail for the n o n
technical person, how they are to be used.
75
(I have not
from Roman
Britain.) The quantities quoted are enormous. I suggest that these flow
rates are based on the maximum height
of
flow anticipated
for the
aqueducts, which would probably have been the flow rate during flood
periods.
For
the
Dorchester
aqueduct,
calculated,
based
on
the
dimensions provided in the RCHM(E) article (p.587), that for one quarter
of the depth of flow given, the flow yield would have been one fifth of
that at the suggested depth of 61cm (2ft) . Even so the yield would have
been nearly 12 million litres/day.
Provision of overflow
common
in
Roman
facilities
water
for discharge
engineering.
Such
of
facilities
were
generally
homes,
11.1.
and
then
discuss
possible
background
to
its
original
information
has
been
added
since
Thompson
reported
the
giving
the
relative
elevations
76
of
suggested
piers
above
the
LINCOLN
ROMAN
D E - PIPELIN E
OP
AQUEDUCT
OVEBHEAO - L I N E
s u p p o r t in g
s to ne
p ie a s
C IT Y
B C - P lP fL IN E
'GHOST*
BOUNDARY
UN06RCROUN0-
TR1HCH
ONLY
STONf
PACXII
SECTION
ALONG
LINE
A8CDEF
r j M
Fig. 3.7a: Route of Lincoln aqueduct (after Thompson 1954, Fig. 3).
W a te r p ip e in
c o n c re te jacket
Em bankm ent
'R o a rin g
JUeg
S o lid fill
M o d e rn g ro u n d level
R o m a n g ro u n d level
P o s t-R o m a n floo d plain,
n o w silt-filled
I Surviving m ason ry
q
S c a le of m etre s
30
77
Wacher
(1995,
141,
Fig.
60)
"a possible
(OD c.
40m)
is unproven,
actual siphon portion. From the contour map (Fig. 3.8) it would appear
that there were several places where the valley elevations were lower
than the delivery point, which would require that a very long section of
the aqueduct would have been an inverted siphon raising the water to the
final required level. If evidence should turn up to confirm this route,
it would be the longest inverted siphon built by the Romans and would be
unique in ancient hydraulic engineering. Wacher also shows a proposed
restoration
drawing
of
the
aqueduct
bridge
across
the
Roaring
Meg
(Wacher 1981, 300, Fig. 18), but it is not clear whether this formed
part of his extended aqueduct.12 By implication the bridge would have
extended further north than the Roaring Meg for some distance,
about
this
remains
an unproven
theory,
until
new
evidence
can
of
an
extended
aqueduct
is
nevertheless
an
attractive
11 Wacher 1981,
and discussion,
297-304.
12 It is noted that he has not included this reference in the second edition
his book though he included the discussion on the extended aqueduct.
78
(1995)
of
237-
ANCHOLME
CATCHMENT
226M a rk e t
Rasen +
lo w e s t p oint
219W'
W ITH A M
CATCHMENT
228-
216 C,
LINCOLN
ioolOO-
6 MILES
- - 2 5 3 W addington
261
ROMAN ROADS
C o le b y -2711
Fig. 10 The Jurassic Ridge and the Wolds.
(A)
(1981),
first referred to
in 1700 by a Yorkshire
commenting that "a small
(Thompson,
1954,
79
refers
to
short
section
of
earthenware
piping
encased
in
concrete which was found in 1857 towards the east of the lower town
above the spring line, beside Greestone Stairs
11) . Since
it was
also encased,
he
suggests
(Lewis 1984,
it could have
71, Fig.
been a
pressure main from the south, either from beyond the Witham river, or
from somewhere along the spring line at about the 50m contour. It would
seem there is not sufficient evidence to be more positive about what the
purpose of this short section of pipe-line was, and where it
entered the
Roman town. If it was a second aqueduct, it could only have served the
lower part of the colonia below the 50km contour line, since most of the
upper colonia lay above the 55m contour. He points out thatother pipes
found in Lincoln indicate that some running water supply was available.
However, there seems to be no direct evidence of how the aqueducts
(if
there were two) did provide water to the upper and lower parts of the
town.
Thompson (1954, 106-28), suggested that the Roaring Meg was the water
source for the aqueduct from the north (Fig 3.7a, position F) . It was
about 21.2m (70ft) below the highest point of the rising main aqueduct
along Nettleham Road, where it is assumed to have reached the Roman town
at the water tank just north of the East Gate. For water to flow upwards
along the pipe, Thomas
suggested that a
Meg. Its level would have had to be greater than 21.2m above the stream
in order to function. From this tank, water could be discharged into the
aqueduct through a down-pipe,
thus
forming a closed
inverted siphon
1806),
ii,
366, and
Pl.X.
Presumably the
map
14 There seems to be some vague evidence of piping which brought piped water to the
lower colonia site, with a suggestion that its source may have been along the Witham
river valley (Lewis 1984, 71).
15 Sympson T, 18th c., Itinerary Curiosuw, i, 88 and his Adversaria,
and Queries, ix, 65-90, mentioned by Thompson 1954, 108.
80
see Lincs.
Notes
(1954,
121-2), but
the quality of force pumps at that period is not considered to have been
of a standard that could have raised sufficient water for the needs of
the Colonia (Smith 1991, 125-6).
FT Baker, who was involved in the original excavations, suggested to me
in 1995, that he is of the opinion that the aqueduct crossed the Roaring
Meg and followed a line towards the left of the main road to the north
(B) , along the Lincoln Cliff, the Jurassic ridge above the 61m (200ft)
contour, where apparently there are springs. However he says, this has
not been looked into. Lewis refers to this ridge in his discussion of
Wacher's
of
(Lewis
the aqueduct
1984,
68),
having
but
its
he does
source
not
in this
(1981,
107-10)
1:10,000
OS
map
contours
indicate
that
Thompson's
line
would
(65.5m) as suggested
It thus would
seem reasonable to conjecture that the Roman engineers were aiming for a
different entry point to the colonia than the one suggested by Thompson.
The 1:10,000 Ordnance Survey map shows there is a valley (now filled in
and built over, carrying a road) between point X and the bath of the
colonia, which would imply the possible need of another siphon,
not
mentioned
the
by
Wood.
At
the
present
time
this
last
portion
of
81
'The
Roman
At Lincoln,
Roaring Meg source, OS surveying shows "that the pipe may have lain 1.3m
(4ft) below the Roman period surface at this point. It then follows that
there is a rise between source and supply-point of some 21.2m (70ft),
attained over a distance of approximately 2km (1H miles) . It is thus
clear that, between point C and the upper Roman colonia the aqueduct
constituted what water-engineers term a
construction,
waterproof
in
lengths
concrete,
of
'rising main';
earthenware
indicates
pipe
the measures
the method of
heavily
taken
sheathed
to withstand
in
the
(Thompson 1954,
I could not
source by mechanical
means
using
an
endless
chain
and
bucket
shows some possible suggestions of how the tank arrangement would have
provided water to the aqueduct.
In order to calculate the amount of water that could be made available
using Thompson's suggestions I have produced a schematic diagram (Fig.
3.9)
of
the
tank
solution
based
on
Thompson's
descriptions
and
schematic
representations,
merely
to
illustrate
the
principles
involved. They are not meant to indicate what the Romans constructed,
if, in fact, they did use this as a solution. Gough
(1806) apparently
reported the tank as being circular. The size of the tank is not known.
82
'inverted siphon'.
(Thomson's
foundation No.VIII, p.120, PI.XI). The size of the foundation was about
4.88m long by about 2.9m wide, giving an area of about 14m2. It would
have been adequate to carry a tank of similar area. The tower, on top of
Pier VIII, may have been of wood construction as suggested by Thompson
because of many nails found around the foundation bases, but it is not
certain.
Lewis,
supporting
the
in
his
tank
drawing,
and
the
implies
down
pipe
masonry
towards
type
the
structure
intake
of
the
aqueduct.
If the outlet of the tank at its base was at least 21.2m above the water
supply level
aqueduct near the colonia, the choice of height of the tank would have
been
important
depending
on
how
quickly
it
could
be
filled.
For
instance, if the tank height was 2m, then water would have had to be
raised from ground level to a height of 23.2m, or 24.2m if the tank was
3m high. Table 3.3 gives some statistics related to the tank supply
configuration and the aqueduct as given by Thompson.
It is necessary to know the heights of such a tank to provide sufficient
quantities of water. The Table 3.2 below gives height and capacity in
litres:
Table 3.2: Capacity of aqueduct supply tank.
Height (m)
Volume (m3)
Capacity (litres)
Mass (kg)18
14
14,000
14,000
28
28,000
28,000
42
42,000
42,000
56
56,000
56,000
The minimum head of water between the tank and where the suggested downpipe enters
the aqueduct
is
16.3m
(OD 61.2m-44.9m)
giving
a water
18 The tank with a height of 4m would have exerted a pressure on the foundation of
between 400 to 500kPa, which would have been well within the bearing capacity of the
foundation.
19 A head of water of
1 63,000kgf/m2 = 163kPa.
16.3m
is
equivalent
83
to
16.3m
1000 k g / m 3
1 6 ,300kg/m2
Table 3.3: Statistics for tank and aqueduct (Thompson 1954, 112-22).
Heights above sea level are given as Ordnance Datum - OD.
Letters A-F and X are given on Thompson's Fig.2
OD of foundation I at 42. 4m
C:
OD at 43.6m.
B:
OD at 44. 0m.
A:
OD at 44.5m.
X:
OD at 45.4m.
bath(?).
Top of 3m high tank at OD 64.2m.
The relative levels of the bottom of Piers I-VIII and 1, 2 and 3 south
of pier I,
resting on the
excavations, are shown on his Fig. 2 and I have reproduced them on Fig.
3.9.
Thompson (1954, 121) suggested that the tank could have been filled with
an endless chain carrying bronze buckets
14 litres per
person for a
population of
5000 inhabitants.
10 litres the
system could have delivered about 140,000 litres per day. Also, if there
were two sets of chain and bucket systems the delivered water could have
been doubled, and it also would have been insurance against breakdown of
one of the systems. The assumptions suggested are speculative and is
intended merely to provide an order of delivery of water at the colonia.
The suggestion
1981, 107-10),
about half a million litres of water, part of which could have been
distributed to the rest of the town.
20 The height of
bucket every 2m.
the
installation would
have
84
been
of
the order
of
25m,
suggesting
H*
OJ
iQ
OH 64.2m
N
Tank
O D 61.2m _ \t/
05
iQ
a
(D
w
rr
(D
a
rr
6
03
U1
Concrete casii
05
rr
0)
OD 44.9m
6
rr
OD 42.9m
O
ri
O D 39.9m
IQ
S
n>
IQ
D
Pier Vin
strengths
of
about
(in tension)
1200-1800kPa.
concrete with
Lamprecht
gives
some
(1995,
(630kPa) at which the pipe jacket cracked, indicating that the concrete
had
adequate
strength
(assuming
that
the
concrete
sample
was
I doubt whether a
Roman pump of the 1st century could have provided a pressure that would
have been necessary to force water up the aqueduct against that back
pressure in the pipe. With the technology available to the Romans it is
inconceivable that they could have produced pumps with packing between
21 If the rate of flow in pipe is assumed to have been lm/sec the pipe would have
delivered about one million litres of water per day, indicating that with the proposed
tank scheme it would have been delivering water well below its capacity.
86
the cylinders and pistons that could hold a pressure of that magnitude.
Today special O-rings are used in pressure pumps with close tolerances
between cylinders and pistons in order to transmit high pressures and
the casings would be made
in the
earlier
wooden
cylinders. I
they
used
estimate
from
alloys.The
whereas
replicas
of
reconstructed Roman force-pumps that they could lift water at most about
4.5m,
provided
cylinder
there
casings
are
reasonably close
the
tolerances
surfaces
are
between
the
smooth and
the
pistons have good quality packing like strong leather to provide a seal
between piston and cylinder. Lewis (1984, 64-6) also comments on the use
of a pressure pump
by
the
Romans
at
Lincoln,
indicating
that
the
(37
suggests
delivery
quantities
by
pump
of
the
order
of
does
not
Based
on
seem
to
these
have
been
arguments
typical
the
of
Roman
proposed pump
engineering
system
seems
provided
more information can become available about the elements of the system
as proposed and evidence of some remains of the equipment used. However,
the two proposed solutions would have been very
22 Lewis 1984,
inefficient ways of
57-73.
23 Thompson mentions allowance for leakage of 50%, which would require no stoppage
the pumping cycle to prevent filling the pipe again, a vicious circle.
87
in
providing
running
water
supply
to
deliberately
planned
large
colonia.
If the earthenware pipe sections that have been recovered are examined
they appear clean with no apparent deposits of calcium carbonate. Pipes
carrying water from geological limestone environments similar to that of
the Lincoln area usually are encrusted with lime deposits. An obvious
investigation that should
pipesin
flowedto
from which it
could be
in
inferred
that
lime-rich water
did
not
flow
the pipe.
1810
'History of
of
drawn
(Nichols 1810,
11(2),
505-6).
up
storm the
A slightly
remains
of
the
oppidum
of
'King
Leir'".
The
first
reliable
(Fig.
(1948,
3.10)
40-1)
consulting
engineer,
who
was
on
the
excavation committee for the Leicester Roman Forum during the early
1930s.
Archaeological Section in 1933, where he gave his opinion that the Raw
Dykes was
likely
aqueduct
supplying
the
town of
Roman
Leicester
(Ratae) with running water. He commented that no wells have been located
for the Roman period in Ratae (several have been found since) and with a
population of the order of 5,000, an external water supply was essen
tial. The SMR records have some details of his work on the subject of
the Raw Dyke, including a brief report on his opinion and a plan showing
details of the so-called aqueduct site, with survey details.24
Fig. 3.10: The Raw Dyke, Leicester, (after Wacher 1995, Fig.58) .
The conjectured construction of the Raw Dyke consisted of a leat type
excavation from the Knighton Brook (A) along the 61m contour for about
2.4km to the Roman town, entering the Ratae town walls through the south
gate
(B)(Fig.3.10).
A portion
of
this
ditch
(leat aqueduct)
still
survives in the town near the junction of Ay lest one Road and Saffron
Lane (C) . At Saffron Lane the ditch was cut along this contour and the
material
side was
dumped along
the west
forming an
embankment. At the top the ditch was about 15m wide and 4m at the
bottom, and about 2.4m deep (Wacher 1995, 350). William Keay prepared a
plan showing the route of the aqueduct, suggesting that it was fed from
the Knighton Brook and that there may even have been a dam constructed
across the stream west of the intake of the aqueduct.
Kathleen Kenyon said there were several problems with relative levels
relating to the channel and the baths, as it skirts past the baths into
the town centre.
"not entirely
satisfactory, but the best that can be put forward on present evidence"
(1948, 41) . At a door-sill of the Jewry Wall, she points out the level
is at 62.3m and the channel bottom is at 57.1m, about 5m too low to
provide water directly to the baths
experience showed that the level was too low, an effort was made to
24 A copy of this report and the plans, together with a copy of a statement he wrote
for the Leicester Mercury of 27 May 1938, 'Leicester's Roman Aqueduct', is held in the
SMR records at the Jewry Wall museum.
89
correct it by raising the bottom 6ft 6in" (2m) (1948, 41) , but she does
not show how this was achieved.
At the Knighton Brook end Keay also shows a length of about 320m (A-C)
of the Dyke as being conjectural. The 1888 OS map showed a length of
740m
(C-D) of the Dyke south of the short length of the Dyke still
visible near Aylestone road. In 1804 a portion of about 3 00m (D-E) was
reported to be still visible. Keay's firm did a tachy survey of the
remains of this surviving length of the Dyke and from city plans of the
area produced the line for the rest of the Raw Dyke.25 Keay mentions on
his plan that Dr Stukeley reported the length of the dyke as being 1,904
feet (580m) , which is considerably shorter than his plan shows, that is
2.4km to the south gate.
The
interpretation
of
the
Raw
Dyke
as
an
aqueduct
raises
several
questions.
Firstly, was it an aqueduct, and was it constructed by the Romans, or
was it constructed at a later date? There appears to be no literary or
archaeological information to confirm any of these questions, except for
Nichol's unsatisfactory reference given above. Secondly, assuming it was
of Roman construction, then they must have been aware from the start
that it would not be able to serve as a supply for the baths (F), which
were at an elevation about 5m higher. That would not have deterred the
Roman engineers. I think, if the structure was an aqueduct, the supply
of water to the baths became a secondary issue, the main purpose being
to bring water to at least the lower parts of the town.
reasonable
to assume
that
It would be
constant flow of water, the reason for it being built. If so, it must
have been purposefully built to supply a large number of inhabitants,
other
public
buildings
and probably
fountains.
Some
of
the
public
buildings were, like the baths, at a higher elevation level than the 61m
contour,
would also have been a problem. This seems to raise the issue of where
the majority of people lived in Ratae who could have benefitted from the
aqueduct. Thirdly, assuming that the dyke was the Roman aqueduct and
there was this problem about elevation, would it not have been sounder
engineering practice to have built a central tower with a raised tank
near the south gate and distributed water from there to the various
buildings including the baths and inhabitants rather than have the tower
90
to provide
his
Figure
158
of
the
line
of
within the precinct of the bath? Kenyon suggested that a tower was built
in the courtyard of the baths and water raised from the river below
(Fig. 5.6). The River Soar was another 8m below the level of the Raw
Dyke at the east gate, requiring water to be lifted to the raised tank
on a tower to a height of about 15-18m. It was not the policy of the
Romans to draw water from main rivers for towns. It would seem to me
that if the Raw Dykes was indeed the water supply of Roman Leicester,
then a raised tower in a more central place would have been the Roman
solution, for which there are many precedents, such as at Pompeii. Such
a tank would have had similar problems to the one assumed for Lincoln,
except that an inverted siphon would not have been part of the supply
system. So far there has been no evidence for such a structure.
Kenyon
suggested
that
the
Raw Dyke
as
an
aqueduct
was
failure,
implying that the town did not have a running water supply until the 4th
century when a tank supply was provided in the south-west corner of the
baths
(Wacher 1995,
349).
It would
seem that
there
is no physical
evidence for such a tank; only Kenyon's attempt to justify the anomalous
situation of a large bath needing plenty of water and an aqueduct of
doubtful use or even existence. Wacher comments on the large drains
virtually round three sides of the bath-building, which seem to indicate
that plenty
on the
east side of Southgate street northwards away from the Raw Dykes (352).
Belairs (1899, 40-4) described a sewer which started from the vicinity
of the baths, then discharging into the river. This would imply that
from somewhere excess running water was being drained, which must have
somehow reached the bath. The meagre evidence certainly supports the
existence of an illusive aqueduct and distribution system.
However,
recent archaeological
R J Buckley of ULAS
1997),
91
junction of the Roman road coming from the east and passing south of the
forum on the south-north road between the forum and the baths. Based on
this evidence it would seem the solution to the water supply of Roman
Leicester is far from resolved.
Neither of the two aqueducts discussed above has provided satisfactory
answers to the question whether they actually functioned as intended
during Roman times. If they were intended to do so but did not actually
deliver water as planned, it would indicate poor planning, ingenuity and
surveying
on
the
part
of
the
Roman
engineers.
However,
lack
of
(A) (SU 5060 3286, Fig.3.II)26 along the route of the motorway. The Trust
for Wessex Archaeology, through the M3 Archaeological Rescue Committee,
then did a geophysical survey of a portion of the aqueduct near Grace's
Farm, and an airphoto study of the whole route. The importance of this
recently discovered aqueduct is that it is the longest aqueduct so far
known in Roman Britain. Its length is reported to be 23.75km (p.8)(Fig.
3.11), following a winding route along and across contours to maintain
an acceptable flow grade. The source is at several springs above the
Manor Farm near Itchen Stoke (B) (Fig.3.11) some 6 kilometres in a direct
line from the delivery point at the north-west corner of the Winchester
defences (C)(Fasham & Whinney 1991, 9, Fig.7).
The report on the structure so far can only be considered as an interim
one. There would have to be a more detailed examination of the aqueduct
over its full length to establish its exact route and the profile of the
structure along its length. The geophysical work was done on sections
along its assumed route based on detailed 1:10,000 OS maps and from this
information a complete route has been provided.
several problems with the location of the route, but is not clear on the
details. Further investigation of the structure will be necessary to
5-11.
92
Abbotstone
King's
Worthy
W oodham 's Farm
G r a c e 's Farm
.
Itchen Stoke
springs
itchen
Abbas
Head bourne
Worthy ,
Ovington
Tichbome
Area o ft
Fig. 9
A 31
WINCHESTER
Chilcomb
Twyford Down
C o n to u rs in m etres 0 . 0 .
'
Possible route of aqueduct
Edge of Rom an W inchester
o
t
km
Fig. 3.11: Route of Winchester aqueduct (after Fasham & Whinney 1991, 9,
Fig. 7).
11.4. London 'a water supply.
London
obtained most
of
its
water
supply
from
springs
and
wells.
However, Wacher suggested on the evidence of wooden pipes found near the
Temple of Mithras and on the forum site that it implies "some form of
distributive system for running water" (1995, 101, 110). Wilmott rejects
Wacher's suggestion that an aqueduct must have supplied the baths (1974,
48-51, and reiterated in his 1995 edition, 90, 95). The Billingsgate and
Huggin Hill baths
(Marsden,
1980,
Thames river front, and the Cheapside bath was in a high water table
area, from which Wilmott concludes there was therefore no apparent need
for an aqueduct type supply and none has been found (1982, 16) . It is
probably correct that an aqueduct did not supply water from an external
source. This cannot rule out the possibility that water may have been
tapped from the abundant springs reported in the literature (including
93
Wilmott's
reports),
from within
the walls
of
Roman
Londinium,
and
Wacher
makes
the
interesting
observation
that
"as
seems
(1995,
110);
however,
no
system supplying
but
the
evidence
seems
to
indicate
that
because
of
leakage
problems, wooden pipes may often have replaced them, the channels then
being
used
supplies,
as
supports
based
on
for
the
pipes.
the archaeological
Sites
with
evidence,
represented
that
the
number
should
be
aqueduct
considerably
supplies
higher.
water
only
and I
In
the
issues
their
external
water
supplies
requires
urgent
further
research.
94
unknown
for
Roman
Britain.
For
instance
how
is
often were
sources
were
generally
from
springs,
or
the
head
waters
of
streams. Dams were often built at stream heads as sources for water, but
evidence for them in Britain is scarce. Many sites relied on wells for
their
water
supply,
and
other
less
secure
systems
such
as
tanks
95
CHAPTER 4.
WATER SUPPLIES: 2) RAINWATER CATCHMENT AND WELLS
1. GENERAL
Rainwater catchment will be discussed first, because the main part of
this chapter will be about wells.
2. RAINWATER CATCHMENT
Storing of rainwater, such as at hillforts, was probably the earliest
form of man-made water supply. Although water could be carried in skin
bags and pottery vessels,
so
Wooden tanks
truncated oval shape because it was found that this added to their
strength and improved the sealing between staves,
but
possible with larger vessels used for water storage. The one redeeming
factor about wood is that when it is kept wet
swells. Lead
it self-seals as
it
generally seem to
including some used for Christian religious purposes. Stone tanks were
the most popular water storing facility in forts, public buildings and
private homes. They generally were
wood-lined,
Variousmeans
lead-lined,
have
required
constant
maintenance,
especially
the
clay-
and
mortar-lined ones. The clay must have been a specially heavy clay with a
high plasticity in order to stick to the walls. With the elapse of time
the clay would have become completely saturated and would have slaked
from the wall, thus requiring resealing. Mortar-lined stone walls would
96
have been very sensitive to minor movements which could have been caused
by constant buffeting by buckets and being bumped by humans or animals,
or settlement of the foundation slabs. The mortar across a joint between
stones has no tension strength, so that the slightest movement caused
either
by
an
external
agency
or
temperature
variation,
could have
(metal-1ined?), Halstock
(stone),
North
Leigh,
North Fleet
(clay-lined) and West Wickham (wood) . At the small town of Ashton a lead
tank is reported. Tanks at forts were generally of stone construction,
even at
the Lunt
fort,
where most
of
the
buildings
were
of wood
construction.
I have recorded the remains of 102 tanks at all site types, but there
must be many more not recorded, or they have not been found. Table 4.1
gives the distribution of sites with tanks (and lists the total number
of sites included in the database below).
Colaniae
Civitas
capitals
6
Shall
towns
8
Settle
ments
16
Villas
32
Municlpia
-
24%
50%
43%
12%
8%
10%
135
14
67
212
372
38
Total
sites
102
a necessity.
The
remarks relating to wells and the number of fort sites recorded in the
database compared to the total number of forts known, applies to tanks
as well. Even the small fortlets would have required more water storage
than could be stored in amphorae alone.
The cavalry fort of Lunt at Baginton had 15 tanks, some of them used for
drinking water for horses, although they possibly could have been taken
to the River Stowe to drink. It had 6 wells which would have been the
water supply for the tanks.1 Housesteads had 5 tanks made of stone slabs
and lined with mortar, including the still well-preserved remains of the
1 There are three excavation reports
the first two.
97
fort,
to
one north-east of the latrine near the South Gate. How these tanks were
filled is not clear, although a few of the possibilities of how water
was supplied to the fort have been mentioned in Chapter 3.
Some tanks were filled from the roofs of buildings that had gutters with
down-pipes, but how general this would have been in forts is not known.
Storing
water
obtained
from
rain
water
could
have
been
only
It is most
probable that tanks would have been used at these sites but have not
been discovered yet,
In Frere's
(1972, Vol.l;
1983, Vol.2).
In London,
the
remains of three tanks have been found. A tank is associated with the
Billinsgate baths south of the frigidarium. A large timber tank preceded
the 3rd century bath-building. At the Cheapside baths the remains of a
timber-lined 28m3 tank (Fig.4.1 below) was found, situated to the north
of the baths. It was not revealed how water was transferred from the
tank to the baths or how the tank was filled. The tank was partly sunk
into the gravel layer and it had no bottom, so it may have filled by
seepage from the bottom, since the water table was high and ground water
was abundant in the area (Wilmott 1982b, 239) . At Silchester only one
tank is recorded,
considering there
were 76 wells. The villas at Keston and Whitebeech had 6 and 5 tanks
respectively,
98
B la c k e a r t h
G ra v e l &
s a n d y s ilt
N a tu r a l
if;?
Fig. 4.1:
N a tu r a l
Rom an
B ric k e a rth
s tra ta
Where there are records of sites with tanks and aqueducts and other
water-related structures, the possibility may also be that the tank (or
at least one if there are more) is associated with the aqueduct, as for
example
at
Aldborough,
Darenth,
Halstock,
Lincoln,
Prestatyn
and
Stanwick.
Obtaining water from the roofs of buildings other than at forts was
probably also a common practice, but the evidence would be difficult to
substantiate,
unless
guttering
can
confirm
it
and
has
been
found
99
3. WELLS
3.1. General.
Wells were a common form of water supply for both public and domestic
use in many parts
It was
In
dry climates like North Africa and the Near East wells were an important
means of water supply because of the short rainy season. In many parts
of Africa it is still a basic form of water supply. Wells were used as
water supplies in Britain until the early part of the 19th century, when
boreholes increasingly replaced wells, which continued the principle of
winning water from underground sources.
For wells of about one metre in diameter one person usually had to do
all the work of digging the soil or chiselling the rock material and
filling the containers. Wells of about 2m diameter and larger could
be dug by more than one person at the same time. As the depth of a
well
increased
from
about
10m
the
problem
of
fresh
air
became
important and diggers would have had to be relieved more often than
at shallower levels.
once when
the
bottom
because
of
dizziness
due
to
foul
air.
In modern
(HN
31.49).
Few wells
from
antiquity,
certainly
for
Roman
Britain,
intrinsic water
supply
have
been
function.
Where they have been excavated, the reason generally has been for finds
found in them (Chapter 2, p.41, item 3) . However, some wells have been
excavated in sufficient detail to reveal their lining by steining, such
as examples at London, Silchester, Scole, Southwark and Lancaster (see
Table 4.3, p.108).
lined.
100
that were
For Britain I have been able to trace 350 sites with Roman wells, but
this underrepresents
the
Table 4.2
shows the
number of sites with wells found at the different site types. Column
2 gives the number of sites with wells for each site type, and column
4 gives the number of all sites with water-related (w/r) features for
each site type.
Table 4.2: Sites with wells
Forts
Sites
with
wells
43
Percent*
of well
sites
31%
All sites
with w/r
features
137
Percent** of well
sites in relation
to total sites
12.0%
Coloniae
100%
1.2%
Municipia
100%
0.6%
Civitas
11
78%
14
3.1%
Small towns
52
79%
66
14.8%
Settlements
135
63%
212
38.6%
Villas
104
28%
372
29.7%
Total Sites
350
Site types
features
The
'fort' has been used which includes fortresses, forts and fortlets.
The percentages are not statistically significant, but they do show some
trend in the archaeological record. For instance, the 43 fort sites with
wells is a relatively low percentage of the total number of all site
types with wells, whereas 135 settlements had wells recorded,
nearly
three times as many as at forts. Villas with wells also show a larger
percentage of sites with wells than for forts. An obvious question seems
to be, did a lesser proportion of forts in reality have wells than other
types of site?
101
however,
Table
4.2 lists the sites with 10 and more wells given in Appendix 1. At
222 sites only one well was recorded and at 128 sites more than one and
less than 10 wells were found.
Table 4.3: Sites with 10 and more wells.
Site name
Site type
Number of wells
Newstead
fort
107 +
Southwark
fort
38
London
municipium
51
Caerwent
civitas capital
16
Silchester
civitas capital
76
Wroxeter
civitas capital
17
Tiddington
small town
14
Stonea Grange
settlement
13
Stanwick
villa
12
Thetford
villa
10
source of water below the ground surface, but it is not known to what
extent this was practical guidance to Roman water engineers charged with
deciding on a site for a well. The range of geological and hydrological
conditions will normally determine where wells should be dug to produce
artificial
sources
of water,
in ancient
It
accumulated
was
experience
over
long
period
is also
passed
times
have
likely that
on
to
new
that
some
forts
and
towns
were
deliberately
It would
established
in
geological conditions where water from wells was easily available, such
as the fort at Newstead or the town of Silchester, both of which had
large numbers of wells.
102
Wells are dug to a depth where seepage from a perched aquifer occurs,
such as at a spring line along a slope, or down into the water table
zone, which would provide an adequate level of water in the bottom of
the well so that water could be drawn from it continuously. Shallow
aquifers usually
occur
in gravel
and
sand
layers
overlying
clay,
land or near the coast. Inland, the water table can be at depths of
10m and deeper.
Fissured
limestone,
arenaceous
rock and
sandstone
as mentioned above.
It would be
interesting to check the relationship between Roman wells and the geolo
gical stratigraphy, to discover whether these limestone and other water
bearing material aquifers were exploited by the Romano-Britons when they
dug their wells. How successful the choice of sites for the wells have
been, will depend on how accurately the stratigraphy of the geological
formations have been described during archaeological excavations. Local
modern
borehole
data
would
be
useful to
determine
the geological
formation in which such wells were made. These government reports can
assist interpretation of sites where no water supply features have as
yet been recorded.
An
example
is
Northamptonshire,
Brixworth
since
no
(SP
Roman
had
a villa
been
found
in
during
needed.
The
HMSO
report
four post-Roman
from
3.66m
(12ft)
and
3.35m
(lift)
respectively.
The
report
states: 'There are numerous wells in and around Brixworth deriving water
103
villa probably did have a well(s), but that it has not been found yet.
If the spring was functioning in Roman times it may have been the water
supply for the villa.
A
study
of
the
HMSO
reports
could
also
in Roman times.
supply
information
on the
during archaeological excavations that some of the Roman wells are now
dry, it is possible, based on principles of soil mechanics and knowledge
of under-ground hydrology, in some instances to estimate the drop in the
water table since Roman times. This could indicate what the likely yield
of wells during the Roman period would have been.
There are a number of sites where remains of baths have been found but
no water supply of any kind. The reason for this may be that for many
sites their complete areas were not excavated to detect wells or other
forms of water supply, or the well-heads may have been covered with so
much overburden that it was not possible to find them. Examples of sites
where baths have been found but no water supply are Acton Scott (villa),
Badbury (settlement), Bearsden (fort), Bignor (villa), Braughing (small
town), Castell Collen (fort), Rivenhall (villa).
material
was
removed
by
hauling
it
to
the
surface
in
of
and sandstones,
the
wells
were
and
did
not
as
In rock the
rule
require
were dug were stable, such as at Colchester, although later wells with
no lining have been found. There are examples of unlined wells found at
forts such as at Brecon Gaer
85,
1930,
77),
Newstead
(Curie 1911,
104
33-6),
Richborough
(Arch. Camb.
(Bushe-Fox,
which have been reported as being unlined, the original timber linings
have perished without trace, is an issue which is unlikely to be proved.
Wells in rock were generally not lined because their walls were usually
stable.
material were steined with clay to waterproof them (Hanson 1970, 397). I
cannot see the reason for this, because steining with clay can hardly
improve the stability of a rock well wall, and one of the functions of a
well was to allow water to seep into it. Also, if there was an aquifer
higher up a slope and seepage from it flowed towards the well wall, a
clay surface would be penetrated by the seepage because of seepage
pressure. Wells with wood and masonry linings are discussed below.
4. STEINING OF WELLS
Revetment of well walls is referred to as steining, which consisted
of lining of the well walls with wood or masonry
(White 1984,
157).
Steining of wells was usually carried out using three methods: timberlined,
(Archaeologia
62(1),
1911,
417)
1872,
218-9; Donel, 1993, 1-2, Figs.3-5), or lined with brick (Wilmott 1982a,
2; Hodge 1992, 52). Wilmott doubts the Roman dating of a supposed stonelined well (No.7) at Aldermary House in London, "as no other stone-lined
Roman wells have been found in London" (1982a, 22) . Sometimes discarded
wine barrels were used to line wells,
1982a,
23)
and
Silchester
such as at London
(Archaeologia
61(1),
(Wilmott
1908,
15).
At
Silchester it is also reported that three wells had the bottom part
lined
with
wicker
(Archaeologia
61(1),
baskets
1908,
below
15).
the
The
masonry
opening
of
lining
a well
above
often was
at St Paul-in-the-Bail
1993,
1-2,
Figs.3-5),
in Lincoln
serving
also
as
(Camidge
a
means
1984,
to
15-21;
prevent
92) . The well-head also could serve as the support for a water-lifting
device
such as a pulley
and bucket
arrangement,
or
even
the more
105
those at London
(Queen Street)
(Wilmott 1982a,
9-19),
and
Scole2.
At London there is evidence that at least 13 wells were lined with dis
used wine barrels (Wilmott 1982a, 10, 18, 22, for example well 19 pi.6
and well 37 pi.5, p.12). The use of barrels to line wells must have been
cost-effective, because there would have been no need to construct the
lining, as was the case with the corner-post vertical timber-lining or
box-frames used widely over Britain. Because of the shape of barrels
they were able to withstand active earth pressures which develop as
material
is removed
Sometimes
box-frames
and barrel
linings were used in the same well. The barrels seemed always to be
below
the
box-frames,
the
probable
reason
being,
that
as
the
construction reached the water table, the gravel layer became unstable
because of seepage, so that a preformed lining had to be inserted to
prevent collapse of the walls as further penetration into the water
bearing stratum continued (Fig.4.1, Wilmott 1982a, 27, Fig.19, Well 31).
Corner-post construction of linings was used at many sites such as at
the baths at Cheapside and at Queen Street (Wilmott 1982a, 25, Fig.18,
No.24).
Other
examples
of
corner-post
linings
occur
at
Colchester,
especially regarding
depths and sizes of wells and their lining. The four wells, numbers 20,
24, 35, 36, (see Table 4.4 for depths) indicate how relatively thin the
gravel layer was in some areas of London and how high the water table
was in the layer. Where the depths were greater than about 3m the wells
were in the Warble valley where the gravel layer was about 6m thick. The
gravel layer overlay the very impervious London clay, which seemed to
have determined the depth to which wells were dug.
2 EAA 5, 1977, 108, Fig,46, well I: 111, 112 Fig.48; 113 Fig.49;
F ig.50, 116 Fig.51, and pages 116-7 give dating of the wells.
106
well
II:
114,
115
Fig. 4.2: Box-frames and barrel lining (well 31) from Queen Street Roman
wells, (after Wilmott 1982a,
27,Fig. 19). Wells 22, 31 with lap joints,
and wells 19, 36 are bridled andbraced.
The dimensions of some of
probably
upwards.
Where instability occurred in the gravel layer, barrels were used during
the digging process, which would have prevented sudden collapse.
Another method was to have four corner posts which were mortised at top
and bottom and also in a face at right angles,
so that it formed a
107
or that
need for steining of wells was no doubt learnt from hard experience of
well collapse. Why at Caerwent all the wells were flint-lined
(Hanson
1970, 84) , is not clear, but it could have been because flint stone was
readily available and that wood had become scarce because of other uses,
and expensive. It would seem that the practice was to make all wells
with stone-built
linings round,
because the
surrounding active earth pressure would tend to compress the stone ring,
in that way making it more stable.
some examples
steining which was used in them. Only the wells from London for which
dimensional information has been given are included in the table; many
more wells were excavated,
dimensions
of
indicating
that
some
the
of
the
lining
box-frames
was
are
quite
pre-fabricated
in
small,
The
probably
sections.
Where
instability occurred in the gravel layer, barrels were used during the
digging process acting like caissons to prevent sudden collapse.
Timber-lining of the corner-post type construction was an early type of
steining, used during the 1st to early 2nd centuries, such as the two
wells at Scole.
108
Location
Feature
London,
Queen St
2
16
lst-2nd c.
35
9
20
22
33
1st c .
24
3rd-4th c.
19
36
Great Dunmow
Sll
Well I
Scole
Well II
Caerwent
Lincoln,
St
Paul-in-theBail
16
wells
stone
and
brick
Lining
type
barrel
barrel
L(m)
B(m)
D(m)
Depth(
m)
0 .63
0.84
barrel
box0.66
0.53
frames
box0.68
0.68
frames
1.30
1.07
boxframes
b/f +
0.58
0.49
barrel
timber,
0.99
0.99
vert.
box0.99
frames
0.68
box0.53
frames
box0.76
0.76
frames
Upper part not known
boxc .1.
c .1.
frames
0
2
boxc .1.
c .1.
frames
0
0
stonelined
0 .93
Reference
(Wilmott 198
7, 8, 21,
Figs. 5, 6,
14)
1.68
1.19
6.25
5.24
shallow?
4 .14
2 .44
6.9
41, 1988
25
Figs.4 & 13.
E A A 5, 1977,
112,113
E A A 5, 1977,
115,116
Hanson 1970,
84 .
C L A U 1993,
Rep.63, 1-2
EAA
>4
>3.5
2.4-14
2.4c.16.5
1.2
(sq)
LAT
1984,
15-
21.
Margidunum
Silchester
Templeboro'
76
wells
stonelined
stonelined
(52)
timberlined
(24)
cO .9
stonelined
3.1
3 .65
JRS
2.49.1
Archaeologia
16, 192 6,
37, Fig.3
53, 1893 to
61, 1909
c .9 .1
L and B are the lengths and breadths of timber type linings and D is
the diameter of barrel linings.
with
further
became
the
favoured style
improvements
during
the
of
3rd
lining wells
century
as
progressive
improvement
in
quality
109
of
workmanship.
Further
is required to determine
whether the lining of wells improved over time during the Roman period,
and whether it occurred only regionally or all over Britain.
03m .
0-3m
03m
Villa Faustini,
referred to
in the Antonine
Itinerary (Britannia 1, 1970, 47,) (TM 146 786). The villa originated in
the Flavian period and was still active in the late 4th century (EAA 5,
1977, 107-17). The original excavations were carried out by the Norfolk
Archaeology Unit, who identified four periods for the villa. The timberlining of the two wells (Fig.46 in the excavation report) was identified
covering two phases, the earliest phase dating from the late Flavian
period and the 2nd phase dating
to mid-Antonine
110
which the wells were dug shows that the clays must have been soft
unstable soils,
rotting of
the drier upper part of the shaft (EAA 5, 1977, 111) . For
the lower parts of the timber lining were remarkably
Ill
West
6wi5iK??\VAV*
WELL I
PLAN
of
Phases I and II
1 "
vi *
112
* . .
llmoo
22a00
WELL I
.^^^
60
30
inn
c e n t im e t r e s
Fig.
4.4b:
1977,
Fig.
Well
at
Scole:
elevations
49).
113
of
timber
lin i n g
(from
EAA 5,
West
iw^V'-v'V V.'^V,'
WELL I I
;X>J'<-s
:3
PLAN
Fig.
4.4c: We l l II at Scole:
114
WELL I I
20
tO
60
60
100
C E N T IM E T R E S
12m00
Fig.
1977,
4.4 d : W e l l
Fig.
II
at
Scole:
elevations
51).
115
of
timber
lining
(from
EAA 5,
The method of steining was to insert four vertical wooden corner posts
with flat planks placed between them forming roughly one metre square
boxes. The planks had corner-cuts at their ends to fit round the two
outside diagonal corners of the vertical posts, which also stabilized
the posts in their vertical positions.
thickness from about 5-10cm and had varying widths. The thickness of the
planks shows that the contractors realized there was a need to support
very unstable soils.
The plan view of Fig.
(hatched) which were left in place when they were replaced by the 2nd
phase timber corner posts and lining. The reason for this was probably
that the soil was
so unstable
that
it would have
resulted
in the
collapse of the wall if they had been removed before the new lining was
inserted. The size of the upper portions of the excavation holes (about
3.5m across) to investigate the two wells, seems to indicate the poor
quality of the surrounding soil.
4.3. Styles of steining.
The
styles
of
steining applied
to
wells
seems to
have
been
both
However,
types
of
were used,
but the
quality was of a very low standard in many of them. For London no stonelined wells have been reported,
but,
as Table 4.4
shows,
box-frame
lining with wood was the accepted form of steining, combined in a number
of cases with barrels.
Southwark a very well constructed wood-lined well was found, whereas the
lining at Scole the quality of the lining was of a poor quality,
as
shown by rough hewn timber and the construction had a shoddy appearance
(pp.112-5).
There is a need for further research into these issues of regional
differences, geological conditions and quality of workmanship for lining
116
whether dug
in soils
or
in the
could easily have slumped as water tables rose and the soil lost its
cohesive shear strength in the capillary zone, or due to swelling and
shrinkage of the soil resulting in flaking of the walls of the well.
Many wells must have become inoperative as a result of collapsing soils,
especially
in
the
vicinity
of
the
water
table,
unless
they
were
supported by steining.
The primary reason for steining of wells
against
inward
collapsing
due
to
the
earth
pressure
acting
on
an
some moisture
pressures.
in
it,
which
will
impart
to
it
negative
pore
become
positive and it is these saturated pore pressures which reduce the shear
strength of materials. Soil is saturated immediately above (or within)
the water table and the positive pore pressure acting on the soil within
this zone will reduce its shear strength to a critical low value. At a
certain level above the water table the degree of saturation decreases
due to evaporation
from the
negative,
capillary
providing
surface and
suction between
soil
particles,
become
which
increases the friction between them and thus provide the soil with its
shear strength. If there is a perched water table at a higher level than
the well shaft from which water can seep downwards and sideways to its
117
wall
surface,
this
positive pressures
seepage
in that
flow will
zone.
reduce the negative capillary pore pressures, thus reducing the shear
strength of the soil surrounding the well shaft. If the negative pore
pressure is reduced to a very low value (of the order of about -5kPa to
zero), the soil friction will be unable to resist the horizontal earth
pressure due to the weight of the soil above and will collapse. The
mechanics of the hydrological regime and shear strength of soils around
well shafts are complicated, but have been formulated mathematically,
and their physical implications are nowadays well understood. The Romans
would not have had this knowledge, but they must have had experience of
the collapse of soil surfaces and through trial and error learnt how to
prevent it by supporting it with various forms of revetment.
The stability of the walls of wells can also be determined by the
methods described by Avery (1993, III, 1-10), though there is the added
complication that the well geometrical configuration is often circular,
a factor not dealt with by Avery.
instability,
special
empirical
To analyze a
techniques
have
linear
been
feature
for
developed
to
determine where the active earth pressure will act on a free surface and
it is combined with the water pressure that acts against or on the
surface of a structure or a wall face. A net resultant thrust against
the wall can then be calculated using the procedures described by Avery.
Included
in
be
the
procedure
for
determining
the
direction and slope of the slip surface. This theory highlights the
reason why steining became necessary when wells were dug in potentially
unstable areas such as London, Newsteads,
118
that became
silted up (Clarke and Jones 1994, 117). The shafts of wells at Newstead
seemed generally to have been poorly lined. Curie recorded that of 19
lined wells six were lined with river cobble stones and one with poor
masonry, and others simply with stakes, so that most of the lining was
ineffectual against the unstable nature of the local material
(Clarke
and Jones 1994, 117). The shapes of the wells at Newstead were unusual
because many had a conical configuration of different kinds (Clarke and
Jones 1994, 115, Fig.6), which was probably due to the difficulty of
digging into the cobble and gravel drift materials.
affected the period over which it was safe to draw water from them.
it could be possible
that
seepage
from
refer to the poor cobble revetment of 6 wells and two unlined ones. One
of the unlined ones lasted for about ten years before it was used as a
rubbish pit (1994, 117) . They suggest that if this life span was typical
for the Newstead wells,then not more than 15 to 20 would have been open
at the same time. Similarly, the proliferation of Roman wells in London
in the Queen Street area may have been due to their limited life span.
4 'Germania Romana',
119
Fig.xxx(2).
but whether there was a difference in the quality of the linings found
within and those found outside in the surrounding vicus, is not clear.
Presumably
the wells
inside
the
fort
been
dug by
soldiers, but who would have dug the large number of wells outside the
fort?. It is likely that this type of problem applied to the other major
sites such as London,
Caerwent,
where large
numbers of wells have been found. The problems of stability of the wells
at Newstead emphasizes the need for further research to be carried out
to provide an overall picture of areas of unstable conditions, and how
the problems relating to instability of wells were dealt with in Roman
times.
about
with a
found
3m below
street
level
were
at
the
Bon
stone
Marche
site
at
56) . At Wroxeter
blocks.
Bushe-Fox
reported a well-head with two blocks forming a square with sides of 1.3m
and a hole in the centre of 0.75m diameter (1912, 3-4). At another well
the well-head block consisted of four separate stone blocks,
but no
further details are given (Wright, 1872, 218-9). It has been suggested
that the number of squared oak timber posts found in the well at Bar
Hill may have been remains of a wooden well-head structure (Macdonald &
Park 1906, 40) . The well at St Paul-in-the-Bail at Lincoln had a stone
and brick arched well-head - described below (Donel, 1993, 1-2, Figs.36). The Templeborough fort had a large well
(3.1m diameter)
in the
principia, which had stone blocks from the surface down to the water
surface about 3.1m below ground level
carriers could walk down to the shallow water level to fetch water.
120
A wide variety of methods for drawing water from wells was available,
but there is not sufficient evidence to indicate whether all of them
were
used
discusses
in
Britain
the
remains
during
of
the
many
Roman
forms
period.
of
water
P Oleson
lifting
(1984)
equipment,
individual.
Even
at
settlements
full
and villas,
121
time
water
occupation
for an
demand may
have
from
wells.
This
system
required
well-head
from
which
structure to support the axle and wheel arrangement round which the
chain
could
be
operated.
Various
ways
of
turning
the
axle
were
available, generally a large wooden wheel supported on the axle for the
bucket and chain system. The wheel was powered by a person 'walking'
along the inside of the rim turning it about the axle. The chain would
move around the axle lifting a set of containers attached to it. Even
this method would not deliver a vast amount of water during a day's
work.
It can be assumed that normally one well could serve a single household
with all its needs. Where the smaller establishments of non-military
sites such as small towns, settlements and villas had activities such as
pottery manufacturing,
smithing, dyeing,
may not have been sufficient for the needs of the community, indicating
that the wells are under-represented at such sites. For the major towns
and military
sites
this may
not
have been
a problem
if
they
had
aqueducts and tanks, they would probably only have been standby water
sources. Many of the forts would have been small and a well water supply
would have been the obvious source if the soil or rock formations were
water bearing materials.
but
it is
unlikely that the army would have relied on external water sources for
many of its forts. A possible reason could be that many forts are known
122
hence the lack of evidence for wells or any other features. For
many
sites only the outlines of the defences are known from cropmarks
and very little else have survived. Wells that have been covered over
with deep soil would probably not show up in airphotos.
were
none were found within the town (Crummy 1984, 26-8, Fig.14). The reason
for this may be that there are still large areas of the modern town
covering the Roman town which have not been explored.
At Gloucester a number of wells were found in the Bon Marche area, of
which two were stone-lined
1934,
73; 80,
1962, 56).
At York only 2 wells have so far been located (Yorkshire Philosophical
Soc. Annual Rep. 1901, 104; RCHM Roman York, I, 1962, 53, 59, 61). This
is probably because of limited excavation, because the colonia was built
over gravel and sand formations, where,
(pers.
comm, from M. Jones) , and the other in the east range of the forum. They
must have been important sources of water, considering the uncertainties
about the aqueduct.
123
Limestone Brash
_ U . L _ l_L.I_I.
.l_UI
!_!_
_ L .L J _ L .I_ L
I \L LLLL
L - l- L L L L L
.1 - I L .L L U
L-LLLLLLL.l 1 u|
_ l l .l l .l . i_ l . i_ l . l :
Limestone
LIl
IS
1
tm
_______SM
__________
K^Ba
! 1 1 :II I
1!L
jfl
>W0 m
L.L.1 L
I '1L L
nll\)'
'
-L U L.I
Northampton Ironstone
Lias Clay
Mj=llJ=LL!=iii=lU=n^
'!E I 11=111=11 E l l l S II 1= j
~! 11=111=111=11 IE III =1113
E ill= IIIE illE III= lll= lil
] metres
124
forum
elaborate well-head
beneath
(Fig.
over
post-medieval
4.5)
the
was about
opening.
masonry,
but
discovered
in
state
reasonable
of
preservation. The shaft of the well was 2.4m in diameter in its upper
section and narrowed down to 1.2m square lower down. The masonry matched
that of this part of the forum, and is dated to the 2nd century, which
may have been constructed after the aqueduct. However evidence seems to
indicate that the original shaft may have been sunk during the legionary
period,
which
probably
legionary headquarters.
at
that
time
The forum,
may
have
formed
constructed later,
part
of
the
platform was
four
constructed with
arches
opening a stone-built
resting
on
the
platform
smaller, to
1) . The two
its
long use as
two
(1993,
to the well,
so that what
has
been
found. In Chapter 3 I
presented problems,
so
must make it almost certain that there was an aqueduct supply to the
town. Several Medieval wells have been found in the town as well as
nearby St Albans with average depths of from about 15m to 23m (Hanson
1970, 70). Why so few wells have been located at Verulamium in contrast
to the similar town of Silchester, is strange. It was an important town
with a well developed street system like Silchester and it also had many
private dwellings, and the usual public buildings and two baths, which
would suggest a demand for a plentiful supply of water. An aqueduct is
conjectured from piping along Watling Street within the town but with no
certainty.
1982b, Fig.l).
125
least
51 Roman wells
(Wilmott
For
Wilmott
London,
comments
on the
'special
importance'
of
the
the earliest
Roman activity
in suggesting
an
explanation for the large number of wells found on these sites" (Wilmott
1982a, 3) . The Thames valley in the London area has been formed by the
meandering Thames river over a long period and during that time terraced
gravels
were
deposited
over
the
London
clay
derived
from
up-land
erosion, probably during the major ice age. These gravels were covered
by soil layers through which the Walbrook stream eroded the valley that
bisects the site of
Cornhill
and
Ludgate. London itself was built over the lm thick brick earth layer
overlying the water-bearing gravel terrace
deposition
continued
subsequently and
(Wilmott
the Walbrook
1991,
14) . Soil
valley
almost
for springs
to emerge
relatively easy to sink wells to the water table. There were several
spring lines along this area and the surrounding soil would have been
saturated and had little shear strength, so that wells would invariably
have had to be lined to prevent collapse.
126
'tolling Cl
B oreholes
10
6
Minimum
slope
London Clay
M e tre s
G ra v e l
____________
V OD
-2
Section line
4 b oreho les.
Approx
+B
Level
on
b r ic k e a r th
+G
Level
on
g ra v e l
Edge
levels above
0 .0
J uju
B ric k e a rth .
to p
of gravel in this
of b rickearth
127
100m.
Fig. 4.7: Barrel-lined well from Queen Street, London (site 26), (after
Wilmott 1982b, 239, Fig. 8).
John Wacher had suggested that from the evidence of Roman water pipes
found in several areas in London, the town may have been supplied by an
external aqueduct supply (Wacher 1995, 90, 101)5. Wilmott comments that
the evidence for the abundance of well and spring water sources "argues
strongly against an aqueduct water supply for London"
(Wilmott 1982b,
241) . The evidence of Roman water pipes, for instance, the pipes found
at the Guildhall Extension (site 9), at the Bank of England (Site 20)
and lead pipes found in a well (Site 26) (Wilmott 1982b, 241, Figs.10 &
11) , and pipes near the Mithraeum (near Site 28) and the forum (Site 36)
(Wacher 1995,
101)
(see Fig.
6.19),
suggests
that
a running water
aquae in the vicinity of Cornhill. No evidence has yet been found of any
5 Wilmott would have used Wacher's 1974 edition of The Towns of Roman Roman Britain,
where the same comment is made on London's water supply.
128
coloniae
and
municipia.
Site name
No. of wells
Comments
Colchester
Gloucester
Lincoln
Likely
that
2+?
Wells
51
over most
of
southern
1 (+?)
as
several
medieval
wells
found.
7.4.
Civitas
capitals.
It
seems strange that all the other listed civitas capitals should have
wells but not these four, and again it would seem there could be some
reason why they had not been found.
reported at the
At Caerwent 16 wells have been recorded (Hanson 1970, 84) and it also
had a running water supply as implied by a sophisticated distribution
system (Archaeologia 61(2), 1909, 157). Four of the wells were situated
in houses in the south-west of the town, and the rest in the middle
north, mostly in the premises of private houses.
129
undiscovered
wells
because
the
plan
by
Brewer
(Wacher
1995,
380,
Fig.170) shows that the areas of the north-west and south-east and east
parts of the town had not been excavated by the 1980s.
Both Chichester
and Cirencester
had
four wells,
but
the
other
six
civitas capitals had only one or two recorded at each site6. Four wells
at each of these two relatively large towns seem low, which suggests
that excavation have not revealed the true number, for whatever reason.
At Dorchester 3 wells have been recorded and a further 3 at Colliton
Park Villa
(RCHM(E)
1970,
556) . At
Leicestser
wells
have
been
recorded, though apparently some more have been found, but are not yet
published.
excavation during the late 1800s and early 1900s. Although there was
this large number of wells in the town, neither the public baths, nor
the mansio with its substantial bath, obtained water from wells, and the
indications are that they were not supplied from an external water
supply (Hanson 1970, 126). However, abundant springs in the vicinity of
both baths were their likely sources of water, but it is not clear how
water was transferred from the springs
to the baths.
The wells at
At Wroxeter Bushe-Fox
(1912-4)
(1942)
130
at
the
civitas
capitals.
Site name
No. of wells
Site name
No. of wells
Aldborough
nil
Cirencester
4+
Caerwent
16
Dorchester
2 + 4
nil
Leicester
3+
1+
Silchester
76
Carmarthen
nil
Winchester
1+
Carlisle
nil
Wroxeter
17
Chichester
Caistor
Norwich
Canterbury
by
Wells at the small towns, settlements and villas are poorly described in
the records. Out of a total of 66 small towns listed by me, 52 sites had
wells; however,
small towns (97) listed by Burnham (1986, 187) . The distribution of the
small towns where wells have been found is concentrated in the south and
midlands regions of Britain. The number of wells found at small towns
varies considerably,
to as many as
14 at Tiddington
(Burnham and Wacher 1990, 312, Fig.107). At six sites I recorded W1+,
indicating that there were more than one well at each site,
since I
found some records not clear on how many were found at those sites.
Tripontium and Scole both had 7 wells, Ewell 6 and Ancaster 5, the rest
with more than one well all having two or three wells. Whether the sites
with one well reflect the true situation for an urban community
is
of how closely
unified group.
small
town community
cooperated as
number of wells,
sites with only one well, the remains of others have not been found.
7.6. Settlements.
Wells were found at 135 settlement sites out of a total of 212 recorded
in the database. However,
131
database
is
small
compared with
the
total
number
of
sites
in the
functions
some
form of an
administrative function (Britannia 13, 1982, 366). Also, wells may have
become disused for some reason and new ones dug. At some of the other
sites all the wells may of course not have been found. The number of
sites with wells
supply and the
is
that
form of water
were
far more
(136ft)
deep and the other 50.6m (188ft) deep. These were major structures cut
into chalk, hard enough to have required considerable effort to sink
them, both monetarily and in terms of labour. Comparing the sections
given of the shafts by Pitt-Rivers with more modern drill holes made by
him, seems to indicate that they may have reached the water table when
originally dug.
7.7. Villas.
Villas also used wells as their main supply of water, but the record
again underrepresents the likely number that had wells. The evidence of
remains at many sites is often limited to a few finds and no information
is available about their water
gazetteer on villas
Stanwick
complexes
with
with
12,
supply,
as clearly
shown
in Scott's
several
Thetford
with
households
10,
and
probably
where
indicate
certain
villa
commercial
The
majority of the other sites had records of only one or two wells, and
132
only four with 3 wells as at Ashill, Barnsley Park, North Leigh and
Rockbourne, and two with 4 wells at Colliton Park and Greetwell Fields.
Six sites have been recorded as having had more than one well, but how
many wells I did not discover. At Scole 2 wells were found with their
wood-lined steining still in a good state of preservation (see pp. 1114). Since the excavations reported in 1977 further excavations reported
in Britannia (25, 1994, 278) states that "an area of the 'small town'
was excavated...other possible structures were found and seven wells".
It is not clear to me how far the villa is from Scole, and whether these
new discoveries belong to it, or to the small town.
in wells
and
suggesting
(Ross 1968,
springs
by the
(1997,
136) . She also comments that "despite the fact that wells and shafts
are
firmly
entrenched
in
the
'Celtic
ritual'
corpus,
numerous
Ann Ross (1968, 255-85) gives details of 59 sites where she considers
that 'shafts, pits and wells' had a religious significance. She emphas-
133
izes the great importance religious ritual had for the soldiers of the
Roman armed forces and that many Roman army artifacts are found in wells
of the Romano-British period. Few of these structures are dated, but the
implication for her is that many of them are in origin of pre-Roman
date.
As
previously
result
of
classified
analysis
as
Iron
(Webster
of
their
Age,
fills,
"cannot
1997,
136).
in
wells
fact
be
that
were
shown
Webster mentions
and quotes
Jordan Hill
to
"some
as an
only a
few of
the
structures
is
there any
indication that water was found in them. This may have been due to it
not
having
been
recorded by
the
original
excavators,
many
of
the
excavations dating from the 18th and 19th centuries, or the water table
may have dropped. Also, the topography and geology through which wells
were cut have rarely been discussed. However, the deep structures were
often cut through an upper layer of soil cover to the underlying flint
and chalk rock formations. It would be unusual for shafts deeper than
10m not to have had seepage water from the fractured chalk and flint
formations, particularly if they were not situated on the top of hills.
Many of the sites are generally lower down the slopes of hills or even
in valleys, which would provide conditions for seepage from the aquifers
in the upper slopes where perched water tables could be present.
Ross makes
ritual
shafts,
such
as
the
shaft
at
Minnis
specifically sunk as
Bay
(Birchington)
and
Bekesbourne (Fig. 66, 276). She observes that during the Viking period
at the fortress of Trelleborg "two of the ritual wells appear to have
been cleaned out and used as water wells". She then suggests that might
have happened in "certain instances
towns such as Caerwent,
in Roman Britain,
especially
in
her entries for each one of the three sites with the comments that they
were
"extremely
difficult",
difficult
to
classify",
or
"selection
is
sometimes
cult importance and those having none" in that order. In none of the
Caerwent wells she quotes were the small finds specifically placed or
found in an ordered fashion
she is
much more general in the description of finds in the wells. She mentions
as evidence the ritual use of a well in Insula I, in which a sword blade
8 Ann Ross does not specifically define her interpretation between shafts, pits and
wells. However, she comments 'A shaft of over 120 feet in depth, for example, may also
be a well; and pits seem to have begun their existence as wells which had subsequently
dried up or where digging had been abandoned before water was reached' 1968, 284.
134
broken in two and two iron bars were found, with below them a collection
of almost 60 iron objects.
depth, a collection of over 100 iron objects was found (Ross 1968, 273) .
At Wroxeter, the three wells she gives as examples of possible ritual
use, the type of finds are no different to what have been reported in
other wells
(Ross 1968, 274) . At Bar Hill she refers to the 13m deep
objects and other finds, and interprets this as possible ritual use of
the well. An alternative explanation would be that the Caledonians were
using the well
as
convenient
hole
down which
to
throw anything
suggestive of Roman origin. None of the cases given here suggests the
type of shaft such as found at Bekesbourne, Kent, (about 7m deep), where
urns seem to have been found placed in an ordered fashion on a platform
with other ritual items round them, probably indicating a ritual site
(Ross 1968, 260). Wait is critical of Ross' interpretation of some of
the religious implications of the finds in shaft/wells, but agrees that
some finds represent ritual deposits in the shafts (1985, 51) . It seems
to me that Ross was reading into the finds from the wells more than is
justified. Many of the examples she gives are typical of wells filled
with rubbish or the deliberate discarding of material such as happened
when the army vacated a fort, or when civil sites were abandoned,
or
excavated
sites,
have
produced
only
post-conquest
fills,
in
contrast to what was reported for excavations from the earlier excavated
sites
(135) .
This,
in
fact,
seems
to
be
problem
with
the
makes
it
clear
functional origins
firmly states
'Celtic'
that
tradition,
that
"many
(being cut
wells
and
shafts
"Wells and
shafts
but
to the development
attest
had
principally
(136)
of
an
She very
Iron Age
of Romano-British
structures,
were
partly
instrumental
135
in
the
development
of
cultural and religious interaction between the British and Romans after
the
conquest.
Many
of
the
remains
that
are
described
in
the
Roman innovations and evolved a new material culture which blended with
their own British traditions.
Springs
converted
into wells
practice.
Probably one of
wells
that
is
(Carrawburgh),
Britannia
of
the
first
Romana.
were
the
centres
well-spring
reported
Many
often
votive
in
of
of
of
religious
these
religious
at
Brocolitia
Coventina,
1732
by
offerings
John
were
Horsley
recovered
in
his
including
a bronze dog
and horse,
well were
1985). The
at Aquae Sulis,
(Ross
1967,
30;
Allason-Jones
& McKay
important, derived from the native deity or Sulis. This local god and
the
ritual
classical
associated with
Minerva
(Ross
it
1967,
became
30).
equated
Other
with
springs
the cult
or wells,
of
where
Celtic and local goddesses were important, in Roman times, and seem to
have been adapted as Romano-British religious figures, are: at Buxton,
Derbyshire (spring, for Aquae Arnemetiae) ; at Well, Yorkshire (spring,
local cult); High Rochester,
Chester,
Kent
worshipped.
of new,
idiosyncratic
rites
which
should properly
be
(1997,
Clarke
interpretation by Curie
(1996,
of many
76)
suggests that
artefacts
as
ritual
objects from the wells/pits from the Newstead fort is unlikely. The
more recent
situated
in forts, towns or
136
they were used as wells for the supply of water or were used only as
ritual wells or shafts. In other instances wells normally used as a
water
supply
also
Wilsford shaft
seem
to
in Wiltshire
have
had
religious
associations.
excavated
in 1961-62
The
is
(Ashbee,
et al, 1989) .
that
it could date
period, and also on the evidence of a bronze broad-bladed axe that was
found at the bottom of the shaft. It was found to have a large surface
opening of about 6m (20ft) in diameter, with sloping sides down to about
6m, after which the shaft diameter became constant at about 1.8m (6ft)
to a depth of about 30.6m (101.6ft). Urns with Celtic religious motifs
were found in the shaft at a level of about 18m which seemed to indicate
to some of the excavators that the shaft may have had some religious
significance,
ever utilized
is not
clear.
The
Another
shaft/well
is
from
Wolfhamcote
at
Sawbridge,
undated (Webster 1997, 143). The shaft at Ashill, mentioned above, with
its many objects including urns placed in layers embedded between leaves
of hazel and hazel nut, fibulae, some iron utensils, bones, and flints,
seems to reflect some Celtic custom (Ross 1967, 28). What is puzzling
about some of these comments are that certain artifacts were apparently
found in an ordered state and in a horizontal position.
This would
indicate that they were placed in the shaft rather than thrown down as
9 In the HBMC(E) English Heritage Publication, 1989. Wilford Shaft : Excavations 196062, the excavators who were in charge at different times during the excavations, give
separate reports, each one with their own interpretations of the shaft's function and
what parts they played in the excavations itself. Each one has definite opinions about
the original function of the shaft.
137
is found in many other wells. The question here seems to be not their
dating or whether they originated from a particular ethnic group, but
how they came to be in those specially placed positions and why? For
such instances where the finds down a well or shaft seems to have been
deliberately placed at a certain level,
that these shafts were in fact used for a ritual purpose, whatever their
original function may have been. I think that in the confined space of a
well from which water was drawn, it could hardly also have been used for
ritual purposes with objects placed at specific
movement
of buckets
levels,
because the
them.
Votive objects may have been thrown down a well, but that would not be
the same as using the well as a ritual shaft.
encounter with
native force, which had overrun the site or forced a hasty withdrawal"
(Clarke 1996, 73) . Ann Ross, in contrast, suggests that most of the pits
were ritual shafts (Clarke 1996, 73). Finally, he observes that Manning
takes a view opposite to both Curie and Ross, arguing that 'by and large
the deposits were quite normal for Roman period sites, representing the
simple disposal of rubbish'
main issues: 1) that very competent scholars using the same raw data can
reach
totally
opposing
interpretations
based
on
archaeological
significance
of
site,
it
seems
to
depend
on what
one's
'Celtists'
archaeologists.
Clarke
suggests
that
there
is
(Clarke
1996, Table 2 68). Clarke says that from their recent analysis of their
own excavations
that of the
hundred pits and wells found, only a limited number appears to have been
in use at any one time.
138
cast
into wells
for
some
ritual
purpose by
individuals or groups, but this would not necessarily mean the well was
a religious well. Merely the association of deep down in the unknown and
the presence of water would have provided the motivation to cast votive
objects into it, which would not have detracted from the use of the well
as a water supply. Evidence that some shafts were used primarily as
religious features have been suggested, but I am not sure that it can be
categorically accepted as having been positively proven. It would seem
that much past evidence needs to be reexamined in order to provide more
rational answers to this complex problem.
9. CONCLUSION
Wells and springs were important sources of water supply to many sites.
Although many of the wells were lined,
effectively
supplied to baths
and
other
wells
their collapse.
towns
like
Caerwent,
London,
Silchester
and
Wroxeter,
and
for
139
and settlements had wells, but it is considered that there were likely
to be many more sites with wells, and more wells within sites than shown
by the archaeological record.
This raises the issue whether there were professional well diggers who
tendered their services to owners of homes to dig their wells, because
although the operations are simple, there is a certain amount of general
knowledge required to be efficient at well digging and to be aware of
the dangers involved. The same question may be asked about well digging
for public use, and the contractual implications that would have been
involved. Was this part of the duties of the local builders? Where
geological conditions were not difficult for digging of wells, private
owners of homes may have dug their own. Most men would probably have had
the skills to build their own homes and may also have been able to
construct their own wells and their linings. The constructional aspects
of well digging for the Romano-British period are poorly understood and
needs to be investigated.
Very little evidence is available on how much water was needed by a
community, and how much water could be drawn from wells to provide their
needs.
from wells
and
from rainwater
catchment could have been adequate for only the smaller baths. In the
next section I shall be discussing baths over a large range of sizes and
the question of how effectively they were supplied with water, which at
present is not well understood.
140
CHAPTER
5.
aspects
of Roman baths
in
runoff.
Some
military
sites
and
towns
had elaborate
sewer
systems for the removal of waste materials. In order that such systems
could function efficiently many sites had a running water supply to
flush waste materials into the sewers, and to facilitate cleaning of
drains and streets.
Excavation reports regularly present the notion of continuous dynamic
change throughout the
but
The
reasons for the changes at sites are varied, of which the most important
ones were wear and tear of structures, destruction by fire - especially
of earlier wooden constructions, abandonment of a site for a period, new
occupation
or
ownership,
the
need
for
improved
facilities,
and
to
villas.
is
epigraphic evidence (all from forts) for repairs to baths (Welsby, 1980,
89-94). An inscription from Bowes
probably
from the
fort
of
drains,
alsoallows
which
one
to
141
consider
developments
in
bath
I have traced evidence for the remains of 488 Roman baths in Britain at
all site types. Baths were introduced to Britain by the army as they
advanced west and northwards.
(1990) and
Yegill (1993), both accounts being based on the principles laid down by
Krencker,
et
al.,
(1929).
I shall
used
at
both
not
discuss
the
elaborate
bath
military
and
other
site
types,
with
The superstructures of
baths in Britain are poorly known and most of the reconstructions that
have been suggested usually are based on counterparts from elsewhere in
the Empire.
I have found dating of structures a problem because it is often not
known when during the life of a site the various structures were built.
Although baths can be easily identified because of their distinctive
layouts (heating, drainage and decorations) , in the absence of detailed
excavation, it is not generally possible to provide reliable dates for
their construction, useful life and abandonement. Inscriptions usually
indicate when a structure was dedicated, not necessarily constructed,
but the information is nevertheless useful. Dates are often referred to
as being during the early, mid or late parts of a century. For some
sites circumstantial evidence may be known, such as when events took
place during a governor's period of office, for which dates are known.
The sites related to the Antonine Wall can generally be dated to within
narrow
limits
because
of
the
two
short
periods
of
occupation
of
Scotland. Some lead and ceramic pipes or tiles can be dated because of
stamps found on them.
Millett
Roman towns, only by the century during which they went out of use, were
destroyed
or demolished
(Millett
1990,
130,
Table
6.1).
For
these
142
referred to
in broad periods.
For
instance,
it is
It
known that
the
Hadrian's visit
to
the
given by excavators
based
on
but often
these only give an end date for the structures. Reference to dating of
structures has therefore been limited to those that are fairly certain.
2. ROMAN BATHS
The subject of Roman baths is vast. It has been documented in detail
during the 1980s by Manderscheid in a number
plemented
by
Inge
Nielsen's
seminal
books
of publications1, sup
on
Thermae
et
Balnea2,
covering baths from all over the Empire, and Fikret Yegul's Baths and
Bathing in Classical Antiquity.
baths,
sections
of
the
civilian community.
The bath
institution
became
in Roman Britain
on the
same commercial
basis
as
were
their
communal
pleasures
and
companionship,
and
where
they
could
socialize, drink, eat, have massages. The ritual of daily bathing was
generally popular among the dlite and wealthy members of communities
throughout
the Empire,
but
the poorer
classes
also used
them.
The
entrance fee was low enough for all people to have access to bathing
(Nielsen 1990, 1.132). It would seem that baths were lucrative business
enterprises which attracted many other
within their precincts
forms of business
activities
the other nations conquered by the Romans adopted the bath as a part of
their own culture under Roman rule, hence the large number of baths
traced by me from the literature for Britain.
1 Manderscheid 1981,
1983,
In Vol.II
143
Natascha Zajac, has posed the question, "Did the Romans go to the baths
to get
clean or be dirty?"
(Omnibus 32,
1996,
16-20),
which
is a
also notes that the baths were nicknamed "the palaces of water" (p. 19).
Zajac
the
public baths, where all sorts of traders, pimps and the like, came to
attract business from the bathers. New research into the customs of the
Roman world is providing a panoramic view of the ordinary daily social
behaviour of Romans as exemplified in their use of the public baths and
other amenities. Katherine Dunbabin has commented on the large amount of
epigraphic evidence relating to baths, which extols both the pleasures
and dangers of bathing, and also its curative value (1989, 7-12) . There
seems
behaviour that
Juvenal
(Let.
56,
86)
complains
about
the
(Omnibus, 1996,
19) .
'deafening noises',
the
panting and grunting of men swinging weights, and the smacking noise of
body massage, the shotting of traders advertising their wares, and other
complaints.
Such
contemporary
comments
give
an
insight
into
the
and
lacked the grandiose atmosphere that was so typical of those of Rome and
possibly also those of Pompeii and Ostia. On the other hand, many of the
vici at forts in Britain had Roman baths and, since their inhabitants
provided some of the commercial and service needs for the army,
likely that
similar activities
it is
baths
in
144
but simplify them to fit in with the layout patterns found for Roman
baths in Britain. Krencker
within
designated a code of
letters related to each one of the areas to indicate them on plans, and
most scholars who have written about baths have adopted his conventions
(Yegill 1992, 130-2). The typical Roman bath had a caldarium, tepidarium,
frigidarium, the usual sequence in which it is believed they were used.
Additional
facilities
were a
laconicvm or
(changing room),
sudatorium
room),
apodyterium
latrines,
and a
(dry
sweating
apraefurnium(furnace)
Britain. Heated water introduced special problems and the large baths
would require greater volumes of water which would have had
to be
Roman baths developed from the YUfivaoiov (gymnasia) and Pc&ocverov of the
Greeks. Vitruvius gives a basic theoretical approach to Greek baths and
gymnasia for the Hellenistic period
gymnasia) , but his model was superseded by the development of the Roman
bath during the later 1st century BC and the 1st century AD.
specifically characterize Roman baths
during
the
early
first
century,
What
which
brought a whole
new
social
Yegttl 1992, 356-95). Nielsen and Yegul4 both seem to agree that Roman
baths
were
broadly
divided
into
two
types: large
public
baths
145
these additional
definitions
the
facilities
eleven
(Nielsen 1990,
large
monumental
3) . According
baths
in
Rome
to
were
thermae, as were many other baths with palaestrae found elsewhere in the
Empire. Baths without palaestrae were classed as balneae,
as used in
this thesis.
Inscriptions
generally attest
the
construction
of
baths
by
private
citizens and that they had a commercial purpose (Yegttl 1992, 43 no.97),
but none has been found for Britain. During the Republican period all
Roman baths used by the public were described as balneae,
and were
thermae
(approx.
lha)
in 25 BC,
which
146
may have had palaestrae. There were many variations of each type and the
intention is not to discuss details of classification,
but merely to
London,
Silchester and
Wroxeter. The reasons for these towns also having thermae type baths are
probably
because
of
their
large
populations,
the
healthy
economic
climate during the 2nd and 3rd centuries and that the inhabitants of the
towns had become used to their daily bathing ritual. I have not been
able to ascertain whether any other sites had thermae, but it is likely
that some of the other main towns and legionary fortesses had them.
The
large
Roman
baths
were
notable
for
the
richness
of
their
Both
of bathing.
The
There has
of
(Fig.5.1)
architectural refinement
achieved
by
He makes
comparisons
between the
four
the
Roman
army
(1986) describes
similar baths
at
(115-29) ,
all
of
the
same
architectural features.
147
vintage,
with
very
similar
Fig.
148
The reconstructed exterior of the thermae and the palaestra with its
swimming pool and sports area are illustrated in Fig.5.1.
Legionary
(Zienkiewicz 1986,
27) . None of the baths found in Britain compares with the eleven large
imperial baths of Rome whose construction seems to have been motivated
to show the power of the Emperor and to win the confidence of the
people.
(1990,
II,
19-20) . I have
added
6 The areas calculated are only estimates, since they are based on plans given in
various texts and it depends on how the dimensions were used. Different texts will
seldom give the same areas for the same sites, as for instance the total areas for
Wroxeter are variously given as 0.81ha, 0.53ha and 0.45ha, depending on the areas that
are included as part of the baths, such as the marcellum and the piscina, both within
the bath precicnt.
149
Table 5.1: Areas of some Roman baths in Britain (after Nielsen 1990, i:
19-20, and an additional selection).
Viroconium
Ratae
Cornoviorum
Corieltauvorum
(Wroxeter)
(Leicester)
Castell Collen
(Corbridge)
Corlosopitum
Gelligaer
Calleva
Atrebatum
(Silchester)
civitas
civitas
0.42 0ha (t h e r m a e )
fort
0.103ha
fort
0 .230ha(p a l a e s t r a O . 140ha)
fort
0 .2 1 0 h a ( p a l a e s t r a
civitas
0.069ha
Hunnum
(Halton)
fort
0.065ha
Aesica
(Great Chesters)
fort
0.054ha
fort, v i c u s
0.049ha
fort
0.044ha
fort
0.044ha
Cilurnum
(Chesters)
(Carrawburgh)
Brocolitia
Condercum
(Benwell)
(Chesterholm)
Vindolanda
fort,
vicus
0.130ha
(t h e r m a e )
0.031ha
(Caerleon)
fortress
Deva
(Chester)
fortress
0.64ha
fortress
c .0.4 0 h a (t h e r m a e ,
small town
c .0.22ha
municipium
0.131ha
(t h e r m a e )
civitas
0.12 6ha
(palaestra
Isca
Dumnonlorum
(Exeter)
Bath
Londinium
(London)(Huggin Hill)
(Dorchester)
Durnovaria
Red House
Chedworth
(north bath)
Caerhun
Rockbourne
(West range)
0.109ha(0.046ha, p a l a e s . )
0.032ha
fort
0.03 0ha
0.028ha
0.029ha
fort
0.024ha
fort
0.021ha
Bearsden
fort
0.016ha
Bothwellhaugh
fort
0.014ha
Bewcastle
Brecon Gaer
(internal)
?)
fort
villa
(S-E bath)
Bidwell)
villa
villa
Great Witcombe
(t h e r m a e )
The listed baths from Britain are a whole order smaller than the baths
of Rome or some of those from North Africa. In contrast the areas of the
large British baths were relatively small in comparison with the bath
complex of Caracalla in Rome which occupied a total area of c. 11.05ha,
of which the area without the palaestra was 2.508ha, and about 1200m2 of
this area would have been used to hold water (Nielsen 1990, vol II, 3,
item C.8). The large thermae at Carthage covered a total area of 3.99ha
of which the bath-suite had an area of 2.625ha (item C.208). Most of the
large public baths in Rome and elsewhere were imperial baths sponsored
and funded by emperors, whereas the public baths at urban centres in
150
Britain
were
probably
funded
locally
by
the
towns
and
financial
is that
For instance, the total area of the Caerleon fortress bath is calculated
to be from c.
whereas the area of the three rooms of the bathing area is only about
O.lha (12.9% or 14.8%).
At Red House fort the area of the bathing block was c. 0.0634ha and the
area of the palaestra was 0.0458ha, which occupied 42% of the total area
of the
complex
0.133ha,
while
total
area
including
the
basilica
is
about
(palaestra)
Nielsen's
list
are of
five forts
in
at
Wroxeter and the legionary bath at Caerleon. The baths at villas are
also very much smaller than those at fortresses, forts and towns. The
area of the S-E bath at Great Witcombe villa is only about 4% of the
total area of the baths at Caerleon and only about
Wroxeter.
6% of that at
bathing-suite area of Caerleon, then its area is nearly 30% of the area
at Caerleon and about
22%
of
To
illustrate
at
those
of
Wroxeter,
Caerleon,
Carthage
and the
Baths
of
7 Wroxeter's second public baths may have been sponsored and paid
Hadrian (Barker 1990, 2).
for by
the Emperor
8 Another problem is that plans given in the literature are often copies of either
original plans or even copies of copies, so that measurements made from these and
using the scales provided can introduce errors. Therefore areas calculated from these
plans can only be approximations of what the remains actually represent on the ground.
151
An important aspect of the size of baths would have been the cost of
running them, of which the cost of fuel would have been a major factor.
Fuel requirements of the smaller more compact baths would have been much
less than for the largerbaths because the volumes of rooms to
be heated
would have been smaller. The larger baths would have been more difficult
to heat efficiently because of the large size and height of the rooms.
The ratios between the larger baths of Britain compared to the large
baths
of the Mediterranean
indicate that
their heating
costs would
possibly have been much lower, a probable reason for the more modest
size of baths in Britain. To what extent the colder climate of Britain
would have influenced heating costs of large sized baths compared to
those of the Mediterranean countries would require special study, but
certainly during the winter months more heat energy would have been
required per unit area in Britain than in the southern Mediterranean
warmer areas.
2.3. Layout of baths in Britain.
Both Nielsen and Yegill
across
the
Empire
discuss
with
often complicated
classifications. From
an
baths
in
the
other
provinces
took
on
exotic
plans
with
the
152
However there were exceptions, such as the later small internal baths at
Brecon Gaer and Caernarfon for reduced garrisons when the areas of the
forts were reduced (Johnson 1983, 193-4). Excluding the bath-suites of
commanding officers, which were always within the defences,
the bath
houses for other ranks were usually outside forts. Sixteen bath-houses
are known in the zone of Hadrian's Wall, of which four were inside the
forts, Bewcastle, Carvoran, Halton Chesters and Risingham
(Gillam,
et
al, 1993, xv), the others all being extramural in the nearby vici.
Some of the smaller forts had the linear type (1) layout with all the
rooms along a single linear axis, with apses sometimes projecting from
the sides, or some small rectangular rooms appended to it, and often an
apodyterium was added at either end. Examples are the forts of Halton
Chesters, Bearsdon and Bothwellhaugh (Fig.5.2); block type (2) baths are
found at the forts of Benwell, Carrawburgh,
(1983,
221), but this seems to be unusual. There is a large room to the right
of the frigidarium which is unmarked on her plan and also on Roger
Wilson's
plan
(1980,
62,
Fig.75A) , with
its entrance
on
the
right
leading from the large enclosure (see Fig.5.3). This unmarked room may
be the apodyterium, and the large room is the palaestra, or an enclosure
with some other function, perhaps of a religious nature, since it has
the
portico
in
its
centre.
Daniels,
the
excavator,
called
this
baths is 1:900).
The size of these changing rooms vary considerably. At the four block
type sites,
Benwell,
Carrawburgh,
(Fig.5.3)
they were the largest rooms in the bath-suites. The four bathing rooms
(hot, tepid, cold and dry hot rooms) at Vindolanda forms a symmetric
square
block,
as
shown
in
Fig. 5.3.
Next
to
the
cold
room
is
rectangular cold plunge bath, and the rectangular apodyterium with its
entrance to the bath-house,
lies at
to the cold
153
154
(b)
Fig. 5.3: Block type layouts: (a) Red House, (b) Carrawburgh, (c)
Vindolanda, (d) Benwell, (from Wilson, 1980, Figs.75(a), (b),.(c), (d)) .
(Scale 1:380)
155
Fig. 5.4:
(1:900)
156
78, Fig.87),
Bewcastle,
Chesters,
Great Chesters and Vindolanda were in the active military zone, and for
those sites probably the important need was to provide a basic service,
rather than additional facilities such as palaestrae. Also, space within
many fort enclosures did not allow for large sports areas. At Strageath
the bath-house was of the simple row type with only a caldarium and a
frigidarium,
ramparts and the barracks (Frere & Wilkes 1973, 1989, Fig. 53).
Chesters is an example of a fort which went through several phases of
changes during its occupation. George Macdonald produced a study of the
site (1931, 219-304) in which he gives details of the construction of
the baths and its changes,
report.
At several places Macdonald refers to the changes that were made to the
structure over time, as at the entrance to the apodyterium. He comments
on the drain which starts in the frigidarium and flows out towards the
apodyterium and into the drain of the latrine adjacent to the east wall
of this room. There seems to have been a change in the direction of flow
since the original drain flowed in the direction of the north-west
corner. It would seem that new floors were laid in the hot rooms at this
time, most probably because the original floors had deteriorated due to
the heat from their hypocausts. There appears to have been evidence that
at a later stage the drain area in the frigidarium settled and that
repairs were made so that flow could be maintained to flush the latrine
drain.
Evidence shows that the lintels of the doors between the caldarium and
the plunge bath to its north were altered at some time. The floor of the
tepidarium was also replaced and so was the roof of the hypocaust of
that room. The floor of the sudatorium west of the apodyterium also was
replaced.
Little
supestructure,
Although
no
can
be
said
about
repairs
or
changes
to
the
dates
are
given
by
Macdonald,
it
is
clear
that
many
alterations and repairs were made over the useful life of the baths.
Such repairs must have been standard operations at most military sites,
and especially after there was a period of non-occupation.
157
The
civitas
Silchester
capitals
of Wroxeter
(0.533ha),
Leicester
(0.42ha)
and
classed
as
balneae,
probably
dictated by
economic
limitations
Large towns.
At WJroxeter the original three bathing rooms of the Hadrianic baths was
of
the
row
type
pair
of
lobbies
(6 and
7)
separating the frigidarium from the tepidarium. During the later 2nd
century,
(Wacher 1995,
367),
another caldarium (17), converting the baths into the complex type (3)
layout. The second laconicum (11) was enlarged (shown as a tepidarium by
Wacher, 1995, 47, Fig.11), and changes to the heating system was made to
accommodate the additions. The bathing rooms all had rectangular plunge
baths at both ends,
rectangular plunge bath. Why the second set of smaller baths were added
is uncertain, but they may have been used to also accommodate bathing by
women. How long the baths survived is uncertain but by the late 4th
century the basilica had been dismantled, as was part of the baths. It
is difficult to say whether the dismantling process was a deliberate act
to destroy the structure, or whether the building had deteriorated to
the extent that it became unsafe and had to be pulled down. The bath
dating from the fortress period also had a linear type layout (Atkinson
1942) .
158
SITE ENTRANCE
ANO MUSEUM
B A TH S
B A SL ICA
FRGOARIUM
10
CALDARIUM
COLO
PLUNGE
d lA C O N IC A
TEPOARIUM
C C fl
Fig. 5.5: Wroxeter baths layout (Webster & Barker 1991, 16-7).
The baths at
Leicester
next
(3)
caldaria and
layout.
The
excavators did not speculate on the reason for this triplication of the
warm baths. They may have served different sections of the community,
but there
an area of about 140m2, had apsed plungebaths along their outside walls
and at their west ends. The centre hot bathroom had a rectangular plunge
bath at its west end. Each caldarium had its own furnace, which must
have used an enormous quantity of wood fuel to heat them. Then follows a
tepidarium and what appear to be two plunge baths at its north and south
ends, and two laconicae next to them.
159
M etres
Fig. 5.6: Leicester baths layout (from Wacher 1995, 45, Fig.10(1)).
The frigidarium was a large room (c.220m2) whose entrance to the baths
was from the west wall of the adjoining palaestra, of which the existing
remains is known as the Jewry Wall. The baths seemed to have had an
extensive drainage network with a major drain running along three sides
of the complex.
The
Silchester
(0.069ha)
public baths
including
its
palaestra
than
the
baths
at
Wroxeter
and
Leicester. Its baths also had a sequence of hot, tepid and cold rooms,
but they were not in the true row type alignment as can be seen from the
plan. The layout suggests that the original bath may have had the row
type (1) layout, but subsequent additions and alterations converted the
baths into a complex type (3) structure. The small caldarium may have
been the original hot room with its furnace at the rectangular plunge
bath end, but a later larger caldarium seems to have been added with a
larger furnace. The room also had two plunge baths, but their positions
have been reversed.
while the
frigidarium had a plunge bath larger than the cold room itself.
PA
Palaestra
LA
Laconicum
Latrine
Frigidarium
PA
Tepidarium
Caidarium
Plunge Bath
PR
Praefurnium
Drain
PR
Service
Apodyterium
PR
PR
100
10
Fig. 5.7: Silchester baths layout (from Wacher 1995, 45, Fig.l0(2)).
Next to the frigidarium is a room referred to by Wacher as a service
room, which may at one time have been part of the frigidarium.
The
entrance to the baths was through the apodyterium, which was adjacent to
the largest room, the palaestra.
The size of the individual bath rooms at Silchester was smaller than
those of Wroxeter and Leicester, probably indicating that the number of
people using the baths were not as numerous as they were at the other
two towns, and/or that the town could not afford to have a larger public
bath.
The evidence indicates that known civilian public baths were of the
complex type layout, even though they may originally have had a linear
layout. Some
of
the
other
larger
towns
have
not
been
excavated
sites
to
provide
better
structures.
161
understanding
of
these
complex
Small centres.
The needs of small towns, settlements and villas would have been quite
different to that at forts and large towns, and the smaller size of
bathing areas indicates that at these sites much lower numbers of people
used the baths at any one time. The baths were usually part of a single
unit of buildings, seldom freestanding, especially at villas, and their
size had to conform to the size of the rest of the buildings.
The small town of Godxnanchester9 had a bath-building occupying an area
of about 365m2 and the bath-building at the small town of Well10 was
even smaller
(c.
Neither of the baths at these two sites had anything more than the basic
requirements for bathing. It is not clear whether baths in small towns
were available to the inhabitants of the towns or whether they were the
property of private owners.
An exception seems to have been the early bath at Gadebrldge Park villa,
which initially was separated from the main house, and later was linked
with secondary rooms.
The baths at the small town of Bath were a special case. They developed
as spa type baths and were probably run on a commercial basis. The baths
were adjacent to the temple of Minerva, which was associated with hot
springs with ritual connotations, which may have affected the status of
the baths. The baths had a long history of development, there being four
periods of major alterations. The bath started as a small spa at the
west end near the spring end as a bath-suite with the complex block type
(3)
layout
(Cunliffe 1984,
140,
Fig.85)
(Fig. 5.8a).
Eventually the
Great Bath was constructed with an area of c.300m2 inside a hall with an
area of about 875m2. The hall was magnificently decorated with portico
pillars holding up a high roof. A smaller bath was added east of the
main bath,
caldarium and tepidarium and other rooms were added at the extreme east
end, which was of the linear type (1) layout. The total area of the bath
complex during the 3rd period was about 0.22ha. The bath complex had
elaborate
water
supply
and
drainage
systems
(Cunliffe
1984,
145),
(Fig.5.8b). A lead pipe was still in place when found, which shows the
soldered joint.
9 JRS 49,
1959,
116,
Fig 13.
10 Gilyard-Beer R, 1951,
162
11,
fig 2.
Figs 5.8a, b: The baths at Bath, and their water supply and drainage
systems, (after Cunliffe 1984, 140, 145).
Villas.
Villas developed from the early 1st century simple cottage type houses
to the variations in sophisticated layout of the 2nd to 4th centuries,
which would have depended on the wealth of their owners.
The small
villas of the early period would have had simple bath arrangements of
one or two rooms. With the passage of time many villas would have
developed into the elaborate layouts found all over Britain south of
Durham, whose occupation extended into the 4th century, indicating the
considerable wealth of their later owners. They, like other site types,
went through continual change and this was reflected in the different
phases of bath construction at many sites. Several villas of the later
period had more than one bath-suite,
(Wilson 1988, 128), and
R ockboum e
for example,
N o rth
L e ig h
had two
163
more than one household lived on the site, or to separate the family
bath-suite from that used by the workforce, or simply because the owner
wanted an improved bath layout. In the case of Rockbourne, over its long
period of occupation, the first two sets of baths went out of service.
The more sophisticated layout and fancy construction of the east bath of
the
4th century
indicates
a wealthy
owner.
Villas
varied
in their
layout, but must have evolved over a long period at some sites such as
Northleigh, where it started as an L-shaped site and eventually, in the
4th century, became a courtyard type villa enclosed on all four sides.
It was probably during periods of alterations and additions that some
villas acquired more than one bath and had mosaics laid out on the
floors.
Rockbourne villa had a complicated history of changes and development
extending over three and a half centuries. The RCHM(E)
report
(1983,
129-50)
through
to
the
lower
construction of the
levels,
villa,
that
is
to
the
earlier
phases
of
Pottery finds indicate the site was continuously occupied from the late
Iron Age
layout)
building
for about
was
350 years.
The villasite
(Fig.5.9
for partial
being a
round
hut
the
first
It was
modified
(phase 4) with
recommissioned
modified
house. During
to provide
for rooms
to
(phase 6)
development took place on the north range, and additions were made to
the west bath.
164
Fig. 5.9: Rockbourne villa, west and east baths (from RCHM(E) 1983, Fig.
2)
(3rd
During the 8th phase, probably during the 4th century, the east baths
were extended with the addition of a plunge bath and drainage and an
extension towards the east. This brief statement is intended to indicate
the evolution of this villa site over time,
number of reasons, the most important being that the structures would
have needed continuous repairs and replacement because of deterioration
over the 300 years of Roman period occupation. Also,
165
under new ownership at different periods who may have added or replaced
structures at the villa.
The Great Witcombe villa's original bathrooms were linearly aligned with
the west wing of the house, constructed c. AD 250-270. Later alterations
were made over the period c. AD 270-400,
I lypOCiHist
7rmrr"
X lA C jO tl
.tk lio n )
-isjauuDt;
C OHItlDOH
(Sol I le v ie s
buttresses
BAUC OF P O H flC O
CA I DA I H U M
mosaic
fj
I MlClUAHItlM
Fish mosaic
slops
itn.oiuMry
b
'.Ot t I I IIN ,| llA III
Drain
a L A T R IN c ^ '
C A L D A R IU M
mosaic
Plucked
SUOATC
to
m o s a ic
FRIGIDARIUM
Fish mosaic
s te p s
W \ 'TEPI^RlU^Jp^
^Tepid.^0afuS
Mbath
rj
s te p s
! 7u
0 ;
" W
U.
s e c o n d a ry
'B
C O L D P L U N G E B /V T H S
O .u iu P ju d b i b ut> bulc*iiu
Drains
Fig. 5.10: Great Witcombe villa baths, (Britannia 1, 1970, 294, Fig. 9)
166
The layout seems to be poorly planned and not squarely aligned with the
rest of the house. The new caldarium (Room 5) , within Room 4 of the
original house, was constructed with a new entrance and the old one was
blocked. It had no hot plunge bath. The original Room 6 now became the
frigdarium with
an apsed plunge
bath
(7)
on
its
east
side
and
rectangular plunge bath (7a) along the south wall of this room, which
was
damaged
by
subsidence.
The
impression
seems
to
be
that
the
rectangular one was built first and when it was damaged the new apsed
plunge bath was built. On the west side is a large sudatorium (10) with
a hot plunge bath (lib). A large tepidarium (8) with a plunge bath (8a)
is sandwiched between the sudatorium and the frigidarium.
The space
between the north-west wall of Room 4 and the caldarium was a latrine.
The bath-suite had an overall area of about 290 m 2. At least four drains
were associated with it. There is a change from the west baths linear
type layout to a more complex one for the east baths. This was a major
change
to
private
estate.
Over
period
of
time
other
major
Gadebridge Park Roman villa was also a large villa estate developing
through six phases from a simple homestead of only a few rooms. The
original freestanding baths (Fig.5.11) of the linear type (1) were built
during the 1st phase, and improved during the 2nd and 3rd phases. They
were considerably enlarged during the 4th period in the late 2nd or
early 3rd century and it was during this time that the house and bath
were connected by other rooms. During the 6th period, c. AD 325 a large
bathing pool was built approached from the baths.
They were in Use for nearly 150 years; the excavator suggests that they
may also have been used by the public11. This would have been an unusual
use of baths at a private estate,
correct.
The
villa
was
served
by
well
water
supply
until
the
enlargement during the 4th phase, when the leat water supply from the
1st century was improved (Neal, 1970, 69-71).
11 Neals D S (ed.), 1974, 68-9, 73-75. At p . 75 Neal says that "In the light of present
evidence the pool can be assumed to have been for public use, but whether it had
religious associations is far from clear".
167
Fig. 5.11:
life styles
of a
the Britons
had their round houses and within a short period they adopted the linear
form of Roman type housing, which finally culminated in the elaborate
villas of the 3rd and 4th centuries,
the
villas
discussed,
which
were
changed
significantly,
was
168
villas which started off as simple homes and developed into prestigious
villas.
Table 5.2: Prestigious villas.
Site name
County
Dating
Chedworth
Gloucestershire
Great Witcombe
Gloucestershire
Woodchester
Gloucestershire
Cosgrove
Northamptonshire
Fishbourne
West Sussex
AD 75 - c. AD 280
Gadebridge Park
Hertfordshire
Lullingstone
Kent
Northleigh
Oxfordshire
Rockbourne
Hampshire
mid-lst c - 4th c.
of
their baths
of the
floors with
"expressions
of
only
foundations,
is
problematic,
because
what
characterizes
buildings are usually their outward appearance. However, from the floor
plans of the remains of Roman baths some characteristics are apparent.
All site types with baths showed the three basic room types of hot,
tepid and cold rooms. The earlier baths for both civilian and military
baths seemed to have had the row type layout, with or without plunge
baths.
Military baths seemed to be built in rigid layout patterns of the row
and block types with not much variation other than the size of rooms.
169
fortresses and
for exercise
and
sport
layouts
compared to the military baths. Their plans appear to be more bulky, and
more ornamental than the plans of the military baths, partly because of
their enlarged apsed plunge baths. They also often duplicated bathing
rooms such as at Leicester, Silchester and Wroxeter. The military baths
seem to have a rigid style of layout, while the civilian public baths
seem to indicate an entrepreneurial freedom. The initial layout of the
civilian Hadrianic baths at Wroxeter seem to be not very different from
the earlier military baths described by Busche-Fox, both of the linear
type. This may imply similar origins in the earliest plans of civilian
and military baths. The duplication of bathing rooms may have been due
to the continual changes that were made to the layout of civilian baths
because of a demand for more bathing facilities, and also to create
separate facilities for men and women. Except for the large legionary
baths at Caerleon and Exeter, the bathing rooms of the civilian baths
are somewhat larger than their military counterparts. When town public
baths were modified they were often extensively altered, which do not
seem to have been the case with military baths. Repairs seem to have
been the main additional work carried out on military baths.
Baths at villas were generally small compared to those at military and
town sites. The early villas started with simple bath layouts and they
were often of the row type,
complex type (3) layout as new owners reconstructed and made additions
to
them.
feature
that
characterized
many
villa
baths
were
the
that were
170
circumstances
of
later owners.
have been
setting within the Romano-British landscape, both regionally and for all
of occupied Britain.
and
drains
have
specific
functions
in
modern
municipal
for health
reasons
and possible
subsequent
use.
are
Stormwater
This
either
and
usually had a vaulted roof; occasionally the channel was covered with
171
stone slabs,
arched design and the whole structure would be large enough to allow
access for labourers to clean or repair them. Regular manholes were
provided for entry into them and also for the escape of gasses that can
develop in closed sewers, and to provide air and light for workmen.
KINGS _
SQUARE
z& w i n e g a t f
P A T R IC K P O O L
ST. SAMPSONS
SQUARE
LINE
r
ROMAN WALL."['
,0
10
20
[iin'rinil" ' I 1
30
40
I- -!
50
60
I -L
70
I
80
I
90
100 Metres
I ..J
Fig. 5.12: York (Church Street) sewer system (after Whitwell 1976, 4,
Fig. 2).
12 I have not seen the sewers at York, but have been in a section of the sewer at
Colchester. This sewer was very well constructed and similar in shape to the aqueduct
of KOln which was an arched structure.
172
Some military
sites
and towns
sewerage
systems for the control of their sewage. The major sites where sewers
have been found are at Bath, Caerleon, Canterbury, Chester, Chichester,
Colchester, Dover, Lincoln, London, Wroxeter and York13.
has
been
reported
(Whitwell
1976)14
(Fig.5.12).
The
primary
function of the sewer seems "was to drain the various services within a
legionary bath building",
with dating material, but the excavator reports that it was probably
constructed during the 2nd century while York was still a legionary
fortress. Because the rescue excavation was limited, it was not possible
to relate the sewer with other buildings, except to show that the branch
sewers pointing to the west was in the direction of the baths. The floor
of the
sewer was
sandstone slabs
laid closely
together with joints sealed with mortar and set in clay. The channel was
covered with Millstone Grit slabs, presumably for their strength. The
walls of the sewer generally consisted of limestone ashlars also set in
clay, to render the system water proof.
large
Fig.5.13.
vaulted
'drain'
Zienkiewicz
at
the
calls
the
of
Caerleon
is
frigidarium
drainage
structure
fortress
shown
in
a
'drain', which was in fact the Roman vaulted sewer. Both the Caerleon
and York sewers were constructed with massive stone blocks,
but the
Caerleon sewer was arched in the standard military fashion, whereas the
sewer at York had a roof of flat slabs.
13 References are given in order of site in text: Brit. 20,312: Hanson 1970,
Brit. 10, 335: 3, 313: 3, 350; 10, 332; 11, 398: 4, 304; 9, 451: 23, 306-7:
1992, F ig.10; 1995, 138: 8, 396; Kelsey, 1840, 138: 9, 438; 10, 297: 4, 281.
14 A Y 3/1,
15
1976,
5-13,
Figs.2 & 3.
173
14/1,
1976,
1.
181-3:
Wacher
Fig. 5.13:
According to John Wacher, Lincoln "is known to have possessed one of the
best developed sewerage systems in any town in (Roman) Britain",
(1995,
138). Figure 5.14 shows the sewer lines and drains of Lincoln (Wacher
1992, Fig. 10).
The discovery of the sewers in Lincoln below the Bailgate was reported
during the the 1850s16, but they have since been covered as a result of
development in the area. The sewers varied in width from 0.71m to 1.22m
16 Arch. J.
1970, 54-5.
19,
1862,
169;
40,
1883,
319;
103,
174
1946,
317;
Hanson
and were from 1.37m to 1.52m high. A series of cross sewers and small
house drains discharged into them.
Bellairs reported on one of the "main sewers" of Roman Leicester, which
ran from the Jewry Wall baths towards the north-west where it discharged
into the river Soar.17 The
1793,
when
sections
of
sewer was
the
sewer
first
was
reported by Throsby
still
in
place.
in
Stukeley
who
installed a new sewer system during the mid 19th century, when the Roman
structures
were
discovered
(Kelsey
1840;
Hanson
1970,
29,
32).
At
N
: BATHS
' '
I L_
ii
1r
SHOPS
S rtO P S
SHOPS
Fig. 5.14:
17 Bellairs 1899,
40-44,
175
3.3. Drains.
One of the prime purposes of specially constructed drains
towns,
channels
buildings
or
in forts,
stormwater run-off.
timber
lined
to dispose of
waste
from
running
along
from
the
streets
and
latrines and
drainage systems
materials
gullies
in forts,
stables
also to remove
where
cavalry
excess surface
water from aqueduct supplies and waste materials of outflows from baths
and
houses,
and
were
sometimes
discharged
into
conduit-constructed
sewers, or often into ditches. They were constructed as both open and
covered channels, particularly along the outside of buildings and along
streets. Inside buildings they were invariably covered with stone cover
slabs.
sides had stone channel drains 0.81m wide and 0.81m deep. It would seem
that water was collected from roofs and discharged into these drains
which were then channelled into tanks. Along both sides of the officers'
temporary compound were drains, and a stone-lined drain is shown leading
from the latrine. From the bath-building an internal drain leads into an
outfall timber drain towards the River Tay. Drains around the hospital
also converge into a main drain (Pitts and St Joseph 1985, 191-4). The
drainage
at
Housesteads
was
also
designed
to
collect
water
from
25,
1904,
211).
A number
of the forts on the Antonine wall from both the Antonine periods (c.AD
142-55, c.AD 158-63), had standard type drainage facilities, similar to
those at the slightly earlier fort of Housesteads (AD 124-6), probably
indicating
construction
that
by
the
mid
technical methods
2nd
century
had become
most
military
standardized.
masonry
Figure
5.15
176
picture.
Fig.
5.15:
lead pipe
seem
instance
at
to
have
had
poorly
Caistor-by-Norwich
developed
and
surface
Silchester,
drainage,
where
the
as
for
drainage
177
with particular
(Merrifield
of a problem of disposing
1965,
of
82) .
excess
baths,
latrines,
and
for
control
of
surface
run-off
of
used cesspits for disposal of waste materials, whereas during the later
phases of villa development they generally incorporated drains for the
removal of waste products.
An example of Hingley's
second phase of
which would
indicate the direction of fall of the land. If this is the case then the
feature shown at the top of the page may well not be a drain, but could
have been the water supply line towards the older bath-suite, collecting
water from springs above the site. Lysons reported that the room (number
178
1 on the plan) adjacent to the baths was at a lower level than the rest
and had a tank in its centre, which was fed by a pipe, indicating that
the villa probably had a running water supply (Antiq. J. 19, 1939, 194).
However, a detailed study of the contours and the juxtaposition between
the villa site and the springs will have to be done to confirm this
suggestion.
drain
is
shown
to
come
from
the
latrine
above
the
north-east range, both of which may have been small sewers. This may
suggest that the two wings were originally independent and used by two
families, hence the duplication of drains. All the drains were stonelined. Another example is the villa at Rockbourne (Fig.5.9), for which
the plan shows evidence of extensive drainage round the site, both for
draining the three bath-suites and for controlling run-off
from the
buildings.
The evidence of such developed drainage systems at towns and villas
seems to indicate an awareness of health considerations, which was a new
concept to the indigenous people of Britain, particularly during the 2nd
and 3rd centuries when considerable building development took place in
the province. However, archaeological evidence seems to indicate that
after a while,
indifference
from inhabitants,
deteriorated.
Hanson
(1970)
drainage was
at many
forCaistor-by-Norwich.
that where wells
of
the
sites
or
drainage
sources
and
at
shallow levels, seepage into the permeable gravel and sand layers could
easily contaminate well-water, such as at Caerwent (p.133), Caistor-byNorwich
(p.133),
(pp.l30ff).
179
over
long
period
1990,
81-91).
The
excavation report for Corbridge (Bishop & Dore 1988), a former fort and
later a small town, says nothing about the extensive drainage system of
the various phases of the forts and later town. However Fig.3 of the
report shows a number of features which indicate drains, including those
on both sides of the Stanegate and in West Dere Street. It is difficult
to understand why there is this lack of information on the drainage of
the smaller urban areas.
Military sites and early coloniae seem to have had the best constructed
drains, mainly because they were constructed to fixed standards. They
were
generally
constructed
at
constructed.
incorporated
the
same
during
time
as
the
the
planning
fortress
stage
or
and
the
were
fort
was
spring water
to
remove and the most suitable drain type that could serve the purpose was
used. At sites where there were poor drainage systems there was the
danger of pollution of wells from cesspits such as at Caerwent where
there were no drains for private householders (Chapter 4, 128), Newstead
(Chapter 4, 138), and
at non-military
structures
and
reconstruction
sites,
drains
became
necessary
for
simple
often dictated by
new
or
the
altered
structures.
5. CONCLUSION
The importance of Roman baths as an institution in Roman Britain is
demonstrated by the speed with which they spread as urban and rural
development took place soon after the conquest. Every site type had
baths, classified into linear, block and complex layouts. The military
baths were mostly of the first two types, but civilian baths varied in
style, particularly those constructed after the mid 2nd century. However
not enough is known about these aspects of baths and whether there were
any
regional
differences.
It would
seem
that
baths
are
very
much
180
The army built baths at many military sites and made provision for
efficient drainage, and disposal of sewage and stormwater. This was not
always the case at non-military sites. The Britons were quick to adopt
the style of housing and institutions,
baths were built in almost all the towns and many of the houses at
villas had baths, even at the smaller cottage type houses,
indicating
that
However
they
were
the number
of public
popular with
the
baths
public.
found
Drainage
seems
to
was
an
A particularly useful
investigation
would be the layouts and status of baths at all site types for Roman
Britain.
181
CHAPTER 6.
THE DISTRIBUTION OF WATER-RELATED FEATURES IN ROMAN
BRITAIN
1.
GENERAL
I have traced 815 sites in the literature, but eight sites1 on closer
examination of
the available
no water-
related features have been proven at those sites. The final number of
sites in the database with water-related features is therefore 807.
Table 6.12 gives a breakdown of the number of sites with different
types
total
of
structures.
sample
of
The percentages
sites
listed
in
given
the
are
calculated
database.
They
on the
are
not
factors
(incomplete excavation,
poor preservation,
Table
6.1:
Distribution
of
number
of
sites
with
water-related
features
Aqueducts
Baths
Wells
Drains
134
488
350
179
84
47
102
Sewers Total
sites
listed
20
807
16.7%
60.5%
43.5%
22.2%
10.4%
5.8%
12.6%
2.5%
Pipes
Springs Tanks
relatively
recorded
at
small
13 4
for
sites
each
but
seem
archaeological
record.
Wells
supply
sites)
for
(3 50
category
to
be
of
site.
Aqueducts
under-represented
were
the most
the
Roman
common
period
form
but
were
in
the
of
water
are
also
undetected.
Even
though
springs
have
been
included
in
the
only 47 sites
1 The
eight
sites are:
Ardoch,
Limestone
Corner,
Neath
and Whickham
Coldharbour (settlement); Brantingham, Frampton and Kingscote (villas).
2 The information
special tables.
for
Table
6.2
has
been
182
extracted
from
Table
6.1
and
(forts);
the
other
aqueducts
spring
have
been
sources.
traced
Also,
along
springs
their
were
complete
usually
routes
outside
the
to
their
areas
of
excavation and may not have been recorded, particularly at villas and
settlements,
supply.
several
such
as
at
Bath
(Ch.5,
of water
p.162),
supply within
London
(Ch.4,
(487), although the total fell far short of the number that I would
have expected. The total of 179 sites where drains are recorded is
also low, perhaps because they are not always mentioned in reports,
though in some cases I may have missed them in plans of the sites.
The number of pipes is particularly low (84) , but this probably can
be ascribed to the fact that wooden pipes have long since rotted and
the usual iron rings may have been robbed or have been completely
corroded away;
many
lead pipes
would
also have
been
robbed,
and
structures.
For
example,
the
number
of
aqueducts
Hanson.
She
reported
aqueducts
at
42
forts
(29
certain,
13
aqueduct water supplies for military and town sites and thus there
are some omissions in her work regarding rural sites.
Table
6.2 gives
related
a summary of all
structure
collectively
been
types.
the sites
The different
referred
to as
forts,
categories
though
in
the
have
text
and
minor
towns,
settlements
3 Hanson 1970,
358-74;
Table 2.1a,
p . 28.
183
but
and
villas
features
were
invariably
such as aqueducts,
baths and wells were also found at forts north of Hadrian's Wall.
This is shown for the distribution of sites with baths in Figs.6.6 to
5.10.
pp.239-253).
Forta
Coloniaa
Hunicipia
Civltataa
S-towna
Sattlananta
Villaa
Total
Aqueducts
66
11
19
24
134
Baths
85
13
30
49
305
487
Wells
43
10
52
135
104
350
Drains
52
17
30
67
179
34
84
23
47
Pipes
28
Springs
10
Tanks
32
16
38
102
Sewers
10
20
Number Sites
137
14
66
212
372
807
AQ * BA
45
10
16
AQ + W
16
10
AQ + TA
23
AQ - WP
27
AQ * DR
22
14
BA - W
26
18
10
55
3A r TA
24
30
3A * DR
31
11
57
+ TA
19
13
AQ + 3A + W
14
AQ - WP - W
15
AQ S P
3A r SP
10
combinations
have been
No water supply source has been recorded for 258 sites with baths
(53% of all sites with baths). Similarly, no baths have been recorded
at 231 of the sites (67%) where wells have been noted, while 50 sites
with aqueducts have yielded no trace of baths (37% of all siteswith
aqueducts).
These
figures
show
up
complex
problem in
the
to
factors
relating
to
coverage
of
site
by
excavation
or
reasons:
firstly,
where
there
184
are
only
small
number
of
particular site type and the same feature has been found at all the
sites,
the proportion
statistically
sample
of
normal
data
is given as
population,
represented
by
100%.
because
the
The
data
there
listed
is
sites
is not
from a
in
bias
in
the
the
database
hence the
Colonlae
Munlcipla
Civi-
tae
Small
towns
Set t l e
ments
Villas
All*
sites
134
capitals
Aqueducts
Baths
Wells
Drains
Pipes
Springs
Tanks
Sewers
Total sites
66
11
20
24
48%
100%
50%
78%
13%
9%
6%
85
13
30
49
305
62%
100%
100%
93%
45%
23%
82%
43
11
52
135
104
31%
100%
100%
78%
79%
63%
28%
52
17
30
67
38%
75%
100%
57%
24%
14%
18%
28
34
20%
75%
50%
50%
4%
4%
9%
10
23
7%
50%
100%
7%
4%
6%
32
16
38
23%
50%
43%
12%
8%
10%
10
7%
75%
50%
75%
1%
<1%
14
66
212
372
137
488
350
179
84
47
102
20
807
*Note: The column for forts include fortresses. The 'all sites' in the last column
represents all types of sites where each category of feature have been found. The
'total sites' in the last row represents the total number of each type of site in the
database.
However,
the likely relationships between the types of features and the total
number of sites on which the database sample
is based.
They also
indicate that there is a bias in how the information was obtained and
what archaeological information was available to be recorded. Also,
the sample of sites recorded in the database bears no relation to the
actual total number of archaeological sites that have been found to
185
analysis
of
the
distribution
of
the
different
water
features
follows. Each type of site will be treated separately, but there may
be some overlap in comment because of the geographical relationships
between certain types of sites. Forts are a special category of site,
because as military sites, the water-related features were intimately
bound up with their administration and would have been financed at
provincial level,
accounts for a particular fort. This was not so for towns, perhaps
other than coloniae which may also have been financed by Rome.
The Ordnance Survey map of Roman Britain, 1991, lists 23 5 Roman forts
and 75
fortlets.
legionary fortresses
(1990,
91, Map
4.24)
show 9
this
gives
total
of
c.350
military
establishments
known in Roman Britain, though they were not all active at the same
time. Only c. 40% of these forts (137) have yielded specific evidence
of water-related
features,
the
remaining
60%
of
known
sites
are
66
sites (48%) are recorded as having had aqueducts out of the total 134
sites with aqueducts. At 28 forts, water pipes of some kind have been
found, in every case at sites also having aqueducts.
Tanks
Sewers
Total
sites
10
32
10
137
7%
23%
7%
Pipes Springs
Aque
ducts
Baths
Wells
Drains
66
85
43
52
28
48%
61%
31%
38%
19%
4 A special problem arises when fortresses and coloniae are discussed, because the
four coloniae developed on the same sites as the fortresses or adjacent to them.
However I shall make it clear when it is a fortress or colonia that is being
discussed. The same problem occurs with some civitas centres, such as Exeter, which I
have entered as a fortress, but it eventually became the capital of the Dumnonii as
Isca Dumnoniorum.
186
It
is
not
always
clear
whether
these
pipes
formed
part
of
the
the
aqueduct
type was.
At
some
forts
it
is mentioned
that
evidence of a water pipe was found at one of the gates, implying that
the pipe was the aqueduct. Both Hanson
(1985b)
(water
(wooden pipelines),
Burrow-in-Lonsdale,
Dalginross,
Dalswinton,
Great
Chesters,
Tomen-Y-Mur, Trawscoed and Wetwang, and there may have been more. The
aqueducts serving Chesters most likely were leats but they may have
been stone channels or even pipelines. Because of its high elevation,
Housesteads was not served directly with an aqueduct {Arch. Ael.4 12,
1935,
from the north to near the site of the external bath, from where it
was raised by unknown means to the fort
(Birley 1961,
181) . Leat
at
Lanchester,
would
also
have
been
(1989) as
construction
projects. For these large quantities of dressed stone would have been
necessary,
Leat type
aqueducts may have been resorted to during the 2nd century for water
supplies because of the high cost of quarried stone and particularly
that of transport. Even so, the comparatively high number of forts
with aqueducts indicates significant military investment in a running
water supply.
Figure 6.1 shows the distribution of the 66 forts with aqueducts as
listed in the database. These sites seem to cluster in three regions.
There are 12 widely spaced sites south of the central Wales-
187
Birmingham axis. The remainder all lie from north Wales to as far as
the Antonine wall with a concentration of sites round Hadrian's Wall.
The
chronological
distribution
of
forts
and
fortresses
changed
considerably from the pre-Flavian (AD 43-68) through the Flavian (AD
70-96) period to the third century (Jones and Mattingly 1990, 89, 98,
100, 132), but the distribution given in Fig.6.1 shows all the sites
with recorded remains of aqueducts for the whole of the occupation
period. They appear to be evenly spread over Roman Britain with the
largest
concentration of
sites
in the middle
north,
the military
zone.
(c.AD 79-85)
(c. AD 138-61),
region around Hadrian's Wall (c.AD 117-38), and a southern region south
of
Hadrian's
Wall.
For
this
latter
region
baths
would
have
been
constructed over the period from the later 1st century to the early 4th
century, depending very much on when forts and fortresses were first
commissioned. Some sites received additional baths at later periods, as
for instance a bath was built in the early 4th century in a barrack at
Housesteads.
I have recorded wells at only 9 of the possible 32 fortresses5 which
were
active
at
various
times
(of which
4 became
coloniae
later).
been
destroyed,
or
I may
have
missed
them
in
the
reported
188
listed in the database including the 4 coloniae. Five are in the north,
6 are associated with forts along Hadrian's Wall and 2 along the west
Cumbrian coast. Seven were in Wales and the remaining 15 are in England
south of Hadrian's wall. I only traced 2 fortlets with wells. Why so few
wells are recorded in the literature for military sites is difficult to
understand, because we are considering wells within the fort walls.
2.1. Dating of water-related features at military sites.
This raises the issue of when did forts acquire their aqueducts? An
aqueduct at a fortress/fort suggests that they may have originally
been planned as permanent bases,
as would well-built
defences and
wall (c. AD 142-55) would have had their aqueducts constructed during
this phase of occupation. With long occupied forts precise dating is
much more difficult and a further complication is that some forts
went through several phases of occupation and abandonment,
and the
From
the
evidence
at
Inchtuthil
where
construction
approach
to
the
building
'reserved plots'
where
of
such
the baths
stone
structures
at
in stone and
189
such as at
It would appear,
for
there
is
some uncertainty
about
it.
The
legion
stationed
at
indicates
the
start
of
their
infilling
and
their
likely
abandonment.
Many forts from the more southern region would have been established
early
as
the
occupation
progressed
northwards
and
provision
for
(1995,
regiment
including
water
supply,
and
RIB
605
found
at
190
(Salway 1993,
2.2.
for water-related
structures
at military
s ite s .
43
(12.6%)
(341) 66
Of the total
for. In other words, for 72.7% of the number of known military sites
no water supply has been recorded. Clearly this shows that for waterrelated features there is a distinct bias in the record, which could
be due to any of the following reasons.
1.
The historical
background
important element.
of
excavation
at
sites
would
be
an
In many
instances
remains
of water-
also
no
evidence
form of
supply
has
been
identified.
5. The excavation within a site was constrained to limited areas so
that remains still extant have not yet been found.
6. Many of the sites have not been excavated at all, and are known by
means other than excavation
(such as air-photography)
and water-
supply has been found, there is an expectation that the others would
have had either a running water supply or at least a well because
water is such an important commodity.
but
produced
normally
problems
with
the
various
facilities
it would have
requiring
a large
fort
(2.1ha,
running water supply nor any internal wells have been found,
two external wells have been recorded.
191
no
though
found (one each near the North and South Gates, one at the north-east
angle
and
large
one
at
the
south-east
angle,
and
one
in
the
(Arch.
Ael .3 25, 1904, 248-9). Inside the commander's house was a small bath
for his personal use, and a bath was built during the 4th century in
one of the barracks. An external bath-house existed near the Knag
Burn west of the fort, which must have been used by both the 1,000
soldiers and the inhabitants of the vicus. In addition to the two
wells it had been suggested that a spring, which appears to have been
converted into a well
379,180m,
(Fig. 6.16:
1904, 253-4) .
LONG 2 19' W
570000m
Boat, House
Sewingshiel
Cr*<r &
f
r'
I,"t u r r e t
Broom ee Lough
3S8T-:
bo., ,,'te0,, *
'
.^Spr
V
...
Ridley Common
Sheepfold4
to
f n c lo # u r * j
-Ji
King s W icket^'r^r
J3^\3
HlLECAST'LE
. 36
>J
^ iS ^ ^ C Ie w X king's
Kennel
Crags.^
TU R R E T
L A T
55 01'
36A
(site of)
v
o r1
: Cuddy's
ETUiA Cri.?j
VER C O yiC lU M
RO M AN
>'
ROMAN
\&EL
\Housesteads
\
_
Museum'
.
<'
>%>^>ROMAr^v^
A yS -W E L L
Informat
Centre
C u m u iu *
rriounj
192
Bousanquet suggested that the Romans chose the site because of the
ready availability
of water
(Arch.
Ael.3 25,
1904,
207)
and
Birley suggested that water was pumped up from the Knag Burn
Eric
(1961,
181). Manning suggested from finds of iron-pipe collars and the fact
that the fort had a range of water related structures (a fountain, an
internal bath for a period and an external bath-building)
fort
must
24.151).
have
been
However,
served
by
an
aqueduct
(1976,
that the
40.151,
found.
Fig.
The
piping may have been water distribution pipes from one or more of the
5 tanks found in the fort.
Raymond Selkirk
Lough (A, Fig.6.16), about 7.5km due north of the fort (Selkirk 1995,
2-3).
(In about
level 255m above mean sea level) . Selkirk told me that he had found
air photographic evidence of the route of the aqueduct running in a
south-west direction from the Broomlee Lough until
it reaches
the
ridge north of the Knag Burn, then turned south-east towards the gap
where the Burn passes below the wall east of the fort (OS level about
245m). Based on this evidence he and his associates "calculated the
spot which would entail the shortest possible bore" When they arrived
at this position they found the "huge cutting through the hill....Its
purpose was obvious - the Romans had tapped Broomlee Lough for their
water supply to the Knag Burn" (p.2).
He does not give the depth or the length of the cutting. From the OS
map
the contours
seem to
indicate that
of
the
(1:25,000)
length of the suggested cutting for the aqueduct that would have had
to be made through the high ground between the lake and the Knag
Burn.
He also describes an intricate water-wheel and pump system, which he
suggests could have been used to raise water from the level of the
Knag Burn to the fort c.15-20m above. According to Selkirk the waterwheel would have been rotated by water
activating
flowing
The pump
drew
water from a sump, which he suggests was the supposed Roman well and
it pumped water some 20m up to the fort. As discussed in Chapter 3
193
proposes
would
need
to
an aqueduct
would
be
investigated
more
thoroughly,
to
been within
the
capabilities
of
the
interest
inspection,
to
of
have
the
Selkirk's
opinion
suggested
of
an
archaeologist,
aqueduct
route.
after
have
not
of
water-related
difficult to assess,
features
within
forts
has
been
dealt with in reports. At the fort of the Lunt there was no external
source of water supply. Six wells have been reported, and water may
also have been drawn from the river.
tanks
barracks,
shows
that
7 were
located within
tanks were distributed round the site (Hobley 1973, 35-8, Plan). The
tanks outside the buildings were either filled by water carriers, or
from rooftops, evidence for which was found of a gutter-fed tank from
the
north-west
granary
roof
with
an
overflow
into
another
tank.
Possibly some of the tanks within the barracks could also have had
catchment arrangements, but this is not mentioned in the report. One
of the wells was
and
another well was located near the north-west angle of the fort. The
fort was active from c. AD 60 to AD 74, and was demolished in AD 75
(Hobley 1973, 15), a short time for such an elaborate establishment,
indicates the pragmatic approach of the Roman army to its military
needs.
The military
camp
of
Lyne
is
reported
to have
had
an
aqueduct,
probably a pipeline from springs about 1.5km north of the camp, which
7 If these pumps were indeed used to remove water from Roman ships, the height that
water would have had to be lifted would not have been more than a couple of metres. J
P Oleson (1984), in his book on remains of 'Greek and Roman Mechanical Water-Lifting
Devices', does not give a single example of the type of "Roman pump" Selkirk refers to
and shows in his article. It seems that Roman pumps had to stand in the water source
to draw water into the cylinders by the action of the pistons. There has been no
record of a Roman pump with a suction pipe attached to it to draw up water into a
cylinder from a source as shown by Selkirk.
194
but
later excavation
present
understanding
of
the
water
related
problems
of
Housesteads are, the fort functioned for a long time, so that a water
supply system existed that worked, however awkward it may have been
at times for the 1,000 soldiers. The remains at the fort of the Lunt
indicates that it functioned with a water supply system of 14 tanks
and 6 wells. In the case of the fort at Lyne there seems to have been
an
over-supply
of
contradictions
problems
in
facing
water
demand
and
no
bath
versus
archaeological
supply
has
of
been
found.
water,
investigations
These
are
indicating
the
relating
to
water-
related features.
Verulamium.
The
large
thermae
bath
at
Huggin
Hill
(site
14,
Fig.6.19), built during the Flavian period (AD 69-96), was demolished in
the 3rd century because it became flooded, and the bath at Cheapside
(12), probably dating from the late 1st century, went out of use during
the 3rd century, which may have been related to its water supply. The
bath at Billingsgate (39) was constructed before the 3rd century and was
still in use in the late 4th century, while the bath at Pudding Lane
(38)
(Rowsome 1995,
418,
420).
still
8 Exeter is listed as a fortress in the database, but later it became the civitas capital
of the Dumnonii.
195
fortress baths, and the other one was probably public baths built after
the town obtained civitas status. Wroxeter had a large second bath built
(A, Fig.6.17) during the Hadrianic period and expanded during the late
2nd or early 3rd century (Webster 1965). Atkinson reported an earlier
bath (B) which was built soon after AD 60, but was abandoned during the
AD 90s, and was demolished during c. AD 120 to make room for the Forum
(1942, 25-54) .
Wroxeter
Fig. 6.17: Roman Wroxeter showing baths (A, Hadrianic, and B, from AD
60) and wells (circles), and aqueduct C. (Barker 1990, Fig. 3).
At Canterbury and Dorchester remains of two baths were
found at each
at
Trajanic or Hadrianic
in
of
196
eitherFlavian-
325-6). Silchester also had two baths, the public baths in the south
east of the town (insula xxxiii) , and the mansio with its baths, located
in the south near the south gate (Archaeologia 59(2), 1905, 133-4).
3.1.
Coloniae.
modern
development
overlies
the
sites,
because
and
this
particularly applies to the four coloniae. The water supply for each
of the four coloniae also presents special problems because for three
of the towns
(Colchester,
only inferred from evidence of piping near the entrance gates to the
towns or from distribution systems within them. For Lincoln (Chapter
3,
section
11.2)
there
is
the
problem
about
the
source
of
the
earthenware aqueduct from the north. Little is known about the second
aqueduct, of which minor remains have been found east of the lower
town. A discussion of the water supply of the coloniae follows.
6.18
(Crummy
1984,
Fig.14)
shows
that
the
colonia
of
Colchester lies between the 8m contour along its northern wall and
rises above the 3 0m contour to the highest point at the Balkerne
Gate. A large part of the town lies above the 23m contour (hatched on
the plan), with the highest point at about 30-32m near the Balkerne
Gate. For the Romans to have been able to provide a running water
supply to the town, water must have been brought to at least a higher
elevation than the 30m level.
consisting of wooden pipes, held together with iron bands, have been
found at the Balkerne Gate
direction outside
the
town
(1984,
they
is not
clear.
uphill, and this would imply some system that could force water up
the pipes,
of which no evidence
has been
found.
Intheexcavation
197
91;
1992, Fig.13, 109). This seems to suggest that when the fortress was
replaced by a colonia (c.AD 50-55), the colonia developed an internal
water distribution system,
effective
running water
supply.
reached
the
Balkerne Gate site has not been resolved, however some possibilities
are discussed below.
About
450m to
the west
the
strong
springs at Chiswell meadow, which were the source of water for later
Colchester.
\ \,
25
40'
50'
'S prings*
Jr
25"
sp rin g
100
c o u r s e d >707- < 7 3 7 svsrem
-100
(after
Crummy suggests it could also have been used by the Romans to bring
water by means
of an aqueduct
contour
From there
level.
into a tank
placed on top of a tower near the Balkerne Gate, the highest point in
the town, from which water would have been distributed to the lower
parts. However, here there seems to have been the same problem that
existed at Lincoln, how water was raised to a tank about 15m above
the delivery point of the aqueduct. This tank would have had to be
198
higher than the tank at the Balkerne Gate. No evidence has been found
of either the means of raising the water or of raised tanks.
Crummy
discusses
the water
supply
of
Colchester
in
some
detail,
he suggests that
(1997),
he gives
an
illustration of
the town as
it was
likely to have appeared in the 2nd century. Below the Balkerne Gate
is
shown
conjectured
distribution tank
elevated
aqueduct
delivering
water
to
This
a
is
based on his discussion of the water supply for the town in Report 3
(1984, 27-7). A second suggestion is, that the water supply may have
been from wells. Nine Roman wells are shown on his plan located from
north of Sheepen to the three wells at Middleborough north of the
north-west corner of the colonia wall. No Roman wells have been found
within the walls, though a number of Medieval ones are shown on the
map.
This
because
it is unlikely
where
(40ft)
is shown to
(Crummy 1984,
that the Romans brought water to the north-west corner of the early
fortress
(c.AD 43-60/1).
the 15m contour, one about 4m higher near the north wall, and another
at the mithraeum at about the 21m level. How much water would have
been
available
from
these
springs
is
questionable,
because
the
199
6 water mains
the 3 0-32m
level,
2) water mains
3) a leat from the Sheepen springs seems to run towards the fortress,
4) an aqueduct
level,
5) 9 Roman wells outside the colonia,
6) at least 5 springs within the colonia.
If
item
was
possibilities:
part
firstly,
of
an
aqueduct
system,
there
are
two
trestle
support
for
conduit,
or
as
stone
or
earth
again there is no evidence for pipes found further west outside the
town. The area is extensively cultivated, so that any evidence of a
wooden pipeline would have been destroyed. Also, wooden pipes used in
an inverted siphon would have presented problems of bursting under
pressure, unless they were encased in concrete, or well protected in
the ground.
The
evidence
of
water
mains
elsewhere
in
the
colonia
is
very
suggestive that a running water supply did exist, but from where it
was distributed is unclear.
in
several
parts
of
the
town.
It
seems
to
me
there
is
sufficient evidence to accept the theory that the colonia did have an
aqueduct water
point,
However,
supply,
even
for distribution
much more study
to
lower parts
of
the
town.
200
within the colonia. There is also a record of two wells, which hardly
seems to represent the number that should have been available for a
large town.
Hanson
suggests
could
possibly have been the springs near Matson on Robinswood Hill about
3km south of the town (1970, 367). Wacher adds the suggestion that an
aqueduct supply could also have come from north of the city
(Wacher
1995, 159) . It has been reported that a large "9m-wide tank has been
located in the fortress" and he suggests that this may have been the
terminal reservoir for an aqueduct
(Stephens 1985b,
223;
JRS 1967,
195). Other tanks have also been found. A bath-house is implied from
remains of an apsidal hypocaust at Westgate Street, but the limits of
excavation on the site prevented more of the structure being revealed
(Wacher 1995, 158) . The colonia had a sewer, but it is not clear how
extensive
running
it
was,
water
archaeological
confirm a
and
this
supply.
would
It
evidence
probably
point
seem
that
not
from
outside
the
would
is available
running water
also
supply,
but,
as
towards
sufficient
colonia
for Colchester,
to
internal
The
problems
of
the
discussed in detail
aqueduct
water
supply
Lincoln
at
have
been
A summary of the
found,
The
aqueduct
Roaring
Meg
was
stream
rising-main
towards
the
of
about
colonia,
2km
with
length
a
from
difference
the
in
found.
4. Early antiquarians of the 18th and 19th centuries reported having
seen a tower with a tank near the Roaring Meg stream.
5. In 1786, a Swiss artist, Grimm, drew two sketches (now in British
Museum) of a bridge structure that coincides with the position of
the foundation bases; they show the line of the aqueduct pipe on
the structure (Thompson 1954, PI.VII A, B).
6.
At
base VIII
reports
nearest
finding
in
the
the
Roaring Meg
filling
of
the
(E,
Fig.3.7a),
construction
Thompson
trench
'a
(1954,
201
of
the
lower
part
of
colonia
above
the
spring
line
(F)
19th century in the upper part of the town, the main sewer running
along the Bailgate.
The evidence of items 2 to 5 seems to indicate that a) the source may
have been at the Roaring Meg; b) that there was some evidence of a
substructure carrying the concrete encased pipe to near the Roaring
Meg stream,
and c)
of
an
inverted
siphon
(the probable
reason
for
concrete
encased pipe) . Two main problems, as yet unsolved, are firstly, was
the Roaring Meg stream the water source for the postulated inverted
siphon, and if so, how was the system supplied with water? Thompson
(1954)
discussed these
on the
Suggestions have
been made that the water source was further north (Chapter 3, section
11.2, Fig.3.8).
Limited further investigation was carried out during the early 1980s
in the vicinity of the masonry piers, but nothing new was discovered
(personal communication from Mick Jones, May 1997). The public bath
(A)
(Fig.6.19)
It has been
suggested by Lewis that water may not have flowed in the aqueduct
(D), based on the fact that no encrustation shows in the ceramic pipe
(Chapter 3, p.86), which would have been expected from the lime-rich
water coming from this limestone region
(Lewis 1984,
68-9).
It is
uncertain whether the second earthenware pipe (n.7 above) was in fact
an aqueduct supplying water to the lower part of the colonia.
202
jj
W a te r T a n k
c\ 0
Cemetery
N E W PO R T ARCH
1
.
^ A S IL IC A I
s
S h ops
S h o p s ii
W EST
L E G IO N A R Y
fortress
i i
iI
i
A 1I
T o w e rs Isite of hotell
ii
FO RUM
east
AND
UP PE R
G ATE
TOW N
C a th e d ra l f
LO W ER - i
TO W N
'" V
Terracing
A q u e d u c t p ip e-lin e 1 8 5 7
11 9)i : I i
n i ji n
11 S i I I j
;J/* ,,
N '
GATE?
nr I
C 3tnedrai S treet
Cem etery
C e m e te ry
urhs .sno
p - - - |
R o m a n S tre e ts
1000
M e tre s
203
134,
Fig.
57
From its position it seems unlikely that it could have supplied water
to the northern part of the lower town, which may have obtained its
water from the upper colonia aqueduct and that in turn would require
an effectively functioning upper aqueduct.
There are other indicators which demanded a plentiful running water
supply such as evidence of an elaborate sewer system (G) in the upper
town and possibly a water distribution system. However, as discussed
in Chapter 3 (pp.82-6), calculations, based on delivering water from
the Roaring Meg stream (OD level 39.5m) to a tank near Newport Arch
(OD level 65.5m), show that only about 72,000 litres of water per day
was likely to have been supplied by this aqueduct. The distribution
of this
a fountain,
sewers,
seems
periods when the water was available to the various consumers, public
and private"
(1970,
system,
it
likely
but
is
that
the
Romans
had
developed
water
regulating systems.
If the observations by the early antiquarians and Grimm's drawings
are acceptable as evidence for a tower structure near the Roaring Meg
and that the aqueduct's intake was in the vicinity of this tower, as
suggested by Thompson (1954), then it implies that the Romans created
an elaborate water raising facility in the vicinity of foundation
VIII. These early sources do not describe the tower or the material
of
its
construction.
In my
calculations
(Chapter
3,
pp.82-6),
ridge limestone to reach the water table; this complication may have
been the reason for building an aqueduct.
9 Even 4 systems would have delivered only a little more than one quarter of a million
litres per day, which is still many times less than what the aqueduct at Dorchester
delivered per day.
204
have had a working running water supply, because the town was still
economically active up to the mid-fourth century (Wacher 1995, 149) .
Further investigation is necessary to resolve uncertainties about the
water supply of Roman Lincoln,
early evidence.
York
is
built
formations,
on
gravel
and
sand
layer
which
overlies
clay
is reported to have had several wells and springs (RCHM(E) 1962, 59).
Well-water could therefore have been a significant source of water
for the Roman town. Only two wells have been reported.
An aqueduct for York is only inferred from evidence of other remains
such as a fountain, a bath (of which only a small part has been found
because of the limited area of excavation) and lead pipes leading to
it, and a massive lead pipe crossing the River Ouse
(RCHM(E)
1962,
49-51; Wacher 1995, 176, Fig.79). The fountain was in the shape of a
lm square tank fed with water from a pipe in a vertical slab at its
back, with an outlet pipe
(75mm diameter)
This
bath,
said
to
have
had
the
largest
caldarium
in
Britain
3.2. Municipia.
Roman London,
developed around the Walbrook valley where springs and wells were the
principal sources of water supply
come
from
Highgate
or
Heampstead
(1978,
104-8),
because
he
it unlikely that
water supply of fresh water, but there is no need to assume that they
205
were aqueduct fed" (1984, 241). This would imply that the pipes could
have collected water from either springs or wells,
refers to as
"intermural aqueducts"
in what Wilmott
(1968,
97)
believed the wooden pipes to have been used for drainage, which seems
to have been a costly way of providing drainage (Wacher 1995, 101) .
Because of a rising water table, a major problem was the disposal of
excess water, not collection. Both Wilmott (1984, 241) and Merrifield
(1965,
146)
argue that the evidence for the well and spring water
supply for all London's needs outweighs the evidence for an aqueduct
system. Wilmott
bath-houses
(Cheapside
Billingsgate
39)
were
12,
Huggin
close
to
Hill
14,
Pudding
spring-line,
which
Lane
would
38,
have
collectors.
This
leaves
the
question
of
an
external
closely
related
to
the
development
of
the
area
around
the
Walbrook valley.
Wilmott reports 51 wells found within the walled area of Roman London
(1982a,
1982b,
1984,
1991),
shown on Fig.6.20
(Wilmott 1982b,
235,
Fig.l). Except for three reported wells to the east of the forum, all
the wells were found along a central band mostly west of the Walbrook
stream, with concentrations in the Queen Street area
26)
(1982a,
20 wells)
Walbrook area
Compton Court
(Wilmott
1991,
(sites 25 and
latter area at
Bucklersbury and
20-52,
Fig.7
&
22)10. Later
another
wells
were
16 single-well
10 W i l m o t t 's 1991 paper is to some extent a summary of the three CBA Research Reports
N o s . 69, 70 and 88 on the Archaeology of Roman London in the Walbrook valley area.
206
Fig. 6 .20: The wells and other water-related features of Roman London
(after Wilmott 1982, 235, Fig. 1).
207
Som an
te a iu 'e s
Q ueen
V ic to r ia
jn o a ie o
S tre e t
Fig. 6.21a and b: Localities of Roman wells in the Queen Street area,
London, (after Wilmott 1982a, Figs 5, 6).
Thirteen of the wells dated from the lst-2nd centuries and 9 to 3rd-4th
centuries
less
(Wilmott 1982b,
secure
dating.
Whether
additional requirements,
the
later
wells
were
new
because
of
ones that may have become disused or polluted needs further study. A
well is shown within the supposed Governor's palace area (Wilmott 1982b,
235, Fig. 1 Site 33), the only building shown to be associated with a
well. Near the Cheapside bath a well is shown, but it is not clear
208
whether it was directly associated with the bath, because the bath
seems to have its own watertank supply
(Fig. 4.8).
In addition there
were other wells/pits which have not been dated and with uncertain
origin, so they are not included here. The information about most of the
wells was published in a gazetteer compiled by Merrifield
(1965), and
(Chapter 4,
the
considerable
evidence
for
internal
distribution
of
water,
which implies a running water supply for the Roman town, but he says
that an aqueduct has not yet been identified (p. 19). The municipium
was the third largest town of Britain during the Roman period, but
little seems
to be known about
(Frere 1964,
favour wells
as an easy way
Silchester.
Frere mentions
that
as at London or
of
wells, e.g. Insulae IV, XIV, XXVIII (p.20), but only describes a well
in Building 3 in Insula XXVIII
iron collars
Street,
209
presumed to have been a water main (Frere 1962, 154). Further along
in Insula XVII a stone conduit was discovered by Richardson
(1938,
85-6), who suggested it may have been the water supply that came from
higher up the River Ver that flows along the north-east boundary of
the town (Hanson 1970, 70). Whether there is a connection between the
conduit
this
latter one
is difficult
to
say,
Near
the
centre
of
the
town evidence
of
a number
of wooden
pipelines have been found, which implies a running water supply for
the town.
century,
Frere comments,
skill
and
piped
supply
from the
(Frere
1960,
20-1).
This probably
to
install
available"
internal
distribution
urban water
aqueduct
pipeline,
may
have
castellum
indicating
that
"we find
brought
water to
divisorum was
represents
continuation
town
has
new
of
the
for any
not
been
important,
264,
Fig. 117,
the
The supply of
being an aqueduct
is perhaps
evidence
of
an
aqueduct
and
also
had
an
elaborate
water
Some of the
town sites may have had aqueducts during their certain or suspected
military phases
of
the
Iceni, Caistor-by-Norwich.
210
On
the
other
hand,
remains of wells have not yet have been found because of insufficient
detailed excavation.
Canterbury,
Dorchester,
have
At
been
recorded but the archaeological indications do not show that they had
many wells,
but
the
Wroxeter,
Atkinson
and
excavations of 1923-7
the
town had
an
reports
further
excavations at
wells
during
the
aqueduct
water
supply,
for
number
of
private
There seems
supplies,
the
centres
as
systems. Silchester
revealed
(40ha)
by
their water
supply
and
drainage
dispose
Silchester
Caerwent
Fig.11.4),
of
their
had
was
own
large
more
waste
public
modest
water
bath,
(Wacher
and
sewage
whereas
1995,
the
45,
into
cesspits.
public
Fig.10.2
bath
and
at
47,
have been
which
seems to imply that some other private houses may have had baths, but
have not been
identified as
Caerwent was
at Caerwent,in particular
the availability
seem to have
of army
developed
a period well
it has
into
town
the
hampered
streets
late
excavation
second
of
12 Millett is critical about the suggestion that military engineers were involved
the construction of civitas capitals (1984; 1990, 69-75; Blagg 1980, 1984).
211
the
in
century
(Brewer
distribution
1990,
system
75),
also
so
that
expanded
it must
with
be
the
assumed
street
that
the
development
(Britannia 16, 1985, 201-2). Whether the aqueduct supply was modified
as the town developed is not recorded.
4. SMALL TOWNS
The 66 small towns listed in the database with water-related features
represent 68% of the total number of 97 suggested by Burnham
187) . Of the 9 which had aqueducts,
(1986,
and
Ewell
6,
Ancaster
and
Heybridge
each.
Wells
were
the
(Black 1995,
of
of
excavation
information
small
towns
in Britain
it
is
quite
of
site
types
related
features
Hingley
states
have
been
75).
Below I give
some
reasons why
I consider
some
sites
135
minor sites,
13 Black
towns.
refers
(1995)
to vici,
many
of which are
212
classed by Burnham
(1986)
as
small
were
recorded,
of
which
8 were
leats.
In most
cases
these
were
too
far
distance
from
the
settlement.
Only
sites
with
aqueducts also had baths, whereas 10 sites had baths and wells. The
implications of these combinations are discussed below.
An important problem
having
been
responsible
for
their
planning
and
or
rural
necessarily
be
communities
an
indicator
Romano-British period.
size
(the
standard
immediately
we
are
of
a special
at,
wells
would
not
organization
during
the
suggests
looking
that
bath-suite)
there must
have
and/or
been
some
an
aqueduct
organization
have
invested
in
such
expensive
structures.
Of
the
212
of
these,
(Table 6.5)
had
both
small
baths
in private
Estate,
and 49 with
Ivy
Chimneys,
but
it
is
not
always
clear
whether the larger baths were public baths or private ones. Table 6.5
gives a summary of the aqueducts and baths at settlements.
20
9%
49
23%
2%
69
32%
16
8%
45
21%
212
213
seems not
so that
it is not
these
so-called settlement
sites with
inhabitants
apparenly
did
not
have
an
outward
structured
be
given
the
status
of
small
towns.
I suggest
that
if
would
communal
indicate
some
decisions
for
aqueduct,
and
political
the
therefore
organization
construction
such sites
and
should be
which
could
management
accorded
make
of
small
an
town
status. The problem settlement sites are those with baths only, which
need
to be
examined whether
the
indirect
evidence
could
indicate
is
possible
structures,
that
some
of
these
aqueducts
were
rather
minor
they are classed as major or minor towns. At items 3(c)-(h) and 4(a)(d)
he
lists
the
requirements
that
were
necessary
to
classify
or aqueducts
that
these
two
as
parameters
structural
types
classification.
should
be
included
I
as
1,
the
41
sites
214
they
baths. If a site
the
site type may also be more than a simple settlement, and could fall
in the
category
Eleanor
Scott's
definitive
of
to
instance
High
Several of
'Gazetteer'
statement
sensible
a villa.
that
assume
(1993) on
they were
that some of
Ham,Rowlands
sites
are
listed
in
with
no
would
be
villas, though
villas.
these
Castle,
the
However
were
it
indeed
Sandringham
and
villas,
for
Stonesfield.
at
Lyminge,
Orton Longueville
and
Scawby,
based
on
Scott's
so I
settlements during the early Roman period and later developed into
villa estates, as at Boreham, Durham, Haddon(?), Somerford Keynes and
Yarwell,
but
I have
listed
them
as
settlements,
because
of
the
uncertainty. Others were vicus sites associated with forts where the
baths were outside the forts,
here were public. Whether these vici later developed into settlements
or towns
in their
especially when
own right,
as
the associated
to
vici,
function,
mansiones
Chigwell,
may
Cold
Knapp
Point,
Romford
off
and
as
vici,
Tilston.
for
Many
example
settlement
sites are associated with villas and it is clear that the villa had a
bath.
However
some
of
the
settlements
near
villas
had
their
own
baths, usually small baths with only one or two small rooms, such as
at Aldbourne, Asthall, Castor(2),
listed
in the
developed
into villas
(or may
be
even
towns)
are
probably
likely to have had their baths built during the later Roman period.
215
Vici often had mansiones and hence would have had public baths which
were used by
the
(Black
1995,
14,
118,
120) . The vici baths were most probably built by the army if they
were
built
soon
after
and
Romford
Leintwardine
the
associated
vici
are
forts
likely
to
were
have
built.
had
Both
official
Chigwell,
Cold
Knapp and Tilston also fall in this category. Settlements near villas
probably had private baths
the
site
of
Castor (2)
(my
database
number),
subburb
of
small
or by Burnham.
Although the
pre-date the
In Table 6.6
shown in the
first column.
define'small towns'.14
as
entered
as
villas,
though
because
of
the
uncertainty
I have
14 Todd 1970, 114-30; Rodwell & Rowley (eds.), 1975, several authors; Rivet 1975,
14; Brown (ed.), 1995, several authors, in particular Burnham, 1995a, 7-17.
216
111-
Proposed
classification:
as villas:
Site names:
Badbury, High Ham, Rowlands Castle,
(previous as
settlements)
as villas:
(settlements that
later may have
developed into
villas)
vici:
Leintwardine, Romford.
vici + mansiones:
settlements:
Somerford Keynes.
(near
villas)
Salford Priors.
Settlements: without
a particular
perplexing
question
is,
how
one
would
determine
whether
the
them
building
structures
of
a public
indicators,
nature,
or
such as
evidence
of
other
common
regions
populated.
of
Rodwell
Roman
refers
Britain
to
was
significantly
Richmond's
217
comment
on
less
"the
densely
enormous
number
of
numerous
settlement
sites"
settlements
pretension
(Hingley
were
1989,
in the Essex
probably
75),
region.15 Many
simple
farmsteads
of
with
these
little
sufficient
for
their own needs. Many British settlements would have produced surplus
crops which would have been available for distribution to markets, or
may have been associated with villas
to whom they
supplied
their
excess products.
settlements,
not
only
must
their
social
and
economic
6. VILLAS
Of the 372 villas recorded in the database 305 had baths (82%) , but even
this seems to be low considering that there were probably about 2000
known villas. The remains of villas with baths are spread over the whole
area of Roman Britain,
County Durham. The large number of villas recorded to have had baths
seem to indicate that many more must have had them.
Table 6.7 shows that 104 villas listed in the database had wells and
24 had aqueducts. This represents 34.4% of the sites for which some
information
is
Fig.6.15
the distribution
of
available
on water-supply
structures.
Comparing my
it
is apparent that the sites with wells are fairly evenly spread over
the
area
of
Roman
Britain
where
villas
have
been
found.
From
where
wells
have
been
found
and
those
without
wells
are
similar, so their water needs could have been the same. It would seem
reasonable therefore to conclude that since the 104 sites with wells
were randomly scattered amongst the population of known villas,
the
likelihood is that most of the other villas also had wells. The 305
baths (82%) at villas (Fig.6.10) represents the most prolific feature
at all the sites recorded in the database and the same conclusion is
likely
to be
true
about
their
distribution as was
suggested
for
wells.
15 Rodwell W, 1975, BAR 15, 'T r i n o vantian Towns and their Setting:
Richmond I, V CH Essex, iii, 1963, 55-6.
218
A Case
S t u d y 1, 85;
Aqueducts
24
6%
AQ + BA
16
4%
Baths
305
82%
AQ + W
1%
Wells
104
28%
AQ + TA
2%
Drains
67
18%
AQ + WP
10
3%
Pipes
34
9%
BA + W
55
15%
Springs
23
6%
BA + TA
30
8%
Tanks
38
10%
BA + SP
15
4%
Sewers
1%
13
3%
+ TA
is very low
that
excavation
building areas,
did
not
generally
extend
outside
the
villa
is indicated at
5) seems unrealistic.
the
farmstead
settlements
and
the minor
villas
would
require
special study. One of the reasons why the figure for wells at villas
is low relates to the past history of villa excavation,
concentrated
features.
villas,
on
the
main
Rarely has
building
extensive
investigating
and
its
mosaics,
which has
and
subsidiary
structures,
yards
and
other
around
associated
16 A r c h a e o l . J.
140,
1983,
129-50.
219
future excavations
Figures
5.9
(Rockbourne)
and 5.10
(Great Witcombe)
their baths and it is clear that they are much less complex than the
larger baths,
larger baths such as the caldarium, tepidarium and the frigidarium, and
the usual hypocaust. Great Witcombe is a large villa with an original
bath complex in the west range dating from c. AD 250-270, and a second
more complex bath-suite was added c. AD 270, which lasted till c. AD
400. This bath had a hot water tank which supplied heated water to the
hot plunge bath next to it. There are existing springs above the site,
which may have been the source of a water supply to the villa during the
Roman period. Excavations at Bignor villa, with two bath-suites
1982)
(Frere
There
of water-related
6.8 gives
shows various
but
have
features were
a summary of water-related
combinations of features.
chosen
the
ones
listed
features
at
forts
and
which
had
special
than
relying
on
wells.
Nearly
and
half
the
listed
forts
had
These
figures
220
seem to
indicate
several
Aqueducts (AQ):
66
48%
AQ + BA:
45
33%
Baths (BA) :
85
61%
AQ + W:
16
12%
Wells (W) :
43
31%
BA + W:
26
19%
Wells only:
26
20%
AQ + BA + W: 14
10%
Pipes (WP) :
28
20%
AQ + WP:
20%
Tanks (T) :
32
23%
AQ + W + WP: 9
7%
TA + W:
19
14%
AQ + TA:
23
17%
27
137
134
The probability must be high that most auxiliary forts with baths had
running water
Shields,
Benwell,
Corbridge,
had
supplies.
Halton
Chesters,
aqueducts.
Wallsend,
It
Chesters
Vindolanda,
would
Rudchester,
Bowness-on-Solway
Along Hadrian's
also
Wall,
(with
the
an
forts
at
elaborate
South
bath),
therefore
be
unusual
Carrawburgh,
Carvoran,
not
had
to
have
for
the
Old
Carlisle
running
forts
water
all
at
and
supplies.
are
also
Balmuildy,
but
forts
at
the
with
aqueducts
forts
of
at
Inveresk,
Crammond,
Falkirk,
Mumrills
Rough
and
Castle,
here)
only
baths
have
been
recorded.
Again,
why
not
aqueducts at these
at
many
forts.
of
of
the
water-related
generally
other
features
The
seems
concentrated
on
to
the
lack
of
consistent
distribution
indicate that
fort
defences
internal
and
excavation has
buildings.
so that 27
16
(20%)
wells as a means of water supply. The data indicate that about one
third of the forts had wells, but, again, since few sites have been
extensively
excavated,
it
is
likely
that
many
more
wells
may
database
with
wells
not
associated
with
either
baths
or
221
waterpipes.
systems.
The pipes
Thirty two
(23%)
of the aqueduct
(16%)
sites
Five forts
and a fortlet had tanks which were not associated with any other form
of listed water supply. This indicates to me that these forts would
have had either wells or aqueducts, if not both. Three of these tank
sites
were associatedwith
baths
and
therefore
strengthens
the
of 85 baths were
found at forts
This
and only
forts
would
(33%)
that they could provide bathing facilities for anything from about 80
to 500 soldiers, suggesting that baths would have had to be more than
mere tubs. At a number of forts, baths were located in the associated
vici, though no doubt these were used primarily by the soldiers, but
perhaps also by the inhabitants of the vici.
have had bath arrangements for the commander, but these do not seem
to have been listed separately as baths at forts. Some of the smaller
forts did have internal baths for the soldiers, but they do not seem
to have been large and could have had their water supply from wells.
However
army
policy
would
seem
to
have
favoured
running
water
The 45 sites that had both baths and aqueducts at military sites seem to
confirm that aqueducts were the preferred form of water supply, though
there are still a large number of sites listed in the database where
evidence
and aqueducts
is lacking.
Wells
are poorly
forts would have had them and especially if a site also had a bath.
6.7
also
shows
combinations
of
different
water-related
were
quite
substantial
or
the
owners
of
these
villas
were
wealthy. The relatively low figure of 55 sites (18%) with both baths
and wells,
222
By contrast,
listed small towns with aqueducts and 5 associated with baths. I discuss
the implication of this anomalous
typographical
combinations
of
features
these.
recorded
Some
of
the
in
the
database
combinations
do
and also
not have
the
numbers
archaeological
of
sites
three
record.
which
do
What
not
have
also
any
is
of
of
the
interest
above
are
the
mentioned
types. The difference will give the number of sites with the chosen
feature.
Table 6.9
shows a summary of
and also
other associations.
Table 6.9: Summary of associations of water-related features
Forts
Manicipla.
1
Civitaa Small
capitals towns
10
5
Settle Villas
ments
4
16
AQ + BA
45
Colonlae
3
AQ + W
16
AQ + TA
23
AQ + WP
27
10
BA + W
26
18
10
55
BA + TA
24
30
+ TA
19
13
AQ+BA+W
14
AQ+WP+W
The combinations of
features
for
forts and
villas, but is applicable for all the data. For instance, using the
data from Tables 6.2 and 6.9 for forts, and calculating differences
223
by
taking
associations
of
two
features
at
time
from
amongst
aqueducts, baths and wells, and subtracting the number of sites with
that combination from the total number sites of each type, will give
the number of sites with each feature. There are 6 permutations for
the three features:
(1)
(2)
(3)
AQtotal
( A Q + B A )
A Q oni y ,
i .e .
66 - 45 = 21, and
BAtotal
( B A + A Q )
BAoniy /
i .e .
85 - 45 = 40;
BAtotal
( B A + W )
BAoniy,
i .e .
85 - 26 = 59, and
Wtotal
( W
+B A )
W only ,
i .e .
43 - 26 = 17;
AQtotal
( A Q + W )
AQoniy/
i .e .
66 - 16 = 50,
Wtotal
( W + A Q )
i .e .
43 - 16 = 27.
W 0nl y /
and
the
following
information
is
given:
2 4 Aq
to ta l
1 6
aq
+BA
In summary,
Table
between
associations
the
6.10
AQ o n ly #
a n d
3 0 5 Ba
to ta l
1 6
aq
+BA
2 8 9 Ba
o n ly *
totals
for aqueducts,
baths
and
wells. The last column shows the totals of each type of structure.
The associations for the other sites are not as significant as for
forts and villas.
There
are
other
associations,
such
as
combinations
of
all
three
so that the
52,
74 and 29
respectively.
Table 6.10: Number of fort and villa sites with aqueducts, baths
wells only.
Forts:
Villas
(3)
(2)
(1)
Totals
AQoniy
21
BAoniy
59
AQoniy
50
AQtotal
66
BAoniy
40
Wonly
17
Wonly
27
BAtotal
85
Wtotal
43
AQoniy
BAoniy
250
AQoniy
19
AQtotal
24
BAoniy
289
Wonly
49
Wonly
300
BAtotal
305
Wtotal
104
Tables 6.9 and 6.10 show that there are anomalies in the excavated
data for forts and villas,
types,
For forts
the figures
indicate to
me
it
is unusual that in group (1) 45 sites should have had both aqueducts
and baths, but 21 sites had only aqueducts out of the total of 66.
Similarly, 40 sites had baths only while there were 66 aqueducts and
85 baths recorded. The same questions are applicable to group (2): 26
sites were recorded as having had both baths and wells, yet out of
the 85 baths and 43 wells recorded,
and 17 had
wells only. For group (3): 16 sites had both aqueducts and wells, but
only 50
sites
of
database,
the
at
data.
Of
the
137
military
71 no aqueducts were
Although this
sites
found and at
recorded
in
the
94 no wells were
found,
supply
argued
that
spring
have
been out
of
character
even
if it is
sites not
to
It would
have
had
provision for some specific water supply. These 137 military sites
represent
only
about
half
sites,
which would
imply that only about a quarter of the known sites had aqueducts and
about
one
This seems
to
associated with both aqueducts and wells, which shows that more than
half the villas listed in the database had no water supply,
they
all
were
supplied
by
springs,
which
are
then
also
unless
grossly
involved.
information,
Clearly
indicates
that
the
record,
based
the
anomalous
225
on
available
gaps
in the
discussion
features
of
the
geographical
distribution
of
water-related
follows.
8. AQUEDUCTS
The data indicate that aqueduct water supplies to military sites in
Britain
were
common
(48%
of
my
sample),
indicating
rational
Roman
Britain
especially
from
concentration
military
Wall.
(Fig. 6.1),
the
was
early
where much
second
associated
of
century AD
with
Hadrians
The other
types
of
the
army
was
onwards.
Wall.
deployed
The highest
The
remaining
and at
forts,
became
fully
operational.
Agricolan period
during
number
the
later
of
these would
first
century,
the
date
but
sites
from
the
those along
Hadrian's Wall would have been constructed between c.AD 122-38, such
as the aqueducts at Chesters, Great Chesters and Halton Chesters. The
chronology of the fort at Vindolanda is fairly well attested because
of the numerous writing tablets that have been recovered in the last
two decades. The first timber fort at Vindolanda appears to have been
constructed c. AD 85, and rebuilt twice in timber from about AD 95 to
102. When Hadrian's Wall was started the fort was again rebuilt (c.AD
122-30) and was capable of housing 1,000 men. It may have been during
this
period
constructed
that
the
stone
(Birley 1931,
firm evidence
channel
aqueduct
to
the
fort
was
140s
the
first stone fort was constructed on the same site, but it was smaller
than
its
predecessors.
probable
reason
why
the
forts
were
reconstructed so often was that the wood did not last for more than
about 10 to 20 years. Around AD 220 the fort was rebuilt for the last
time, which is the remains that are now visible.
It seems that at
this time a civilian population also grew up outside the western wall
and a mansio with a bath was also constructed, which obtained water
from a tank at the terminal end of the aqueduct and from a well near
the tank.
226
the 66 at forts, accounts for 61% of the total number of sites where
they were found. The proportion of sites with aqueducts at all the
known forts (341) and major towns (20) amounts to 22.7%. As mentioned
above
have had running water supplies, so that the 22.7% proportion at both
forts and major towns underestimates the likely real situation.
In contrast to forts and major towns, comparatively few small towns,
settlements and villas have been listed as having had running water
supplies, representing only 8% for all those site types
(650) in the
database.
The
small
(Fig. 6.3)
number
is
of
only
low considering
small
that
towns
with
30 sites
(Burnham 1986,
aqueduct
had baths.
remains
Of
the
97
wells, of which only two were recorded as having had both. A number
of small towns had mansiones for which well water supplies would have
been adequate,
and baths are recorded whereas only 5 sites had both aqueducts and
baths.
Settlements also had a low number of aqueducts
24
villas
proportion
(6.5%)
and
that
that
had
aqueducts
aqueducts
were
notthe
also
normal
indicate
form
low
of water
supply to villas, since only 6.5% of the 372 listed villas had them.
It is likely that only the more prosperous owners would have gone to
that
expense.
number
of
uncertain.
However,
villas
Many
spring sources,
that
villas
whether
would
may
the
have
have
low
had
been
figure represent
running water
sited
to
the
supplies
exploit
real
is
localized
is
wells as
Caerwent,
London and Silchester. Even at towns which had aqueduct water supplies
as at Dorchester and Wroxeter, wells have been found at private homes.
227
presumably coming
9. BATHS.
Baths were the most common single type of water-related structure in
Roman Britain
representing
database.
of
the
Not unexpectedly
total
number
of
recorded
sites
in
the
and a
frontier
zone
Wall,
and
the
remaining
Thirty of
is likely
the sites
with baths are associated with aqueducts, whereas 18 sites had both
baths and wells,
which
seems
towns with baths, wells or aqueducts were situated below the HumberNorth Wales axis.
Eighteen villa sites
sites,
Grimstead,
Littlecote
17 For example at Rome and Pompeii, and Ptolemais in Libya (Ward-Perkins 1986, 109-53)
and Volubilis in West North Africa
(Wilson 1995,
52-6). Wilson comments on the
importance of social status that p r o v ision of running water supply brings to owners of
houses.
18 R C H M 1970, 534, 587, (533-89). The report of the villa excavatons (1961-3) does not
mention any aspect of the water supply of the villa. Not even the well (Pro. of the
Dorset Nat. Hist. & Archaeo. S o c ., 1982, Vol.104, 'Excavations on the Library Site,
C olliton Park, Dorchester, 1961-3', by G & N Ait k e n ) .
228
each, though at none of these have any form of running water supply
been found. At Grimstead the only water-related structures recorded
are the three baths, whereas both Littlecote Park and Northleigh had
in addition three wells each.
10. WELLS
Some aspects
of
the problem
of wells
at
forts
have
already been
and baths.
but it hardly
seems possible that it would have supplied all the water required for
the establishments of 500 and 1000 soldiers. At 18 of the 43 forts
where wells have been found more than one well has been recorded: 4
had 2 wells, 3 had 3 wells, and 5 records merely show more than one
well. The fort at Templeborough had 5 wells, and Derby and the Lunt
each had 6 wells, while Newstead had a record 107 wells/pits already
discussed. The coloniae had wells which may have been dug when they
were still fortresses. The 43 fort sites where wells have been found
underestimate the number of forts which would have had wells.
Wells were found at all four coloniae and the two municipia, and at
10 of
the
civitas
their
water
capitals,
supplies
have
and
been
the
implications
discussed
above.
associated with
For
the
civitas
(16), Silchester
(7 6) and Wroxeter
reasons why wells may not have been found, such as unsuitable geology
and therefore the hydrology of the site, hence a greater reliance on
aqueducts, springs, river water and rainwater. Usually wells were of
a sufficient depth that they would have been difficult
to destroy
shows
the most
(52), settlements
important
form of water
(104), and
this trend would likely be even more pronounced if more data at sites
on wells were available.
Aqueducts
the preferred
whereas 104 (28%) wells were found, but even this is unrepresentative
for the large number of known villa sites. Many villas with wells
seem to lie in a broad band along the Jurassic Ridge from the East
Midlands to the Gloucester/Cirencester area (Fig.6.15). Whether there
229
distribution is not
clear,other
axis. The
baths, which seems to be confirmed by the data of Table 6.10 for the
limited sample of sites in the database. There were 182 villas with
baths which were not associated with either wells or aqueducts.
It
seems to me that there were many sites that had both baths and wells
but they have not been found for any one of the reasons given in
section 2.2. Springs may have been an important source of water for
many villas, but they have also not been recorded, and the conduits
linking them have disappeared.
The distributions
settlements
of
(Figs.6.11,
6.13,
6.14)
for
also
forts,
small
indicate
that
towns
they
and
were
(Millett,
1990,
47)
(Millett, 1990, 143) for small towns, indicate that these sites with
wells were also uniformly spread over the landscape, so that it would
be reasonable to assume that many of the other sites were also likely
to have had wells.
The
of
Britain and
Fig.6.14
showing
settlements
with wells
the sites where no water supply has been found are very likely to
have had wells but they remain undiscovered.
listed
with
water
pipes
is
very
low.
Water
pipes
were
The materials
for pipes (see Chapter 2.5.5 and lc-f p.30) were wood, lead, ceramics
and stone. Stone pipes were not used in Britain, presumably because
of cost and the skill
230
Ceramic pipes
and were
Chichester,
Inchtuthill,
and Newstead.
Evidence
of lead pipes were found at 7 sites and lead would appear to have
been
used
to
repair other
forms
of
pipes. Lead
pipeswere
most
encased
hermetically
in
concrete, probably
sealed
because
they
could
be
other
forms
of
also
pipes.
an
expensive
material
to use
and
costly
to manufacture
Chester,
Hardknot
Castle
and
Winchester,
and
for
Carpow,
internal
lists
aqueducts or
1985b,
20
sites
where
wooden
pipes were
used
222-3 0) . He
(1985a,
either
forts,
as
198-202;
whereas
8 settlements and 34
villas.
Distribution water pipes must have been one of the most vulnerable
features
on all
phases
of
piping
existed.
types
of
reconstruction
Wooden
ancient
of
pipes
structures
would
along
street
generally
have
had
where
limited
robbed
development
removed
would
fronts
also
have
(Chapter 2, p.26).
of
value.
piping,
Later
hence
the
Where
they do occur they suggest that a running water supply existed, but
it has been difficult to associate some of these distribution systems
at
sites
that
also had
aqueducts.
At Wroxeter where
distribution
pipes have been found and it is known that the town had an aqueduct,
it has not so far been possible to directly link the two systems.
231
However tanks
drinking
(and wells)
water.
They
to
were
or
built
above
ground
level
top
(1992,
fed from
as a rule above
ground level both at forts and elsewhere, however some sunken tanks
have been reported, such as at Lyne.
At Colchester five tanks have been found, two clay-lined tanks were
at private dwellings outside the Balkerne Gate (Report 3, 1984, 1412) , and three are reported at building sites inside the colonia in
Report 6 (1989): a timber-lined tank (89-90, Fig.3.48), a tile-lined
tank (89, Fig.347, 255), and a lead-lined tank (355-6). This variety
of tanks
at
one
site
is unusual.
Figure
4.8
shows a timber-lined
cannot
really be
supply
sources,
that were at the terminal end of a running water supply, such as the
very large tank at the North Wall Gate of the Lincoln upper colonia.
At Housesteads fort,
Roof-top
runoff was
of
filling the
tanks. The fort at the Lunt had 15 tanks of which some were probably
filled from rooftops for drinking water supply.
The more simply constructed tanks in Britain where they were at the
terminal
end
of
an
aqueduct,
served
Italy,
the
same
function
as
the
However
their relatively small capacities did not allow for back-up storage
if an aqueduct was damaged or out of service.
Tanks used as catchment for rainwater were probably more common than
the archaeological record indicates because excavators cannot always
be sure whether a tank near a building received its water
from a
rooftop.
in
Although
102
sites
with
tanks
have
been
listed
the
232
purposes, so that the lack of evidence of more fort sites with tanks
indicates a problem of tank survival.
Similarly,
of 212 settlement
with no obvious
indication from the evidence that the less Romanized sites such as
farmsteads
and
minor
villas
had
tanks.
This
may
be
due
to
stored
pits
in
clay-lined
or
in
some
form of
ceramic
container.
Wooden barrels could also have been used extensively as water butts,
but would normally leave no traces in the archaeological record.
removed
water
and
waste
products
from
domestic
properties.
and excess overflow from fountains, waste water from baths and often
from latrines,
and
from
domestic
waste
water
outlets.
Whether
they
were
called
Drains
from homes,
slabs,
baths
and
latrines
often were
channels
into
recorded at
179
sites.
Remains
of drains
are usually
In towns,
sides
of
drains
streets.
Forts
would
have
had
network
A d a m 1984,
262, p i s . 612,
233
613).
of
drains,
London,
Melandra,
Strageath
and
Verulamium.
At
Corbridge
the
remains of elaborate stone channel drains were found along both sides
of Dere
Street,
and the
When
drains were along the centre of streets they were sometimes provided
with a kind of stone grating with openings cut into a paving slab so
that water can flow directly into the drain below. A number of such
gratings
have
been
found,
examples
of which
are
shown
by
Hodge,
is
likely
recorded,
that
most
of
the
sites
where
drains
have
not
been
had some form of drains but they have not been found or
(197 0, 254)
refers to the
deep"
at
Housesteads.
conventional sewer.
This
is
an
open
drain
and
not
but this
in towns
is a problem,
because they
were often cleaned out, so that datable finds of pottery or coins may
not necessarily reflect the earliest deposits in the sewers closer to
the time when they were constructed. The sewers at Caerleon, Chester,
Colchester,
Lincoln
and
York
were
major
underground
arched
stone
a main sewer as
shown
in Fig.5.12 (Whitwell
(Whitwell,
1976, 4-5,
main sewer
its branch lines. The sewer system, probably constructed during the
early part of the 2nd century,
the
outfall
of
the
sewer
234
is
likely
to
have
discharged
(Whitwell 1976, 23-5) . Branches 2 and 6 of the sewer may have formed
part of the draining system of the baths about 40m south of the main
sewer.
The remains of the main sewer at Lincoln now lie buried below the
present Bailgate,
usually
show
two
(Wright
branch
1852,
235-6).
Plans
of
the sewer
sewers
running
into
the main
system
sewer
(see
Fig.6.19) and one sewer parallel to the main sewer along a street to
the east of Ermine Street. Figure 6.22 shows a sketch T. Wright made
of the sewer and the
1970,
33,
linking
sewers
from cross-streets
(Whitwell,
(AD 81-96), so it is likely that the sewers may have been constructed
during this period or early during the 2nd century. The distribution
of these sewers in the town implies that the running water supply at
Lincoln was able to provide
Atkinson
(1942)
refers to the
'main sewer'
at Wroxeter,
and so do
reports in Britannia (8, 1977, 323-4, 394-6; 10, 1979, 297-8). I have
not been able to establish whether
it was
Lincoln and York, or whether the reference is to the large open drain
along Watling Street through the town between the forum and baths
insulae.
Fig.
6.22:
235
(after Whitwell
1970,
33,
short
periods
before
construction works,
or
start
that when
could
the
be
later
made
forts
on
such
were
large
built
the
The four
not
building
funding would
have had
the
them
as
their
to come
from
inhabitants of the towns, and sewers did not really contribute to the
status of the community or town.
which
was
arched
(1985,
39),
the
main
outfall drain (p.53). Its main function seems to have been to dispose
of large volumes of water from the springs. Similarly,
sewer at
the Great
the reported
likely a major
drain
military
sites
and
coloniae,
but
there
seem
to
be
some
236
14. CONCLUSION
The distribution of water-related
features
implies
that
many more
sites with aqueducts, wells, baths and the other features are still
to be
found.
found with some features but that associated features are missing,
such as baths without
limited
sample
of
a water
sites
supply.
listed
It
is also
in the database
shown
that
from the
the
favoured
means of water supply for forts and coloniae were aqueducts. Wells
were
the
villas,
preferred
although
water
supply
at
small
towns,
settlements
and
sites and
major towns. In some towns such as London and Silchester wells appear
to have been the only form of water supply.
more
widely
used
as
water
sources
at
all
of
sites
of all
than
types,
settlements, and even at villas where many sites have been found with
baths, they are few compared to the known number of villas. Evidence
seems to indicate that water catchment from rooftops may have been
common,
implying
that
roofs
must
Tanks
were
simple and effective way of storing water, so that the low number of
sites recorded with tanks seems to be unrepresentative. Water pipes
at many sites indicate their wide use, but they are underrepresented
in
the
record.
Drains
are
also
poorly
represented
at
all
sites
of
information.
structures
Some
has
structures,
been
difficult
because
especially at military
of
sites
lack
of
can be
237
239-243.
244-248.
249-253.
238
N
250
~
Fig. 6.1:
500 Kilometers
~
71
239
IK
N
250
500 Kilometers
240
IK
N
500 Kilometers
250
241
N
500 Kilometers A
250
Pig. 6.4:
242
N
250
500 Kilometers
243
N
250
"
Fig. 6.6:
500 Kilometers
-
244
IK
N
250
~
Fig. .7:
500 Kilometers A
"
245
l \
N
250
"
500 Kilometers
7
Fig. 6.8:
i - "
"
246
IA
N
250__________500 Kilometers
Fig. 6.9:
247
o.
.V
N
250
500 Kilometers
=
248
N
250
500 Kilometers
Fig. 6.11:
249
IK
r\,
250
Fig. 6.12:
500 Kilometers
250
,r\
N
250
500 Kilometers
- '""
'
Fig. 6.13:
251
\Jtf
*
'
0
N
500 Kilometers A
250
Fig. 6.14:
252
N
250
500 Kilometers
=
Pig. 6.15:
253
CHAPTER 7.
SOME SOCIAL, ECONOMIC, ORGANIZATIONAL AND CONSTRUCTION
ASPECTS OF WATER-RELATED FEATURES
1. INTRODUCTION
This chapter will briefly deal with some aspects of romanization and
the
social
impact
that
the
provision
of
water
and water-related
urbanize
the
local population.
administration
of
water
and new
supplies
and
the
the
organization and
public
baths
became
case
study
discussed.
of
of an
aqueduct
is
to manage
2. ROMANIZATION DEBATE
It would appear that the romanization debate for Roman Britain has taken
two directions:
The
and
change between that of the conquered societies of the Empire and the
established social order of Rome.1 The second deals with the material
culture introduced to those societies in the form of urbanization and
Roman buildings, housing, baths, water supplies and imported wares. The
Romans had a long history of romanization of their widespread empire and
Britain was one of the last areas to be exposed to their influence.
Recent research2 indicates that the British, particularly in the south,
were exposed to Roman goods and imports even before the conquest and
their response to the material culture introduced by Rome followed a
natural development of their own culture as they encountered the newly
imported romanized cultural influences. After the conquest romanization
became a subtle form of acculturation brought about by the exposure of
communities that had come under the commercial or political influence of
Rome,
and
the
importation
of
many
material
culture
products
that
1 Brendel 1979,
2 Cunliffe 1982;
Hingley,
1982,
17-52;
254
1-39.
poses
romanization:
work?
three
key
questions
relating
to
this
process
of
(ii) how did things actually change under Roman rule? and (iii)
how deep did the veneer of romanization percolate?" (Burnham 1995, 121) .
The
Romans
government"
actively
"promoted
growth
as
the
focus
of
self-
town
(Burnham 1995,
divisions
within the indigenous British tribal communities, but was careful not to
break up the existing social
structures unless
security demanded
it
"'romanization'
at a remove"
(Haselgrove
1989,
acculturation
22).
When
conquest, more
into
profound
discussion of
the
I do not intend to
romanization
process,
but
offered to
occupation of the country. The Britons, especially the 41ite, did absorb
some of the material aspects that the Romans introduced, such as stone
buildings for their homes, baths, aqueducts and wells.
Keith Branigan (1994, 9-16) comes out strongly in support of the theory
that
theamenities
which
the
Romans
introduced
during
their
3 Haselgrove,
1989,
17-30,
and n o s . 16-22, provides an analysis of the so-called
' r o m a n i z a t i o n 1 process before and immediately after the conquest period. The whole
process of romanization is very complex, but it is evident that the acculturation of
the British (and the Gauls) was a deliberate process, without it being enforced by
coercion,
as an expression of Roman values for their hopeful adoption.
In large
m easure the romanization was a c h i e v e d at certain levels and may have changed the life
styles of Britons in many respects, but did not change them into Roman Britons.
255
The
poorer
Britons
may
not
have responded
to the
romanization process with equal enthusiasm, since they could not escape
entirely from having contact with the Romans, mainly because of their
obligations for the taxes demanded by the occupation power. Even so the
construction of Roman forts, stone housing,
and fountains must have had
commenting
on
the
analysis
of
the
traditions.
data from
the
fort at
is so
seem
Roman society, or Roman culture, but rather many sub-cultures each with
its own aspirations and values"
(1995,
separation between Romans and Britons, but there may have been selective
differences
of the
remained
which
is
Nevertheless,
confirmed
however
by
Zosimus'
separate
the
writing
two
in
groups
the
may
6th
century.
have
remained
Roman baths
and
running water
supplies.
Villas were
the
Romans
the
Britons
imposed
new cultural
departed
values
on a
early
in
the
5th
century
leadership.
256
a unifying
the
absence
of
literary
or
epigraphic
information
on
the
evidence
law, the Lex Coloniae Genetiviae Juliae, setting out the charter
for the colonia of Urso in Spain and the details of how the town had
to be administered,
in which various
decrees regarding carrying out of public work and for the building of
an aqueduct
civitas capitals developed from the 2nd century BC into the Empire
period,
their
administrative
structures
being
based
on
that
of
Republican Rome and of the Empire period, but their municipal status
was ranked at a lower level than that of Rome (Abbott & Johnson 1926,
3-20).5 The models for those towns seem also to have applied to the
towns in Britain.
Chartered
towns
were
able
to
raise
revenue
through
the
incolae
to
(Duncan-Jones 1960,
civitas
capitals.
The
160,
164,
174-8),
four
coloniae were
preceded
by
have been
layout. Whether the army was ever involved in the construction of the
aqueduct for the colonia at Lincoln is uncertain, but the quality of
the
workmanship
of
considerable skill,
indications
are
the
concrete
encased
pipe
seems
to
indicate
that
Verulamium
was
municipium
and
the
army
surveyors may have been involved with setting out its street grid,
and possibly also were involved in the construction of its sewers and
original street drainage
257
town having had a walled fort of uncertain date, but founded about AD
100
(Frere 1974 125) and incorporated into the city walls probably
(Morris 1982,
171). There is no
suggestion that the army was involved in the planning of the town,
though the construction of some structures such as the basilica and
the
Huggin
Hill
baths
may
indicate
that
there
was
imperial
involvement and hence probably the army (Selkirk 1995, 329). The fact
that London had a fort within its walls complicated the question of
its administration,
municipium and would therefore have had an ordo like the coloniae.
Ways of raising revenue by civitas capitals could have been from minor
local
taxes
of different
kinds
(Duncan-Jones
1960,
160 n.8),
or by
and
baths,
including
funding
for
costly
aqueduct.
170(5.2).6 A
town
council
(ordo
decurionum)
consisted
of
on
brought"
"the
lands
through
(Hardy
1912,
44-5,
magistrates,
the
equivalent
which
(sect.
of
streets,
the
water
an
aqueduct
XCIX)).
aediles,
who
The
may
lawfully
duoviri
were
be
appointed
responsible
for
supply
and
drainage.
It
would
seem
reasonable to assume that this model for town councils also applied
to the public towns of Roman Britain and that decisions to build an
aqueduct would have followed a similar procedure as that at Urso. All
the civitas capitals would have been administered based on the same
model as that of the chartered towns without the benefit of raising
taxes for its development projects,
Durnovaria (Dorchester).
6 Duncan-Jones 1990, 174-8 and 1982, 84. 'Pliny mentions that the summae honorariae of
n ew councilors at one town, C l a u d i o p o l i s , were immediately put to use in building new
town baths'
176). Summae h o n orariae applied
to chartered towns like coloniae and
municipia. There are inscriptions from Lepcis Magna and Subratta implying very large
individual contributions towards the cost of aqueducts and fountains. The town
of
Durnovaria must have relied h e avily on private donations from inhabitants and local
taxes.
258
3.1.
S o c i a l a s p e c t s a n d d a t i n g of the D o r c h e s t e r a q u e d u c t w a t e r supply.
c.
AD
70
(RCHM(E)
1970,
534),
but
later,
because
of
(Wacher 1995,
decided
to
build
aqueduct
is
uncertain.
There
was
certainly
prosperity amongst the native British during the 2nd to 4th centuries
(Salway 1981, 235-8). Based on the remains of the public buildings such as a forum, basilica, public baths and amphitheatre - the economic
situation at Durnovaria was prosperous during the later 2nd and 3rd
centuries
aqueduct
2nd century period of prosperity, even though the town already had wells
to supply drinking water.
constructed about
two decades
after
1968 and 1980, at least in the vicinity of Poundbury (A, Fig. 7.2a).
7 According to Bill Putnam (1984, 32-3), the Durotriges were originally not all that
interested in the building of a new civitas town at Dorchester during the Flavian
period,
but by the 3rd century their wealth had considerably improved,
as for
instance, can be seen by the extensive villa of Colliton Park.
259
\x\nniii
ve>*
>** t 4 7 8
Jrov q n t
oreat >9M-942
SECTION 16 (Aqueduct)
TWA
Fig.7.2a: Poundbury Camp and aqueduct (after Green 1987, 16, Fig.l);
Fig.7.2b: The aqueduct channel showng the three levels of construction:
1st, 2nd and 3rd phases, (after Green 1987, 50, Fig. 23).
260
Dating is dificult, but is based on samian ware of the late 1st or early
2nd century found in the channel silt, and a slightly worn silver coin
of Vespasian (AD 69-79) found in the diversion conduit on Colliton Park
(RCHM(E)
1970, 588-9; Green 1969, 172). Samian ware was also found at
one of the places sectioned along the east side of Fordington Bottom (B,
Fig. 7.4). Based on this evidence, the narrow channel of the aqueduct
(Fig.7.2b) of the first phase was cut during the late 1st century. Green
suggests
aqueduct
was
reconstructed
0.5m
higher
along
this
section,
to
the
a
(RCHM(E)
1970, Vol.2, 589). There seems to be evidence also for such changes at
Bradford Peverell (see C, Fig.7.4) (Green 1987, 51). This last phase may
have coincided with the building of the Roman baths.
The tenuous nature of the dating evidence for the first two phases of
construction
of
the aqueduct,
raises
the
issue of whether
newly
indicates
that
the
earthwork
structures
(1987,
were
started
'within
28 to 32ha)8 with
little regard
for
(by
the political
division between the northern and southern groups9 and this must have
created intense rivalry between families from the north and south. The
northern group centred round Lindinis (Ilchester, 8.9ha) may well have
been the wealthier as indicated by the number of villas surrounding it
8 RCHM(E)
970,
n.3,
533.
9 Ilchester seems to have become a separate civitas centre at a later date (Wacher
1995, 21) , but see Rivet and Smith,
giving the alternative views of Stevens and
Bogaers
(1979, 392-3). Stevens suggested, based on inscriptions,
that the Civitas
Durot r i g u m was subdivided w i t h Ilchester as the capitol in the north and Durnovaria
that of the south. Bogaers disagreed, suggesting that civitas Durotrigum simply meant
"town in the territory of the D u r o t r i g e s ".
261
(Wacher 1995, 324) . Since the initial civitas capital was at Durnovaria
(32ha) it is possible that there was competition between the two groups
to hold office within the ordo, as rival factions vied with each other
to provide benefactions for expensive building projects to gain civic
status. The material remains indicate clearly that the inhabitants of
the civitas did indulge in considerable expenditure at Durnovaria to
improve their social standing, material culture, and civic status, but
this aspect has received minimal attention.
This would have applied to other public structures constructed during
the same period. If the calculation of water delivery suggested for the
aqueduct is valid (RCHM(E) 1970, 587), then much
funding
amongst
an aqueduct.Firstly,
to raise fundsby
and
facilities had
imposing taxes, so
to be
raised
from
Secondly,
did
this
contribute
to
overstretching
of
resources? The answer to the first question was complex, because the
Durotriges community as a civitas stipendiarius also had tax commitments
to the province, but as suggested above, rivalry between the northern
and southern groups could have been a motivation. The Durotriges tribal
area had several important LPRIA hillforts with rival 61ite families who
had accepted the romanized building styles, but retained their former
wealth
and
competition
overspent
power. An
importantelement
between
large
sums
of
money
on
could
public
well
buildings,
have
been
baths
the
having
and
an
public buildings by the fourth century (Millett 1990, 130, Table 6.1).
Inflation and increased tax also became significant factors during the
4th century and the decuriones became less inclined to contribute to
development of cities during the 3rd century (Millet, 1990, 128, 204).
Either
the wealth
of
even
the
rich declined
during
the
later
3rd
262
situation
of
the
late
third
century
resulting
from
the
troubles on the northern and eastern frontiers in the 4th century (3468) , which could have adversely influenced the markets that provided the
wealth of the previous century. New demands were made on the wealth of
inhabitants to provide defences, as at Durnovaria and other towns in the
civitas, so that many public facilities must have suffered because of
lack of maintenance. The expenditure of the earlier period in many areas
of Roman Britain could not be sustained later as available resources
became overstretched by channeling it into defences. The rich also spent
more on providing themselves with more luxurious homes, further removing
financial resources from towns and inevitably the towns declined as
their buildings, water supplies and drainage deteriorated, in spite of
having become walled towns. Durnovaria seems to have declined in this
way during the 4th century when its public baths went out of use,
probably because the aqueduct was also no longer in use.
4. D O R C H E S T E R A Q U E D U C T C A S E - S T U D Y
4.1.
D i s c o v e r y of D o r c h e s t e r aqueduct.
(8)
s e c r t o a/ 5
263
section ofwhat he
The
leat
occasions,
aqueduct
but
the
to
Durnovaria
actual
source
has
been
investigated
has
never
been
on
reliably
three
located
1925 (RCHM
1970,
Artillery)
subsequently suggested in
Notton Mill (G, Fig. 7.4),
major
in the Royal
theaqueduct was at
IRIMSTONE'
F R AM PTO N
V.
Roman ru in * ,
f la m p io n
iiC oui I
lO T T O N
Jm u c k l i f o r o
JUTHOVER
IH R O O P
Step*
fa rm /
*.l Poimribuf]
BRADFORD;
P E V E R E LL /
W H itflp M
ipphoA
Poundbuiy,
Form
lA itc io n t t U l d is ro n p W l
.... .
h a rd i
3000
1500
45 00 FEET
I M IL E
3 K I LO M ETRES
ETRES
264
why the source of its supply has not been located. Bill Putnam has been
doubtful that the source was at Notton Mill and has been excavating
along the aqueduct
sociated with
the
since 1992,
locate
its
source.
Four
reports have been published so far in the Dorset Proceedings10 and there
is an interim 5th report relating to the 1996 excavations. Previous to
Putnam's work the aqueduct's route had been firmly established as far as
Penns Plantation
the route as far as Littlewood (F) , that is about 2.5km further than was
previously
known
(Putnam
1995,
128-31).
In
the
interim
report
he
suggests that a dam was constructed across the Steps Bottom stream just
south of the Littlewood farm buildings
source of the aqueduct, and will be investigated during 1997. This would
reduce the length of the aqueduct by about 4km, to a length of about
14km instead of the length given in the RCHM(E) article (1970, 585) of
about 18km. However, in my analysis below I use the latter length.
There are a number of factors which determine the type of construction
of an aqueduct,
such as
cost,
available
revenue,
labour
resources,
suitable water source, and topography. The topography along the chosen
route of the aqueduct was such that for the Dorchester aqueduct it was
decided to have an open leat aqueduct, cut into soil and in places into
flint rock to maintain the flow rate.
The straight line distance from the town to Notton Mill
about 9.1km (5.7 miles)
(G) is only
is known,
it shows
aqueduct.
Usually
the
necessary
expertise was
supplied
by
the
architects and engineers in the Roman army. If the army was involved,
what would have been the relationship between it and the civitas? A pos10 Dorset Proceedings, 114, 1992, 239-40; 115, 1993, 152-3; 116, 1994, 123-4, and 117,
1995, 128-31; Putnam W G & Hewitt I, Interim Report December 1996 'Excavations and
fieldwork on the Dorchester Rom a n Aqueduct 1996', Bournemouth University.
11 R C H M ( E ) , County of Dorset,
England,
2(3),
265
1970,
585.
sibility is that some retired army veterans with the required knowledge
might have settled in the town and helped with the technical aspects
related to the aqueduct. Alternatively, the town ordo might have hired
qualified engineers and contractors retired from the army, or could have
called in the army to assist.
individuals
had
construction,
also
acquired
the
technical
skills
necessary
(Campbell 1996,
for
77-80).
It
would seem that the provincial authorities allowed the army to assist
with technical help for towns if they were not occupied on military
campaigns,
private
as
for
instance
encouragement
and
Tacitus12 suggests
official
assistance
that
to
"Agricola
the
gave
building
of
temples, public squares and private mansions" and presumably other types
of construction such as aqueducts.
finished
materials
for
structures
building
the
and
how
the
structures
labourers
can
manipulated
conjectured.
Adam
the
(1994)
Wall
some
possible
sequences
of
construction
have
been
suggested, but the details are still partly obscure. We can conjecture
about
them
by
applying modern
technological
knowledge
(itself
very
G e r m a n y , trans.
by M Mattingly,
1948,
The
Penguin
266
2.
3.
6.
7.
8.
9.
10. arrangements for the control of the workforce and allocation of work
to various sections of the project;
11. arrangements for the inspection of the work;
12. execution of the work;
13. approval by senior authorities of the progress of the work;
14. final acceptance of the completed work.
some
involved
1800 years
I am aware of
some
of
the
implications
of
the
planning,
construction
and
cost
of
an
been instructed to make all the necessary arrangements to plan and have
the aqueduct constructed. The next step would have been to identify
appropriate water source,
an
267
channel.
Many aspects of Roman surveying remain uncertain. However, we know the
type of very simple instruments they used such as the groma, chorobate
and
dioptra
(Dilke
1980) . Various
possible
methods
of
surveying
technique have been suggested that could have been applied. Whatever
methods they actually used,
difficult routes is attested all over the Roman Empire16. The surveying
of the routes of the Nimes aqueduct in France and the KOln aqueduct must
be among the more outstanding surveys done by the Romans. The fall in
elevation for the Nimes aqueduct, in quite broken and hilly countryside,
R o m e 1, from
268
was
also
of
high
quality,
as
for
instance
those
of
the
speculative,
but
some
idea of how
planning and budget costs are arrived at in practice today (very much
simplified) . It is not intended to imply that this was the procedure
followed by the Romans.
It is to be noted that for the fairly low degree of sloping ground of
ratio 1:3, about 20 degrees to the horizontal, the area is nearly double
that for level ground. However,
17 The details of the field w o r k of these cross-sectional surveys and the surveys by
the Royal Engineers are in the archives of the Dorset County Museum, Dorchester.
269
Area: 2.839m'
1..524m ---------1
Area: 1.694m'
1.324m
These are large volumes of earth and rock to move, considering that the
tools used were principally pickaxes and spades, hammers and chisels to
loosen the rock and soil. I have chosen for convenience of calculation a
value near the mean of those volumes, that is 41,400m3.
In the RCHM(E) article (1970, 587)18 it was suggested that the slopes of
the walls of the actual cut were at the ratio 1:0.5. This is very steep,
giving a vertical to horizontal slope of about 63 degrees. For soils in
general,
at steep slopes
like this,
place over a period of time during wetting and drying cycles, one can
expect slumping of the sides to the natural angle that would be stable
for such soils.
270
that a maximum slope would have been at most 45 degrees (ratio of 1:1).
This would have increased the chosen volume by about 9,000m3 (22.2%) .
When we try to estimate the labour force required to excavate such
quantities
of
soil
one
can apply
of
labour-
which he
absurdly
low
(see
below) . He
does
not
give
details
of
his
19 Based on the 138,400 labouring hours and the volume of material of 30,750m3 it would
have taken 4.5 hours to dig a cubic metre of compacted material or 1.78m3 per day. if a
present wage of 4/hour is taken, it would cost 18 to dig one cubic metre of material,
which seems low.
271
hours" to be too low for the reason mentioned, and that he did not allow
for side-cut into sloping ground, of which there was a fair amount.
when loosened,
occupy
between 1.65 to 1.75 times its compacted volume; I have used a factor of
1.71. Hence the uncompacted volume calculated from the cross-sections
shown in Fig. 7.5 should be 41,400m3 multiplied by 1.71 giving 70,700m3
to be dumped as uncompacted spoil. For calculation purposes I used the
slightly higher volume of 70,800m3 (c. 2.5 million ft3) to make the
calculations clearer. At 353 spadefuls per m 3 (10 spadefuls per ft3),
this would have required about 25 million spade movements21.
Based on personal experience of unskilled labour-intensive road- and
dam-building projects, I shall consider two time lengths that it would
have most likely taken to complete a spade movement: a half-minute and a
one-minute time span. These time spans vary considerably in practice,
but it provides an order of work activity for estimation purposes. These
also take account of waiting time between different activities, breaks
and changeover time.
To start with, let us consider the basic spade movement of half a minute
to set out the calculation procedure for the time spent on shoveling the
loosened earth into containers. The figures thus produced can then be
adjusted as multiples of the half-minute base value. On this basis it
would have taken of the order of 12.5 million minutes or roughly about
8,680
man-days
for
24
hour
day
(23.8
man-years),
to
fill
the
20 Initially I calculated everything in the feet/pound system and then converted the
values
to me t r ic values.
The original
figures were all rounded figures but on
conversion to metric equivalents tens and units enter into the values which is not
really j u stified as one cannot usually estimate large bulk volumes so closely.
21 See p r evious note.
272
containers with
ft3) of loosened
soil. For an 8 hour working day the above will come to 26,040 man-days
(71.3 man-years). If the period was one minute for completion of a spade
movement the time taken would be doubled. The above applies only to the
filling of the containers. The diggers loosening the soil would have
taken a comparable time and so would the carriers of the spoil.
So
of
containers
and
the
removal
of
soil.
have
used
the
time ofconstruction,
as
estimate
excavation,
of
and
the
time
taken
for
200
men
to
complete
all
the
summarized in Table 7.3 (p.276) . If we assume that the large force unit
of 200 'labourers' was available to fill the containers each day, that
would give a figure of 43.4 labour-force-days for a 24 hour work-day. If
a days work was based on 8 hours and the efficiency 75% (43.4 x 3 -r
0.75), then the time taken would have been 174 days (about half a year)
to complete the excavation if all the work was only in soil. With a
labour force of 100, the work would require 348 days, or about 1 year.
If the labour force was only 50, the work would take 696 days or nearly
two years, and with only 25, the work would have taken 1,392 days or
nearly 4 years. If the time period is one minute per spade movement on
average, then the above figures would all be doubled. Therefore, for the
labour force of 50 the project could easily have taken about 4 years,
and with half that number the project could have taken about 8 years.
The logistics of the movement of the containers,
experience,
would
have
required
proportion
of
based on practical
carriers
directly
related to the distance they had to walk to dump the spoil. This ratio
multiplied by the 200 labourers who filled the containers would give the
number
of
carriers
for
each distance.
Table 7.2
(no.23)
gives
the
22 For m o d e r n construction projects, apart from the preliminary planning period, the
actual construction period on earthworks projects is usually governed by the time
excavation and moving of earth takes to complete. The same would apply to a building
project where the controlling operation is usually chosen to estimate the length to
complete a project.
273
Time
Pickup
Total
Filling
Ratio
Filling
Ratio
walked
taken
and
time to
c o n t a i n e r at
1:
c o n t a i n e r at
1:
(mins)
dumping
dump
0.5min per
1.Omin per
soil
(mins)
spadefull
spadefull
(mins)
from
pickup
position
(min)
(m)
(mins)
Ys
Y2.5
200
6.4
0.4
6.8
2.5
2. 72**
5.0
1.36**
143.75
4.6
0.4
5.0
2.5
2.0
5.0
1.0
100
3.2
0.4
3.6
2.5
1.44
5.0
0. 72
65.625
2.1
0.4
2.5
2.5
1.0
5.0
0.5
50
1.6
0.4
2.0
2.5
0.80
5.0
0.44
25
0.8
0.4
1.2
2.5
0.48
5.0
0.34
10
n*
0.16
0.4
0.56
2.5
0.29
5.0
0.14
This situation will pertain when the soil is dumped onto the embankment
directly by s p a d e . For only this situation will no carriers be needed
The ratios show that the time taken to walk the distances can be either more
than, equal,, or less than
the time taken to fill a container. When the ratio
is equal to 1, then the number of carriers required match the number labourers
filling containers. If the ratio is greater, more carriers are required by the
proportion indicated by the ratio. If the ratio is less than 1, then the
number of carriers needed will be less by the proportions shown by the ratios.
If the ratio is zero then no carriers are involved and the spoil is dumped in
its final place by spade. The critical distances shown for the ratios to be 1
is shown in bold numerals. On a labour intensive project tables of this kind
(usually more elaborate) are used to regulate labour distribution for this
type of work. Equations of the linear relationships are given by 1) Y 2.5 =
0.4x, and 2) ys = 0.2x. Equations relating the distances walked to the relative
ratios are given by 3) y 2.5 = 0.0144d, and 4) y 5 = 0.0072d. From these any other
values of the ratios can be calculated based on the time taken and distances
to walk to dump the spoil, from which estimates can be made to balance number
of labourers filling containers against number of carriers.
23 The mass p er unit volume of in situ soil is about 1800kg/m3 , i.e about 112 l b / f t 3 .
Therefore, in its loosened state it will be about 112/1.65 pounds or nearly 70 lb/ f t 3 .
It is e s timated that a carrier will on average carry about 35 pounds of earth at a
time. For a rate of 0.5min per spade movement and 10 spades per cubic foot of soil, it
will take 2.5 m inutes to fill a carrier's container of about one-half a cubic foot. If
the spade mov e m ent is 1 min/spade it will take 5 minutes. The distance the soil will
have to be carried will vary. Table 7.2 shows distances from 200m to zero and assuming
that the carrier walks at a pace of 16 min/km. The times taken to walk to the dumping
p o s i t i o n and b ack are shown in second column, to which must be added the times shown
for items 1 and 2 below, which remains constant. The sequence of his movements will be
as follows:
1 2 3 -
0.2 min.
0.2 min.
l/2min/spadefull; 2.5 mins (constant).
1 min/spadefull
5.0 mins (constant).
Therefore,
proportion of
filling
time
: collecting,
walking
and dumping
time,
expressed as a ratio of y:x, wh e r e x are the times given in the 4th column, and y2.5
and y 5 the ratio values given in column 6 as l:y 2 .s, and in column 8 as l:ys.
274
There would have been a labour contingent who loosened the soil with
pickaxes. It would probably have consisted of at least as many labourers
as those who filled the containers, that is another 200 labourers. As
there would have been any number of miscellaneous duties to perform I
also add another 100 men of more professional status. There were likely
to
have
been
other
functionaries
such
as
accountants,
paymasters,
been at the lower end of the scale shown in the table. And, of course,
the number could have varied from time to time during the work period.
Who the labourers were that worked on the aqueduct at Durnovaria is not
known. Pliny the Younger refers to criminals condemned to work as forced
labourers on public works
in
24 Pliny,
EP.
10.31-2.
275
25 million minutes
= 8,680 man-days
(24 hour-day)
= 23.8 man-years
17,360 man-days
(24 hour-day)
71.3 man-years
= 26,040
52,080
8-hour man-days
8-hour man-days
1 min/spade
Labourers
Days
Years +-
Days
Years +-
200
100
50
25
174
348
696
1,392
0.5
1
1.9
3.8
348
696
1,392
2,784
1
1.9
3.8
7.6
in the text.
without any
construction
carried out.
generous of HS2.525 per day for labourers doing the digging, shoveling
and carrying on the project.
Hel&ne Cuvigny
(JRS 86,
1996,
13 9-45)
in
of c. HS1.6
276
25 denarii per day and food, which seems to reflect an inflation rate
of
17-19
prices
times
of
(Duncan-Jones
the wheat-price
1982,
366).
compared with
If these rates
the
of
2nd
century
inflation are
This
is 3.5
times the Mons Claudianus rate of HS1.6/day and 2.25 times the rate
of
HS2.5/day.
The
construction
of
the
Dorchester
aqueduct
is
suggested as having been built in the late 1st or early 2nd century,
or about 200 years before the DPE had effect. Therefore the rates of
pay provided for in the DPE cannot be used as a common labourer's
wage
174
days/men =HS348,000
and
b) : {(HSl. 6 /day x 600) + (HS5 x 100)} men x 174 days/men = HS254,040.
If
the
aqueduct
was
actually
4km
shorter, as
suggested
by
recent
excavations, the costs would have reduced by a factor of about 22%, that
is, to HS278,400 and HS203,200.
It is difficult to compare these figures with costs for aqueducts in
other parts of the Empire where costly stone channels and bridge struc
tures carrying the aqueduct conduits were involved.
By AD 212 the rates of pay had increased by 4 times and by AD 235 the rates had
increased by 8 times to HS7,200 and HS6,000/per year, or HS19.7 and HSl6.4/day. (M A
Speidel in J R S 82, 1992, 106). The author states that the bold figures are based on
direct d o c umentary or literary evidence.
27 Duncan- J o n e s 1990, 176-7: W a cher 1978, Roman Britain, 215, says that the basic
agricultural wage for a day labourer was 25 denarii, but he also received fringe
benefits of f o o d ' . I have not included food costs, asuming that the labourer provided
his own food.
277
aqueduct.28 Coulton
(1987,
84,
n.43)
gives
costs
of
(Aspendos)
several
to HS30
HS350
million' .
(Macmullen does
not
seem
to
give
costs
for
it was lined
Fig. 7.6 Photograph of rock cut of the Dorchester leat, near Bradford
Peverell (AB).
This activity would have included another sequence in the construction
operation, not allowed for in the cost estimate of the aqueduct. I have
no figures for the length of channel that was in rock-cut, but if one
takes a stab at 5% (I suspect it was higher) , and at a rate of four
28 Pliny,
EP.
10.37.
278
times the cost of excavating the soil, then we need to add another fifth
to the above cost amount giving an excavation labour cost of a) about
HS365,400, and b) about HS266,700. To this should be added the equipment
and material costs,
and
bucket
containers,
rope,
pegs,
levelling
staffs,
sack
surveying
equipment, and other peripheral costs. This is usually taken for estima
tion purposes at 40%. So an overall rough estimate for the Dorchester
leat aqueduct for the two rates of pay is a) about HS511,500, and b)
about HS373,400.
However, if one considers the inflation that has taken place indicated
by Duncan-Jones'
later 3rd to early 4th century, the rate of pay for labourers could have
increased by a factor of 2, which would have doubled the cost of the
aqueduct. This aqueduct was a simple earthworks type structure only, but
the cost of digging and moving of one cubic metre of soil/rock would
have been about HS15, which is low considering that a fair amount of
rock was present in the excavations. The cost calculated applies only to
direct labour costs for digging and dumping the material. There usually
are hidden costs such as surveying and problems with the peculiarity of
the site, which would have added to the total cost. The assumption is
that the town hired the labour for the project.
private contractor who tendered for the work, the costs would probably
have been considerably higher, because a contractor has to allow in his
tender for some profit and for hidden costs, and the constraints placed
on him.
The above cost estimate is based on a number of assumptions which might
not have applied at all during antiquity, but it gives an order of cost
for
such
a major
represented
notwithstanding
project
large
sum
in
the
of
money
costs
This
town
suggested
like
in the
would
have
Durnovaria,
sources
for
either way,
but
the
implication
is that
constructing
an
aqueduct like the Dorchester one was a major undertaking. The appointed
aedile would initiate, plan, budget, organize the labour force, provide
technical
expertise,
construction
of
food,
the aqueduct.
equipment,
and
finally
The maintenance of
supervise
the
the aqueduct
once
constructed, would have been his responsibility, but even so must have
been a constant concern for the town administrators. There however seems
to be no evidence of maintenance activities on the Dorchester aqueduct,
279
which,
of
course,
The RCHM(E) article (1970, 587) quotes the discharge volume calculated
according to Lewis and Jones
depths in the leat. For a depth of flow of only 0.305m (1ft) the water
quantity would be less than half the previously suggested discharge
rate, that is only 27,040,000 litres/day. If the flow depth was 0.24m
(l/2ft) the quantity of water delivered would be only about l/5th the
suggested figure at 11,781,000 litres/day. The storage capacity at the
source end would have had to be considerable to provide even this low
quantity of flow. This low flow rate is still a large amount of water
for a town,
of say,
A2= 1.11m2
0.5
0.61m
A, = 0.51m2
0..305m
0.152m
A* = 0.243m2
!:524m
A 2 : Ai : A* = 1 : 0.46 : 0.22
Approximately 1 : 1 / 2 :
Many other
280
litres/day.
One
of
the
reasons
why
aqueducts
provided
such
large
(cut-off
wooden
planks
which
could
be
slipped
into
grooved
deposits
which
The phenomenon
had
to
be
removed
is caused by a chemical
calcium carbonate.
from
the
channels
of
reaction between
in the atmosphere
to
form
to
clean
the
drains
along
streets
and
public
latrines,
and
community.
We know very
set about
paintings do give some indication of men at work and what tools they
281
in Wiltshire
(pp.123-4).
(p. 137)
Pitt-Rivers
refers
to
the
two
deep
wells
he
wells
were
likely
and about
the remains
of the
least
one
person
could dig
and place
the
loosened material
in
containers for hauling to the surface. For that reason wells are seldom
recorded as having a diameter less than one metre. A number of wells
have been found with diameters of 2m, and larger.
In large diameter
29
wells more than one person could work down a shaft at the same time .
Wells with a square cross-section may not necessarily have been dug
square, but their steining timbers give the impression that they were
square. There were advantages to digging a round well to insert square
steining timbers, because if the diameter was slightly larger than the
diagonal of a square steining structure it would assist in supporting
2 9 In areas where shafts were sunk for recovery of lime in the chalk formations of
southern England the shafts were usually larger than 2m in diameter so that more than
one p e r s o n could work below at the same time.
282
the corner posts and would have provided the space to place the lateral
planks
in position.
special
problems,
Working
particularly
of
view
(see
extrapolation
Modern
of
costs
costs
based
for drilling
on
modern
practice
would
be
large diameters
order of 3000 to 6000 per metre drilling in soft rock, with at least
another 25% additional costs to establish the equipment and personnel on
a site to do the work. A shaft of 30m depth could cost from 100,000 to
250,000. Drilling in hard rock formations could increase those costs by
several
orders.
In
antiquity
digging
shaft
in
soil
or
gravel
start
had
to be carried out
at
regular
depth
intervals before it would have been safe to proceed further down. Major
problems in digging a shaft by hand
about 6m down.
amateurs whodig
It is customary for
wells by hand to
have air cylinders blow fresh air into the hole with a pipe lowered to
the bottom. The ancients did not have this advantage.
The rate of digging a well is difficult to assess because it depended on
the material to be penetrated and the quality of the chissels used to
loosen the rock. Soils and gravels would be relatively easy to dig, but
as the depth increases removing the loosened material would change the
rate of effective penetration of the shaft. Table 7.4 gives the volumes
of materials that have to be removed for wells of different diameters
and depths.
If a container of
0.5ft3
material to the surface it would require well over 2000 movements of the
container for a lm diameter well 10m deep.
283
Depth(m)
1
10
20
30
0.78
3.9
7.8
15.7
23.5
3.14
15.7
31.4
62.8
94
7.07
35.3
70.7
141.4
212.1
Effectively the rate of digging the shaft could decrease from lm for the
first
metre of
digging per day to l/5th and less per dayfor deep wells.
could
become
would be several times slower and here the rate of progress would depend
more on the rate to penetrate the rock and usually steining was not
necessary.
so that the
rate of progress of sinking a well would have taken even longer than
shown above.
This brief discussion is intended to show only some aspects of well
digging, as there are other problems that would have faced the ancient
well digger. An important issue always was where to dig and be certain
of getting to water-bearing layers. Although Vitruvius
his views
on
finding water,
in practice
it seems
carried out on a site where water was required, hoping that water will
be found. In areas like London and Silchester the materials were water
bearing gravels so that water was easily found. The two deep wells at
the settlement at Rushmore and the one at Lincoln were in rock.
quarrying,
working,
Agriculture
would
building,
tanning,
have
lime production,
pottery
depended
284
and
on
the
tile
manufacturing,
mosaic
rainwater,
so
industries.
that
special
provision
of water
supplies
to
irrigate crops
and market
gardens
water,
adequate.
while
for
some
Roman
pottery
others,
well
manufacturing
supplies
was
would
major
have
been
industry
in
Britain,
military
zone
1990,
206,
Map
6:24).
Several
44).
At
Colchester
there
is
indication
of
an
aqueduct
at
Sheepen, which possibly could have served the extensive pottery works
at Warren Fields
kilns have been found. The pottery production centre at Alice Holt
Forest,
was another sizable operation. Two short leats have been found taking
water
from small
Jeffries
1979,
3-4,
Figs.
1 & 2).
(Lyne &
(Gilmour
1955, 536-45; YAJ 39, 1956, 33, Fig. 1). If this was the only well it
must
have
been
public
well,
which
introduces
the
question
of
It is
likely that alternative water supplies must have been available. If,
for the production of pottery at each kiln, a minimum of 100 litres
of water per day was required,
would have had to be drawn from the well each day. A 2m diameter well
with a constant water depth of about 2m would have been adequate to
provide that quantity of water each day. The constant preparation of
the potting
seem
to
indicate.
Swan
has
indicated
that
The
water
from the
storage
extensive pottery
285
industry
along
the
Nene
valley
most
likely
obtained
its
water
aqueducts
were
specifically
constructed
for
industrial
use.
Since the mines would have become imperial property once the people
of west Wales were
subdued,
constructed by the army with the help of slaves. The main aqueduct
was
11km
long
and
tapped
the
Cothi
river
Several
near
the
Pwll
Uffern
and
Haverfield commented that some sections may have had a flat ledge on
which
wooden
troughs
or
pipes
were
laid,
which
have
long
since
about
quickly
11
million
released
technique
to
litres
in great
expose
capacity,
volumes
gold-bearing
to
from which
remove
quartz
soil
veins.
water
could
in the
Excess
be
hushing
water
was
directed into a reservoir above the main workings, which was used for
ore-washing
(Jones,
the Annell and Gwenlais valleys provided water for other parts of the
mine workings
crossed a low point and it has been suggested by the excavators that
a conduit of some kind may have been carried on wooden trestles to
maintain a flow gradient, but there is no visible evidence of remains
to confirm the suggestion. It has also been claimed that an aqueduct
near Linley (Shropshire) also supplied water for sluicing purposes to
expose lead and silver ore bodies. There may have been other sites
where this hydraulic technique was used
1990,
Table 6.1) .
Tanning and metal working industries would also have needed water,
for which wells and other simple water storage facilities would have
been adequate.
Wroxeter
Evidence
found at
at Alcester
(Hingley 1989,
probably the water supply, but no water supply has been recorded for
the fortlet at Brithdir, for which the water supply was probably the
River Wnion.
Metal-working
widely spread over Britain, and at many of the sites water supplies
have been recorded,
but
286
The need for water in the building industry would primarily have been
used
for
mixing
of
mortars
to
lay
bricks
and
bond
masonry
in
structures, plastering of walls, and for mortar sealant for tanks and
channel type aqueducts. The water could have been obtained from any
source such as local springs,
dedicated
channeled
from
water
springs
supplies,
or
which
streams.
could
Industrial
easily
have
enterprises
been
would
for
emperor, as for
of Caracalla,
Ostia, baths used by the public could be privately owned (Nielsen 1990,
119). Towns were the owners of public baths
have been employees who would have been specialized to run such an
enterprise. Nielsen suggests they were often freedmen or slaves (1993,
125). If there were traders, tavern-keepers and prostitutes at baths,
they no doubt paid some form of rent to the owner. In Rome many baths
were constructed by private persons and run as commercial enterprises
(Nielsen 1990, 120) . In Britain it is not certain whether there were any
privately owned baths that could be used by the public. Privately owned
baths in Britain usually were associated with villas or houses in towns.
However there were large public baths in towns such as the baths at
Wroxeter, Leicester and Bath. It is not known how these British baths
were administered and whether they were run on the same basis as their
counterparts elsewhere.
287
It would seem that the supervision of public baths came in the early
days under the aediles, one of the town magistrates, and during imperial
times were supervised by an appointed curator, and still later became
the responsibility of the city prefect.31 There was an official called
the balneator who seemed to be the person who was in daily charge of the
baths (Nielsen 1993, 127), with many responsibilities, and it was he who
had to oversee that the bath temperatures were correct, and that there
was
enough
fuel
complaints were made to the owner if there were some aspect bathers
objected to in the running of the baths. However, we do not know whether
the same sort of control existed at public baths in Britain. Conceivably
one of the aediles or an assistant would have exercised that duty at
baths in British towns. Who would have been the employees at baths in
Britain is difficult to ascertain, because the type of person in Britain
who would have worked at a bath would not likely have been freedman or
slaves as suggested for Italy or other parts of the Empire. Whether
baths were run by the town itself or whether they were leased out as a
commercial enterprise we will probably never know. Small private baths
probably
had
one
attendant,
mainly
to ensure
that
(Plondon 1166,
This is about 13 times the amount paid annualy for water, also for Egypt
(see below) . The cost of water to baths is a difficult issue, because
public baths run by a town may have owned the aqueduct and therefore the
supply of water would normally have been supplied free to the bath. But
there are several references
supplied to some
public baths was paid for, for which the evidence is cited by Nielsen
(1993, 124, ns.26-29). For instance water for the Severan baths in Egypt
cost 18 obols per day, the equivalent of HS924 per year (Nielsen 1993,
124, n.30)32. It was not uncommon for a benefactor to invest an amount
31 N i e l s e n S, 1993, 125, n.7, refers to 0 Robinson who published an article 'Baths, an
aspect of R oman local government l a w ,
in: Sodalitas. Studi in
onore A.Guarino, 3 pp.
1065-82. I have not been able to refer to this reference.
32 Plondon 1177,
lines 30ff,
{AD 113}.
288
of money that would cover the cost of water supply and fuel for public
baths. If baths were leased out, the lessee would have had to bear the
cost of supplying water to the bath and have had to obtain a license to
draw water
from an aqueduct
(Nielsen 1993,
123-4).
for bathing
to charge
for this
ostraca,
(Nielsen 1993, 132). An entry fee was charged to Greek gymnasia and the
king of Egypt received a large income from public baths. If the baths
were leased out, the king received a third part (trith) of their income
(Rostovtseff 1941, I, 312) . As commercial enterprises no doubt the rents
charged for the various facilities offered within the baths would have
been based on some profit making basis, because the admission charge
would be insufficient to cover all the costs of running the baths.
There are many references to the payment of admission charges,
first
referred to by Cicero (Cic. Cael. 62), who stated that the cost to use
baths
in
(= 1/4 as).
From Diocletion's
time
289
a small portion of the running costs of a bath. The leaseholder met his
financial
obligations
by
sub-licensing
certain
functions,
such
as
cloakroom facilities, shops and other facilities, and the supply of oil.
However,
baths functioned (Nielsen 1993, 124). For Roman Britain we also have no
information on the financial implications of running bath institutions.
YegOl refers to the text of the management of a small bath belonging to
the state mines and the mining community of Vipasca (modern Aljustrel?)
in Portugal. The bath was leased out under contract to a lessee, giving
the details of his obligations (1995, 47, and CIL II, no. 5181, lines
22f) . The contract stipulated that the lessee were '... required
to
heat the baths and keep them open for use entirely at his own expense
every day from day break to the seventh hour for women [until c. 12.001.00pm] and from the eighth hour of the evening to sunset for men [from
c. 1.00-2.30pm to c. 6.30-8.30pm; the first set of figures being winter
hours], at the discretion of the procurator in charge of the mines'. The
lessee
was responsible for 'a proper supply of running water for the
heated
rooms, to the bath-tub at the highest level and the basin, for
women as well as men. He shall charge men 1/2 as and women 1 as each
[1/4 of a sestertius] . Imperial freedmen and slaves in the service of
the procurator of the mines are admitted free;
soldiers...'.
woman
to have
different
bathing hours,
and
in some
of
the
larger
prohibition
separavi t")34.
The
against
hours
of
of
increasing scandals,
mixed bathing
bathing
reserved
("lavacra
for
pro
men
Hadrian
sexibus
during
the
may have
been
other
reasons why
the
sexes
were
separated,
a probable
reason for the increased entrance charge to them in this instance. The
33 Martial,
N.H.
24,
26,
(transl.
revoked
Satires, 6, 412ff.
under
Marcus
Rackman) ; Seneca,
290
Aurelius,
Letters,
86,
but
10,
raised
under
(transl.
RM
lessee was also required to keep the bronze equipment of the baths and
its fixtures clean by washing, drying and coating them with fresh grease
every 30 days. He had at all times to keep a supply of wood sufficient
for 30 days, and was not allowed to sell wood. If the baths were not
kept open and properly furnished he could be fined by the procurator up
to HS200 (Yegill 1993, 47). There clearly were very strict regulations to
control the running of baths and to ensure the safety and health of the
public.
It is not known whether there were any baths used by the public in
Britain and leased out to private lessees,
here.
There
is
very
little
information
on
complex
structural
features,
the
actual
because of
costly
little as HS60,000
(30-31,
124-30).
HS300,000-350,000 for his baths, excluding the cost of the land (Nielsen
1990, 121). Pliny gave a donation of HS300,000 for the decoration of the
baths at Comum, his home town and a further HS200,000 for the tutela or
maintenance of the baths. He probably also paid for the construction of
the baths.36 We have no similar cost figures for baths in Roman Britain,
but when baths of comparable sizes are compared from Britain to those in
other parts
of
the Empire,
(Nielsen 1990,
at a
36 CIL V 5262 Comum, c. 111-113 AD. It is not clear whether the tutela generallyin cluded cost of staff and running costs, such as provision of fuel and water (Nielsen
1993, 122-3) .
291
8. CONCLUSION
Romanization was a two-way process: Rome who initiated a course of
action to persuade her provincial subjects to adopt the Roman way of
life and in the case of Britain, the Britons who adopted aspects of
the
process.
There
was,
however,
large
section
of
Britons,
development
of
Romano-British
particularly reflected in
imposition
from above. To provide these romanized buildings with the mosaics and
wall
paintings,
deliberately
professionals
hired.
Italy must
culturally and
have been
religiously
the Dorchester
aqueducts in Britain,
leat,
Dorchester
construction
and
leat
indicates some
provides
relating
to
dating,
remained
in
service
an
order
maintenance and
is
practical
unknown.
of
costs.
length
The
aspects
of
Additional
of time
implications
its
issues
the
aqueduct
for
civitas
capital when a utility like an aqueduct went out of service must have
been serious,
loss
of
their
running
water
supplies
must
be
an
important
consideration.
Nothing is known about
the
single most
popular
How the
compared to what is
is
real
need
importance water-related
communities,
for
study
to
structures had
determine
what
in the economics
level
of
of urban
life of Britons.
292
CHAPTER 8.
CONCLUSIONS AND RECOMMENDATIONS
1. CONCLUSIONS
Roman water supplies and baths were dominant structural features at
many sites in Roman Britain, but their relevance to the social and
economic
development
included
in
of
urban
academic
and
discussion.
rural
The
centres
economic
has
rarely
been
implications
of
providing running water supplies and baths would have been important
for the communities involved, yet discussions on them do not enter
debate on the development of towns or rural settlement
(Todd, 1970,
1978; Burnham 1986, 1987; Hingley 1989; Burnham and Wacher 1990). To
understand fully the development of urban and rural centres
it is
social
villas.
and material
The most
recent
impacts were
on towns,
settlements
and
for
the
inhabitants,
who
became
dependant
on public
also have
(White
1990,
5)
More
rainfall
in Britain has
importance
attached by scholars for the need of water supplies, but even in the
wet countries they were priorities for both the Roman rulers and the
public.
it seldom enters
used for washing of the person, washing of clothing, and was needed
for
bathing
facilities,
for
specialist
293
production
in
textiles,
of
which
towns
households,
water
will
usage
need
supplied
or also
based
for
on
special
water
study.
only
specialist
contemporary
to
It
is
also
public
production,
pre-industrial
not
known
facilities
which needs
and
to be
investigated.
Aqueduct water supplies usually came from rural areas, but it is not
known what effect the taking of water from a particular area to a
town would have had on settlements based near the water source, which
needs to be investigated.
Aqueducts found in Roman Britain were mainly of the leat, pipe and
simple channel types
been made of their typology in Roman Britain. They were used at all
site types,
they were used at rural sites like villas, they were usually of pipe
construction and relatively short in length, tapping nearby springs.
Some
leat
Gadebridge
period,
type
aqueducts
have
Park) . Although
little
has
been
there
been
are
published
reported
many
about
for
leats
villas
from
their
(as
the
at
Roman
physical
and
been made
However,
that flowed in leats, or for how long the leats supplied water,
is
needs
to
be
researched
to
understand
294
how
effectively
they
that
for Colchester,
leats,
wooden,
Gloucester or York.
It has been
pipes
have been
influenced
to urban communities
the development
of
towns
and
indirectly had
some
impact on their economy, but no comment has been made on this aspect
of water
supply
in the
literature.
Determining the
life spans
of
towns. The life span of wooden aqueducts must have been very limited,
implying it was not a long-term means of providing water. Leats could
silt up or their banks could have collapsed, so that unless effective
maintenance was
carried
out
regularly,
an aqueduct
could
after a
to
indicate that
related to the end of the effective supply of water from the aqueduct
(Green 1984,
51).
several
Wroxeter)
sites
provision
of
public
wells
London,
must
Silchester
have
had
and
special
(Chapter 1, 7
and n.9). Wells were not only sources of water but often they were
also
used
artifacts
for
ritual
purposes
and
as
depositories
for
religious
been found at Silchester where they were the main source of water
supply,
their
internal
distribution
and
use
have
not
been
specifically studied since they were first reported during the 1890s
and early this century.
determine how many wells were in operation at the same time, and who
295
used them.
but
large areas have not yet been examined, with the likelihood that more
wells will be
found.
At Newstead,
recent
revealed more wells/pits than the original 107 found by Curie (Clarke
and Jones 1994, 110, Fig.l). At Wroxeter wells were found grouped in
two localised areas
the Roman town still unexplored. No Roman wells have been found at
Colchester within the colonia walls. At the other major Roman towns
there is also a lack of evidence for wells. Further archaeological
work will be needed to provide a more balanced distribution of wells.
Wells
are
not
easily
categorized,
basically
being
a hole
in the
linings.
3 0m. When dug into gravel, sands and clays their walls may have been
unstable
linings
have been found but the distribution of sites with different types of
linings and inter-site differences need to be correlated. Box-framed
linings changed in construction over time and also need correlation.
This
isparticularly
well
illustrated
by
the
difference
(Fig.4.4a-d,
pp.112-5)
in
from
the 1st century to that illustrated for London (Fig.4.2, 107). This
could have been due to improvements in carpentry skills and tools, a
topic well worth investigation.
Archaeological evidence indicates that at a number of sites the life
spans of wells were often of short duration.
distribution shown for Newstead (Clarke and Jones 1994, Fig.l), both
in the fort and the vicus to the south, suggests that many wells were
in use for short periods and only
one time
any
the
area.
Silchester
with
Whether
wells
had
similar
short
life
span
at
separate
buildings.
study
of
the
life
span
of
wells
may
provide information on changes that may have taken place within sites
and could be significant for villas.
Who dug wells is not known, but with the difficulties of well digging
it seems possible that itinerant professional well diggers may have
296
performed
that
task.
There
is
no
direct
evidence
to
prove
this
assertion. Builders may have contracted to do the work when they were
hired to build public buildings or homes for private owners. A study
of well digging and their steining could bring to light other aspects
of specialist skills in the building trade during the Roman period,
especially
possible
the
that
progress
the
in
skills
carpentry.
in
For
instance,
construction
of
it
steining
may
of
be
wells
water
related
structure,
importance as social
the
centres,
main
reason
having
been
their
More baths have been recorded at villas in Britain than at all the
other site types grouped together, indicating their social importance
to the private sector during the Roman period. Although public baths
had the practical function of bathing, perhaps their major function
was as centres where people could meet socially and obtain services
which
were
not
otherwise
available.
However,
for
Roman
Britain
is
supplemented
structures.
by
finds
found
at
baths
or
associated
the people who used baths in Britain from a study of such finds in
the context
of their clothing,
with
imaginative
some
supplemented
from what
interpretation.
is
known
This
can
of behaviour
and
be
judiciously
at the
continental
but
their
significance
not been
evaluated.
short
periods,
indicating
specific
social
need.
Chesters,
Vindolanda.
requires
important structures,
Halton
Chesters,
Great
At
many
especially at
such as at Caerleon,
Chesters,
Benwell,
and
research,
and
will
require
imaginitive
297
The remains of only two Roman baths in Britain have been described in
detail,
but the
carried
out
the
duties
of
attending
to
the
baths,
and
the
provision of the large amount of fuel that was needed for heating.
Baths at military sites may not have had the typical image of public
baths, but there was a social aspect to them as well, which needs to
be explored.
large
staff
considerable
evidence
was
social
shows
that
available
to
activities
the
baths
run
the
were
institution
available.
required
and
that
Archaeological
regular
maintenance
and
affected
the
use
of
the
baths.
There
is
similar
lack
of
baths
systematic
had
a variety
study
has
been
of
shapes,
published
layouts
of
their
and
sizes,
but
no
characteristics
in
An
overall
distribution
impression
from
the
data
presented
is
that
the
is
under-represented.
Summary of conclusions relating to Roman Britain:
1.
2.
3.
4.
5.
6.
298
2. RECOMMENDATIONS
The
following
set
providing a basis
of
recommendations
are
made
with
view
to
of
has been
whereas
in
The reason
In
it is therefore recommended
served
all
the
different
types
of
sites.
For
better
it is
so-called
Wroxeter.
Leicester
have
made
an
aqueduct and
those
at
attempt to
arrive
at
299
Dorchester and
a
cost
of the
context.
Research,
however,
is necessary
to
determine
how
in Britain,
A comparative
York,
research
wells.
needs
and
others,
is necessary
Caerwent, Gloucester,
large
areas
are
unexplored,
and
further
London, Wroxeter
correlated using
existing
and by
archaeological
site type,
information.
they
served
at
particular
structures,
also
needs
to
be
on baths
the
found at
International
all
Association
for
the
in Roman
Study
Britain.
of Ancient
appear one for Britain. Although baths have been studied extensively
for the Empire, detailed work on the typology of baths in Britain has
not been done except for a few sites. Classification of bath types
based on the simple three layouts used in this thesis, of row, block
and
complex
types
(Chapter
5,
300
p.145),
oversimplifies
their
configurations.
systematize
There
the way
aspects of baths
is
need
baths
for
research
are described
into
in the
this
topic
literature.
to
Other
distribution of bath types, how the public baths were managed, their
costs, and their place in Romano-British society. This research would
be a major study and it is recommended that it be allocated as a
thesis study.
to be
Colchester,
investigated,
Caerwent,
Gloucester,
Verulamium,
supplies
Dorchester,
of
Lincoln,
is especially significant,
only
be
resolved
by
further
archaeological
research.
The
it
is
considered vital
that
the
funds
be
found
for
further
lack
of
scholarly
debate
on water-related
structures
in
for
301
level,
on an
Volume 2: Appendices
by
October
1997
APPENDICES
(302-317)
(318-482)
(483-498)
Appendix 4: Abbreviations
(499-505)
Appendix 5s Bibliography
(506-545)
APPENDIX
Is
DATA
FROM
DATABASE
tip* opting
4**4
A b b o ts L a n g le y
A b e rg a v e n n y
A b in g d o n
BAl
fo rt
BA1
A b r id g e
v illa
AQl
It
W1
s e ttle m e n t
Low
v illa
A c to n S c o t t
Adel
A i l s w o r t h (1 )
A i l s w o r t h (2 )
BAl
v illa
s e ttle m e n t
BAl
AQl
v illa
s m a ll
A lc e s te r
W1 +
BAl
s m a ll
A ld b o r o u g h
A ld b o u r n e
W2
to w n
c iv ita s
BAl
AQ l
+ s e ttle m e n t?
A ld e n h a m
A ld e r m a s to n
A lfo ld e a n
TA3
BAl
BAl
v illa
BA l
+ s e ttle m e n t
BAl
W1
BAl
W1
s m a ll
A lr e s fo rd
to w n
v illa
AQ l
BAl
A lw in to n
s e ttle m e n t
W1
A n c a s te r
s m a ll
W5
to w n
A n g m e r in g (1 )
v illa
BA2
A n g m e r in g (2 )
v illa
BAl
A p e th o r p e
v illa
BA l
A p p le s h a w
v i 1la
BAl
A r b u r y Road
A rd o c h
DR1
SPl
SP1
W1 +
W1
s e ttle m e n t
fo rt
DRl
W3 +
to w n
fo rt
A lc h e s te r
(fa ls u s )
A rm o th
v illa
BAl
A rre to n
v illa
BAl
Ash
v illa
BAl
A s h -C u m - R id le y
v illa
BAl
DRl
Ashdon
v illa
BAl
DRl
A s h ill
v illa
A s h te a d
v illa
A s h to n
s m a ll
A s th a ll
W3
BAl
W1
to w n
+ s e ttle m e n t
BAl
A tw o rth
v illa
BAl
A u c k le y
s e ttle m e n t
A x m in s t e r
s m a ll
B a d b u ry
B a d g e w o rth
B a ld o c k
to w n
v illa
BAl
B a n c ro ft
v illa
v illa
TA2
W1
S P l , b ro o
k
AQl
It
BA2
BA 2
DR2 +
WPl
SPl
DR 2
WP1
SPl
BAl
fo rt
BAl
v illa
BAl
B a rm in g H e a t h
v illa
BAl
B a rn ac k
v i 1 la
P a rk
TAl
DRl
W3
A Q l-c l p i
B a rk b y T h o rp e
B a r n s le y
to w n
B a n w e ll
Bar H i l l
BAl
BAl
fo rt
DR 3
W1
+ s e ttle m e n t
s m a ll
B a l m u il d y
tom*
W1
s e ttle m e n t
A b i n g e r Hammer
A c k w o r th
v illa
W1
TA l
W1
v illa
BAl
B a rn w e l1 (1 )
v illa
BAl
B a r n w e l l (2 )
v illa
BAl
TAl
W3
LP1 +
302
Sxfce- isaraa
jSxfe* Typos
A<3XJB<312<31
Batch.
Weil
fo rt
B a ro c h a n
B a r r in g to n
v illa
B a r tlo w (l)
v illa
B a r tlo w (2 )
v illa
B a r t o n C o u r t Farm
to w n
Id p i
v illa
BAl
A Q l? -Id p i
s e t tle m e n t
fo rt
BAl
BAl
Beenham
v illa
BAl
B e r in s f ie ld
s e t tle m e n t
LPl
AQl
W1
BAl
WP1 ?
W1
DR l
B ib u ry
v illa
BAl
DRl
B id d e n h a m
v illa
s e ttle m e n t
B ig n o r
v illa
B ille r ic a y
s e ttle m e n t
B in c h e s te r
fo rt
W1
BAl
A Q l-
st
ch
B ir r e n s
fo rt
A Q l- s t
ch
to w n
DRl
BAl
v illa
BAl
B itto n
v illa
BAl
B le d lo w - C u m -S a u n d e r t o n
v illa
BAl
B le tc h in g le y
v illa
BAl
s e ttle m e n t
s e ttle m e n t
BA 2
B oreham
s e ttle m e n t
BAl
v illa
BAl
fo rt
BAl
B o th w e llh a u g h
B o tte s fo rd
B o u g h to n
DR2
W1
B o u rn e
v illa
AQ l
It
v illa
AQ l
It
AQl
It
Bowes
s m a ll
BAl
BA l
DRl
?
W1 +
to w n
fo rt
Box
v i 1la
B o x te d Farm
v illa
W1
DR1 +
BAl
B ra d fo rd -o n -A v o n
v illa
BAl
B r a d in g
v illa
BAl
H ill
B r a in tr e e
B r a is h f ie ld
W1
s e ttle m e n t
s m a ll
DR2 +
tow n
v illa
TAl
BAl
BAl
v illa
B r a d f o r d Down
DRl +
BAl
s e ttle m e n t
B ozeat
LPl
WP1 +
v illa
B r a d le y
W2
s e ttle m e n t
B o u r n e /M o r to n
TAl
TAl
DRl
v illa
B o u g h to n M o n c h e ls e a
B o u rto n -o n -th e -W a te r
WP1
WP2 wd
W1 +
B ly b o r o u g h
B o ro u g h H i l l
W1 +
BAl
B itte s b y
B lis w o r th
WP+
W1 ?
fo rt
s m a ll
DRl +
BA 2
B i r d o s w a ld
B ite r n e
TA2
DR 1 +
BA 2
v illa
B in s te a d
TAl
W2
BAl
B ig lis
SEl
DR l
fo rt
B e w c a s t le
SPl
TAl
v illa
fo rt
SPl
AQl
v illa
s e ttle m e n t
LP1 +
WP1 +
BAl
B e d d in g to n
B e n w e ll
W1 ?
BAl
B e d d in g h a m
B e r e R e g is
LP1
DR2
BAl
fo rt
B e c k fo o t
Sewer
TAl
LPl
W1
BAl
B e a d la m
fo rtre s s
Tank
W2
A Q l-
BAl
B e a u p o rt P ark
BAl
v illa
B e a u f r o n t R ed h o u s e
Spacing-
BAl
BAl
B a th fo rd
B e a rs d e n
fifH a
DRl
v illa
s m a ll
B a th
Pram
W1
BAl
303
TAl
Sxfce- Name
A iaifedtfcjt
; J S x t*
s m a ll
B ra m p to n
to w n
liiiliil l l p
fo rt
B ra n d o n
B r a n tin g h a m
B r a u g h in g
B re co n G a e r
v i l l a ( fa ls u s )
s m a ll
BAl
BAl
v illa
BAl
B r i s t o l (1 )
v illa
B r i s t o l (2 )
v illa
BAl
fo r t le t
BAl
v illa
BAl
B r o a d fie ld s
B rom h a m (2 )
+ s e ttle m e n t ?
BAl
B r o u g h - b y - B a in b r i d g e
fo rt
AQ l
B r o u g h - o n -H u m b e r (1 )
fo rt
AQl
B r o u g h - o n -H u m b e r (2 )
s e ttle m e n t
A Q l?
fo rt
AQ l
B u c k n o w le Farm
B unny
TAl
DRl
DRl
DR 2
fo rt
DR2 s t l n
v illa
BAl
AQl
BA l
DRl
B u rg h
v illa
BAl
v illa
BAl
B u rto n
s e ttle m e n t
B u x to n
s m a ll
Cadder
AQ l
It
AQl
WP1 c o n e
tow n
BA 2
C a e rh u n
fo rtre s s (A )
C a e r le o n
fo rtre s s (L )
A Q l-
Id
pi
C a ern a rv o n
fo rtre s s (A )
A Q 2- p i
ch
C a e rs w s
fo rtre s s (A )
C a e rw e n t
c iv ita s
C a is t o r -b y -N o r w ic h
AQ l
C a m b rid g e
c iv ita s
s m a ll
s m a ll
C a m p h ill
A Q l-w d p i
C a n tle y
BAl
W1
DR 14-
BA 5
W2
DR 1 +
DR3
PL2
DRl 4
PL14
BAl
W1
BAl
W1 +
BA 2
W16
BAl
DRl
BAl
C a r is b r o o k e
C a r lis le
C a r lis le - O ld
C a r m a r th e n
C a rp o w
C a rra w b u rg h
fo rt
WP24
TA3
W1
A Q l-p i
ch
WP1
BAl
DRl
BAl
AQ l
AQ l
AQl
DR l 4
BAl
BAl
It
TAl
BAl
DRl 4
BAl
W1
BAl
s e ttle m e n t
W2
C a rv o ra n
fo rt
W1
C a s t e l l C o lle n
fo rt
C a r s in g to n
WP1
W1 +
BA2
fo rt
c iv ita s
TAl
BAl
AQl
v i 1la
fo rtre s s (A )
SE1
W1
fo rtre s s (A )
c iv ita s
TA l
T A l-Id
DRl
to w n
f o r t le t
C a s t le
SPl
TA2
W3
s e ttle m e n t
Cappuck
Id
TAl
to w n
c iv it a s
WP1
BAl
v illa
C a n te r b u r y
SPl
DRl 4
fo rt
C a m elo n
C a m e rto n
DR 3
fo rt
v i 1 1a
C a in e
SP2
BA2
fo rt
C a is te r -o n -S e a
WP1
?
W1
Burham
fo r t
P L l4 ,S tch
W1 +
s e ttle m e n t
B u r r o w -in -L o n s d a le
SPl
W1 ?
s e ttle m e n t
fo rtre s s (A )
B u c k to n
W1
W1
BAl
B u c h le y
TA4
DRl
s e ttle m e n t
v illa
B ry n c ro s s
T s it k
W1
B rom ham (1 )
B r o u g h -o n -N o e
S jp rx n g
W1
fo rtre s s (A )
v illa
B r ith d ir
P ijp a
DR1 +
to w n
B re w o o d
B r ix w o r th
ftr a tir t
DRl
B r is lin g t o n
C a r d iff
ll
W3
BAl
BAl
304
DR24
WPl 4
SE1
x t *
Ty><3:
S fx fc e -
J S a ta e
C a s tle
C a ry
fo rt
C a s tle
G re g
f o r t le t
C a s tle
H i l l (1 )
C a s tle
H i l l (2 )
C a s tle
N ic k
C a s t le -D y k e s
s m a ll
B S th .
W3
BAl
v illa
fo rt
C a s t le s te a d s
fo rt
A Q l-s t c u l
BAl
BAl
D R ls tln
W1
DR1 +
v illa
BAl
W2
BAl
W2
C a ts g o re
v illa
s e ttle m e n t
C a v e rs h a m
s e ttle m e n t
C h a lk
ch.
DRl
W1
Chapel H i l l
fo rt
BAl
Charm y Down
v illa
C h e d w o r th (1 )
v illa
C h e d w o r th (2 )
v illa
C h e lls M anor
s e ttle m e n t
C h e lm s f o r d
s m a ll to w n
W1
W1
BA 2
BAl
BAl
fo rtre s s (L )
AQl
BAl
fo rt
AQ l
BAl
C h e s te r h o lm
fo rt
A Q l- s t
fo rt
A Q 2,
tow n
C h e s te rto n (2 )
s m a ll
tow n
C h e s t e r t o n (3 )
BAl
W3
DRl +
BAl
Wl
DR l +
W4
W2
C h i l g r o v e (1 )
v illa
BAl
C h i l g r o v e (2 )
v illa
BAl
Wl
C h ip p in g W arden
v illa
BAl
Wl
C h i s e ld o n
v illa
BAl
C h u r c h ill
s e ttle m e n t
C i r e n c e s t e r (2 )
C la y t o n
C liffe
House
c iv ita s
AQ l
v illa
BAl
BAl
BAl
BA 2 ?
BAl
C o ld K nap P o i n t
C o ld h a r b o u r
C o le rn e
C o le s h ill
C o llin g h a m
D R l-s t
WP -
ea
TA2
AQ l
W9
BA 2
DRl 0 +
WP9 +
v illa
SP5+
TA2 +
SPl
+ s e ttle m e n t
v illa
SEl
Wl
tow n
s e ttle m e n t
TAl
It?
v illa
C o ld B r a y f i e l d
SEl
TAl
DRl +
AQ l
v illa
c o lo n ia
W2
Cobham P a r k
C o lc h e s te r
TA3
SPl
SP3
W4 +
s e ttle m e n t
s m a ll
WPl
W3
s e t tle m e n t
Cobham
Coddenham
SEl
SE1
Wl
s e t tle m e n t
s e ttle m e n t
SP2 +
SPl
v illa
C n u t ' s Dyke
WP2+ I d
B A l?
BAl
C la y d o n P ik e
TA2
W1
+ s e ttle m e n t
H o s p ita l
st
ch
3rd
BAl
C h u r c h ill
SPl
DR 2
BAl
C i r e n c e s t e r (1 )
Id
DR l
D R l-
c iv ita s
C h i g w e ll
W P l-
W1 ?
fo rt
C h ic h e s te r
DR 4
WP1 +
W1
C h e s te rs
s m a ll
DRl
W1
A Q l- wd p i
C h e s t e r t o n (1 )
TAl
s e ttle m e n t
C h e s te r - le -S tr e e t
TAl
WP2 +
DRl
BAl
BAl
Cheshunt
SPl
BAl
AQl
v illa
C h e s te r
WP1
DR1 +
+ s e ttle m e n t
C a tty b ro o k
TAl
DRl
W1
C a s t o r (1 )
fo rt
S e w e r
DRl
BAl
C a s t o r (2 )
C a tte r ic k
T & ft-te
D R ls tln
fo rt
C a s t le s h a w
S fp r x ix g
W3 + ?
f o r t le t
fo rt
r a x ji
W1
to w n
C a s tle fo rd
BAl
BAl
7
BAl
+ s e ttle m e n t
BAl
v illa
BAl
DRl
Wl
305
TA3
WP2 +
TAl
SEl
Sifcs- KEattf
C o llito n
P ark
C o ls te r w o r t h
B ite
B a th
W4
v illa
C o m b e rto n
v illa
BAl
C o m b ie y
v illa
BAl
C o m p to n (1 )
v illa
BAl
C o m p to n ( 2 )
v illa
BAl
C o m p to n (3 )
s e ttle m e n t
C o r b r id g e
C o rh a m b ry
C o rto n
C o s g ro v e
C o tte rs to c k
BAl
to w n
SPl
BAl
fo rt
BAl
Wl
AQl -
fo rt
AQl
It
-
WPl
BAl
DR1 +
Wl
v illa
BAl
Wl
fo rt
BAl
W6
s e ttle m e n t
W2
D e s b o ro u g h
s e ttle m e n t
W3
D e s fo rd
s e t tle m e n t
Wl
D e w lis h
v illa
BAl
D i c k e t M ead
v illa
BAl
W2
v illa
Wl
to w n
D o r c h e s te r (1 )
c iv ita s
D o r c h e s t e r (2 )
c iv ita s
D o rn
Dover
D ow nshay Wood
D ow nto n
D r a y t o n (1 )
D r a y t o n (2 )
D r o itw ic h
s m a ll
WP2 +
W1 + ?
BAl
B A l?
DRl
BA 2
Wl
to w n
D u n c to n
v illa
D u nsby
s e ttle m e n t
D u n s t a b le
s m a ll
Durham
W2
v illa
v illa
DRl
AQl
v illa
D u n to c h e r
Wl
v illa
D u c k li n g t o n
TAl
W1 +
BA2
to w n
D r y h ill
W2 +
AQ l
s e ttle m e n t
s m a ll
TA2
AQ3
fo rt
v illa
fo r t le t
DRl +
BAl
Wl
BAl
AQl
Wl
It?
Wl
to w n
+ s e ttle m e n t ?
TAl
DR1 +
DRl
s e ttle m e n t
s e ttle m e n t
WPl
DRl
s e ttle m e n t
s m a ll
It
BAl
D o n c a s te r
DR1 +
AQ l
D o la u c o t h i
DR 2
DR1 +
fo rt
D e r s in g h a m
D id d in g to n
f o r t le t
v illa
TA2
W8
Wl
BAl
v illa
D a v e n try
D itc h le y
TAl
DR1 + +
Wl
to w n
v illa
D e rb y
Wl
v illa
D a re n th
D e n to n
DR1 +
Wl
BAl
s m a ll
D ean H a ll
BAl
v illa
s e ttle m e n t
D a ls w in to n
LPl
s e ttle m e n t?
C o w b r id g e
D a l g in r o s s
BAl
Cow R o a s t
C ro y H i l l
AQ2
v illa
v illa ?
Cram ond
S p rin g l i i i i ! ! ! ! $ ! ! !
BAl
C o v e h it h e
C ox G re e n
1 1 1 !!
DRl +
W1 +
v illa
s m a ll
ftra tn
Wl
s e ttle m e n t
Combe Down
C o n g re s b u ry
j l l l l l l l l
v illa
AQl
BAl
DRl
BAl
D u s to n
s m a ll
to w n
W1 +
E a r ith
s e ttle m e n t
W1 +
306
TAl
SEl
Sifc& JStejae
SxtB Type
Bafefe.
v illa
BAl ?
E ast C oker
East
Ils le y
E a s t e r H a p p re w
v illa
E a to n
BAl
s e ttle m e n t
AQl
v illa
BAl
E c c le s
v illa
BAl
E c k in g t o n
v illa
BAl
E ly
v illa
BAl
s e ttle m e n t
W4
v illa
W2
Em pingham
E p p e r s to n e
E w e ll
v illa
BA l
v illa
BAl
s m a ll
E w h u rs t
to w n
BAl
E x e te r
fo rtre s s
E x n in g
v illa
H u n g e rfo rd
F a r m in g t o n
F a rm o o r
BA 2
DRl
BAl
v illa
BAl
v illa
BAl
DRl
W3
WP1 +
v illa
v illa
v i 1 1a
F a r n h a m (3 )
s e ttle m e n t
F a rn in g h a m
v illa
BAl
AQl
LPl
BAl
Wl
BAl
DRl
v illa
F e lt w e ll
v illa
Fendoch
fo rt
F fr ith
fo rt
BAl
F ife h e a d
v illa
BAl
F illin g h a m
v illa
BAl
F in c h in g fie ld
v illa
BAl
F in d o n
v illa
Wl
fo rt
W2
F is h b o u r n e
v illa
F la m s te e d
s e ttle m e n t
F le tto n
s e ttle m e n t
LPl
Wl
s e ttle m e n t
F a w le r
F in g r i n g h o e
WPl
BAl
fo rt
F a r n h a m (2 )
F a rn w o rth
AQ2 ?
s e ttle m e n t
F a rn h am R o y a l
F a r n h a m (1 )
BAl
BAl
AQl
SP1 +
BAl
AQl
BAl
W1+ ?
DR2 +
WPl
DR 4
WP1 +
Wl
Wl
v illa
F o rd c ro ft
v illa
F o s c o tt
v illa
BAl
LP1 +
F o x c o te
v illa
BAl
LPl +
F ra m p to n
F r is k n e y
v illa
(fa ls u s )
s m a ll
AQl
F ro c e s te r C o u rt
v illa
F u n t in g t o n
v illa
BAl
BAl
TAl
to w n
s e ttle m e n t?
TAl
LPl
F o lk s to n e
F r ilfo r d
SEl
W6
v illa
F a lk ir k
TAl
DR1 +
Wl
s e ttle m e n t
E n g le t o n
SPl
BAl
v illa
F a r le y
DR2+- s t
W1 +
fo rt
E ls te d
E n fie ld
DR1 +
BAl
E b r in g t o n
E m b e rto n
Sewjr
W2
BAl
fo rt
E lg in h a u g h
Tank
SPl
Wl
v illa
E b c h e s te r
Sj>KX21-ff
s e ttle m e n t
E a t o n -b y -T a r p o r le y
eijpe
BAl
v illa ?
E a s to n
ftra x u
Wl
s e ttle m e n t
E a s tb o u r n e
s t t il
BAl
Wl
DRl
BAl
W2
DR 4
AQl ?
TAl
WPl
307
S it e - IJaiaa
G a d e b r id g e P a r k
A = g 'u e < 3 iK 2 t
Bafefo.
v illa
A Q l- I t
BAl
s e t tle m e n t
G a rd e n H i l l
G a rto n
J S ite TyjCH?
1i
Wl
BAl
s e t tle m e n t
s la c k
v illa
G atcom be
G a y to n T h o r p e
G e llig a e r
G e s tin g t h o r p e
AQl
BAl
v illa
c o lo n ia
s e ttle m e n t
G o d m a n c h e s t e r (1 )
s m a ll
to w n
AQl
ch
?
DR1 +
BAl
BAl
AQl
BAl
BAl
Wl
G o r in g
v illa
BAl
Wl
BAl
AQl
K im b le
G r e a t B u lm o re
G re a t C a s te r to n (l)
G re a t C a s te rto n (2 )
BAl
+ s e ttle m e n t
BAl
fo rtre s s (A )
v i 1 1a
fo rt
s m a ll
G r e a t Dunmow
BAl
v illa
G re a t C h e s te rfo rd (2 )
W2
BAl
AQl
It
BAl
v illa
BAl
v illa
BAl
G r e a t Tew
v illa
BAl
G r e a t T o th a m
v illa
G r e a t W itc o m b e
v i 1 1a
G r e e t w e ll
v illa
fo rt
Wl
BAl
BAl
v illa
BA 3
G r im s to n
v illa
BAl
on t h e
H ill
H acconby
Haceby
s e ttle m e n t
s m a ll
BA2
to w n
Wl
H a d s to c k
v illa
BAl
H a le s
v illa
BAl
H a lto n
C h e s te rs
v illa
AQ l
BA 2
fo rt
A Q l- wp
BAl
H a m b le d e n
v illa
BA!
H a m i lt o n
v illa
BAl
H a m p s te a d N o r r e y ' s
Hanham A b b o ts
W4
Its
BAl
H a ls to c k
Wl
W2
s e ttle m e n t
Wl
H a r d in g s t o n e
s e ttle m e n t
Wl
tow n
TAl
TA2
Wl
tow n
fo rt
Wl
v i 1 1a
s m a ll
SPl
WPl
s m a ll
H a rd k n o tt
DRl +
DR3 +
H ardham
H a r lo w
TAl
Wl
AQ2 -
+ s e ttle m e n t
Haddon
SPl
s e ttle m e n t
v illa
H a c h e s to n
DRl 3
BAl
A Q l?
G r im s te a d
G r in le y
TAl
Wl
G r e a t S ta u g h to n
F ie ld s
to w n
G re a t L in fo r d
G r e ta B rid g e
SPl
DR 2
BAl
to w n
G r e a t C h e s t e r f o r d ( 1)
G r e a t C h e s te rs
DRl
v illa
s m a ll
WPl
DRl
BAl
s e ttle m e n t
G re a t & L i t t l e
TAl
Wl
v illa
s e ttle m e n t
WPl
W1 +
v illa
E s ta te
S ew er
TA2
WPl
W2
G o rh a m b u ry
G ra fto n
LPl
Teuak
BAl
AQl
G o d m a n c h e s te r (2 )
G ra n d fo rd
DRl +
S-PCiHff
TAl
BAl
G lo u c e s te r
WPl
Wl
fo rt
fo rt
F ifH ?
DR2 +
v illa
G le n lo c h a r
:j& r a x r
W2
s e ttle m e n t
G a y to n
P llll
AQ2
BAl
BAl
H a r p o le
v illa
BAl
H a rp s d e n
v illa
BAl
H a r tfie ld
v illa
BAl
Wl
TAl
308
SEl
JSifc-e Typ^
Bsfch.
H a r tlip
v illa
BAl
H a v a n t (1 )
v illa
BA1
H a v a n t (2 )
v illa
H a ye s
v illa
Stfcs- Naawe
W ell
a n d R e ac h
W5
H ib a ld s t o w
s m a ll
tow n
W1
H ig h C r o s s
s m a ll
tow n
W1 +
s e ttle m e n t
DRl
fo rt
H igham F e r r e r s
+ s e ttle m e n t
BAl
v illa
BAl
v illa
BAl
H olc o m b e
H o ld itc h
s m a ll
H olm e House
H o lt
A Q l- s t ch
W1
tow n
v illa
f o r t le t
BAl
DRl
BA l
AQ l
EPl
H o r n c a s tle
s e t tle m e n t
W1
H o r n in g s e a
s e ttle m e n t
W1
fo rt
BA 2
v illa
BAl
H u c c le c o t e
v illa
BA l
H u n s b u ry
v illa
BAl
H o u s e s te a d s
H o v in g h a m
P ark
H u n tin g d o n
v illa
H untsha m
v illa
H u rc o t
v illa
Ic k h a m
s e ttle m e n t
Ic k le t o n
Ic k l i n g h a m
I fo rd
In c h tu th ill
In v e r e s k
Ir c h e s te r
Is lip
W2
TAl
TA3
BA l
I v y C h im n e y s
W1 +
BA!
W1
W1
AQ l
AQl
s e ttle m e n t
K e lv e d o n
s m a ll
K e m p s fo rd
s e ttle m e n t
K e m pston
s e ttle m e n t
K e n c h e s te r
s m a ll
K e s to n
K e tte r in g
tow n
WP1
BAl
K in g s W e s to n P a r k
v illa
K in g s c o t e
v illa
K in n e il
S in k
L a k e Farm
La m bourn
AQl
BAl
W1
BAl
W1
SP1 +
TA l
TAl
W1
AQ l
It
B A l?
W2
W1
BAl +
W1
BAl
W1
DR1
WP1 +
DRl
WP1
TA6
W3
to w n
v illa
K in tb u ry
W1
v illa
s m a ll
SPl
W3
to w n
Keynsham
DR1 +
BAl
v illa
Jo rd o n H i l l
TA l
BAl
tow n
s e ttle m e n t
SPl
to w n
fo rt
s m a ll
TA5
v illa
fo rtre s s
WP1 +
DRl
s e ttle m e n t
s m a ll
DR2+
W1
v illa
Ix w o r th
K ir k
AQl
v illa
Ilc h e s te r
DR1
W1
BA l
H ig h R o c h e s t e r
H in t o n C h a r t e r h o u s e
Sewer
BAl
tow n
H ig h L e gh
T&nk
TAl
DRl
s m a ll
+ s e ttle m e n t
SpaciiiSr
W1
H e y b r id g e
H ig h Ham
v illa
v illa
H e m sw o rth
t i p
BAl
fo rt
H a y to n
H e a th
BAl
W1
BAl
AQ l
BAl
SPl
DR 2
fo r t le t
DRl
v illa
BAl
v illa
AQ l
fo rtre s s (A )
AQl
BAl Sc2
DR1 +
SPl
?
TA2
v i l l a ( fa ls u s
309
SEl
sifc Jfeaae
Sit Jype
A gxseduct
fo rt
L a n c a s te r
L a n c h e s te r
fo rt
AQ3
st
Bath
W e ll
BA2
W1
ch
+ s e ttle m e n t
BAl
L a n g to n
v illa
BAl
L a tim e r
v illa
BAl
L e a C ro s s
v illa
BA2
L e a d e n W e ll
v illa
Landw ade
W1
W1
v illa
L e c h la d e
L e c k h a m p to n H i l l
W1
AQl
v illa
BAl
W2
DR1 +
TAl
BAl
W1 ?
DRl
TAl
wl
L e i c e s t e r (2 )
c iv ita s
BAl
L e in tw a r d in e
+ s e ttle m e n t
BAl
L e y to n
c o lo n ia
AQ2 ?
L in le y
s e ttle m e n t
AQ l
L in to n
v illa
DR1 + +
BA2
v illa
Dunmow
v illa
L ittle
P o n to n /S tr o x to n
v illa
Wl
L ittle
W a lth a m
s e ttle m e n t
Wl
P ark
L ittle to n
L la n d d e w i B r e f i
W2
tow n
v illa
BA3
v illa
BAl
BAl
v illa
BAl
L la n d o u g h
v illa
BAl
L la n fr y n a c h
v illa
BAl
L o c k in g
L o d d in g t o n
fo rt
BAl
v illa
BAl
v illa
BAl
BAl
L ondon
m u n ic ip iu m
(c iv iu m )
BA4
Long M e l f o r d
+ s e ttle m e n t
BAl
L o n g th o r p e
Wl
fo rt
BAl
v illa
BAl
L u llin g s to n e
v illa
Lydney
L y m in g e
Lympne
L y n c h Farm
Lyne
M a id e n h e a d
M a id s t o n e
M a ito n
AQ l
AQ l
Wl
BAl
W6
AQ l
BAl
DRl
TAl
?
DR1 +
TAl 5
WPl
TAl
BAl
AQl
BAl
W2
AQl
v i 1la
DR1 +
SPl
TA3
S E l?
BAl
v illa
fo rtre s s (A ? )
SP2 + +
DRl +
s e ttle m e n t
fo rt
DR2 + +
W ll
+ s e ttle m e n t
fo rt
SPl
BAl
fo rt
v illa
DRl
BAl
s e ttle m e n t
v illa
TA2
BAl
fo rt
Lunt
W55 +
fo rtre s s (A )
L u fto n
WPl
W7
v illa
Loughor
L o w e r S la u g h t e r
DRl
s e t tle m e n t
Low B o r r o w b r id g e
Low Ham
TA3
Wl
v illa
L o n g s to c k
DR1 +
W3
s e ttle m e n t
L oddon
L ong W itte n h a m
SE1 +
WPl
v illa ?
L la n tw it
TAl
BAl
AQl
L la n d d o w r o r
L la n io
SPl
BAl
L ittle
s m a ll
DR1 +
W2
L itlin g to n
L ittle c h e s te r
SEl
W2
s e ttle m e n t
L in c o ln
TA3
W1
BAl
fo rt
c iv ita s
L e ic e s t e r
L e i c e s t e r (1 )
L ittle c o te
DRl
DR 3
fo rt
L e a s e R ig g
: UraxH
BA2 ?
AQ l
Wl
BAl
310
WPl wd
SEl
fo rt
M a n c e t t e r (1 )
s m a ll
M a n c e t t e r (2 )
fo rt
M a n c h e s te r
Wl
AQ l
AQl
s e ttle m e n t
M a rs w o rth
s e ttle m e n t
M a ry p o rt
fo rt
M e la n d r a C a s t l e
fo rt
DRl
v illa
AQ l
BAl
Wl
BAl
Wl
AQl
Wl
s e ttle m e n t
Wl
v illa
W1 +
M u c k in g
M u m r ills
fo rt
M unthem C o u r t
N a n tw ic h
AQl
BAl
Wl
s e ttle m e n t
TAl
s e ttle m e n t?
N e a th
fo rt
N e a th a m
(fa ls u s
s m a ll
N e th e r W ild
W3
s e ttle m e n t
N a z e in g b u r y
Farm
to w n
N e th e rb y
fo rt
N e ttle to n
s m a ll
to w n
fo rt
N e w c a s t le
N ew haven
v illa
Newnham
v illa
N e w p o rt
v illa
N e w s te a d
fo rt
N o rd e n
Wll
BAl
e /w
AQl
AQl
BAl
It
?
BA l?
BAl
AQ l
W l07
BAl
N o r t h Mundham
v illa
BAl
N o rth
v illa
BAl
N o rth W r a x a ll
v illa
BAl
N o r th a lle r to n
s e ttle m e n t?
N o rth c h u rc h
v illa
BAl
v illa
BAl
DRl
Wl
W1 +
s e ttle m e n t
Wl
s e ttle m e n t
N o r t o n D is n e y
v illa
BAl
N o rto n S t.
v illa
BAl
v illa
BAl
Nunney
W3
Wl
s e ttle m e n t
P h illip
Wl
N u r s lin g
s e ttle m e n t
W1 +
Oakham
s e ttle m e n t
W2
O a k le y
v illa
W7
O a k r id g e
O d e ll
+ s e ttle m e n t
TAl
Wl
N o r th a m p to n
N o r th m o o r
DRl
DRl
BA 3
N o r th fle e t
Wl
BAl
L e ig h
Wl
BA 2
v illa
S ta in le y
TA 1+
BAl
AQ l
s e ttle m e n t
S tre e t
DR l +
BA 2
v illa
v illa
N e th e r a v o n
N o r f o lk
DRl
AQl
s e ttle m e n t?
M o u n ts o rre l
TAl
BAl
?
fo rt
M o re s b y
TAl
DRl
BAl
v ill.
N o rth
Wl
BAl
s m a l1 to w n
M ild e n h a ll
WPl
Wl
v illa
M e th w o ld
SPl
Wl
M a rs h fie ld
M id d le h a m
LP1
BA2
v illa
M a r k e t O v e r to n
TA l
v illa
M a n s f i e l d Wood H ouse
WPl
W3
to w n
Wl
BAl
W3
s e ttle m e n t
O diham
v illa
BAl
O ld Durham
v ill.
BAl
DR1 +
311
TA1 +
S x f c e - js fatftff
fo rt
O ld P e n r i t h
fo rt
O l d S o u th Eau
O r p in g t o n
O rto n L o n g u e v ille
iWdXH
Type
O ld K i l p a t r i c k
s e ttle m e n t?
BAl
+ s e ttle m e n t
BAl
Wl
O x fo rd
s e ttle m e n t
Wl
O x fo rd (2 )
s e ttle m e n t
Wl
Pagans H i l l
s e ttle m e n t
Pen L l y s t y n
P e n -Y -D a rre n
P ennal
P e n tre
F fw r n d a n Farm
BAl
+ s e ttle m e n t?
BAl
BAl
BAl
s e ttle m e n t
P in c h b e c k
Wl
AQl
TAl
DR1 +
BAl
BAl
s e ttle m e n t
P la x to l
SEl
wl
Wl
v illa
P o o le
P o rtc h e s te r
P o r tis h e a d
Is la n d
Wl
fo rt
W3
v illa
Wl
P u c k e r id g e
s e ttle m e n t
P u lb o r o u g h
+ s e ttle m e n t
fo rt
P u tle y
AQl
wl
BAl
BAl
AQ l
DRl
s e ttle m e n t
s e ttle m e n t
W1 +
R a d f ie ld
s e ttle m e n t
W2
AQl s t - l d
AQ l
WPl
s e ttle m e n t
R a v e n g la s s
fo rt
BAl
Wl
s e ttle m e n t
R e ac h
v illa
BAl
fo rt
BAl
Farm
R e ig a t e
R ic h b o r o u g h
R id g e w e ll
R ip p in g a le
R iv e n h a ll
R o c h e s te r
W1+ +
AQl
v illa
s e ttle m e n t?
v illa
s m a ll
to w n
v illa
R o c k b o u rn e
v illa
R o c k in g h a m
s e ttle m e n t
v illa
DRl
W2
v illa
Rock
R ode rs ham
W1+ +
s e ttle m e n t
fo rt
W2
Ra yn e
R e c u lv e r
TAl
DR2
Q u in to n
fo r t le t
TA1+
Wl
BAl
Q u e rn m o re
R a in h am
DRl
W1 +
s e ttle m e n t?
R a e b u rn fo o t
DR1 +
wl
s e ttle m e n t?
s e ttle m e n t
P u m s a in t
Wl
BAl
s e ttle m e n t
P re s ta ty n
TAl
WPl
v illa
P ia s Coch
WPl
BAl
v illa
s e ttle m e n t
TAl
Wl
P itm e a d s
R e d la n d s
WPl
WPl
BAl
P it n e y
P o r tla n d
s e ttle m e n t
P it s f o r d
TAl
W2
v illa
fo rt
SPl
DR l
BAl
fo rt
s e ttle m e n t
P ie r c e b r id g e
BAl
+ s e ttle m e n t
P e t e r b o r o u g h (1 )
P id d in g t o n
AQ l
fo rt
P e t e r b o r o u g h (2 )
S e v e r
wl
v illa
fo rtre s s (A )
Tank
AQ l
v illa
v illa
P a u lto n
S p r in g
BAl
O tfo rd
P a m p h ill
S lp e
Wl
DRl
BAl
TAl
BAl
A Q l-
It
BAl
BAl
Wl
DR1 +
BA2
W3 ?
DR 7
DRl
wl
BAl
312
SPl
Si fee- astaaia
R o e c liff e
+ s e ttle m e n t
R o t h le y
v illa
Rough C a s t l e
sil
BAl
Wl
BAl
R u d s to n
v illa
BAl
Wl
R u s h to n
v illa
BAl
W1 +
R u t h in
s e ttle m e n t
R y to n on D unsm ore
s e ttle m e n t
S a lfo r d
P r io r s
S a n d rin g h a m
S a n d w ic h
S an d y L ane
DR1 +
BAl
+ s e ttle m e n t ?
BAl
v illa
BAl
BAl
tow n
S a p c o te
v illa
BAl
v illa
BAl
+ s e ttle m e n t
BAl
Wl
S c o le
s m a ll
to w n
W7
S c u n th o r p e
s e ttle m e n t
W4
S ea b e g s Wood
S e a to n
S e ls e y
S e w in g s h ie ld s
DR2
v illa
v illa
BAl
AQl
v illa
S h a w e l1 /C h u r c h o v e r
v illa
BAl
S h e p r e th
v illa
BAl
S h ip t o n G o rg e
S h ir e o a k s
S horeham
S h o rtla n e s e n d
S ib e r to n
S id d in g t o n
S id le s h a m
S ilc h e s te r
S ix p e n n y H a n d le y
S la c k
S le a fo r d
S n o d la n d
S o m e r fo r d K e y n e s
S o u th S h i e l d s
S o u th w ith a m
S o u th w a r k
S o u th w a r k S t r e e t
BAl
Wl
v illa
BAl
s e ttle m e n t
Wl
v illa
Wl
s e t tle m e n t
DRl
s e ttle m e n t?
Wl
v illa
BAl
c iv ita s
BA 2
BAl
WP1 +
SP4 +
TAl +
Wl
v illa
TAl
BAl
+ s e ttle m e n t
BAl
AQ2
?
DRl +
BAl
v illa
WPl s t
TAl
Wl
fo rt
W3 8
AQl
BAl
v illa
BAl
v illa
BA 2
S p a r s h o lt
v illa
BAl
DR 3 +
WPl
DR1 +
WPl
Wl
TAl
WPl
s e ttle m e n t
v illa
DR1 +
Wl
tow n
s e t tle m e n t
(m a n s io )
Wl
fo rt
fo rt
SPl
DRl
W7 6
s e t tle m e n t
s m a ll
Wl
s e t tle m e n t
S o u th w ic k
S p o o n le y Wood
W2
s e ttle m e n t
S o u t h w e ll
S p i ls b y
S hakenoak
S h e p to n M a l l e t
Wl
BAl
MC34
SPl
DRl
f o r t le t
f o r t le t
TAl
BAl
+ s e ttle m e n t
S ca m pton
Scaw by
DR1 +
Wl
+ s e ttle m e n t
s m a ll
Tank
Wl
fo rt
BAl
S am bourne
Sppiw
Wl
s e ttle m e n t
v illa
R o w la n d s C a s t l e
PIPS
BAl
+ s e ttle m e n t ?
Rousham
CkrOkJETfl
W3
fo rt
R o m fo rd
R o t h w e l l H a ig h
Estk
.Sifc Typs
BAl
Wl
S p r in g h e a d
s m a ll
tow n
BAl
Wl
S t.
M a ry C ra y
s e ttle m e n t
BAl
S t.
S te p h e n
v illa
BAl
Wl
313
SPl
Site- i&Mae
asfek
Site Typs
s m a ll
S t a in e s
S ta in f ie ld
tow n
Wl
s e ttle m e n t
S ta m fo r d B r i d g e
s e t tle m e n t
BAl
Wl
S ta n c il
v illa
BAl
S ta n d o n
v illa
BAl
in
th e V a le
S ta n to n
F itz w a rr e n
S ta n to n
Low
S ta n to n
S t.
Jo h n
BAl
v illa
v illa
v illa
v illa
S t a n w ic k
v illa
S ta n w ix
W!
v illa
S ta n to n b u ry
Wl
s tib b in g to n
v illa
A Q l-
W12
ch
DRl
s e ttle m e n t
S to k e R o c h f o r d ( l )
D R' +
DRl
Wl
v illa
BAl
S to k e R o c h f o r d ( 2 )
v illa
BAl
S to k e R o c h f o r d ( 3 )
v illa
BAl
S to n e
W2
s e ttle m e n t
S to n e a G ra n g e
S to n e h am A s p a l
S to n e s f ie ld
s e ttle m e n t?
BAl
v illa
BAl
s e ttle m e n t
BAl
S to r r in g to n /P a r h a m
v illa
S to w e
v illa
S tra th g e a th
v illa
S t r o u d (2 )
v illa
S u tto n C o u rte n a y
S u t t o n V en y
Lea
S w a n w ick
S w in d o n
T a llin g t o n
T a r r a n t H in to n
T e m p le b o ro u g h
T e s to n
Wl
DRl
AQl
BAl
A Q l-
Wl
Wl
v i 1 1a
W2
v illa
fo rt
AQ l
BAl
Wl
BA 2
W5
BA l
v illa
s e t tle m e n t
Wl
T h a tc h a m
s e t tle m e n t
Wl
T h e a le
s e t tle m e n t
T h u rg a rto n
T h u r lb y
Th u rn h a m
T id d in g to n
TA2
W3
W10
BA l
TA1 +
W2
BA l
v illa
s m a ll
SPl
BAl
s e ttle m e n t?
v illa
DR1 +
DRl
T e w k e s b u ry
v illa
W2
s e ttle m e n t
v illa
TA2
It
s e ttle m e n t
T h e tfo rd
DR 2
Wl
s e ttle m e n t?
T h is tle to n
TAl
BAl
s e ttle m e n t
v illa
DR 4
T h e n fo rd
TA1 +
Wl
v illa
S w a ffh a m B u lb e c k
S w a lc liffe
v illa
DR1 +
BAl
s e ttle m e n t
S t r o u d (1 )
W13
BAl
BAl
fo rt
B r id g e
TA4
s e ttle m e n t?
S to k e O r c h a r d
TAl
Wl
BAl
s e ttle m e n t
WPl
Wl
S to k e G i f f o r d
TAl
WPl
BAl
v illa
S to k e D 'A b e r n o n
SPl
DR1 +
BAl
s e t tle m e n t
S to c k p o rt
S tre tto n
Wl
AQ l
v illa
S te p h e n M a l l e t
Wl
BAl
BAl
fo rt
S te b b in g
W1 +
BAl
v illa
S ta in le y N o rth
S ta n fo rd
Well ;Grain
BAl
+ s e ttle m e n t
S ta d e n
A Q l-
Wl
It
BAl
tow n
W14
314
DRl +
Sewer
i M
S i t e Ty
B ath
+ s e ttle m e n t
BAl
T in g w ic k
v illa
BAl
T its e y
v illa
BAl
v illa
BAl
T ils to n
T ix o v e r
T o c k in g t o n
v illa
P ark
T o tte rn h o e
A Q l-
T o w c e s te r
fo rt
fo rt
T re m a d o c
T u r f W a ll M ile
C a s tle
BAl
Wl
DRl
W7
DR1 +
T w y w e ll
s e ttle m e n t?
( b u rg i )
BAl
DRl
Wl
BAl
Wl
Wl
v illa
BAl
Wl
U p m in s te r
s e ttle m e n t?
Usk
fo rtre s s (L )
AQl ?
BAl
W1+ ?
DRl
V e r u la m iu m
m u n ic ip iu m
AQl ?
BA2 ?
W3+ ?
DRl 3 +
v illa
W a d fie ld
W a in fle e t A l l
S a in ts
W a le s b y
v illa
W a lls e n d
W a lto n H e a th
W a lto n -L e -D a le
W a lto n - o n -T h a m e s
W a l t o n - o n - t h e - H i l l (1 )
W a l t o n - o n - t h e - H i l l (2 )
W a nborough
(L o w e r)
W a n s te a d P a r k
W are
W a s p e r to n
W a tc h fie ld
W a te r N e w to n
Hanger
W e ld o n
BAl
s e ttle m e n t
W2 +
s e ttle m e n t?
BAl
v illa
BAl
to w n
BAl
Wl
v illa
wl
to w n
s e ttle m e n t?
AQl
It
s e t tle m e n t
W2
s e t le m e n t
Wl
to w n
W2
v illa
BAl
Wl
BAl
A Q l-
v illa
W e lto n W o ld
v illa
W elw yn
v illa
W e n d le b u r y
st
ch
TAl
BAl
Wl
s e ttle m e n t
W e ilo w
BAl
Wl
W2
BA 2
Wl
v illa
Level
W e s t D ean
W e s t D e e p in g
W est K e a l
W e s t N e w to n
s e t tle m e n t
DR4 0 +
v illa
BAl
v illa
BAl
s e ttle m e n t
v illa
v illa
DR +
DRl
v illa
v illa
We1 1 i n g b o ro u g h
W e s tb u r y
BAl
v illa
W e ll
TAl
DR1 +
BAl
s m a ll
SE2
SPl
BAl
s e ttle m e n t
W e e k le y
W e s t W in c h
Wl
AQl
fo rt
s m a ll
SPl
It
v illa
s m a ll
W a s h in g b o ro u g h
W e n tlo o g e
A Q l-
fo rt
W a lto n
W a te rg a te
s e ttle m e n t?
fo rtre s s (A ? )
W a lls
WP2 +
BAl
v illa
W a ll
TAl
It
f o r t l e t MC50
v illa
U pm arden
TAl
DR 2
BA!
to w n
v illa
Sewer
TAl
AQ l
A Q l-
T w y fo rd
Upham
Tscrak
BAl
v illa
s m a ll
T r ip o n tiv m
3$>riitgr
It
v illa
T r a w s -C o e d
i lir a in
BAl
fo rt
Tomen - Y - M u r
W ell
Wl
BAl
Wl
Wl
v illa
315
Naste
JS xt* Tyjxs
W e s te rw o o d
W e s to n U n d e rw o o d
v illa
W harram P e r c y
v illa
W h a r r a m -le -S tr e e t
W h ilto n
BAl
SPl
(fa ls u s )
s m a ll
to w n
W h i t c h u r c h (1 )
s m a ll
to w n
?
Wl
W2
BAl
W h i t c h u r c h (2 )
v illa
W h it e
v illa
BAl
v illa
BAl
S ta u n to n
W h itfo r d
?
LPl
v illa
C a s tle
TA5
W h ittin g to n
C o u rt
W h ittle b u r y
W h ittle s fo r d
BAl
fo rt
AQl
?
SPl
BAl
v illa
BAl
v illa
BAl
s e ttle m e n t
W h i t t o n (1 )
v illa
W h i t t o n (2 )
v illa
W ic k fo r d
v illa
W i g g in t o n
v illa
W ig g o n h o lt
DRl +
WPl
BAl
Wl
Wl
BAl
v illa
to w n
BAl
DRl
AQ l
W ild e r s p o o l
s m a ll
W ilm c o te
s e ttle m e n t
Wl
W ils fo rd
s e ttle m e n t?
W2
W in c h e s t e r
c iv ita s
W in d e rm e re
s e ttle m e n t?
w ingham
TAl
AQl
BAl
K in g s to n
v illa
W itc h a m p to n
v illa
W i t h in g t o n
BAl
DRl +
s e ttle m e n t
W in te r to n
Wl
AQ l
BA 2
DRl
v illa
Wl
s e ttle m e n t
W3
W o lfh a m c o te
s e ttle m e n t
Wl
W o lla s to n
v illa
v illa
W oodcock H a l l
BAl
v illa
W o o ls to n e
v illa
BAl
+ s e ttle m e n t ?
BAl
W o rc e s te r
s m a ll
s e ttle m e n t
to w n
BAl?
v illa
BAl
v illa
BAl
c iv it a s
Wycombe
+ s e ttle m e n t
(H ig h )
Wykham P a r k
Wymondham
W ym ondley
Y a r d le y
Y a r w e ll
(G r e a t)
H a s tin g s
s m a ll
tow n
W2
AQl
It
DRl
W17
BAl
Id
WP1 +
WPl
BAl
Wl
v illa
Wl
BAl
Wl
s e ttle m e n t
+ s e ttle m e n t
DR2 +
WPl
BAl
s e ttle m e n t
v illa
SPl
Wl
W r a x a ll
W ro x e te r
DRl
BAl
W o r t le y
Wyck
SPl
W1+ ?
s e ttle m e n t
W o rth
TAl
DRl
BAl
AQ l
W o o la s to n
W o o tto n H i l l
WPl
BAl
W itte n h a m
W o o d c h e s te r
WPl
W1 +
Wl
v illa
w in te rb o r n e
SPl
LP1
s e ttle m e n t?
W h it le y
W h it le y
BAl
L odge
W h it e b e e c h
DRl s t l n
BAl
v illa
fo rt
Tank
It
s e ttle m e n t
W h e a tle y
W hickh am
S$>rXi*g
U raxiJ.
BAl
A Q l-
fo rt
W harram G ra n g e
BAl
+ s e ttle m e n t
W etw ang
mil
Aguedttdfc
fo rt
BA l?
316
SP2 +
SPl
TAl
SE1 +
JSifee T yjxt
BSfefi
Y a tto n
S ite - Jteiae
v illa
BAl
Y e o v il
v illa
BAl
Y o rk
c o lo n ia
AQl ?
BA2 +
317
W e ll
W1 +
?PKxag
j&raxji
DR2 +
WP1 +
T an k
SE1++
1914,
147.
WSWHANS,
1973,
20- 2 2 .
20,
17,
fort
342.
Nash-Williams
established c.
granaries, bath-building.
AD
1969,
50,
45.
PNRB,
abandoned
c.
AD
TIRCGLL
150.
Walls,
(Silures).
20,
excavated well
is
1989, 296;
square,
stone-lined,
Fig.17,
2.8m deep,
p.335, the
and
with
timber
Pottery finds
suggestive of prosperity.
and
Field,
1991-2, 91-5,
leat. Fulford
1977,
35-84.
Scott
1916, 154; 81, 1991-2, 91-5. The Builder, 5th Jan., 1878.
Abridge, Essex.
Britannia 22, 1991, 261 timber-lined well.
84-6,
plan
Adel,
Leeds,
West
Yorkshire.
Simpson 1879.
YAJ 32,
229,
1934-6,
459;
34,1938-9,
231.
TIRBS
1981,
1987,
1,
major
RIB 629-633.
143-5.
possible
(Brigantes).
J.
30,
pl.l,
Artis'
Town,
1969,
1873,
127
-40 bath,
survey map.
17.
Scott
well,drain.
PNHAS 1921,
1993,
31,
Artis,
29.
1828,
14-5,
SMR 01583,
CA 2
(TL 10 97),
SMR
Ailsworth(2), Cambridgeshire.
Artis
1828
415,
Condron 1996,
823,
824,
located
415-6,
829,
square
stone-lined
well
at
Swan 1984,
799,
fiche
Normangate.
368,
816,
820,
'...several
c. AD 140.
Alcester, Warwickshire
Britannia
1, 1970,
121,
129,
183, ns.24,
1994, 272 defences. Booth 1980, 'Roman Alcester'. JRS 51, 1961, 172-3;
53,
1963,
134;
1966,
206;
57,
1967,
185;
58,
1968, 187-8. TBWAS 66, 1945-6, 35-48; 1958, 10-8; 1959, 27-32; 1963-4,
139-43.
236-7.
WMANS 18,
1975
(1976),
CBA
'2nd well, 2m deep, was originally lined with staves and later
with a box-framework. In the late 4th c. or later the well silted up'
(Brit. 7, 331).
TIRBS 1987,
with stone
flooding.
(Dobunni).
Alchester, Oxfordshire.
Britannia 1, 1970,
1987,
45,
118-9,
extramural
123;
settlement;
5,
1974,
23,
319
468;
1992,
6,
312;
1975,
20,
256;
1989,
18,
141-7.
Foster 1989.
1939,
281-8.
TIRCGLL
1983,
17;
31,
town
118-23.
founded
VCH Oxfordshire 1,
under
Claudius
(?) ;
(Catuvellauni).
1,
1970,
(Isuriam), 476.1
1975,
237,
40,
42,
47
Ant.
{Isubrigantum); 5,
'the
Helican
Itin.
1974,
pavement
465.3
(Isurium) , 468.3
416 cobble
mosaic';
17,
foundation;
1986,
76,
6,
'Roman
18,
development;
1987,
25,
373,
1994,
lead
sealing;
265 street.
21,
CSIR:
1990,
322
Great Britain
1.3,
ribbon
1983,
Nos. 15-17. JRS 52, 1962, 169. PNRB 1979, 379-80. There is a large and
deep Roman quarry below the civitas capital,
Medieval period. This quarry must have been a major source of stone
since Roman times and supplied other centres with
Romano-British period.
the stone,
as
it was
and to transport
the
stone to other sites. Smith 1852, pls.xviii & xxviii. Wacher 1995 (2nd
ed.),
for a
lined
There
77. TIRBS 1987, 2, RIB 708-10. Defensive earth bank built late 2nd c.
(22ha) ; stone wall added in 3rd c., bastions in 4th c. Decline late
4th c. There was a civilian settlement towards end of 1st c., and the
civitas capital was probably founded under Hadrian.
(Brigantes).
SMR
WZ
4,
SU 26 73
Aldbourne
Corse,
'substantial
1920-1,
389;
1968,
119;
1972,
173;
1973,
135;
1975-6,
134,
Hertfordshire
JRS 32, 1942, 112; 53, 1963, 136; 56, 1966, 209.
91, HT 5.
1.
320
6; 5,
1967,
2; 1970,
early 4th c.
3. TIRCGLL 1983,
{Catuvellauni).
Aldermaston, Berkshire.
Arch. Ex., 1976, HMSO. Britannia 8, 1977, 419, bath, well 4m diameter.
Current Arch. 54, vol 5, Jan. 1976, 220f. 'In the late 3rd c. or early
4th c.
an extensive
laid out.
One of
its ditches
163.
2nd Report,
1985,
456-57,
TIRCGLL
1983,
Black 1987,
112-57;
English
18,
76,
minor
247, bath-house,
1935,
Gower
Sc
120-
183-92.
'Alfoldean
settlement,
near
Posting
bridge
over
Station'.
river
Arun.
(Regni).
Airesford, Essex.
Antiquary 11, 1880, 36.
8.
5, 178.
a spring, which could imply a running water supply; 3(2), 1887, 136-9,
pls.v, vi, site plans.
18,
corridor villa.
TIRCGLL 1983,
(Trinovantes).
Alwinton, Northumberland.
Arch.
Ael.(2)
1877-80,
75-6,
'ritual
well
at
St
Ninian's
at
J.
103,
1947,
17-21.
1962,
1969,
Rowley
214;
(eds.),
67,
1975,
167,
192;
53,
128 camp.
215-23.
1954,
125-7,
1976, 323.
Notes
Arch.
1963,
1955,
1956,
JRS 47,
1957,
1965.
284;
210-11; 51,
131,
167;
54,
1964,
159;
57, 1967,
182;
58,
1968,
184;
Todd
Whitwell
1981.
1982,
183
Todd
Sc
in Rodwell
185,
reports
Sc
wells. TIRBS 1987, 2, 1st c. fort and later civil settlement on Ermine
Street at junction with King Street,
321
at SK 983 436.
(Corieltavi).
1971,
1982,
114, mosaic,
211,
175,
tile c. AD 75.
1973, 76-9.
82-107.
170,
p.233.
180,
184,
Cartwright
Gentleman's Mag.
1, 1832,
building materials;
1832,
Vol.2,
13,
172-3.
Cunliffe
SxAC 84,
82-107.
RAF APs,
Sux.
Collector's
Scott
Mag., 12,
1938,
1938,
SxAC 79,
3-44.
405-10; 16,
1942,
SxAC 86,
1-21.
Northamptonshire.
AASRP
1859,
5,
Supplementary
97-107.
Series
2,
Northamptonshire
1980:
Commentary
Archaeological
on
18
transparent
Whitwell
Atlas,
map
Smith 1868,
Appleshaw,
Redenham, Hampshire.
Britannia 17,
1986,420.
PHFC
9,
1920-5,
215-7,
82, NAR SU 34 NW 7,
item 5,
pi. opp.
HA 10.
in diameter;
49,
which is 3.19m
1956,
13-23;
52,
(10.5 ft)
1959,
69-72,
?pit/well.
(Catuvellauni).
322
Ardoch,
Porthshire, Scotland.
Note: No water-related structures. St. Joseph 1970, Britannia 1, 16378, 201; 2, 1971, 248; 3, 1972, 5, 7, 14, 16, 30, 43: Samian from; 9,
1978,
inscription;
25, 1994,
255,
fort.
Armoth, Trelissey, Dyfed, Wales.
Arch.
Camb.
1963,
192.
1964,
15.
121,
1972,
107.
Carm. Ant.
Scott
7,
1993,
BBCS 18,
1971,
56,
6.
1958-60,
295-303;
Nash-Williams
SMR 3 644,
20,
Part 2,
1954, 87.
DY 2, bath-house.
Savory
Wainwright
1967, 66.
4 Aug.,
1973 .
1993,
102,
KE
4, baths,
TR
32
59.
VCH Kent
3,
1932,
34.
(Cantiaci) .
1, 1970,
Kent.
Haverfield
1915,
25,
fed by springs'.
said
bath
building within the complex. A pond (pit 2) next to the bath; a drain
discharges
VCH,
Kent 3,
house.
1932,
103-4.
TIRCGLL 1983,
Scott
21;
1993,
102-3,
KE
corridor villa;
5.
bath
(Cantiaci) .
Ashdon, Essex.
Arch.
J. 10, 1853,
describes
Bowers' Farm,
4760, ES 4.
14-7,
Essex.
Britannia 24,
Copt
1993,
Hill,
3 02.
The first
a part
part
of Great
Scott 1993,
60,
SMR
EAA
No.5, 1977,
323
EAA Fig 5,
Mercury 24 Oct.
wells.
(40ft),
1874.
Scott 1993,
VCH Norfolk 1,
1901,
129,
295-6.
SMR 8712,
TIRBS 1987,
NF 4, two 1st c.
3,
at TF 908
057.
(Iceni).
c. 240-250.
196; 17,
Britannia 2, 1971,
181ff;
3, 1972,
147;
1994, 104.
6,
SyAC 37,
pts. I & II, 1927, 144-63; 38, 1930, 1-17, 132-48, Fig 1, p.146, bath
housedetached
from
dwelling
house,
storage
tank and
drains
(Regni or Atrebates).
Ashton, Northamptonshire.
Britannia 6, 1975, 253, stone-lined tank, probably for quenching;
8,
1977, 399, pl.29B, cylindrical lead tank, weighing over 3 cwt (152 kg)
with Chi-Rho monogram,
lined well;
274.
10,
1979,
Durobrivae,
1975,
153; 12,
Northants,
S-W,
3, 1975,
Current Arch.
12-5; 5, 1977,
56,
5, May
6-11.
1977, 210-11;
13,
4, 1982,
stone-lined well,
185,
1978,
442, drainage
181;
stone-
Sc
1976,
(1977),
Northants Arch.
14,
1979,
105.
10,
RCHM,
item bc(2-4),
at TL
VCH Oxon 1,
(Dobunni).
Atworth, Wiltshire.
Antiq.
J. 23,
stair- well;
1940-2,
1943,
5,
46-95.
148-52.
1974,
455;
TIRCGLL 21,
Britannia 3,
6,
1975,
278;
bath-suite,
1972,
7,
346;
1976,
c. AD 200,
4,
362.
1973,
317,
WANHM 49,
abandoned c. AD
400. (Belgae).
Sc
1980,
324
Sc
Dolby 1980,
145-64.
Note:
There
seems
to
be
some
confusion
about
this
site
in
the
(SN 41 20)
41,
(1-50
Cosmography1.
pls.I-X),
'The
Britannia 19,
1988,
for this
fort.
British
471-3;
Archaeologia 93,
Section
22,
of
1991,
the
281-2,
Ravenna
Fig 26,
plan showing Woodbury Fort and how Fosse Way Roman road was diverted
towards
it.
Moridunum
Peut.),
Ephemeris Epigraphica
(Ant.
'modo
Itin.
482,
vias huius
483,
(Eph. Ep.)
Rept.
87.06;
tractus
91.14;
flexibus
91.18:
at
'Moridunum,
site
near
Axe
vallium
645:
location of Moridunum,
implicatas
it
is
suggested
identified
as'.
that
For
Smith,
1993, 33-133,
PNRB
comments,
recte
estuary
discussion on the
'Statio
p.281
1913,
9,
(p.79), and
TDAS 17,
1885,
280;
54,
1922, 66-8.
Badbury, Wiltshire.
Arch. Review 1971, summary report.
Britannia 2, 1971,
346
found during
(uses
SU
195
810),
bath-house
282; 3, 1972,
excavations
for M4
24-29.
TIRCGLL
{Belgae).
c. 139m
and a
spring c. 139m (450 ft) S-W of villa, probably near the 182m (600 ft)
contour. Scott 1993, 68, SMR 450, GS 7, ?bath.
Baldock, Hertfordshire.
Applebaum, 1932, 244-58. Britannia 2, 1971, 269, 289; 3, 1972, 329; 4,
1973,
298;
7,
1986,
338;
11,
1980,
412;
12,
1981,
345,
385-6;
15,
1984, 304; 17, 1986, 400; 18, 1987, 327; 19, 1988, 457; 20, 1989, 188,
298-300,
Fig 22,
325
ditches and
wells,...'.
Arch.
86,
Selkirk
20-3.
Gomonde
1849, 7-13. JRS 45, 1955, 72, no.20, tile; 49, 1969, 221-2.. JBAA (2nd
ser.), 38, 1932, 235-69. Witts 1883, 60, no.10. Smith 1987, 212. Stead
1975,
125-9.
TIRCGLL 1983,
British period.
23,
occupied
1st c.
to end of Romano-
(Catuvellauni) .
1971,
122,
diet;
3,
1972,
7,
19,
43-5,
75, principia,
303, Antonine Wall; 5, 1974, 177-8, 201, 303, 405; 6, 1975, 227-8; 11,
1980, 48, fort.
clay pipes.
Hanson
1970,
358.
84-7,
Fig 54,
Robertson 1990,
215-6,
Miller
1922,
24,
1793,
pi. 35,
Fig.7.
1983,
74-6.
Macdonald
(6)
GAJ 4,
1934
(2nd
internal bath,
RCAHMS Lanarkshire
1976,
ed.),
99-102.
312-24.
1978,
114-7,
plan.
Roy
Wolverton, Buckinghamshire.
304, 306 Fig 19; 16, 1985, 290-3; 17, 1986, 399-401; 18, 1987, 326-7,
Fig.14;
14;
7,
1977,
79-85.
Milton
Arch.,
Keynes
Development
Corp.,
Records of Buckinghamshire,
Vols.15-17,
1985-87,
60-79,
Occasional
25, 1983,
171.
2 bath-suites,
small one at west side of house, and a large one at south end of the
house;
drains
forboth baths;
side
of
report
Materials',
Buckinghamshire Arch.
section on water
in
vols,
leat
caldarium,
Vol.l, 'Excavations
and Building
supply, 209-10,
perhaps
wooden water-pipe.
7.
See
TIRBS 1987,
5,
occupation 1st or early 2nd c., when burnt down and rebuilt. Occupied
into 5th c. Shrine at SP 825 405, of 4th c.
{Catuvellauni).
Soc. Newsletter
JRS 58,
1968,
199.
Scott 1993,
326
13,
SMR 214,
AV 5, AV 4,
bath-house.
VCH
1, 1906,
building; pottery
3 07-8,
item 14.
late 3rd to
TIRCGLL
1983,23,
bath-
10,
1979,
276,
'the
bath-house
and
latrine
block',
and
the baths collected water from furnace and bath-rooms; this and a 2nd
culvert from the intervalum road,
11, 1980,
18,1987, 21,
inside the
internal bath-house.
Neither source nor means of supply are certain but the discovery of
channelled stones in the vicinity of the water tank suggest that these
had either formed the water channel or, as at Balmuildy, had carried a
pipeline'.
Macdonald 1934,
Robertson,
et al,
Plate XV.
TIRBS 1987,
5,
at NS 707
RIB 2165-73.
1949,
104.
Scott
59-60;
Occupation
51,
1976,
1993,
109,
TLAHS 28,
58-9;
52,
NAR
1952,
1977,
60 NW
47;
BH,
32,
87-8.
LE
1956,
Whitwell
4, bath;
94-5;
1982,
50,
189.
1st to 4th c.
VCH Kent,
small
3, 1932,
(1.2m)
deep,
9.5ft
(2.89m)
(1.37m)
4ft
wide....Its walls
(0.3m x 0.4m)'.
47,
1973,
145-6,
pi.
XV
A.
Britannia
2,
1971,
180,
quarries; 5, 1974, 257, '...A square stone-lined well lies on the west
side close to the presumed position of the
front door,
of a small
house presumed to be near the front door, 2.5km W of the villa. JRS,
55, 1965, 74-89.
327
112.
TIRBS 1987,
(Catuvellauni).
1980,
1951,
135; 53,
1963,
143,
Current Arch.
164; 54,
352; 9,
1964,
171;
Figs, at pp.9b, 10, 11. It is noted that in the RCHM article figure, 3
wells are shown,
while
two wells
are
shown and two stone-lined pits, of which the previously mentioned well
3 is one. TBGAS 86, 1967, 74-87, Webster comments that the villa 'is
sited on almost
supplies
near by.
and
bath-house,
It
associatedwith
domestic
...the main
occupation,
were
(25ft)'.
380.
2nd
c.
building
and bath-house
associated with
domestic
into 5thc.;
99, 1982,
121-78.
Scott 1993, 68, SMR 1, GS8. TIRCGLL 1983, 24; built 2nd c., flourished
especially between AD 33 0 and 375.
c. 7.6m
22,
252;
Advertiser, 13-11-1973.
25,
20,
1989,
Advertiser,
Hadman
&
290
13
Upex,
434,
villa;
Nov.
Peterborough
1990,
Citizen
&
Northamptonshire.
bath-suite;
21,
1973,
1974,
224-6.
1994,
20,
1990,
news
27-8.
18,
332;
report.
1987,
324;
22,
1991,
Durobrivae
Northamptonshire
19,
1988,
252.
450-2;
Citizen
2,
1974,
27-8.
SMR
395,
1312.
Barochan, Strathclyde.
328
&
4th
Britannia 1,
1970,
224;
4,
1973,
275;
5,
1974,
144,
148;
16,
1985,
Barrington, Gloucestershire.
TBGAS,
n.d., MS Notebook of Dr
1976,
J Moore,
March 1882,
14-16.
RCHM
VCH Gloucestershire 6,
1965, 17.
Bartlow(1), Cambridgeshire.
Antiq.
J.
17,
463.
1937,
Arch.
Archaeologia 25,
J. 10, 1853,
Britannia 4, 1973,
252.
138.
157-8.
14, 17-21;
Fox 1923,
1834,
19,
1; 26,
1862,
24.
1835,
279; 21,
1913,
310,
1864,
1.
249-56,
at
32-3,
SMR
Scott 1993,
Bartlow(2), Essex.
Arch.
J.
10, 1853,
were opening
17-21,
opp.
p.18:
(9.43m)
'...we
reached
the water'. Well structure not described. The well was 0.91m from the
edge of the building on its N side. A 0.15m long piece of lead-pipe
still in place in the excavated wall
Frere & St Joseph 1983,
(p.17).
Dunnett 1975,
101-2.
TIRBS 1987, 5,
1975, 74;
6, 1975, 279; 7,
square built of
found a well 50 cm
a clay-lined
and
the
frames
recovered';
8,
earlier quarry.
base
of
1977,
of
the
oak(?).
419-20,
steining
well-hook
Fig.28
Current Arch.
rested
30,
shows
1972,
and
on
superimposed
bucket-binding
a well
332.
two
discovered
Miles
1986.
was
in an
Scott
4,
1973,
311;
7,
1976,
1-32
(Cunliffe) , pis
I-VIII;
8,
1977,
444, Roman reservoir; 10, 1979, 148; 11, 1980, 387-8, Roman reservoir;
329
12, 1981 370-81; 13, 1982, 396-406; 14, 1983, 336-7; 15, 1984, 333-6;
16, 1985, 199-200, water supply, 322-4; 17, 1986, 415, 430; 18, 1987,
341-3,
364-5,
369;
20, 1989,
178,
312-3, sewer;
21,
1990,
348;
126-7.
Hanson
22,
1970,
358,
comments,
Cunliffe
feeding
fresh water to the baths, in addition to the mineral water supply, was
discovered but apart from its existence nothing else is known'.
1979, 255-6.
Ross
TIRCGLL
24,
1983,
1967, 30ff.
shrine
of Celtic
goddess
SULIS;
1975,
temple
PNRB
131-8.
of
SVLIS
Avon.
Scott 1993, 13, SMR 1739, 1742, AV 11, bath, AV 12. Skrine 1871, 23,
bath.
300-1,
30.
1973,
201.
1970,
277-9; 4,
JRS 57,
1973,
1967, 179.
Yorkshire Arch.,
90,279,
334;
8,
5.
YAJ. 43,
1977,
96, 106.
1971,
24/06/1976.
381;
9,
Tyler
178-86;
TIRBS 1987,
45,
6.
(Pari si).
Bearsden, Strathclyde,
Breeze 1984, 32-68.
Dumbarton.
Britannia 5, 1974,
405-6, bath-house,
drain; 7,
1976, 1-32, 302-4 vicus; 8, 1977, 365, 433-4, inscription of Legio XX.
Macdonald 1934
1793,
pi.xxxv.
TIRBS,
1987,
Robertson
6, small
1960,
1990,
88-92.
fort unoccupied
Roy
in Antonine
period.
153;
bath-house,
2, 1971,
127,
141, bath-house;
Britannia 1, 1970,
6, 1975,
230:
'Agricolan
330
Current Arch.
1988, 217-74.
350;
10,
1979,
139-56;
Built during 1st half of 2nd c., but fell into disrepair during 1st
half of 3rd c. Appears to have been abandoned by c . AD 250. SyAC 29,
1879, 168.
(Regni).
Beckfoot, Cumberland.
Arch. J. 132, 1975, 28.
Bruce 1978,
(13th ed.);
idem. 1867, 365, plan at 536; idem. 'The Roman Stations in the West1,
Arch. Ael.(2) 5, 1861, 137-41.
aqueduct,
with
its unique
construction
Hanson
of
flag-stones
sloping
outwards over a hewn channel which carried the water in a wall, which
is raised towards the terminal end as it enters the fort. At p.359:
'The terminal
channel,
made
of channelled
stones, was found entering the fort near its S-E angle. The water was
delivered
Robinson
into
1880,
two
stone
138-9.
only
TCWAAS{1)
tanks
5,
fragments
1880,
of
which
136-48;
36,
remained'.
(2nd ser.),
the S-E corner associated with stone troughs. From his description of
it, I suppose it to have been the end of an aqueduct leading water to
the fort at that corner'
1987, 6-7.
(Brigantes).
19, 1988,
1990,
289; 23,
358-9;
SxAS News
22,
Letters
1991,
53,
1988, 4,
bath;
306;
54,
319
+ pl.xxii; 21,
24, 1993,
1988,
10;
307,
57,
Fig.21.
1989, 11,
Sc
Adkins
1986.
'excavation...relocated
(no.221),
and
showed
large ditch
dated
the
it
to
Britannia
bath-house
have
13,
first
undergone
at
1982,
discovered
least
one
375-6,
in
1871
phase
of
the
late
2nd
c.,
and
recut
at
least
331
1986,
409-10,
464; 22,
1991, 273.
PSA 5, 1870-3,
18,
149-50.
1987,
337;
19,
SyAC 6, 1874,
1988,
117-121;
VCH Surrey 4,
Beenham, Berkshire.
Berkshire Arch. J. 69, 1977-8,
1976, 221.
1-36, bath-house.
Arch. Ael.(A),
135-92;
52;
watertank,
54,
drains,
Current Arch.,
4,
19,
pipe,
1927,
1941,
1-43,
2 drains;
5,
1928,
pis. 1-5,
38,
Northumberland.
1960,
46-74,
plan
233-5.
mansio,
which
bath,
shows
Birley
tank,
1961,
bath,
163-5.
Breeze & Dobson 1976. Britannia 1, 1970, 150; 2, 1971, 127, 129-30; 3,
1972, 183, 193; 10, 1979, 280 vicus not located; 11, 1980, 319, 338-9;
12, 1981, 22, 24; 22, 1991, 234; 23, 1992, 112.
NCH 13, 1930, 521-7.
147.
1-
43, comments, a possible aqueduct water supply from Denton Hill Head,
from a spring, 3 miles away.
Fig.2
TIRBS 1987, 7,
1352.
(Brigantes).
was
the upper 3m to 3.6m (10 to 12ft) revetted with regularly laid blocks
of chalk and
Mansell-Pleydell
Winterbourne
Kingston
in
1888
in
Parish'. There
"neighbouring
appears
to
have
field"
been
in
Roman
103-5; 86, 1963, 100; 87, 1964, 111; 88, 1966, 116. Warne 1853, 85-92.
10,
water holes.
1979,
Original
302-3,
well,
site dates
wattlewood-lined
and
number
of
Bewcastle, Cumbria.
332
Birley 1961,
1978,
(13th
38,1938,
Gillam,
120,
231-3.
ed) .
CW(2)
195-237;
49,
and
functioned
22,
1922,
1949,
216-8
1993.
Dating
to c.
early
19, 1988,
169-85;
baths;
31,
54,
474. Bruce
1931,
1954,
137-40;
265-7
baths.
to mid
4th
c.,
dated
by
c. AD
coin
of
Constantine the Great AD 309-310, and a pot rim dated c. AD 360. The
drain consists of stone slabs (Fig 5, Sect. 7), and is 0.84 m wide and
0.1 m deep. The fort is on Hadrian's Wall and the bath is within the
ramperts.
997.
(Brigantes).
Bibury, Gloucestershire.
Britannia 16, 1985, 298: '...a stone culvert, 0.6m wide and 1.07m high
running S-W to N-E on the west side of the site. A length 12.5m was
traced...;
19,
Vol.I, 282.
12.
1988,
RCHM
465.
Glos.
Glevensis 21,
1976,
14-5.
1987,
42-5.
Scott 1993,
Gough
68-9,
1789,
SMR 366, GS
Biddenham, Bedfordshire.
AASR 4, 1858, 283-90, well, stone-lined.
4.
Biglis, Glamorgan,
Wales.
hardstanding,
upon
which
stood
stone
plinths
probably
21,
1990,
occupation.
306.
Dating
Summary
of
from c.
2nd c.
excavations
in
BC to end of
Glamorgan-Gwent
1975.
Archaeologia
18,
1817,
203-21;
19,
1821,
176-7.
Black 1987, 153. He suggests an earlier bath to late 1st to early 2nd
c. nearby.
6; 5, 1974,
20, 1989,
457;
17,
1986,
421-3;
320-1;
22,
1991,
189-90:
24,
1993,
continued excavations,
Frere
but
SxAC 67,
1987,
351-3;
19,
25,
still
1994,
no
288;
with pls.vi-xiii.
333
84-88;
26,
indication
1988,
478-9;
307;
1982,
18,
121,
1983,
1995,
370,
of a water
Herbert
1927,
Oxford J. Arch.
203-8.
Scott
Steer n.d.,
(timber building),
'The Letters
of John Hawkins
replaced c. AD
and
Samuel
and
March 1995.
{Regni).
Billericay, Essex.
Britannia 2,
1971,
clear whether
1988, 457;
Rowley,
27.
331; 3,
from
1972,
Roman period;
331;
4,
8,
1977,405; 9,
Rivet 1964,
1973,
330,
well,
but
1978,
not
449;19,
Rodwell &
TIRCGLL 1983,
(Trinovantes).
40 Ant.
1977, 229,
1978, 425-6,
379;
9,
Itin.;
2,
477;
1971,
127-8;3,1972, 355;8,
347-9;
11,
1980, 361, 367, 409; 12, 1981, 327, 383; 18, 1987, 318; 20, 1989, 277;
22, 1991, 238 final bath in praetorium,
bath dates
from c.
2nd
c.
Hooppell
1891, 'Vinovia'.
253-87.
pis.5, 6, vicus.
1988, 343-4.
and
final
JBAA 43,
1887,
(eds.), 1979,
TAASDN 10,
constructed
111-23,
233-54,
1953 , 363 ;
11,
c.
299-306;
plans.
1958,
AD
46,
3 60.
1890,
Sommer 1984,
115-27.
Wilson
{Brigantes).
Britannia 16,
1985,
327; 20,
1989,
319;
22,
14, bath.
Birdoswald, Cumbria.
Arch.
Ael(l)
4,
1855),
63-75,
141-9;
19, 1988,
436-7; 494;
1990,
317, 365;
22, 1991,
296-7; 23,
1993,
284;
ed.),
198-206.
345-76;
205; 31,
120-30.
1994,
298.
TCWAAS(l)
20,
1989,
1992, 315-6,
Bruce 1867,
14,
Ael(2)
1974,
25,
Arch.
1897, 413-6;
4,
274-5 drains;
318-9,
365;
32,
1932, 141-5;
334
33,
1933,
172-90,
246 62;
21,
24,
1978, (13th
197-210,
249.
4, 1973, 152; 5,
(3rd ed.),261;
15, 1899,
1860,
169-
34, 1934,
item 31 he comments,
stone channel
This
261).
(Bruce
(de architectura,
1892-93,
166).
The shape of
the channel
suggests
the
is
undated'.
(Bruce).
Water was
TIRBS 1987, 8,
filtered
through
RIB 1872-1929 .
charcoal
in
the
tank
(Brigantes).
This
probably fed a distribution tank behind the gate, from which water was
distributed
(Richmond
in
pipes
1937-38).
probably
Another
wooden
channel,
held
in
stone
"large drain"
of
casings
triangular
be
(St Joseph,
re-interpreted
as
1951,
57).
casing
of
The
a
triangular
wooden
pipeline
67 n .96, 201; 3,
channel
that
Britannia 1,
well
in fort; 4, 1973, 317; 5, 1974, 154, 159; 11, 1980, 19; 16, 1985, 326;
23, 1992,
317-8.
Bruce 1978,
Bath
Fig.13 p.298.
are of wood,
1897. Miller
1952,
1793. RCAHMSDumfriesshire
(ed.),
1920,
Fig.19 p.304.
100-6.
St
Joseph 1951,
57.
Bitterne, Hampshire.
Antiq. J. 27, 1947,
103,
n.48,
86,
118,
70 Ant.
192-3,
It.
materials; 4, 1973,
393; 22,
1991,
287,
Fig.70,
478.1
small bath.
Black 1995,
Britannia 1,
1970,
(Clausentum XX);
2,
1971,
169-94,
10,
1979,
161-2;
stone-lined well;
23,
1992,
302;
25,
93,
49-50,
building
11,
1980,
1994,
287.
335
(SU 474
132),
or at
port near mouth of river Itchen; dates from Claudian period to 4th c.
(Belgae) .
1993,
Leicestershire.
remains
of a bath.
Bitton, Avon.
Britannia 1,
1970,
facing p.106
and p.119.
7, ST
58-60,
Ant.
Itin.
Scarth 1864,
125-6,
see maps
Bledlow-Cuzn-Saunderton, Buckinghamshire.
Britannia 5, 1974, 254, 258; 20, 1989, 298.
28, 1938, 185; 29, 1939, 210-11, 288-9; 55, 1965, 88.
19,
67-8.
large
villa
with
no
water
398-426; 17,
1965,
supply.
410;18,
Branigan suggests a
Scott 1993,
25,
SMR
PSA 31,
in
JRS
1918-
Records
1969,
261-76:
the villa,
6, SMR 0498,
of
272.
BU 7, SMR
0878, BU 8, SU 79 99.
Bletchingley, Surrey.
Scott
of
1993,
Surrey',
Camberwell.
MS
letters
1917,
and
plans
188-90, Appendix:
at
(n.d.)'
Minet
'Roman Baths',
History
Library,
item 9,
p.189.
Blisworth, Northamptonshire.
BNFAS 4, 1970, 38.
West 4, 1982, 17-19, items 3 (some wells SP 735 530) & 7, Fig.32. VCH
Northamptonshire, 1, 1902, 216.
Blyborough, Lincolnshire.
Directory of Lincolnshire,
(3rd ed.),
1972,
15.
Dudley
1949,
181.
1994,
Esssex.
279-80;
26,
1995,
359,
farmstead
century date, was examined, when structural remains (4th c.) of aisled
building 12m by 4.5m were found associated with field boundaries and
trackways, and a boxed cremation of the 2nd c. The trackway leading to
a large enclosure c. 140m square and to a field...., its 4th c. timber
successor 15m by 28m, a bath-house,... Two Roman wells, a pond, and a
cremation burial'.
Borough Hill, Northamptonshire.
Condron 1996, 322.
Northants.
Northamptonshire SMR
Woodfield 1978.
1975,
'decorating cover
20-35;
7,
1976,
3 04-5,
Fig.8,
feeding a drain
outside the east wall....The cold plunge bath drained by a lead pipe',
(pi.XXVI A);
phases of construction,
ended in destruction.
172-87.
both
Miller 1942,
Bottesford, Lincolnshire.
Arch. J. 91, 1934,
NE 8, LI 18.
Boughton, Northamptonshire.
BNFAS 4, 1970, 16;
item 5 no.6,
5, 1971, 22.
Fig.24.
Well,
RCHM Northants,
located south
of
N-W, 3, 1981,
Boughton
14-6,
Grange, on
OD.
Boughton Honchelsea,Kent.
Archaeologia 29, 1842, 414-20,
96, coin of Adminius.
Bourne, Lincolnshire.
AASRP 22, 1989-90,
10.
Society 76,
Vol.2, 3 00.
1883,
bath.
Bourne/Morton, Lincolnshire.
Whitwell
1982,
Stukeley,
Surtees
198, aqueduct/leat,
Stukeley,
Whitwell 1982,
198.
354;
13,
1982,
Gloucestershire
walls, wells,
377;
14,1983,
& Cotswolds,
shallow pits,
314;
17-20,
20,
Fig.,
1989,
p.18,
234.
RCHM
structures
TBGAS 56,
1976,
include
1934,
99-
Itin. 468.1 (Lavatras) ; 2, 1971, 251; 3, 1972, 41-2; 11, 1980, 18; 12,
1981, 327;
16, 1985, 203, 325; 19, 1988, 491; 20, 1989, 277.
1970,
comments,
3 59,
(modern Levypool,
'It brought
NY 9675
1535),
CSIR 3, 1983,
the Flavian
water
to
the
fort
Hanson,
from Laverpool
of
the
fort'.
about the levels along this leat. The leat is not dated, but probably
dates from time when the bath-house was restored
during Roman times discharged into the pond next to the bath, which
possibly could have been the water supply for the bath.
119.
5.
Scott
Scarth 1864,
1930-2,
60, 104, 422; 63, 1968, 109; 64, 1969, 123-4; 66, 1971, 194, 197; 69,
1974, 185.
Cant.
15,
1883,
104-7.
Britannia 1,
1970,
Payne 1883,
3 04, Romano-Celtic
105.
Payne 1893,
KAR 17, 1969, 32; 18, 1969-70, 9, well. PSA 9, 1882, 162-3, a
very long building 217 ft, 50 ft wide, with many rooms, but no trace
or record of a hypocaust system.
Fig.23.
(Cantiaci).
338
106-9,
item 10,
Bozeat, Northamptonshire.
Beds. Arch. J. 3, 1966, 3; 5, 1970, 57-65.
1965, 210; 56, 1966, 207.
RCHM Northhamtonshire,
10, SP 906 587.
BNFAS 6, 1971, 3.
Central,
II,
1979,
3-5,
JRS 55,
77-82,
drains.
drains.
TIRBS 1987,
98;
90,
1968,
171;
91,
1969,
189,
the
PDNHAS 76,
structure
of a
final
phase
the
villa
proper
was
constructed.
small
During the
The
site
was
Bradford-on-Avon, Wiltshire.
Scott 1993, 198, SMR SW 3 09, WZ32, bath.
J.
23,
1943,
153.
(Ashmolean Libr.,
Aspinall-Oglander
& Oglander
(rev.
comment
bath,
ed.)
1)
'But
and a well-
house...', A figure shows the plan of the villa with a small bath and
the position of the well (p.10). 'The well itself - the superstructure
of which is a reconstruction - was found to be 30.5m (100ft) deep;...'
(p.25). 2) the 'semi-circular fountain, the bath-house, and the well,
were covered in again after being excavated'
53.
Rivet
1969,
43-4.
1987, 19-28.
Scott
1993,
101,
SMR 1017,
Tomalin
(Belgae).
1970,
177-252. Leech,
et
295-6;
al,
2,
1971,
276; 4,
1981, 177-252,
1973,
310-11;
12, 1981,
and
and a
a cistern
stone-lined
drain with
p.188.
339
in
building
3,
itemF109,
Braintree, Essex.
Britannia
6,
1975, 262;
457,
9,
1978,
449,
(TL
settlement;
20, 1989,
302,
7685
1985,
2310
241-2, 295,
a timber-lined
Essex 3,
30,
major
Blyth's
327; 19,
well.
at
Essex
147. VCH
settlement
on the road
(Trinovantes).
Braishfield, Hampshire.
Antiq. J.
PHFCAS 41,
house, c.
1st c. to c. AD 370.
Scott
1985,
69-80, bath
19 .
Brampton, Norfolk.
Britannia 1, 1970,
1973,
300;
1977,
213-6,
wells,
6,
19, 27,
1975, 260,
209-11,
405;
9,
1978,
448;
11,
1972, 330,
7,
1976,
1980,
375:
well; 4,
209-16;
'(a)
8,
Three
1.15m long and 0.25m wide. A clay-lined cistern lm by 1.15m deep and
probably dating to late 3rd or 4th c.';
307-8;
15,
1984,
306;
16,
1985,
294;
13,
18,1987,
369-70;
44, 330.
EAA
1982,
14,
1983,
CBA Norfolk
5, 1977, 31-95.
JRS
Scott 1993 ,
(Iceni).
53,
1979,
51-5.
Black 1995,
26-7,
62,
139,
Figs.20a,
b.
Britannia 13, 1982, 360; 15, 1984, 294, Fig.13; 17, 1986, 393, Fig.16
shows a drain about 57m long, traced for 45.7m
(p.294)
on side of a
11, 49-92.
TIRBS 1987,
1949,
32.
Britannia 1,
1970,
118,
120,
123,
127,
313; 2, 1971, 180, 188, 192; 4, 1973, 299; 5, 1974, 437, pi, xxxviii
B, 3-roomed bath-building 22m x 4.5m and was abandoned during the 3rd
c.;
10,
1979,
349.
Burnham
&
(eds.), 1975,
Figs.27,
222,
occupation
28.
HT 19.
through
to 4th century, and probably later. There was a villa at Mentley Farm,
north of the
small
town at
evidence of a street
the
system.
A bath-house,
dating
where
there
is
period (c.AD 70-90), went out of use in early to mid 2nd c., situated
in the bend of the River Rib, presumably its water supply.
TIRCGLL
1983, 30, major settlement on the road between London and Durovigutim,
Godmanchester; occupied 1st to 4th c.
(Catuvellauni).
257; 2, 1971,
112-3;
14, 1983, 280; 16, 1985, 183-4; 18, 1987, 10; 22, 1991, 224; 25, 1994,
246.
Hanson 1970,
396.
Nash-Williams 1969,
48-51.
9;
together
fort
occupied
c.
AD
75-200,
with a
PNRB 1979,
307.
(Silures) .
1938,
267-93.
JRS 28,
1938,
Britannia 20,
1989,
1972, 78-85.
Scott
1901.
183.
Poulton
1993,
14,
SMR
Branigan 1969a,
18,
Fig.3, well.
Sc
1390,
AV
16.
PSANHS
VCH Somerset
116,
1,
1906,
3 03-5,
item 9,
(Belgae).
Bristol(1), Avon.
Britannia 26, 1995, 366 stone-lined well.
302;
21,
1990,
348,
ST 5870 7339.
1993,
Smith
TBGAS 69,
1978,
1950,
139.
Scott
5-58.
Villa
JRS 51,
1993, 269.
1961,
130,
bath-house,
cistern,
slate-lined.
Nash-
(Ordovices).
Brixworth, Northamptonshire.
BNFAS 2, 1967, 7-8.
1978,
371;
16,
1985, 324;
Newsletter 2, 1972,
9.
1966,
207;
186;
2968,
3009.
57,
1967,
22,
1991,
252,
6914.
CBA G r .9
1968,
1981,
26-31,
VCH Northamptonshire
1,
JRS 56,
SP 7498
9,
1902,
Scott
200.
item 16,
TIRBS 1987,
11,
small
5-
roomed house built late 1st c., partly rebuilt late 2nd c. or early
3rd
c.;
additionsincluding
bath-house
at
beginning
of
4th
c.
(Catuvellauni or Corieltavi).
320;
5,
1974,
457;6,
1975,
282.
This
site was
occupied from 1st c. to 4th c., and had 36 iron smelting furnaces, a
large industrial enterprise,
settlement.
{Regni).
6, 1859, 260; 35, 1907-8, 441; 45, 1930-2, 178-9, bath; 72-3,
180 well (?) ; 74-5, 1979-80, 205. At ST 96 66 WZ 37
1977-8,
(Scott) and
WAM
ST 97
(Belgae).
Bromham(2), Bedfordshire
Bedfordshire Arch. J. 6, 1971, 84.
Newsletter, 1972, 12.
155,
279;
3,
1972,
192-3;
59, 1969,
41-2 occupation,
destruction,
506-7.
1983, 113Stephens
pipeline'
(Wade,
1952,
Soc.
1, 1928,
261-84;
8, 1955,
153-6.
Proc. Leeds
56,
c.
CSIR 1.3,
1983, Nos.
106-7.
TIRBS 1987,
4 & 12,
RIB 722-724,
{Brigantes) .
Brough-on-Humber(1), Humberside.
Hanson 1970,
360,
item 9:
formed by a wooden water pipeline entering the fort by the West Gate.
Its source may well have been the upper reaches of on of the streams
that run down into the Humber to the west of the fort'
16);
item 10:
(Wacherl969,
PNRB 1979,
437-8,
443.
2km (1.25
miles)
to the
Stephens 1985b,
227,
'Gullies found in both the Vespasianic fort and the later naval base
may have been pipe trenches'.
RIB 707
(aediles
supervenientium
Flavian
fort
Wacher 1969,
vici Petuariensis) ;
Petueriensium);
c. AD 70-80;
Not.
?Itin.
stores
depot,
41-2.
Ant.
c.
TIRBS 1987,
466.4
AD
12,
(Pf. numeri
(Praetorio)?.
80-125.
Late
2nd-c.
(Parisi).
Brough-on-Humber(2), Humberside.
Britannia 24,
defences
present.
etc....,
at
1993,
SE
287;
9410
26,
2675
1995,
347 '(a)
considerable
evidence
agricultural
period'.
Brough-on-Noe, Derbyshire.
activity
activity
was
through much
of
the
Roman
432;
10,
1979; 293;
11, 198,
318-20;
15, 1984,
1974, 24;
289,
16, 1985, 282, 326; 18, 1987, 321, 370; 22, 1991, 244-5.
vicus;
Dearne 1993,
The site lies on a low spur of land 170m (545ft) OD, in a bend of the
river Noe,
near
the
confluence with
the
Bradwell
Brook.
1st phase
Flavian c. AD
i.e.
Derbyshire Arch.
4th c.
1965,
Fig.l,
123-6;
86,
99-101;
59, 1938,
87,
1967,
53-65, Figs.1
154-8;
88,
& 2; 85,
1968,
89-93,
1968',
drain;
water
channel
represent the
29,
1966,
J.
1939,
206;
89,
1969,
99-106.
Hanson 1970,
360,
tank
1940,
168;
42,
1952,
204;
the
fort
Hart
44,
could
1981,83-7.JRS
1954,
108ff,
full
account; 49, 1959, 108; 50, 1960, 216; 53, 1963, 160; 56, 1966, 201-2;
57, 1967, 181; 59, 1969, 211, Fig.33, aqueduct.
23 0, 'The fort was supplied by a stone channel pasing through the s-w
gate.
This
fed a distribution
showing that
it was
level,
(Taylor &
- probably
wooden in view of the size of the hole - although water may have been
distributed either
in stone channels,
or
in pipelines
supported by
such channels. The aqueduct dates from Phase II, c. AD 158 (RIB 283)'.
TIRBS 1987, 12, RIB 281-283, 2243 (A NAVI ONE MP XI). (Brigantes) .
Bryncross, Gwynedd, Wales.
Britannia
24,
1993,
269,
two
round houses,
with
external
drainage
gully, and the 2nd house had a stone-lined drain internal to a number
of concentric wall slots.
Robertson 1979,
stone-lined culverts
apart,
2 drains,
12.5m
roughly at
its
12,
1981,
359 bath;
13,
1982,
384;
15,
1984,
318-20,
bath,
water supply, Fig 25, drain; 16, 1985, 306; 17, 1986, 417 'An Iron Age
round house together with pits
underlay
this
group of buildings'; 18, 1987, 345-6; 19, 1988, 474; 20, 1989, 314-5;
21, 1990, 350;
98,
1976,
54; 99,
1980, 88;
1977,
103,1981, 88;
120;
100,
104,
1978,
112; 101,
1979,
133;
146; 106,
1984,
116-7; 107,1985, 164; 108, 1986, 181; 109, 1987, 129; 110, 1988,
2; 111, 1989, 107; 112, 1990, 117-9;
113, 1991,
173-4.
102,
151-
Scott 1993,
250, with a bath-suite, lying to the east of the main range of rooms'.
Buckton, Hertfordshire.
Black
1995,
certainly
1979,
were
118,
have
21,
lists
official
1983, 98-102.
mansio,
which
found passing
represent
an
360,
almost
'Remains of a
would
wooden pipeline
fort
and could
Joseph
JRS 50, 1960, 222; 51, 1961, 123-4; 52, 1962, 169; 67,
Nash-Williams 1969,
93-4.
Stanford 1968.
TWNFC
Bunny, Nottinghamshire.
EMAB 9, 1966, 38, pottery found dating from c. AD 250-3 00; remains in
very poor state;
no buildings found.
1967,
Alvery
1967;
72,
Well 12.2m
(40ft)
deep,
0.76m
42-9.
Whitwell
1982,
203,
well.
Burgh, Suffolk.
Britannia, 2, 1971, 172, 188, 191; 6, 1975, 261 village; 7, 1976, 341;
10, 1979, 307; 18, 1987, 46, vicus; 25, 1994, 305, lead sealing.
40, 1988, 73, bath.
Moore 1988.
Johnson 1975.
EAA
BUG 002, SU 4.
Burham, Kent.
Arch. Cant. 34, 1920, 155. Britannia 2 1971,
Scott 1993,
103,
KE 16.
VCH Kent 3,
1932,
Burrow-in-Lonsdale, Lancashire.
Arch. J. 142, 216-236.
418.
Bruce 1851,
Britannia 1, 1970,
TCWAAS 54
Burton, Lincolnshire.
Arch. J. Ill, 1954, 106-28. Settlement, ?water-pipe, ?aqueduct.
Buxton, Derbyshire.
Britannia
JDANHS
25, 1903,
161;
254-5.
PNRB 1979,
of natural springs,
cold,
one
St.
3, 1972, 7,
12, 19;
8,
1977,
3 stone-lined drains.
365,
Clarke 1933,
10,1979,
62-3. Hanson1970,
3 60, comments that there is no known water supply, but that the supply
could have come from 'the east, and the existence of a spring cased in
Roman masonry was noted
312.
3,
Fig.52,
fort.
2 baths,
(2)
fort,
41,
297-
(6)
external
to
Macdonald 1934,
Robertson 1990,
JRS
Wales.
Britannia 1, 1970,
6, 1975,
Wooler 1917,
54;
190.
94;
7,
CSIR 1.5,
(A KANOVIO M P VIII);
{Ordovices) .
107.
1-7. Arch,
Itin.,
484.4,
51-2.
484.10,
2,1971,127,
1976,
1979,
279-80;
273-4;
11,
8,
130 n.54,
1977,
1980,
198,
351;
12,
202-6,
263-6;
9,
1981,
288-9,
317,
5, 1974,
1978,
379,
147;
409-10;
10,
395-6;
13,
1982, 334, 420-1; 15, 1984, 260, 337-41, 348; 16, 1985, 257-9, 262 Fig
9, 324; 17, 1986, 367-9; 18, 1987, 305 Fig.4, 307, 377; 19, 1988, 1804, 421-2,
the colonade in brick; a colonade with drain was also now added along
the side of the basilica....; an original lead water-main was replaced
and the street replaced. At the end of the century the drains became
filled, the porticoes were dismantled and the lead pipe dug out; new
metalling
of
Valentinian
the
street
and
the
porticoes
after
replaced by stone
AD
33 0;
from c.
AD
coin
of
74-75,
but
8-9.
1 internal,
Hanson
Glamorgan-Gwent Arch.
197 0,
2 wells,
180-1,
spring,
aqueduct,
Stephens 1985b, 223, 229, 'Wooden and lead piping dating from c. AD 80
has been found near the amphitheatre (Wheeler & Wheeler 1928, 144, pi.
xx). They imply the existence of an aqueduct. Distribution was by lead
mains
(Nash-Williams
1929,
ed.),
29-33.Zienkiewicz
145,
Fig.5).'
1986.
Nash-Williams
TIRCGLL
1983,
1969,
32,
RIB
(2nd
316-94;
Cambr. 109,
1960,136-72;
111,
1962,
111-24;
123,
1974,
Arch, of Wales, CBA Gr, 2, 18,1978, 75, 53; 76, 54; 77, 29.
fort,
dismantled
482.5, Seguntio;
c.
AD
300-25.
Britannia
1,
1970,
54ff.
Flavian
54 Ant.
Itin.
356-8, Fig.2; 9, 1978, 404-6, wells; 10, 1979, 30, 60, 269-71: '... in
the annexe or ordnance depot WNW of Segontium,
other
leather
inscription,
430,
recording
the
inscription
1963,
dating
125,
from
160.
AD
of
JRS 43,
fort's
1953,
104;
50,
228,
230,
'An
Stephens 1985,
198-209
the
'The discovery of an
reconstruction
1960,
in a well containing
(RIB
43 0)
commemorates
"riuos
was
more
than
"uetustate conlabsos"
one
shows
aqueduct
or
that more
fort.
channel,
restoration
No doubt
347
whilst
one channel
or channel
supplied the
Wheeler 1924,
126-7,
1921 excavations;
32,
267,
1922
Y Cymmrodor,
excavations.
110-11,
CSIR 1.5,
(Ordovices).
been occupied from c. AD 75 to the late 3rd c., or early 4th c. The
vicus seems to have ceased functioning commercially after about AD 13 0
(p.4) . It is
base for
364-5 vicus;
18,
1987,
303-4;
20,
132-3; 9,
1989,
343,
1978,
22,
vicus; 23, 1992, 258; 24, 1993, 273-4; 25, 1994, 246-8.
150.
Montgomeryshire.
1967-8,
64-6;
61,
1987,
15,
well.
(Ordovices).
Coll.
1969-70,
SO 029
920,
42,
1931,
37-42.
17-67;
59,
Nash-Williams
occupation to 4th c.
Britannia
406;
1991,
JRS
224-5,
67, 1977,
1965,
1969,
17,
112-5;
66-70.
60,
TIRBS
58,
1903, 1-38;
59, 1905, 1-22: Ashby; 60, 1906, 1-20; 1907, 1-14; 1908,
?;
62,
1910,
(1901-3),
1-20;
1911,
1901,
1-34;
405-47,
1902,
plates;
1-16:
Ashby,
Britannia 1, 1970, 58
et al;
shows 9
108 n.13,
190; 6, 1975,
190-2,
10, 1979, 102, 168, 174-5; 13, 1982, 132-4; 14, 1983, 283-4; 15, 1984,
241, 271;
16,
1985,
259-61;
17,
1986,
367-70;
18,
1987, 45, 201-2, 307-9; 19, 1988, 422-3; 20, 1989, 264; 21, 1990, 30710; 22,
251-2.
1991,
225;
Hanson
23,
1970,
1992,
3 61,
258-9;
coments
24,
that
1993,
275,
'the
322;
remains
25,
of
1994,
a wooden
89), 2 baths.
TIRCGLL 1983,
it
(Silures).
1978,
394-5;
16,
1985,
189;
18,
1987,
330,
drains and
60, 1993,
c. AD 180
97;
45,
occupation from
136;
52,
1962,
175-6;
53,
1963,
137;
56.
1966,
209.
Rivet 1964,
(2nd
{Iceni).
Caistor-by-Norwich, Norfolk.
Antiquity 3,
1929,
182-5.
Arch.
1970,
52-3, Ant.
1971,
1973,
Itin. 474.6
J.
106,
1949,
(Icinos), 479.10
270,
6, 1975,
300 baths;
176, 207,
62-5.
Britannia 1,
(a Venta Icinorum); 2,
3, 1972,
293-6,
222; 7, 1976,
175,
359;
4,
180; 9,
1978, 358; 10, 1979, 307; 13, 1982, 408-9, inscription; 15, 1984, 110,
344; 16, 1985, 201, 203; 17, 1986, 447; 18, 1987, 42-4, 263-272, 330;
20, 1989,
EAA
in
the
reorganization
early
of
3rd
both
c....and
army
and
may
naval
be
forces
built on unoccupied
related
than
more
purely
to
coastal
as
forts
in
the
Notitia
Dignitatum
on the
Saxon
Shore' .
Frere,
river Tas. See reconstruction isometric view of the Caister forum, Vol
ii, 17, Figs.6-8, p.19, showing bath, and draining ditch leading into
a timber-lined sump,
Roman
Britain in 1914-28,
83-9;
22,
1932,
JRS 18,
33-46; 29,
1928,
1939,
201,
214,
'A
line of flanged iron rings for jointing wooden water-pipes was traced
some 4ft south of the road running apparently from a spring westwards
and eastwards'; 51, 1961, pl.X,I (vert).
230. Stephens 1985b, 197-208,
a single wooden pipeline'.
15,
(Iceni) .
TIRBS 1987,
VCH Wiltshire
Britannia 1,
1970,
47 Ant.
Itin.
474.8
{Duroliponte) , 116,
290;
2,
1971, 178, 188, 191; 4, 1973, 300; 5, 1974, 438; 7, 1976, 340-1, 'item
e: of several wells, one was timber-lined,
opposite
shafts...',
sides....Another
1984,
296;
11, 1980,
19,
JRS 54,
Cambridgeshire,
16,
collection
of
bone
produced
1988,
12,
1981,
271,
450;
20,
1989,
293-5.
167-8;
7, 1978,
TL 445 595,
17;
55,
39-43,
1965,
273,
275,
1987,
213;
58,
pre-Roman settlement;
lead tank;
194.
TIRBS 1987,
VCH
fort
or
{Catuvellauni).
Britannia 1, 1970,
202;
4, 1973,
273
9, 1978,
275;
11, 1980,
47;
13,
65. RCAHMS
337;
9, 1919, 128,
1982,
1985, Nos.160-4.
1984,
15,
JRS
112-8, 151-
Robertson 1990,
(4th ed.),
CSIR 1.4,
Camerton, Somerset.
Britannia 1,
1970, 117-25,
129;
10,
1979,
176; 20,
1989,
174,
183.
Smith
(Belgae).
Camphill, Northamptonshire.
Britannia 11,
Canterbury, Kent.
Antiquity 23,
1976,
1949, 153.
235-44.
Britannia
(Duroruerno), 473.4
1,
(Durarueno), 473.9
1970,
sewer;
42,44-6
Arch. Cant.
Ant.
(Duraruenno); 6,
Itin
1975,
10,
92,
472.5
78;
8,
1979,
16, 1985,
TIRCGLL 1983,
33-4,
RIB 15;
61-80.
important town,
Tatton-
Wacher 1995,
capital of the
CIVITAS CANTIACOR UM, following a Claudian fort which succeeded a preRoman settlement.
(Cantiaci) .
1960.
More
than
industrial complex,
Age
and
Roman
South Yorkshire.
40
kilns
were recorded,
enclosures,
pottery
536-45,
kilns;
6,
1975,
large
pottery
excavated to a depth of
works,
active
pottery
237-8
(1.52m)
from
Gilmour 1955,
27ft
(8.21m)
in diameter.
mid
2nd
(also
large
to
without
TIRBS 1987,
late
4th.
Also
(Corieltavi or Brigantes) .
J.,
142,
1985,
216-36,
Stephens
at
p.224,
comments
'...a
runs
through the
courtyard
of
the
central
for
there are too many water channels for them to have been connected with
drainage'.
AD
163-196.
121,
largemarching
camp
south of
the
fortlet at NT 698 209; 63, 1973, 216. Stevenson & Miller in PSAS 1911,
446-83,
This
figure shows
ten
unspecified type
drains or branches of the main drains marked as such. Three pits are
shown, one in the main building and two in the S-E corner just inside
the inner rampart; both appeared to have been used as drainage pits.
RCAHMS Roxburghshire 2, 1956,
381-3.
CSIR 1.4,
1985, No.44.
TIRBS
1979,
272;
11,
1982, 331-2.
1980,
349,
12,
Carlisle, Northumberland.
Archaeologia 64,
115-6.
Arch.
J. ,1893, 20-36;
476.6
supply;
13,
1913, 299-301;
8, 1977,
1982,
376;
79-89,
10,
149,
155; 5,
1979, 281;
343-4 fort;
15,
1978,
(Luguvallo), 474.1,
1974,
11,
135,
142,
410-1
water
1984, 113;
16,
1985, 197-207,
at
p.202 Stephens states that 'At Carlisle, where a public fountain was
still in working order when St Cuthbert visited the city in AD 685,
the evidence comprises a wooden pipeline and a fountain';
112-3,
360-7;
400 561,
fort,
Wacher
22,
1991,
drains;23,
112-3,
1992,
141-58,
1990,
235,
299-301,
45-109,
319; 24,
at p. 55 it
is
aqueduct,
excavations
the
1994,
1990,
bath building at NY
of
21,
first
263.
famous
Flavian
Burnham &
aqueduct
seen
working by St Cuthbert remains elusive, but the water tank inside the
portico of a building...was presumably fed by it', and presumably the
fountain was
1893, 344-64.
TCWAAS(l)
12,
fact
that
from a supply
of
JRS 42,
1952, 104;
48,
1957,
202,
to have grown
1957,
28ha
(70 acres).
'St Cuthbert
perambulated its walls in the 7th c., and also saw a fountain': Bede,
'Vi ta Sancti Cuthberti', iv, 27; also 1995, 2nd ed., 21, 32, 88, 421.
TIRBS 1987, 17-8, RIB 943-64;
after;
continuing
held
to
c.
3 00;
S-E
c.;
an
c.
early3rd
extensive
AD
c.
(but
settlement
{vicus)
(c.
(Brigantes).
Carlisle-Old, Northumberland.
Arch.
J.
Carlisle',
122,
1975,
271-4.
1959, 15-31.
Frere
24-5,
plan.
CW{ 2) 28,
1928,
Britannia 8,
103-19;
1977,
15,
1951,
'Ateco of Old
16-39;
59,
probably a mansion,
Sommer 1984,
BAR 129.
Wilson 1980,
or Augustiana.
272) . Its date and the fact that it underlies the 2nd c. defences of
the civitas capital show that it is military'.
(Muridino) , 483.5
243-4 amphitheatre;
17,1986,
1969,
366-7
2-5;
6,
1970,
Williams 1969,
(1975,
4-14.
389-93).
TIRCGLL
260; 25,
James
23-6. Ross
(Muriduno) , 270;
1980,
1967,88.
1983,
1994,
34,
248.
No.2,17,
2, 1971, 102,
Carm. Antiq.
Fig.
2.3.
412-3;
fort
and
5,
NashFig.175,
town,
the
town
from late1st
to
4th
c.;
walls
enclosing
c.
6ha.
(Demetae) .
1985,
216-236,
at 217,
223,
entered the
229,
Stephens comments
fortress
through the
(JRS 59,
'A
south
1969,
202). A channel measuring 2.6m wide by 0.6m deep has also been traced
approaching the S-W angle (JRS 48, 1958, 91)
a pipe trench, which suggests that it was a leat cut to supply the
extramural bath-house.
Birley 1967.
Britannia 1,
1970,
273; 2, 1971,
248; 3, 1972,
29,
36,
12, 1981, 305-6; 15, 1984, 57; 18, 1987, 27; 21, 1990, 310; 24, 1993,
278.
Hanson
1970,
'The
aqueduct,
first
recognised
from
air
deep and 8.5ft wide. No trace of any lining for this ditch was found,
and it would seem therefore that it had contained a pipeline, although
no remains of piping were discovered....excavations on the S Gateway
in 1969 did
reveal
channels,
The 'ditch'
clayey soil.)
1962, 163; 53,
127,
202.
1958, at p.91;
52,
47-62.
the Roman
Wilkes
fortress
at
Flavian
enclosure;
Carpow,
Perthshire,
CSIR 1.4,
Severan
184-207,
1985,
'Excavations of
1961-2',
by
Nos.171-2.
vexillation
R E Birley.
TIRBS 1987,
fortress
(9.7ha);
18;
annexe
(0.7ha), aqueduct.
Carrawburgh, Northumberland.
Allason-Jones Sc McKay 1985.
60-87, 88-107.
Arch. Ael.(4) 36, 1958, pi. 25; 40,1962, 59-81 nymphs; 45, 1967, 1-16;
50,
1972,
81-144.
Britannia 1,
1970,
136,
275-6;
2,
360; 8, 1977,
420;
20,
11,
112-5.
1980,
318 fort;
349;
9, 1978,
JRS 51, 1961, 193 ; 55, 1965, 222; 56, 1966, 218.
Longstaffe,
60-87.
TIRBS 1987,
1989,
3,
Hope 1893,
Wall 1877-80,
1983,
198;
122;
178.
1877-80, 88-107.
14,
1971,
Rav.
Cos.
107.26
baths, well.
333-5,
15,
1984,
290;
16,
1985, 282; 23, 1992, 233-6; 25, 1994, 270. Occupation appears to have
been from Period I c. AD 125-50 to Period III 4th c. A well 'at the N
end
of
the
site,
large
cobbled
yard
associated
with
timber
Arch. Ael.(3) 13, 1916, 85-102; Arch. Ael.(4) 31, 1953, 82-
'Paulinus baptised many people at this well during the 5th c.', 346,
Cohors I Batavorum quingenaria unit left an inscription on Hadrian's
Wall
Dig.),
Castle 46; 3, 1972, 41, 193, this fort additional to Hadrian's Wall,
and rebuilding at Carvoran and other forts in c . AD 163; 4, 1973, 275;
10,
1979,
346;
12,
1981,
380,
near MC 46;
17,
1986,
354
381;
19,
1988,
1913,
448-9;
Radnorshire.
1914,
411-20;
12, 1912,
183-98;
46, 1956, 119-20, bath-house, 110 ft (33.5m) long, to the south of the
fort.
Trans,
(1.44ha).
Castle Greg, Lothian Region, Scotland.
PSAS 1, 1852, 58-9; 52, 1917-18, 221.
1977,
18, 263.
Oswald 1941,
119-22,
JRS 31,
Smith
1987, 285-7,
36-7, 68.
62-5, Figs.24,
25, 29
73, 1969,
7-112.
183-96,
41-4 bath-house.
124-6; 2, 1971,
238;
6, 1975,
98; 8,
3 wells, stone-lined.
PNRB 1979,
Margidvnvm.
Britannia 1,
413-4,
Ant.Itin.
Todd 1969.
TTSN 31,
VCH Nottinghamshire 2,
1910,15-7.
TIRBS 1987,
55.
(Corieltavi).
1942,
142-51,
183-96,
at 144,
comments,
'Felix Oswald
Oswald
1948;
1952;
1956.
PNRB 1979,
413-4.
1987, 55.
1910,
15-17.
were
Todd
(eds.), 211-
Webster 1980,
162.
TIRBS
(Corieltavi).
Ael{ 1) . 4,
1984,
280;
16,
1855,
1985,
273.
271;
17,
Britannia 14,
1986,
355
378-80,
Northumberlandshire.
1983,
290-1
Fig.11;
18,
Fig
6;
1987,
15,
316,
Bruce 1978,
J. 32,
1928,
197;
1875,
19,
135-54.
1929,
190.
1955,
Heslington 1867.
257-9.
Lukis
1875,
JRS 11,
135-54 plan
TIRBS 1987,
1921,
I,
83;
18,
& plan
II.
1980,
33.
Tyler
North Stainley,
SE 292 756;
baths
(Brigantes).
J. 142,
1985,
have been a stone culvert which carried a piped water supply from a
spring.
478.7
Britannia 1,1970,
(Lagecio, var.
47, 51-2,
Laiecio),
64 Ant.
Rav.
Itin. 475.6
Cos.
238; 7, 1978,
107.8
{Legeolio) ,
(Lagentium,
var.
288-9,Fig.7,
bath-house; 11, 1980, 18 as Lagentium; 12, 1981, 330; 13, 1982, 349-50
vicus;
14,
1983,
294-5,
337;
16,
505; 20,
1985,
279-80;
1989, 278;
24,
17,
1986,
385;
1993, 287.
18,
West York
(Brigantes).
Britannia 2,
1971, 132,
253; 9,
17, 1986, 385 fortlet; 19, 1988, 444; 20, 1989, 280-2.
114,
1989, 225-9,
'Castleshaw'.
Hanson,
1970, 396.
154;
67,
1957,
118; 71,
1961,
1978, 429;
Current Arch.
The Archaeology
163-5;
77,
1974,
1-18.
Bruton,
1st Interim
Report 1908 &2nd Interim Report 1911. See Fig.6 for well
and drains.
a bath establishment.
097;
Flavian
(Trajanic).
fort
100-3.
reduced
to
TIRBS 1987,
a
fortlet
20, SD 999
size
0.34ha
11.
Bruce
(Brigantes).
228-9.
Northumberland.
Britannia 5, 1974,
463;
12,
1981,
Uxellodunum,
and
formed part
of the fort's
159-65; 85,
1985, 77-80.
Hansoninterprets:
'andpresumably
The OS
these
'Camboglanna'; this
fort
is unique,
gives
since it was
built between the Vallum and the Wall opposite MC 57. The remains of
the fort was completely destroyed when Castlesteads House was built in
1779.
TIRBS
1.46ha).
1987,
20,
RIB
1976-2009;
fort of
Hadrian's Wall
(c.
(Brigantes).
Arch.
J. , 131,
(plan).
1974,
Mag.
1974,
152,
164-5.
Britannia 1, 1970,
431; 7, 1976,
286;
332-4.
2, 1826, 366.
Artis 1828,
2,
1971,
Durobrivae
JRS 53,
1963,
pis.iii-viii, xi-xiii
243-89;
9, 1954,
135; 54,
1966,
61.
Rivet
1958, 114.
367,
RCHM 1969,
TL 1165 9764,
1154 9770,
1153 9781,
1972,
22-5.
1964,
3,
320;
Gentleman's
164,
2 wells;
source Bede.
24,
27,
CA 45.
1149 9790,
5,
59,
Lewis
sites
21-2.
Swan1984, fiche
2 wells,
bath
house .
Huntingdonshire.
Peterborough Museum record no. 899, 945, 946, 950, 953, 954, 959, 960,
966, 968.
Swan
1984,
fiche
368,
TL
1175
Catsgore, Somerton,
Arch.
9781-
1185
9770,
Wild 1976.
2 wells.
VCH
Somerset.
Britannia 2,
1971, 276-8;
4, 1973, 311-2;
3, 1972, 343; 10, 1979, 323-4; 11, 1980, 388-9; 15, 1984, 316.
1982.
Leech
1,
1970,
40,
42-3,
47:
Ant.
Itin.
465.2
{Cataractoni) ,
468.2, 476.2, 418; 2, 1971, 200-2; 3, 1972, 260-1; 7, 1976, 137-8, 314
drain, 380; 8, 1977, 288; 9, 1978, 481-2;
363;
1986,
13,
83;
1982,
20,
defences, 306-7.
1960,
420;
16,
1989, 277;
1985,
21,
197-207,
1990,
water
51,
supply,
327-8;
10, 1979,355;
277,
22, 1991,
1953,
1961,
90; 45,
330;
238-40
1954,
17,
fort
82; 50,
204; 58,
1968, 208;
63,
1970,
362:
1973,
214.
179-81.
Hanson
found supplying a
Stephens
1985a,
199, 200,
Occupation
207.
Wacher 1971,
170.
1987,
21,
RIB 725-9;
extensive civil settlement from midle of 2nd c., walled in mid 3rd c.
(Brigantes).
Cattybrook, Almondsbury,
Avon.
excavations started.)
Caversham, Berkshire.
Britannia
1, 1970, 210; 12, 1981, 271-6; 20, 1989, 319, 333, no.13,
pi.26, circular lead tank and timber-lined well, with 4 square timbers
each about 1.5 m long supporting vertical sides.
Chalk, Kent.
Arch.
Cant.
73, 1959,220-3.
1971,
186,
building materials;
Black
1987,
3, 1972, 112-48;
282-5,
interim Report.
Wooler 1917,
190.
Charmy Down,
Avon.
blocks.
2,
1975, 282-3,
130-1.
TIRCGLL 1983,
35;
(Cantiaci).
Grampian, Scotland.
1991,
6,
Chapel Hill,
Britannia 22,
Britannia
145,no.35.
278:
'nearby
'...further buildings
date';
23,
1992,
cylindrical
296,
water-
Chedworth(1), Gloucestershire.
Archaeologia 59,
1905,
207-32.
Arch.
J 44,
1887,
322-36;
59,
1905,
1972, 270-5;
4, 1973,
318-9, hypocaust;
1984, 312;
227; 7, 1976,
178;
9, 1978,
18,
329-30;
13,
10,
1979,
1982, 377;
1987, 337-9;
19,
15,
1988,
465, quarry and hypocaust; 20, 1989, 309-10, 2 bath-suites; 23, 1992,
294, floor-swilling drain. Goodburn 1972.
215-27,
139,
56, 212.
JRS 45,
1955,
149 no.27;
55,
215-6;
Courtyard type villa. Early 2nd c. to late 4th c., probably destroyed
by fire.
The
Chedworth
is
78,
Gentleman's
propbably
one
1959,
5-23
Magazine
of
the
1957,
(reinterpretation of the
1865,
best
TBGAS, 76,
i,
kown
595;
villas
ii,
in
302-3.
Britain,
c. to the late 4th c., with evidence of fire in the early and late
phases.
N.W. corner of the site; it was channelled into a cistern just outside
the
suggested
N.W.
angle
of
the precint
wall.
roughly
occurs at H,
identified.
circular
some
and a latrine,
are
given but many relate to small finds and mosiacs. McWhirr 1981, 90-12,
150-3. Ross
centre,
1967, 50ff.
two
temples,
Romano-Celtic,
the
other
nymphaenum.
(Dobunni).
20,
possible bath.
1989,
309-10.
JRS
21,
1931,
240,
water
pipes
and
Mr C E Key,
"90
pillared hypocaust'.
Britannia 19, 1988, 455; 21, 1990, 338-9 Fig 20; 22, 1991, 259, well,
4.4m diam at top, narrowing to 1.1m at depth 14.5m. With such a large
diameter at the top it appears as if the mouth was enlarge by caving
in of the soil.
Chelmsford, Essex.
AEx
1972,
47-8.
Britannia
(Caesaromago), 480.6
207-9
1987,
Fig.95,
37,
3-29.
3,
capital
1972,
1978, 449;
457; 20,
1989, 302,
Wacher 1990,
4, 8,
CBA Gr.7
vicus, mansio,bath,
474.3
342-3; 9,
Sc
1969,
'possibly
Itin.
4, 1972,
1988,
Burnham
& Hist.
Ant.
249,270, bath;
19,
Essex Arch.
52
Bulletin 1988, 5.
47,
(Cesaromago);2, 1971,
1, 1970,
drains,
of
the
223.
Rodwell
Sc
212-3
aqueduct.
Civitates
TIRBS
Trinovantum'.
(Trinovantes) .
Cheshunt, Hertfordshire.
Smith 1987,
177.
OS NMR,
TL 30 SW/5,
(personal communication by J
Chester, Cheshire.
Britannia
1,
1970,
42-3,
54,
65,
180,
JRS 45,
203;
3 63, comments,
58,
'The
482.8
1972, 313,
1968,
87- 4thc.; 2,
sewer;
8,
of
1971,
1977, 387;
11,
(Deva), fortress
Legio XX, c. AD
469.2,
146;
207;
46, 1956,
59, 1969,
(legionary)
235.
50,
1960, 221;
Hanson 1970,
57,
185-91,
from springs two miles to the east of Chester. Finds of clay piping
along the projected
indicating the
been necessary
of Deva'.
Stephens
1985(b),
223,
229,
comments,
'Nymphis et Fontibus"
in the plural suggests that more than one spring was tapped.
Distribution was by lead mains dating from AD 79', and in 1985a 206,
he comments,
in
that
AD
79
(RIB
463),
the
same
year
its
water
mains
were
360
the service pipes supplying it were contemporary'. Thompson 1965, 512, lead pipe.
460,
2434;
(Cornovii).
13,
1936,
218-57;
15,
1938,
fort;
2,
1971,
249;
222-37;
48,
1970,
3, 304,
306-7;
4,
97-155;
Arch.
Britannia 1, 1970,
1973,
275-6,
347-8;
5,
1974, 360-73, 408-9, 471-80; 7, 1976, 342; 8, 1977, 198; 9, 1978, 4801; 10, 1979, 346; 11,
at
1993,
17, 1986,
418-9;
453;
Stephens 1985a,
Chelmsford,
the
mansio
197-208,
was
110-2,
comments,
supplied
by
19,
'...,
wooden
item 23,
fort /vicus/mansio,
'An aqueduct
of
Hanson 1970,
channelled
stones
that came from a spring just to the west of the fort, refs. RIB 1049'.
See
Birley's
new
excavation
reports(1993) .
Aqueduct
RIB
1049;
Ael.(4) 22,
9,
street;
1978,
15,
426
1984,
1944,
83-90;
sewer;
281,
10,
342;
this
fort
is
known
46,
1968,
1979,
22,
285;
1991,
from
11,
238,
Hanson 1970,
only
75ff.
1980,
361-2,
in the N-E
362,
an
Britannia 2,
1971,
cobbled
room
'...a
inscription
attesting
its
construction during the early 3rd c., RIB 1049, inscriptional evidence
of
the
aqueduct.
TIRBS
1987,
23,
RIB
1043-
50;
CIL
vii,
1234
{Brigantes).
Chesters, Northumberland.
Archaeologia 46, 1878?, Iff: refers to a 'finely made stone-lined well
in the principia courtyard..., p.247.
vicus; 19, 1942, 163-4;
26.
Birley 1960,
16.
Arch. Ael.(A)
8, 1931, 219-340
Britannia 1, 1970,
140,
264;
2, 1971,
127-9,
199-200;
10, 1979,
346;
11,
1980,
314,
319,
338;
18,
1987,
45
vicus; 22, 1991, 234. Bruce 4th ed., 1895. He comments on a rain-water
tank in the North Chamber of West Gate, with gutter
place,
pp.89-90.
stones still
we come
in
to one of
inner
receiving
side.
On
the rainfall
the
previously
now', p. 90,
2nd para.
the
Hanson
1850s.
from
roof, which
were
in the position
found,
for
when
the
in which they
are
1970, 243-8,
item 7,
363, items
25
& 26,
'The
was inserted
latter's
construction)it
referred
to
Severus';
by
an
see no.l
aqueduct, p.247.
1972, 193.
is
possible
inscription
p.245.
RIB
Scott
that
1463,
this
is
the
aqueduct
dating
to
the
reign
makes comment
on
suggested
of
3rd
JRS 41, 1951, 55; 45, 1955, 146; 47, 1957, 229; 62,
Turret 27A.
water-mill;
fort
on
Hadrian's
Wall
overlying
(Brigantes) .
Chesterton(1), Cambridgehire.
Artis 1828.
Condron 1996,
Swan 1984,
415, well.
Chesterton(2), Cambridgeshire.
Condron 1996, 416-7 ?bath-house.
913, 926, 927, 951,
961,
1868.
RCHM (E ) 1969,
264-5,
site 42.
Swan
104-17,
Flavian
bath.
fort
VCH Staffordshire 1,
(1.9ha,
or
more
JRS 48,
1908,
probably
1958, 150.
NSJFC
189.
TIRBS 1987,
2.2ha);
bath-house.
(Cornovii).
1970,
5, 1974,
9,
1978,
49-50 Ant.
457;
Itin.
6, 1975,
466-7 wells;
282;
10,
477.10
7,
1979,
(Regno);
1976,
332-3;
3,
1972,
372-3 baths;
11,
8,
1980, 396-8;
350-1
1977,
14,
1983, 332-3;
372;
19,
289.
16,
1978,
1985,
202,
479; 20,
207 leat;
1989,
89-96.
255-71,
Down
1951-2,
164-220;
Fig.121, bath,
320; 22,
1991,
& Rule
SxAC 90,
1995,
5,
17,
27, 290;
100,
drains,
1962,
13-19,
2 wells.
1977, 353,
25, 1994,
1974;
III,
Ross 1967,
80-110.
Wacher
TIRCGLL 1983,
37, RIB
{Regni).
Chigwel, Essex.
Black 1995, 102 Route 9, 118 item 42, 119 Fig 1 position 42, 120 item
18, 121, Fig.2 position 18, mansio with bath.
3, 65, 73, Ant. Itin. 480.7
1979, 102.
idem. 1985,
47, 'In each of these two main sites a well has also been found,
Rivet 1964, 147.
61,
SMR 4057,
ES
17.
...'.
Scott 1993,
TIRCGLL 1983,
37.
(Trinovantes).
Chilgrove(l), West Dean, Wellmeadow, West Sussex.
Black 1987,
153.
Down 1979,
bath-house.
Black 1987,
1972, 351;
1975,
282;
153.
Britannia 1,
4, 1973,
8,
Committee, 'Report
321;
1977,
5, 1974,
422.
302-3 bath;
458 well,
Chichester
Cunliffe
3,
6,
1973,
102,NAR SU
286;
Civic
2,
1971,
Soc.,
Excavations
1970,
81 SW 60.
TIRCGLL
87-8.Down
to 5th c.
1979,92-3,96-
83,
S-W,
27-
4, 1982,
363
JBAA 5, 1850,
(Catuvellauni).
9.
Scott 1993,
199,
117;
57,
1958-60,
24-9;
67,
1972,
174;
69,
1974,
SMR SE
WAM 49,
185;
70-1,
Churchill, Oxfordshire.
Britannia 13, 1982, 3 67, spring.
63.
30;
5, 1974,
No.4, 1974,
21-2.
436, well.
CBA G r .9
Oxoniensia 37,
1972,
possibly pre-Roman.
Cirencester(1), Gloucestershire.
Antiq.
J. 41,
Britannia
1961 to 47,
1,1970,
57-8,
1967 inclusive;
64,
Rav.
Cos.
49,
106.31
1969,
?;
53,
1973,?.
(Cironium Dobunorum),
184, 196 early fort, 227-36 defences, 293 amphitheatre; 3, 270-5, 33940 houses; 4, 1973, 122-269 pottery, 307-9 defences, bath; 5, 1974, 7,
283;
6,
1975,
271-3;
7,
fifth-century plague,
112-3,
355;
118,
14,
148,
1983,
413,439-40; 9,
167-8,
343-4;
176,
15,
319;
1984,
384;
1978,
11,
68;
1977,
198,
323-4
359, 362,
455;
10,
1979,
299-300,
411;
12,
1981,
190,
202-3
1980,
16,
8,
1985,
181,
183,
water supply; 19, 1988, 465-7; 22, 1991, 274-5; 23, 1992, 217-8, 2945;
24,
1993,
'Cirencester'.
303;
25,
1994,
285. Wacher
1995,
302-22
Fig.136
McWhirr
military
occupation
c.
AD
45 to 75 (?)
(RIB
108-9);
town developed
under the Flavians and eventually became the second largest town in
Roman Britain in the 4th c.; probably the capital of BRITANNIA PRIMA
(RIB 103); survived into 5th and 6th c. when taken by the Saxons after
the battle of Dyrham in AD 577.
70,
SMR 2092,
(Dobunni).
Gloucestershire.
GS 26, wigged-corridor villa.
TBGAS 33,
197-98;
4th c. settlement.
Tyler
106-9. Antiq.
Cottenham, Cambridgeshire.
Whitwell
1982,
4th ed.,
212,
TL
1949,
145-63,
excavation at
J. 29,
Phillips 1970,
1991.
South Lincs.
28 NEl.
TIRBS
Durobrivae 5,
Arch.
4,
1980,
17,
'A
canal,
1987,
19-23.
flat-
bottomed and 10-13m wide, running from the river Cam near Cambridge to
the Witham at Lincoln and making use of rivers where convenient'. Much
of its earlier sections pass through the Fens. Whether it flowed or
held water permanently is uncertain. The altitude along the length of
its profile must vary up and down in elevation.
Car Dyke.
Cobham, Surrey.
Black
1987, 150,
constructed
c.AD
constructed c.
No.89,
320-30,
suggests
abandoned
possibly 2
c.
AD
bath-buildings;
355-65.
Frere
1947,
Coddenham, Suffolk.
Britannia 1, 1970,
Combretovio) , 71;
1975, 207;
1980, 19;
45-53,
52-3,
Ant.
Itin.
480.2
(Conbretovio, var.
7, 1976, 341-2;
18, 1987, 8.
8,
1977,
Dunnett 1975,
405 fort; 9,
43, 62, 89.
1978,
448;;
11,
(Trinovantes) .
Colchester, Essex.
Arch. J. 123, 1967, 32; 142, 1985, 216-36,
identified
west
at
the
Sheepen
site
to
the
of
the
(Colonia) , 480.4
fortress'.
(Camoloduno) ,
149, 181-2, 258, 290; 2, 1971, 168-94; 3, 1972, 164-81; 4, 1973, 302-4
fortress and colonia;
5,
1974,
210-1;
6,
1975,
79-83,
176,
198-9,
263-4; 7, 1976, 176-7, 180, 182, 189-90, 343-4; 8, 1977, 92-5, 97, 100
water supply,
198-9,
407,
437-8;
9,
1978,
449-51
sewer;
10,
1979,
157-63, 148, 308-9; 11, 1980, 378-9; 12, 1981, 289-90; 13, 1982, 371;
15, 1984, 105, 343; 16, 1985,
178-90,
324-5.
Crummy
1977,
65-105,
at
wells timber-lined,
26-8,
19, 1988,
24,
1993,
100,
295-6;
1-6,
Figs.11,
Idem.
196; 20,
302;
20,
1984,
17,
25,
where
5, leat,
wells,
leat,
spring and aqueduct, Figs.84, 96, 99, 101-4, 107-9, 111, 115-7, watermains and iron collars,
146;
idem.
tanks
140-2,
Fig 102,
104,
106,
131,
baths
1992, baths 71, 268, tanks 31, 40, 63, 78, 89-90,
105,
255, 355-6, water-pipe junctions collars 72, 101, 105, 358 and micro
fiche 995, wells 36, 335-6, 365, 388, 390-2, and micro-fiche 799, 820,
876, 879, 970, 1003 1019, 1048, water supply 24, 30, 40, 44, 47, 67,
69,
72, 101,
27-61,
at
31,
105, 355-6,
Fig.2,
358;
pl.IIIA;
idem. 1990-1,
bath(?).
idem,
128,
1975,
Dunnett 1966,
n.17,
it has
been
conjectured that water was pumped uphill from the Sheepen springs but
the
aqueduct
was
most
certainly
a pipeline
aqueduct
employing
an
inverted siphon to convey water from the springs to the south or west.
It is conceivable that
an
elevation
Gosbecks. The
to
colonia
have
was
first
laid to supply
282-4,
pi.XII.
supplied the
fortress
supplied
by
the
battery
of
the
four
fort
at
wooden
pipelines which entered through the Balkerne Gate. These date back to
the pre-Boudicean period.
Hull 1958,
fiche 277, Colchester (7) TL 9865 2517, well V; fiche 286, Colchester
(13), well nearby at TL 9874 2566, Sheepen, east-side of Hull's region
4; Colchester (14), Middleborough Castle Market site
timber-lined wells
in
nearby,
(6 wells).
the vicinity of
kiln
36,
TL 9926 2556: 3
with
TIRCGLL 1983,
another
well
fort,
legionary fortress
(Legio XX Valeria),
then colonia AD 49; destroyed by Boudica AD 60; rebuilt and given name
Victricensis;
Chelmsford);
possible
temple
enclosed c. 40ha.
capital
of Claudius,
of
the
Civitas
Trinovantum
and
7 Roman-Celtic
temples;
(cf
walls
(Trinovantes).
1993,
26,
SMR 1280,
BU 22:
'head of spring'.
Wolverton and
(?)bath.
Coldharbour, Gloucestershire.
Britannia 16, 1985, 93f.
Colerne, Wiltshire.
Scarth
1864,
120-1.
VCH Wiltshire
1,
1957,
59.
TIRCGLL
1983,
39.
(Belgae).
Coleshill, Warwickshire.
Britannia 10, 1979, 300 bath, 3 tanks shallow mortared-lined,
c. 1 x
1982, 419;
Cornovii) .
in rock-cut,
2m diam.,
19,
1988,
505-6.
Faull
17m deep,
Sumpter
with
& Moorhouse
(eds), 1981,
(eds.)
10,
(Brigantes) .
1979,
195,
288
well-house,
1981, 33 0; 18,
Scott 1993,
428;
substantial
9, 1978,
147.
1987,
Proc.
SMR 1972, WY
YAJ
1-14;
60, 1938,
51-65;
84,
1963
113,
another well
at the New
Clinic site, 17.33m deep and 1.2m in diameter, of which the upper 2.1m
was
steined
with
flint and
limestone blocks.
RCHM
197 0,
County
of
in vicinity of building
III: 1) 1.1m
(3.5ft)
diam. , 10.4m(33.5ft
(p.560a),
3)
stone well-head of hexagonal plan with circular shaft 0.6m (2ft) iam.,
contained
filling,
9055;
late
SY
at
Roman
material
in
SY
6904
9053,
its
partially
excavated
upper
stone-lined
covered
drain
stone-lined,
(2.13m,
7ft,
SY6909 9050. Figure also shows several drains. Colliton Park Villa was
situated too high to obtain water from the Dorchester aqueduct.
Colsterworth, Lincolnshire.
Antiq. J. 12(3) 1932, 262-8;
164.
Grantham
Whitwell
1982,
Journal
132-3,
13
February,
213,
1932.
SK 92 SW 17;
Tylecote
idem.
1970,
1962,
113-4.
230.
TIRBS
(Corieltavi) .
1,
1906,
TIRCGLL
1983,
Somerset, 309-10,
39,
RIB 179;
item 19,
probable
villa, courtyard;
fort,
coins
bath(?).
of
4th
c.
(Belgae).
9,
1849,
1842, 526.
Cambridge Antiquarian
Gentleman's Magazine,
(n.s.)
8,
7.
pipe.
Samianpottery,
also Vespasian
and
VCH Cambridgeshire 7,
coins:
1978,
45. RellanDrawings,
Gratian, c.
4thc.
pottery.
TIRBS
probably
1987,
24.
(Catuvellauni).
to 3rd c.;
456;
5,
1974,
7, 1976,
271-82;
1980, 393;
22,
Scott 1993,
'The Roman
Tomalin, 1987.
154.
1991,
290 settlement;
WS33. Occupation from earlier rather than late 3rd c., probably c. AD
90-110 based on stamped tiles, and as late as 4th c.
Compton(2), Surrey.
Scott 1993, 176, SMR 1630, SY 11.
baths.
Compton(3), Berkshire.
TNFC 7, 1934-7, 211-6.
(Atrebates) .
394,
AV
34,
PSANHS 108,
small
1963-4,
buildings,
one
172-4.
with
bath
and
hypocaust.
Corbridge, Northumberland.
Arch. Ael. (2) 6, 1865, 189;
303;
5,
1909,
305-424;
6,
205-72;
7,
14,
1937, 95-102;
33,
1967,
17-26;
5,
1977,
47-74
143-267;
1938, 243-94;
36,
46,
1958, 227-41
1968,
forts.
115-26;
Bishop
8,
1912,
1940,
1955, 218-52;
45,
15,
1911,
31,
vicus; 37,
49,
1994.
1971,
Bishops
85-115;
21,
1953, 116-26,
1959,
59-84
1-28;
Arch.
& Dore
1988.
occupation;
6,
1975,
66-8
fountain
and
276;
5, 1974, 409
granneries,
74-5
16, 1985,
two aqueducts:
1991,
18,
1987,
1) stone-lined, 2) leat.
146, Fig
45 town plan.
Dore
the Stanegate,
when
the
forts
were
moved
to the Wall,
and
in
became
an
It eventually became
a small town covering some 40 acres... The civilian areas are in the
fields
to
the
west
and
south
of
the
fort...
The
water
in the
centre of the
the
fort.
supply
is
It fed a great
Thefountain
head was
the
river'.
TIRBS
1987,
24-5,
RIB1120-97,
2296-7.
(Brigantes).
Corhambry, Hertfordshire.
Britannia 6, 1975,
Original
timberstructures
Period of occupation
at least
probably
15 phases can be
burnt
down
during
99,
142-51,
183-96,
at 192,
Well, brick-steined.
4, 1970, 7-8.
Britannia 1,
Excavations',
1969,
11.
JRS 48,
1958,
1993,
140;
1959,
115.
Quinnell
Scott 1993,
WDASN 4, 1959, 7;
TIRBS 1987, 26.
293.
49,
1970,
7, 1962,
(Catuvellauni).
Cotterstock, Northamptonshire.
Britannia
Research
211-2.
Comm.
1991,
252.
Annual
Rep.
Durobrivae
1989-90,
5,
4.
1977,
24-5.
Northants.
Nene
Arch.
12,
Valley
1977,
NH 29.
49-53
22,
Stuckeley,
(spring?).
Covehithe
192.
Whitwell
1887,
1982,
(Catuvellauni).
Suffolk.
438;
6, 1975,
257,
8 well shafts;
7,
1976,
338-9;
8, 1977, 401; 9, 1978, 444, well, 3.7m deep; 10, 1979 338-9; 21, 1990,
338;
Smith 1987,
pits.
235 wells.
BDAS 1978,
Scott 1993,
'Excavations
154-5 Fig.l,
No.81, 7(10),
Britannia 9, 1978,
12,
3 08-9.
484.3
409;
13,
1966,
(eds.),
11,
District
1980,
1982,
349,
333-4;
14,
220.
Current Arch.
1981, 316-7;
bath-building.
1981, 308-9.
and
Wales.
1994,
by Berkhamstead
Parkhouse 1981,
41; Ant.
It in.
(Silures).
60, 1962,
(under Maidenhead),
62-91.
c. AD 2nd to 4th c.
99;
Scott 1993,
TIRCGLL 1983,
41,
(Atrebates).
8,
1.4m diam.;
289;
19,
1972,
1977,
10,
1988,
304 vicus;
5, 1974,
429;
278-9;
23,
11,
1992,
163-224:
9,
1978,
1980,
264-5,
Fig.7:
1976,
241-5;
59, v(12),
occupation,
1977,
378-81
latest
1976,
418 well,
2,
1981,
305-6 bath
c.
321;
3.5m deep,
14,
1983,
Crammond road'.
tank Fig.12.
and
354;
7,
Current Arch.
(2 pictures).
v(8),
55,
indicates
in late 3rd or
4th c.
22,
Britannia 3, 1972,
9, 1978, 413-5 Fig 5; 10, 1979, 276-8; 11, 1980, 47; 13, 1982,
1991,
228.
DAES 1977,
12f.
Macdonald 1934,
371
258-71.
PSAS
1984,
1969,
109;
63,
1970,
196,
3 65,
55.
1973,
224.
to
further
the
west
comments,
(south
been
of
41-5.
JRS 55,
item 34,
Crawford 1949,
fed
the
'the
129-31.
camp) comments:
'This
by an aqueduct
coming
site...', item
same
35(north
1965, 81;59,
Hanson
camp),
Hanson
writer,...alsorecorded
that
the
the
1985b,
site'.
Roy
1973,
224, comments,
pl.xi.
PSAS
97,
163-4,
196-8.
Stephens
shows that it was not a leat. Undated but presumably 1st c.'. SAF
1981, 30 pl.l.
12,
mill-pond
south
noted
to the
crossing
the
of Dalswinton
site
of
the
1-48.
'An open
GAJ 4,
leat feeding a
Roman
fort.
Since
the
Brandy
is the nearest
been
Burn,
source
from which the external gravity-fed water supply could have been laid
on to the site,
it
is possible
that
...Woodhead...It
the
59; 48,
Maxwell
is possible
that
1961,
122;
labour-camp,
the southernmost
(NX
986
839)'.
Antiquarian Soc.,
anexes
structure
Trans.
34,
of
1957,
of
the
the military
Dumfries
10-11.
&
site until
recently
complex at
Dalwiston
Galloway
TIRBS 1987,
28,
Nat. Hist.
Large
&
fort with
Darenth, Kent.
Arch.
Cant.
22,
1897,
49-84.
Britannia
1,
1970,
304: ' a
small
periods
inscriptions;
voussoir
c. AD
180-350;
4, 1973,
tiles;
12,
176,
1981,
2,
1971, 184,
pottery;
164-5.
KAR 18,
1969,
372
6,
building
1975,
Bromhead
18-21,
176;
1942,
drains;
matls.,
297-8,
10, 1979,
186
148,
aqueduct
19,
1970,
16.
Rivet 1969,
132-49.
villa,
111,
pis xviii,
bath-house,
r a Kent 3, 1932,
cistern,
xix,
Morris
1979.
cistern.
xx,
111-3
1983,
43.
Edgar
1813,
40.
(Cantiaci).
35,
1932,
1853,
383-95.
177.
Barker
1830,
345.
N-W 3,
1981,
62-72,
item 18, No.23, Figs.54-58, villa near S-W corner of the northern fort
on a small knoll at 191m above OD;
Fig. 55. Not clear if villa occupied before 3rd c., but went into 4th
c.
Scott
208.
1993, 142,
195.
Smith 1,
1848,
113;
3,
1854,
(Catuvellaunni).
16,
1985,
probable springhead
Dating
299-300, Fig.24,
paved
drain
leading
from
the Hall.
Denton, Lincolnshire.
Allen 1834, 2, 215.
1949-50,
1973,
189.
138.
29-57, the
Arch.
Rev.
3,
1968,
to examine the
small
deep.
1950-1,
JRS 40,
179.
Britannia 2,
EMAB 2, 1959,
8.
1971,
176;
Greenfield,
4,
1971,
finds;
a square well,
100;
50,
1960,
221-2,
plan at 221.
LHA 5,
6,
1971,
29-57;
7,
1972,
7.
53.
bath,
Scott
Figs.72-4,
1973,
drains,
Vol.2,
Fig 72.
and well
6m deep,
Fig.74,
steined,
111,
220,
(Corieltavi).
Brassington
1980,
8-47.
Britannia
419-20,
1,
1970,
283,
vicus; 6, 1985,
well
(SK 353
242-4,
6 wells,
373
fort;
92, 1972,
14,
101,
1989,
5-14.
1924,
plan.
20,
283-6.
DAJ 87,
EMAB 10,1974,
224,
(SK 252
JDANHS 47,
1981,
90,
102,
1968,
1988,
115-9;
1982,
19,
1925, 256-8;
plan;
6,
1967,
2 tanks;
74-86.
Webster
1961,
82,
JRS
211-3,
1962,
85-110.
110;
TIRBS
(Corieltavi) .
1990,
340,
Desborough, Northamptonshire.
PSA 22,
1908, 333.
Central,
2, 1979, 33-4,
Desford, Leicestershire.
Liddle 1982,
vol.l,
42.
TLAS 37,
(15ft) deep.
1961-2,
67 well,
stone-lined 4.6m
(Corieltavi).
Dewlish, Dorset.
Britannia 1, 1970, 298; 2, 1971, 279; 3, 1972, 345; 4, 1973, 314-5; 5,
1974, 453-4 bath, water containers in the floor Figs 21 & 22; 6, 1975,
277 bath;
Dorset,
7,
1980,
1976,
360-1.
PDNHAS 94,
1972,
Scott
81-6.
RCHM
County of
1993,
52,
SMR 1 040
011,
DO
11,
(Durotriges) .
Cambridgeshire.
Britannia 18, 1987, 367; 19, 1988, 450; 24,1993, 296; 25, 1994, 274-5,
Fig. 11,
3 phases
identified,
2 wells,
tank.
1970,
a period
331-2;
10,
289;
of
2,
1971,
disuse
1979,
269-70,
before
349-50;
Fig.10,
final
11,
compact bath-suite,
demolition
1980,
410;
c.
12,
Ditchley, Oxfordshire.
374
AD
300;
1981,
4,
386.
Britannia 2, 1971,
houses, well.
c.AD
70-200,
and
c.AD
30-400.
VCH
Oxfordshire
I,
1939,
311-13,
(Dobunni).
1,1970, 270,
3rd aqueduct
2nd aqueduct
recognised;
4,
1973,
271;
Boon
Williams 1966,
Sc
H,
Sc
33,
1979,
16-9.
Where
JRS 8, 1918,
Jones
Sc
Jones
1982,
Pliny
(The Elder),
Nos.
Maude
113,
10-4.
115,
produces
Lewis
Nat.
128.
gold and
victory'.
210-11.
1966,
Lewis
1976-77,
Hist.
and
Walker 1992,
Sc
41-7.
Jones
Sc
16pp.
c.
171, 288-
198-9,
c.
AD,
Fig 25.
244-72.
Nash-Williams
other metals
RIB 406.
1969,
1969,
365,
Macpherson,
Sc
Lewis
1950,
RCAM Carmarthenshire
1st
Davies
37 is incorrect.
Sc
Jones, Blakey
59,
268
Boon 1993,
Sc
122-7;
33.74.
Tacitus
silver
51-7.
Fig.3,
1981,
Sc
53-102; 56,
1991,
Bick
Burnham 1994,
note:
300.
1936,
244,
12,
Burnham, Burnham
94-139,
Jones
1971,
168;
1935,
idem.
10,
2,
122-7.
1993,
at 101-2;
recognized;
the
79-84.
Inventory,
Agricola 12,
which are
Sc
'Britain
rewards
of
{Demetae) .
Doncaster, Yorkshire.
Britannia 1, 1970,
47,
51, Ant.
Itin.
2, 1971,
253; 3, 1972,
311; 4,
8,
331,
1978, 247-70;,
1977,
415-6;
13,
Magilton
384;
1982,
1972,
Valentinian
9,
420;
25,
34-5,
II coin.
275.
Smith
PNRB 1979,
282;
428;
10,
1979,
267.
Current
from
Flavian
329:
207.
Ant.
TIRBS 1987,
(over 2.6ha)
v ar. Ano) ;
It.
29,
290;
Arch.
to
11, 1980,
273-9.
AD
388-92,
Iter V;
No.32.
RIB
317;
33,
c.
475.5
1987,
1994,
occupation
1973,
618.
478.8
Selkirk
Flavian
to
{Brigantes).
Dorchester(1), Dorset.
Antiq.
p. 440,
J.
20(4),
1940,
429-48,
(K M
Richardson),
60-1 Ant.
Itin.
345; 4, 1973,
483.6,
486.15
10,
water supply; 9,
375
facing
Britannia 1, 197 0,
building, wells;
pl.LXXII
1978,
3, 1972,
462 bath-
75-100, disused
18,
350-2.
1987,
3 45;
20,
1989,
315; 21,
comments
199 0, 3 50-2;
Frere,
22,
1967, 246.
14, 1983,
1991,
284-7,
about
water
start
of
which
the
has
aqueduct
not
investigations.
is
as
been
JRS 30,
suggested
175-6;
27-9,
drainage system.
PDNHAS 77,
90,
1969,
171ff,
pp.172-3,
of
where
the
at
aqueduct
next
to
the
early
by
35,
ed.,
excavation
by
substantiated
1940,
Hanson assumes
1955,
C
80.
but
archaeological
Moule 1906,
133-4;
the
engineers,
later
1945,
that the
78,
Green
Poundbury
1956,
80-3;
describes
cemetry
2nd
and
his
the
He concludes
that the information obtained from sherds shows a primary phase of the
1st c. AD with a channel of a flat base and sloping sides and lined
with clay in the lower portion
a second phase when it was cleared out in the later 1st century to a
new U-shaped profile,
lining,
PDNHASFC 22,
80-3,
1901,
article by R A Coates
46,
1925,
1-
13, article by P Foster and the plot of his speculated line towards a
source for the aqueduct;
59, 1938,
13.
description
of
(p.586, pi.221),
the
aqueduct and
includes
the
map
numbered
and P
of
on
the
Royal
Currently
Engineers
in
c.
1925, based
the Liverpool
datum.
last
existence
attested
into
section
the
plantation
TIRCGLL 1983,
5th c.
of
the
aqueduct.
for
about
He
250m,
RIB 188-90;
occupied
has
when
established
it
its
disappears.
(Durotiges) .
91,
197 0,
182-3,
a drainage
376
ditch
leading
into
quarry;
Dorn, Gloucestershire.
Cox
1720,
31-41.
postscript.
Haver f ield
MeWhirr 1981,
60-2.
2 wells.
1963,
St Joseph
18-24.
1901.
Nash
1781,
Oswald 1963,
I,
18-24.
1961,
119-135.
VCH Worcestershire
1,
II,
RCHM 1976,
1962,
12-
194-5;
TIRBS 1987,
(Dobunni).
1782,
TBGAS 81,
199-221.
101;
30,
82,
small
Dover, Kent.
Archaologia 93,
1970,
45-6,
1949,
65 Ant.
32.
Arch.
J. 126,
Itin.
473.2,
473.5
bath-ouse,
earthenware 2+,
aqueduct
cistern;
366;
16,
23, 1992,
1973.
was
13,
1985,
306.
angle.
17,
KAR 23,
by
tank
1986,
1971,
14,
426;
chalk
block
20,
have
7,
pipes
15,
323-5;
1971,
5, 1974,
two+,
8,
11, 1980,
334-5;
1989,
2,
197 6, 376;
Philp 1981,
401;
1984,
21,
1977,
12,
217-23,
199 0, 3 64;
76, 79-80.
Philp
bothin ceramic
must
wood
473 vicus;
1983,
76-86.
Britannia 1,
322, 332;
channel,
283 sewers;
comments
Distribution was
distribution
stone-lined
6, 1975,
Stephens 1985b,
supplied
393 annexe;
315;
78-100.
1982;
1969,
traces
the
N-W
lain close to
approaching
this
angle.
One
of
the
have been capable of delivering at least 45.6 cubic metres per day'.
SxAC 107,
45, 3rd
c.
fort of the
'Litus
Saxonicum'. (Cantiaci).
Proc. 60,
1938,
66-72;
115.
69,
1947,
RCHM
42-4;
70,
1948,
29-59;
75,
620:
B Calkin.
Downton, Wiltshire.
Britannia 2, 1971,
1990,
353
settlement.
Wiltshire 1, 1957,
64.
Scott
1993,
WANHM 55,
201, SMR
1953-4,
377
SE
176-8;
4, 1973,
301,
58,
WZ
163; 21,
71.
1961-3,
VCH
303-41,
bath-house.
TIRCGLL 1983,
flourished
(Belgae) .
Drayton(1), Leicestershire.
EMAB 1978,
10.
SP 831 931.
8-11.
Scott
(Corieltavi).
Drayton(2), Leicestershire.
Britannia 20,
1989,
286;
22,
191, 245;
24,
1993,
289,
idem.
1995,
Condron
1996,
338.
1989, 7-17;
161-2.
7-17.
Scott
TLAHS
64, 1990,
1993,
111,
52, 1976-7,
13 near
Burnham 1993,
SMR SP 89 SW Q,
98;
Fig
55,
1979-80,
66, 1992,
LE 27.
95-7; 63,
1972,
287-8
well
317,
(Salinis) ; 2, 1971,
181 n.61;
4,
1973,
330-
'the circular
the
439;
shaft';
10,
5,
1979,
1974,
299;
430;
12,
6,
1981,
1975,
337-9;
34,
1963,
17-8;
42 ff ; 20,
55-8.
64,
1973,
1977,
52ff.
15,
12-3;
17,
1977,
1984,
396-7;
448;
19,
9,
1990,
1967,
1974,
8,
PNRB 119-20.
WMANS 10,
16,
248;
7;
14,
49-50;
1959,
1-3.
TWAS 1925,
1971,
17-9,
plans;
18,
VCH Worcestershire 1,
1975,
1901,
58;
212-6.
19,
15,
1976,
Occupation
(c.
5ha)
TIRBS 1987,
30,
and villa,
(1)pre(SO 898
639) .
Ducklington, Oxfordshire.
Britannia
1989,
5,
1974,
386,
297 settlement;
22,
436;
1991,
7,
1976,
258.
378
336,
well,
stone-lined;
1974,
42.
20,
CBA
Gr.9
Newsletter
5,
1975,
41.
Oxoniensia
40,
1975,
171-200.
Scott
185,
154.
NE 9, WS 36.
materials.
VCH Sussex 3,
Scott
1935,
24.
Dunsby, Lincolnshire.
Britannia 10,
1979,
293,
well.
EMAB 1978,
23.
Whitwell
1982,
224,
well.
Dunstable/ Bedfordshire.
Bedfordshire Arch.
J.
7,
1972,
21-34.
Britannia 1,
1970 42,
49,
51
Ant. Itin. 471.2 Iter II, 476.9 Iter VI, 479.7 Iter III; 2, 1971, 267;
8, 1977,
399-400;
11,
1980 17,
6.
TIRCGLL 1983,
46,
minor settlement;
55-66.
Mathews &
VCH Bedfordshire 2,
occupation
1st to
4th
{Catuvellauni).
12,
1981,
321;
13,
1982,
item 39,
101; 16,
comments
1985,
197-207.
DAES 1978,
26.
5;
comments
ornamental
idem.
'...;
fountain
at
in
1990,
96-8.
Duntocher
the
on
Roy 1793,
the
extra-mural
pl.xxxv.
Antonine
Wall,
bath-house
of
Stephens
where
shows
206.
an
that
CSIR 1.4,
Britannia
house
16,
with
1944,
1985,
small
276.
1-21;
29, 1951,
TIRBS 1987,
bath-house,
203-12;
31;
cold plunge,
(Brigantes) .
Duston, Northamptonshire.
379
31,
1953,
116-26.
2 heated
rooms.
small
Two
Assoc.
Archit.
Condron
Soc.
1996,
Rept.
327.
8,
53-61.
Northamptonshire
1885,
Swan 1984,
Britannia 7,
SMR
4946.
fiche 519,
1976,
334.
Dryden
1885.
538, wells.
Woods
TIRBS 1987, 31, major Iron Age and Roman settlement, wells,
Occupation c.
(Catuvellauni).
Earith, Cambridgeshire.
Britannia 6, 1975, 250; 11, 1980, 375; 12, 1981,
194.
to 4th c.
256-7.
132.
Phillips,
1970,
(Catuvellauni).
11,
1980,
290-2.
1,
TIRCGLL
1906,
1983,
5-6.
329-31,
46,
villa,
Collectanea
Scott 1993,
item
51,
Antiqua
167,
very
hypocausts,
2,
1852,
SMR 53911,
beautiful
coins
SO 22.
mosaic
mostly
51-2.
from
VCH
Fig
4th
88.
c.
(Durotriges) .
Berkshire.
290-2,
was closeby'.
1993,
EA 9-12.
SxAC 2,
1848,
3,
121.
1972,
9,
PSAS 90,
settlement.
1956-7,
JRS 47,
93-101.
1957,
200-1,
RCAHMS Peeblesshire
plan;
1,
51,
1967,
Easton, Suffolk.
JBAA 8, 1853,
159-60;
10,
1855,
383.
(reprinted
Eaton, Leicestershire.
Britannia 11, 1980,
367.
380
63.
Whitwell
Eaton-by-Tarporley, Cheshire.
Britannia
12,
1981,
383;
13,
353
villa,
bath-suite
and
drainage
(Cornovii).
Ebchester, Durham.
Arch. J. 142, 1985, 216-36.
Trans.
R.S.23,
date?,
1129-32:
burried
channel
is
in
at
the
Stephens
south
aqueduct
that
Phil.
supplied
the grass.
known
Hunter C,
probably
...an
undated'.
1985b,
angle
(Steer
217,
226,
(Hutchinson
1938,
228;
'A stone
1794, 434),
Jarrett
1960,
...
200);
the
1982,
cold
210.
plunge
JRS 49,
bath
to
1959,
the
127 bath,
latrine
and
51,
other
1961,
32, villa,
'stone-lined drain
drains,
186.
and
PCNFC 36,
near a spring,
bath-block with 9
(Dobunni).
Cant.
1976.
1973,
78,
1963,
Britannia,
322;
Jarrett
&
5,
Dobson
JRS 53,
56,
217,
1956,
2, 1971,
1974,
village.
459,
79,
1964,
121-35;
114, 170-93,
bath;
1965,
6,
105-8.
Scott
JBAA
1993,
TIRCGLL 1983,
196 bath;
4,
1849,
1972,
17,
81ff,
103,
47,
80,
286-8, 297;
1975,
224.
Roman buildings.
125-41;
1986,
4,
428.
probably
220,
KE6; KE 7, TQ 73
351;
224, 226;
60:
various
1st to 4th
(Cantiaci).
1,
4th Ed.,
1901, 211,
1991.
Scott,
3 stone-lined
1993,
or qoined
89,
HE10.
wells,
VCH
drains.
(Dobunni).
118;
Flavian period.
14,
1983,
172-6;
16,
1975,
264-5 bath-house,
1988, 428-
1987,
154.
West Sussex.
Britannia 7,
1976,
373 ; 20,
290, bath-house
186,
SMR 0817,
1989,
230;
1975,
21,
1990,
Scott 1993,
58-66.
SxAC
67-85.
to late 3rd c.
TIRCGLL 1983,
{Silures).
Emberton, Buckinghamshire.
Condron 1996, 347.
Empingham, Leicestershire
AEx 1969,
59; 1970,
69.
Britannia 1, 1970,
excavated
1966,
46.
6, 1975, 246.
1969-71,
McWhirr
286; 2, 1971,
1972,
75,
TLAHS
60, 1986,
314;
7,
1976,
351 well;
9,
Scott
NW P, LE 30: SK 94
1970,
EMAB
up,(1993).
1993,
258 a well
08.
3-18, 258.
(Corieltavi)..
Greater London.
11, 1980,
381;
25,
1994,
Engleton, Staffordshire.
Staffordshire Record Soc.
1969,
1938, 267-93.
winged
corridor
villa,;
Webster 1975,87-9.
bath-suite;
183-4.
TIRBS
occupation
2nd
Rivet
1987,
to
33,
4th
c.
(Cornovii) .
Epperstone, Nottinghamshire.
EMAB
1959,
1993,
154,
1963,
13-4;
SMR 01848,
134; 54,
65,
1961,
2nd
c.,
house.
1961,
1064,
1963,
15;
1964,
NAR SK 64 NE 3(2),
159,
6. Whitwell
timber;
14;
later
162;
1982,
rebuilt
55,
25;
NT 11,
1966,
bath-house.
1965, 207.Todd
111, 323.
TIRBS 1987,
in stone foundations
(Corieltavi) .
382
40-1.
1973,
33,
Scott
JRS 53,
33.
TTS
villa, early
Ewell, Surrey.
Archaeologia 32, 1847, 451-55.
23,
1992,
PSAL,
304.
Coles
(2nd ser.),
85-6.
& Simpson
1, 309-13.
9-46;
1968,
(eds.),
Pemberton 1973,
69, 1973,
1-26.
284.
idem,
85-6.
2(4),
Smith 1987,
274-5,
361-3.
settlement,
Warne
1859-61,
1847.
1974,
61.
1-26;
Diamond
309-13.
TIRCGLL
6 wells,
VCH Surrey 4,
1983,
49,
large
{7Regni or Atrebates).
Ewhurst, Surrey.
SyAC 65,
1968,
Farm, Ewhurst,
1-70.
Scott
Cranleigh.
1993,
176,
SMR 447,
TIRCGLL 1983,
SY
12,
at Rapsley
{Regni).
Exeter, Devon.
Arch,
of
Exeter,
Bildwell 1980,
53-4;
idem.
1983-4,
bath,
1979,
2-4,
52-3,
1-66,
p.3,
Fig
Figs.28, 30,
3,
aqueduct
of
AD
100-1.
iron collars is shown as the water supply for the bath which required
a daily consumption of water of about 320,000
1970,
452;
60-1 Ant.
6,
1975,
Itin.
483.8,
276; 7,
55-75,
to early
324-6;
11,
165,
5th c.;
1980,
3 89;
72, 441-2;
21,
199 0,
3 48-50;
Fox
1973,
166-9.
detail
18,
22,
1981,
1987,
Hanson
comments
415;
1973,
5,
1974,
Legionary defences
c. AD
1978,
1982,
459,
3 82;
313;
476;
14,
281-2;
the
water
23,
111-6,
Roman
473-4; 20,
1992,
366,
290-7;
item
aqueduct
as
its
1979,
1983,
32 0-3,
276,
303-5;
1989, 313-4;
25,
41,
10,
1994,
286.
discusses
in
existence
is
Stephens
1985b,
60). These denote an aqueduct supply. The source was perhaps two
springs
situated
clay,
9,
4,
358-60,
3 58; 13,
1970,
of
486.17;
Britannia 1,
1991,
the uncertainty
item 9,
1977,
318;
278-80,
8,
12,
486.8,
litres.
c. 1km
to the
(Tucker
1858);
N-E,
the
where
the medieval
basin overlies
catchment
4.2m deep
basin was
platform
of
tapped, the aqueduct was probably a pipeline, even though the civitascapital was part supplied by a leat
335-43,
Fig.151,
33 6 shows
Wacher 1995,
383
TIRCGLL
1983,
49-50,
established
settlement
baths,
a legionary
c.
AD
55
developed
fortress
and
c.
probably
AD
80
evacuated
and
continued
c.
until
AD
(15.4ha)
75;
the
was
civil
5th
c.
New
(Dumnonii).
Moore 1988.
1987,
41-66.
TIRBS 1987,
33, villa,
44-5.
stone and small bath-suite added. In early 3rd c., complete rebuilding
in stone and baths modified. Destruction by fire in 4th c.
{Iceni).
Falkirk, Stirlingshire.
Britannia 3,
1972,
18,
248-62.
51-3,
303; 8,
1977,
1994, 255.
364;
12,
Macdonald,
RCAHMS, Stirlinghamshire,
Figs.35,
36.
CSIR
1.4,
320;
1934, 214-6.
1,
1985,
1981,
1963,
No.72.
99.
13,
1982,
PSAS 111,
Robertson,
TIRBS
1987,
34,
RCHM 1976,
SMR 2562,
GS
(Dobunni).
Farmoor, Oxfordshire.
Britannia
6,
197 5,
Unit Rep.
2,
1979
item F43),
F1050)
27 9-80.
(CBA Res.
0.80m diam.
Fig.19,
and
Lambrick
Rep.
1.6m deep;
stone-lined,
32),
1.06m
& Robinson
(1) well,
(2)
(3.5ft)
1979,
stone-lined
two wells
&
Oxford Arch.
(Rep.2,
(items F1046
1.37m
(4.5ft)
&
deep
28 .
Farnham(l),
Suffolk.
384
SMR 1629,
BU
Moore 1988.
Farnham(2), Surrey.
Britannia 2, 1971,
1939,
61-118.
JRS 37,
1946-7,
50,
4th c.
villa,
Lowther
175.
1955,
47-57,
Scott 1993,
176-7,
aqueduct
SMR 1715,
detached bath-building,
aqueduct;
TIRCGLL
occupied
3rd
and
depth 28.2m,
av.
(Regni or Atrebates).
Farnham(3), Dorset.
PDNHAS 104,
1982,
179;
106,
1984,
116:
Roman well,
66; 1976,
85.
Britannia 7, 1976,
817,
lead-pipe.
Rivet
1969,
376;
JRS
134-50.
channelled drain.
1932,
8,
Cant.
1977,
15,
424.
1925,
Scott
61,
245;
1993,
1948,
180-2;
76,
TIRCGLL
1983,
18,
1928,
104, KE
50
1961,
Gentleman's Magazine 1,
208;
39,
30-33, baths
corridor
villa,
inJope
(ed.). 34ff,
188.
1866,
1949,
110.
113-4.
occupation.
Arch.
VCH Kent 3,
bath;
5th
c.
(Cantiaci) .
Farnworth, Cheshire.
Hanson 1970, 399-400.
O'Neil 1961,
well,
stone-
lined .
1,
Newsletter
1927,
13,
478-9.
1983,
Britannia
127,
18,
1987, 324-5.
CBA G r . 9
bath-house,2ndc.
origin.
4f.
Scott 1993,
Oxford Arch.
1294, OX
No.31,
1986,
1-48,
bath-house
pp.9-14.
Gregory
1982,
369.
1979,
1971,
21-5,
28;
123 n.7;
12,
3,
1981,
1972,
21-4,
385
69-70,
27.
79,
86-7;
Hanson 1970,
5,
1974,
436;
196-200,
366,
Fig. p.23 0, comments 'Part of the aqueduct channel which seems to have
contained a water pipeline
west
of
stretch
either
the
of
fort.
The
low-lying
springs
pipeline
ground
to the
was
between
south-west
required
the
of the
in
source
fort
order
of
to cross
the
aqueduct
or perhaps
the higher
reaches of the Fendoch Burn - and the plateau on which the fort was
sited',
item 42.
PSAS
1985b,
112,
224,
(Richmond
138-40,
discussion of
McIntyre 1938-9,
Sc
pipelines', 224.
138-40).
1991,
222.
Camden
1951.
Davies
172.
RCAHM Flintshire,
19A-F,
e.g.
CL
1949,
Note:
Ffrith,
1586,
226-38.
8-14.
Frith,
JRS 58,
Wales, 1912,
1968,
55-6.
Nash-Williams
Scott 1993,
thereare different
spellings
1969,
of
this name,
(Deceangli).
Fifehead, Dorset.
PDNHAS 24, 1903,
1883, 66-70.
2 018 013,
172-7;
PSA
(ser.2)
TIRCGLL 1983,
50, villa,
8, 1881,
543-5; 9,
(Durotiges) .
Fillingham, Lincolnshire.
Archaeological Notes 1959-60.
Scott 1993, 120,
Lincolnshire Magazine 3,
LI 58, bath-house.
1936-8,
91-2.
Finchingfield, Essex.
Coverton J G, 1933.
Coverton J G, 1937.
Rodwell in
VCH Essex
3, 1963, 129-30.
TEAS
TIRCGLL
(Trinovantes).
1954-55,
204-5;
Black 1987,
154.
6,
1955-60,
101-2.
Antiquaries J.
8,
286, well,
79m deep.
386
Scott 1993,
1928,
173.
NAR TQ
10 NW 39,
WS
42.
Sussex Notes
& Queries
15(7),
1961,
250.
(Regni).
Fingringhoe, Essex.
Britannia 1, 1970,
1989,
181-2; 2,
and 9.73m
1983,
51,
VCH Essex 3,
deep; ii)
fort
8, 1977,
87; 20,
Essex Naturalist 1,
?,
181.
1.1m diam., of
and naval
192-3;
base
unknown
of
Aulus
depth.
TIRCGLL
Plautius, AD
43.
(Trinovantes).
84-6.
Britannia
2, 1971,
materials; 3,
1,
168, 174,
1970,
180f,
90-1 military
buildings;
1981,
364;
333;
1986,
424-4;
14,
1983,
18,
1987,
15,
353
10, 1979,
1984,
20,
base
to
c.
189,
Drains: (xvii,
1976, 164; 8,
aqueduct,
423f)
21,
WS 25.
TIRCGLL
to
4th
c.;
19,
1990,
12,
189;
17,
1988,
406
360 mosaics;
Margery 1971,
117-21.
the pipes
Site dates
timber-lined.
1983, 51,
palace
160-1;
1985,
43
320-2;
supply of palace;
palace,
AD
29; 16,
311 glass
192f building
328 Fig
1989, 11,
Cunliffe 1971a.
ditches N of road,
Chichester),
186,
related
183,
1977,
from c. AD
Scott
1993,
built
AD 75; burnt
(under
military
c.
AD
280.
(Regni).
Flamsteed,
Kent.
19-20,
c. 3 0m (100ft) deep and between 2.1m and 2.4m in diameter, cut through
about 18m of upper strata and continuing for more than 10m into chalk.
Fletton, Huntingdonshire.
Peterborough Museum Record Nos.
item 1,
RCHM 1926,
the Nene,
was
spot
mile
1/2
1520, 1627.
occupied early
S
of
St
Margaret's
387
Church. . .A
particularly at a
little
south
was
circular,
steined
diam...'.
stone-lined well,
VCH Huntingdonshire
1,
1926,
249.
in external
Whitwell
1982,
233,
well.
1925,
209; 38,
13,
291.
c.
J. 5,
1982,
1926,
211;
Bromehead
AD
3 50;
Arch. Cant.
45-50,
20,
10, 1876,
1989,
193-200,
41
p.45.
(TR 22,
Britannia
325; 21,
1990,
drains, waterpipes,
(see Winbolt
Scott
for details
'Guide1.
1993,
Winbolt
37,
6,
1975,
3 64; 22,
1991,
complete
or drain-pipes
of piping.
sections
105,
1925,
JRS
14,
KE 36.
109.
1924,
242, 246.
VCH Kent,
TIRCGLL
Rivet
3, 1932,
1987,
of
found in room
37);
bath-house,
earthenware
27',
65-7.
Burough of
1969,
114-5,
57-64.
p is.xxi-xxii.
51, corridorvilla,
bath.
(Cantiaci) .
Cant.
223; 89,
83,
1968,
127;
South London.
84,
1969,
39-77; 86,
1971,
306-7, bath-house;
5, 1974,
239; 88,
1973,
Britannia 4, 1973,
446; 6, 1975,270; 9,
197,
471; 10,
1979,
336 .
Foscott, Buckinghamshire.
Gentleman's Magazine,
303.
1838,
p t .1,
RCHM Buckinghamshire 2,
baths,
tank.
Records
302;
1913,
115,
of Buckinghamshire
1841,
p t .1,
No. 134,
5,
81;
item 1,
1885,
p t .1,
1843,
p t .1,
lead pipes,
3 55.
Scott
Foxcote, Buckinghamshire.
RCHME
Buckinghamshire 2,
1913, 115.
VCH Buckinghamshire
TIRBS
2,1908,
7.
(Catuvellauni).
Frampton, Dorset.
Antiq. J. 57, 1977, 312-3.
11,
150.
1980,
298;
22,
1991,
Britannia 2, 1971,
156;
23,
1992,
123-4.
Frilford, Berkshire.
388
(Durotriges).
Arch. J. , 42,
opp.
342,
190,
temples;
1987,
1809,
well,
47.
417-85;
45,
1880,
405-10;
11,
1980,
396,
VCH Berkshire,
1906,
1897,
Britannia 7,
bath-house;
1,
54,
13,
207-8,
1976,
1982,
Fig.
340-54 Fig.
175,
305-9,
TIRCGLL
178,
368;
18,
1983,
52,
(Atrebates/Dobunni).
Friskney, Lincolnshire.
Whitwell 1982, 234, TF 45 NEC.
9,
1983,
1971,
1978,
316-7;
275 villa;
455;
10,
1979,
3,
1972,
319;
11,
1985,
1980,
300;
7,
1976,
357;
384;
12,
1981,
17,
1986,
8,
412;
1977,
355;
15,
1984,
314;
shows
16,
339;
18,
14,
1987,
22,
Glevensis 11,
1977, 24; 14, 1980, 24-5 + Fig; 18, 1984, 53: report a '...square 1.5m
timber-lined tank with a stone slab floor and a drain emptying to the
south as did others... These drains ran into a ditch (No.35), which had
at
least
two
distinct
1964,
183;
alignments,
JRS 50,
55,
1965,
dated
to
1st-
216;
56,
1966,
212.
3rd
centuries'.
182; 53,
McWhirr
1963,
1981,
143 ;
83-7.
Scott 1993, 72, SMR 5198, GS 50 bath-suite, not earlier than c.AD 275.
TBGAS 77, 1959,
57 & 58.
23-30;
89,
1970,
early 5th c.
(Dobunni).
Herts.
1852,
Arch.
315,
Rev.,
394-7;
1971,
JRS 55,
extensive
additions
swimming pool;
59,
1969,
1965,
56,
221.
211-2,
include
1966,
Neal
56.Britannia 5,
bathing pool
1974,
added by AD 325.
At
reconstructed
1974,
out-buildings, stockade,
71,
35 1853,
1967,
'Winged-corridor
bath-house
and
bath
in
187-8;
villa
enclosure
4th
58,
(AD
c.
and
1968,
194;
140-60)
with
ditches.
Earliest
(SMR 186
Sc
SMR 0088)'.
Scott 1993,
389
SMR
517,
HT 33,
53.
SMR 0088,
HT 35.
15.
PSA 1849,
1987,
1977,
339-50,
Settlement
pls.xix
203,
Fig. 8 .
Britannia
xx, bath,
period ist
1973,
323;
5,
1974,
458;
8,
at Garden Hill,
&
4,
Sussex:
drains,
Interim Report
lead-pipe
from bath;
1968-76',
Fig.5 &
five periods,
3rd
c., and 5th period end 2nd c. or early 3rd c. ore brought
in for smelting;
9,
1978,
467,
481;
10,
1979,
333-4;
11,
1980,
398-
51,
v(4)
400, Fig.23.
1972, 310;
1975,
article by T C M Brewster,
104-116,
4, 1973,
28,
well.
Current Arch.
shows picture
of the well,
at a depth of 5.5m after loose gravel sides had fallen in (iv, 281).
1975, 175-81,
Occupation
Fig.l;
1977,
1986, 92-5.
246-7, Fig.5 at
Gardiner 1976,
p.25,
168,
pl.l5A.
172.
Greene
Gayton, Northamptonshire.
Archaologia 30, 1844,125-31.
Nat. Hist. Soc.
Britannia 22,
1991, 252.
J. Northants
S-W, 4, 1982,
Scott 1993,
143,
NH 45.
TIRBS 1987,
SMR 574,
217.
(Catuvellauni).
235-6.
166-209.
15, 1984,
Gregory
1982,
362-364.
47.
EAA
SMR 3743,
bath-suite,
...Detached building,
(Iceni).
390
Ward 1903.
Ward 1909,
Gestingthorpe, Essex.
Draper 1985, villa,
bath.
Rodwell in Todd
133-4.
Scott
(Trinovantes).
1971,
26,
145
1995,
of a water channel,
fort;
45.
3,
1972,
11;
Hanson 1970,
5,
366,
1974,
154 Antonine
comments,
II
1983,
27-8,
plan
and
"a
covered
36,
Rav.
Cos.
1951,
60;
pipeline"
have been
43,
found
TIRBS 1987,
JRS 41,
water-channel or
126-9.
224
item 21.
duct
10-1.
Gloucester, Gloucesteshire.
Antiq.
J.
52,
21-58.
27 5;
1972,
24-69;
54,
Britannia 1, 1970,
3, 1972,
1974,
57, Ant.
339-40; 4, 1973,
9, 1978,
11,
12,
346-402,
aqueduct;
Itin.
169, 202-3,
8-52.
1981,
485.4
309;
455-6;
20;
(Clevio); 2,
1971,
1975, 88-90,
192;
6,
10, 1979,
13,
1976,
1982,
177, 319-21;
413 ; 14,
1983,
316; 15, 1984, 314, 333; 16, 1985, 202-3, 300; 17, 1986, 414, 429; 18,
1987, 339; 19, 1988,
275-6;
23,
Leggatt,
44,
1992,
1968,
comments,
colonia,
in
469; 20,
295.
11.
Fullbrook-Leggatt
Glenvenis 23,
1952,
1989, 3-15.
16-7.
1942,
1962,
197
the
form
system
39-52;
item
1985b,
item
(Hurst
1974,
Fullbrook-
of
number
was
of wooden
pipelines, has
item
8,
47,
probably
40;
1957,
57,
comments,
31;
233
1967,
'The
Hassall
been
near
Matson
1991,
&
item 31;
195.
51,
colonia was
Rhodes
391
1974,
Robinswood
1961,
McWhirr
on
196
1981,
supplied
26-7).
item 32;
52,
21-58.
Stephens
by
aqueduct
an
This
may
have
seems
rainwater
large
for
tank
and
II,
III, well;
Wacher 1995,
87,
150-66.
1969,
(Wilson 1967,
could
easily
have
113ff;
93,
1974,
31,
195);
this
been
347-58,
Fig 2,
aqueduct,
TIRCGLL 1983,
the
well.
24 .
81-2,
86,
237.
TIRBS 1987,
36,
minor settlement
(c.
(Corieltavi).
Black 1995,
Britannia
1,
1970,
41-2,
287;
4,
CA
84.
Smith
7210)
early
2nd c.
during
to
well
2ndc.
7,
118,
Mansio
1976,
fort
c.
it must
well,
have
1.5m diam.
Street.
mansio;
in
10,
38, SMR
370,
(TL 2590
later
1st
c.
to
AD 3 00,probably
had
bath
and
TIRBS 1987,
36,
Development
fire
190-1.
well;
Scott 1993,
destroyed
on Ermine
Hadrianic
333-4,
Occupation
a mansio,
supply;
151 Fig.20,
Swan 1984,fiche
1.5m diam.
c.
'massacre' . Since
and
325-8;
1987, 182-6.
period
1973,
91,
late 4th
in Flavian
late3rd
c.
(Catuvellauni).
?; 58,
1968, ?.
PCAS 61,
1968,
19-43.
Scott 1993,
37, SMR
Gorhambury, Hertfordshire.
AEx 1972, 54-5; 1975, 61.
121;
4,
1973,
299;
5,
1974,
401;
9,
1978,
445;
10, 1979,
437;
6,
305-6;
1975,
11,
181-258,
Newsletter
1, 1983,
51, 1961,
504,
HT
180-2.
52,
Neal 1980.
bath-house.
Neal,
Todd
et al, 1990.
1978,
8,
1980,373-4 Figs.11,
1981, 165,
20.
339-40;
33-58.
1994,
12; 12,
104.
1983, 115-121.
Scott 1993,
TSAHAAS
1977,
1961,
CBA
JRS,
95, SMR
21-30.
TIRCGLL 1983, 54, winged-corridor villa with bath; from late Iron Age,
succeeded
by
timber
house,
burnt
in
1st
c.,;
early
2nd
c.
masonry
1987,
333 ;
15,
155, No.
1984,
132.
328-9;
Rudling 1983,
19,
1988,
481,
45-7.
Britannia 14,
pl.xxixB.
SxAS 54,
1983,
1988,
10,
1970,
42,
69,
Ant.
Itin.
aqueduct.
469.3
(Bovio) ; 12,
1981,
333;
236.
location,
Waddelove 1984,
255-7,
must have been a Roman tile and pottery factory. From the information
I conclude that the site must have been at approximately SJ 445
near
suggests.
For
it to have
been
at
514
Holt
would have required a crossing of the major River Dee to get to Deva
(Chesters) from Holt.
Grandford, Cambridgeshire.
Britannia
Potter,
1981,
79,
81,
Roman
drain;
85-93,
fort,
16,
126ff,
1985,
190.
Phillips 1970,
originally
Figs.4-6,
pl.v.
settlement
TIRBS
197.
probably
1987,
37,
rural
1967,
151.
Head
(Iceni).
160.
14,
Arch. J. 124,
Records of Buckinghamshire 1,
1946,
316.
RCHM 1,
1912,
165.
Thopson
1848-55,
39,
140;
Scott
1993,
27,
404-5;
15,
1984,
2,
1857,
1957.
16,
8, 1977, 360,
1985,
429;
260-3 ; 19,
11, 1980,
1988,
181; 20,
1989,
263 ; 25,
271
bath,
1994,
252
121;
Britannia 1, 1970,
5, 1974, 11,
27;
10,
1979,
116,
184,
Coder
393
119-24,
1961, 49-50.
197; 3,
48-9;
12,
1972,
1981,
102;
41,
1951,
127-30, plan,
section;
42, 1952,
1953, 115-7; 44, 1954, 92-3; 47, 1957, 212-3; 48, 1958,
137; 51, 1961,
Scott
1993,
59.
VCH
Wright
(2.43ha)
175; 53, 1963, 134-5; 56, 1966, 204-5; 57, 1967, 183-4.
112,
LE 38.
Stukeley 1724,
Leicestershire
1684,
96; 43,
1,
1907,
(reprinted 1974),
with
vicus, on
88-99.
35.
Ermine
24.
Whitwell
Street;
reduced
1982,
in
239-40.
Claudian fort
size
c.
AD
70
(Corieltavi) .
(ed.),
interim
Joseph 1961b,
bath.
reports,
13-17.
TIRBS
1st,
Todd 1973,
1987,
37,
1951,
78,
villa,
89,
92,
coins
2nd, 1954,
3rd,
94. Whitwell
from
Marcus
1961.
1982,
St
102,
Aurelius
to
{Corieltavi).
180,
temple;
Collins
Arch.
126,
1978,
10,
Arch.
& Hist.
11-18,
183; 3, 1972,
Soc.
290-3;
309-11;
12,
G r . Bull.
2.
19,
1988,
PCAS 75,
348-50,
Fig
Crossan
264-5,
12;
23,
1992,
et al, 1990,
item 19
1988,
Britannia 1
3-41.
11-8.
1972,
21,
290.
Essex
1990,
290-3.
VCH
Essex 3, 1963, 72-88, Fig 22, p.77, 2 wells. TIRBS 1987, 37, suspected
Neronian
vexillation
fortress
(14.15ha);pre-Roman
shafts;
civil
356-78.
Scott 1993,
ES 15, comments,
'A
Ael.{2)
24,
1903,
205,
soldier;
destruction
12,
NZ 717 688,
1981,
of
268;
19-64;
Arch.Ael. (3)
5, 1909,
158-67 baths;
4,
1973,
200,
3rd
c.
inscr.;
tombstone
18,
1987,
318 at
in advance
394
of
560mm
comments:
of an open leat,
consisting
fort from a source on the Caw Burn only 2.25 miles away from Aesica in
a direct
line'.
JRS 30,
1940,
161-4;
35,
1945,
80-1 aqueduct,
plan.
'The fort
was supplied by a 9.65km long leat fed from Haltwhistle Burn/Caw Burn
4.4 km to the N-E
1945,
80).
This
(Bruce, 1851,
will
have
257-62,
fed an
The
leat
Hadrian's
Wall,
extramural
after AD
distribution
tank
from
is undated'.
built
Fig 99;
TIRBS
128,
RIB
1987,
37,
1736;
fort
(1.38ha)
overlies MC43.
on
Civil
(Brigantes).
49.
Britannia 2, 1971,
272;
1972, 333,
356;
4, 1973,
'A 3rd c. well c. 7m deep had a timber lining made of oak planks
41,
1988,
'Excavations
oak
planks
17,
1986,
at Great Dunmow,
c. 6.9m deep,
3,
on gravel
442; 25,
Essex',
1994,
a well,
280.
Period
Sll).
Construction
shaft
300mm
50mm
dovetailed
at
the
corners.
It
was
originally lined with flint and green clay, which collapsed inwards as
the wood
1975,
started to
85-101.
rot',
(p.11).
125.
Wickenden 1988,
& Rowley
11.TIRCGLL
settlement. {Trinovantes) .
7,
1976,
41ff ; 10,
337-8;
1980;
12,
12,
1981,
1982,
344-5.
78-81.
CBA G r .9 Newsletter,
Mynard 1987,
97-104.
6,
Scott
1959,
118;
50,
1960,
224-5,
corridor house,
Fig.26.
Scott 1993,
38,
CA 90.
TIRBS
of 2nd to 3rd c.
395
(Catuvellauni).
Scott
1993,
159,
SMR
Oxfordshire 1, 1939,
bath.
2336,
OX 27.
310-11.
Oxoniensia 31,
TIRCGLL 1983,
1966,
153.
VCH
(Dobunni).
21,
1990,
Saxon
settlement.
342,
period
Roman
(?),
well,
Howell's
timber-lined;
Farm,
TL
22,
857
1991,
098,
262,
Iron
Age
CBA Gr.7 Bulletin 10, 1989, 8-9; 11, 1990, 17-8, plan.
161-66.
1966,
212.
Scott
1993,
JBAA 1,
73,
1961,
1846,
system,
275; 73,
1954,
53,
1963,
GS
53.
54,
92-3.
1964,
54-65;
RCHM 1976,
56,
60-1.
1907,
246;
57,
1935,
141;
McWhirr 1981,
probably a sewer.
5-69.
1983,
186;
56.
SMR 423,
drainage
TIRCGLL
Archaeologia 19,
(Red Portfolio).
courtyard
villa,
late
Witts,
1st
c.
to
5th
c.;
baths.
(Dobunni).
21,
1891,
48-52.
Lincolnshire.
Arch.
J.
Rivet
48,
1891,
Scott 1993,
49,
187;
121,
1892,
258-62.
Whitwell 1982,
242.
(Corieltavi).
3,
1972,
?mansio;
(1.48ha),
reconstruction.
possible bath.
42,
work
6, 1975,
late
235.
Flavian
Extensive
on
or
civil
fort
c.
AD
Hartley 1971,
Trajanic
205-7;
58.
5,
413-5,
TIRBS 1987,
foundation,
settlement
1974,
early
(vicus),
3rd
?mansio,
38,
c.,
i.e.
(Brigantes).
Grimstead, Wiltshire.
Scott 1993,
1957, 75.
202,
Occupation,
Grimston, Norfolk.
Sumner 1924.
VCH Wiltshire 1,
Britannia
16,1907,
18,
1987,
219-27.
330.
Gregory
Scott 1993,
1982,
133,
362-4.
Norfolk Archaeology
TIRBS 1987,
39,
{Iceni).
1964, 25.
Hacconby, Lincolnshire.
LAASRP
8 (n.s .)
1959-60,
44,
62.
the
1993,
Stukeley
Car
Dyke
121,
TF
1724, 8.
& TF 12 SW F.
in
07
the
24,
Phillips
91,
256-7;
South
NAR
Lincolnshire
TF
02
1970,
SE
1.
Wash
of
Simmons
Two
running
the
Fens.
1975,
33-6.
{Corieltavi) .
Haceby , Lincolnshire.
Antiquity 3, 1929, 486.
7.
Scott 1993,
1818,
634.
124,
SMR 7712/7789,
Greenfield
1971,
LI 116.
29-57.
Gentleman's Magazine 1,
Hindley
1961,
LAASRP
Transactions 35,
428-32.
9(1),
Whitwell
158,
Stukeley 1719,
1982,
16.
280.
Wooler
(Corieltavi).
Hacheston, Suffolk.
Britannia 1, 1970,
1974,
439,
468;
20,
6,
29; 2, 1971,
1975,
261-2,
271 settlement;
288:
well,
3, 1972,
lined with
oak
361;
5,
boards,
dated to c. early 1st c. to mid 2nd c., 5.2m deep, and measured 0.83m
by 0.76m between the boards;
Rowley
well;
1975, 85-101.
occupation
evidence.
c.
8 ,
1976, 403.
20,
1989, 301.
Rodwell
Sc
c.
to
late
3rd or
early
4th
c.
from
coin
(Trinovantes).
Haddon, Cambridgeshire.
Britannia 23, 1992, 286; 24,
1996,
418.
AD 50-410?
Condron
Scott
1993,
38, CA 95.
Hadstock, Essex.
Rodwell
1978, 32.
1963, 135-6.
397
VCH Essex
38,
201,
I960,
204;
Bruce 1966,
153-60. AA(5)
4,
1973,
200-1,
1974,
461-2;
15,
1984,
lift.
in
the
periods.
1914,
51,
1961,
MacLauclan
468-73.
OS
1959,
5,
JRS
Jarrett
276;
193-
177-90,
clay.
69-70.
164,
1857,
describes
22.
Historical
Northumberland
Map
and
1.94ha)
burnt
on Hadrian's Wall;
down
late
2nd
c.,
Guide:
over
County
Hardrian's
several
History
Wall,
fort
rebuilt
10,
1989,
151-
buildings
in
early
3rd
c.,
(RIB 1427);
but
partly
(Brigantes).
259, bath-house;
5, 1974, 426,
Staffordshire
165,
VCH Staffordshire
ST 7.
TIRBS 1987,
1, 1908,
2 wells,
drains. N.
189.
WMANS.12
,1969,
SMR
26.
(Cornovii).
1979,
327,
15,
101.
Britannia 4, 1973,
316; 5,
1974,
455;
covered drain;
tank;
68, 80,
slabbed-drain;
11,
1980,
390,
12, 1981,
14,
1983,
1984,
320;
259;
16,
1985,
1976, 417.
Current
RCHM
(Durotriges) .
Hambleden, Buckinghamshire.
Archaeologia 71,
Britannia 14,
1921,
141-98,
1983, 256-9.
bath.
Arch.
J.
124,
1968,
152,
158.
430,
Hamilton, Leicestershire.
Clarke 1956,(ed.).
1951;
231,
Museums.
1956.
Condron 1996,
Liddle 1982,
379.
Archaeology Report
4,
(possible bath-house).
398
40,
Leicester
1861, 336; 19, 1863, 60-3, well house; 36, 1880, 27-8.
Scarth
1973,
TIRCGLL 1983,
122 bath-house.
69-70.
SxAC
16,
1864,
(eds.),
52-64;
68,
earthen rampart
1968,
1927,
265.
89-132.
enclosing 1.6ha;
(Regni).
Hardingstone, Northamptonshire.
Britannia
5,
1974,
265,
270-2,
276,
280.
Dryden
1885,
well,
found
1968,
fiche 523-4.
1885,
61.
Swan 1984,
Hardknott, Cumbria.
Collingwood 1928, 314-52.
comments,
Garlick 1985,
(1973).
rising just to the south of the Parade Ground, which lay to the N-E of
the fort and nearly 3 0.4m (100ft) above it, was found in the 19th c.',
(item 48) . At
item 49,
'The bath-house
lay just
Excavations... revealed
traces of a dam across the stream close to the bath-house and this may
well have been constructed to feed a channel or pipeline to supply the
building with water.
227,
comments,
short
(c.
Hutchinson 1794,
578.
Stephens
1985b,
item 42,
40m)
leat
fed
from
an
adjacent
hill,
'Camp
Sike'
(Hutchinson 1794,
Harlow, Essex.
Bartlett 1991, 22, 154-5, Roman settlement.
a building with hypocaust and mosaic floor,
1973,
304,
325;
5,
1974,
442;
8,
1977,
407,
timber-lined;
4,
20,
399
&
Hist.
SMR
Harpole, Nottinghamshire.
BNFAS 2,
1851,
1967,
126.
144,
11,
(SP 69 62).
JRS 57,
SMR 375,
1902, 197.
1967,
NH 61,
JBAA 2,
186.
1847,
JNNHS 11,
364;
5,
1901-2,
1850,
7-8.
375;
6,
Scott 1993,
VCH Nottinghamshire 1,
Harpsden, Oxfordshire.
Scott
1993, 159,
VCH Oxfordshire
1,
1939,
323-4.
(Catuvellauni) .
1985,
Scott
327.
Money
1993, 59,
1973,
Current
5, 1974, 458;
Arch.
41,
15,
185-8.
SxAC 108,
1984,
Money
1970,
39-49;
330;
1980.
111,
1973, 27-43.
1987,
Hartlip',
No.
37,
Hartlip,
Appendix
Ten,
'The Bath-building
at
124-5,
246,
Fig.49.
JBAA 1,
1845,
314; 4,
5,
1849, 370.
II,
&
Figs.,
xxiv(2).
bath-building.
VCH
1848,
3,
1932, 117-8,
398;
pl.xxiii,
(Cantiaci).
1930,
286-7,
villa,
bath.
JRS 12,
1922,
273.
Scott
1993,
(AD 335)
(Regni).
1987, 158,
287.
JRS 16,
1926,
Britannia 8,
84,
1977,
418; 25,
NAR SU 60 NE 1.
1994,
TIRCGLL
(Regni).
Hayes, Kent.
Black 1987,
145, No.38,
400
Philp 1973.
315-6,
1979,
1978,
84.
Johnson
57-114;
'Timberlined
49;
acting
17,
as
an
(p.69).
1979,
drain
10,
Arch.
J.
6,
1971,
71-2,
well,
Bedfordshire.
stone-lined,
3.65m
(12ft)
deep,
1.06m (3.5ft) diam., dated to not later than 1st half of 2nd c.
Scott
Hemsworth, Dorset.
Britannia
of Roman Britain,
with bath,
4th ed.,
1991.
TIRCGLL 1983,
58,
courtyard villa
of Constantine I to Gratian.
(Durotriges) .
Heybridge, Essex.
Britannia 26, 1995,
large
well...
360 item 9,
timber-lined
preservation of
wood', (b)
11m due
of
the
have
been
excavated
Langford, at TL 845
TEAH 17,
to erosion
wells
near
1986,
7-68;
with
082,
19,
good
'a simple
1988,
243-8,
high water
table
and
sandy
soil.
J.
1977,
389;
Arch.
77,
1981,
168-71, well.
14; idem.
1977,
SMR
91,
9,
2354,
HU
settlement,
c.
1934,
169.
1978,
6.
Britannia
433;
11,
Whitwell
1980,
7,
1976,
366;
20,
Smith 1987,
1989,
189-98,
settlement;
278-80.
8,
Current
Figs.11-4,
pis.8-
1970,
pre-Roman occupation;
324-5,
67-8.
TIRBS
1987,
42,
minor
(Corieltavi) .
1,
1970,
42,
184;
49,
51
2,
1971,
Ant.
258.
1987,
42,
minor
Itin.
10,
470.4,
1979,
477.3 (venonis),
296.
Liddle 1982,
settlement,
with
well,
at
crossing
of
SP 464
(Corieltavi) .
44, 99-100.
PSANHS 92,1946,
25-
25,
1950,
1-6,
61-4,
Somerset
141-3; 27,1961,
58-61.
30 .
282;
4, 1973, 284;
19,
1988;
445
drain;
21, 1990,
331.
13, 1930,
171-84.
(Bremenio) , 150-1;
337; 20, 1989,
ed.),
301
stones
2,
Britannia 1,1970,
1971,
3, 1972,
entering
the
'Bruce
fort
the remains
by
recorded
the
seeing
the
Richmond
1940,
94.
remains
Itin. 463.3
14, 1983,
143
fort,
of an aqueduct
South-Gateway'.
9;
67 Ant.
'recorded seeing
comments,
127-8;
40-1,
of channelled
Hanson
1970,
of anaqueduct
by the S. gateway'.
3 68,
of
NCH 15,1940,
(2ha) to c.
13,
19.
1978,
79.
252,settlement. Northants.
RCHM County of Northants.,
Hinton Charterhouse,
Fiford Plantation,
Scott
1635,
1993,
16,
SMR
Avon.
AV 44, bath-house.
VCH Somerset
1,
1906,
363 .
J.
11,1854,
53,
1973,
49-51.
16-41.
Britannia
1970,
297, bath;
Arch.
J.
2, 1971,
278-9;
3,
PSA
2, 1852, 265.
402
1958,
223;
137
51,
2, 1962,
vicus;
1961,
174,
60-71.
AD70-160;
3rd
49,
1959,
197;
112 well
52,
1962,
6.08m
195-6.
(20ft)
deep;
50,
NSJFS 1, 1961,
26-
c.
settlement
on
higher
ground
occupation
to
the
West.
(Cornovii) .
1,
1970,
279-80,
dismantled
end
Fig.6,
2nd
c.,
bath,
drain
dates
from
c.
1973,
Clack 1982,
1970,
58.
381-4.
JRS 41,
Scott 1993,
Piercebridge.
1951,
52.
5,
140s,
indicates
AD
1971,
that
251-2,
MOW 'Excavations',
1969,
51-2;
Tyler 1980,
60.
Holt, Cheshire.
Britannia 1, 1970,
42,
12,
1970,
1981,
368,
395;
item
43 n.29 Ant.
Itin.
2,
1971,
127-9,
15,
1984,
255-60.
51,
comments
that
469.3
130 n.
54,
Davies
182,
1949,
303;
3,
143-55.
1972,
Hanson
supply depot
bath-house.
1.5,
Waddelove,
1986,
Nos.
tile works
55,
1988,
98.
Scott
1993, 47,
TIRBS 1987,
43,
SMR
193 0,
RIB 439-443 ;
supplying the
1177c,
CL 7,
14-5.
CSIR
pottery and
legion at Chester;
bath-
(Cornovii).
Horncastle, Licolnshire.
Arch.
J. 90, 1935,
129-30,
169-79;
103, 1947,
9,
1966,
15, well.
Todd
1982, 76-7.
TIRBS 1987,
Roman
called
road
21-3 ; 112,
1994, 269.
High
Street;
walls and
38-9.
EMAB 5, 1962,
1970,
72-4;
(c.2.5ha),
towers
1958,
c.
idem.
on the prelate
3rd
c.
Corieltavi).
Horningsea, Cambridgeshire.
Britannia
well.
12,
1981,
PCAS 10(iv)
340.
1904
Fox
1923,
(1898-1903),
403
210-11. Hughes
237-40.
17, 1912,
1904,
14-69.
237-40,
Phillips
1970,
200.
1972,
71-2,
well at TL 4983
Walker 1912,
143,
6348.
Swan 1984,
VCH Cambridgeshire 7,
1978,
pottery works
near
Eye
Hall;
active
71-3.
TIRBS 1987,
mainly
in
2nd
c.
(Corieltavi).
Ael(l) 1,
1822, 263-320;
253-4,
Arch.Ael.{3)
255-63;
25,
1904, 193-300,
at
10,
1933, 85-96;
11, 1934, 185-205; 12, 1935, 243-4; 38, 1960, 61-71; 39,
1961,
40,
279-99;
1962,
127-43.
1971,
127,
13 0 n .54,
1974,
410;
6,
83-96,
105-15,
117-33,
Birley
1961,
153,
181.
Britannia
27 6;3,
1972, 193-6,
197 5, 232;
7,
431-2;
1978,
155,
268;,
1863,
9,
197 6,
372-3,
latrine
and
420-1;
369; 19,1988,
fountain. Bosanquet
214ff. Bruce
41,
1963,
263-4,
1985,
277-80;
1875,
170,
40.151,
1, 1970,
161,
11,
1980,
14,
234.
207, 248-9,
Hanson
Fig 24.151.
3 90;
359;
1992,
2,
3 60;
5,
8,
1977,
12,
1981,
1904,
3 06-8,
3 09,
18,
17-30; 7,
153, 276,;
2 02,
165,
37-
16,
111-2,
315,
255.
Bruce
253,
Stephens 1985(b),
comments,
channelled
stones,
226.
It
is highly probable
that
the
fort was
also supplied by an aqueduct, although not from the adjacent Knag Burn
(contra Birley 1961,
181).
from the
N and N-E
stone
of
an
(Manning
1976,
40.151,
Fig.24.151). The pipeline suggests strongly that the fort was supplied
by
an aqueduct.
Collingwood
nymphs
(Lap.
This
could
only
have been
a pipeline
aqueduct
and
17 0,
23 4)
stood at
Roman
external wells at NY 7897 6869 and NY 7918 6894. A third well of later
construction
is
TIRBS 1987,
43-4,
at NY
7909
RIB 1594;
6857
fort
near
and outside
(2.06ha)
the
South
on Hadrian's Wall,
gate.
earlier
than the Narrow Wall but overlying the demolished foundations of the
Broad Wall and Turret 36B; civil settlement and Mitraeum; occupied 2nd
-4th c.
(Brigantes).
404
Clark
Ramm
1935,
1978,
88-92.
87,
Drake
89,
90,
1736,
106.
1980,
62.
361,
582,
Scott
587.
1993,
Elgee
150,
SMR
TIRBS
1987,
44,
occupation
1933,
00314,
354-6,
1st
168-9.
to
NK
14.
Roman bath.
late
4th
c.
(Parisi).
Hucclecote, Gloucestershire.
Britannia 21, 1990, 347.
drain and part
127.
TBGAS
Liversidge
Scott
of a bath-building';
55,
1933,
323-76;
48,
79,
1958,
1961,
153
73,
SMR 468,
a graffito
possibly a villa,
GS
sketch
59,
comments,
80,
49,
1959,
1962,
42-9.
VA wall plaster
of a timbered building
no.23;
159-73;
1993,
depict
on
stone
seems
to
footings,
bath-suite,
building
and
228-9,
Northamptonshire.
231,
Iron Age;
n.172
mentions
name
5, 1974, 268,
12,
change
1981,
342,
434;
11, 1980,
describes main
from Wootton
Hill
Farm
to
Hunsburry Villa;
13, 1982, 366; 16, 1985, 182; 20, 1989, 200. CBA G r .
1974,
Newsletter
Bedfordshire,
4,
22;
9,
1979,
'A
review
22, 1988-9,
Jackson,
1988, 3-21.
of
Archaeology
10 & 22.
in
Northants.
Swan 1984, fiche 527. Occupation, dated from pottery, c. 2nd c. to 4th
c.; boundary ditches and a pit from 1st c. AD.
5,
1985,
pi.3,
424-6,
RCHME Northamptonshire
Scott
1993,
148-9,
(Catuvellauni).
Huntingdon, Cambridgeshire.
Britannia
Paqueduct.
166.
6,
1975,
251
settlement;
Todd 1978,
12,
115.
TIRBS 1987,
1981,
3.
273-5,
lead
Medieval Arch.
tank,
12 1968,
(Catuvellauni) .
Huntsham, Herefordshire.
405
later combined to
52,
1962,
167,
Fig. 15,
169,
'...,
partly built into the walls of the larger tank sunk into the floor of
the east aisle. The last remained in use after the barn was destroyed,
in or before the 4th century....';
205,
Fig.11,
TIRCGLL 1983,
to late 4th c.
206.
TWNFC
37,
55,
1961,
1965,
208,
179-91.
Fig. 13;
WMANS
56,
12,
1966,
1969,
26.
occupied c. AD 200
(Dobunni).
from
the
ruins.
of
hypocausts,
baths
and
mosaic
Ickham, Kent.
Britannia
1980,
413:
6,
1975,
283:
inscribed
wells,
2 water-mills;
lead seal;
12,
1981,
10,
387;
1979,
20,
1989,
350-3;
11,
178,
188.
Fox 1923,
183-
Ickleton, Cambridgeshire
Arch. J. 6, 1849,
4.
14--26.
JBAA 4, 1849,
Britannia,
356-68.
Scott
1971,
1993,
187,
39,
192.
SMR 04168/04153,
CA 102,
(Catuvellauni) ..
Icklingham, Suffolk
Britannia 1,
1975,
262;
1984, 307;
8,
444-5;
1978,
293 309-10.Swan,
3rd c .
9,
1972,
330;
448, well;
12,
5, 1974,
1981,
439;
271-5;
6,
15,
19, 1988, 456; 20, 1989, 235; 21, 1990, 341; 25, 1994, 279
settlement.
1911,
1977,
30-1; 3,
1984.
NA 28,
TIRBS
1945, 213.
1987,
VCH Suffolk 1,
(Inceni).
Iford, Wiltshire.
Scarth
covered
1864,
up
120,
after
TIRCGLL 1983,
claims
it
was
it
was
'a perfect
investigated.
coins of
(Belgae).
Ilchester, Somerset.
406
Roman
bath',
VCH Somerset
1,
but
1906,
was
3 00.
4th c.
313,
drains
noted'.
Leach
1982.
Leach
1985.
11-84,
Figs.1 & 2.
Wacher,
Leach
1992,
Shefield
1995,
20,
21,
32,
324,
SANH 135,
331.
TIRCGLL
town occupied
(Durotriges) .
226,
5,
12,
233-4,
1981,
241;
18,
25,
1974,
32,
13,
20-1,
51;
discusses
the
56,
water
pipes
distribution
of
fortress
completed,
Before
was
these
water
could be
once
and
Pitts
trenches
the
future
constructed
the
&
which
aqueduct
the
1966,198.
and
St
trenches
because
completed,
if
of
the
indeed
legionary fortress
short
itwas
bath-house
fort was
water
inside
1985,
been
for
to
the
the
demolished
establishes
fort.
c.AD
period
of
that
the
fortress
1987,
104;
occupation'.
TIRBS
85.
'The discovery of
started.
successive construction-camps
have
supplying
Joseph
could
and pipe
55-7;
27,
200.
213;
1984,
21,
1970,
1960,
15,
1979,
Hanson,
63 pi. viii;
284;
10,
12-3,
27; 19, 1988, 170-1; 21, 1990, 310; 25, 1994, 103, 159.
50,
1983,
33-5;
4, 5, 7,
1987,
104;
14,
27,
3, 1972,
18,
287;
123 n.7;
It
was
45.
not
Flavian
V V (21.74ha). Two
3,
8,
DAES 1977,
the
remains
water pipeline
29-30,
42-4,
304,
vicus
at
NT
of
1990, 29-30.
stone-lined
Hanson 1970,
channel
which
345
720
339-40;
16,
368,
had
comments:
carried
JRS 36, 1946, 109; 37, 1947, 165; 38, 1948; 81199;
pipe channel"'
support
22;
(probably of wood)
the
20,
265.
1956
1972,
of
(JRS 1966,
a wooden
199;
1967,
pipeline
407
from
one
of
the
Antonine
phases'.
CSIR 1.4,
RIB
2132-3.
Fort
(c.
2.84ha)
TIRBS 1987,
occupied
twice
45 at NT 344
in
the
Antonine
Irchester, Northamptonshire.
Arch. J.
100-28.
BNFAS 5,
1971,
(1),
19;
6,
1971,
14 Irchester
Britannia 1, 1970,
(settlement);
BAJ 7, 1972,
occupation AD 68
114,
116,
118,
1985, 200, water supply, 289; 23, 1992, 285; 25, 1994, 273.
Wacher,
1990,
ramparts
147.
JRS
16,
1926,
223,
3 wells;
14.
excavations
43, 1953,
123,
16,
Burnham &
by W W
92.
to
Robb on
Nat. Hist,
of
Northants 1712, 517, J Morton reports that walls was still standing in
his time, but now completely destroyed by cultivation.
RCHM County of
Islip, Northamptonshire.
Britannia
5,
Archaeology
same site
1974,
13,
1978,
257,
villa
examined
179.
PAS 9,
1882,
from
air.
90-91,
'Possibly from
Northants.
the
(limestone slabs).
11,
379;
1982,
13,
1980,
377-9, plan
412;
15,
1984,
Fig.14,
307-8,
bath-house;
reservoir c.
12,
260
to AD
settlement
360-400.
dates
from
An
area
Iron
Age
of
springs
period
and
north
of
350,
1981,
Dating from c.
the
settlement;
Romano-British
religious
site.
Ixworth,Suffolk.
Britannia 1,
settlement,
1911,
1970,
vicus.
183 ;
2, 1971,
187,
189,
193 ; 17,
1986,
(reprinted 1975),
404 fort,
VCH Suffolk
pottery.
408
266.
1970,
616.
Ross 1968.
(Durotriges).
Kelvedon, Essex.
Britannia 1, 1970,
52,
5, 1974,
442-3;
31,
Current Arch.
199.
settlement
9, 1978,
with
mansio
Chelmsford Arch.
Rodwell 1988.
1963, 149.
451;
480.5 {Canonio) ;
10,
1979,
311;
48,
v(l),
1975,
25-30,
and
presumably
Trust.
Stephens 1985(a),
16,
1985,
fort
bath.
198.
4,
1973,
265;
197-208,
item
and
CBA
JRS 59,
Res.
vicus
Rep.
1969,
later
63,
223,
245.
VCH Essex 3,
(Trinovantes) .
Kempsford, Gloustershire.
Britannia 17, 1986,
414,
paddocks c. 20m x 15m. . .dating to the late 1st c. and early 2nd c.;
2nd phase
1957,
2nd
c.
and
early
3rd
c.
Gloucestershire
1985, 2-3.
Countryside
O Neil 1961,
9,
36.
1993,
timber structures,
to 7th c.
1908, 8.
293
& 295:
a well,
(ii)
Simco,
503.
1984,
108.
VCH Bedfordshire 2,
(Catuvellauni) .
Kenchester, Hertfordshire.
Branigan
484.7
9,
5.
1976,
(Magnis) , 118-9,
1978,
Jack
& Fowler
438;
1916.
10,
189;
22,
1981,
Herefordshire 1,
bath-house;
1983,
1908,
other
169-70.
175-83.
houses
Britannia 1,
1970,
1979, 298;
with
99-102.
31; 23,
Ant.
Itin.
283-4,
villa.
TIRBS 1987,
also
55,
with
46,;
1964-6,
192-5.
settlement
possible
93VCH
(c. 9ha)
bath-houses;
at
kilns and
409
AEx.
1976,
91.
69,
1955,
Cant.
Britannia 2,
12,
1981,
58.
Archaeologia
96;
1971,
366;
82,
187;
16,
22,
1967,
5,
1985,
1828,
184-91.
1974,
459;
178, 315;
Philp,
et al ,
1991.
8,
17,
Philp
1986,
359;
1926,
870
795,
SP
278,
875
493-501
Central,
536,
2,
wells
Whitwell
1979,
at
101-5,
259.
9,
1978,
471-2;
1815,
45-
bath-suite,
drains,
119.
TIRCGLL
(Cantiaci).
29,
51,
1939,
90;
12,
1911-2,
147,
208.
223-4.
206-13.
SP 8715 8028,
1977,
No.38, Fig.96;
8055.
Britannia 1, 1970,
9, 1978,
JRS
24,
SP 8718
1982,
94-8,
SP 8723 8009,
9, 1974,
(illustration);
, 1893.
II, 241.
434-5;
796.
Northamptonshire Arch.
28
Dunkin,
1972,
SP
Arch.
Oct.
409;
426.
120.
Northamptonshire.
9. Bridges 1791,
5, 1974,
1855,
1977,
1973,
1993, 79, GT 6.
36,
Athenaeum
21-3.
336;
SP 876 798,
SMR
3957.
PSA(2) 23,
1911,
Swan
VCH Northamptonshire
TIRBS 1987,
JBAA. 32,
Northants
212.
129; 3,
1984,
1,
1902,
fiche
194.
(Catuvellauni
or Corieltavi).
1926,
J. 2,
Britannia 10,
1922,
415.
371.
Bulleid
Fritter 1991,
16,
SMR 1214,
&
Horne
1926,
1979,
175;
11,
109-38.
1921, 210-4;
1980,
bath. Antiq.
293-6;
Collingwood,
14,
1924, 234.
17,
1986,
1926,
135.
Scott 1993,
at
Durley
Hill,
the
so
called
'graveyard'
site.
She
comments:
TIRCGLL
1983,
62,
large
courtyard
villa;
coins
of
c.AD
265-375.
(Belgae).
410
Britannia 1, 1970, 184, possible fort; 8, 1978, 456; 9, 1978, 456; 10,
1979, 322;
378,
380,
11, 1980,
quarries
1979,
294.
1981,
'The Chessals
73-80,
385;
and buildings;
Glenvensis
2 drains,
12, 1981,
10,
1976,
61.
21,
1990,
17f,
Excavations,
201.
plans.
1975-80.
1982,
Arch.
McWhirr
70-3, plan.
13,
Current Arch.
Kingscote
Kingscote,
one stone-lined,
of Gloucestershire 1, 1976,
Scott 1993,
69,
Ass.
1981,
RCHM County
73,
SMR 325, GS
10,
1979,
275;
12,
1981,
143-61;
1982,
338.
Glasgow
41,
13,
TIRBS 1987,
Macdonald 1934,
Robertson 1990,
47. Fortlet
(0.5ha)
191-2.
PSAS 94,
Kintbury, Berkshire.
BBAA 31, 1950, 2; 41, 1961, 1.
Roman bath-house.
353-68.
2nd c.,
144.
7, 1976,
JRS 59,
1969,
207.
YAR 1974,
3.
probably of 2nd c.
TIRBS 1987,
(new ser.), 3,
Scott
1993,
150,
NK 10, bath,
Tyler
1980,
52-3.
YAJ 46,
47, villa,
c.
1974,
(Brigantes).
188-9, Field,
'Phase
(1.2m)
( c.
entering the camp from the south... Phase II, the 1st tank was reduced
to a smaller volume,
Occupation appears
of similar construction
JRS lix 1969,
411
to 60s.
228, no.
Webster
Todd
M, (ed.),
Antiquity,
about
PSA 2, 1885,
410-11.
1989.
(See
their
generation' . A
authors of articles
Richards 1978.
Scott's
TNDFC 1, 1870-1,
review
of
this
book
in
very
critical
approach
Scott
of
some
1993,
of
23,
the
BK 27.
Lancaster, Lancastershire.
Arch. J. 127, 1970, 239-40.
2,
1971,
254;
1975, 239;
331;
14,
3,
1974,
312-3;
4,
1974,
206-9;
5,
plan, Fig.16.
15, 1929,
27 0-1;
15,
Hanson
33-40;
17,
1984,
1988,
1930,
284;
179-83.
57-72.
17,
1974,
Britannia
418,
1979, 290;
1986,
43 6;
JRS 49,1959,
Leather 1973.
19,
465;
12,
6,
1981,
1987,
443,
106-8, plan.
LAAA
Leather G M,
1979.
Shotter & White 1990, Occasional Paper 18. One bath north of fort and
2nd one east of fort. Well,
1953,
1-23.
Wall'
late
TIRBS 1987,
Roman
construction.
Civil
THSLC 15,
1st
'Wery
to
4th
c.
(Brigantes) .
177-8
with
aqueducts.
fort
at
Steer
1938,
Longovivium
includes
the
channels.
All
mining
during
plan;
10,
(2.52ha)
pl.xvii (plan).
two
the
1953,
where
(Lanchester)
dams
three
211-23,
Clack
at the
channels
19thc.
394-5.
established c.
are
sources
were
the
discussed
of
1936,
Gosling
held till
1976,
213-25
three aqueducts
the
destroyed
TAASDN 7
AD
&
JRS 28,
two
in
detail,
longest
during
200-15;
for the
which
aqueduct
open-cast
9,
1939,
RIB 1072-98.
late 2nd c.,
coal
110-22
Large fort
re-occupied
(vicus).
(Brigantes).
38,
1960,
Britannia 4, 1973,
1-38;
39,
1961,
371-3 ;
41,
1963,
19-35,
211-3.
178, 187.
1932,
8.
38,
151,
Ramm
1978,
71-4,
03000.071/03000.75, NK 18.
Webster
1969,
246-8.
80-7.
Scott
YAJ 44,
1972,
32-7.
1993,
255SMR
TIRBS 1987,
49,
corridor
villa, with small bath-house and well; defensive ditch established 1st
or 2nd c.; expansion from 3rd into 4th c.
(Parisi).
Latimer Buckinghamshire.
Britannia
Branigan
Fig.11,
1,
1970,
1971,
315;
169.
2,
110,
Medieval
112,
Arch.
Rees,
114,
11,
116;
1967,
of Bucks.
6,
1975,
263;
19,
12,
1973,
c.
into
1968,
1-11,
340-3.
Scott
plan.
AD 80 lasting to c. AD 120;
through
197
medieval
period
and
into
then occupied
modern
times.
(Catuvellauni).
2,
1971,
260,
villa,
'Three periods
of
construction...The
first building possibly a bath-house, was already ruinous when the 2nd
building was erected on the same foundations.'.
2nd centuries.
Scott 1993,
Gentleman's Mag.
164,
SMR 1057,
1793.
SNL 39,
Trans.
Shropshire Arch.
TCSVFC 4, 1905,
26-7.
TIRBS
1987,
Rickman,
SH 8.
plan.
1970,
36.
Nov.
Dates from c.
49,
villa
of
1970,
49;
1st to
1838,
40,
pl.iv.
1971,
Soc. 56,
7-10,
1957-60,
3 periods
with
bath-house;
(Cornovii).
Britannia
1,
1970,
Gloucestershire
walling,
126,
1976.
2.58m deep,
n.78.
TBGAS
Donovan
55,
1933,
0.67m diam.;
56,
H
377,
1934,
1934,
pl.l,
99;
99-128.
well,
57,
RCHM
dry
1935,
stone
234-59,
period c. AD 120 to c. AD
381;
10,
1979, 287;
11,
1980,
363;
12,
1981,
328,
'A road-side ditch drained into the butt end of of the fort ditch...;
a gully ran along the west side of the via praetoria,
413
perhaps leading
to a stone drain immediately behind the north rampert'. Hayes & Rutter
1964, 69-75. Spratt
(Brigantes).
Lechlade, Glocestershire.
Britannia 22, 1991, 276, villa.
2442, GS 65, bath, SMR 3191, GS 66, well, dating from c. 2nd c.
96,
1978, 22-3,
Champion 1971.
located
Burrows, et . a l ., 1925.
of
Leicester, Leicestershire.
Bellairs 1899,
40-4,
3,
wells;
336;
13,
444-5;
25,
262-4;
4,
6, 1975, 77-8,
9,
1978,
1982,
20,
479;
415-6;
16,
271.
bath-house.
286,
339-40;
JRS
49, 1959,
156,
2;
5,
1979,
1985,
PNRB 443-4.
1973,
10,
38 n.17,
246; 7, 1976,
435,
1989,
1994,
Britannia 1, 1970,
(Ratas) , 479.3
1972,
fortress;
2 Figures.
1974,
7 possible
160-2; 11,
1991,
155 n.44,
vexillation
1980,
245-6;
24,
1986,
1993,
290,
1968-9,
1-10.
390,
317-8;
TLAHS 44,
51,
22,
49,
1948,
Wacher 1995,
343-
AD 43, but the Coritani was unlikely to have enjoyed self-rule after
the battles of AD 60. Wacher also discusses K M Kenyon's
Jewry Wall,
baths, watertank,
drains,
(A RATIS MP
Possible
Flavian
fort
II) .
but this
TIRBS 1987,
vexillation fortress
on Fosse Way.
Civitas capital
on the river
and/or pre-
of
to early
Corieltavi
from
(Raw Dykes,
and later
Claudian
is still
and sewer,
work on the
late
not yet
continued to
(Corieltavi).
68. Britannia 7,
12,
1981,
1976,
337-8,
327-8;
Fig.9.
8,
1977,
Current
392;
Arch.
11,
81,
1980.
1981,
367
314.
8-9,
SK 575
Ratae Coritanorum' .
JRS 29,
SMR SK 50 SE G C , LE 51.
Leicestershire
1,
1939,
207-9,
Fig 13.
1907,
196.
The
58;
Scott
1993,
55, 1979-80,
watertank
113,
83.
VCH
was
timber-lined.
occupation 2nd
Leicester(2), Leicestershire.
Britannia 24,
to 3rd c.
1993,
290,
Leintwardine, Hertfordshire.
AEx
1971,
18,
Britannia 1,
summary
1970,
report.
55 Ant.
Black,
Itin.
484.8
1995,
26-7,
30,
(Bravonio) ; 3,
83,
1972,
93.
66,
69,
189, 317; 12, 1981, 33 9; 13, 1982, 3 60-1; 23, 1992, 284; 24, 1993, 292
fort at SO 415 740; 25,
39,
1969,
1994,
271.Nash-Williams 1969,
40, 1972,
318-20; 41,
large 2nd c.
fort or stores-base
94-5.
TWNFC
1975,
297-
(4.55ha),
with bath-house.
(Cornovii) .
Leyton, Essex.
London Arch.
S-W,
1921,
7(15),
item
1995,
60,
166,
397-401.
Roman
Period,
'one
Cental and
or more buildings
were
Lincoln, Lincolnshire.
Archaeologia
colonia;
70.
104,
111,
1973,
1954,
61,
49,
Itin.
51 Ant.
1971,
257-8,
1981,
Arch.
111, 1954,
J.
103,
1946,
106-28; 117,
475.3,
tank,
gates.
10-28 aqueduct;
lower city;
129-207
extramural.
476.7,
colonia;
5,
477.9,
1974,
Britannia
478.10
1,
1960,
1979,
1970,
(Lindo) , 184,
422 water-pipe;
26-56
40-
50-91
38,
47,
260;
6, 1975,
2,
275,
16,
southernmost
Colyer
1975.
197ff,
water
supply;
24,
1993,
288
10, 1979,
refers
to
294the
&
Hockley 1993,
415
Lincoln
Arch.
Unit 5th
Annual
lengths
long;
also
Hanson
197 0,
unfounded,
drains,
sewer,
43-55,
that
bath,
3 69.
She
the aqueduct
tank
at
to a
stone-lined well
(Proc.
Thompson
ICE,
1981,
have
very
its source
(1.75 miles)
been
positive
some
2 km
located.
statement,
(1.25 miles)
1979,
2.8km
and well
makes
It is c.
EMAB 11,
at
SK
9777
1977,
7211;
3 0 drain or gully
12,
1978,
25,
East
1954,
1985,
106-28.
298-300),
Wacher
proposes
1995,
a much
86-93.
PNRB,
132-50.
longer
Wacher
aqueduct
with a
Wilford,
Lincoln Arch.
Unit,
1982,
unpublished summary
AD 90.
Further
Monumental gates
changes
to
colonia during
3rd and
4th
centuries.
including sewers,
aqueduct,
(Corieltavi).
1982,
1908,
Webster
257-8.
358,
aqueduct;
1975,
15,
100-2.
1984,
Wright
291.
1872,
VCH Shropshire 1,
24-9.
TIRBS 1987,
(Cornovii).
Linton, Cambridgeshire.
Archaeologia 32, ?, 352.
1853,
163.
64,
227;
14,
Arch. J. 1, 1850,
1857,
14,
Gentleman's Magazine 2,
13-4.
CA 107,
147, 193;
2,
63.
1859,
Britannia 22,
1991,
418.
171.
PCAS 3,
PSA 1,
bath-suite.
10,
1905,
VCH Essex 3,
1963,
135-6.
Arch.
1868,
32.
TEAS 1,
SMR
91,
62.
Litlington, Cambridgeshire.
Archaeologia 26,
1, 1829,
546.
183 6, 376.
Babbington,
J. 25,
416
Fox 1923,
Gentleman's Mag.
184-5.
JBAA(3)
CA
109,
bath.
Courtyard
VCH Cambridgeshire
villa,
said
to
have
7,
at
1978,
46.
least
30
TIRBS
rooms;
1987,
51.
bath-suite.
(Catuvellauni).
1963,
126.
Wooden
water-pipe,
iron
collars
VCH Essex
located,
possible
aqueduct.
LI 102.
Whitwell 1982,
266, SK 93
SW 10, well.
Drury 1978.
Littlechester, Derbyshire.
Antiq.
J.
51,
1971,
36-9.
Brassington
1993,
24,
who
refers
to R G
Britannia 1, 1970,
283; 4, 1973,
11,
1981, 335,
285,
330;
5, 1974,
324,
1980,
365;
12,
445-6; 20,
EMAB 1960,
2.
1968, 3; 59,
pi.
420 vicus;
Watkin
confirmed
by
JRS 58,
1985,
Mr
70-91.
Brasher
comments '...which
informing
me
that he
came
further
upon
what
to a depth of 3.6m
he
(12ft).
10,
1979,
329-30,
of
buildings
development
20,1989,
of
boundaries
after
c.
AD
1990,
350-70;
16,
1985,
as
excavation
functional at any
VCH
Wiltshire
(Littlecote 1979);
dating
15,
from mid
to
391f;
308;
1,
1983,
1984,
19,
project
322f,
1988,
completed.
1981,
1-5
328f, Fig.29,
Fig.26
407-10,
Hoare
shows
Fig.27;
1819,
117-21.
1957,
98.
Walters
&
Phillips
12,
one was
c.
1980,
11,
villa;
317 f; 21,
Fig.17;
1st
c.
to 4th
417
c.
indicating
(1)
1978
3 bath-houses,
8 periods
of
flint-lined,
late 3rd
3 tanks,
(Atrebates) .
Littleton, Somerset.
VCH Somerset 1, 1906, 323-4: 2 villas, at villa No.2, item 37, 'a bath
found'.
(Durotriges).
1949, 6, 17,
25.
11,
Scott 1993,
57, SMR
5999, DY 7, bath-house.
J.
17,
1937, 138.
Arch.
Camb.
83-7.
(6th ser.), 7,
1907,
175-212,
58-60.
in Wales,
Cambria
9(1),
Rep.
1978-9,
29-38.
32.
Robinson 1988,
Britannia 11,
Glamorgan-Gwent
1980, 349-50,
Arch. Trust
29-38.
Ltd,
Fig.3.
Annual
Robinson 1980,
GL 4.
27-
Bath-suite, water
Llanio, Cardiganshire.
Arch.
J.
121,
1965, 29.Arch,
of Wales
CBA G r .2,
1970,
10;
1971;
fort,
vicus; 3,
400.
after
AD
120.
1972, 300
JRS
67,
baths; 4,
1977,
154-5.
1973, 271praetorium;
Bath,
drain,
5,
spring.
1974,
Nash-
(Ordovices).
418
1949,
Camb.
5,
1888,
413-7;
1808,
67, GL 5.
50ff.
102,
1953,
Britannia
Hogg 1974,
TIRCGLL 1983,
89-163.
65-6.
225-50.
5,
1974,
Cardiff Naturalists
Morris
Courtyard villa,
1979.
Scott 1993,
(Silures).
to late 4th c.
(Belgae).
Loddington, Northamptonshire.
RCHM County of Nothants, Central 2, 1979,
105-7, No.39,
item 3, well.
Loddon, Norfolk.
Scott 1993, 135, SMR 17982, NF 132, bath-house.
London
Britannia 1, 1970,
42,
44,
49,
7, 1976,
1982,
373-4; 16,
461,
Fig.21,
39-49,
(Note:
The
only
some
baths;
65, Ant.
6,
1975,
202, water
amphitheatre;
1975,
51-3,
'Seal
on Roman
refers
1990,
342-4.
by
1987,
related
structures
19,
Current Arch.
13,
1988,
49,
Grimes
(Londinium) is
pipe,
baths;
335;
Scholfield.
passim,
water
London
to water
471-480,
268,timber
supply; 18,
House', report
literature
that
21,
Itin.
v(2)
1968.
extensive
will
be
and
cited).
1993,
Marsden 1980.
Milne 1985.
Morris
1982
No. 70.
JRS
55,
Perring,
1965,
215.
2) Cheapside; 3) Billingsgate,
and
Wacher,
3 public baths:
1)
an apparently private
1995,
88-111. Wilmott
list
of
55.
Many
excavation
reports
419
have
been
reported
in
in
London Archaeology and TLMAS during the last four decades. During the
last decade extensive excavations at the Guildhall Yard has revealed
the amphitheatre of Roman London and its elaborate drainage works to
drain water which was a problem as a result of the high water table in
the area.
Three systems
This
work has still to be wriiten up. Wilmott 1982a, TLMAS, 33, 1-31; idem,
1982b London Archaeologist Vol.4(9), 234-42, 4 baths;
Special Paper 13.
50,
enclosing 132ha,
added c. 4th c. A bridge over the Thames linked London to the suburb
at Southwark.
(Londinienses). London.
38,
1862,
327-52;
39,
1863,
135-43.
Berks.
Buck,
and
302; 8, 1977,
Britannia 1, 1970,
'several wells'.
1922,
270-3,
bath-house.
PHFC 9,
1920-4,
288-90,
388,
398.
Longthorpe, Cambridgeshire.
Dannell
Durobrivae Vols.1-8,
BC
Frere
& St Joseph,
through
to
Britannia
Iron Age
5,
1974,
1-129.
Well,
Occupation
drain,
c.
tank.
(10.9ha) on bank
(Catuvellauni).
245-6; 3, 1972,
300-2;
420
282, bath-house;
15,
1984,268-
9;
19,
1988,
421;
23,
1992,
258.
TIRCGLL 1983,
67,
(Silures).
1970,
53-4 Ant.
Itin. as
'Alone';
4,
1973,
260 n.45
'Alione' in Not. Dig.; 8, 1977, 377 bath; 23, 1992, 272, 312.
1952,
47,
91;
49,
1947,
1-19;
Lune gorge
built
1959,
51,
1951, 40-66.
(1.13ha).
c.AD
120;
TIRBS 1987,
99-100.
53,
as
JRS 42,
TCWAAS(2)
RIB 756.
Fort
in
rebuilt
Extramural bath-house.
in
early
3rd
c.
Occupation
to
end
4th
c.
(Brigantes).
SDNQ 25,
1947-50,
1-6,
61-4,
141-3.
TIRCGLL 1983,
67, timber
c. AD 3 67. Baths.
(Durotriges).
11 wells.
Scott 1993,
GS 71.
1958,
59-77.
coridor villa;
225.
PSANHS
92,
Scott
1993, 166,
1946,
41-3 ; 97,
SMR 53634,
SO 5.
1952,
91-112;
TIRCGLL
116,
1983,
67,
(Durotriges).
67,
1954,
1970,
15-36;
312;
2,
70,
1971,
169,
176 no.41,
274-5;
175,
1979, 309-13;
Meats
186,
189; 9,
1979,
Vol.l,
xlvii-1.
181 no.60,
426.
187,
Britannia 1,
190,
192-3;
196-7; 7, 1976,
Meates
3,
171,
1955 (1963).
supply, suggesting that from the well-head outside the bath building a
line of wooden pipes
from
the
well
by
delivered water
slaves
or
others.
However
the
as
it was
problem
is
drawn
not
resolved. Well
Scott 1993,
421
{Ca.ntia.ci) .
Lunt, Warwickshire.
Applebaum 1967,
21-34.
34,
40-1;
339; 14, 1983, 295; 16, 1985, 286-7; 21, 1990, 294.
Interim
Report,
TBWAS
85,
(1.21ha)
7-92,
1981,
pis.1-14.
Hobley
1975,
1971-73,
12,
4.
Large
c. AD 64-80;
fort of unknown
size,
reduced
civilian occupation.
in
The
Fig.11, p.39 of TBWAS 85, which also shows all the water structures.
(Corieltavi).
1,
1970,
138,
148;
wooden
excavated'.
9,
1932,
1973,
Scott
Wheeler
1932,
52-55.
priests
accommodation,
151,
207;
McWhirr 1981,
132-7.
4,
1993,
13, 1982,
184-5;
380;
16,
baths
at
Lydney
7,
1985,
has
'A
been
74,
TIRCGLL
1975,
supplying
152-5.
6,
GS 72.
1983,
guesthouse
68,
and
Ross,
1967,
RIB
305-8.
baths.
22,
176,
Temple,
Aqueduct
339.
with
wooden
(Dobunni).
Lyxnpne, Kent.
Britannia
1,
1970,
Lemanis) ; Rav
1977,
235,
Collingwood
Poole, Lambert
107,
1969,
Cos.
238,
Sc
45-6,
65,
106.3 5
246,
Itin.
425;
11,
51.
1980,
Johnston
2,
227-88;
Cunliffe 1980,
102-25.
473.7,
(Lemanis) , 241-2;
Richmond 1969,
Sc
Ant.
1977,
473.10,
(ad portum
1971,
172,
16,
1985,
227-88.
Johnson
8,
209-36.
Hutchinson,
188;
55-9. SxAC
1979,
53-6.
(LEMANA FLUMEN,
Rav. Cos. 108.39); 2nd c. port and probable naval base of the Classis
Britannica
(RIB 66);
bastions; baths.
3rd c.
(Cantiaci).
422
(Manning W ) ; 2, 1974, 23
252.
Northants Arch. 8, 1973, 9-12, 2 stone-lined wells; Lynch Farm is lowlying, with lots of flooding seen by numerous enclosures and drainage
ditches. Quite an elaborate farmstead of the 3rd-4th c, with a smithy
and a fishtank.
3,
1972,
9,
14,
44-5;
5,
1974,
153.
Hanson
1970,
370,
sewers
recognised
154-86,
1983,
79;
and
cesspits
stone-built
channels
179-81,
148.
63,
the
Figs.8
JRS 41,
1973,
Sc
15,
1951,
216-7.
PNRB
Christison,
3 cisterns,
57; 45,
2 drains.
1955,
1979,300-1.
85; 51,
plan.
RCAHMS Peeblesshire
1985(b),
224,
and
cisterns
the
fort
comments
(2.66ha);
were
that
1,
1967,
two annexes.
Maxwell
Sc
121;
55,
1965,
PSAS 75,
39-
Feacham 1959-63',
Stephens
a stone-lined channel,
TIRBS 1987,
Antonine I (?)
be
Hanson
173-5, plan.
stone-lined.
Sc
can
1900-1,
1961,
Steer
tanks
PSAS 35,
Richmond 1941,
and
53,
fortlet
Antonine
II(?)
(O.llha)
150m to
1890-1,
North.
Maidenhead, Berkshire.
BBOAJ 30, 1926, 76.
Maidstone, Kent.
AEx 1972,
58.
Arch.
1972, 357; 4,
Cant.
10,
JBAA 2,1847,
172-4; 3,
106,
Smith 1876.
baths.
(Cantiaci).
1970.
Antiquity 2,
466.2
fortress,
1928,
(Derventione),
vicus
and minor
69-82.
313-4;
settlement,
Britannia 1, 1970,
2,
1971,
291,
252-3,
302-3;
3,
40-1, Ant.
vexillation
1972,
361;
382;
10,
1979,
10,
35;
17, 1986,
423
449;
19,
1988,
504;
20,
4,
9,
1989,
178.
Hanson
1970,
370,item
59B,
comments,
'A
wooden pipeline,
probably of mid to late 2nd c. date, was found running down the centre
of the main street in the civilian settlement at Malton.
It supplied
YAJ 41,
5.47m deep,
0.6m diam. at
45,
209-61 at,
212-3,
comments,
Mitchelson
'The
road running
to the
gate
of
item
the fort
was
flanked by a wooden pipeline from about the second quarter of the 2nd
c.
(Mitchelson
1964,
213,
p i .iv). The
excavator
believed
that
this
pipeline supplied two small buildings in the "hop area" of the vicus.
Since
the
supplied
road
is
official
military,
the
structures,
so
chances
that
are
the
that
buildings
this
may
pipeline
have
been
military fabricae. The pipeline could not have conveyed water from the
nearby Lady spring. It presumably conveyed water either from the fort,
or
from an
external
source
have
bath-house
2,
Periods
(Britannia
cover
Trajanic,
Wenham 1974.
from which
supplied
1971,
252;
Severan,
the
Possible
Large
Agricolan
early
3rd
fort
c.;
held
fort
Constantian
37,
till
late
4th
c.
located
and
100,
evacuated
also
& Hassall
c.
supplied.
extramural
1971,
291.9).
Theodesian
dates'.
116.
was
recently
Wright
RIB 711-9.
the
TIRBS 1987,
fortress
120-158.
Extensive
(c.
54,
8.9ha).
Reconstruction
civil
settlement.
(Parisi).
Mancetter(1), Warwickshire.
Britannia 1, 1970,
1972,
319;
8,
42 Ant.
1977,
Itin. 470.3
379-80,
fort;
occupied
to
4th c.;
15,
91,
1981,
Smith 1987,
54.
1-24.
13,
1978,
1982,
1984,
361,
295-7,
Oswald
224-6.
9,
(Manduesedo) ; 2, 1971,
440;
12,
1981,
339,
dated to c. AD 40-60,
vexillation
5-17.
& Gathercole
263; 3,
fortress;
TBWAS 74,
1958,
TBWAS
16,
1958, 3074,
30-52.
TIRBS 1987,
(Corieltavi) .
Mancetter(2) , Warwickshire.
AEx 1971.
water
Britannia
channels;
9,
2,
1978,
1971,
263;
440-1.
3,
1972,
309,
Swan
1984,
fiche
319,
650,
3 wells
651,
424
and
well,
JRS
Manchester
Britannia
1,
1970,
42-4,
53,
Ant.
(Mamcunio, var.
Mancunio) ; 4,
1973,
10, 1979,
11,1980,
17,
1909a.
291;
Greater
364;
Manchester
?aqueduct, ?well.
Itin.
283,
vicus; 7,
1986,
Arch.
J.
468.7
385; 19,
1,
(Mamucio) ,
482.2
197 6,
fort;
1988,
1985,
319,
444.
13-18,
Bruton
Figs.8
&
9,
Arch.
J.
1886,
1993,
155,
28f f .
SMR
farmhouse.
80-c.AD
1973,
1787; 363-76.
Antiq.
J.
22,
18, 1938,
176-8.
PPS 5, 1939,
2 baths,
187,
one in villa,
192.
Scott
other in
aisled
450-475,
89.
spring,
TTS
53,
leadpipes,
1949,
1-14.
baths,
separate
latrine.
VCH Nottinghamshire
1,
Todd
1910,
28.
1957-8,
52,
172,
1962,
90,
192;
137,
55,
212;
50,
1965,
207.
Whitwell 1982,
224;
MJTL
51,
1966,
1961,
84,
175,
93-4,
plan;
146,
St
1960,
89, 90-
271, SK 91 NW 3. Well.
Marshfield, Avon.
Arch. Rev.
1986,
4, 1969,
415;
18,
41.
1987,
343 ; 19,
watertank,
drain,
1976,
Scott 1993,
80.
ST 78 73,
Britannia 14,
foot-bath,
The Hams,
1988,
1983,
471.
lead-pipe.
16,
1985,
Blockley
302;
1985,
17, at ST 79 76,
SMR 2 051,
317;
BAR
1982.
17,
141,
RCHM
limestone slabs.
Marsworth, Buckinghamshire.
Records of Buckinghamshire 18,
1970,
440.
Scott 1993,
29,
SMR 1520,
Maryport, Cumbria.
Archaeologia
10,
Britannia 2,
1971,
295-6.
Bruce,
1787,
128-9;
1978,
140.
3,
273-8.
Settlement at Maryport'.
Birley,
1972, 355;
1961,
4,
Collingwood
Jarrett 1976.
425
129-31,
1973,
?,
112,
'The
Stephens
220,
260;
Roman
1985(b),
no.41.
22,
Fort
226,
1991,
and
item
34,
comments,
Lapidarium
cistern
of
'A
sculptured
Septentrionale
901)
aqueduct
supply
in
26,
fort;
1926,
39,
118;
70,
415-22;
36,
1939,
19-30;
1970,
42.
the
from
analogous
Bath
a
(Bruce,
fountain
55,
panel
or
from
High
142-
bath-house
268-71;
1987,
the
85-99,
1954,
at
come
Housesteads'.
TIRBS
of
1936,
54,
Venus
have
167).
view
of
may
an
53;
panel
58,
RIB
and well,
1958,
63-7;
808-79.
within
65,
1965,
Hadrianic
fort
(Brigantes).
1975,
293;
244;
12,
283-4,
7, 1976,
1981,
335;
323;
14,
8,
1977,
1983,
299;
5, 1974,
387f;
16,
9,
420, bath-house;
1978, 432;
1985,
282-4,
10,
1979,
Fig.18,
'an
'a stone-lined
Bruton
Fort
288,
1-
c. AD
(Brigantes).
Methwold, Norfolk.
Phillips,
56.
1970,
250.
297-8,
bath.
1993,
151,
TIRBS 1987,
(Iceni).
1941,
1980, 70.
131; 47,
1957,
208.
Scott
NK 22.
Tyler
Mildenhall, Wiltshire.
Annable 1976,
126-7,
well,
c. AD 50-60.
Burnham Sc Wacher,
& St Joseph
1983,
1963,
'Excavations
30-8,
house; 70-1,
88-9.
1975-6,
and
126-7.
Britannia 1, 1970,
1990,
007,
Fieldwork
148-52,
009,
in
0013,
0014.
Wiltshire,
58, Ant.
WANHM 57,
I960',
bath
(Atrebates) .
Mag.
2,
1856,
500.
HBMC
1983,
Nat.
426
No.
13158.
Scott 1993,
257.
100-2 (1975),
Webster
1975,
VCH Shropshire 1,
114-5 (1991).
1908,
224-6.
Britannia 4,
1973, 335,
fort relocated; 5,
1974,
286-7; 8, 1977, 179; 10, 1979, 283-4; 15, 1984, 234, 470, inscription.
Bruce
1861,
139,
'There
is a well
in the
south brow
49,
1949,
218-9;
(1.46ha)
on
51,
1951, 176-7.
Cumberland
(Cabrocentio) .
late
Bruce,
the
camp,
1978, 224-
TIRBS 1987,
coast,
of
Hadrianic
to
4th
fort
c.
civil
(Brigantes).
1950-60,
62.
leat.
Alian.
?,
?,
172-99,
at
189-9,
bucket.
LAS 2,
114,
1907,
pi.,
well,
bucket,
215.
rectangular
Whitwell
106;
Scott 1993,
172,
1870,
1982,
277,
2.13m
SK
by
51
133-41.
VCH Leicestershire 1,
1.22m,
SE
5,
1.
18m
deep,
and
TIRBS
1987,
57.
(Corieltavi) .
1976,
344;
8,
4, 1973, 305;
334-5;
1977,
451;
12,
1982,
Current Arch.
50,
v(3),
5-6.
Hanson
1970,
370,
item 60,
comments,
'Badly sited,
from the point of view of water supply, on a dry plateau the water for
the garrison of this fort must have come from the only nearby source,
the Westquarter Burn. The water could either have been pumped up from
a point below the fort or, more likely, have been piped into the fort
by means
about
of an inverted
0.8km
(0.5
mile)
siphon
to
the
west'.
396--575.
Macdonald
PSAS 1928-9,
427
of the Burn
1934,
396-575;
94,
194-214.
1960-1,
86-132.
47-51.
96-9.
Robertson 1990,
6, 20,
fort (2.88ha).
5,
1954-5,
settlement
204-5;
from
6,
1955-60,
Iron Age
to
101-2.
Roman,
Britannia
192f
11,
Romano-Brit ish
1980,
173,
ritual
site
(Regni) .
Nantwich, Cheshire.
McNeil & Roberts, 1987, Britannia 18, 1987, 287-8, Fig.5, tank.
1976,
344-5,
3 wells,
8, 1977,
one c.
407.
2m
deep,
c. to
after
1984,
Williams
founded
269;
1969,
c.AD
16,
1985,
98-101.
75,
and
256;
20,
Webster,
abandoned
263;
1981,
c.
AD
57.
21,
1990,
TIRCGLL
130.
306.
1987,
(Silures) .
Nash-
74,
fort
(no water-
related structures).
Neatham Hampshire.
Black 1995, 118, mansio.
283; 3,
1972,348;
tanks, 286;
Millett
4, 1973,
317, pl.xxxv;
1975a.
Millett
1975b,
Current Arch.
watertanks
as
indicated
52, v(5),
origin,...but after
213-6,
278, water
11, 1980, 393, bath; 12, 1981, 384-5; 20, 1989, 178,
5th
6, 1975,
by
timber-lined.
135,
comments,
the conquest
it formed
pottery.
TIRCGLL
186.
Bath-house,
1983,
to the 4th c.
74,
well,
drains,
minorsettlement,
(Belgae/Regni).
2 baths.
Netheravon, Wiltshire.
428
7, 538. Bath.
Netherby, Cumbria.
Birley,
1970,
1961,
42
229-30.
Ant.
(13th ed.),
Itin.
311-4.
Breeze
467.1,
& Dobson
144,
Stephens
1976,
155; 17,
1985(b),
passim.
1986,
441.
225, item
23,
Britannia
1,
Bruce, 1978,
comments,
'The
1969,
16;
seetlement.
fort
and
extensive civil
(Brigantes).
Nettleton, Wiltshire.
Britannia
6, 1975, 45;
TIRCGLL
20,
1989,
74,
possible
JRS 52,
1962, 191.
fort,
succeeded
by
shrines.
(Belgae/Dobunni).
Northumberlandshire.
Britannia 5, 1974, 410; 9, 1978, 419; 10, 1979, 279-80 fort; 11, 1980,
3 58, 410; 12,
Birley 1961,
329, 376-8;
161-3.
CSIR 1.1,
1977, passim.
13, 1930,
TIRBS 1987,
?aqueduct. (Brigantes).
tile
fragments
4,
1973,
458;
6,
282.
Scott
Britannia 2,
1993,
59,
EA
1971,
16;
187,
EA
17.
1976,
Saxon
335.
times.
Simco
Roman
1984,
26-8.
occupation
c.
Occupation
2nd
(Catuvellauni) .
429
to
4th
from the
c.
Iron Age
TIRBS
1987,
to
58.
Antiq. J . 9, 1929,
1971,
173-93;
141-51,
12,
1981,
345-71.
303-4;
Basford 1980,
22,
1991,
123.
303 ; 25,
Britannia 2,
1994,
30.
Scott
Tomlin 1987.
(Belgae).
flat-bottomed
Clarke S, 1996.
Hanson
197 0,
drain1. Clarke
J,
1933.
Clarke
'...a
S,
1995.
371,
comments,
'A
line
of
earthenware
pipes
carrying
water to the external bath-house and nearby latrine from the direction
of the
for the
pipeline is unknown,
excavations
the
within
fort'.
PSAS
84,
1949-50,
1-38.
RCAHM
Roxburghshire 2, 1956, 313 Fig 424 & 316 Fig 426. Jones, Clark, Clarke
& Rush,
17,
ceramic
comments,
pipelines,
(Curie 1911,
99,
'The
one
extramural
of which
bath-house
appears
to
was
run
supplied
towards
224,
by
the
two
latrine
The
fact that the block served only the auxiliary troops of the garrison
(Richmond
1949-50,
bath-house,
58,
Antonine
implies wider
and perhaps
23),
also
(Antonine I)'.
RIB 2120-30.
I
to the
&
II
continued after
distribution
fort.
CSIR 1.4,
Large Flavian
(5.95ha)
retreat
forts;
The
the
system
(4.2'9ha)
and
(c. AD
and
to c.
to date
66.
II
mansio.
163)
legionary
seems
annexe
from Scotland
to
TIRBS
(5.78ha),
Occupation
AD
180(7).
(Selgovae).
Norden, Dorset.
Britannia
1,
1970,
299;
4,
1973,
316,
two
inscribed
TIRCGLL 1983,
75,
altars
shrine,
in
well;
(Durotriges).
1981,
337,
SMR 278.1851,
1982, 38-9.
11.1912,
Condron 1996,
287.1975,
907.1978,
430
374,
Leicestershire
526.1980.
Liddle
15;
436;
372-3.
Antiquity
7,
1976,
337 bath;
8,
219.
1977,
Hands 1968.
Medieval.
1943,
197-8;
1970, 107.
3 wells.
Britannia 4,
400;
9,
1978,
1973,
444;
16; 4, 1974, 9.
297;
11,
5,
1980,
Hakewill 1826.
JRS 34, 1944, 81; 52, 1962, 175; 54, 1964, 166; 55, 1965,
210.
Wilson
?, 1962,
Arch.
23,
11,
1958,
Scott 1993,
1967,
133-4;
268;
24,
160-1,
14,
1959,
1970,
162.
13-21;
33,
Oxoniensia
1968,
138;
8,
35,
Sherlock
1980,
HMSO.
TIRCGLL
1983,
75,
large
courtyard
188,
62,
1846?,
'...where
133.
the
1st
Arch.
J.
period
32,
1875,
baths,
134-54.
excavated
SxAC 81,
in
1866-74,
include a frigidarium drained in one corner...'. JRS 11, 1921, 83; 18,
1928,
197;
19,
1929,
190.
Scott 1993,
151, NK 26.
Tyler 1980,
33.
Scarth 1864,
Scott 1993,
92.
TIRCGLL
well
is 20.7m
deep, circular and lined with masonry and 1.1m in diameter trimmed to
diameter of 1.21m.
from early 4th c.
Coins
but mostly
(Belgae).
Northampton, Northamptonshire.
Dryden 1885, 53-61.
Northchurch, Hertfordshire.
Antiq. J. 2, 1922, 379-80. Berkhamsted Parochial Review 1971, 94 no. 9.
Britannia 5, 1974,
135;
8,1973,
438,
148-9.
464.
JRS 19,
BDAS 1978.
1929,
431
Herts. Arch.
196; 28,
1938,
185.
4, 1974-6,
1-
Scott 1993,
earthworks
of
Berkhamsted Golf
Roman
Course
building,
site.
SMR
TIRCGLL
1337,
1983,
HT
26,
42
at
TL
corridor
00
09,
villa
of
(Catuvellauni). Occupation c.
Northfleet, Kent.
Black
137,8;
395.
1987,
147,
9, 1978,
no.60.
472;
Britannia
10,
1979,
2,
336-7;
1971,
12,
190;
1981,
1973
368, bath;
130
13,
4,
240ff;
56,
1979,
140.
Scott 1993,
69, 1982,
n.8,
1982,
Ansell
201-2;
50,
to
Valens'
period,
from coin.
TIRCGLL
1983,
75.
(Cantiaci).
Northmoor, Oxfordshire.
Arch.
News,
Britannia
17,
15,
4,
Dec.
1984,
1989,
300-1;
4. Benson
17, 1986, 399,
&
Miles 1974,
wells;
21,
map
22.
1990, 334.
Oxford Arch. Unit's Newsletter, 10, 1983, notes 2-4; 16, 1990, ?.
Norton, Northamptonshire.
BNFAS 7, 1972,
Newsletter 2, 1972,
of Northants,
9.
4,
1973, 296.
North West,
3,
1981,
149-158,
CBA G r . 9
RCHM County
item 4, No.46,
Figs.114-
LHA 7, 1972,
284.
10.
TIRBS 1987,
Arch. J.
1964, 12,
Whitwell
(Corieltavi).
VCH Somerset 1,
1906,
Scott 1993,
bath.
432
236.
Gentleman's Mag.
'...in the
1,
1838,
1st half
435;
2,
ii,
1, 1906,
1839,
38; 21,
77.
JRS 49,
1959,
129:
including a bath-building
1875,
67; 35,
1889,
50.
VCH Somerset
(Belgae).
Nursling, Hampshire.
Britannia 26, 1995,
1995,
35-41.
1900, 311.
Composite
distribution
centre.
VCH
Hampshire
1,
AD 70 to AD 380,
Oakham, Leicestershire.
Britannia 18, 1987, 322, 2 wells; 19, 1988, 447-8,
1994,
278,
planked
shaft;
1995,
357-8,
one square,
26,
of an oak cask
(many of
them appeared
to be young
spring-cut
oak)
crossed a
dating
to middle
Iron
Age
(6th-4th c.
to
3rd-lst
c.
BC)
settlement
c.
i.e.
occupation
late
early
(period 3),
Roman
times
(period
2),
to
later
Roman
A large deep well has been reported, see Figs.7 & 7a, opp. pp. 72-77.
Deep well,
21.95m
26.7m
(72ft)
(87ft
when
it
6in)
at a constant diam.
widened
slightly
433
to
of
a width
1.22m
of
(4ft),
1.52m
to
(5ft).
Water was
reached
at
22.48m
(73.75ft),
but
dropped
to about
23.27m
(73 .3ft) .
Odell, Bedfordshire.
BAJ 7,
400;
1972,
9,
engaged
1-16.
1978,
442, Fig.11,
in mixed
wicker-lined
Britannia 6,
'Farming
economy'.
well
in
1975, 256;
based
7, 1976,
on a
single
2 stone-lined wells
area
A.
Tanks,
for
336;
8,
1977,
family
in area
watering
B,
of
unit
and
animals.
Odihaxn, Hampshire.
Britannia 2, 1971, 282-3;
Proc.
(Atrebates).
22, 1944,
31, 1953,
116-26.
Britannia
Scott, 1973.
1970,
402, well,
and
refers
to
buckets
and
water
raising
1,
1970,
1978,
424-5,
54.
TCWAAS(2) 13,
47, 1947,
4th
c.
Ant.
474 vicus.
1913,
166-82, plan;
42
2287
Itin. 467.3
34,
1985,
(milestone);
fort
and
civil
4, 1973,
9-128.
1934, 217-8;
1982,
(c.l.53ha),
(Voreda);
200;
JRS 41,
36,
1936,
51-71.
9,
1951,
132-41;
TIRBS 1987,
late Flavian(?)
or
settlement,
occupied
into
4thc.
(Brigantes).
1982,
TF 292 094,
aqueduct leat.
Cant.
40,
1928,
46;
71,
1957,
1959,
1; 76, 1961; lii; 88, 1973, 223 bath; 89, 1974, 220.
434
xlvi,240;
72,
1958,
210;
73,
Britannia
196;
1993,
79,
baths.
94,
GT
1988,
9,
75-8.
bath-suite.
CBA Calendar
TIRCGLL
Summaries
1983,
76,
1973,
corridor
9.
Scott
villa
with
(Cantiaci) .
1982,
334;
285 bath.
8,
Cambridgeshire
1977,
SMR
398-9;
1808.
22,
JRS
1972,
1991,
50,
1960,
322;
257;
1975,
24,
224-5;
6,
item 3(c),
58,
1968,
occupied by Saxons.
occupation
7,
298.
189-90.
villa;
252;
1993,
Britannia 3,
TIRBS 1987,
4th c.
Later
(Catuvellauni).
Otford, Kent.
Arch. Cant. 39, 1921?,
153ff;
43,
1931,
157;
47,
1935, 236-7; 61, 1948, 182; 68, 1954, 44-5; 70, 1956, 172-7; 81, 1961,
lxi.
Britannia 2, 1971,
1978, 309;
14,
1928, 208.
Comm,
1983,
169,
KAR 6, 1966,
1926,
238,
187; 3, 1972,
121;
244;
4-5.
6, 1975,
17,
118;
1927,
209;
9,
18,
1928-9.
Rivet,
1969,
91,
145.
TIRCGLL 1983,
(Cantiaci).
1972,
1972,
33-4.
327-8,
well;
5,
Current Arch.
1974,
3,
1971-2,
301-3.
?.
TIRCGLL 1983,
436.
CBA Gr.9
211,
Fig.
fiche 589.
1952/3, 224-
VCH Oxfordshire
of site shows
Oxoniensia 17/18,
Swan 1984,
Newsletter
occupation pre-
(Catuvellauni).
1941,
9-21.
Oxoniensia
6,
1941,
9-21;
38,
1973,
215-7,
1976,
1951,
175,
112-42,
178,
180;
15,
p.134 no.60;
435
1984,
101,
336;
1956-7,
20,
1989,
15-57,
201-17.
a temple
well,
stone-lined,
cut
in rock,
5.17m
(17ft)
2 6c 7.
deep.
Rahtz
& Watts
TIRCGLL 1983, 77, Romano-Celtic temple and well. Built late 3rd c and
in use until early 5th c.
(Belgae).
Rev.
3,
1968,
14; 4,
Britannia 4, 1973,
p.82.
PDNHAS 76,
1969,
315.
1954,
36-7;
7,
1972,
Field 1982,
98;
90,
1968,
1982, 71-92.
28.
Branigan 1976,
PDNHAS 104,
171;
91,
71-92,
1969,
Scott 1993,
189;
bath at
92,
1970,
DO 23 .
85-7;
12,
5,
1981,
1974,
21,
13,
20-2,
23-4,
27.
27;
10,
Hanson
1979,
1970,
21-2,
371,
27-8,
item
62,
41,
52-3,
comments,
'The wooden pipeline bringing water into the fort from an unidentified
source to the N-W of the site has been recently discovered' .
1968,
101-92,
Figs.19-24
square
fort.
channel
and
pl.xii..
Stephens
At
1985,
pp.124-6
228,
identified as a water
he
describes
item 51,
pipe
Hogg,
the
comments,
entered
the
'A
fort
intervals of about 7.6m , the pipe trench was found to increase from
0.6m to 0.9m in width for a length of about 1,2m;
solidly packed with clay
to the
junctions
rectangular
pipes
are
known
at
London
(Wheeler
193 0,
PI.
been provided.
channel fashioned from planks, rather than a wooden pipeline. The fort
is late-Vespasianic but the channel must be later than its foundation
since the channel's trench was cut after the 'via decumana' had been
metalled
(Hogg 1968,
reign of Titus
1980,
34-6,
50,
125).
(AD 79-81),
Fig.40.
(Ordovices).
62,
from the
of Domitian'. Wilson
Flavian
fort
(1.8ha)
and
Nash-Williams 1969,
bath.
106-8.
Wilson 1980,
AD 120.
92.
1976, 84-6,
fort occupied AD 74 to c.
(Silures).
1984,
266.
JRS 67,
1977,
15.
bath-house;
Nash-Williams 1969,
TIRBS 1987,
62,
10,
fort
1979,
104-5,
272;
130-1.
(1.71ha)
and civil
settlement.
(Ordovices) .
18,
1987,
Blockley
302-3;
& Musson
19,
1989,
406, 482-3;
1988,
171,
BAR 207.
10,
416;
1979,
23,
269;
1992,
TIRBS 1987,
63,
13,
256.
1982,
O'Leary,
after
which
it
final abandonment.
was
used
as
lead-processing
workshop,
before
(Deceangli).
15,
299.
RCHME
'Occupation dates
3rd c.
BC to the
to 4th c.
19 69,
Peterborough
(6th c. BC)
New
and most
Town,
7,
intensive from
1st c.
13, 1982,
AD...
briefly
1973,
14-5.
Britannia
1990,
342-4.
Friendship-Taylor
12,
1981,
342;
Current Arch.
1989:
Upper
19,
1988,
452;
Nene
Archeological
20,
1982,
Society.
Occupation c. 3rd c. Northants. Arch. 14, 1979, 106; 24, 1992, 99-101.
RCHME Northamptonshire
2,
1979,
63,
site
11.
used
as
Scott
198.
437
1993,
144,
SMR
8m diam. by 3.5m
water
supply
for
bath.
drain
and
corridor villa.
timber
water
pipe-line
found.
TIRBS
1987,
63,
(Catuvellauni).
118-9.
313, 380-1;
11,
9(1)
1958,
large
55, 1976,
1939,
Wooler
9(2),
1917.
(4.58ha) and
6, 1975,
234-5;
9, 1978, 474;
100,
1941,
TIRBS
1987,
63,
1976,
285;
pi. 25.
127-38;
7,
1979,
Sommer, 1984,
43-68 plan;
165-75.
fort
245.
461;
Current Arch.
235-77;
Britannia 5, 1974,
1991, 302.
TAASDN 7,
10,
RIB
1950,
1936,
285-309;
1021-7,
2293;
(well preserved),
(Brigantes).
1973,
128.
VCH
Wiltshire
1, 1957,
110.
TIRCGLL
1983,
79,
(Belgae).
Pitney, Shropshire.
Britannia 11,
11,
1863,
1980, 296-7;
22-4.
Scott 1993,
12,
79,
1981, 163-4;
14,
1983,
13ff.
SMR 54410,
SO 47.
PSANHS
SO 46;
326-8.
(Durotriges).
Pitsford, Northamptonshire.
RCHM County of Northants, S-W, 4, 1982, 204, item 10, well.
Clwyd, Wales.
Britannia 26,
326,
1995,
dependent on
326.
Kent Arch.
bath-house.
bath.
Well,
VCH Kent
Newsletter,
3,
1932,
constructed with
11,
122-4,
timberwork
1987,
item 40,
(p.123).
Poole, Dorset.
438
1-3.
Scott
Fig. 29.
107,
163,
TIRCGLL
KE76,
pi. 26,
1983,
80,
Britannia 12,
1981,
383, well;
19,
1988,
475;
16,
1985,
308.
RCHM
1971,
46-7,
wells;
3,
283,
n .149,
1972,
n.189.
318.
Antiq.
J.
Britannia
270-5,
348;
4,
43,
1,
1974,
1963,
1970,
222-3.
300-1;
160-72,
227,
2,
Arch.
Rev.
1971,
283,
318.
Cunlifffe
16,
1985,
303.
Scott
1993,
17,
SMR
429,
439,
AV
65-67.
Well.
28,
1971,
1872,
180,
32-3,
Roman quarry.
161-80.
Dorset
PPS 24,
Proc. 84,
1958,
118.
112.
RCHM County
1962,
of
'Guide to Early
(Durotriges)
Cambr.
210.
92, 1937,
The
214-20.
settlement
dates
12, 1981,
RIB 444;
BBCS 23,
from
c.
1969
AD
187-9.
80-140.
Bath,
two
Bockley
possible forts;
1989,
aqueduct,
8, 1977,
358-
(Deceangli).
Puckeridge, Hertfordshire.
Britannia 2,
442;
1971, 270;
3,
1972,
329, wells;
4,
1973,
299;
8, 1977,
234.
Cunliffe,
294-6; 2(23),
1974, 101.
06 18,
1909,
1973,
377-8.
174,
77-8.
193;
Hull,
2, 1981,
1963,
46.
142; 75,
271,
273-5;
PSA 2(8)
1934,
20,
1899,
a drain amongst
67, at Nutbourne,
bath structure.
Sussex
122; 2, 1971,
County
Magazine
8,
1934,
677-8.
439
VCH
SMR 23 67, WS
indicating likely
Mins.
93,
& 95,
1970.
Sussex
3,
1935,
63-4.
TIRCGLL
1983,
83,
villa at
Pulborough Farm;
baths;
coins
of
1st
c.
(Regni) .
Pumsaint, Carmarthenshire,
Wales.
273;
lined,
19,
lm
1988,
420;
diameter
and
21,
1990,
304-6
over
4.8m
deep;
Carmarthenshire Antiquary 9,
10, 1974, 3-12.
1973,
3-27.
1974.
1991,
203-7,
210.
(1.89ha),
stone-
Carmarthenshire Antiquary
vicus; 10,
reduced
Bath-building on
(Demetae).
1993,
90,
SMR 3228,
HE 17.
TWNFC 1822,
258;
36,
1958,
84-7,
Quernmore, Lancashire.
Britannia
2,
Leather
254;
1973.
(?aqueduct)
kiln.
1971,
3,
Swan
1972,
313 ; 4,
1984,
fiche
1973,
423,
283;
7,
1976,
stone-lined
403.
culvert
(Brigantes).
Quinton, Northamptonshire.
BNFAS 5, 1971,
1975,
233
25; 7, 1974,
Sc
235?,
Northamtonshire Arch.
building B.
Fig
9,
30-1.Britannia 3,
9;
7,
1974,
95;
1972,
1976, 334-5;
12,
1977,
216,
322, Fig 8; 4,
8,
1977,
399.
well
near
plan,
Radfield, Kent.
Britannia
5,
1974,
459;
KAR
14,
1968,
10,
Radfield', 2 wells.
166;
50,
1960,
214;
51,
'The main
1961,
68-70,
161.
plan.
PSAS 97,
1963-4,
Stephens
189-9,
1985(b)
224,
440
TDGNHAS
39,
1960-1,
24-49.
TIRBS 1987,
(0.64ha), within
451-2;
11,
1980,
379;
13,
1982,
374-5,
Fig.8, two
Ravenglass, Cumbria.
Britannia 1,
107.2
1970, 53-4,
65, Ant,
1-138.
19,
1988,
256; 20,
TCWAAS(2) 38,
Potter 1979,
134-5;
Cos.
1928,353-66;
TIRBS
58,
1958, 14-30;
85,
Hadrianic
1985,
81-5.
fortlet
(c.
{Brigantes).
Rayne, Essex.
Britannia 19, 1988, 459.
lkm west
of
Rayne,
Smoothy 1989,
Fig.l.
Well,
diam.
at
top,
Reach, Cambridgeshire.
Antiquity, 45, 1971, 224-5.
1973,
245.
Scott
7,
1978,
RCHM County
45,
pl.ivB.
of
Cambridgeshire
CA
TIRBS 1987,
2,
1972,
88-9,
JRS 63,
Fig.80.
VCH Cambridgeshire
large villa,
with bath-suite.
(Iceni ?).
XXVIII, 8,
1986,
444.
1977,
15.
84;
bath,
drain;
2,
296.
1971,
BMQ 24,
172-94;
1969,
4,
58-63.
1973,
257,
Britannia 1, 1970,
261 n.47 Not.
Dig.
1969,
233, 242.
drain.
Johnson,
1970,
1979,
(6th ed.).
Pre-Roman occupation.
45-8.
Johnston
TIRCGLL 1983,
Small Claudian
fort
of
mid 1st c.; then fort of the 'Litus Saxonicum' , 3rd to 4th and 5th c.,
with vicus.
(Cantiaci).
Redlands Farm, ?
Britannia 22, 1990, 252-4, 2 stone-lined wells.
441
Reigate, Surrey.
Arch. J. 6, 1849,
Scott 1993,
288.
177,
Britannia 25,
SMR 1054,
SY 38.
1994,
288.
Williams
Hooper 1945,
1984,
SyAC 75,
16-7.
111-53.
Well, drain.
Richborough, Kent.
Britannia 1, 1970, 42-3, 65, Ant. Itin. 463.4, 466.5, 472.6
Ritipus), 240-8 date of fort; 2, 147 n.26
3, 1972,
393-5;
238,
245,
178-9;
1975, 175,
287-8,
17,
177,
442-3;
comes
from
1979, 22,
comments,
10,
1986, 427;
157-8,
43; 15,
basin or cistern
1926, 19,
unlike
much larger
at
basin, tile
Lincoln
1969,
1926,
built
pi.
is known,
and
made
1985b,
of
but
102- 125.
less ornamental
1956, 32-6).
TIRCGLL 1983,
invasion
1993,
1987,
1863,
27,
villa
19.
hexagonal
228,
in
comments:
coated with
248),
fountain
In
Later
with
'basin'
was
SxAC
fort(0.2ha) and
fort of the
'Litus
(Cantiaci).
Essex.
319.
Rodwell
1978,
bath;
not
lower
227-8)'.
RIB 46-65;
in AD 43.
is
the
fact the
(Brown 1971,
85,
and
in
Archaeologia 14, 1803, 62; 16, 1812, 364; 35, 1853, 62.
JBAA 19,
1985,
Vol.I,
tiles
(Thompson
bridgehead of Claudius'
98-9.
16,
Stephens
180;
8, 1977,
(Bushe-Fox
colonia
200;
the
1984,
193, 195,
fountain or water
136,
180,
Bushe-Fox
'An hexagonal
(ad portum
coins
from
31,
Dunnett 1975,
No.
170-1,
Claudius
38.
Scott,
Fig.34.
to
TIRBS
Arcadius.
(Trinovantes).
Rippingale, Lincolnshire.
Whitwell 1982, 294, TF 12 NW 15, bath.
Rivenhall, Essex.
AEx 1972, 51.
2, 1971,
174-93;
3,
1972,
33 5-6; 4, 1973,
442
3 05;
5, 1974,
444;
7, 1976,
200;
8,
1977,
Rodwell in Todd,
ES 60.
(ed.),
VCH Essex 3,
with bath.
439;
13,
1982,
1978,
1963,
211.
11-32.
Scott 1993,
65,
55,
1986.
SMR 8059-91,
(Trinovantes).
Rochester, Kent.
Arch. Cant. 18, 1889, 193-5; 21, 1895, 1-16; 27, 1905, lxvii-lxx,
177-
92;
110;
28,
1909,
lxxxviii- xcii;
29,
39,
1927,
159-64;
84,
1969,
1,
472.3
1970,
473.8
42-44-6,
Ant.
Itin.
1987,
359;
22,
1991,
{Dubobrius) ,
(Durobrivis) , 473.3
Britannia
7, 1976, 377;
292.
JRS
1971, xvi-
14,
1924,
138.
1980, 17;
Cassius
Dio
'Historiae Romanae', LX, 20, the decisive battle in AD 43, between the
British
Dudley
forces
and
& Webster
Frere
1978,
the
1965,
80-1.
Roman
invading
army
under
67-70.
Fisher
1772,
History
VCH
Kent
3,
1932,
81-2,
105,
109.
TIRCGLL
1983,
85,
Pre-Roman
Wight.
Britannia 7,
Fig.25,
1976,
367-9,
drain,
Aulus
Plautius.
of Rochester.
Fig.14,
p.81.
'Rochester
Wacher
1966,
oppidum with
mint.
(Cantiaci).
spring140m west
of house.
(Belgae).
2,
1971,
179,
181,
183,
187,
190,
193 ; 3, 1972,
348,
Fig.18; 10, 1979, 148 vousoir tiles,168, 174, 176; 11, 1980, 394, date
mid 1st c. to c.AD 200, roof collapsed c.3rd c.; 14, 1983, 329f, bathsuite
(east wing), c. end 3rd c., perhaps replacing earlier baths; 18,
1987,
1990, 355,
1963,
well,
150,
Hewitt,
164;
55,
1965,
e t . a l ., 1983,
c.4th c.,
5.5m,
water
first,
level
at
J.
228;
140,
56, 1966,
129-50,
0.92m stone-lined,
2.75m;
SU 095
211;
217,
Arch.
at
east
21,
88; 53,
214,
219-20,
225.
2 baths,
2 wells;
west
well 2.3m
deep,
0.92m
diam.,
Hewitt M A T ,
idem,
3rd
Interim
Report,
443
1971;
idem,
Final
Report,
1974.
Rockingham, Northamptonshire.
Northants.
Archaeology
Northamptonshire,
9,
Central,
1974,
2, 1979,
75-7.
RCHM
126-30, No.49,
county
of
item 3, Figs.113-
Rodersham, Kent.
Britannia 17, 1986,
c.
25,
1994,
265-6,
fort
previously
unknown,
2.5 ha
(6.2
acres). 3 wells.
Romford, Essex.
Black
1995,
120
settlement and
It in.
480.7
site
43,
121
Fig.2 position
mansio,?bath.
(Durolito)
Britannia
( Note:
there
43,
vicus
1, 1970, 52-3,
seems
to
be
or
roadside
65, 73,
some
Ant.
uncertainty
both
sites
identification
are
given
discounted;
the
same
20, 1989,
TIRCGLL 1983,
location);
304.
86,
6,
TEAS 7,
1975,
93,
95.
VCH
1900,
(cf. Chigwell).
(Tinovantes) .
Rothley, Leicestershire.
Liddle 1982, 38-9, Fig 26.
TLAHS 9, 1904-5,
157-8.
Scott 1993,
216-7.
Whitwell
YAR 1977, 3.
3,
1972,
7,
18,
75-6.
JRS 50,
quarters
building.
PSAS
1978-80,
230-85.
RCAHMS
1793,
1990,
pi.xxxv.
45;
7,
1976,
St Joseph 1983,
Robertson,
29,
56-9
Wooler
39,
I960,
1905,
1917,
213;
442-99;
well,
190 bath.
military way;
85-6,
Stirlingshire
bath-house,
301,
1,
52,
67,
CSIR
1962,
1.4,
Frere &
163,
head
1933, 243-96;
110,
1963,
located
53.
14,
100-2,
inside
1985,
plan.
fort.
Roy
Nos.
73-5.
Roushaxn, Oxfordshire.
Scott 1993,
Bath.
310.
Bath.
12.
Britannia 2, 1971,
253; 3,
1972,
310-11;
1; 54,
Richmond
SMR
1964,
4139,
HU
156;
55,
17
1965,
bath.
204.
Scott
1973,
1973,
90,
1963.
Vol.l,
4,
Scott 1993,
70-95.
Stead
100,
1980.
Yorkshire Archaeology 1973, 6. YAJ 31, 1934, 366-76; 32, 1936, 214-20;
33,
1938,
81-6,
222-4,
320-38.
TIRBS
1987,
67,
courtyard
villa,
(Parisi).
67-73.
watertanks.
1981,
333;
Scott
13,
1993,
1982,
46,
VCH Cheshire 1,
353-4.
SMR
1987,
CAB 8,
882,
210-1.
CH
1982,
3,
49-52;
bath,
9,
drainage,
Dated to c. AD 170-200 to
13,
1982,
329;
20,
tank,
timber-lined latrine;
256.
21,
1989,
1990,
258-9,
304,
299-302,
1989, Britannia
fort,
vicus,
20, 249-54.
timber-lined
304;
23,
1992,
Flavio-Trajanic
period.
23,
1992,
284
settlement;
24,
1993,
272-3;
25,
1994,
272
Sambourne, Warwickshire.
445
Sandringham, Norfolk.
Gregory 1982,
360.
Scott 1993,
137,
SMR 3254,
bath,
NF 164,
3rd to
4th c .
Sandwich, Kent.
Arch. Cant. 94, 1978, 191-4.
402;
12,
1993,
1981,
317.
368;
19,
56,
1979,
KAR
484
133-4;
settlement;
63,
1981,
22,
56-60,
1991,
292;
Parfitt
24,
& Jones
1981, bath.
TIRCGLL
58, Ant.
1983,
92,
Itin. 486.4.
minor
settlement;
coins
of
53-4,
Victorinus.
(Belgae).
1996,
38-40,
319;
Leicestershire
bath-suite.
1, 1907, 207.
SMR
408.1860;
Pickering 1935,
520.1968.
162-7.
Liddle
VCH Leicestershire
{Corieltavi).
Scampton, Lincolnshire.
Britannia 5, 1974,
424,
58.
vol
Scott
Whitwell
1987,
c.
1973,
1970,
68,
Vol.2,
466.
1,
56,
courtyard villa,
Illingworth 1810,
151-4,
82-3,
Vol.2,
144.
6-13.
Fig.60.
Whitwell
Northants SMR
Todd
1982,
semi-detached bath-house,
1973,
81-3.
105-6.
TIRBS
probably of 4th
(Corieltavi).
(4th ed.),
settlement.
Stead 1976,
657.
257, 259.
Trollope 1872,
TIRBS 1987,
57-8.
White
68, extensive
(Corieltavi).
Scole, Norfolk.
Britannia 1, 1970,
187; 5, 1974,
2 timber-lined wells,
1975,
80,
439,
corrected
to
Stoke Ash;
"small
town"
7,
1976,
388;
9,
1978,
446
2, 1971,
of buildings
327;
6,
17,
'an area of
found and 7
wells,
of
which
2were
number of plans,
show wells.
timber-lined.
EAA
5,
1977,97-224,
and
xvii,
TIRBS 1987,
occupation
4th c .
68,
minor
settlement;
(Iceni).
J.
1,
1970,
222-45.
& Miller
wells
1979, 210,
shown.
1964, 13-4; 9,
235-6.
TIRBS
Whitwell
1987,
30,
68,
Britannia 20,
1982,
Iron
79-80,
Age and
1989,
Loughlin
301, Fig.14a,
Roman
settlement;
(Corieltavi).
4,
12,
1973,
1981,
Antonine wall';
1972,
40-1;
107-12.
274;
5,
143-9,
13,
1974,
Magdonald
JRS 59,
1969,
1934, 239-40.
102-3.
9,
(0.68ha),
fortlet
2 culverts;
culvert
Stirlingshire 1, 1963,
primary
407,
'stone-capped
1982,
51.
1974,
PSAS
(NS 822
Hanson
Sc
105, 1972-4,
Robertson 1990,
1978,
passed
288,
202-4.
9,
60-2.
415-6,
through
the
797).
DAES
Keppie
157.
1979,
RCAHMS
TIRBS 1987,
small
68fort
J. 49,
1892,
297;
10, 1979,
326;
180.
19,
Britannia 1, 1970,
1988,
476;
JRS 11,
1921,211;
66-8; 102,
1970,
247-8.
(Durotriges).
Scott
9, 1978,
420;
Hadrians
Wall, 11,
1980,
10,
1979,
358-9;
12,
280,
Mile
1981,
Castles
269,
394;
19, 1988, 456; 21, 1990, 341; 23, 1992, 288. ?aqueduct.
447
34 and
17, 1986,
35
on
403;
Shakenoak, Oxfordshire.
Britannia 1, 1970, 288, 311; 2, 1971, 268, 301; 3, 1972,
1973,
297-8,
villa,
bath-house;
166; 9, 1978,
332;
73,
TIRCGLL 1983,
4 vols.
10,
1979,
5, 1974,
160-2.
94,
436;
6,
Brodribb,
7,
1976,
1968-
1975,
121, 328; 4,
188;
site A occupied AD
(Dobunni).
Shawel1/Churchover, Leicestershire.
Arch.
J.
5,
1848, 328.
RSNHSR 1939,
JRS 43,
1953,
?;
118,
pi.23;
31.
PSA
1949,
53,
(2nd
1963,
134.
ser.), 8,
Shepreth, Cambridgeshire.
Fox 1923,
185.
PCAS 6,
1885,
60-1.
Scott
1993,
42,
SMR 03364,
CA
TIRBS
(Catuvellauni).
5,
PSANHS 134,
1974,
1990,
346;
22,
47-56.
1991,
Scott
279-82,
1993,
170,
Fig.25;
24,
SMR 25160,
1993,
SO
303.
52.
VCH
Somerset 1, 1906, 318, well, shown on plan, but not mentioned in text.
TIRCGLL 1983, 94 settlement.
(Belgae).
Shireoaks, Nottinghamshire.
EMAB 1959, 13; 1961, 15; 1962, 22.
156.
Scott 1993,
SxAC 116,
1978,
Shortlanesend, Cornwall.
Britannia 11, 1980, 389, drain.
448
Siddington, Gloucestershire.
RCHM 1976,
101-3,
Scott 1993,
c. 2nd c. to 4thc.
Sussex
{Regni) .
Silchester, Hampshire.
Archaeologia 53,
30, pi. xxiii,
1983,
6; 54 (2),
1895,
485.7,
485.8,
486.7;
107-14; 5, 1974,
1976,
152, 162,
1978, 307-8,
5;
12,
165,
100, 330-1;
362;
15,
486.8
3,
124,
1990 272;
13,
1984,
76 wells,
22,
1972, 146,
283;
409-
1982,
1991,
122,
255; 2, 1971,
357;
78, 91-2,
201,
Fig.24,
16,
drain,
1985,
200-1,
418-9;
124-32,
9,
129, 39414,
1983,
water supply,
Hanson 1970,
1973, 38
209-10; 7,
389-91;
187, 316;
4,
324-6;
21-2.
114,
302-4,
69, 241,
(Calleva Atrebatum) ,
211, 233,
6, 1975,
10, 1979,
(Calleva),
171, 368-71; 8,
338, 464-5;
1981,
1897,
55(2)
and drains.
1958,
101-19.
Pitt-Rivers,
sinking).
deep,
1.2m
(36ft
diam.,
small
cavities,
27-8,
pls.i,
2 wells,
1)
(4 ft) diam. ,
and about
called
Vol.I,
1887, 7-239,
ii,
iv, v,
evenly-cut
originally
1.25m
PPS 24,
2 deep wells.
193,
well
198,
shaft c.
into
The well
which
Pitt-
(100ft)
located as a depression
(4ft) deep.
"put-logs"
of
11m
timber
was
inserted,
probably for climbing down and up the shaft. The timber was so located
that a bucket could still be lowered down the well.
There seems to be
because of the draw-down from modern wells in the vicinity which are
deeper than the Roman well. A Roman type wooden bucket was found at a
depth
of
56.5m
(186ft)
down
the well,
449
probably
indicating
that
in
did
For both the wells water could have been present at the time but that
the
(2)
small well,
0.91m(3ft)
(193).
modern times;
2(2),
44,
67 n.96,
151; 2,
10,
1971,132,
254; 7,
1976,
231,
JRS
190, bath.
YAJ 26,
(Brigantes).
Sleaford, Lincolnshire.
Britannia 17, 1986, 390.
167; 54,
1964, 159.
EMAB 1961, ?.
Whitwell 1982,
305-6,
well.
Cant. 46,
1983,
335;
1985, 4.
1934,
17,
1986,
202-4;
427;
Scott 1993,
82,
20,
1967,
1989,
192-217,
326.
tank.
Kent
Britannia 14,
Arch. Newsletter
VCH Kent 3,
1932,
124,
9,
item
16,
1985,
302f;
341;
Cotswold
20,
1989,
Community
310,
3;
2,
TBGAS 107,
SU
1988,
033
Age
12.7ha site.
School
Iron
22,
1988,
277,
1987,
occupation;
16,
1991,
18,
257.
963,
Bath-
1980,
83-102.
206-9; 5, 1974,
318-9,
355;
Britannia 2, 1971,
165,
12,
407 vicus;
1981,
321;
450
7,
15,
127,
1976,
1984,
130,
306;
277,
n.54,
295;
8, 1977,
371;
342;
16,
1985,
268, 325-6; 17, 1986, 374-6, 448-9; 18, 1987, 314-5, Fig.8; 1988, 4313; 20,
112,
1989,
267-9;
272-3;
24,
1994, Vol.1.
1993,
1962,
267-9,
315; 22,
284;
in a courtyard,
1954,
1052-1071.
25,
1991,
1994,
232-3,
Fig.6; 23,
264-5.
1992,
color photo
193.
1990,
South Shields';
suite',
21,
11.
p.25.
JRS 50,
Hanson 1970,
1960,
237;
52,
CSIR 1.1,
1977,
passim.
TIRBS 1987,
71,
RIB
(Brigantes).
7, 1976, 352,
354;
376-7;
292;
11,
14,
19, 1988,
344-5,
313-4;
464; 20,
1973,
15, 1984,
1974,
12,
310; 16,
466;
6,
9, 1978,
1981,
8002 a well,
timber drain,
5,
409-10;
381-2;
307;
388; 8,1977,
1980,
1983,
4,
1991,
270-1,
473-5;
10,
1979,
443;
13,
1982,
17,
1986,
353-4,
1985,
1975,
298;
273; 23,
1992,
409;
293-4,
timber-lined,
c.
3rd c.;
revetted channel,
iron collars.
Cowan
1992,
3-191.
At p.101,
piped,
possibly
to the bath
mentions
a wooden
a probable indication
of
the mansio.
Sheppherd J D,??.
London
Britannia
3, 1972, 265, 267-8, large villa, original bath replaced by larger one
with elaborately
13-54;
71,
apparently
1981,
172.
1967,
13-7;
79,
1975,
in
3rd
c.
451
1959,
14.
Whitwell
13-4.
TTSN 5, 1901,
58-9; 70,
VCH Nottinghamshire
EMAB
2,
101-2,
308-9.
TIRCGLL 1983,
4th c.
71,
pottery of 2nd to
(Corieltavi).
Scott
1993,
126,
LI
NAR SK 91 NW 5.
1982,
309,
well.
1993,
191,
1983,
SMR 4344,
95-6,
WS91.
90;
17,
VCH Sussex 3,
Cunliffe 1973,
1966,
1935,
SxAC
280-1.
25,
70.
Scott
TIRCGLL
and
(Regni) .
Sparsholt, Hampshire.
AEx 1972, 10.
170-193,
building
1972,
348,
Fig.15,
1981,
163;
22,
1972b,
materials,
bath;
1991,
283,
4,
288.
1973,
Age
318-9,
&
Roman
no.151;
Current Arch.
12,
occupation;
5,
1969,
1974,
318;
14-8.
3,
12,
Johnston
that a damaged structure near the stoke-hole may have been a cistern.
Scott 1993, 87, HA 93, NAR SU 43 SW 18.
Spilsby, Lincolnshire.
Arch. J. 91, 1934, 181,
52,
1889-90,
13ff.
1953,
162-6.
651-68.
JBAA 38,
96,
courtyard
1882,
J.
215-6.
47,
1890,
JRS 49,
420-1.
1959,
127.
TBGAS 14, 1889, 208; 32, 1909, 300; 33, 1910, 12;
RCHM
Gloucestershire
Arch.
villa
Cotswolds
1976,
with
bath;
coins
of
3rd
and
113-4.
TIRCGLL
4th
c.
(Dobunni) .
44 Ant.
Itin.
2,
1971,
452
170-92,
Britannia 1,
building materials,
288;
3, 1972,
351;
4, 1973,
323 ; 6,
9,
3, pl.xiv.
TIRCGLL
1983,
Romano-Celtic temples.
96,
settlement;
187;
8,
16, 1985, 184, 189; 20, 1989, 178; 21, 1989, 297.
Rodwell Sc Rowley 1975, 39.
183,
cult
of
springs
with
(Cantiaci) .
J.
Excav.
102,
1945,
rept.
21-110,
Excavation
Britannia
1,
1970,
6,
1975, 198; 8,
1970,
62-5.
4838,
HT
Scott,
57.
1993,
TEHAS
95,
Report;
289-90;
50.
1971,
1977,
SMR 1469,
1945-6,
2,
118,
HT 54,
TSAHAAS
1961,
169-93,
328.
100-35,
building
Herts. Arch. 2,
SMR 4264,
(n.s),
4,
HT 56,
1932,
SMR
212-4.
WSWHASB 3, 1955, 2, Bricket Wood; 4, 1956, 2, Park Str.; 18, 1973, 123, well.
site of native farmstead and enlarged in masonry c.AD 150, and again
late
3rd
c.,
and
abandoned
mid
4th
c.;
detached
bath-house.
(Catuvellauni).
Staden, Lincolnshire.
CBA Gr. 14, 3, 1988, 43-5, bath, occupation 1st to 4th c.
Staines, Surrey.
Britannia 1, 1970,
359;
6, 1975,
49,
64 Ant.
92; 7, 1976,
Itin.
374; 8,
478.4
(Pontibus) , 292;
1977,
408-9 well;
Current Arch.
161-2;
2(14),
1976,
Burnham
& Wacher
Pontes'.
Smith
52,
1975,
362-5;
1990,
134-8.
3(7),
306-10,
1987,3 65.
TLMAS 27,
Roman name
1972,
1979,
468;
London Archaeologist 1,
?, 180-6.
gives
18, 1987,
10,
3,
1976,
1970,
71-134.
'Pontibvs' or
73.
"Ad
TIRCGLL 1983,
Occupation
from late
(Catuvellauni).
Stainfield, Lincolnshire.
AN Lincs.
1967.
453
LI 155.
Whitwell
Stainley North, ?
Wooler 1917, 190. Bath.
Scoreby.
corridor
between,
and
well.
13ha
enclosure adjacent
to
River
1943,
261-9,
Derwent'.
Stancil, Yorkshire.
JRS 30,
1940,
bath.
TIRBS
167.
1987,
71,
villa
170-2.
with
YAJ 35,
bath-block.
(Corieltavi
or
Brigantes).
Standon, Hertfordshire.
Scott 1993,
1890.
96,
SMR 1101,
HT 62,
bath,
found
in 1756,
and again in
20,
1989, 297;
1989, 8-9;
21,
1990,
334.
Oxford Arch.
Unit News
1988,
17(2),
Scott 1993,
206,
SMR
bath.
1987,
72,
1990,
366.
SP
842430,
well,
Fig.16,
1953-60, 198-215.
associated
with
TIRBS
buildings.
(Catuvellauni).
J.
3, 1846,
116-29.
Scott
454
1993,
161,
SMR 1357,
0X52.
VCH
Stantonbury, Buckinghamshire.
Antiq. J.
1976, 337-8,
farmstead;
Arch. J. 146,
11,
1980,373 ;
1989, 135-278.
Britannia 7,
13,
Stanwick, Northamptonshire.
Britannia
wells,
17,
1986,
396-7;
19,1988,
culverts(drains), watertank,
452-3; 20,
spring,
1989,
149-71,
possible aqueduct,
12
165-7
longer working; 21, 1990, 333; 22, 1991, 285, bath and watermill; 23,
1992,
285,
312;
25,
1994,
306.
Neal
1989,
Britannia
20,
149-71.
Stanwix, Cumbria.
Britannia 1, 1970,
1153; 3, 1972,249 n.
271;
19,
17,
TCWAAS(2)
154;
85,
that
this
1986, 441;
1988,
499;
the
148; 7, 1976,
22,
1991, 234;
25,
1994,
263-4.
wall
and
Horsley's
observation that the fort must have had a bath-house with an aqueduct
water
supply.
Hanson
('Britannia Romana,
the
remains
of
1732,
stones
197 0,
155)
372,
item
68,
comments,
'Horsley
which
"resembled"
Stephens 1985,
those
226,
from
other
shown
Roman
'A
(1732,
TIRBS
the crossing at the river Eden which was the main route to the west of
Scotland. Occupied from 2nd c. till 4th c.
(Brigantes).
Stebbing, Essex.
Britannia 9, 1978,
66
452; 20,
(Rodwell's no.33).
social
aspects
such
1989, 305.
Rodwell
as
1978,
status
and
Scott 1993,
1-38,
ethnic
65,
at p.31,
origin
SMR 1259, ES
comments on the
of
the
owners
Burials seem
limitations
(p.14)
of
he mentions
such details
are obvious.
In his
but
last paragraph
for
interpreting
455
of
1988-9,
18-9.
imbrices;
a water
system
using
wooden
pipes,
of
which
iron
rings
survive.
1970,
1964,
287;
33-54.
00170,
3,
1972,
Artis 1828,
226,
236,
'Durobrivae', p i .1
238.
JRS 48,
1958,
139.
231.
Scott
1993,
CA 142, well.
Swan 1984,
fiche 377
Britannia
LHA 2,
42,
SMR
80,
1896-7, 143.
SyAC
Scott
1993,
126,
SMR 7164,
LI 159,
illustration,
bath.
Scott
plan and
1973,
Vol.l,
1831,
385-7;
(Corieltavi).
1959-60.
Archaeologia 22,
EMAB 2, 1959,
9.
7163/6691,
160/161,
1982, 314.
LI
LAASRP 9
(n.s.),
bath.
Scott
1829,
1961,
1973,
26-32;
16.
456
Scott
Vol.l,
23,
1993,
159-60.
126,
SMR
Whitwell
(Corieltavi) .
158,
bath.
Stukeley
1883,
Scott 1993,
Vol.76,
323.
126,
Turner
Whitwell 1970, Vol.II, 80. Whitwell 1982, 314. TIRBS 1987, 72.
(Corieltavi) .
Stone, Buckinghamshire.
Archaeologia
32,
1847,
451-5
Buckinghamshire 4, 1870,
wells;
34,
122; 9, 1909,
1852,
209,
21-32.
270;
10,
Records
1916,
87.
of
RCHM
1980,
375,
Fig.16,
shows
9 wells
and
12,
1981,
bath;
16,
1985,
341;
13,
Fig.20,
Stonesfield, Oxfordshire.
Archaeologia 11,
bath.
Gough's
54, bath.
1794,
37; 37,
'Camden'
1857,
2, 1806,
434-5.
15.
Oxoniensia 6, 1941,
Scott 1993,
162,
1-8
SMR 1232,
OX
7,
13-4.
Scott 1993,
1938-9,
1963, 20-2;
VCH Sussex
3,
1935,
64.
Occupation
of
site
site
Strathgeath, Perthshire.
Britannia 2, 1971, 248; 5, 1974, 402-4; 6, 1975, 225; 7, 1976, 300; 8,
1977,
361,
plan,
1979,
274;
11,
iron spade;
1980,
351-2;
9,
1978,
411,
12,
1981,
20,
457
amphora-urinal, 482;
319;
13,
1982,
33 6-7;
10,
14,
1983, 284-7; 15, 1984, 274-5, 348; 16, 1985, 263-4; 17, 1986, 371; 18,
1987,
36
camp,
309-10,
bath-house,
drains;
24,
1993,
278.
GAJ 4,
Barton 1958,
Trans.
in BASTP 74,
1958,
Soc.
74, 1958,
10-11.
Jones
1975,
6-9.
St
Webster 1958,
149,
158,
177-78.
Stroud(1), Gloucestershire.
Glevensis 11, 1977,
GS92/3,
38.
Cashe's Green
Scott 1993,
75,
SMR 3563,
75,
SMR3563/3588,
(SO 85 05).
TBGAS
Stroud(2), Hampshire.
Williams 1908, Arch. J. 65, 57-60; Williams 1909, Arch. J. 66, 33-52,
pls.i-iv.
TIRCGLL
1983,
97,
corridor
villa
with
detached
bath,
(?Regni).
5,
198.
1974,
457
enclosures
Scott
1993,
162,
with
SMR 2852,
stone-lined
OX
55.
well.
JRS
57,
TIRCGLL
1983,
97,
(Atrebates).
207,
VCH Wiltshire 1,
1957,
110-1.
11,
1980,
Cambridgeshire 2,
Lode',
375,
1972,
Roman
112-3,
buildings.
No.8,
item 82,
RCHM
County
of
probably
of
(3.33miles)
Roman
long
origin,
extending
an
in
artificial
a
direction
watercourse
across
the
c.
5.4km
Fens
from
Commercial End', p.114, co-ords TL 5559 6322 to TL 5219 6725, pi.6, &
Fig 104, p.114. In Medieval times it was used as a drain for the Fens.
458
Bloxhamist,
well.
66,
1960,
Scott 1993,
No.477.
VCH Oxfordshire
1,
1939,
308-9,
Swanwlck, Hampshire.
Ashbee 1978,
225.
4.25m
30-33;
Fox
depth 2.74m.
Scott 1993,
206,
SU 15 83,
2 wells.
WAM 68,
WANHM 50,
1973,
SMR SE
134;
1942-44, 100.
72-3,
TIRCGLL
(Dobunni).
Tallington, Lincolnshire.
Britannia 20, 1989,
176.
128.
LAASRP
Whitwell 1970,
2 wells.
Rev.
1976,
36.
5,
1970,
20;
Britannia 1,
7,
1972,
1970,
28.
299;
BAA
2,
1846,
1971,
281;
179-82.
3,
1972,
Branigan
346-7;
4,
1973, 316; 5,1974, 455; 6, 1975, 278, bath; 11, 1980, 391-2; 12, 1981,
172,
359-60,
383 well;
13,
17.
1973,
PDNHAS 91,
91-3;
96,
1982,
411;
15, 1984,
343; 19,
1988,
Current Arch.
80, 1981,
260. Hutchins
410.
1861-
1969,
189-90;
1974,
64-6;
92,
98,
1970,
1976,
151-2;
61-2;
94,
99,
1972,
1977,
86-7;
124-5;
95,
100,
1978, 117-8; 101, 1979, 140-1; 102, 1980, 90-1; 104, 1982, 184-6; 105,
1983,
146-8;
106,
1984,
4th c. Bath-suite,
058 017C, DO 29.
3rd and 4th c.
118.
well,
9.12m
(30ft) deep.
Scott
1993,
54,
SMR 2
17,
channel
1970 191;
1986,
82.
7,
1976,
Hanson,
157;
1970,
9,
1978,
372,
379,
comments,
382;
'A
10,
1979,
stone-built
leading from springs to the north of the fort fed the early
459
forts I,
II, III; forts II, III in succession were built on the foundations of
fort
I. Aqueduct,
stone-lined channel
from a springhead;
2 baths,
Teston, Kent.
Scott
1993,
108,
KE
90.
VCH Kent,
1932,
125-6,
item
45,
Fig
30.
Bath, drain.
Tewkesbury, Gloucestershire.
Miles
& Fowler
development,
1972,
well, stone-lined.
implications
of
Thatchazn, Berkshire.
Harris
1931,
92-101;
vi (2),
1931,
92-101;
Harris
1937,
vii(4),
219-55.
Smith 1987,
1937, 219-55.
244.
TNFC
TIRCGLL 1983,
98,minor
(Atrebates).
Theale, Berkshire.
Britannia
26,
1995,
368,
'excavation
revealed
Romano-British
post
built structure with adjacent well. Two further wells may be of Roman
date'.
Thenford, Northamptonshire.
AEx 1972,
62.
BNFAS
Age occupation;
4,
1973,
Newsletter 2, 1972,
143-5.
1902,
Scott
201.
1971,32-3.
294,
Britannia 3,
bath-suite;
1993,
147,
Rivet 1969,
5,
1972,
1974,
325,
434.
360,
Iron
CBA G r .9,
SMR 124,
NH 109.
pi.3.17.
Whitwell
VCH Northamptonshire
1982,
320.
1,
TIRBS 1987,
(Catuvellauni).
Thetford, Norfolk
Britannia 12,
306;
1981,
347-8;
13,
1982,
410;
14,
1983,
308-9;
15,
1984,
EAA
122,
parts
tubes,
and
saturation
of
the
checking
was
near
pits
the
or
100%,
by penetrating
degree
or not.
of
saturation.
it would
460
indicate
the
the
that the
degree
pit
of
had
penetrated the water table, and therefore it was most likely a well) .
Scott, 1993, 138, SMR 5683 St Helen's Well, NF 192, SMR 17397, NF 193.
TIRBS 1987, 74, buildings within a tripple ditch enclosure
(c. 3.6ha).
(Iceni).
1985,
pi.15.2,
52,
179.
EMAB
137-8;
1962,
49,
172-3,
Scott 1993,
117,
1959,
Figs.19,
115-6,
11.
JRS 47,
113;
50,
1960,
224;
20;
55,
1965,
207.
SMR SK 91 NW D.
1982,
320.
1957,
212;
51,
4, 1973,
48,
1961,
168,
1958,
98,
175, Fig.22;
Morris,
Smith 1978,
227;
1979,
169,
141.
178.
VCH
TIRBS 1987,
74-5, minor
corridor villa, built early 4th c, and destroyed about 50 years later;
likely bath-suite.
Corieltavi) .
15;
65,
villa built
stone.
1961,
6.
40-1.
JRS 53,
TTS 58,
Whitwell
in timber,
with
1982,
323,
bath-suite,
c.
bath.
AD
TIRBS 1987,
240;
later
75,
built
in
(Corieltavi).
Thurlby, Lincolnshire.
Arch.
J.
91,
1935,
121,
unlikely to be Roman.
AA
44347/1
NMR
probably built
184,
LAASRP 9 (n.s.),
20813,
?aqueduct,
c. AD 125.
probably native,
1961-2, 21.
part of
Car
Section of c.190m
Dyke
in
Park
Wood,
in length, about
east of the Roman road, King Street, on the edge of the Fen.
175m
Whitwell
Thurnham, Kent.
Arch.
Cant.
1993,
108,
74,
KE
1960,
91,
162-70.
probable
Britannia 2,
1971,
174,
187.
Occupation 1st
Scott
to
4th
1,
1970,
129;
11,
1980,
369;
12,
1981,
16,
1985,
238-40;
20, 1989,
288; 23,
461
13,
339-40;
15,
1982,
1984, 296, SP
1992, 285,
310-3,
13,
SP 217 556
show 14 wells,
Fieldhouse,
drains.
May
&
WMA23,
Well
1980 119,
Stood,
121.
1931.
TIRBS
1987,
75,
minor
{Dobunni).
Tilston, Cheshire.
Britannia 1, 1970,
45,
Ant.Itin.
469.3
(Bovio) ; 11,
1980,
365; 12,
1981, 333; 14, 1983, 299; 15, 1984, 255-7, 341; 19, 1988, 445.
1987,
75;
settlement
with
bath;
pottery
mostly
of
2nd
TIRBS
and
3rd
c.
(Cornovii) .
Buckinghamshire.
of Buckinghamshire 2,
1862,
33-50.
RCHM 1913,
Vol.2,
299,
Scott
1993,
30,
SMR
87,
BU
76,
bath.
VCH
Figs.41,
42,
Titsey, Surrey.
Archaeologia 59,
1905,
pt.2,
1344, SY 47.
SyAC 4,
214-37.
367-9.
Scott
1993,
TIRCGLL 1983,
178, SMR
98, corridor
(?Atrebates/Cantiaci).
198,
TLAHS 34,
324.
bath-house.
1958,
84;
McWhirr
35,
1959,
1970-1,
85;
46,
TLAHS
1970-1,
46,
1-8.
1-8
JRS 23,
(reprint).
Whitwell
1982,
(Corieltavi) .
13,
SMR 1472,
Maclean,
AV 2.
TBGAS 12,
comments,
'The
discovered
would
apparatus
and
bath
establishment', 202.
accommodation
1888,
hypocaust
159-69;
13,
chambers
1889,
recently
to
so
large
an
(Dobunni).
117, 1968,
462
further details
tank.
Bowen &
Gresham 1967,
230-5, map.
Houlder C,
1974,
Fig.17.
JRS 47,
1957,
JRS 51,
1961,
Totternhoe, Bedfordshire.
Britannia
1963,
2,
1971,
61-4.
110,
Rainey
113-4,
1973,
116.
149.
Scott
1993,
214-5.
21,
SMR
Mathews
534,
BD
30.
TIRCGLL 1983, 99, courtyard villa with bath; pottery mostly of 4th c.
(Catuvellauni).
Towcester, Northamptonshire.
Britannia 1, 1970,
42,
1975,
corduroy
255,
timber
14,
49,
Ant.
Itin.
470.6,
aqueduct;
1983,
347;
15,
7,
476.11
1976,
1984,
(Lactodoro) ; 6,
335;
300 bath;
8,
1977,
399,
16,
1985,
289;
20, 1989, 292; 22, 1991, 253; 23, 1992, 285; 24, 1993, 293.
Wacher 1990, 5, 9-10, 117, 142, 152-7, 160.
S-W,
4, 1982,
149-60, No.61,
6.
TIRBS
1987,
76,
item 5, Figs.113,
Burnham &
114.
Northants Arch.
town
(settlement)
enclosing
c.
11.25ha.
(Catuvellauni).
Wales 24,
1984,
Occupation
50; 25,
1985,
27-8;
26,
from C.
AD
7 0s
to
1986,
43-4;
28,
1988,
final
abandonment
c.
AD
125-3 0.
Aqueduct leat.
23,
1969,
187-90.
Caernarvonshire 2,
1960,
Nash-Williams
259-60.
Scott
1954,
1993,
81,
33-5.
RCAHMW
GY 1, bath-house.
(Ordovices).
(under
(Tripontio);
Caves
Inn)
1,
79
Ant.
Itin.
477.2
49,
1; 5, 1974, 431;
83,
1970,
130-79,
at Caves
Inn,
near Rugby,
Grid Ref.
SP 57,
535
2)
stone-lined,
circular,
463
bottom courses
rectangular,
c.
9.1m
(30ft)
deep;
3)
(45ft)
in dispute;
level was
deep;
6)
4)
11.3m
top 3m
(37ft);
stone-lined,
(10ft)
5)
12.5m
destroyed,
largest
(41ft)
of
deep;
total
the wells,
7)
stonelined
Twyford, Hampshire.
JRS 49, 1959 131, well, bath.
bath.
VCH
Hampshire
1,
1900,
3 09-10.
TIRCGLL
1983,
100,
winged
(Belgae).
Twywell, Northamptonshire.
Northants Archaeology 10,
1975,
31-93,
full report.
RCHM County of
Upham, Hampshire.
Arch.
J. 6, 1849, 397.
JBAA 5,
Scott 1993,
88, NAR
59,
1971,
1969,
190.
231.
Down 1981.
TIRCGLL
JRS 57,
1983, 100,
1967,
198;
58,
1968,
(Regni).
Usk, Gwent.
Britannia 1, 1970, 55, 57, Ant.
192,
273, wells;
3 02,
3 54;
1976,
3 55;
18,
252-3.
Fort,
1973,
1971,
272;
298-9, 391-2;
8,53,
263;
4,
2,
11,
1980,
1987, 309;
19.5ha,
c.
AD
1974,
8, 1977,
3 51;
Manning 1981.
fortress,
5,
19,
stone-lined;
401-2;
446; 9, 1978,
12,
1988,
7,
1981, 2 69;
423-5; 20,
(Burrio) , 179,
6,
1975,
483-4;
3, 1972,
223-4,
10,
188,
78,
293;
7,
14,
1983,
3 49; 16,
1985,
1989,
245,
265; 25,
1994,
TIRCGLL 1983,
100,
RIB 396;
Flavian
fort abandoned
mid
2nd
c.,
464
civil
Itin.
471.3, 476.8,
110,
196-9;
3,
1974,
267,
283;
1977,
199,
288,
passim 219,
1972,
479.8,
80-1,
122,
146,
passim;
4,
49,
161,
256-7; 2,
1973,
41,
306;
442; 9,
1978, 39,
12,
47,
1957,
455-6;
217; 50,
Figs.24-6,
bath;
1988,
1960,
pls.xvii,
12, bath;
20,
327;
1981, 386-7;
1989, 300;
225-7,
xviii.
21, 1990,
Figs.27,
Frere
passim,
13,
28;
sewers,
Figs.15
&
31,
spring.
1979,
30,
133-4;
51,1961,
1983,
19; 3 wells,
Vol.II,
19-20,
Verulamium was
JRS
178-180,
5,
179; 8,
1982,
276,
1972, Vol.I,
largestcivitas capital
10,
39,
336;
1976, 175,
1971,
10,
several
Figs.89 &
occupied
from
Hertfordshire
166-
8.
Sudeley 1877,
Sc
13 and 15 pis.
1971,
124.
Witts
1883,
Scott,
66,
RCHM 1976,
112-3
No.21.
JBAA
(2nd ser.),
I,
(Dobunni).
E,
183 9,
Whitwell
1982,
Topographical
327,
TF
45
NEJ.
Sc
Historical
TIRBS
1987,
Account
78,
of
Wainfleet.
salt-boiling
site.
(Corieltavi).
Walesby, Lincolnshire.
AN Lincs.
1978, 434;
{Lincs. N
(n.s .),
Sc
683.
Jewitt L, 1878,
LAASRP 6, 1861-2,
59-61.
Scott 1993,
Britannia 9,
141.
135-8;
128,
Whitwell
(Corieltavi).
Wall, Staffordshire.
AEx. 1975,
189,
197, bath-house;
465
2,
1971,
260;
3,
1972,
316;
well; 8, 1977, 263, 394; 9, 1978, 435-6; 10, 1979, 296; 11, 1980, 367;
12, 1981,
287; 21,
49.
337;
13,
1982, 356;
1990, 371.
16, 1985,
1989,
1979,
bath,
spring.
1-50;
8,
1966-7,
12-29,
1-40.
1-14.
TSSAHS
TIRBS 1987,
(Cornovii) .
Proceedings
88,
1966,
107-8.
1985,
55-86,
drains.
3,
1972,
enclosure wall,
193,
388-9,
fort;
7,
1976,
lead sealing;
306-8,
11, 1980,
cavalry
barracks,
270;
8, 1977,
360.
JRS 59,
1969,
121;
63,
1973,
(2.29ha);
extra-mural bath-house.
193.
65-81.
c. 2nd c.
1971,
SyAC 2,
114,
187,
TIRCGLL 1987,
190,
1950,
(?Regni or Atrebates).
Walton-Le-Dale, Lancashire.
Britannia
13
1982,
stone-lined well,
1984,
284-6,
352,
supply base,
phase
III,
Figs.6, 7,
date,... Phase
III,
as
c. AD
drain;
130-140
79,
Flavian
1983,
to 3rd and
296-7,
large
4th c.;
15,
period 2,
the
14,
to
early
3rd
(Brigantes).
Walton-on-Thames, Surrey.
466
buildings
associated
c.,
military
supply
with
TIRBS
base(?).
Scott
1993,
178,
SMR 722,
SY 48.
Whimster
1931,
147,
238,
ceramic
drainpipe.
Walton-on-the-Hill(1), Surrey.
Black
1987,
134-40,
(TQ 23
53)
comments
that
because
there exists
of
high
a problem of
the
114 no.32,
occupied in pre-Roman times, dating from the latter part of the Iron
Age', based on pottery of c. AD 10-43. The bath appears to have become
ruined c. AD 270-274
(coin of Titrius
II),
and reconstructed c. AD
280-300.
Walton-on-the-Hill(2), Staffordshire.
Antiq.
doubt,
J.
14,
the
(2nd
s.),
position
1881-83,
of
the
110-1,
bath
of
the
have
no
establishment',
Edwin
Britannia 1, 1970,
57-8
Sc
terminated c. AD
271-3,
2; 11,
WZ
176.
275; 4, 1973,
1980,
115-26,
227; 7, 1976,
Figs.lSc
282, timber
362; 8, 1977,
2, earlyoccupation
WANHM 63,
1968,
1975,
16,
1985,
2, 1971,
42-3.
3, 1972,
Figs.1
Black 1995,
110;
233-5.
65,
204-5;
TIRCGLL 1983,
104,
66,
1971,
188-9.
major
settlement,
20, 1989,
305. Essex
(Dobunni).
73.
132.
461;
248, No. 98, a pond is said to occupy the site of a 'Roman well',
the present Hermony Pond.
Stukeley,
Lethieullier.
3,
Britannia',
23.87.
VCH
Vol.II,
Essex
50.
1963,
Lysons,
Ware, Hertfordshire.
467
'Environs
Gough
of
1789,
London1,
in
154,
'Camden's
iv,
n.d.,
Britannia
Fig.12, well;
Occupation
from
1st
10,
to
6, 1975, 260; 8,
1979,
350;
5th c.,
12,
when
1977, 401,
1981,
347;
403; 9, 1978,
21,
finally abandoned.
1990,
340.
Large,
deep
Washingborough, Lincolnshire.
LHA 4, 1969, 104; 7, 1972, 9, aqueduct leat, from TF 024 708 to TF 303
706.
Wasperton, Warwickshire.
Britannia 13, 1982, 361-2; 14,
one timber-
Berkshire/Oxfordshire.
Britannia 5,
1974,
277;
11,
1980, 1,
13 as Durobrivae;
Coney Field,
at TL
1161
at TL 1148
9682,
9708,
two wells.
(2)
Swan 1984,
fiche 382,
Billing
VCH Huntingdonshire
1,
(Catuvellauni).
1935,
28-9.
TIRCGLL
1983,
104,
corridor
villa with
(Regni).
Weekley, Northamptonshire.
BNFAS 5,
1971,
26.
Britannia
434. Northants.
Arch. 1,
Northants,
4,
1993,
148,
1982, 329.
c.
S-W,
1982,
SMR 2660,
1902,
202,
2,
1971, 266-7;
NH 114.VCH Northants 1,
5,
41-94.
2),
468
272-4,
181,
RCHM County of
item
1902,
al3 .
194.
Scott
Whitwell
(Catuvellauni) .
Weldon, Northamptonshire.
1974,
1979,
164-6,
115
No.61,
Smith,
Hird
Scott 1993,
NH
Arch.
22, pp.23, 67, Fig. 11, shows 2nd house and bath-suite.
TIRBS
6c Dix
148, SMR
1439,
133-4;
(Great Weldon).
1956,
193.
1988-9,
Northants
TAMS 1
Whellan 1874,
814.
1st to early
2nd
(Corieltavi) .
Res.
were
fed
Rep.
by
1.
a
Gilyard-Beer 1951.
Hanson 1970,
stone-built
373,
item 73,
aqueduct
0.45m
comments,
(1.5ft)
'The baths
wide
by
0.47m
(1.58ft) deep which led from the Well Beck, a stream running from east
to west to the north of the site'.
YAJ 7,
465-6,
1882,
284-5;
aqueduct,
34,
1939,
cistern.
Tyler 1980,
342-9;
35,
TIRBS 1987,
1943,
81,
89-90.
226;
Lukis 1882.
36,
1947,
250,
(Brigantes).
Wellingborough, Northamptonshire.
BNFAS 2, 1967, 20, 28-9; 3,
1969,
219.
Northants,
Northants.
Central,
1969, 6.
Arch.
1, 1979,
10,
166-70,
1975,
No.62,
142-4.
RCHM
JRS
59,
County
of
works and
(Catuvellauni) .
1,
1906,
312-5,
item 20.
(Belgae).
1971,
occupation;
depth 30m.
4.
TIRBS
24-5;
1974, 47;
1976.
Britannia
3,
1972, 311,
Iron
Age
81,
corridor
villa;
(Parisi).
Welwyn, Hertfordshire.
469
occupation
c. AD
YAR 1974,
250-
350.
Arch.
165,
7-9.
27(7),
1971,
106-9.
excavation report.
Herts.
269-70;
Arch.
Rook 1968.
317-20;
18,
1938,
4, 1973, 331.
1,
1968,
117-8;
Scott
1993,
96,
9,
339-
Current
1987,
79-
SMR 99/1913,
HT
166-8. 2 bath-houses,
TSAHAAS 1937.
VCH,
23,
1585, OX 60.
1992,
287-8,
stone-lined well.
Scott
1993,
162,
SMR
25,
1994,
175-211.
Fulford,
19,
1994,
1988,
175-211,
of
pottery
fabric material,
and
which
is
other
said
small
finds,
including
some
preserved
Roman
in the
remains
have
also
been
of Roman Britain.
recovered
and
it
has
Much
been
conjectured that the wetland marshes were used as fields for growing
fodder
for horses
The settlement
seems
to have
been extensive, but not much building remains have been located up to
1992.
208,
Wiltshire 1, 1957,
119.
VCH Hampshire 1,
WAM 13,
1872,
33-5;
22,
1900,
1885,
311-2.
243-50,
(Belgae).
1995,
350:
'4ha excavation,
but enclosures
and fields
from 0. 5ha to lOha in extent. Many buildings with several rooms each,
and some hypocausted, propbably indications of bath-structure'.
470
TF 36 SE OS. Well.
401.
TIRBS 1987,
81-2; villa near Icknield Way; bath-house, with pottery of the 3rd and
4th c .
(Iceni).
Westbury, Wiltshire.
Scott 1993, 208, SMR SE 300, WZ 177.
1971,
133;
6,
GAJ 5, 1979,
228;
7,
1976,
12-8.
1975,
20
178.
301;
10,
1979,
278,
Macdonald 1934,
253-8.
TIRBS
321,
bath-house.
Settlement
8,
1977,
384;
14,
1983,
35-44,
the
early
unenclosed
16, 1985,
Dent 1983.
excavations
1964-80.
1985a,
comments
to
(Wetwang)'.
it
aqueduct.
Loughlin
'...the
Ramm
& Miller
1979,
identification of a
1980,
95,
summary of
140.
Stephens
82,
settlement;
{Parisi).
1980,
Geophysics
36/78.
Rahtz 1979.
363 ; 12,
Scott
drain.
1993,
00984/04028,
153,
SMR
David A, 1978,
NK
38/40.
1978,
1981,
427;
12,
329-30;
13,
1982, 349;
471
15,
1984,
282;
16,
1985,
278-9; 19, 1988, 440. Occupation from Iron Age through to 5th and 6th
c.
Scott 1993,
153,
farming; ?bath..
TIRBS 1987,
82,
settlement,
(Parisi).
Wheatley, Oxfordshire.
Scott 1993, 162, SMR 2760, OX 62, bath.
3.
TIRCGLL 1983,
4th c.
(Catuvellauni) .
1971,
120, pl.xi.
Britannia 2, 1971,
17.
250 pl.xxxivB,
TIRBS 1987,
82,
fort;
fort.
(Bannaventa) , 477.1
(Isannavantia) , 479.5
(Bannavanto) ; 3,
1972, 9.
Northamptonshire
1,
RCHME Northamptonshire 3,
1902,
186-7.
WMANS 1971,
CBA Gr. 9
1981, 150-2.
14.
VCH
TIRBS 1987,
82,
Whitchurch(1), Shropshire.
Arch. J. 34,
53, 64-5, Ant. Itin. 469..4, 481.1, 482.4 (Mediolano), 190 early fort;
8,
1977,
394;
9,
1978,
436-7,
1845,
Herefordshire
537.
197.
TIRBS
1931,
253.
82,
Flavian
282.
254.
RCHM County
of
23, 1992,
Herefordshire
1931, 253.
timber-lined wells:
1900,
RCHM
2 timber-lined wells;
fort
VCH Herefordshire 1,
and
later settlement.
(Cornovii).
1877, 363.
RCHM Herefordshire,
South-West,
472
1,
1931,
122.
Lewis 1845,
253.
Scott 1993,
91,
SMR 08494.
TWNFC 1882,
Herefordhire 1, 1900,
258;
122.
1900-2;
36,
TIRCGLL 1983,
1958-60,
105,
227-33.
iron working.
VCH
?bath.
(Dobunni).
1929,
98-103.
Somerset 1, 1906,
334,
Scott 1993,
172,
SMR 53262,
spring,
SO 69.
lead piping.
VCH
TIRCGLL
(Durotriges).
Cooper T S,
Scott
1993,
Nevill
175
Leveson-Gower
(typed copy
1534,
SY
1883,
9.
The
Whitley, Derbyshire.
Britannia 26,
1995,
3 50,
and considerable
1925, 249-60.
Stephens
1985b,
226,
item 35,
comments,
'An altar
{RIB 1198)
bath-house
(Blair & Carr 1925, 254). The altar may be of the mid 3rd
c.
date
that,
(Wright
although
1943,
38).
a dedication
to Apollo
bath-house
rather
rather
than
the
suggests
Nymphs,
the
than
the
fort,
since
altar
the
fort
Wooler,
1917,
would
190.
(Brigantes).
RCHM 1976,
TIRCGLL 1983,
105,
126-8, Figs.
TBGAS 71,
473
1953,
bath-house
of
earlier villa of
2nd c.;
latest
coins
of Honorius,
AD
393-423.
(Dobunni).
Northamptonshire 4, 1982,
168-9.
Scott 1993,
148,
107-14.
RCHME
1993,
45,
Cambridgeshire
SMR
7,
04308/04309,
1978,
CA
46.
181/182,
TIRBS
drainpipes.
1987,
83,
VCH
settlement.
(Catuvellauni).
190;
15, 1984,
courtyard villa.
51,
1988.
212.
Moore
Bath,
1990,
341,
water pipe.
301; 21,
TIRBS 1987,
83, corridor
(Trinovantes).
summary report.
reconstructed c.
late 3rd c.
or early
4th c.;
16,
1985,
181.
Jarrett & Wrathmell 1981, Celtic Studies 5, Fig.2, 40, Fig.23, well,
44, Fig.25 section of the well, 7.4m deep, an octagonal frame over the
well with 8 posts.
59,
26.
1983,
1969,
201,
Fig
TIRCGLL
174; 58,
105-6,
1968,
pre-Roman
176;
farmstead;
(Silures) .
Wickford, Essex.
Antiq.
J.
pl.lxx.
(2nd
ser.), 50,
1970,
268-71,
274-5;
52,
176, 273; 3 1972, 335; 10, 1979, 349; 17, 1986, 356-8.
CBA Res,
1972, 338-40,
Rept. 34,
Roman village,
lined well,
68.
dated
1980,
65-8,
Pre-Flavian
late
3rd c.
or
Wigginton, Oxfordshire.
474
fort
2, 1971,
Buckley
and
1980.
prehistoric
4th
c.
remarkably
well
Beesley
1848,
41-3.
193.
Scott 1993,
bath.
TIRCGLL
JRS 56,
163,
1983,
1966, 208.
OX 64.
106,
Oxoniensia 29-30,
VCH Oxfordshire
villa;
coins
of
1,
1939,
1964-65,
309,
Victorinus
to
No.3,
Valens.
(Dobunni).
bath building;
6,
209;
220.
55,
1965,
1975,
Black
1987,
163-70, 195;
MOW Excav.
23,
13-36.
SxAC 78,
1992,
Annual Rep.
64. Britannia 1,
306.
1964,
JRS 19,
10.
122
1929,
Scott 1993,
1970,
tiled-drain; 81,
1974,
(Regni).
Wilderspool, Lancashire/Cheshire.
AEx 1974,
50-1;
1976, 36-7.
Britannia 1, 1970,
281-2;
6, 1975,
240;
8, 1977, 385 clay-lined tanks; 23, 1992, 282; 24, 1993, 285; 25, 1994,
269.
3, 224,
228-32,
179;
58, 1968,
1903-4,
209-37.
1886, 260-73.
182.
320,
May
Fig.73,
1900.
Thompson 1965.
?aqueduct,
May
THSLC
1903-4.
52, 1900,
115-236.
Watkin
(Brigantes).
144,
Roman well,
2.74m diam.,
1841,
Wilsford, Wiltshire.
Antiquity 8, 1934, 459-61;
1962, 16.
1935,
pl.xxxlx,
Fig.2.
Hoare Vol.l,
1812-
21, 206.
Winchester, Hampshire.
Antiq.
J.
Britannia
44,
1964
1,
1970,
to
55,
49-50,
1975.
60-1
Ant.
475
Itin.
1964,
478.2,
bath-house.
483.2
(Venta
Belgarum,
486.11
wooden water-pipe;
2,
1971,
168-93, 283-4; 3, 1972, 271-6, 348-9 wells; 4, 1973, 40, 162, 318; 5,
1974, 456; 6, 1975, 279; 7, 1976, 182, 204-6; 8, 1977, 419;
465-6;
10,
1979,
331-2;
11,
1980,
395-6;
12,
1981,
9, 1978,
362-3;
13,
1982,
391; 15, 1984, 326; 16, 1985, 201 water supply, 311-2; 17, 1986, 421;
18,
1987,
288;
349-51,
370;
20,
1989,
316,
318;
21,
1990,
356;
7.
During
investigations
the
indicated
a possible
1991,
Current Arch.
reconstruction
22,
of
line
the
for
M3,
the
archaeological
aqueduct
for
the
from
small
scale
maps
given).
Wacher
1995,
291-301,
Nicolson
Wingham, Kent.
Donker G, 1882, Part I, Arch. Cant. 14, 134-9, 2 figures of the bath;
15,
1883,
villa;
351-7,
82,
1967,
Part
lx.
61-2,
II,
sketch
drawing
of
excavation
1971,
187,
193 ; 20,
Britannia 2,
at
Wigham
1989,
199.
(Cantiaci).
J.
33,
Dorset Proc.
County
1953,
11,
74-5.
1890,
of Dorset,
Britannia
1-6.
300-5,
19,
1988,
Bromhead 1942,
item 21,
476;
143.
21,
1990,
RCHM 3(2),
'Romano-British Well
353.
1970,
on Kingston
(60.8m), and cut into chalk limestone and is 1.1m in diam, and
25.8m
Winterton, Lincolnshire.
Antiq. J. 41, 1966 72-84.
Britannia 2,
1971, 258 villa; 3, 1972, 315; 4, 1973, 286; 5, 1974, 424 water-leat;
6, 1975,
11,
1980,
245-6;
366-7;
7,
1976,
12,
327-7 well;
1981,
330-1;
13,
476
8,
1977,
1982,
391;
351;
10,
14,
1979,
1983,
295-6;
296;
15,
1984,
283;
16,
pl.viii(l);
50,
early 3rd,
51,
1961,
1985,
281;
1960,
17,
221,
1986,
'bath-suite
387.
in use
JRS
49,
1959,
from
late
109,
2nd c.
52,
1962,
167;
54,
1964,
157,
159,
Fig.10;
55,
1965,
item 237,
which
to
item 236,
Stead 1966.
234,
supplied water
1965, 21.
to the baths,
drain,
25,
Fig.14,
tank.
EMAB
104.
5,
PBNSS 1,
1908-9,
63-4.
RCHM Dorset
1975,
104-10.
Scott
109, well,
Finberg 1955.
spring.
Smith 1969,
97-101.
TIRCGLL 1983,
(Dobunni).
Wittenham, Berkshire.
VCH Berkshire 1, 1906, 219-222, Fig., p.221, 3 wells.
Warwickshire
purposes,
1,
1904,
1.3m square
249,
well,
and greater
than
probably
used
13.5m deep,
for
but
ritual
bottom
not
reached. At 6.7m 'was a large square stone with a hole in it, on which
stood urns of grey ware'; 12 were intact and 12 were broken.
Wollaston, Northamptonshire.
Arch. J. 3, 1966, 1-6.
ditches.
1987,
settlement.
84,
RCHM(E)
corridor
Northamptonshire
villa
with
2,
Hall &
bath-house,
180,
site30.
near
native
(Catuvellauni).
Woodchester, Gloucestershire.
Britannia 5, 1974, 451; 13, 1982, 197-228, with plans and figures and
pis.xiv-xvii. Clarke
figure.
1982.
Lysons
1797.
RCHM
1976,
132-4,
477
and
TBGAS
1956, 172-5.
TIRCGLL 1983,
107,large
(Dobunni).
p.4 comments
'Fresh
water
river,
(16ft),
or from
to draw water
from the top of the underlying clay'. The site is along Peddar's Way
Roman
Road. Appears
Claudian
fort
on a
"Primi
in
C(enturiae)
Britannia, 9,
Primi"
1978,
(Property
480.
of
Smith 1987,
the
157,
a pit
of
comments,
'A
century
made of
(Norwich Museum
Cunliffe 1974,
285.
Woolaston, Gloucestershire.
Arch.
311-2;
97-9.
93-125;
102,
1953,
100.
Britannia 20,
1989,
McWhirr 1981,
(Dobunni).
Woolstone, Oxfordshire.
Berkshire Arch.
J. 57,
1959,
83-5.
JRS 46,
1956,
144, bath.
Scott
Worcester, Worcestershire.
Branigan
&
Fowler
(eds.),
9,
1978,
1976,
114-5.
439;
16,
1985,
Britannia
8, 1977, 397,
1988,
water
287; 23, 1992, 284; 24, 1993, 293; 25, 1994, 272.
70,
105-6.
312-3.
PNRB 1979,
TWAS(3)
2,
496,
1968-9,
Rav.
7-116,
Worcestershire 1, 1901,
203-8.
extra-mural
coins
settlement,
Cos.106.30
at
dating
478
from
1970,
191
water pipeline
supply;
20,
1989,
(Vertis) .
15-19,
TIRBS 1987,
1,
63,
84, town
late 3rd
75,
Smith 1987,
84-5.
VCH
with extensive
to
4th
c.
and
evidence
of
iron
smelting
during
this
Age
ocuppation
upper part
period;
Iron
(Dobunni).
Worth, Kent.
Antiq. J.
was in brick earth and was lined with flint and clay,
part was in chalk and unlined.
TIRCGLL 1983,
temple.
107,
Lewis,
1966, 3.
198.
(Cantiaci) .
Wortley, Gloucestershire.
Britannia
278
18, 1987, 341 villa; 19, 1988, 470; 20, 1989, 312;
bath-suite;
23,
1992,
295-6. Glevensis
20, 1986,
22, 1991,
41-4.
Bath,
large
bath-suite;
occupied
c.
TIRCGLL 1983,
AD
250
to
c.
107,
AD
350.
(Belgae).
(369);
2 drains,
water
Antiq. J. 46,
cistern(water tower)
distribution
Fig.167.
The
1966, 229,
1970), aqueduct
(133-
Fig.11; major
sewer
water
supply
network
this
structures
1975.
also
applies
other
Barker
than
(n.d.),
Britannia 1, 1970,
261-2 baths,
428-9;
1966-74),
other
247-8;
City:
Itin.
the
1962,
1976,
469.6
4,
27-3 9.
328-30;
8,
1973,
1977,
Barker
1966-80'.
(Urioconio), 482.9-10
287;
2, 1971, 20,
5,
316-7;
water-related
Excavations
1975, 78,
7,
of
TBAS 78,
Roman
basilica; 3, 1972,
6,
aspects
aqueduct.
'Wroxeter
(Viroconiorum) , 484..9
197,
the
to
323-4,
394-6
1974,
158,
excavated
reservoir,
14, 1983, 302-3, 337; 15, 1984, 345; 16, 1985, 204-5 water supply; 17,
1986, 391-3;
1993,
292;
20,
25,
during 2nd c.
271.
JRS 43,
1953,
1991,
Bromhead
118;
44,
250;
1942,
1954,
23,
1992,
185, many
93;
48,
321-2;
wells
1958,
24,
closed
98;
50,
1960; 51, 1961; 52, 1962, 169-70, '... N of the post office, a cistern
479
constructed after AD
155, and robbed of its tile lining within the Roman period,
in the late 3rd century to early 4th century AD';
1964, 162-3, 165, Figs.14, 15.
Wacher 1995,
159,
bath-house,
362-77,
389,
433,
Fig.11,
53,
probably
1963,
?;
166,
168,
sewer
distribution
369,
bath-house,
channel
Fig.167.
1965-8,
197-219;
59,
drains,
RIB
1969-74,
cistern
291.
TSAS 56,
133-7;
70,
167,
(369),
Fig.11,
Webster
24-31;
54,
57,
water
1988,
120-44.
1962-3,
112-31;
1975-6,
5-39.
TIRBS
Wyck, Hampshire.
Britannia 20, 1989, 319, bath.
2nd c. to 4th c.
Parker 1878, 3.
28,
1908,
SMR,
BU49.
VCH Buckinghamshire
2,
17-9,
plan.
Scott 1993,
TIRCGLL
late 2nd c. to
(Catuvellauni) .
1,
1939,
331
(sv Banbury),
well.
TIRBS
1987,
85,
(Dobunni).
889.
1976,
340 settlement.
bath-building,
coins
date
JRS 12,
rooms
from
AD
1-3 pi.
193
(Catuvellauni).
480
to
xx,
1922,
256.
Scott
1993,
392.
Samian
TIRCGLL
ware
1983,
of
2nd
55,
c.
4,
1970,
14-5;
6,
1971,
19,
Yardley
Hastings
3.
Central,
Swan
1993,
318.
Northants. Arch.
Durobrivae 3,
11, 1976,
1975,
15-8.
JRS 27,
114-5;
13,
1937,
1978, 77.
Scott
1993,
149,
SMR 1711,
NH 125,
site dates
(TL 061
(Catuvellauni).
1887, 353-62.
JRS 55,
VCH Somerset 1,
1906, 306-7,
1965,
74, 1828,
122-
item 2, Fig.66.
(Belgae).
1991,
TIRCGLL 1983,
107;courtyard villa
late 4th c.
1928,
with bath;
153.
occupiedlate 2nd to
(Durotriges) .
York, Yorkshire.
Antiq. J. 44, 1974, 200-17.
1970,
41-3,
47,
49,
52
Itin.
466.1,
Britannia 1,
468.4,
475.7,
478.6,
sewer, 325,
280-
83-6, 236-
7, 289; 7, 1976, 165-6, 168, 315 well; 8, 1977, 330-2, 382-3; 9, 1978,
426-7;
10,
1979,
160-1,
436;
11,
1980,
318;
12,
1981,
13, 1982,
349; 14,
1983,13ff,
16, 1985,
279; 17,
1986,74-6,
7; 19, 184-9,
275,
267.
325-7;
22, 1991,
Hanson 197 0,
watermains
that
341; 23,
1992, 273;
the
24,
'A section
water
394-5;
373, comments,
distributed
328-9,
147-57, 221,
1993,
286; 25,
of one
of
the lead
an
aqueduct
brought
by
1994,
101-7;
481
61-99; 44,
Whitwell
shows
1976,
1-55,
high,
pls.ia-xvib,
fall 0.74m,
covered
with
which
the
sewer
Millstone
slabs
1.20
channel
of
0.45m wide by
0.75
0.3
m,
(scale 1:100).
355-7.
SheahanSc Whellan,
Stephens
1985b comments,
(scale 1:200)
Vol.l,
1855,
308.
PNRB 1979,
found
outside the S-E angle together with a second to early third century
cooking
pot.
Intramural
Richmond 1960,
distribution
was
I, 38, pi.
by
lead
mains
the
system
Trajanic
suggests
rebuilding.
that
the
The
aqueduct
environmental
may
have
RIB
640-706;
fortress
(20.23ha)
laid not
evidence
been
Sc
(Taylor
from
covered
c. AD
that
channel
established
later
71;
85-6,
probable
(Brigantes).
482
lillliillliiiililli
iiiiiiiiiiii
Abbots Langley
villa
fort
Abergavenny
A bingdon
202000
TL
08
02
329000
214000
SO
29
14
499
settlement
449900
197200
SU
villa
509750
147200
TQ 0975 4720
settlement
545700
196400
TQ
A binger Hammer
Abridge
Go bannivm
508000
457
972
964
17
Ackw or th Low
villa
444000
417000
SE
44
Ac to n Scott
villa
345800
289800
SO
458
settlement
427000
441000
SE
27
A i l s w o r t h (1)
villa
410000
298000
SP
10
98
A i l s w o r t h (2)
small town
511000
297000
TL
11
97
fort
408900
257200
SP
089
572
small town
457200
220300
SP
572
203
440400
466300
SE
404
663
Adel
Alcester
Alchester
Aldbor ou gh
Aldbourne
civitas
426250
173550
SU 2625 7355
villa
514000
201000
TL
14
+s ettlement
460500
168100
SU
605
681
511700
133000
TQ
117
330
+settlement?
Alde n ha m
A ld ermaston
Al foldean
Isvrivm Brigantvm
898
41
small town
(Upper Upham)
(minor settlement)
01
villa
601600
219900
TM
016
199
A lwinton
settlement
391600
606200
NT
916
062
Ancaster
small town
498100
343600
SK
981
436
104000
TQ
05
08
Al resford
(early fort)
A n g m e r i n g (1)
villa
505000
A n g m e r i n g (2)
villa
508000
104000
TQ
Ap ethorpe
villa
502630
294930
TL 0263
villa
429200
148450
SU 2920 4845
settlement
545220
260490
TL 4522
284100
710100
NN
841
villa
217000
207000
SN
17
07
villa
453000
086000
SZ
53
86
59
Apples ha w
A rbury Road
A rdoch
A rmoth
A rreton
fort
(falsus)
Alavna
04
04
9493
6049
101
Ash
villa
628000
159000
TR
28
As h-Cum-Ridley
villa
560800
165000
TQ
608
650
A shdon
villa
557800
243500
TL
578
435
Ashill
villa
590900
305800
TF
909
A shtead
villa
518000
157000
TQ
18
504800
204800
TL
048
893
SP
286
115
856
665
As h to n
small town
(in du s tr ia l)
058
57
Asthall
+settlement
428600
211500
At worth
villa
385600
166500
ST
Au ckley
settlement
465680
400320
SE 6568
Axminster
small town
329700
097400
SY
297
974
+settlement
419400
181000
SU
194
810
Badbury
0032
vi lla
393170
216890
SO 9317
1689
small town
524730
233570
TL 2473
3357
fort
258300
671700
NS
Bancroft
vi lla
482700
240370
SP 8270
Banwell
villa
340200
157900
ST
402
579
fort
270800
675900
NS
708
759
Ba rkby Thorpe
villa
464000
307000
SK
64
07
Barming Heath
vi lla
572000
154000
TQ
72
54
Barnack
vil la
505600
306500
TF
056
Badgeworth
Baldock
Balmuildy
Bar Hill
583
Barnsley Park
villa
408100
206150
SP 0810
B a r n w e l l (1)
vil la
507600
284800
TL
483
076
717
4037
065
0615
848
.... ................
villa
505000
285000
TL
076
848
fort
241300
669000
NS
413
690
Barrington
villa
420400
213800
SP
204
B a r t l o w (1)
villa
558000
244000
TL
58
B a r t l o w (2)
villa
558800
245000
TL
588
450
villa
451000
197800
SU
510
978
375000
165000
ST
75
65
65
B a r n w e l l (2)
Barochan
Bath
small town
Aqvae Svlis
138
44
Bathford
villa
378000
165000
ST
78
Beadlam
villa
463400
484200
SE
634
842
fort
254500
672100
NS
545
721
fortress
397100
565100
NY
971
651
B earsden
Beaufront Redhouse
578600
114000
TQ
786
140
fort
309300
549500
NY
093
495
Be ddingham
villa
545800
107400
TQ
458
074
Beddington
villa
529700
165800
TQ
297
658
Beenham
villa
460000
167000
SU
60
Benwell
fort
421500
564700
NZ
215
Bere Regis
settlement
385140
097140
SY 8514
9714
Berinsfield
settlement
458400
196800
SU
584
968
747
Beauport Park
Beckfoot
settlement
(iron mine)
67
647
356600
574700
NY
566
Bibury
villa
412200
206400
SP
122
Biddenham
villa
501000
250000
TL
01
Biglis
se ttlement
314200
169400
ST 1420
Bignor
villa
498700
114700
SU
567500
193800
TQ
675
938
421000
531300
NZ
210
313
Bewcastle
fort
Billericay
settlement
Binchester
fort
Binstead
Fanv m Cocidi
Vi n ov ia /Vinovivm
6940
147
476000
141000
SU
76
361600
566300
NY
616
663
321500
575400
NY
215
754
443400
113200
SU
434
132
449000
285000
SP
49
85
villa
367000
169000
ST
67
69
villa
479900
269000
SP
799
villa
fort
Banna
Birrens
fort
Blatobvlgivm
Biterne
small town
Bittesby
villa
Bitton
Bledlow-CumSaunderton
Bletchingley
Birdoswald
987
064
50
41
990
50
villa
532000
150000
TQ
32
settlement
471900
254000
SP
719
540
Blyborough
+settlement
491900
394400
SK
919
944
Boreham
+settlement
575150
211900
TL 7515
Blisworth
B orough Hill
B othwellhaugh
B ottesford
B oughton
1190
villa
458000
263000
SP
58
63
fort
272900
657800
NS
729
villa
489900
407000
SE
899
070
settlement
474500
265700
SP
745
657
78
578
51
Bo ughton
M onchelsea
Bourne
villa
578000
151000
TQ
vi lla
510000
320000
TF
10
Bourne/Morton
vi lla
510800
320500
TF
108
205
small town
416100
221000
SP
161
210
399300
513500
NY
993
135
82
Bourton-on-theWa ter
Bowes
fort
Lavatris
20
68
Box
vi lla
382000
168000
ST
Boxted Farm
vi lla
585160
166190
TQ 8516
6619
5998
set tlement
489590
259980
SP 8959
Bradford Dow n
v ill a
397700
104400
ST
977
Br adford-on-Avon
v ill a
381000
161000
ST
81
61
Brading
villa
459000
086000
SZ
59
86
s ettlement
348000
130340
ST 4800
Bozeat
484
044
3034
.ttwiiitifttt
Bit*
Braintree
Braishfield
Brampton
Brandon
B ra ntingham
small town
S75500
223100
TL
755
villa
438000
126000
SU
38
small town
622410
223780
TG 2241
340100
272400
SO
401
724
493200
428800
SE
932
288
539140
224330
TL 3914
2433
300000
229000
SO
00
29
fort
(camp)
v i l l a (falsus)
Braughing
small town
Br econ Gaer
fortress(A)
Cicvcivm
231
26
2378
Brewood
villa
389000
310000
SJ
89
Brislington
villa
362000
171000
ST
620
B r i s t o l (1)
villa
355000
178000
ST
55
78
villa
353000
177000
ST
53
77
fortlet
277300
318800
SH
773
188
villa
474700
271900
SP
747
719
B r i s t o l (2)
Brithdir
Brixworth
10
710
settlement
525800
135300
TQ
258
353
Bromham(l)
villa
399800
164200
ST
998
642
B r o m h a m (2)
+settlement?
502700
252300
TL
027
523
902
Broadfields
Brough-byBainbridge
Brough-onH u m b e r (1)
Brough-onH u m b e r (2)
Brough-on-Noe
Bryncross
Buchley
Bucknowle Farm
Buckton
fort
Virosidvm
393700
490200
SD
937
fort
Petvaria
493000
426000
SE
93
494100
426750
SE 9410
settlement
2675
418100
382700
SK
181
827
settlement
221400
331600
SH
214
316
fort
Navio
fort
259500
672000
NS
595
720
villa
395400
141800
SY
954
418
fortress(A)
339000
273000
SO
39
Bunny
settlement
457950
328900
SK 5795
Burgh
villa
622400
252300
TM
224
Burham
villa
572000
161000
TQ
72
361730
476150
SD 6173
Burrow-in-Lonsdale
26
fort
Calacvm
73
2890
523
61
7615
Burton
settlement
496100
374700
SK
961
747
Buxton
small town
405700
373400
SK
057
734
Cadder
fort
261700
672600
NS
617
726
Caerhun
fortress(A)
Canovivm
277500
370400
SH
775
704
Caerleon
f o r t r e s s (L)
Isca
332300
190900
ST
323
909
Caernarvon
fortress(A)
Segontivm
248200
362500
SH
482
625
Caersws
fortress(A)
303000
291800
SO
030
918
346700
198700
ST
467
651700
312310
TG 5170
623100
303400
TG
231
398000
170000
ST
98
Caerwent
Caister-on-Sea
C ai stor-by-Norwich
Caine
Cambridge
Camelon
Camerton
Camphill
Canterbury
civitas
Venta Silvrvm
fort
civitas
Venta Icenorvm
villa
small town
Duroliponte
fort
small
town
villa
civitas
Dvrovernvm
Cantiacor
987
1231
034
70
544400
259300
TL
444
593
286500
680600
NS
865
806
368000
156000
ST
68
56
473000
258000
SP
73
58
614740
157680
TR 1474
5768
Cantley
settlement
461100
401480
SE 6110
Cappuck
fortlet
369500
621300
NT
695
213
fortress(A)
318100
176600
ST
181
766
villa
448000
088000
SZ
48
339600
556100
NY
396
561
326000
546400
NY
260
464
Moridvnvm
241000
220000
SN
41
320000
717000
NO
20
385900
571200
NY
859
Cardiff Castle
Carisbrooke
Carlisle
Ca rlisle-Old
Ca rmarthen
Carp o w
Ca rrawburgh
civitas
fort
civitas
Lvgvvallvm
Maglona
fortress(A)
fort
Brocolitia
485
0148
88
20
17
712
IIIIIIIIIP III
8it+
C arsin gt o n
settlement
425000
352400
SK
250
524
Carvoran
fort
366500
565700
NY
665
657
Castell Collen
fort
305500
262800
SO
055
628
Castle Cary
fort
279900
678300
NS
799
783
Castle Greg
fortlet
305000
659200
NT
050
592
Margidvnvm
470000
341500
SK
700
415
Carvoran Magnis
Castle H i l l (1)
small town
Castle Hill(2)
fortlet
Margidvnvm
470010
341510
SK 7001
4151
fort
376100
567800
NY
761
678
765
255
Castle N ick
428500
476500
SE
285
Castleford
fort
Lagentivm
442500
425500
SE
425
Castle sh a w
fort
Rigodvnvm
399880
409650
SD 9988
Castlesteads
fort
xellodunum
351200
563500
NY
Castle-Dykes
vil la
512
C a s t o r (1)
villa
511550
297780
TL 1155
C a s t o r (2)
+settlement
512400
298400
TL
350700
126600
424100
497200
Catsgore
Catterick
villa
fort
Cataractonivm
0965
635
9778
124
984
ST
507
266
SE
241
972
593
Cattybrook
settlement
359300
197200
ST
Caversham
settlement
472860
174090
SU 7286
834
7409
villa
567700
173000
TQ
677
Chapel Hill
fort
406000
835000
NK
06
Charmy Down
villa
377600
169100
ST 7760
C h e d w o r t h (1)
villa
405300
213500
SP
053
135
C h e d w o r t h (2)
villa
405200
213400
SP
052
134
Chells Manor
settlement
526600
225900
TL
266
259
Chelmsford
small town
570920
206210
TL 7092
Cheshunt
settlement
535000
202000
TL
35
340400
366200
SJ
404
662
513
Chalk
Chester
fortress(L)
Caesaromagvs
Deva
730
35
6910
0621
02
Chester- le-Street
fort
Concangis
427600
551300
NZ
276
Ch esterholm
fort
Vindolanda
376800
566300
NY
768
663
Chesters
fort
Cilvrnvm
391200
570100
NY
912
701
11
C h e s t e r t o n (1)
small town
511000
296000
TL
C h e s t e r t o n (2)
small town
512000
296000
TL
12
C h e s t e r t o n (3)
fort
383200
348900
SJ
832
489
486500
104600
SU
865
046
545000
196000
TQ
45
Chichester
Chigwell
civitas
+settlement
Noviomagvs
Regnorvm
Dvrolitvm
C h i l g r o v e (1)
villa
483000
112000
SU
83
C h i l g r o v e (2)
vi 11a
484100
113600
SU
841
96
96
96
12
136
Chipping Warden
villa
451000
248200
SP
510
Chiseldon
villa
419000
180000
SU
19
Churchill
settlement
428500
225500
SP
285
255
Churchill Hospital
settlement
454600
205700
SP
546
057
402500
201700
SP
025
017
villa
401600
202200
SP
016
022
settlement
419000
199600
SU
190
996
C i r e n c e s t e r (1)
C i r e n c e s t e r (2)
Clay don Pike
civitas
(Grounds Farm)
Corinivm Dobvnorvm
482
80
villa
530000
113000
TQ
30
Cliffe House
settlement
473000
465500
SE
730
Cnut's Dyke
settlement
526000
293000
TL
260
C obham
villa
508000
159000
TQ
08
59
Co b ha m Park
villa
567000
168000
TQ
67
68
Combretovivm
611690
252810
TM 1169
Camvlodvnvm
599000
225000
TL
99
villa
492000
252000
SP
92
+settlement
310090
166490
ST 1009
355000
203000
SO
Clayton
C oddenham
Colchester
Cold Brayfield
Cold Knap Point
Coldharbour
small
town
colonia
settlement
(falsus)
486
55
13
655
930
5281
25
52
6649
03
Colerne
Coleshill
Collin gh a m
C olliton Park
Co lsterworth
villa
381800
171500
ST
818
+settlement
419500
290500
SP
195
villa
440340
446150
SE 4034
villa
368950
090960
SY 6895
9096
settlement
491800
324000
SK
918
240
768
622
715
905
4615
Combe Down
villa
376800
162200
ST
Comberton
vi lla
538450
254890
TL 3845
5489
Com biey
villa
453000
087000
SZ
53
87
797
C o m p t o n (1)
villa
479700
112400
SU
C o m p t o n (2)
vi lla
495000
147000
SU
95
C o m p t o n (3)
settlement
447600
111400
SU
476
114
villa
344500
162700
ST
445
627
Corbridge
small town
398400
565300
NY
984
653
Co rhambry
villa
511700
207900
TL
117
079
se ttlement?
654000
297000
TM
54
SP 7947
4212
Co ngresbury
Co rton
124
47
97
Co sgrove
villa
479470
242120
Co tterstock
villa
503260
291070
TL 0326
9107
Covehithe
villa?
552200
282000
TM
522
820
C ow Roast
settlement
595800
282000
SP
958
Cowbridge
small town
299410
174750
SS 9941
7475
Cox Green
villa
487000
179800
SU
870
798
fort
318900
676900
NT
189
769
fortlet
273400
676500
NS
734
765
Dalginross
fort
277300
721000
NN
773
210
D alswinton
fort
293600
583900
NX
936
839
villa
556400
170600
TQ
564
706
villa
458800
263200
SP
588
632
settlement
368200
211900
SO
682
119
Cramond
Croy Hill
D arenth
Daventry
D ea n Hall
Denton
Derby
villa
fort
Derventio
Dersingham
settlement
102
487400
331300
SK
874
313
435200
332800
SK
352
373
567600
332800
TF
676
328
Desborough
settlement
479420
282590
SP 7942
8259
Desford
settlement
447800
303500
SK
035
Dewlish
villa
376800
097200
SY
768
972
Dicket Mead
villa
523400
215900
TL
234
159
settlement
520200
265100
TL
202
651
villa
438700
200900
SP
387
009
266400
234400
SN
664
344
Diddington
Ditchley
Dolaucothi
small town
Doncaster
settlement
Danvm
457800
403200
SE
578
032
civitas
Dvrnovaria
369400
090500
SY
694
905
civi tas
(S o m e r le ig hC o ur t)
D o r c h e s t e r (1)
D o r c h e s t e r (2)
Dorn
Dover
Downshay Wood
Downton
D r a y t o n (1)
D r a y t o n (2)
Lventinvm
478
small town
369090
090390
SY 6909
420700
233900
SP
207
31
9039
339
41
631000
141000
TR
settlement
397820
079010
SY 9782
7901
villa
418000
121000
SU
21
fort
Portvs Dvbris
18
483000
293000
SP
83
villa
481700
291800
SP
817
918
639
villa
93
small town
389800
263900
SO
898
Dryhill
villa
393000
216000
SO
93
Ducklington
villa
436300
207200
SP
363
Duncton
villa
496000
117000
SU
96
Dunsby
settlement
512200
327100
TF
122
Dunstable
small town
501000
221000
TL
01
Duntocher
fortlet
249400
672700
NS
494
727
+settlement?
428900
541600
NZ
289
416
Droitwich
D urham
Dvrocobrivis
487
16
072
17
271
21
i'ft&atiapii"
Duston
small town
473500
260200
SP
735
Earith
settlement
539100
275900
TL
391
villa
354000
112000
ST
54
settlement
449000
180000
SU
49
80
villa
561000
099500
TV
61
99
settlement
319400
640100
NT
194
villa?
628200
258700
TM
282
587
settlement
477900
326600
SK
779
266
villa
357170
363410
SJ 5717
East Coker
East Ilsley
Eastbourne
Easter Happrew
Easton
Eaton
Eaton-by-Tarporley
759
12
401
6341
Ebchester
fort
410400
555400
NZ
Ebrington
villa
419010
239900
SP 1901
Eccles
villa
572200
160500
TQ
722
605
Eckington
villa
392000
241600
SO
920
416
Vindomora
104
602
554
3990
fort
332100
667300
NT
321
673
Elsted
villa
481800
115300
SU
818
153
Ely
villa
314500
176700
ST
145
767
settlement
488000
250000
SP
88
943
Elginhaugh
Emberton
Empingham
Enfield
Engleton
Epperstone
Ewell
Ewhurst
villa
494300
307700
SK
settlement
534410
199990
TQ 3441
villa
389500
310300
SJ
50
077
9999
895
103
496
villa
467200
349600
SK
672
small town
522500
162500
TQ
225
625
villa
509000
140500
TQ
090
405
291920
092800
SX 9192
Isca Dvmnoniorvm
9280
Exeter
fortress
Exning
villa
561200
267600
TL
612
676
fort
288200
679900
NS
882
799
Farley Hungerford
villa
304000
140000
ST
04
Farmington
villa
413230
215850
SP 1323
settlement
445000
206000
SP
45
villa
496000
182000
SU
96
82
Farnham(l)
villa ?
636000
260000
TM
36
60
F a r n h a m (2)
villa
484600
146600
SU
846
466
F a r n h a m (3)
settlement
395300
115300
ST
953
153
Farningham
villa
555500
167500
TQ
555
675
settlement
351000
387000
SJ
510
870
Fawler
villa
437200
216900
SP
372
169
Feltwell
villa
569990
292070
TL 6999
9207
Fendoch
fort
293000
728000
NN
93
28
Ffrith
fort
328400
354800
SJ
284
Falkirk
Farmoor
Fa rnham Royal
Farnworth
40
1585
06
548
Fi fehead
villa
377280
111210
ST 7728
Fillingham
villa
495400
386400
SK
954
864
Finchingfield
villa
568500
238500
TL
685
385
Findon
villa
511900
109700
TQ
119
097
fort
604700
219400
TM
047
194
villa
483700
104400
SU
837
044
Flamsteed
settlement
538900
177300
TQ
389
773
Fletton
197
Fingringhoe
Fishbourne
1121
settlement
519700
296500
TL
Folkstone
villa
624260
136950
TR 2426
3695
Fordcroft
villa
546780
167580
TQ 4678
6758
Foscott
villa
472000
235000
SP
72
Foxcote
villa
472200
235300
SP
722
361000
095000
SY
61
Frampton
Frilford
Friskney
Frocester Court
villa
(falsus)
965
35
353
95
97
town
442000
197000
SU
42
settlement?
546300
345500
TF
463
455
villa
378500
202900
SO
785
029
small
488
.
Funtington
Gadebridge Park
G arden Hill
Garton Slack
Gatcombe
Gayton
G ayton Thorpe
Gelligaer
Gestingthorpe
Hotftan Stott*
villa
f Skat la#*
482000
108000
SU
82
08
villa
505500
207508
TL
055
075
+settlement
544400
131900
TQ
444
319
settlement
495300
460100
SE
953
601
villa
352600
169800
ST
526
698
471270
254880
SP 7127
5488
573500
318000
TF
180
settlement
villa
735
fort
313000
196000
ST
13
villa
588200
238700
TL
882
96
387
Glenlochar
fort
?Levcovia
273500
564500
NX
735
645
Gloucester
colonia
Glevvm
383600
218900
SO
836
189
477600
327000
SK
776
270
Dvrovigvtvm
524700
270500
TL
247
705
Goadby Marwood
settlement
G o d m a n c h e s t e r (1)
small town
G o d m a n c h e s t e r (2)
villa
525000
271000
TL
25
Gorhambury
villa
511700
207900
TL
117
079
Goring
villa
510600
103900
TQ
106
039
71
settlement
342500
352200
SJ
425
522
Grandford
settlement
539300
299600
TL
393
996
villa
482000
206000
SP
82
+settlement
336100
191500
ST
361
915
small town
500100
308800
TF
001
088
Bovivm
06
villa
500400
309300
TF
004
093
Great
C h e s t e r f o r d (1)
Great
C h e s t e r f o r d (2)
Great Chesters
fortress(A)
550300
243000
TL
503
430
villa
550000
243000
TL
50
370400
566800
NY
704
668
Great Dunmow
small town
562600
221900
TL
626
219
413
Great Casterton(2)
fort
Aesica
43
Great Linford
villa
484400
241300
SP
844
Gr eat Staughton
villa
513600
263500
TL
136
Gr eat Tew
villa
440000
210000
SP
40
Great Totham
villa
587000
210000
TL
87
Great Witcombe
villa
389900
214400
SO
899
144
Greetwell Fields
villa
499600
371600
SK
996
716
fort
408400
513100
NZ
084
villa
423000
127000
SU
23
Greta Bridge
Grimstead
635
27
10
131
27
villa
571800
321600
TF
718
216
settlement
473700
392000
SK
737
920
settlement
511700
325000
TF
117
250
villa
502000
336000
TF
02
town
631200
256700
TM
312
567
+settlement
514000
293200
TL
140
932
villa
553300
207600
TL
533
076
Hales
vi lla
372200
333800
SJ
722
338
Ha lstock
villa
353300
107600
ST
533
076
399700
568400
NY
997
vi lla
478000
185000
SU
78
Grimston
G rinley on the
Hill
Hacconby
Haceby
Hacheston
Haddon
Hadstock
Halton Chesters
Hambleden
Ham ilton
Ha mpstead Norrey's
Ha nh am Abbots
small
fort
vill a
464000
307000
SK
64
07
464000
176000
SU
52
76
villa
364000
170000
ST
64
70
503000
117000
TQ
03
17
476000
257000
SP
76
57
321000
501000
NY
21
01
small
Hardingstone
settlement
Ha rlow
Harpole
684
85
settlement
Hardham
Ha rdknott
Onnvm
36
town
fort
M edi obogdvm
small town
549500
211200
TL
495
112
villa
468400
259900
SP
684
599
489
..
....
Harpsden
villa
475000
180000
SU
75
Hartfield
villa
545600
138300
TQ
456
80
383
Hartlip
villa
582000
164000
TQ
82
H a v a n t (1)
villa
471000
105000
SU
71
05
H a v a n t (2)
villa
469000
107000
SU
69
07
Hayes
villa
540000
166000
TQ
40
fort
481800
445600
SE
818
456
villa
493600
229200
SP
936
292
Hayton
Heath and Reach
64
66
Hemsworth
vil la
396000
106000
ST
96
Heybridge
small town
584700
208200
TL
847
H ibaldstow
small town
495900
402800
SE
959
028
High Cross
small town
447300
288600
SP
473
886
+settlement
343000
128000
ST
43
settlement
368900
383100
SJ
689
831
383300
598600
NY
833
986
High Ham
High Legh
High Rochester
fort
H igham Ferrers
+settlement
496460
268810
SP 9646
villa
379000
158000
ST
79
Hinton
Charterhouse
Holcombe
Holditch
Holme House
Bremenivm
331500
092800
SY
315
102000
ST
34
villa
422100
515200
NZ
221
152
545
340500
354500
SJ
405
525700
369500
TF
257
Horningsea
settlement
549830
263480
TL 4983
fort
Vercovicivm
villa
Hucclecote
villa
Hunsbury
villa
Huntingdon
58
334000
fortlet
Hovingham Park
6881
villa
settlement
Housesteads
28
small town
Horncastle
Holt
06
082
villa
928
02
695
6348
379000
568800
NY
790
688
466300
475700
SE
663
757
387650
217350
SO 8765
473700
258200
SP
737
582
TL
240
716
565
524000
271600
1735
175
Huntsham
villa
356500
217500
SO
Hurcot
villa
351050
129780
ST 5105
Ickham
settlement
523100
159100
TR
231
591
villa
549400
243500
TL
494
435
settlement
578200
272200
TL
782
722
villa
380000
157500
ST
800
351000
123000
ST
51
fortress
312500
739600
NO
125
396
fort
334600
662200
NT
346
722
Ickleton
Icklingham
Iford
Ilchester
Inchtuthill
Inveresk
Irchester
Islip
Ivy Chimneys
Ixworth
small
small
town
Lindinis
town
villa
2978
575
23
491700
266700
SP
917
667
497000
278200
SP
970
782
136
settlement
581100
136000
TL
811
vil la
593000
269000
TL
93
69
82
Jo r do n Hill
settlement
369000
082000
SY
69
Kelvedon
small town
586200
219000
TL
862
Kempsford
settlement
417200
199650
SU 1720
Kempston
settlement
501200
247800
TL
Kenchester
small
344000
242000
SO
44
541400
163200
TQ
414
487220
280480
SP 8722
366800
161700
ST
668
617
534
776
Keston
Kettering
Keynsham
town
Magna
villa
small
town
villa
012
villa
353400
177600
ST
Kingscote
villa
380650
196080
ST 8065
fortlet
297700
680300
NS
977
Kintbury
villa
439000
167000
SU
39
K ir k Sink
villa
393900
453600
SD
939
Kinneil
490
190
9965
478
42
632
8048
9608
803
67
536
illllliiiliiiiiM
$ |g |g !
iilillll
Lake Far m
L ambourn
fortress(A)
399700
099000
SY
997
v i l l a (falsus)
434000
181000
SU
34
990
81
Lancaster
fort
Ca lacvm ?
347400
461900
SD
474
Lanchester
fort
Longovicivm
415000
546000
NZ
15
46
563000
267000
TL
63
67
Landwade
+settlement
619
Langton
villa
481600
467500
SE
816
675
Latimer
villa
499800
198600
SU
998
986
417
Lea Cross
villa
341700
308500
SJ
Leaden Well
villa
416210
220980
SP 1621
fort
481400
504200
NZ
814
villa
421000
121000
SU
21
394800
218300
SO
948
183
457900
304200
SK
579
042
043
Lease Rigg
Le chlade
L eckha mp t on Hill
Leicester
fort
civitas
Ratae Co ritanorvm
085
2098
042
60
L e i c e s t e r (1)
villa
457500
304300
SK
575
L e i c e s t e r (2)
civitas
458600
304500
SK
586
045
Leintwardine
+settlement
340600
274000
SO
406
740
538500
186100
TQ
385
861
497400
372000
SK
974
720
Le yton
Lincoln
Bravonivm
set tlement
colonia
Linley
set tlement
334700
292700
SO
347
927
Li nton
villa
557100
246200
TL
571
462
Li tlington
villa
531300
242500
TL
313
425
Li ttle Dunmow
vil la
566400
221200
TL
664
212
Li ttle
P onton/Stroxton
Li ttle Wal th am
vil la
492700
333400
SK
927
334
set tlement
570000
212000
TL
70
353
376
765
12
435300
337600
SK
Littlecote Park
villa
430100
176500
SU
301
Littleton
vil la
349000
130000
ST
49
vi lla?
264000
256000
SN
64
56
vi lla
225000
212000
SN
25
12
Llandough
vi lla
316850
173150
ST 1685
Ll anfrynach
vi lla
306000
225000
SO
06
264400
256200
SN
644
562
169900
SS
958
699
Littlechester
Llanddewi Brefi
Lla nddowror
Ll anio
Llantwit
Locking
Loddington
small
town
fort
De rventio
Bremia
vi lla
295800
villa
336000
160000
ST
36
settlement
481500
278000
SP
815
Lo dd on
v ill a
Lo ndon
munici pi u m
(civium)
+settlement
Londinivm
30
7315
25
60
780
636000
298000
TM
36
98
532000
181000
TQ
32
81
586100
245100
TL
861
settlement
455000
195000
SU
55
95
v ill a
434000
136000
SU
34
36
fortress(A)
515700
297700
TL
157
977
Loughor
fort
256200
197900
SS
562
979
fort
361110
500910
NY 6111
451
0091
v illa
343000
129000
ST
43
settlement
417400
222600
SP
174
Lu fton
v illa
351000
117000
ST
51
Lu llingstone
v ill a
553000
165100
TQ
530
651
fort
434700
274600
SP
347
746
v illa
361600
202500
SO
616
40
34
L ow Ham
Lower Slaughter
Lunt
Ly dney
Lym inge
Lympne
L ynch Farm
Lyne
M aiden h ea d
+settlement
fort
Portvs Lemanis
se ttlement
fort
Carbantoritvm ?
v illa
491
29
226
17
025
616000
140000
TR
16
611000
134000
TR
11
514900
399700
TF 1490
9970
318800
640500
NT
188
405
48 8000
181000
SU
88
81
M aidstone
villa
Mait o n
156300
""
TQ
756
..
563
471600
SE
792
716
fortress(A?)
Derventio
479200
fort
Mandve ssedvm
431840
296490
SP 3184
9649
432600
296700
SP
967
M a n c e t t e r (1)
M a n c e t t e r (2)
575600
small
M anchester
town
fort
M an sf ie l d W ood
House
Market Overton
Mamvcivm
326
383450
397780
SJ 8345
9778
villa
452300
364700
SK
647
523
villa
488000
316000
SK
88
Marshf ie ld
settlement
379800
176100
ST
798
M arsworth
settlement
16
761
493000
214000
SP
93
Maryport
fort
Alavna
303800
537300
NY
038
M elandra Castle
fort
Ardotalia
489000
395000
SK
890
950
573100
295800
TL
731
958
M eth w ol d
villa
M idd l eh am
villa ?
Mildenhall
small town
More
fort
Morton Bourne
Mountsorrel
413000
487000
SE
13
87
421000
169000
SU
21
69
334000
292000
SO
34
Cab rosentvm
298400
521100
NX
984
211
513900
323300
TF
139
233
settlement
457900
314900
SK
579
149
villa
567300
180300
TQ
673
803
fort
291600
679400
NS
916
794
settlement
407000
012000
SZ
07
settlement?
364800
352300
SJ
648
387
Mumrills
Nant wi c h
Ne a th
fort
(falsus)
small town
N e at ha m
N et he r W ild Farm
N ettleton
small town
Newcastle
fort
206600
TL
197800
SS 7476
9780
V i n d o m i (s)
474200
141200
SU
742
412
514200
201200
TL
142
414000
148000
SU
14
359600
571700
NY
596
382000
176000
ST
82
425000
563900
NZ
250
Castra
Ex ploratorvm
Pons Aelii
N ewhaven
villa
544000
101000
TQ
44
N ew nh am
villa
507500
249200
TL
075
Ne wp ort
villa
450000
088000
SZ
50
NT
571
N or fo lk Street
012
48
717
76
639
01
492
88
344
357100
634400
settlement
395800
082600
SY
958
826
vil la
457500
304300
SK
575
043
154
fort
N ewstead
N orden
066
538700
fort
N etherby
523
274760
villa
v ill a
N ether a vo n
12 ?
N idvm
se ttlement
N azeingbury
92
settlement?
Mucking
Munt he m Court
373
Cvnetio
villa
Moresby
14
Trimontivm
N or t h Leigh
villa
439700
215400
SP
397
N or t h Mundham
villa
487000
103000
SU
87
N or t h Stainley
villa
428500
476500
SE
285
N or th Wraxall
v illa
383000
176000
ST
83
76
Northa ll e rt on
settlement?
436000
494000
SE
36
94
N or th ampton
s ettlement
474480
258830
SP 7448
N or th church
v illa
497000
209000
SP
97
No rthfleet
v illa
561600
174100
TQ
616
N orthmoor
settlement
Nort o n
settlement
Bannaventa
442000
202000
SP
42
461200
264500
SP
612
360000
SK
85
03
765
5883
09
741
02
645
60
N o r t on Disney
v illa
485000
N o r t on St.
v illa
379000
158000
ST
79
58
villa
374000
146000
ST
74
46
Nurs li n g
settlement
436600
116150
SU 3660
O akham
settlement
486700
309500
SK
Oa kley
villa
661500
127850
TM 6150
+s ettlement
464200
153500
SU
N unney
Oakridge
Odell
Phillip
495600
settlement
492
256800
SP
867
1615
095
2785
642
535
956
568
Odih a m
villa
473000
152000
SU
73
Ol d D u r ha m
villa
427500
542100
NZ
275
fort
246000
672000
NS
46
fort
349400
538400
NY
494
settlement?
528400
309900
TF
284
099292
094
6583
TQ 4540
Ol d K il p at ri ck
Old Penrith
Old South Eau
Orping to n
52
421
72
384
villa
545400
165830
+settlement
517600
295600
TL
176
O tford
villa
553000
259000
TQ
53
O xford
settlement
454650
205750
SP 5465
O x f o r d (2)
settlement
454430
204080
SP 5443
0408
Pagans Hill
settlement
355600
162600
ST
626
Pamphill
v illa
397000
162600
ST
97
Pau lton
+settlement?
365000
156000
ST
65
fortress(A)
248100
344900
SH
481
Orton Longueville
Pen Llystyn
556
956
59
0575
04
56
449
Pe n-Y-Darren
fort
305300
206800
SO
053
068
Pennal
fort
270500
300100
SH
705
001
Pentre Ffwrndan
F arm
P e t e r b o r o u g h (1)
+settlement
325500
372300
SJ
255
723
settlement
521500
298900
TL
215
989
P e t e r b o r o u g h (2)
set tlement
520600
298900
TL
206
989
vi lla
479650
254000
SP 7965
5400
157
Pi ddington
fort
421500
515700
NZ
215
settlement
524000
325000
TF
24
Pitmeads
villa
437000
116000
SU
37
16
Pitney
vi lla
345000
130000
ST
45
30
Pitsford
set tlement
474200
267200
SP
742
672
Plas Coch
se ttlement
332600
351700
SJ
326
517
villa
561400
153700
TQ
614
537
se ttlement
399800
099000
SY
998
990
046
Piercebridge
Pinchbeck
Plaxtol
Poole
25
fort
462400
104600
SU
624
vil la
346000
176000
ST
46
set tlement?
368650
072010
SY 6865
Prestatyn
set tlement
306200
381700
SJ
062
817
Puc keridge
settlement
538800
223700
TL
388
237
Pulborough
+settlement
507000
118000
TQ
07
Portchester
Portishead
Po rtland Island
76
7201
18
265600
240600
SN
656
set tlement?
364000
237000
SO
64
set tlement
352220
459340
SD 5222
Q uinton
se ttlement
477500
253500
SP
775
535
R adfield
set tlement
593900
162800
TQ
939
628
fortlet
325200
599100
NY
252
991
set tlement
554500
182000
TQ
545
820
088
Pumsaint
Putley
Quernmore
Ra eburnfoot
R ai nh a m
R avenglass
fort
495800
SD
222350
TL 7127
2235
557260
265180
TL 5726
6518
622700
169400
TR
227
694
s ettlement
495800
270500
SP
958
705
villa
526000
150000
TQ
26
50
TR
32
60
733
R each
villa
Reigate
Rich b or ou gh
Ri dgewell
Rippin ga l e
Ri venhall
Roch es t er
R ock
958
308800
set tlement
R edlands Farm
37
5934
571270
fort
Ra yne
Re culver
Lventinvm
406
fort
Clanoventa
Re gvlbivm
632000
160000
villa
573300
240200
TL
s ett lement?
512600
328500
v ill a
582000
217000
TF
126 285 275
150
17
TL
82
574200
168500
TQ
742
685
442400
084100
SZ
424
841
fort
small
town
Rvtvriae
Dvrobrivae
villa
493
402
Rotfcaji Naans
a it *
..
Ro ckbourne
villa
412000
117000
R o c k in g ha m
settlement
486660
291280
SP 8666
villa
592000
161000
TQ
92
438700
466500
SE
387
551000
188000
TQ
51
Rode rs h am
Roecliffe
fort
120
Ro mford
+settlement
Rothley
villa
456900
312200
SK
569
settlement
438000
428000
SE
38
Rothwell Haigh
R ough Castle
R ousham
Rowlands Castle
Rudston
Rushton
R uthin
R yt on on Dunsmore
Salford Priors
Sambourne
Sandringham
Sandwich
Sandy Lane
Sapcote
Scampton
Scawby
(cf Chigwell)
SU
fort
297200
680100
NS
972
villa
446000
224000
SP
46
+settlement?
473000
109000
SU
73
villa
508800
466700
TA
088
villa
357000
363000
SJ
57
settlement
310300
358200
SJ
103
settlement
437000
272000
SP
37
SP 0795
170
9128
61
665
88
122
28
801
24
09
667
63
582
72
+settlement
407950
251820
+settlement
408000
261000
SP
08
61
+settlement?
569000
328000
TF
69
28
631900
157300
TR
319
396000
168000
ST
96
68
villa
449000
293000
SP
49
93
villa
495400
378600
SK
954
786
+s ettlement
496900
404600
SE
969
046
786
villa
small town
Verlv cio
5182
573
Scole
small town
614700
278600
TM
147
Scunthorpe
settlement
490500
413800
SE
905
138
fortlet
281200
679200
NS
812
792
909
Seabegs Wood
Seaton
Selsey
Sewingshields
vi lla
vi lla
fortlet
?
(MC34)
323800
090900
SY
238
485000
094000
SZ
85
380800
570300
NY
808
94
703
villa
435000
212000
SP
35
S hawell/Churchover
villa
453300
279500
SP
533
Shepreth
vil la
539000
248000
TL
39
se ttlement
363000
142400
ST
630
vil la
348400
091900
SY
484
919
se ttlement
454300
379800
SK
543
798
21
Shakenoak
Shepton Mallet
Shipton Gorge
Shireoaks
12 ?
795
48
424
vil la
521000
105000
TQ
se ttlement
180500
047560
SW 8050
4756
settlement?
507000
289000
TL
07
89 ?
se ttlement
402800
198800
SU
028
vil la
485000
097000
SZ
85
Calleva At rebatum
463800
162800
SU
638
628
396300
118100
ST
963
181
Camvlodvnvm
408400
417600
SE
084
176
507700
345800
TF
077
458
570800
162100
TQ
708
621
402000
194500
SU
020
945
436600
567900
NZ
366
679
villa
492000
319000
SK
92
fort
532500
179900
TQ 3250
532000
180000
TQ
32
Southwell
settlement
(mansio)
v illa
470300
353700
SK
703
So uthwick
villa
524000
105000
TQ
24
Sparsholt
villa
441500
130100
SU
415
settlement
540000
366000
TF
40
v illa
404500
225700
SP
045
S hor eham
Shortlanesend
Siberton
Siddington
S idlesham
Silchester
Si xpenny Handley
Slack
Sl eaford
Sn odland
Somerford Keynes
Sou th Shields
Sou th Witham
Southwark
So uthwark Street
Spilsby
Spoonley Wood
civitas
settlement
fort
small
town
villa
+se ttlement
fort
Arbeia
494
05
988
97
19
7990
80
537
05
301
66
257
..
Sp r in gh ea d
small town
Stephen
S taden
Staines
Vagniasis
'''"LObafciijiii'''1'"
561700
172500
TQ
617
s ettlement
547000
167000
TQ
47
67
v illa
514000
203000
TL
14
03
+settlement
471220
377240
SK 7122
503700
171500
TQ
small town
Stainf ie ld
Ad Pontes
(Pontibus)
037
725
7724?
715
villa
507000
324000
TF
07
St ainley Nor th
settlement
415000
495000
SE
15
Stam fo r d Bridge
settlement
470000
454000
SE
700
540
v illa
461000
396100
SK
610
961
Stancil
24
95 ?
Standon
v illa
537000
217000
TL
37
17
S tanford in the
V ale
S tanton F itzwarren
v ill a
434000
194000
SU
34
94
417400
190100
SU
174
901
427
S tanton Low
villa
villa
483900
242700
SP
839
v illa
457000
210000
SP
57
S tantonbury
villa
484000
243000
SP
84
S tanwick
vi lla
497100
271100
SP
971
S tanton St.
John
Stanwix
S tebbing
Stephen Mallet
Stibbington
10
43
711
340300
556800
NY
403
568
villa
567100
223200
TL
671
232
fort
Vx el od vn v m
v illa ?
340000
150000
ST
40
settlement
508500
298600
TL
085
50 ?
986
villa
389000
380000
SJ
89
80
settle ment?
512000
158000
TQ
12
58
settlement
361600
180000
ST
616
800
Stoke Orchard
settlement
391300
228900
SO
913
289
492800
327900
SK
928
279
493000
328700
SK
930
287
SK
918
276
Stockport
Stoke D'Abernon
Sto ke Rochford(l)
vi lla
Stoke Rochford(2)
vi lla
St oke Rochford(3)
villa
491800
327600
Stone
Stonea Grange
S toneham Aspal
(North Stoke)
set tlement
478000
212000
SP
78
+settlement?
544900
293700
TL
449
12
vil la
613000
259000
TM
13
59
17
937
+settlement
440000
217000
SP
40
S torri n gt on /P a rh am
villa
506000
117000
TQ
06
17
Stowe
villa
330000
273000
SO
30
73
fort
289800
718000
NN
898
Stonesfield
Strathgeath
389000
310000
SJ
89
S t r o u d (1)
vil la
382000
205000
SO
82
S t r o u d (2)
vil la
472500
123600
SU
725
Stretton Bridge
Sutton Courtenay
S utton Veny
S wa ff ha m Bulbeck
Sw alcliffe Lea
Sw anwick
Sw in don
T allin gt o n
Tarrant Hinton
T em p ie bo ro u gh
T es to n
set tlement
Pennocrvcivrn
449700
159600
SU
497
vil la
390000
143000
ST
90
se ttlement
352190
167250
ST 5219
v illa
180
10
05
236
596
43
6725
se ttlement
438000
238000
SP
38
38
se ttlement?
451000
109000
SU
51
09
v illa
416000
182000
SU
16
82
settlement
509000
308000
TF
09
08
118
392600
111800
ST
926
fort
441000
391000
SK
41
91
v illa
569000
153000
TQ
69
53
32
v illa
Anicetis
T ewkes b ur y
settlement
389000
232000
SO
89
T ha t c h a m
settlement
350000
267000
SO
50
Theale
settlement
467300
170200
SU 6730
7020
415
67
T hen f or d
v il la
452500
241500
SP
525
T het f or d
v il la
586600
284900
TL
866
849
Th i st le to n
v il la
490800
317000
SK
908
170
Th u rg ar to n
v illa
467400
349500
SK
674
495
495
<ti ifama
Thurlby
T hu rn ha m
Tiddin gt on
Ti lston
mm
... .................
settlement?
v illa
small town
+settlement
Bovivm
510520
316110
TF 1052
1611
380000
157000
ST
800
570
421900
255600
SP
219
556
345300
351700
SJ
453
517
Ti ngwick
villa
465800
293000
SP
658
Titsey
villa
540000
154000
TQ
40
Tixover
villa
498200
301900
SK
982
019
To ckington Park
villa
361000
186400
ST
610
864
fort
270600
338600
SH
706
villa
498000
220000
SP
98
Tomen - Y- Mur
Totternhoe
386
20
Towcester
fort
469000
247900
SP
690
479
Traws-Coed
fort
267100
272700
SN
671
727
villa
255700
340100
SH
557
401
small town
(burgi)
fortlet (MC 50)
453500
279500
SP
535
795
360800
566000
NY
608
660
villa
448000
124000
SU
48
settlement?
495250
278080
SP 9525
7808
Tremadoc
Tripontivm
Turf Wall Mile
Castle
Twyford
Twywell
Upham
Lactod orvm
930
54
24
villa ?
354000
122000
ST
54
22
villa
479000
113000
SU
79
13
Upminster
settlement?
557000
184500
TQ
570
845
Usk
fortress(L)
337900
200700
SO
379
007
Verula mi um
mu nicipium
514030
206160
TL 1403
0616
villa
502310
226040
TL 0231
2604
Wainfleet All
Saints
Walesby
settlement?
549800
358900
TF
498
589
514600
392700
TF
146
927
Wall
f o r t r es s( A? )
409600
306500
SK
096
353990
087300
SY 5399
8730
430000
566000
NZ
30
66
325700
260500
SO
257
523000
153000
TQ
23
Upmarden
Wa dfield
Walls
villa
villa
Wa llsend
fort
Walton
fort
Walton Heath
Walto n-Le-Dale
W alton-on-Thames
W alton-on-theHill (1)
Wa lton-on-theHill (2)
Wa nborough (Lower)
Wans te ad Park
Ware
W as hi ngborough
Wa sperton
Watchf ie ld
Wa t er Newton
Bvrrivm
Le(c)tocetvm
?
Segedvnvm
villa
065
605
53
355100
428200
SD
551
settlement?
509000
164000
TQ
09
64
villa
522000
155000
TQ
22
55
villa
395000
320000
SJ
95
20
settlement
small
town
(supply base)
Durocornovivm
villa
419500
185200
SU
195
541000
187000
TQ
41
282
852
87
town
535200
214500
TL
352
145
settlement?
502400
370800
TF
024
708
settlement
426100
258100
SP
261
581
setlement
454000
190000
SU
54
511480
297080
TL 1148
9708
14 ?
small
small
town
Dvrobrivae
90
villa
475000
114000
SU
75
settlement
488400
281800
SP
884
W eldon
villa
492940
289990
SP 9294
8999
Well
v illa
426500
481100
SE
265
811
680
W atergate Hanger
W eekley
818
settlement
487500
268000
SP
875
W e l l ow
villa
572000
257000
TL
72
W el t o n Wold
villa
497400
427900
SE
974
We l wy n
villa
523000
216000
TL
23
16
W endlebury
villa
456000
220000
SP
56
20
settlement
323000
177600
v illa
425000
127000
ST
25E
SU
W e i 1 ingborough
W entlooge Level
West Dean
496
57
279
776230
790
27
25
ijlggiiS^
v illa
551150
331000
TF 5115
fiiii
3100
settlement
535920
363090
TF 3592
6309
276
West Dee pi ng
W est Keal
570000
327600
TF
700
563000
313000
TF
63
villa
386000
152000
ST
86
fort
276100
677400
NS
761
+settlement
454000
250000
SP
54
fort
493300
459100
SE
933
591
villa
484700
465700
SE
847
657
villa
485800
464200
SE
858
642
se ttlement
486600
465000
SE
866
villa
460000
204000
SP
60
421900
560200
NZ
19
602
461300
264700
SP
613
647
415
Wes t New to n
We st Winch
villa
villa
W estbury
We s te rw oo d
W e s t on Under wo od
Wetw a ng
W ha r r a m Grange
W ha r r a m Percy
W ha rr am -le-Street
W heatley
W hi ck ha m
fort
(falsus)
52
774
50
650
04
W hi l t o n Lodge
small town
W h i t c h u r c h (1)
small town
354100
241500
SO
541
Whitchurch(2)
villa
354000
217000
SO
54
17
White Staunton
villa
328000
110000
ST
28
10
villa
397840
236100
SO 9784
3610
settlement?
312000
376000
SJ
12
76
345600
309600
SJ
456
369000
548000
NY
69
W hitebeech
Wh itford
Whitley
villa
Wh itley Castle
Bannaventa
13
fort
Epiacvm
096
48
Whitti ng to n Court
villa
401570
220510
SP 0157
Wh ittlebury
vil la
427300
244500
SP
273
445
settlement
545200
247600
TL
452
476
Whitton(l)
villa
614000
246000
TM
14
W h i t t o n (2)
villa
308100
171300
ST
081
Wi ckford
villa
576200
193700
TQ
762
937
Wi gginton
villa
439200
233600
SP
392
336
Wh i ttlesford
2051
46
713
506000
116000
TQ
06
Wilderspool
small
town
361100
386400
SJ
611
Wilmcote
settlement
416000
258000
SP
16
Wils fo rd
settlement?
410860
141480
SU 1086
4148
2962
Wiggonholt
villa
16
864
58
Winchester
civitas
448050
129620
SU 4805
Windermere
settlement?
340000
498000
SD
40
98
villa
624000
157000
TR
24
57
Winterborne
Kingston
Wi nterton
settlement
384500
097500
SY
845
villa
491100
418200
SE
911
Wi tchampton
villa
396320
158700
ST 9632
0587
14
W i n gh a m
Venta Belgarvm
975
182
villa
403000
214000
SP
03
Witt en ha m
settlement
455500
193500
SU
555
W olfhamcote
settlement
452000
265000
SP
52
villa
490100
265000
SP
901
383970
203110
SO 8397
settlement
589000
301000
TF
vi lla
359700
198700
ST 5970
9870
v illa
329000
187000
ST
29
87
+ settlement ?
443000
161000
SU
43
61
385000
254000
SO
85
54
633400
151400
TR
334
514
ST
765
918
Wi thington
Woll as to n
Wo odchester
W oo d co ck Hall
Woolas to n
Wo olstone
Woot t on Hill
villa
Worcester
small
town
Wo r th
settlement
Vertis ?
89
935?
65
650
0311
01
Wort ley
villa
376500
191800
Wraxall
v illa
347000
171000
ST
47
356800
308300
SJ
568
083
475700
139300
SU
757
393
487300
192300
SU
873
923
civitas
W roxeter
+ settlement.
Wyck
W ycombe
Viroconivm
Cornoviorvm
(High)
small
town
497
71
jpaiai*
W y k h a m Park
W ym o n d h a m
W y m o nd l ey
(Great)
settlement
443000
237000
SP
43
villa
484900
318500
SK
849
185
37
villa
521800
226800
TL
218
268
settlement
487400
258100
SP
874
581
+settlement
506690
297900
TL 0669
Y atton
vi lla
340000
165000
ST
40
65
Yeovil
villa
355000
115000
ST
55
15
460500
452000
SE
605
Yard l ey Hastings
Yar well
Yor k
colonia
Eboracvm
498
9790
520
APPENDIX 4: ABBREVIATIONS
499
Archaeom. - Archaeometry
Arch. Rev. - CBA Groups 1 to 13 Archaeological Review
AW - Ancient World
BAA - British Archaeological Association
BAARG - Bristol and Avon Archaeological Research Group
BAJ - Berkshire Archaeological Journal
Banwell AN - Banwell Society Archaeological Newsletter
BAR - British Archaeological Reports
BARG - British Archaeological Research Group
BASTP - Birmingham Archaeological Society Transactions and Proceedings
BBAA
Bulletinof
the
BBCS
Bulletinof
the
Bulletinof
the
Institute of Archaeology,London
BLIA
Bulletinof
Hesperia
Hist.
Historia
Journal ofthe
MerionethHistoricaland
RecordSociety
JNNHS
Journal ofthe
NorthamptonshireNatural
HistorySociety (Field
Club)
JRA - Journal of Roman Archaeology
JRAFCC - Journal of the Royal Air Force Collection, Cranwell
JRMES - Journal of Roman Military Equipment Studies (Great Britain)
JRS - Journal of Roman Studies
KAR - Kent Archaology Review
Klio - Klio. Beitrage sur Alten Geschightte (Berlin)
LA - The London Archaeologist
LAAA - Liverpool Annals of Anthropology and Archaeology
LAHS - Leicestershire Archaeologocal and History Society
LAASRP - Lincolnshire Architectural Society Reports and Papers
LAT - Lincoln Archaeological Trust
LDASB - Loughborough and District Archaeological Society Bulletin
LHA - Lincolnshire History and Archaeology
Lines. Mag. - Lincolnshire Magazine
501
NMR
NA - Norfolk Archaeology
NAJ - Norfolk Archaeological Journal
Northants. Arch. - Northamptonshire Archaeology
Not. Dig. - Notitia Dignitatum Occidentis
OAHSNS - Oxford Archaeology and History Society News Sheet
OJA or OxJA - Oxford Journal of Archaeology
Ordnance Note Book - Ordnance Survey
Oxon. - Oxoniensia
PAI
PAS
History Society
502
Society of Antiquaries of
Newcastle
Society of Antiquaries of
Scotland
??
the University
Society of London
Monuments
inScotland
503
Society
505
APPENDIX
BIBLIOGRAPHY
Abbott F F & Johnson A C , 1968, Municipal Administration in the Roman
Empire. New York.
Adams J-P, 1994 (English Ed.), Roman Building: Materials and Techniques.
London.
Adkins L & R A, 1986, Under the Sludge: Beddington Roman Villa.
Aicher P, 1995, Guide to the Aqueducts of Ancient Rome. Wauconda, 111.,
USA.
Aitken G & N, 1982, 'Excavations on the Library Site, Colliton Park,
Dorchester, 1961-3', PDNHAS 104, 93-126.
Alcock L, 1955, 'Castell Collen Excavations', Trans. Radnorshire Soc. 25,
46.
Allcropt A M ,
Sc
Sc
224. Oxford.
Andrew S Sc Lees W, 1911, The Roman Forts at Castleshaw. Manchester.
Annable F K, 1954, 'The Roman Pottery Kilns at Cantley Housing Estate,
Donacaster: Kilns 1-8', YAJ 38, 403-6, pis.1-5.
Annable F K, 1960, The Romano-British Pottery at Cantley Housing Estate,
Doncaster: Kilns 1-8, Doncaster Museum and Art Gallery Publication
no. 24. Doncaster.
Annable F K, 1976, 'A Late First Century Well at Cunetio', WANHM 70-1,
1975-6, 126-7.
Applebaum E S, 1932, 'Excavations at Baldock in 1932', TSAHAAS, 244-58.
Applebaum S, (ed.j, 1967, Roman Frontier Studies. Tel Aviv.
Applebaum S, 1975, 'Some observations on the economy of the Roman Villa
at Bignor, Sussex', Britannia 6, 118-32.
506
507
Sc
Miles D,
1974,
Sc
Biddle M
Boon
Sc
G C,
1993,
R N Quirk R N,
Antiq. J.
19 6 1 ',
119,
24,
247-9.
150-94.
Sc
Military Supply in
508
Birley E, 1931, 'An introduction to the excavation of ChesterholmVindolanda', Arch. Ael.(4) 8, 182-212.
Birley E, 1960, Chesters Roman Fort. London.
Birley E, 1961, Research on Hadrian's Wall. Kendal
Birley R, 1977, Vindolanda: A Roman frontier post on Hadrian's Wall.
Greenhead.
Birley R, 1990, Vindolanda's Roman Records,
King A
Sc
BAR 13 6. Oxford.
Blair R
Sc
Carr
S,
1925,
1,
249-60.
through
Sc
509
(JEAR). Paris.
Bowen E G
Sc
1975,
'Gatcombe', in Rodwell
Sc
15,
175-81. Oxford.
Branigan K, 197 6, The Roman Villa in South-West England. Bradford-onAvon.
Branigan K, 1977, Gatcombe: The Excavations and Study of a RomanoBritish Villa Estate 1967-78, BAR 44. Oxford.
Branigan K, 1994, 'The New Roman Britain - a View from the West Country',
TBGAS 112, 9-16.
Branigan K
Sc
Fowler
P J
(ed.s),
1976,
Abbot.
Breeze D J, 1983, Roman Forts in Britain. Aylesbury.
Breeze D (ed.), 1984, Studies in Scottish Antiquity. Edinburgh.
Breeze D, 1987, Hadrian's Wall: A souvenir guide to the Roman Wall, 2630. English Heritage, London.
Breeze D J
Sc
Brendel O, 1979, Prolegomena to the Study of Roman Art. New Haven, Conn.,
USA.
Brewer R J, 1990, 'Caerwent - Venta Silurum: a civitas capital', in
510
(unpublished notes).
Sc
511
512
1956,
28, 94-5.
Clarke G, 1982, 'The Roman Villa at Woodchester', Britannia 13, 195-228,
pis.xiv-xvii.
Clarke J, 1933, The Roman Fort at Cadder, Strathclyde. Glasgow.
Clarke S, 1995, 'Quantative analysis of finds from the Roman fort at
Newstead - some preliminary findings', TRAC 94, 72-82. Oxford.
Clarke S, 1996, 'Abandonment, Rubbish Disposal and "Special" Deposits at
Newstead', TRAC 1995, 73-81. Reading.
Clarke s 5c Jones R F J, 1994, 'The Newstead pits', JRMES 5, 109-24.
Cleary, Esmonde S, 1987, Extra-Mural Areas of Romano British Towns, BAR
194. Oxford.
Clowne P, 1980, 'A section through the Car Dyke in Horbling Fen', S.
Lincs. Arch. 4, 19-23
513
76-84,
Fig.1.
Collingwood R G & Wright R P, 1965, The Roman Inscriptions of Roman
Britain, RIB, Vol.I, 'Inscriptions on Stone'; Vol.II,
'Inscriptions
(typed
Sc
Sc
Yorkshire. York.
Corder P, 1951, 1954, 1961, The Roman Town and Villa at Great Casterton:
1st Interim Report', 2nd Interim Rep., 3rd Interim Rep.
Nottingham.
Cotton M A & Cathercole P W, 1958, 'Excavations at Clausentum,
Southampton, 1951-4', Min. of Works Arch. Dept., Report No. 2,
HMSO, London.
Coulton J J,
1987,
Thompson,
1987,
72-84.
514
Sc
20,248-53.
Cox T, 1720, 'Magna Britannia et Hibernia', vi, 240, p.12b, CUAPS (?),
31-41.
(Dorn).
Cunliffe B,
in Britain. London.
Cunliffe B,
Portsmouth.
Cunliffe B,
Britannia 7, 1-32.
Cunliffe B, 1978, Rome and her Empire. London.
Cunliffe B, 1980 'Excavations at the Roman Fort at Lympne, Kent, 1976-8',
Britannia 11, 227-88.
515
Sc
Kirby D P, 1994,
275-306. Cardiff.
Davies O, 1935, Roman Mines in Europe,
94-139). Oxford.
Davies 0, 1936, 'Finds at Dolaucothi', Arch. Cambr. 91, 51-7.
de Garros C B, 1989, Aqueduc Romain de Vers. Cahours, France.
DeLaine J, 1988, 'Recent research on Roman Baths', JRA 1, 11-32.
Dearne M J, 1993, Navio: The Fort and Vicus at Brough-on-Noe, Derbyshire,
BAR 234. Oxford.
Dent J S, 1983, 'The Impact of Roman Rule on Native Society in the
Territory of the Parisi', Britannia 14, 35-44.
Diamond H W, 1847, 'Account of Wells or Pits, containing Roman remains
discovered at Ewell in Surrey', Archaeologia 32, 451-55.
Dilke O A W, 1971, The Roman Land Surveyors: an introduction to the
Agrimensores. Newton Abbot.
Dilke 0 A W, 1980, Surveying the Roman Way. Newton Abbot.
Dixon P, 1994, Crickley Hill, Vol. 1, The Hill fort Defences. Nottingham.
Donel L, 1993, 'Roman Well, St Paul-in-the Bail', CLAU Archaeological
Rep. N o .63.
516
Sc
Sc
Quantative studies,
Roman Empire:
517
n.34. Paris.
Etienne R, 1960, Le quartier nord-est de Volubilis. Paris.
Evans E, Dowdell G & Thomas H J, 1985, 'A third century Maritime
Establishment at Cold Knap, Barry, South
57-125.
Eyice S, 1979, 'Byzantinische Wasserversorgungsanlagen in Istanbul', in
Leichtweiss-Institut fur Wasserbau, Mitteilungen, Heft 64, 31,
abb.5 (cistern Yerabatan). Bonn.
Fasham P J & Whinney R J B, 1981, Archaeology and the M3: The Watching
Brief. The Anglo-Saxon settlement at Abbots Worthy and
retrospective sections, Hampshire Field Club and Archaeological
Soc. Monograph 7, 5-11. Winchester.
Faull M L & Moorehouse S A (eds.), 1981, 'West Yorkshire: Archaeological
Survey to AD 1500', West Yorkshire Archaeology Survey. York.
Ferris I M & Jones R F J, 1980, 'Excavations at Binchester 1976-79', in
Roman Frontier Studies, W S Hanson & L J F Keppie (eds.), BAR Int.
Ser. 71 (i), 233-48. Oxford.
Field N H, 1967, 'Excavations on the military site at Lake, near
Wimborne', PDNHAS 89, 143.
Field N H, 1969, 'Excavations on the military site at Lake, near
Wimbourne', PDNHAS 91, 188-9.
Fieldhouse W J, May T & Well Stood, 1931, A Romano-British Industrial
Settlement near Tiddington, Stratford-on-Avon. Birmingham.
Finberg H P R, 1955, Roman and Saxon Withington: A Study in
Continuity. Leicester.
Fitz J (ed.), 1977, Limes: Akten des XI Internationalen Limeskongress.
Budapest.
518
Sc
Sc
St Joseph J
Britannia
Frere
S S Sc
5,
Wilkes
K,
1974,
1-129.
J J,
1989,
519
9.
London.
in Peacock
1977, 35-84.
Fulford M, 1989,
and Drainage
YAJ 38, 53 6-
45, pis.1-4.
Gilmour E F, 1955, 'The Roman pottery kilns at Cantley Housing Estate,
Doncaster, Kilns 9-15', YAJ 38, 407-12, pls.A-E.
520
521
Sc
BAR
Sc
522
523
Hobley B, 1975, '"The Lunt" Roman Fort and Training School for Roman
Cavalry, Baginton, Warwickshire, Final Report: Excavations 1972-73,
with conclusions', TBWAS 87, 1-56.
Hoare R C, 1812-21, The Ancient History of South and North Wiltshire,
Vol.l & 2. London.
Hoare R C, 1819, Roman Aera, 117-21. London.
Hodge A T, 1983, 'Siphons in Roman aqueducts', PBSR 51, 174-221.
Hodge A T, 1984, 'How did Frontinus measure the quinaria? ', AJA, 88, 20516.
Hodge A T, 1989, 'Aqueducts', in Barton 1989, 127-49.
Hodge A T (ed.), 1991a, Future currents in aqueduct studies. Leeds.
Hodge A T, 1991b, 'A Roman Factory', Scientific American 263(5), 58-64.
Hodge A T, 1992, Roman Aqueducts and Water Supply. London.
Hodge A T, 1996, 'In Vitruvium Pompeianum: urban water distribution
reappraised', AJA 100, 261-76.
Hodgson J, 1816, 'Observations on an ancient aqueduct and certain heaps
of iron scoria in the parish of Lanchester in the County of
Durham', Arch. Ael.(l), 1, 118-21.
Hodgson J, 1840, A history of Northumberland, Vol.4, Newcastle.
Hogg A H A ,
Arch.
J. 125, 101-92.
Hogg A H A ,
Hope R C, 1893, The Legendary Lore of the Holy Wells of England. London.
Hopkins K, 1980, 'Taxes and Trade in the Roman Empire', JRS 70, 101-25.
Horsfall N, 1987, 'How Not to Build an Aqueduct', Omnibus-Omnibus Vols.
1-7, 40-1.
Horton M
Sc
Sc
(Horningsea).
524
(originally published
Sc
essays presented
525
Sc
Jones G
Sc
G D
Sc
Shotter
D C A,
1988,
Sc
179-83.
11.
Jones G
Sc
Jones G
Sc
Sc
Hobley (eds.) .
Sc
PAM,
O'Neil. London.
InterimReporton
84, 82-107.
of the Sewers and
DrainsofLondonand the
526
Sc
Laver H, 1887, 'On a Roman Villa at Arlesford Lodge, Excavated June 1885'
TEAS 3(2), 136-9, pls.v, v i .
Leach P J, 1982, 'Ilchester Vol.I : Excavations 1974-5', Western Arch.
Trust Exc. Monograph 3. Taunton.
Leach P J, 1985, 'Ilchester Archaeological Excavations 1985: an Interim
Report', Birmingham Univ. Field Arch. Unit. Birmingham.
Leach P J, 1992, 'Ilchester Vol.2: Archaeological Excavations and
Fieldwork to 1984', Shefield Excavation Reports
Leach J D
Sc
2,
Sc
7. Taunton.
527
Sc
Liversidge
D J
Smith
J,
Sc
Stead
M,
Lukis
Lukis
Surtees
Surtees
Sc
London.
Lysons S, 1797, Account of Roman Antiquiries discovered at Woodchester.
London.
Lysons S, 1813, Reliquiae Britannico-Romanae, Vols. 3, 4. London.
McNeil R
Sc
Roberts
Britannia
R,
1987,
'A
18.
Sc
Park A,
1906,
J Sc
Barbour
J,
1897,
Sc
528
Sc
Mann J C,
&
75-9.
Manning W
in theMuseum of
529
Marsden
21-34.
Mathioulakis C Z, 1966, Crete: Guide and Map. Athens.
Mattingly D J, 1995, Tripolatania. London.
Mattingly D J & Hitchner R B, 1995, 'Roman Africa: An Archaeological
Review', JRS 85, 165-213.
Maxwell G S, 1975, 'Excavations at the Roman Fort of Bothwellhaugh,
Lanarkshire, 19 67-8', Britannia 6, 20-35.
Maxwell G S & Wilson D R, 1987, 'Air Reconnaissance in Roman Britain',
Britannia 18, 1-48.
May T, 1900, 'Excavations on the site at Wilderspool and Stockton Heath
1899-1900', THSLC 52, 1-52.
May T, 1903-4, 'Excavations on the site at Wilderspool and Stockton Heath
1901-4', THSLC 55/56, 209-37.
May T, 1922, The Roman Forts at Templebrough, near Rotherham. Rotherham.
McKay A, 1978, Vitruvius, Architect and Engineer. Bristol.
McWhirr A, 1970-1, 'A Roman Villa at Tixover Grange, Rudland', Trans.
Leicestershire Arch. Hist. Soc. 46, 1-8.
McWhirr A, 1972, Archaeology in Leicestershire and Rutland. Leicester.
McWhirr A, 1981, Roman Gloucestershire. Gloucester.
McWhirr A, 1986, Cirencester Excavations III, Houses in Roman
Cirencester. Cirencester.
Meates G W, 1963 (2nd ed.), Lullingstone Roman Villa. London.
Meats G W, 1979, The Roman Villa at Lullingstone, Kent, Vol.l, 'The
Site'. Maidstone.
Merrifield R, 1965, The Roman City of London. London.
Merrifield R, 1969, Roman London. London.
Mickett R, 1983, The Roman fort at South Shields: Excavations of
Defences, 1977-81', 7-11, Tyne & Wear Museums. Newcastle.
Middleton J H, 1893, 'On a Roman Villa at Spoonleywood, Gloucestershire;
and on Romano-British houses generally', Archaeologia 52, 201-21.
Miles D (ed.), 1982, The Romano-British Countryside, BAR 103(2). Oxford.
Miles D, 1986, Archaeology at Barton Court Farm, Abbington, Oxon, CBA
Res. Rep. 50. Oxford.
530
J R Sc
Maynard
G,
Ipswich', PSIA
1931-33,
21,
240-62.
Money
Moore
531
Nash-Williams V E, 1969,
ed. by M G Jarrett.
Nash-Williams V E, The Goldmines at Dolaucothi', BBCS 14, 1950, 79-84.
Neal D, 1980, Gorhambry Excavations. London.
Neal D, Wardle A & Hunn J, 1990, Excavation of the Iron Age, Roman and
Medieval Settlement at Gorhambry, St Albans. London.
Neal D S, 1974, The Excavation of the Roman Villa Gadebridge Park, Hemel
Hempstead, 1963-8. London.
Neal D S, Witcombe Roman Villa: A reconstruction. Chichester.
Neal D S, 1989, 'The Stanwick Villa, Northants: An Interim Report on the
Excavations of 1984-88', Britannia 20, 149-71.
Nevill R, 1883, 'A Roman Villa at Chiddingfold',
308-9, 334.
Neville R C, 1853, 'Investigations of Remains in the County of Essex,
1852', Arch. J. 10, 14-7.
Nichols J, 1810, The History and Antiquities of the County of Leicester,
Vol.II, Part II, 505-6. London,
(reprint. 1976,
Wakefield).
Nielsen I, 1990, Thermae et Balnea: The Architecture and Cultural History
of Roman Public Baths, Vol.l, Text, and Vol.II, Catalogue and
Plates. Aarhus, Denmark.
Oldfield E, 1829, Topographical & Historical Account of Wainfleet and the
wapentake of Candleshoe, in the county of Lincoln. London.
O'Leary T J,
532
(1991),
ix.2, 110-24.
Peacock D P S
Bromley.
533
Philp B J f et al, 1991, 'The Roman Villa at Keston, Kent: First Report,
Excavations 1968-1978: 6th Res. Rep. Kent Monograph Series. Kent.
Pickenden N P, 1986, 'Prehistoric settlement and the Romano-British
"Small Town" at Heybridge, Essex', TESAH 17, 7-68.
Pickering A J, 1935, 'The Roman Sites of S-W Leicestershire', TLAAS,
1934, xviii, 41-86; xviii, 1935, 157-94, at 162-7.
Pitt-Rivers A, 1987, Excavations in the Cranborne Chase, near Rushmore,
on the Borders of Dorset and
I & V. London.
Pitts
Potter T W,
ScPrice,
A Guide to
1881,
London.
Price J
Sc
Putnam W G,
Aqueduct 1995', Dorset Proc. 117, 1995, 128-31. Also 114, 1992,
239-40; 115, 1993, 152-3; 116, 1994, 123-4.
Putnam
W G Sc
Hewitt I,
Roman Aqueduct,
1996,
1 9 9 6 ',
1-6).
A Sc V ,
1974,
1 9 6 6 ',
Britannia
5,
1954-
163-224.
41-77.
Radford C A R
Sc
Coker. London.
Rahtz
P Sc
Fowler
P,
1972,
P J,
(ed.),
217 .
534
187-
Rahtz P & Watts L, 1989, 'Pagans Hill Revisited', Arch. J. 146, 330-71.
Rainey A, 1973, Mosaics of Roman Britain: a gazetteer. Newton Abbot.
Rakob F, 1979a, 'Wasser als Element rdmischer Infrastruktur', D U Die
Kunstzeitschrift, 3, 66.
Rakob F, 1979b, 'L'aqueduc de Carthage', Doss. Arch., 38,
(Oct-Nov,
1979) .
Ramm H G, 1971, 'The end of Roman York', in Butler 1971, 179-99.
Ramm H, 1978, The Parisi. London.
Reed A H & Austin W, 1976, 'An Archaeological Survey of the Lanchester
Area', in Clack & Gosling (eds.), 213-21.
RCHM 1913, An Inventory of Historical Monuments of Buckinghamshire, 2
Vols. London.
RCHM 1926, An Inventory of the Historical Monuments in Huntingdonshire,
95-7, No. 31. London.
RCHM York, 1962, Eburacum, Vol.I, 49-51, 59, 61. London.
RCHM, County of Dorset, England, 1970, 'Roman Dorchester (Durnovaria)',
Vol 2(3), 533-89. London.
RCHM 1976, Iron Age and Roman Monuments in the Gloucestershire Cotswolds.
London.
Renfrew C & Shennan S, 1982, Ranking, Resource and Exchange Aspects of
Archaeology of Early European Society, 79-88.
RCHM(E), 1983, 'West Park Roman Villa, Rouckbourne, Hampshire', Arch. J.
140, 129-50.
Richards J, 1978, The Archaeology of the Berkshire Downs: An Introductory
Survey. Read in g .
Richardson K M, 1938, 'Report on Excavations at Verulamium', Archaeologia
90, 85-86.
Richardson K M, 1940, 'Excavations at Poundbury, Dorchester, Dorset,
1939', Antiq. J.(4) 20, 429ff.
Richmond I A, 1940, The Romans in Redesdale. Newcastle upon Tyne.
Richmond I A, 1941, 'The Water Supply of the Roman Fort at Lyne, PSAS
75, 39-43.
Richmond I A, 1946, 'The Four Coloniae of Roman Britain', Royal Arch.
Inst. 103, 76-77.
Richmond I A, 1954, The Roman Fort at South Shields. Newcastle.
Richmond I A, 1963, The Roman Pavements from Rudston.
Richmond I, 1968, Hod Hill: Excavations carried out between 1951 and
1958, Vol. 2, 87. The Trustees of the British Museum. London.
Richmond I A, Romans T & Wright R P, 1944, 'A Civilian Bath-house in the
Roman Period at Old Durham', Arch. Ael.{4) 22, 1-21.
535
comments
Rivet A
Rivet A
Britannia 1, 34-68.
Rivet A L F
Sc
S, 1975,
Birrens (Blatobulgium).Edinburgh.
Robertson A
S, 1990,
Glasgow.
Robertson A S ,
et al, 197 5, Bar Hill, a Roman Fort and its finds, BAR 16.
Oxford.
Robertson J
Robins F W,
(1994
pbk). London.
Rodwell K A, 1988, 'The Prehistoric and Roman Settlement at Kelvedon,
Essex', CBA Res. Rept. 63. London.
Rodwell W, 1972, 'The Roman Fort at Great Chesterford, Essex', Britannia
3, 290-3.
Rodwell W, 1978, 'Rivenhall and The emergence of 1st century villas in
northern Essex', in Todd, 1971, 32.
Rodwell W & Rowley T (eds.), 1975, The 'Small Towns' of Roman Britain,
BAR 15. Oxford.
Rook A G, 1968, 'A Romano-British Well at Welwyn', Hertfordshire
Archaeology, 1.
536
conservation and
Salway
Scarth
Schnitter
des
Selkirk A,
Selkirk A,
537
London.
Vol.1. London.
Vol.3. London.
Smith C R, 1876, 'On a Roman Villa near Maidstone', Arch. Cant. 10, 16372 .
Smith D J, 1978, 'Regional aspects of the winged corridor villa in
Britain', in Todd M (ed.) 1978.
Smith J T, 1962, 'The Roman Villa at Denton, Part I', LAASRP 10(1), 75104.
Smith J T, 1978, 'Halls or Yards? A problem of Villa interpretation',
Britannia 9, 3 51-8.
Smith N A F, 1970, 'The Roman dams of Subiaco', Technology and Culture
1(1), 58-68.
Smith N A F, 1971, A History of Dams, Chapter 2, 'The Romans', 25-49.
London.
Smith N A F, 1976a, 'Attitudes to Roman Engineering and the Question of
the Inverted Siphon', History of Technology, Vol.l, 45-71.
Smith N A F, 1976b, Man and Water: A History of Hydro-Technology, Chapter
7, 'Roman Water Supply, 81-93. London.
Smith N A F, 1990, 'The Pont du Gard and the Aqueduct of Nimes', The
Nucomen Society Transactions, 62, 1990-91, 53-80.
Smith N A F, 1991, 'Problems of Design and Analysis', in Hodge (ed.),
1991a, 113-28.
Smith R F, 1977, 'Hibaldstow, Staniwells Farm', LHA 12, 74.
Smith R F, 1987, Roadside Settlements in Lowland Roman Britain, BAR 157.
Oxford.
Smith D J, Hird L Sc Dix B, 1988-9, 'The Roman Villa at Great Weldon,
Northamptonshire', Northants Arch. 22.
Smoothy M D, 1989, 'A Roman Rural Site at Rayne, Essex: Excavations
1987', TEAH 20, 1-29.
Somer S, 1984, The Military Vici of Roman Britain, BAR 129. Oxford.
Speidel M A, 1992, 'Roman Army Pay Scales', JRS, 82, 87-106.
Sprague de Camp L, 1978, The Ancient Engineers. Cambridge, USA.
538
Joseph
S,1958a,
'Air
'The
Joseph
S,1958b,
JosephJ K S,
539
Sumpter A B, 1988, Iron Age and Roman at Dalton Parlours, BAR 193,
Oxford.
Swan V G, 1984a, The Pottery Kilns of Roman Britain. London.
Swan V G, 1984b, A Gazetteer of the Pottery Kilns of Roman Britain.
London.
Tabula Imperii Roman Condate-Glevum-Londinium-Lutetia, 1983, The British
Academy,
(Stroud 1991).
540
London.
Wacher J S, 1966, The Civitas Capitals of Roman Britain. Leicester.
Wacher J S, 1969, 'Excavations at Brough-on-Humber, 1958-61', Research
Report of the Soc. of Antiquaries, No.25, London Soc. of
Antiquaries, 16, 41-2.
Wacher J S, 1971, 'Yorkshire towns in the 4th century', in Butler R M,
1971.
Wacher J S, 1974, The Towns of Roman Britain, London.
Wacher J S, 1978a, Roman Britain, 215, London.
Wacher J S, 1978b, 'The water supply of Londinium', in Collectanea
Londiniensis, London Middlesex Arch. Soc., Special Paper 2, 1978,
104-8.
Wacher J S, 1995 (rev. 2nd ed.), The Towns of Roman Britain. London.
Wacher J S, 1981, discussion,
20, 249-54.
541
542
(rev.
(134-44).
Wheeler R E M ,
543
Wilmott A, 1982b, 'Water supply in the Roman City of London', The London
Archaeologist 1982, Vol.4(9), 234-42.
Wilmott 1984, 'Roman Timber Lined Wells in the City of London: Further
Examples', TLMAS 35, 5-10.
Wilmott A, 1991, Excavations in the Middle Walbrook Valley, City of
London, 1927-1960, TLMAS Special Paper 13.
Wilson A, 1995, 'Running Water and Social Status in North Africa', in
Horton H & Wiedemann T, Centre for Mediterranean Studies, 52-6.
Bristol.
Wilson A E, 1947, 'Angmering Roman Villa', SxAC 86, 1-21.
Wilson D R & Sherlock D, 1980, 'North Leigh Roman Villa'. HMSO, London.
Wilson P R, 1984, 'Recent work at Gatterick', in Wilson, et al, 75-82.
Wilson P R, Jones R F J & Evans D M,
544
Sc
Sc
b.
J,
1969,
Excavations
at
Hardingstone,
Northants,
1967-8.
Northampton.
Wooler E, 1917, The Roman Fort at Piercebridge, County Durham. Frome.
Wrathmell S
Sc
Wright R P, 1943, 'The Whiteley Castle altar to Apollo', JRS 33, 36-38.
Wright R P
Sc
(1995 pbk).
Cambridge, USA.
Yule B, 1982, 'A third century well group, and the later Roman settlement
in Southwark', The London Archaeologist 4(9), 243-9.
Zojak N, 1996, 'Did the Romans go to the baths to get clean or dirty',
Omnibus 32, 16-20.
Zienkiewicz J D, 1986, The Legionary Fortress Baths at Caerleon, Vol.1:
The Buildings, and Vol.2: The Reports. Cardiff.
545