Visible Tools, Invisible Craft
An analysis of textile tools, textile production, and craft organisation along
the south coast of Iron Age Britain
This thesis is submitted for the degree of Doctorate of Archaeology
Submitted September 2020
Lewis Jean Ferrero
Wolfson College
2
3Declaration
This thesis is the result of my own work and includes nothing which is the outcome of work
done in collaboration except as declared in the Preface and specified in the text. It is not
substantially the same as any that I have submitted, or, is being concurrently submitted for a
degree or diploma or other qualification at the University of Cambridge or any other
University or similar institution except as declared in the Preface and specified in the text. I
further state that no substantial part of my dissertation has already been submitted, or, is
being concurrently submitted for any such degree, diploma or other qualification at the
University of Cambridge or any other University or similar institution except as declared in
the Preface and specified in the text.
4
5Acknowledgements
This thesis would not be possible without the help of a great many people, including museum
staff across the south coast of Britain. I would like to extend my personal thanks to Jayne
Wackett and the staff of the Royal Cornwall Museum for helping me access the collections
and gathering data. Without their help, I would not have been able to include Cornwall in this
thesis at all. I would also like to thank Andrew Richardson and the staff at the British
Museum, who allowed me to access and measure textile tools from several key sites across
the study area. This was crucial in including Hod Hill in this thesis and expanding the data
collections in other sites. My thanks go out to Barry Chandler and the staff at the Torquay
Museum for their aid and enthusiasm in accessing and recording the textile tools from Kent’s
Cavern and other sites. They were essential in providing the data for Kent’s Cavern, one of
the most important and intriguing sites in Devon. I would also like to thank Fiona Pitt and the
staff at the Plymouth Museum, who managed to fit my visit in during a museum-wide
renovation and a mass movement of collections. Their time and patience was greatly
appreciated. I would like to extend my gratitude to Thomas Cadbury and the staff at the
Royal Albert Memorial Museum in Exeter. They worked hard to provide the necessary
artefacts that were able to bolster the data collected in Devon, which was both essential and
very appreciated.
I would like to thank Richard Breward and the staff at the Dorset County Museum, who
were also undergoing a reshuffle of the museum’s collections during my visit. They not only
provided me with the textile tools I needed but also made me feel welcome, and made
excellent tea. My thanks also go to Katherine Walker and the staff at Hengistbury Head
Visitors Centre, who worked hard to track down the textile tools and excavation reports for
my visit. Their pride in their site was evident in all their hard work. I am grateful to Jenny
Stevens and the Andover Museum staff, who were a great help in accessing the multitude of
Danebury finds. They also provided an upbeat atmosphere and some very welcome cups of
tea during my winter visit. I would also like to thank David Allen and the Museum of the
Iron Age staff for their hard work in providing access during their own major restructuring. I
am very grateful that they could find time to fit me in such a hectic period. My thanks also go
out to Amy Roberts and the staff at both the Novium Museum and Fishbourne Roman Palace
for providing access to a number of site collections. Their hard work and preservation in
tracking down artefacts was very appreciated. I would like to thank Emma O’Connor and the
Barbican House Museum staff for their time and effort in tracking down collections. Their
6persistence was essential in accessing several collections. I would also like to thank James
Sainsbury of the Worthing Museum and Art Gallery, who did his best to track down every
artefact possible. I was very impressed with his dedication and pride in his museum, and
greatly appreciated the time he took to help me. Finally, I would like to thank Andrew
Richardson and the staff at the Canterbury Archaeological Trust, who worked hard to track
down and provide access to several essential Kent collections. They provided access to
written records as well as artefacts and were enthusiastic in helping my research, even during
a period of renovations.
It would be very remiss not to acknowledge and express my appreciation for my supervisor,
Marie Louise Stig Sørensen, who has been essential in guiding my work and encouraging my
improvement both in this thesis and as a well-rounded academic. Without her, I would have
missed out on many wonderful and crucial experiences and opportunities.
Finally, I want to give my heartfelt thanks and appreciation to my parents, Tony and Allison
Ferrero, for providing patience, love, support, and encouragement for both myself and my
work. I cannot say how grateful I am that you had my back.
7Abstract
Textile production was an essential element of society in prehistory, supplying people with
clothing, furnishings, and utilitarian items like sails, straps, and sacking. The fragility of
textiles means few examples of prehistoric fabrics survive in Britain and northwest Europe
outside of some special conditions. Thus tools that produced textiles are the main source of
information on textiles created and how these were made. Textile tools also provide evidence
that textile fragments cannot: where production occurred and how it was organised.
Identifying areas and modes of production provides insight into how settlements and craft
were organised; e.g. whether a craft and workers were independent or controlled by social
elites. Analysis of tools can indicate levels of skill required for production, quality, and
quantity of goods, providing evidence on the economy these tools contributed to. This study
focuses on textile craft as an economic activity and its organisation during the Iron Age’s
period of social change and development, c.800 BC - AD 43. This period of expansion saw
changes in craft production and tools, and development of new forms of settlement across
Britain and Europe. This thesis presents a unique study on textile tools from various
settlement forms across Cornwall, Devon, Dorset, Hampshire and Kent in Britain, via
analysis of such tools and their regional distribution. Five sites per county were chosen based
on the number of textile tools, except Kent where few excavations discovered more than one
or two textile tools. Comparing tool materials with local geology indicates the importance of
textile production compared to other crafts at each site, indicating the status of textile work
and its workers across the south coast. Data collected from the tools demonstrates what
textiles were most likely produced at the selected sites, identifying areas with specialist
workers and areas of basic household production, and highlighting how these communities
organised this craft.
8
9Contents
Declaration ii
Acknowledgements iv
Abstract vii
Contents ix
List of Figures xiv
List of Tables xviii
List of Appendices xxi
1. Introduction 1
Textiles and textile production 2
Textile tools 5
The British Iron Age 8
Research aims 11
Thesis structure 12
2. Previous Work 17
Textile fibres 17
Fibre preparation tools 22
Spinning tools 24
Weaving 28
Weave patterns 29
Weaving tools 31
Looms 34
Tablet weaving 88
Dyeing 41
Finishing techniques 44
Fulling 44
Embellishing 44
Textile Workers and Organisation 46
Direct evidence of skill 47
Craft specialisation 49
Conclusion 55
Society in the Iron Age 57
Defining development in Iron Age Britain and Europe 57
Iron Age society and organisation 59
Settlement in the Iron Age 61
South West 64
Central South 68
South East 72
Conclusion 74
10
3. Methodology 75
Issues and resolutions 75
Site selection and study area 77
Data collection and recording 80
Data presentation and discussion 80
4. Data 82
Cornwall 82
Bodrifty 82
Carn Euny 83
The Rumps 84
Trevelgue Head 117
Trevisker 111
Devon 88
Blackbury Camp 88
Dainton 89
Kents Cavern 91
Milber Down 92
Mount Batten 93
Dorset 95
Allard’s Quarry 95
Gussage All Saints 96
Hengistbury Head 98
Hod Hill 99
Maiden Castle 101
Hampshire 104
Danebury 104
Easton Lane 106
Kennel Farm 107
Winnall Down 108
Zions Hill Farm 110
Kent 111
East Wear Bay, Folkestone 111
Highstead 113
Sunset Caravan Park 114
Thanet Earth 116
5. Discussion 118
Tool analysis 118
County by county overview - Spindle whorls 118
Loom weights 121
Tools by settlement type - Fortified sites: Spindle whorls 125
Loom weights 128
Enclosed sites: Spindle whorls 132
Loom weights 134
11
Open sites: Spindle whorls 137
Loom weights 138
Other sites: Spindle whorls 139
European textile evidence 140
Tool distribution 144
South West 144
Central South 148
South East 156
Reused pottery spindle whorls 159
Wider context 167
Areas of specialisation and raw material production 167
Specialist workers and centres of production 170
Skill and social status 174
Evidence for textile standardisation 181
6. Conclusions 183
Aims and objectives 183
Advantages 185
Issues 187
Further work 188
Bibliography 190
12
List of Figures
. Top left, Iron Age weaving comb from Danebury (Cunliffe and Poole 1991); bottom left,
Viking Age wool comb from Buiston Crannog (Ashby 2009: fig 2); top right,
Iron Age spindle whorl from Kent’s Cavern (Silvester 1986); bottom right: Iron
Age loom weights from Danebury (Cunliffe and Poole 1991).
. Typical shapes of loom weights from the Bronze Age Aegean (left) and Iron Age Britain
(right). Author’s illustration, after Mårtensson et al 2009.
. The basic textile chaîne opératoire with tools and resources, after Belanová-Štolcová and
Grömer 2010: 9 (author’s illustration).
. Examples of A) flax, B) jute, and C) hemp. (Brockhaus 1908)
. The direction of S-spin (left) and Z-spin (right) threads. Author’s illustration.
. Examples of a bottom weighted spindle (left), middle-weighted spindle (centre), and
top-weighted spindle (right). Author’s illustration.
. Examples of spindle whorl shapes, following Gleba’s (2008: 105) typology. Globular
(upper left), discoid (lower left), cylindrical (middle), conical (upper right), and
truncated conical (lower right). Author’s illustration.
. An example of how direction of spin can alter the appearance of a textile.
s illustration.
. An example of a late Bronze Age bone pin beater from Cyprus (top, after Smith 2001: 84,
fig. 2) and a Iron Age weaving sword from Britain, made of red deer bone
(bottom, courtesy of Museum of London Online Collections). Not to scale.
0. An example of the form and manufacturing process of a weaving comb. A) Areas of
antler used for various purposes. B) Splitting of antler, ready for carving into a
weaving comb. After MacGregor 1985, fig. 42.
1. Examples of a two beam loom (left), ground loom (centre), and warp weighted loom
(right). Author’s illustration.
2. Unevenness in weaving, resulting from too few loom weights (left) and too many (right).
Author’s illustration, after Mårtensson et al. 2007 and 2009.
3. Comparison of Mårtensson’s reconstruction of a Bronze Age Trojan loom weight (left)
and a typical British Iron Age loom weight (right). Author’s illustration, not to
scale.
4. An example of tablet weaving structure, warping, and method. Author’s illustration
5. Examples of typical settlements in the South West. A) Trevisker, enclosure. B) Carn
Euny, open settlement. C) Milber Down, multiple enclosed settlement. D)
Chyauster, courtyard settlement. E) Trevelgue Head, cliff castle. After Cunliffe
2005.
6. Examples of typical settlements in the central South. A) Bramdean, banjo enclosure. B)
Tollard Royal, ditched enclosures. C) Winnall Down, earthwork enclosed
settlement. D) Maiden Castle, hillfort. After Cunliffe 2005.
7. Examples of typical settlements in the South East. A) Caburn, palisaded enclosure. B)
Oldbury, hillfort. After Cunliffe 2005.
8. Map displaying the study area and the sites used in this thesis. 1- Carn Euny; 2- Bodrifty;
3- The Rumps; 4- Trevelgue Head; 5- Trevisker; 6- Mount Batten; 7- Kent’s
Cavern; 8- Dainton; 9- Milber Down; 10- Blackbury Camp; 11- Hengistbury
Head; 12- Allard’s Quarry; 13- Hod Hill; 14- Gussage All Saints; 15- Maiden
Castle; 16- Danebury; 17- Zionshill; 18- Easton Lane and Winnall Down; 19-
Kennel Farm; 20- Highstead; 21- Sunset Caravan Park; 22- Thanet Earth; 23-
East Wear Bay. Author’s illustration.
9. Plan of the areas of excavation at Maiden Castle with areas that produced textile tools
highlighted, from Sharples 1991.
0. The diameter and weight of spindle whorls from all sites, separated by county.
6
7
17
21
24
26
27
30
32
33
35
38
39
40
68
71
73
78
103
13
1. The weights and depths of loom weights from all sites, separated by county.
2. The diameter and weight of spindle whorls from fortified sites.
3. The weights and depths of loom weights from fortified sites in all counties.
4. The diameter and weight of spindle whorls from enclosed sites.
5. The weights and depths of loom weights from enclosed sites in all counties.
6. The diameter and weight of spindle whorls from open sites.
7. The diameter and weight of spindle whorls from other sites.
8. Map of tool types found across all site types in the South West. 1.: Carn Euny. 2.:
Bodrifty. 3.: Trevelgue Head. 4.: Trevisker. 5.: The Rumps. 6.: Mount Batten. 7.:
Dainton. 8.: Kent’s Cavern. 9.: Milber Down. 10.: Blackbury Castle. Author’s
illustration.
9. Map of tool types found across all site types in the Central South. 1.: Hengistbury Head.
2.: Allard’s Quarry. 3.: Hod Hill. 4.: Gussage All Saints. 5.: Maiden Castle. 6.:
Danebury. 7.: Zions Hill Farm. 8.: Easton Lane. 9.: Winnall Down. 10.: Kennel
Farm. Author’s illustration
0. Danebury chalk and baked clay loom weights. Author’s illustration.
1. Map of tool types found across all site types in the South East. 1.: East Wear Bay. 2.:
Sunset Caravan Park. 3.: Thanet Earth. 4.: Highstead. Author’s illustration
2. A map of the geography of Cornwall, with sites highlighted. After Scrivener et al. 1997.
3. A map of the geography of Devon, with sites highlighted. After Devon County Council
2020
4. A map of the geography of Dorset, with sites highlighted. After Thomas 2008.
5. A map of the geography of Hampshire, with sites highlighted. After Hampshire County
Council 2020.
6. A map of the geography of Kent, with sites highlighted. After English Heritage 2011.
7. Two examples of the forms of decoration present on Torbay spindle whorls, from the
Torbay Museum’s collections. After Silvester 1986.
8. A map of the Torbay region, with sites included in the study area highlighted. A - Kent’s
Cavern; B - Milber Down; C - Dainton. Author’s illustration.
9. Examples of common decoration on Iron Age weaving combs from Danebury. Left:
Comb number 3.238; centre: comb number 3.239; right: comb number 3.240.
After Cunliffe 1991.
118
121
125
128
132
134
137
139
144
148
155
157
160
162
163
164
165
175
177
179
14
List of Tables
Thread diameters and tension required for weaving according to spindle whorl weight and
corresponding number of threads per loom weight to achieve required tension,
after Mårtensson et al. 2009: 378.
Showing the evidence required by the various specialisation modes as set out by Costin
(1991) and van der Leeuw (1977)
Sites in this study with population estimates and number of textile tools
The minimum, maximum, and average measurements of Bodrifty spindle whorls.
The minimum, maximum, and average measurements of Carn Euny spindle whorls.
The minimum, maximum, and average measurements of the Rumps spindle whorls.
The minimum, maximum, and average measurements of the Rumps loom weights.
The minimum, maximum, and average measurements of Trevelgue Head spindle whorls.
The minimum, maximum, and average measurements of Trevelgue Head loom weights.
. The minimum, maximum, and average measurements of Treviker spindle whorls.
. The minimum, maximum, and average measurements of Trevisker loom weights.
. The minimum, maximum, and average measurements of Blackbury Camp loom weights
. The minimum, maximum, and average measurements of Dainton spindle whorls
. The measurements of Dainton loom weights
. The minimum, maximum, and average measurements of Kent’s Cavern spindle whorls
. The minimum, maximum, and average measurements of Milber Down spindle whorls
. The measurements of Milber Down loom weights
. The minimum, maximum, and average measurements of Kent’s Cavern spindle whorls
. The minimum, maximum, and average measurements of Allard’s Quarry spindle whorls
. The measurements of Allard’s Quarry loom weights
. The minimum, maximum, and average measurements of Gussage All Saints spindle
whorls
. The minimum, maximum, and average measurements of Gussage All Saints loom weights
. The minimum, maximum, and average measurements of Hengistbury Head spindle whorls
. The minimum, maximum, and average measurements of Hengistbury Head loom weights
. The minimum, maximum, and average measurements of Hod Hill spindle whorls
. The minimum, maximum, and average measurements of Hod Hill loom weights
. The minimum, maximum, and average measurements of Maiden Castle spindle whorls
. The minimum, maximum, and average measurements of Maiden Castle loom weights
. The minimum, maximum, and average measurements of Danebury spindle whorls
. The minimum, maximum, and average measurements of Danebury loom weights
. The minimum, maximum, and average measurements of Easton Lane loom weights
. The minimum, maximum, and average measurements of Kennel Farm loom weights
. The minimum, maximum, and average measurements of Winnall Down loom weights
. The minimum, maximum, and average measurements of Zions Hill Farm loom weights
. The minimum, maximum, and average measurements of East Wear Bay spindle whorls
. The minimum, maximum, and average measurements of East Wear Bay loom weights
. The measurements of Highstead spindle whorls
. The measurements of Highstead loom weights
. The minimum, maximum, and average measurements of Sunset Caravan Park spindle
whorls
. The minimum, maximum, and average measurements of Thanet Earth loom weights
. Measurements of spindle whorl clusters from all sites and counties, the most likely thread
thickness they would produce, and the general speed of their spin.
. Measurements of loom weight clusters from all sites and counties, the thread thickness best
suited for weaving
. Measurements of spindle whorl clusters from fortified settlements, the most likely thread
28
51
79
82
83
85
85
86
86
87
88
89
90
91
92
93
93
94
95
96
97
97
99
99
100
101
102
103
105
105
15
thickness they would produce, and the general speed of their spin.
. Measurements of loom weight clusters from fortified sites, the thread thickness best suited
for weaving
. Measurements of spindle whorl clusters from enclosed settlements, the most likely thread
thickness they would produce, and the general speed of their spin.
. Measurements of loom weight clusters from enclosed sites compared to the thread
diameters best suited for weaving
. Measurements of spindle whorl clusters from open settlements, the most likely thread
thickness they would produce, and the general speed of their spin.
. Measurements of spindle whorl clusters from other settlements, the most likely thread
thickness they would produce, and the general speed of their spin.
107
108
109
111
112
112
114
114
115
117
120
123
127
130
133
135
138
140
16
17
List of Appendices
Appendix 1. South West sites in this study 208
Appendix 2. South Central sites in this study 213
Appendix 3. South East sites in this study 218
Appendix 4. Textile tool data form 221
Appendix 5. Cornwall artifact data 222
Appendix 6. Devon artifact data 226
Appendix 7. Dorset artifact data 229
Appendix 8. Hampshire artifact data 237
Appendix 9. Kent artifact data 242
Appendix 10. Cornwall environmental data 244
Appendix 11. Devon environmental data 245
Appendix 12. Dorset environmental data 246
Appendix 13. Hampshire environmental data 247
Appendix 14. Kent environmental data 248
1
11. Introduction
As a common saying goes, ‘a craftsman is only as good as his tools’ while ‘a bad craftsman
blames his tools’. In both proverbs, there is a subtle link between skill, tools, and what they
produce; be that a good product or a bad reputation. The need of a crafts worker for
functional tools is universal; it is key in modern manufacturing, affects what can and cannot
be produced, and adapts to new technologies and production methods. This is best reflected
by a few crafts in the archaeological record, most notably lithics, pottery, and metalworking.
But what about the textile craft? Textiles are rarely considered in British prehistoric sites
unless fragments of fabrics are discovered during excavations or a particular study is
conducted focusing on the evidence for textiles (Bradley 1978; Bradley et al 2016; Cunliffe
2005; Hill 2011), despite the importance of textiles as a product and textile production as a
craft. There are a number of difficulties in conducting research on this topic, especially in the
temperate climates of Britain and north-west Europe. However, I would argue that the
importance of textiles and textile production in a society’s structure and economy is great
enough to tackle these challenges in order to develop a more detailed and thorough
understanding of ancient societies, particularly those that have left no written records.
The organisation of textile craft in Bronze and Iron Age Mediterranean societies has been
used as an important factor in determining the location of production centres. The use of
specialist textiles in contexts such as religious sanctuaries shows the importance of this craft
and the status of its workers in their societies (Andersson Strand and Nosch 2015; Barber
1992; Burke 2005; Marín-Aguilera et al 2019). No similar work using textile production has
been conducted on British contexts of the Bronze or Iron Age, despite there being no lack of
evidence at these sites. My thesis is the first study to use the methods developed in the field
of archaeological textiles and textile tool research to analyse the tools for spinning and
weaving from various settlements along the British south coast. The data from these tools and
the distribution between different forms of settlement provides the first details of how textile
production was organised, how this corresponds with the current theories on settlement
organisation and use in Britain and north-west Europe, and how this compares to craft and
settlement organisation in Iron Age Mediterranean contexts.
2Textiles and textile production
Textile production was a craft that supplied communities with one of their fundamental
materials used in many contexts, from furnishings and clothing to cordage and netting. In
addition to their utilitarian function, textiles also provided a medium for individuals to
display social status, personal wealth, political allegiances, or cultural backgrounds through
the use of decoration (e.g. beading, embroidery, brooches and pins, etc.), dye colours, and
even weave patterns (still displayed in Scotland’s use of tartan to denote clan membership).
As spinners and weavers contributed critical elements to almost every section of society, they
and their craft arguably held a position of importance in society. This is reflected in literary
evidence from Ancient Greece, where textiles could form part of an expensive gift (e.g.
dowries, diplomatic gifts, etc.) and particularly skilled textile workers were sought-after
slaves (Burke 2016; Marín-Aguilera 2019).
In many countries up to the present day, this is one of the few crafts where women made up
the majority of workers (Marín-Aguilera 2019), so the study of textiles and textile production
allows archaeologists to gain insight into the status, economic contributions, and influence of
skilled women in ancient societies. Perhaps because of this historic link with women,
producing textiles has long been considered a relatively unimportant and uninteresting chore
by antiquarians and early archaeologists (Arnold 1985; Brumfiel and Earle 1987; Childe
1950; Marín-Aguilera 2019: 231-2). Occasionally, these academics would study the textiles
and garments of high-status burials, yet their interest would be confined to judging the wealth
and power associated with a buried individual (Barber 1991). Textile production was
generally considered ‘women’s work’ and the subject and artefacts associated with the craft
were rarely discussed in any detail until the late 20th century (Barber 1994; Marín-Aguilera
2019: 231-2). This dismissal ignored the considerable investment of time, energy, and raw
resources required to produce such fine and high-status materials.
Various experiments have demonstrated how serious and taxing such investments could be.
The experimental recreation of a 90 square meter sail cloth from scratch is a useful indicator
of what is involved. (Andersson 2003). These experiments demonstrated that to create a sail
of 90 square meters, a reasonably experienced spinner could spend 15,000 hours (equivalent
of 2 and ½ years of modern working hours) preparing and spinning the wool (Andersson
2003: 46) and a reasonably experienced weaver 3200 hours weaving it, reaching a total of 4
and ½ years of modern working hours (Bender Jørgensen 2012: 177-8). These hours are
3calculated for a single textile worker and does not take into account time gathering the wool,
finishing the textile, or sewing it into a sail.
Although using evidence from Viking Age textile work, the technology (weaving patterns,
spinning methods, textile tools, etc.) and raw materials used are comparable to those from the
Bronze and Iron Ages. Moreover, the remains of a number of looms spanning 3 to 4m from
Bronze Age contexts, indicate that large textiles could be produced already then
(Belanová-Štolcová and Grömer 2010: 17). Furthermore, Bronze Age textiles have been
found with areas where weft threads cross over one another, demonstrating that these were
worked by two or more people at a time, passing the weft threads between each other when
they reached the end of their section of the loom (Barber 1991). When considered alongside
the time required for other household activities that are traditionally considered ‘women’s
work’ (e.g. childcare and cooking), it is hard to imagine how this sort of textile work could
be done by anyone who was not a specialist and therefore, not free to focus on spinning and
weaving. The dismissal of textile work as a field not worth studying also ignored the
important detail that textile production, although widespread, is not an easily mastered
activity.
Ancient textiles were made from animal or plant fibres that had to be grown, harvested or
shorn from the animal, processed (e.g. washing and combing fleece; softening and splitting
the plant fibres), spun into threads, and finally woven into a textile before it could be cut or
sewn into the finished textile. The stages of processing fibres into woven cloth, the textile
chaîne opératoire, could be long and complicated, particularly when using plant fibres
(Barber 1991). Each of these stages required training, experience and the correct tools to
accomplish, which are the basic elements required for any productive activity to become a
craft activity and a worker to become a craft specialist (Brumfiel and Earle 1987; Costin
1991; Hagstrum 1985).
‘Craft specialist’ is a term that describes a person able to devote their time to their chosen
craft; chefs, blacksmiths, jewellers, and textile workers can all be craft specialists. Craft
specialisation can be full- or part-time, depending on consumer demand, resource availability,
or other demands on the craft worker’s time (e.g. harvest time, religious duties, or other
social responsibilities). By focusing on their craftwork, specialists have the potential to
become particularly skilled in their work and produce high quality items. On the other hand,
a craft worker who produces large quantities of lower-quality items can still be called a
4specialist; after all, the primary requirement of the label ‘specialist’ is the time these workers
are able to devote to craft production (Costin 1991). As previously discussed, textiles were
historically important and valuable assets, which often led high-ranking members of society
to control the organisation of textile workers and specialists (e.g. where they worked, what
they produced, how it was traded or distributed, etc.) in the same way that other crafts were
controlled (for the organisation of pottery producers see Rice 1981; for metal workers, see
Costin 1991).
Evidence from various prehistoric cultures across the Mediterranean has demonstrated the
pattern of specialist textile production falling under the direct control of elite administrative
or religious institutions or high-ranking individuals (see Marín-Aguilera 2019 for specialist
textile production in 5th century BC Iberian religious sanctuaries; see Hruby 2013 for
specialist production at Nestor Palace in Mycenaean Greece). Many crafts, like smithing and
pottery making, were often organised in some form of workshops; this allowed for both
easier control over craft specialists by keeping them in a single location, often near their
patron, and easier production by centralising where to send raw materials and train new craft
specialists (Brumfiel and Earle 1987; Costin 1991). It is interesting to note that there is a
notion among archaeologists that specialist craft production must be separated from domestic
activities (Allison 1999, 8; Hendon 1996, 49; González et al. 2007), as if workshops are the
only space in which specialist production can take place (Marín-Aguilera 2019: 230-1).
Spindles and upright looms do not require much space and can easily be stored or placed out
of the way, so specialist textile workers may not have needed a workshop to store their
equipment (Marín-Aguilera 2019: 231). Certainly, there are examples of Mediterranean sites
that have hundreds of spindle whorls and loom weights all made to the same shape,
dimensions, and weight, which are clustered not in separate workshops but found in the main
domestic structure (Marín-Aguilera 2019; Marín-Aguilera et al 2019). These are clearly
centres of specialist textile production that do not require separate areas for the textile
workers, perhaps because of how little space textile production requires. So the study of
textile production can highlight patterns of how different crafts and craft workers may have
been organised in ancient societies..
There are some issues with the theoretical frameworks and methodologies with which craft
production and organisation is typically studied. The methods to identify areas of specialist
production were developed for use on crafts that left physical remains, either debris from
stages in the crafting process or archaeological features. For example, Tykot et al. (2006)
5identified areas of obsidian craft specialisation in Sennixeddu (Sardinia) by examining the
knapping debris left behind and calculating how much was produced during the Late
Neolithic. Similarly, Cunliffe (1994, 1995) used pieces of crucibles, hammer-scale, and metal
slag at Danebury to identify this as a major metalworking site during the Iron Age. On the
other hand, textile production does not produce debris that survives deposition or easily
identifiable archaeological features. Organic material quickly disintegrates and Iron Age
looms were either simple lean-to frames or constructed from two parallel posts; these are near
impossible to accurately recognise as features of textile production. This leads to the wider
issue of preservation, particularly in areas of temperate climate like Britain, which is the
geographical focus of my thesis.
Organic materials, particularly thin and fragile ones like textiles, rarely survive long in
damp temperate environments before rotting. Occasionally, fragments of textiles may be
preserved in anaerobic conditions that prevent bacteria from deteriorating the material. Peat
bogs and waterlogged lake beds have provided excellent conditions for the preservation of
ancient textiles from Switzerland (see the finds from Hallstatt [Grömer 2005]), Denmark
(Mannering et al 2010), and Britain (see Must Farm, Knight et al 2019). Otherwise, hints of
the form and appearance of ancient textiles may be preserved as impressions on clay (where
it was pressed against a pat before firing) or metal (where rust pressed against the fabric
before it rotted away) (Barber 1991; Ferrero 2014). Because it is rare to find ancient textiles
outside these conditions in temperate regions, we must look elsewhere for evidence of textile
production.
Textile tools
In areas where textiles do not survive, textile tools are some of the best available evidence for
what was produced. The term ‘textile tool’ can cover a number of items used at various
points in the textile chaîne opératoire (for examples of each tool see figure 1, below): combs
to process raw wool, hackles and wooden scutches to process plant fibres, spindle whorls for
spinning these fibres, loom weights for weaving on a warp-weighted loom, and weaving
combs. The materials and uses of each of these tools is expanded on below.
6Figure 1. Top left, Iron Age weaving comb from Danebury (Cunliffe and Poole 1991); bottom left, Viking Age
wool comb from Buiston Crannog (Ashby 2009: fig 2); top right, Iron Age spindle whorl from Kent’s Cavern
(Silvester 1986); bottom right: Iron Age loom weights from Danebury (Cunliffe and Poole 1991).
The shape of some types of textile tools can vary throughout time or across large regions.
Loom weights in particular demonstrate a great deal of variation between the Bronze and Iron
Ages in Britain, as well as between different cultures. The weights of the British Bronze Age
typically have a single hole in the centre and are cylindrical , while those from the Iron Age
are triangular and have a hole through each corner (see figure 2). In comparison, loom
weights from the Bronze Age Aegean have a variety of shapes (see figure 2). The most
common material for loom weights is locally-sourced clay that is moulded to the desired
shape and fired or baked; however, it is not the pliability of this material that causes the
variety of loom weight shapes across regions and time periods. Loom weight shapes appear
to depend on the society in which these are made and the weaving culture, technology, and
techniques that are predominant in these areas (Mårtensson et al 2009). For example, the
disc-shaped loom weights of Troy are light enough to work with very fine threads and thin
enough to have many loom weights on the same loom, significantly raising the textile’s
thread count and producing a very fine textile (Mårtensson et al 2009).
7Figure 2. Typical shapes of loom weights from the Bronze Age Aegean (left) and Iron Age Britain (right).
Author’s illustration, after Mårtensson et al 2009.
Differences in the shape of spindle whorls can also be very pronounced, although
sometimes this can be explained by how these tools are made. Whorls made from reused
pottery sherds (as is common in the British Iron Age) will, per force, have a very different
shape from those made of clay or stone. On the other hand, some sites demonstrate very little
variation in whorl shape and material. Marín-Aguilera (2019) noted that the spindle whorls of
Cancho Roano, an Iron Age site in Iberia, shared the exact same shape, dimensions, and
weight; on closer inspection, she found signs of impressions on the whorls that proved these
tools were made in moulds, so that the spinners had completely standardised tools. Tools
made to set standards are rarely seen in Britain or north-west Europe during the Bronze or
Iron Ages despite being highly beneficial for specialist textile workers or a workshop setting.
Standardised spindle whorls could reduce the amount of time needed for setting up or
working on looms by producing threads that require similar weights. This would allow less
experienced workers to learn the necessary set-up procedures more quickly, which could give
these learners more time to concentrate on the process of weaving itself.
Over recent years a methodological framework for analysing spindle whorls and loom
weights has been developed that allows us to determine the sort of yarn or fabric the tools are
best suited to make. In particular, the arguments and experimental data produced by Linda
Mårtensson, Marie-Louise Nosch, and Eva Andersson Strand (2009) at the Centre of Textile
Research (CTR), Andersson Strand and Nosch (2006) at the Textiles, Tools, Texts and
Contexts Research Program (TTTC) and by Karina Grömer’s work on the Hallstatt textiles
(2005), have been influential in establishing a new platform for the analysis of textile tools.
By studying the materials, make, and standardisation of tools, as well as their links to various
contexts, it is possible to gain culturally specific insights into how textile works/ers were
8organised, how such organisation varied from settlement to settlement or region to region, as
well as indicators into the importance of textile production in the larger economy of a
particular society.
Another common method of identifying specialist production is to find evidence of surplus
production, typically in the form of larger quantities of craft debris or tools than would be
needed to supply a household (Costin 1991; Gleba 2007; Marín-Aguilera 2019: 229-230).
Simply finding evidence of any production, like a single spindle whorl or loom weight, does
not mean there was a dedicated craft worker on site; it only suggests some form of textile
work was present in that household or settlement. Excess production (beyond what is needed
for an individual household) would allow craft workers, or those they work for, to supply
items for local and foreign exchange and control how their products were distributed
(Brumfiel and Earle 1987; Costin 1991); this is the most widely accepted factor used to
distinguish household production from specialist (often workshop-based) craft activities
(Costin 1991; van der Leeuw 1977; Marín-Aguilera 2019). Usually, for textile production to
be deemed an economic rather than a subsistence activity, a site must have a large amount of
textile tools; like Bronze Age Malia, which produced over 600 loom weights (Cutler et al.
2013), or Gordion in Turkey, where excavations produced over a thousand spindle whorls
and loom weights from several buildings (Burke 2005). However, few sites in north-west
Europe and Britain produce such high numbers of textile tools during the Iron Age.
The British Iron Age
I chose to focus on the Iron Age in Britain for two main reasons: first, because little research
has been conducted on its textile tools, despite the numbers that have survived; and second,
because new forms of settlement develop during the Iron Age, the social and economic roles
of which are not fully understood. The Iron Age in Britain spans from 800 BC to AD 43 and
is recognised as a period of great social change and development in Britain. The increasing
use of iron resulted in stronger, longer-lasting tools that allowed farmers to expand their
fields, which could then support a larger population (Bradley et al. 2016). Trade routes with
northern Europe and the Mediterranean are disrupted and altered as demands for resources
shift (Bradley et al. 2016; Fernández-Götz 2018). Even settlement patterns alter. Large,
fortified hillforts and oppidas are constructed, which appear to function in various ways
around the country. Some are occupied for long periods and appear to function as centres of
9trade and production (e.g. Hengistbury Head, Cunliffe 1978; Maiden Castle, Sharples 1991;
and Danebury, Cunliffe 1984a), some appear to be seasonally occupied - possibly during the
winter months when the lower field and hinterlands are unsuitable for agriculture - and others
lack any evidence of occupation and appear to function as refuges in times of conflict
(Cunliffe 1991, 2005; Hill 2011).
In an attempt to explain these major shifts in society and economy, many academics
compared Iron Age Britain and north-west Europe to the settlements and societies found in
the Mediterranean (Collis 2014). These had strict social hierarchies and centralised means of
production, evident in both the archaeological features and the written records of these
cultures that academics believed more ‘barbaric’ civilisations in north-western Europe were
poorly copying (Collis 2014). However, this comparison was fundamentally flawed. Unlike
cultures in the Mediterranean Bronze Age, the society of Iron Age Britain had no royal
residences or wealthy, powerful elites that can be clearly identified by the remains of an
individual household or by unique burial practices (Barrett 2000; Hill 2011; Sharples 2005,
2007). Hillforts and brochs are the closest that British archaeologists could come to
identifying possible elite residences (Cunliffe 1984a, 1984b; Haselgrove 1982) and these
interpretations were debated for a time (for a brief but clear overview, see Hill 2011:
243-254). Domestic structures in the Mediterranean could vary in size and complexity based
on the social rank of its household; in comparison, roundhouses were the only form of
domestic building during the British Iron Age. It is difficult, near impossible, to identify
social rank from structural remains of roundhouses alone (Hill 2011). The structure and
materials are the same across different settlement types and there are no structural dividers
within the space that could denote areas of higher social status (Hill 2011). The common
consensus among archaeologists now is that Britain and north-west Europe did not mimic the
contemporary societies of the Mediterranean, but developed parallel to them instead (Collis
2014). It is possible to use the archaeological evidence from the Mediterranean to inform our
understanding of similar features in temperate Europe, although we must be careful to respect
the differences between the two.
Another issue is that studies on the British Iron Age rarely discuss textile production in
detail, often only mentioning it in passing. It is assumed that textile production in the British
Iron Age often took place in domestic spaces - roundhouses - rather than a separate,
dedicated weaving or spinning area (Cunliffe 1991, 2005; DeRoche 1995). There is plenty of
10
evidence that other crafts took place in British domestic buildings during this period. For
example, the hearths of Hut G in Bodrifty and Hut DB2 at Maiden Castle doubled as forges,
with crucibles and slag found alongside domestic pottery, saddle querns, and spindle whorls
(Dudley 1956: 10-11; Wheeler 1943), demonstrating that craft work occurred alongside
domestic activity. Comparing the evidence of textile production with the evidence of other
crafts (e.g. pottery production, metalworking, lithics, etc.) as well as foreign goods present at
various settlements can develop a broader understanding of the status, organisation, and
contribution of textile workers. This method can provide better insight into what was
produced at each site and how intense that production was, which can then highlight any
similarities or differences in the scale of textile production at these settlements.
As previously discussed, the importance of textile production and the value of its producers
was significant in prehistoric times. With the advancements made in the study of textile tools
and craft organisation, it is now possible to study the British Iron Age and begin to draw
comparisons with its European counterparts. The British Iron Age provides tools that have
not been viewed as sources of useful information, resulting in a large collection of data never
previously studied. In addition to this, there is evidence that Iron Age society in Britain and
Ireland (as well as parts of north-west France) was subtly different from the rest of
north-west Europe. The construction of dwellings in the British Isles are round, rather than
the rectangular structures found in the rest of north-west Europe (Pope 2007). Similarly, the
rectangular houses of Europe frequently had half of the building built to house livestock
owned by the household; in comparison, British roundhouses do not have similar spaces and
used stockade enclosures to hold livestock overnight (Sørensen 2007; Pope 2007). While it
could be argued that these are superficial differences, these examples indicate a deeper divide
between organisation of activities and settlements in Europe and Britain. It is less clear
whether there is a difference in the organisation of craft activities and workers between
Britain and Europe. The study of textile tools, as well as the textile these are best suited to
produce, and the distribution of such tools across various types of settlement has the potential
to provide an excellent insight into the variation of craft organisation that existed between
Britain and north-west Europe.
11
Considering all these issues, my research question is:
“What forms of organisation and specialisation in textile production can be identified from
a close study of Iron Age textile tools taken from various forms of settlement along the
British south coast?”
By mapping the regional distribution of textile tools in a variety of settlements (farmsteads,
villages, hillforts, etc.) and comparing this distribution to a more detailed analysis of the
textile tools themselves, I will be able to highlight areas or settlements that had specialised
production. Such specialisation could be achieved by a settlement working on one part of the
chaîne opératoire (e.g. spinning) or by textile workers using tools with a restricted range of
weights and dimensions in order to limit the thickness of the yarns and fabrics produced. And
thus working within a notion of standardisation.
Research aims
In pursuit of this question I designed three research aims, as my route map. These aims
focused on the main areas of research in this thesis: the tools themselves and what these are
capable of producing; how these tools reflect the organisation of craft workers and the
economy of various sites; and how the distribution of tools across various forms of settlement
(i.e.: fortified settlements vs. farmsteads, etc.) reflect forms of textile production and craft
worker organisation. Thus, the aims are:
● To record and compare the dimensions and weights of spindle whorls and loom
weights from five sites in each of the counties of the British southern coastal area (Cornwall,
Devon, Dorset, Hampshire, and Kent)
● To determine the most likely characteristics of the yarns and fabrics produced by
these tools and within each county; to discuss the organisation of textile production (or lack
thereof), and the possible importance that this craft (and its craftworkers) held in supplying
communities or supplementing trade
● To compare the number of spindle whorls and loom weights found in each site across
the counties, to determine whether these textile tools were confined to certain areas or types
of settlement; to compare the results regarding evidence of textile production and its
organisation between different counties, in order to assess wider patterns and variations
12
Time restrictions made it impossible to expand the study area further than Cornwall,
Devon, Dorset, Hampshire, and Kent, limiting this study to the southern coast of Britain.
Unfortunately, East and West Sussex were excluded from this study due to the lack of useful
data; the settlements that produced five or more textile tools were excavated before the
Second World War and those tools could not be located in any museum collections. This
represents the first study in Britain to conduct a large-scale, thorough analysis of textile tools
and the distribution across different types of sites.
The data will be used to analyse changes in these textile tools  between settlement types
and contexts and how these variations may demonstrate differences in the organisation of
craftworkers and the character of their textile production. This will provide the basis of a
methodology for others working on Iron Age textile production, social organisation, or craft
workers in general in Britain. Furthermore, the research will highlight the importance of
textiles and the spinner/weavers’ contribution to the Iron Age economy and society;
something often overlooked in broader archaeological literature.
Thesis structure
The chapters of this thesis are listed here, along with a brief summary of each chapter’s
contents.
Chapter 1. Introduction
The introduction lays out the basic premise of this thesis - the research question, the thesis
aims and objectives, and the reasons for pursuing such research. It introduces the basic
elements of the coming report including the tools that will be the focus of analysis, the
importance of craft production and specialisation in society, and the character of Iron Age in
Britain. This chapter also provides an outline and summary of the thesis.
Chapter 2. Previous work - Textile production, tools, and methods of analysis
The second chapter focuses on the previous work surrounding textile production, textile
tools, methods of analysis, determining craft production and organisation, the British Iron
Age, and patterns of settlement and economy in this period. This large range of topics is split
into three main sections in order to better explore and evaluate the information contained in
each. The first section covers issues of textile production, such as an overview of the textile
chaîne opératoire. The discussion explores the raw materials used in textile production, the
13
natural range of these materials, and methods of processing prior to further use in order to
establish what was available to the Iron Age textile worker in Britain. Following this are
definitions and descriptions of various methods of spinning, weaving, and finishing textiles,
the archaeological evidence that remains, and the likely range of each of these methods.
Examples used are from the Hallstatt and Dürrnberg salt mines in Austria, as well as other
areas of Europe. During the discussion of each stage of the chaîne opératoire there are
definitions and descriptions of the tools used at these points, the materials used to make these
tools, and the methods or theories of how these tools were used. Archaeological evidence for
each process conducted in the chaîne opératoire, as well as the materials and tools used in
these stages, is provided in order to develop a more rounded understanding of the techniques
used in prehistory. These are presented alongside previous and current theories regarding the
widespread use of each of the materials, processes, and tools. This first part of the Previous
Work concludes with remarks on which stages of the chaîne opératoire, materials, and tools
will be focused on in this thesis, given the archaeological evidence and theories discussed
above.
The second part of this chapter focuses on modes of textile production and the workers who
made these textiles. This section provides an overview of the evidence surrounding the
production of textiles and the organisation of its producers. This includes direct evidence of
textile workers’ skills from archaeological remains, as well as explanations of the
preservation conditions required for such evidence to survive. Various modern investigative
techniques are discussed that are able to determine the range of skills involved in the original
textile’s production. The examples used are primarily from the Hallstatt and Dürrnberg salt
mines, among others. Following this overview of textile production is a discussion of the
various definitions of specialisation, the evidence required for each definition, and its
importance in determining craft production and organisation. This section concludes with an
overview of the current evidence available from Iron Age Britain and the current theories of
textile production and specialisation in Iron Age Britain.
The third and final part of the Previous Work chapter looks at evidence for the structure of
settlements, society, and economy of Britain in the Iron Age. This includes definitions and
discussion of social development in Britain and Europe during the Bronze/Iron Age transition
and an overview of the differences between Iron Age settlements, trade, and social structure
and those in the Bronze Age. The development of society and its crafts between these periods
is important as both undergo significant changes; understanding these developments can help
highlight patterns of craft production and organisation that may not be immediately obvious
14
from the Iron Age evidence alone. This is followed by a discussion of previous theories
surrounding the possible influence that the Mediterranean had on the developments in
north-west Europe and Britain, the evidence used to support these theories, and current
theories on the Iron Age development in Europe. These are contrasted by a discussion and
analysis of the evidence available that settlements in the British Iron Age could support craft
specialisation, including methods of calculating population size. This is further explored with
a brief discussion of the structure and organisation of British Iron Age settlements, including
the roundhouse, and any differences or similarities these settlements have with the continent.
This section also includes an overview of the evidence of trade and occupation from these
sites, that are found in the areas covered in this thesis. These areas are labelled South West
(for Cornwall and Devon), Central South (for Dorset and Hampshire), and South East (for
East and West Sussex and Kent). Finally, the Previous Work chapter concludes with remarks
on the most likely uses for settlement types and the differences between sites of similar
construction across the study area, as well as evidence of trade and contact with the continent
at various settlement types
Chapter 3. Methodology
The methodology of this thesis provides a summary of data collection and analysis
techniques used in this thesis. This chapter provides an overview of issues encountered
during data collection, including incomplete tool assemblages from sites with poor
preservation, lack of context or detail from site reports, missing tools and lost collections, and
tools with (often severe) damage. It details how these various issues were resolved or worked
around. There follows a discussion on the parameters for sites included in this thesis, the
number of sites per county, and the size of the study area were determined. This includes a
brief overview of the sites per county, as well as each site’s minimum population size and
number of textile tools included in this thesis. This provides a more nuanced understanding
of the textile production at each site as well as the demand for textiles to supply each
community. Following from this is an overview of which tools were used in this study, what
techniques used to gather the required data, and the calculations used to analyse this raw data.
Finally, there is a concise runthrough of how the conclusions of these calculations were
presented and discussed.
Chapter 4. Data - Cornwall, Devon, Dorset, Hampshire, and Kent
This chapter presents the main portion of this thesis and focuses on the presentation and
discussion of the data gathered from each site in these counties. Each site is introduced with
an overview of the nature of the site, its location, its date of occupation during the Iron Age,
15
and its excavation. There is also a summary of the efficiency, thoroughness, and detail of
excavation methods and the excavation reports, as well as where these tools were stored, and
a summary of the variety of finds found at each site. This includes evidence of occupation,
other craft production (e.g. metal working, pottery production, etc.), and local or
long-distance trade at each site. Finally, the number of textile tools from each site, as well as
the minimum, average, and maximum weights and dimensions of these tools are presented.
Chapter 5. Discussion - Tool analysis
Similar to the Previous Work chapter, the Discussion is split into three main parts, the first
focusing on the analysis of the tools recorded in the previous chapter. The first analysis is a
comparison of the tool weights and dimensions from all sites, in order to determine common
tool patterns in each county and how these vary in different counties. This is followed by a
comparison of tool weights and dimensions from sites of the same settlement type (e.g. all
spindle whorls from fortified sites, all spindle whorls from enclosed sites, etc.) to determine
tool patterns specific to certain site types. Such data is used to identify the most common tool
weights and dimensions and estimate the most likely threads and textiles that these tools
could produce from each site. By identifying these patterns, it is also possible to identify any
sites or counties producing unusual yarns or textiles (e.g. extremely thin or thick yarns or
textiles, etc.). This section is also where any recorded data that could not be used as a textile
tool due to its weight or dimensions is identified and discussed.
The second section of the Discussion focuses on the distribution of the textile tools
included in this study. This includes the identification of patterns of distribution of each tool
type according to region, the nature of each site, and what was most likely produced at each
site. Several possible explanations for these patterns are discussed in this section, including
land use and available raw materials for textile tools. This section also identifies areas of
specialist production and evidence of textile worker organisation; this discussion also takes
into account whether the population at each site could support a full- or part-time craft
worker. The final part of this section focuses on identifying possible centres of production
and the average nature of textile work in smaller sites in each county. This brings together the
evidence for sites that otherwise might be overlooked in favour of the larger sites or areas of
more complex production, creating a more detailed picture of the variety of production
occurring across the study area.
The final part of the Discussion section looks at the wider context of this study and the
findings thereof in relation to current work on textile tools and prehistoric crafts. This is
accomplished by comparing the evidence of specialist workers and/or centres of production
16
in the study area to selected sites in Europe, such as Hallstatt, Dürrenburg, and Cancho
Roano. Doing so allows a direct comparison of the British evidence against known centres of
production/specialisation and different modes of production in Europe. This section also
discusses whether any sites included in this study could fit any form of craft specialisation as
defined by Costin and Van der Leeuw. Comparing the evidence from British Iron Age sites to
these forms of definitions would demonstrate whether the parameters of these definitions are
well suited to textile craft evidence. Following this is further comparison of evidence for the
skill of British and European textile workers based on the textile tools available (Hallstatt,
Dürrenburg, and Cancho Roano); as skill is difficult to determine from tools alone, this also
connects with the discussion of textile production as a public activity in Britain and Europe
based on tool evidence (e.g. decorated tools and communal spinning) at Kent’s Cavern and
Danebury, compared to Cancho Roano and Dürrenburg. This leads to a discussion of the
evidence for, and level of, standardisation in textile production, including the number of
pottery sherd spindle whorls and range of thread diameters. This evidence is further
compared to that from European sites (such as Hallstatt, Dürrenburg, and Cancho Roano).
Chapter 6. Conclusion
This chapter presents the concluding remarks of this thesis, including a summary of the
aims of this thesis and brief discussion of how well this study met those aims. Following this
is an overview of what this method of tool analysis is able to achieve, including identifying
textile traditions that produce unique textiles (e.g. Devon and Cornwall), regional specialist
textile producers, areas specialising in certain raw materials (e.g. wool instead of plant fibre),
and possible specialist textile traders. There is a brief discussion of the issues with this
analysis method, including the requirement for various settlement types with multiple textile
tools, the difficulties in gathering data from excavation reports that did not record the
dimensions, weights, or total number of tools, and the issue of lost collections. Following this
is a discussion of further work that could benefit or improve this method of analysis, such as
repeating techniques in different British counties or European countries, adding more tools to
this method of analysis (e.g. weaving combs, needles, etc.), and constructing a database with
textile tool weights, dimensions, site type, original location and context, and excavation
reports.
17
2. Previous Work
Textile fibres
A number of tools are used during the complex process of creating textiles, from gathering
and processing the raw materials to finishing the woven fabric. The tools used in the early
part of the textile production process, or chaîne opératoire, are very different depending on
whether the raw material is from a plant or animal. It is unclear when certain processing
techniques (such as retting, explained below) developed. Cultures across different periods
and regions have always shared some stages of the chaîne opératoire, such as gathering raw
materials, processing these, spinning, and weaving (Andersson Strand 2012). Loom weights
and spindle whorls vary very little from Viking Age Denmark to Pharaonic Egypt and textile
fragments also demonstrate that certain weaves, such as tabby (explained below), remained
the same across these periods and distances (Andersson Strand 2012: 22). So it is possible to
piece together a rough outline of the textile chaîne opératoire (fig. 3 below) used during the
Iron Age across northern and western Europe (Belanová-Štolcová and Grömer 2010).
Figure 3. The basic textile chaîne opératoire with tools and resources, after Belanová-
Štolcová and Grömer 2010: 9 (author’s illustration).
The first phase of textile production is the production of the raw materials; rearing the
wool-bearing animals and/or growing the fibrous plants to get the best fibres possible. Sheep
18
capable of producing wool were introduced to continental Europe from the second
millennium BC onwards (Bender Jørgensen 1992; Bender Jørgensen and Rast-Eicher 2013,
2016; Sabatini et al 2019). It is most likely that sheep were initially domesticated in the
Neolithic for their meat, rather than their wool or milk (Sabatini et al 2019). While the skins
of goats and sheep were used to make clothing in this early period, it is not until the Near
Eastern Bronze Age that wool is widely used in textile production, c.3rd millenium BC
(Michel and Nosch 2010; Sabatini et al 2019).
By the 2nd millenium BC, evidence of written and iconographic sources demonstrate that
the Aegean developed equally large and complex centres of woollen textile production as
those seen further east (Burke 2012; Nosch 2014, 2015; Sabatini et al 2019). Unfortunately,
this information relies on written sources rather than archaeological evidence, thus limiting
the ability of archaeologists to identify such complex systems of textile production (Burke
2012; Sabatini et al 2019). Without these sources, archaeologists must rely on the presence
and distribution of textile tools, zooarchaeological assemblages with high proportions of
sheep/goats, and fragments of woollen textiles (Sabatini et al. 2019). However, the use of
woollen textiles may have changed as the quality of the wool improved over the course of the
Bronze Age and into the Iron Age (Rast-Eicher and Bender Jørgensen 2013). The sheep
breeds in the Bronze Age are often called primitive and there are a few primitive breeds in
Britain that still survive to this day: the Soay, Hebridean, Manx Loughtan, Shetland,
Broreray, and North Ronaldsey are all considered primitive native breeds (Ryder 1968,
1983). These surviving breeds can provide a lot of information on the natural behaviour of
the breed and the characteristics of its wool.
Primitive sheep are lighter and hardier than the modern breeds, which has allowed these
animals to live without human husbandry or intervention for roughly fifty years in places like
St Kilda (Clutton-Brock et al. 2004; Ryder 1983). There are three characteristics of the fleece
that contribute to the breed being labelled as primitive (Ryder 1983). Firstly, the fleece is
formed from two types of fibre, long, hair-like kemp that protects the soft, shorter, insulating
woollen undercoat from rain and dirt. These kemp hairs are not easily spun and very scratchy,
making them undesirable for spinners to work with and unsuitable for use in soft textiles.
Secondly, the fleece has areas that are predominantly kemp and areas that have more wool,
which means not all of the fleece can be used (or at least, not for the same purpose) and the
overall quality of the fleece depends on each individual animal, so it is difficult to judge how
much suitable wool a flock would produce. And thirdly, this fleece is naturally shed during
19
the spring and summer months; the wool can be gathered during this time by plucking it from
the sheep, a process known as ‘roo-ing’ (Ryder 1983). It is unclear when exactly farmers bred
sheep without kemp and with non-shedding fleece; iron shears from Iron Age contexts have
been found across continental Europe (Barber 1991), although Sabatini (et al 2019) notes that
shears could have been invented first and driven the selective breeding for longer,
non-shedding wool (see also Barber 1991). By the Iron Age, the sheep’s fleece had longer
wool with fewer areas of coarse hairs than its more primitive forebears and more
standardisation of wool quality between animals (Ryder 1983; Sabatini et al 2019).
These changes allowed for greater quantities of wool to be used in textiles and for higher
quality fabrics. Where preservation conditions allowed bones to survive, osteological
evidence demonstrates that sheep had a large presence across Iron Age Britain (Albarella
2007). Studies on sheep bones have shown it is possible to determine at what age the animals
were slaughtered and therefore what the most likely use for the herd was (Albarella 2007;
Grant 1984: 107; Greenfield 2014; Ryder 1968, 1983); mostly young animal bones indicate a
herd kept primarily for milk production, a mixture of ages demonstrates a meat-producing
herd, and mostly elderly animals indicates a wool-producing herd (Halstead 1998; Sabatini et
al. 2019). It appears herds filled multiple roles, providing wool, meat, and possibly milk to
the communities reliant on them (Sabatini et al. 2019) Furthermore, sheep are able to graze
on poor farming land and even areas that are unsuitable for crops, providing wool with an
advantage over plant fibres (Grömer 2016a: 63). Still, while it is possible to determine where
wool was being produced and the rough quantities being produced, as well as any changes in
the herd structure at any given site (preservation permitting), caution must be taken not to
make sweeping conclusions from such limited evidence.
Identifying the presence of plant cultivation for textile work in the archaeological record is
far more difficult. Plant fibres have been harvested and used in various ways (e.g. textiles,
cordage, netting, basketry, etc.) since the Upper Palaeolithic, as shown by the band of twisted
strings worn by several Venus figurines (Cybulksa and Maik 2007: 185). These earliest plant
fibres were typically bast fibres, harvested from the inner bark lining of plants like jute,
nettle, the lime tree, bast reed, hemp, and flax (Gleba and Harris 2019). Curiously, only some
of these plants were domesticated early (e.g. flax, hemp, and jute - see fig. 4 below) and
others were not (e.g. bast reed and nettle) (Andersson Strand 2012: 23). Why this divide
exists is still a subject of much debate; perhaps some were more suited to the climate and
environment where domestication began (Andersson Strand 2012), perhaps species like flax
20
and hemp were easier to gather and handle than the stinging nettle, or perhaps the textile
workers simply preferred the fibres from the plants they eventually domesticated, or these
species were not suitable for domestication. The growth range of jute centres in southern
Asia, mainly India and China, although there are records of its use as a food crop in Roman
Egypt (Pieroni 2005: 31). Lime trees are also limited in range to hot climates with enough
rainfall to support them, such as the Mediterranean Basin and southeast Asia. The bast fibres
from this can be peeled away from the tree trunk by hand, although the tree is believed to
have originated in Asia before being introduced to Romans (Reuther and Webber 1967).
However, excavations at Must Farm in 2006 revealed that most of the Bronze Age textiles
were most likely made from lime bast (Big Dig Diary 2015), suggesting that either the fibres
were imported as raw material or spun yarn, or it was imported as a finished textile.
Previously, it was thought that these fibres would not have been available to the people of
Iron Age Britain and north-west Europe, due to the geographical range of the plant and the
chronology of its introduction. Must Farm has proven that the trade of textiles and textile
fibres was prominent throughout prehistory, even reaching as far as Britain. Nonetheless, this
remains the only definitive proof of lime bast in Britain; without more evidence of the
prevalence of this fibre, it is impossible to know whether this fibre was still imported or used
in the Iron Age. Because of this, these fibres are not included in this thesis and will not be
discussed further.
In comparison, hemp and nettles (Urtico dioica) grow wild across much of continental
Europe and Britain, need no cultivation, and are difficult to positively identify, often being
mistaken for flax (Linum usitatissimum) (Andersson Strand 2012: 24). Prehistoric examples
of these fibres have been found in temperate Europe, proving that these fibres were known
and used in Sainte-Colombe and Chavéria (France), Prague-Záběhlice and Stehelčeves
(Czech Republic), and the Dürrnberg salt mine (Austria) (Grömer 2016a: 49). Hemp fabric
was discovered in a princely burial at Hochdorf (c.500 BC) where hemp textiles were used as
cushion covers and other upholstery (Banck-Burgess 2012). Interestingly, the hemp fibres
were not processed in the same way that flax fibres were; the bark was spun in strips instead
of individual fibres, suggesting that this was the traditional use of hemp in this area (Grömer
2016a: 49). One positively-identified nettle textile has been found in Denmark’s Lusehøj
burial mount (Bergfjord et al. 2012; Hald 1980: fig. 117; Mannering et al. 2012: 97, fig. 3.3);
this was a tightly-woven tabby weave that was wrapped around cremated remains and buried
in an urn, dated to c.900-750 BC (Grömer 2016a: 50-51). Nettle has little fibre yield per plant
of approximately 5-15% usable fibre compared to up to 40% from flax and hemp (Grömer
21
2016a: 50). Nettle does have other useful properties, for example young nettle leaves are
edible and when dried the entire plant can be used as very nutritious animal fodder (Grömer
2016a). In comparison, flax can produce textile fibres, straw for animal bedding, and seeds
that can be pressed to produce oil or eaten as a food crop (Andersson Strand 2012: 23-24;
Larsson 2013).
Figure 4. Examples of A) flax, B) jute, and C) hemp. (Brockhaus 1908)
Because of this, the presence of pollen or seeds from these plants at a site is not
immediately indicative of the plant’s use in the textile chaîne opératoire. However, studies of
flax seeds indicated that two different forms of flax were cultivated in Germany and
Switzerland by the late Neolithic period, resulting in a shorter flax plant with larger and more
numerous seeds for oil and a taller, lightly branched flax with smaller and fewer seeds for
fibre cultivation (Andersson Strand 2012: 24). So where seeds survive, it is possible to
identify the presence of the fibre-producing flax, although not the quantities in which it was
cultivated in any particular area. The identification of certain processing stages, such as
retting, is also complicated. Retting is the method of speeding up the rotting process, either
by laying the flax on the ground and letting dew form on the stems or by submerging the flax
in water (still or running); the plants are gathered from the damp environment  and dried after
a week or two (depending on the climate) before further processing. Retting is the most
22
efficient, least labour-intensive method of stripping away the unneeded exterior of the flax
and exposing the long internal fibres used in textile work.
Evidence of retting in north-western Europe is available in areas where flax remains are
preserved during the processing stages. Microscopic analysis of flax stems from Neolithic
lake settlements in south-west Germany demonstrated the complete stripping of epidermal
tissues that is characteristic of the retting process (Maier and Schlichtherle 2011: 570).
Reurings (2019) from Universiteit Leiden noted some of the smaller, fibre-producing flax
seeds in lake deposits, during their study of flax seeds and flax use across late Neolithic West
Frisia. Possible retting pits/wells in late Bronze Age/early Iron Age Denmark have been
identified via bundles of flax stems preserved at the base of wells and waterlogged pits from
a number of settlements across Denmark (Andresen and Karg 2011). Collectively, these
findings are strong evidence of retting performed in western Europe as early as 4000 B.C.,
although similar work in Britain is needed before it is possible to determine when this
practice spread here. Even where this evidence is available, it remains unclear when retting
was adopted as a widespread practice in the flax chaîne opératoire and whether early retting
was performed in ponds or natural streams (Andresen and Karg 2011: 570).
Fibre preparation tools
Some tools used in the processing of raw plant material are scarce in the archaeological
record, such as wooden hackles and scutches (see Andersson Strand 2012, fig. 3). Scutches
are long and sword-like wooden implements used to break off the hard inner stem and the
outermost fibres of flax. Hackles are combs used to brush through the flax fibres, pulling out
tangles and rougher sections of flax until the remaining fibres are smooth and aligned in the
same direction. While modern hackles use metal teeth and can come in sets with varying
widths between the teeth, remains of ancient hackles have been found without any metal
elements. Several wide wooden hackles with thorns embedded to act as teeth have been
found from Swiss Neolithic sites (Barber 1992: 14) and from La Tène sites in Slovakia and
Austria (Belanová-Štolcová and Grömer 2010: 11), indicating the combing of plant (and
possibly woollen) fibres before spinning. Unfortunately, the organic materials used for these
tools rarely survive deposition and textile archaeologists are forced to make broad
assumptions of the processing stages based on such limited evidence and close analysis of the
fibres in the surviving textile fragments. Interestingly, closer study of Neolithic and early
Bronze Age flax textiles has suggested that the majority of these threads were not hackled or
23
spun, but spliced together and rolled into a thread by hand (i.e., without tools) before being
plied using a drop spindle (Gleba and Harris 2018). This raises the question of when hackling
became part of the chaîne opératoire.
Splicing is an ancient form of textile production that predates spinning and weaving; two
different lengths of fibre are twisted (separately from each other) in the same direction, then
these twisted lengths are placed together and twisted in the opposite direction in a technique
known as plying (see Gleba and Harris 2018: fig.1). In this thesis, ‘thread’ is used to identify
a single-spun (or twisted) length of fibres, whereas ‘yarn’ is used to identify material made
from several threads that have been plied (wrapped around each other) to create a yarn.
Splicing has been identified by analysing yarns and textiles ranging from 4000-500 BC in
date and differs from spinning in two key ways. Firstly, the fibres used in splicing are thicker
than those used in spinning; these would have been drawn out for use one-by-one, whereas
spinners drawn fibres from a mass of soft, fluffed up fibres that are usually wrapped around a
distaff (Gleba and Harris 2018: 1). Secondly, splicing does not use any tools in the first stage;
the lengths are rolled by hand instead, usually along the thigh. This ancient technique was
still in use during the Neolithic, despite the fact that spindle whorls were available and soft,
retted plant fibres were likely produced in continental Europe. It is most likely that these
whorls were only used for plying the individual threads together, a process that would be
much quicker and easier to control with a weighted tool like the heavy whorls of this period
(Gleba and Harris 2018).
The predominance of splicing of plant fibres in the archaeological record strongly suggests
that the increase in spindle whorl numbers during the Bronze and Iron Ages directly
correlates with the introduction and spread of wool in textile production (Gleba and Harris
2018). By the middle Iron Age, the evidence of splicing in plant textiles greatly decreased
and at the same time, spindle whorls became lighter. It is very likely that the practice of
splicing faded out as these whorls became better suited to working with fine fibres, although
this is difficult to determine for sure (Gleba and Harris 2018). If this is the case, hackling
could have been adopted in the middle to late Iron Age in Britain. However, preservation of
organic tools and fibres is extremely poor in this region and the lack of plant processing tools
and other direct evidence of plant fibre production frustrates any study of this part of the
chaîne opératoire.
24
Spinning tools
Once the fibres have been prepared, spinning can occur. Fibres can be spun clockwise,
known as Z-spin, or anti-clockwise, known as S-spin (see fig. 5 below). Plying plant fibres
was commonly done during prehistory and experienced spinners could achieve extremely
fine plied yarns; in Zürich-Breitingerstrasse, a Neolithic Age site in Switzerland, preserved
plied bast yarns average roughly 0.3 mm (Leuzinger and Rast-Eicher 2011: 538). On the
other hand, wool is rarely plied across northwest Europe and Britain (Barber 1991). It is
possible that because wool cannot be spun as finely as bast fibres, weavers had to work with
single threads to keep the textiles as thin as those woven from plant fibres or they risked
making a textile that was too thick and stiff for what they needed. In northwest Europe and
Britain during the Bronze and Iron Ages, there is little evidence for spinners spinning wool in
only one direction and plying the other way (Barber 1991; Grömer 2016a), although in later
periods spinning in only one direction became set into many textile cultures, with some areas
preferring S-spin and others preferring Z-spin (Banck-Burgess 1999: 84–85;
Bender-Jørgensen 1992; Rast-Eicher 2008: 167–169).
Figure 5. The direction of S-spin (left) and Z-spin (right) threads. Author’s illustration.
During spinning, care must be taken to control the fibres to avoid any tangles from being
twisted into the yarn or pulling the other fibres out of the order that the previous stages had
managed to achieve. This is particularly important (and difficult) when drop-spinning fine
threads from long fibres. One of the easiest methods for keeping long fibres in check is by
using a distaff. Distaffs are lengths of wood held by the spinner, the prepared fibres would be
25
wrapped around the top and held in place with a ribbon while the base is often tucked into a
belt or held under an arm. By wrapping and tying the fibres to the distaff, the spinner can
draw on the raw material they need without the risk that it becomes tangled, lost, or dirtied
and even carry it around with them while both hands remain free to work the drop spindle.
Long distaffs, which can be tucked into a belt, are best suited for longer bast fibres, whereas
short and handheld ones suit shorter woollen fibres (Grömer 2016a: 79). There is a strong
correlation between the increasing presence of distaffs in images from the early Iron Age
Mediterranean (as well as archaeological remains from Austria, see Tomedi 2002: 161–162)
and the development of finer, longer wool and uniform, high quality yarns in this region
during the late Bronze/early Iron Age (Grömer 2016a: 79-80).
It appears that most distaffs in the Bronze and Iron Ages were made of wood or in a few
high-status burials, metal and other luxury materials (Grömer 2016a: 80). Unfortunately, no
remains have been positively identified as distaffs from Britain or north-west Europe;
however, these regions do follow the pattern of higher wool and yarn quality seen in the
Mediterranean during the same period. While it is possible that distaffs were as common in
Britain and north-west Europe as in the Mediterranean, the scarcity of evidence makes study
of these tools in these areas almost impossible. There are certainly no recorded remains of
distaffs from British Iron Age contexts and because of this, distaffs will not be included in
this study of textile tools.
Spindle whorls and looms weights are more robust, being frequently made from fired clay,
stone, or reused pottery shards. This allows for better preservation in the archaeological
record and the development of several methods of study and analysis that can determine the
form of fabric production each tool is best suited to (Andersson 2003; Barber 1992;
Mårtensson et al. 2006a and b, 2009; Grömer 2005, 2006). Spindle whorls sit on a wooden
shaft - the spindle - and work as a flywheel to keep the spindle turning for as long as possible
as it hangs. Its weight also provides extra momentum for the spin and tension of the fibres,
allowing the twist to gather in the material instead of turning uselessly with the spindle.
Spindle whorls can be placed anywhere along the length of the spindle shaft. A whorl that
sits at the base of the spindle while it is in use makes the spindle ‘bottom-weighted’, whereas
a whorl at the top of the spindle during spinning makes it ‘top-weighted’, and one with the
whorl in the spindle’s centre is ‘middle-weighted’ (see fig. 6 below). It has been noted that
top-weighted spindles turn faster and for longer than bottom-weighted spindles that have
similar whorl diameters and weights, and that the spinner must change where they touch the
26
spindle to turn it depending on the whorl’s position; there is little difference in the threads
produced (Andersson 2008: 74-74; Gleba 2008: 106).
Another method of spinning is support spinning, where the turning spindle is allowed to
run resting in a bowl or on the ground (Grömer 2005). This takes most of the weight off of
the threads being spun and onto where the spindle touches the bowl/floor, thus producing a
thinner and softer thread than would otherwise be possible. The supported spinning technique
can use whorls of any size and weight; heavier whorls might even be more beneficial with
this method, since these would retain more momentum than smaller whorls while most of the
weight is kept off of the spun thread.
Figure 6. Examples of a bottom weighted spindle (left), middle-weighted spindle (centre), and top-weighted
spindle (right). Author’s illustration.
The use of clay as one of the main materials for whorls allowed these tools to be formed
into various shapes across Europe and Britain (see fig. 7, below); these shapes appear to alter
very little over time, although some forms are more common in particular areas. Various
experiments have determined that the overall shape of a whorl has little effect on the spinning
process or the character of the finished threads (Grömer 2005; Mårtenson et al. 2006b). This
suggests that the range of whorl shapes are the result of different textile cultures and
traditions, rather than changes in spinning technology or the finished thread. There is a
noticeable increase in spindle whorl varieties around the same time that more intensive sheep
breeding and wool use became widespread across Europe and Britain during the late Bronze
and early Iron Age, (Belanová-Štolcová and Grömer 2010: 14-18; Gleba 2008: 106). This
could suggest that including wool into the textile chaîne opératoire caused a massive change
27
in textile traditions, although this would ignore other major cultural shifts that happened
during this time.
Across north-west Europe and Britain, this was a period of cultural development in almost
all areas: from adopting ironworking and iron tools into society, to increased agricultural
production, new settlement types, and large breaks and changes to long-distance trade and
exchange networks (all these cultural changes are discussed in detail in Iron Age Society,
below). The development of new spindle whorl shapes could be another reflection of the
social upheaval common throughout this period: textile workers adapting to new materials
while helping to establish their local group identities. In addition to this, there is a strong link
between whorl weight and the fibres used in spinning (as well as the finished threads made
by the spinner). Barber (1992:52) noted the use of heavy spindle whorls (between 100 - 150
g) for spinning long-fibre wool in ethnographic studies. Long plant fibres required heavier
spindle whorls; flax tow, cotton, and short woollen fibres required lighter spindle whorls, of
approx. 8 g; and medium-length wool required whorls around 33 g in weight (Barber 1992:
52; Ryder 1968: 81). However, it was not until the work of Mårtensson, Andersson, Nosch,
and Batzer in 2006 that such observations were tested in controlled experiments.
Figure 7. Examples of spindle whorl shapes, following Gleba’s (2008: 105) typology. Globular (upper left),
discoid (lower left), cylindrical (middle), conical (upper right), and truncated conical (lower right). Author’s
illustration.
Through various experiments, Mårtensson et al. determined which properties of various
tools directly affected the nature of the final textile (2009: 378). The weight and diameter of
spindle whorls were most important; the weight directly influenced the diameter of the thread
28
produced, which in turn affected the loom weights and tension required for weaving
(Mårtensson et al. 2007; 2009: 378, 392), as demonstrated in table 1, below. The
diameter:height ratio of the whorl directly corresponded to the speed of the turning spindle
and how long the spindle could turn for (Gleba 2008: 106; Mårtensson et al. 2009: 378-9). A
whorl with a wider diameter and lower height could spin for longer at a lower speed,
allowing a spinner to produce a long, fine thread without constantly stopping to turn the
spindle. A whorl with a smaller diameter and taller height could spin faster but for shorter
periods of time, creating a tighter twist more suitable for woollen fibres and a more robust
thread for use as a warp (Gleba 2008: 106; Mårtensson et al. 2009: 378-9). The direct link
between the weight and dimensions of spindle whorls established by these experiments
allows archaeologists to determine not only the type of threads spun in certain areas, but also
what fibres are most suited to spinning in these places as well. So even if there is little to no
evidence of what raw materials a particular site used due to poor preservation, it is still
possible to take into account all the common fibres available to spinners in the past.
Table 1. Thread diameters and tension required for weaving according to spindle whorl weight and
corresponding number of threads per loom weight to achieve required tension, after Mårtensson et al. 2009:
378.
Spindle whorl
weight
Thread diameter Required tension
for weaving
Threads per 300g
loom weight
Threads per 500g
loom weight
4 g < 0.3 mm c.10 g 30 50
8 g 0.3 - 0.4 mm c.15 - 20 g 20 - 15 33.3 – 25
18 g 0.4 - 0.6 mm c.25 - 30 g 12 - 10 20 - 16.6
44 g 0.8 1 mm c.40 g 7.5 12.5
Weaving
Before we continue further, some weaving terminology must be explained to avoid
confusion.
Warp - Threads placed on a loom before weaving that run the length of the fabric. These are placed
under tension during the weaving process.
Weft - Threads that are placed across the warp by the weaver. These run the width of the fabric and
can be placed over and under warp threads at varying intervals to create different
styles of weaving.
29
Shed - A selection of warp threads that are raised apart from the rest of the warp before the weft is run
through, and lowered before the weft is beaten into place. A loom may have a
number of sheds depending on the weave pattern being created.
Heddle - The means by which the shed is raised and lowered; this may be a series of strings attached
to a stick or a piece of wood with holes for the shed and slits for the rest of the warp
to slide through (known as a ‘rigid heddle’).
Shuttle - The tool that holds the weft while it is passed back and forth between the warp.
Tabby - The simplest form of weave, where the weft goes over one warp thread and under the next
warp in succession.
1/2 Twill - A more complicated weave, where the warp goes over one warp thread and under the next
two, before repeating this pattern.
2/2 Twill - Similar to the above, but here the weft passes over two warp threads and under the next
two, before repeating this pattern.
2/1 Twill - Similar to the first twill, but where the weft passes over two warp threads and then under
one, before repeating this pattern.
Selvedge - The edge of the fabric; this only refers to the side edges of the textile where the weft
changes directions at the edge of the warp.
Starting/finishing border - The start and end of the fabric; these borders tend to be woven differently
to the rest of the fabric, these borders are responsible for spacing the warp threads
and ensuring the fabric does not fray.
Weft-faced - A style of weaving where weft outnumbers warp threads per square cm. The weft is
beaten very firmly, so that it completely (or almost, depending on the tightness of
the weft) obscures the warp from view. Any fabric can be weft-faced, whether the
underlying weave pattern is tabby or twill.
Warp-faced - A style of weaving where warp threads outnumber weft threads per square cm,
resulting in a fabric where the warp threads completely (or almost, depending on the
tightness of the warp) obscure the weft. This can be done with any type of weave
pattern, tabby or twill.
Weave patterns
Textiles can be patterned in a number of different ways by the process of weaving, even
before considering the use of dyed yarns or fabrics. During the Bronze/Iron Age transition,
weavers appear to experiment with decorative weaving techniques. One technique
30
characteristic of the Hallstatt salt mine assemblages is the use of spin patterns (Grömer
2016a: 171), which uses the direction that single thread wool is spun in to create subtle
patterns on the textile (see fig. 8 below). The pattern is highlighted when exposed to light,
although it can be difficult to see in the fragmentary textiles that remain. Weavers arranged S-
and Z-spun threads in alternating groups, both as warp and weft, to create subtle stripes or
check patterns on the textile. This technique does not appear to affect the properties of the
finished fabric (e.g. its strength, pliability, etc.), which strongly suggests that this was done
purely for aesthetic purposes. Spin patterns were a method of creating subtle patterning with
a tabby weave, which is the simplest form of weaving possible. It was only in the Iron Age
that twill patterns were developed, despite being known as a basketry technique since the
Neolithic (Barber 1991; Grömer 2006, 2016: 130, fig. 73; Richter 2010: fig. 34.2-34.3).
Unlike spin patterning, twill is created by the placement of the weft during the weaving
process, changing the structure of the textile. This means that it has a clearer appearance than
spin patterning and alters the physical properties of the finished fabric, allowing the fabric to
drape more fluidly than tabby textiles. Twill weaving also demonstrates the development of
more complicated weaving technology and its appearance in the Iron Age marks a major
development in textile history (Grömer 2016a: 135).
Figure 8. An example of how direction of spin can alter the appearance of a textile. Author’s illustration.
31
Weaving tools
There are a variety of tools used in weaving, some with known functions and other tools with
uses that are debated. Two of the tools with known functions are pin beaters and weaving
swords. During the weaving process, newly inserted weft has to be pushed (or ‘beaten’) into
line with the rest of the woven fabric. Weaving swords are one method of doing so by using
long, blade-like pieces of wood (see fig. 9 below). These swords are able to cover a wide
section of weft while beating and when used particularly on ground looms (where weft is
beaten towards the weaver), these can be very effective at ensuring the weft lies evenly
across the warp. Even on a standing loom, weaving swords can ensure that weft is firmly
beaten into place, making it the most suitable tool for producing densely woven or weft-faced
textiles. Several weaving swords have been excavated from various European Neolithic and
Bronze Age wetland settlements (Bazzanella et al. 2003: 141, 228–229; Grömer 2016a: 118;
Médard 2010: 168, 226), as well as Iron Age sites in Italy, Switzerland, Austria, and
Germany (Gostenčnik 2013: fig. 4.5c; Grömer 2016a: 118; Von Kurzynski 1996: 14–15). So
far, none from similar periods have been found in Britain.
Pin beaters are small pins blunted at both ends that are made from bone or antler (possibly
wood, although none have been found so far) (see fig. 9 below). While used for the same
purpose of lining the weft up properly, pin beaters are less cumbersome and easier to apply to
small areas, which would benefit weavers working on a delicate textile with extremely fine
threads. It is possible that weavers could have only used these to untangle warp threads
during weaving (Grömer 2016a: 117). On the other hand, these may have been easier to use
on vertical looms, where the weft is beaten up and away from the weaver. Unfortunately, few
positively-identified pin beaters have been excavated alongside other weaving tools for
archaeologists to determine the primary use of these tools.
32
Figure 9. An example of a late Bronze Age bone pin beater from Cyprus (top, after Smith 2001: 84, fig. 2) and a
Iron Age weaving sword from Britain, made of red deer bone (bottom, courtesy of Museum of London Online
Collections). Not to scale.
Weaving combs are also used to beat weft into shape, although on what type of loom is
debated (Coles 1987: 105-6; Hodder and Hedges 1977: 17; Tuohy 1999). These tools are
made by splitting an antler beam into a small section and carving teeth into one end (see fig.
10 below). It is, and has been for many years, widely believed that weaving combs were
primarily used on warp-weighted looms for beating the weft (Bulleid and Gray 1911; Coles
1987; Cunliffe 2005; Hodder and Hedges 1977). Recently, however, closer analysis on the
teeth of weaving combs has raised questions about this theory. Tuohy (1999) pointed out that
the wear found on many weaving combs was both unevenly distributed across the teeth and
worn in places that should not be reached if it were used on a standing loom. Tuohy argued
that these weaving combs were more commonly used in weaving bands and belts (1999:
56-8). Whether in tablet weaving or another form of small loom is unclear; however, her
evidence has brought to light the ambiguity in our understanding of weaving combs and their
use.
33
Figure 10. An example of the form and manufacturing process of a weaving comb. A) Areas of antler used for
various purposes. B) Splitting of antler, ready for carving into a weaving comb. After MacGregor 1985, fig. 42.
Tuohy’s theory is further supported by a general pattern of weaving tools seen in British
Iron Age settlements; in sites where organic materials survive, weaving combs are rarely
found alongside large numbers of loom weights. In Glastonbury Lake Village, for example,
weaving combs appear in larger numbers during its middle to later Iron Age phases, yet in
the same phases the number of loom weights drop from 19 in the middle Iron Age, to 3 in the
late Iron Age, and none in the final phase of the site. In comparison, the weaving combs
maintain a steady number; 13 in the middle Iron Age, 16 in both the late Iron Age, and 16 in
the final phase of the site (Coles and Minnett 1995: 33-89). This pattern is also seen in
34
Danebury (Cunliffe and Poole 1991: 354, 372) and Maiden Castle (Laws 1991: 234; Poole
1991b: 210), demonstrating that the presence of weaving combs could reliably indicate a
different form of textile production (although what was being produced is still uncertain).
Many weaving combs are highly decorated and have a single hole at one end where a strap
of leather could thread through, allowing it to have a handle or loop through a belt (Tuohy
1991). Such an attachment combined with the decoration may suggest that these were seen as
personal belongings rather than tools that belonged to a household. Conversely, Tina Tuohy
(1999) argued that these decorations were not added by the weavers using these tools, but the
antler craft workers who produced them as a means of making the tool more attractive to
potential buyers. However, I would argue that this theory takes a very modern, capitalist view
of tools. Throughout history, humans have demonstrated an almost compulsive drive to
decorate items, whether to simply enhance the object in question or to make it more
individual and unique. I would argue that dismissing the decorations on weaving combs
because these tools were common and made by someone other than its final user denies the
agency of the weaver in choosing or personalising their tools, as well as the possibility that
the weaver wanted to display the weaving comb as a symbol of their status or ability as a
weaver. It should be noted that although there are numerous examples of weaving combs in
the British archaeological record, there is as of yet no agreed upon method of recording or
analysing these tools. However, Jennifer Beamer is currently developing a recording system
to record the most important information from weaving combs and testing various methods
of use wear, to determine what causes the wear on the original artefacts and how to best
record such wear (Beamer 2017, 2018, pers.communication).
Looms
There are several types of loom present throughout prehistoric Europe: the ground loom, the
warp-weighted loom, and the two-beam loom (see fig. 11 below for examples of each). The
ground loom is a horizontal form of weaving where the warp is stretched out between two
beams that are staked to the ground, the weaver then sits at one end and as they work, they
roll up the newly woven fabric on the beam closest to them, allowing the weaver to progress
up the warp threads. Horizontal looms like this require a lot of floor space since they are
staked down, as such, these looms are often set up outdoors. This leaves the loom, yarns, and
weaver vulnerable to the elements, which is why it is more common in the drier climates of
the Near and Middle East (Grömer 2016a: 107-108). Archaeological evidence of ground
35
loom use shows that this loom type was common in ancient Egypt (Barber 1991: 83–91, fig.
11.1) but there is no evidence - archaeological, iconographical, or otherwise - that this was in
use in north-west Europe or Britain at any time. The two-beam loom is a form of standing
loom that is also underrepresented in the historical and archaeological record, although there
is some evidence of its use in north-west Europe. The loom consists of two vertical posts
with a beam across the top and base; its posts could be sunk into the ground to make a
permanent loom, or the whole frame could be leant against a wall. The warp would be
threaded in a continuous loop around these beams to create a tube that could be moved up or
down by the weaver, allowing them to weave a continuous fabric around the beams.
Figure 11. Examples of a two beam loom (left), ground loom (centre), and warp weighted loom (right). Author’s
illustration.
A square frame filled with a checked pattern is depicted alongside a female figure on a
conical necked vessel of the Hallstatt period from Rabensburg (Austria), which could be an
36
example of a weaving frame or two-beam loom (Grömer 2016a: 139, fig. 78). Additionally,
textiles have been excavated from Danish bog that are clearly from two-beam looms; these
are woven as a tube, without the typical starting and finishing borders of other looms
(warp-weighted or ground looms) (Grömer 2016a: 139; Stærmose-Nielsen 1999: 124–125).
These textiles date to c.1400-1100 BC and include articles of clothing such as cloaks, skirts,
tops, and blouses (Stærmose-Nielsen 1999: 124–125). Unfortunately, two-beam looms are
difficult to trace without the type of evidence mentioned above; because it does not use loom
weights, the only features left from these looms are the post holes where it stood.  While the
two-beam loom does not appear to be commonly used across prehistoric north-west Europe
and Britain, there are instances where its use is inferred. In some areas of Britain (e.g.
Teeside), spindle whorls are common finds but loom weights are entirely absent; it is
extremely unlikely that this entire area did not weave any textiles, so the two-beam loom is
believed to be the weaving tool of choice (Shaffrey 2017). Elsewhere across prehistoric
north-west Europe and Britain, the warp-weighted loom is commonly used.
Warp-weighted looms are standing looms that could be leant against a wall or created from
two posts. A beam at the top held a tablet-woven border through which the warp fibres were
threaded before hanging down to the base of the loom and threaded through a loom weight.
Depictions from ancient Greece and some more abstract art from north-west Europe
demonstrate how weaving was accomplished on such a set-up (Stærmose-Nielsen 1999:
144–147). During weaving, the necessary heddle is lifted and the weft inserted before it is
beaten upwards and away from the weaver, creating the cloth at the top of the loom. For
archaeologists, the advantage of this form of weaving is the physical evidence left behind in
the form of loom weights. Loom weights first appear during the Neolithic (c. 6th Millennium
BC) in south-east Europe (Barber 1991: 93) and continue through to the Medieval period
throughout Europe. Various in situ finds from prehistoric contexts across Europe have
demonstrated the use of the warp-weighted loom; lines of loom weights have been excavated
from the remains of structures that burnt down while the loom was in use from several sites,
including Virje-Sušine (Northern Croatia, c. 3rd-2nd century BC) and pre-Roman Iberia
(Karavidović and Sekelj Ivančan 2018 and Marín-Aguilera  2019 respectively).
Finding loom weights in situ also allows archaeologists to determine what type of textile
was being woven at the time of destruction, not via analysis of the loom weights but by the
placement of these tools. Ingrid Schierer (1987) conducted a series of experiments based on
37
the discovery of three rows of loom weights at the Late Bronze Age site of Gars-Thunau
(Lower Austria) (for a plan of the finds, see Schierer 1987: 44 and Grömer 2016a: fig. 75).
Schierer repeatedly warped the same loom for tabby and twill, setting up sheds in different
ways each time, before destroying the weave in various ways. The warp was cut and the loom
knocked over and eventually burnt in order to simulate the possible ways that the original
could have been destroyed (Schierer 1987: 70-80). Each time, she recorded the exact position
of the loom weights and compared the data to the evidence from Gars-Thunau. Her results
demonstrated that tabby weaves result in two clear rows of loom weights, while twills result
in the weights clustered closer together in several rows or a single heap (Schierer 1987:
80-87). Unfortunately, no such finds have been discovered in British Iron Age contexts,
forcing archaeologists to identify loom weights based on deposition alongside other textile
tools and similarity to previously identified objects.
Only Anglo-Saxon period features have produced lines of loom weights in Britain; during
the Iron Age, triangular ‘loom weights’ are exclusively excavated from pits, post holes, and
ditches either as a single find or in small groups (Shaffrey 2017: 230). The identification of
British loom weights is further hampered by a recent debate as to whether the triangular clay
weights of the Iron Age are actually oven bricks, due to the common evidence of burning on
these artefacts (Poole 1995: 285; Shaffrey 2017: 230). Despite this, most British
archaeologists agree that these triangular artefacts are loom weights that were either
baked/fired to harden before use or reused as oven bricks (Shaffrey 2017). Another
peculiarity of the British evidence is the weight of the triangular ‘loom weights’: while the
majority of European Iron Age loom weights measure about or less than 1kg, most of the
British weights are between 1-2 kg (Shaffrey 2017). As of yet, there is no agreed upon
explanation for why weavers in Britain used these heavier weights for weaving, although it is
possible that these were designed to work with plant threads (which require heavier weights
than woollen threads).
Because loom weights survive so well in different environments, archaeologists are able to
determine the sort of textiles these tools produced, as well as what properties the threads
attached to these weights would have had. Experiments have shown that warp-weighted
weaving is directly influenced by the weights and diameters of spindle whorls. The thickness
of warp threads affects the amount of tension required for optimal weaving efficiency and
corresponds to the number of warp threads that can be attached to a loom weight (Andersson
Strand 2011: 9; Dimova 2016: 656; Mårtensson et al. 2007, 2009). However, weaving is also
38
influenced by the thickness of the loom weights, which can adversely affect the shape of the
fabric as it is woven (Mårtensson et al. 2007: 8-10, 2009: 374). Too few loom weights and
the combined width of these tools would not be enough to keep the warp straight, instead
pulling it inwards; in comparison, too many loom weights would have a combined width
larger than the starting border of the fabric, causing it to expand towards the base (see fig. 12,
below). While an experienced weaver may be able to counteract these effects to some degree,
the simpler solution is to have an upper and lower limit to the number of warp threads
attached to each loom weight, regardless of how many could be supported by the weights.
Over several experiments using reconstructions of thin loom weights excavated at Bronze
Age Troy, Mårtensson’s team determined that optimal weaving could take place between 4
and 30 threads per loom weight (2009: 389-390); more would place too little tension on the
warps, causing them to move and become uneven, less would result in too many loom
weights and a stretching of the woven fabric.
Figure 12. Unevenness in weaving, resulting from too few loom weights (left) and too many (right). Author’s
illustration, after Mårtensson et al. 2007 and 2009.
With this knowledge of various tool’s limitations and requirements, it is possible to
extrapolate the range of fabrics produced at any given site, so long as enough loom weights
and spindle whorls survive to provide a reasonable average for the types of threads spun and
woven there. However, Mårtensson’s team acknowledged that their conclusions can only be
used as a guideline for further work, particularly in British and northern European contexts
(2009). The loom weights found in these regions during both the Bronze and Iron Ages are
39
far larger, thicker, and heavier than those in the experiments detailed above (for comparisons
see fig. 13, below) (Shaffrey 2017). The two sets of Trojan loom weights that Mårtensson
based her reconstructions on are disk-shaped, with a single hole, and weigh between 108–112
g and 183–187 g, with a thickness of 2 cm in both groups (2009: 381). In comparison, loom
weights from Iron Age Danebury have a triangular shape, with holes through the narrow
sides at each corner, and weigh on average 1370 g and measure 87.7 cm in thickness
(Cunliffe and Poole 1984, 1991).
Figure 13. Comparison of Mårtensson’s reconstruction of a Bronze Age Trojan loom weight (left) and a typical
British Iron Age loom weight (right). Author’s illustration, not to scale.
Shaffrey’s (2017: 241) work on British loom weights noted that while it is possible for
weights around 2 kg to be used as loom weights, the amount of warp threads that need to be
attached can cause complications in weaving and produce an uneven fabric. Alternatively, a
few heavy weights can be used to weigh down a wooden rod that the warp threads are fixed
around, as described by Hoffmann (1974: 87). Such a form of weaving could exist alongside
warp-weighted looms during British prehistory and would require only a few heavy loom
weights for each set-up (Shaffrey 2017: 242), although these weights would have to be
limited to 2 kg or more. The breadth of work focusing on textile tools and the fabrics such
tools produced provides an excellent starting point for developing a suitable method of
studying British Iron Age textile tools. It is clear, however, that the heavy, triangular loom
weights found in Britain during this period present the greatest challenge in determining the
most optimal loom set-up for these tools. For the majority of British loom weights weighing
500-1500 g, Mårtensson’s experiments provide a good basis of analyses and further
adaptation of Mårtensson’s work is discussed further in the methodology section below.
40
Tablet weaving
This form of weaving differs from loom weaving in both how the shed is made and how the
pattern of the fabric is formed. The tablets used in this weaving can be square or triangular in
shape, with holes in each corner for threading; it is these holes that create the different shed
on the loom. As the tablets are turned, the warp threads are twisted around on another - some
lowered and others raised - creating a twisted pattern unique to this style of weaving (see fig.
14 below). Unlike loom weaving, where the appearance of the fabric depends on the number
of warp and weft threads and the weaver intending to produce certain weave styles, tablet
weaving always produces warp-faced fabrics; any dyed threads must be used as warp threads
so that the colours are not hidden underneath the weft. These warp threads are passed through
a tablet and the direction of the threads through the tablet can also affect the final weave
pattern (see fig. 14, below).
Figure 14. An example of tablet weaving structure, warping, and method. Author’s illustration
It is not clear when this form of weaving first began, whether it predates the invention of
the looms discussed above or not; the earliest pieces of evidence available are the ‘Ramses’
belt from Egypt (c.1200 BC) and the tablet-woven starting border of a tabby-weave textile
from the Hallstatt salt mines (c.1500-1200 BC) (Collingwood 1982: 10-11; Farke 1993: 111;
Grömer 2016a: 102). Despite the lack of earlier evidence, tablet weaving appears to be a well
known and fully-developed craft form by the Bronze Age; the range of techniques and
41
complicated weaving patterns attest to how thoroughly this form of weaving was known and
how comfortable the weavers were working with such complexity (Grömer 2016a: 102).
Tablet weaving was most commonly used to create narrow, decorated bands and borders
around other textiles (Grömer 2016a: 103). The major advantage of this weaving technique is
the strength of the weave; tablet woven bands can withstand a great deal of tension and
stress. Even one warp thread breaking does not damage the strength or structure of the band,
since turning the tablets during weaving ensures that there are always about three other warp
threads to take up the strain (Grömer 2016a: 103-4).
The archaeological evidence for tablet weaving mainly depends on surviving fragments of
tablet-woven textiles; without these it is impossible to gauge how complex these weaves
were or the quality of the fabrics produced and the skills of the weaver. Tablet weaving tools
may survive but as these are often very small and thin clay or wood tablets, the chances are
very slim that these could survive deposition - especially in temperate environments like
Britain (Grömer 2016a: 104-5). The only other possible tool connected with tablet weaving
that could survive in the archaeological record are bobbins; these could have functioned as
spools/weights for the warp threads while a starting/finishing border was woven (Grömer
2016a: 104-6, fig.56). It is important to bear in mind that bobbins could be used for a range
of tasks; as of yet, it is not possible to determine how common tablet weaving was in Britain,
only that by the Iron Age it was well-established in the textile culture across Britain and
north-west Europe.
Dyeing
Dyeing is another complex process in the textile chaîne opératoire, requiring specialist
knowledge and experience to produce the most vivid and long-lasting colours. This process
can be performed at multiple stages of the chaîne opératoire depending on the final product
the weaver wants to make.
Dyeing wool prior to spinning can produce coloured felt or spun yarn with an even colour
distribution throughout; dyeing spun wool before weaving would result in several colours
from a single fleece, allowing weavers to make checked or striped fabrics; and dyeing woven
fabric would allow several finished textiles to be dyed at once, ensuring these receive
uniform colouring (Barber 1991). Prehistoric dyes rely on natural sources for pigment, which
can limit what evidence survives in the archaeological record. For example, woad leaves are
42
beaten to produce indigo blue, boiled to transfer the colour to the fibres, and discarded. The
chances of such leaves surviving deposition in any recognisable state are extremely low. In
comparison, purple dye was produced in the Mediterranean from several sub-species of
murex (sea snails); creating enough pigment for dyeing required workers to crush thousands
of snails (Jacoby 2004: 210), which would produce waste able to withstand deposition (e.g.
crushed shells). Excavations across Crete revealed large amounts of crushed shells of various
murex species alongside pottery fragments from the Minoan civilisation (Reese 1987).
Archaeologists were able to use these ceramic fragments to date the earliest production of
purple dye to roughly 2000-1700 BC.
In northwest Europe, the best evidence for dyes in textile production comes from the
Hallstatt salt mines, with its range of fabrics and high quality of preservation offering the best
conditions for dye to survive. Analysis of various micro-samples used high-tech methods
such as scanning electron microscopy with energy-dispersive X-ray analysis (SEM-EDS) to
detect chemical elements of the dye process, such as mordant use, and high performance
liquid chromatography with photodiode array detection (HPLC-PDA) to analyse the dye
colours (Hofmann-de Keijzer et al. 2013: 125). HPLC-PDA is the most accurate method of
analysing dyes to date, determining both the dye’s retention time (time it takes for dye to
separate from the fibres) and UV-VIS absorption spectrum before matching this data to a
world-wide database to ascertain the correct dyestuff used (Hofmann-de Keijzer et al. 2013:
125). These methods indicated that Hallstatt textile workers used woad for blue dye, lichens,
dye insects, and various Rubiaceae plants for red dyes, and either bedstraw or wild madder
for yellow dyes (Hofmann-de Keijzer et al. 2013: 128). It is also possible that weld or
weld-dyed fibres were imported from the Mediterranean for more yellow, although
degradation of the chemical signatures renders it difficult to be certain (Hofmann-de Keijzer
et al. 2013: 128). The presence of dye insects is strong evidence that dyestuffs were traded
across Europe and Asia, as these are not native species and would have to be imported
(Hofmann-de Keijzer et al. 2013: 128). These dye analyses also indicate that colours such as
black, brown, and green were achieved via dyeing the fabric a primary colour with a mordant
of iron, aluminium, or copper (Hofmann-de Keijzer et al. 2013: 128). These mordants
allowed textile workers to ‘fix’ the dye to the fibres permanently and could alter the pigment
of the final dyed fabric; changing which mordant was used with each colour would result in
various colours, allowing textile workers to produce a full spectrum of dyes with a limited
range of natural dyes.
43
In areas without suitably well-preserved fabric remains, chemical analysis of organic
residue in pottery could provide some information on dyeing. So far, little work has been
conducted on finds from Britain or northwest Europe; such technical analysis would be
difficult and costly to perform on the masses of pottery from these regions. Organic residue
analysis has been conducted on ceramic fragments from a Minoan period dyeing workshop in
Alatsomouri-Pefka, Crete (Koh et al. 2016). The Minoan civilisation had a well-developed
textile industry with specialist workers and workshop production at various stages of the
textile chaîne opératoire; Alatsomouri-Pefka itself had a series of seven stone basins cut into
the bedrock for use as dye vats, no domestic structures, a number of tripod vessels with
evidence of scorching around the bases, and large quantities of crushed murex shells (Koh et
al 2016: 536). These remains clearly represented a specialist dye workshop, which provided
an excellent opportunity to test the suitability of organic residue analysis. Unlike the
techniques used for the Hallstatt fabrics, this analysis was performed with gas
chromatography, a non-destructive procedure conducted with samples taken from solvent
applied to the ceramic. 42 samples from 21 ceramic fragments were taken in total; the first
sample was taken using dichloromethane and the other sample using ethanol, taken from the
same fragment (Koh et al. 2016: 537). This analysis revealed the presence of lanolin - an oil
produced by sheep, demonstrating at least some of the materials being dyed - as well as
Murex purple, madder, weld, and urine (Koh et al. 2016: 537). Urine was used to create a
lighter shade of purple, while madder and weld produce red and yellow dyes respectively.
Unfortunately, organic residue analysis cannot gather any information on the mordants used
in the dying process, so the full range of colours produced at this site is still unclear.
These various studies of dyeing processes across Europe demonstrate the widespread use of
some dyestuffs, such as weld and madder and the range of evidence that can provide
information on this stage of the textile chaîne opératoire. While it is not unreasonable to
argue that woad, weld, and madder were used in British Iron Age textile production, the
British archaeological record still lacks the remains able to prove the use of these materials or
other dyes imported from further afield. Because of this, dyeing as a specialist craft will not
be explored further in this thesis.
44
Finishing techniques
The end of the textile chaîne opératoire can involve several techniques that alter the
appearance of the textile (e.g. fulling and embellishing) and adapt it to its final use (e.g.
tailoring). The tools involved in these finishing techniques can also survive deposition,
despite being made from organic materials, although other methods are more difficult to
identify in the archaeological record.
Fulling
Fulling (also known as ‘walking’ or ‘waulking’) is a process of thickening and cleaning a
woven woollen fabric of oils and dirt before being cut. Taken straight from the loom, the
textiles were submerged in water mixed with ammonium-rich urine and repeatedly trodden
on, tightening the weave and whitening the fabric, before being rinsed repeatedly and
stretched. Finally, a teasel was run over the textile’s surface to ‘raise the nap’, a term
describing the process of teasing out fibres on the surface to soften the textile, giving it a
fuzzy appearance. This fuzzy surface was then trimmed with fuller’s shears before the cloth
could be sent on for final embellishments and tailoring. The various stages of this process are
mentioned in Greek and Roman sources (Flohr 2013; Forbes 1956: 80-89; Singer et al. 1962:
216-221; Zawadzki 2013), which attests to the long history of this finishing technique. The
earliest evidence of fulling is debatable. Some historians believe that some Mesopotamian
seals from the Uruk period (c.3500-3200 BC) represent fullers treading cloth, although these
could also represent tanners or other craftworkers at work (Algaze 2008: 81, 85-86, figs.
14g-h). On the other hand, most agree that the first clear evidence for fulling in Mesopotamia
dates to roughly 2500 BC (Peyronel 2004: 72; Soriga 2017: 24). Regardless of the precise
data, it is clear that fulling was a long-established process connected with wool use in textile
production. Roman sources demonstrate the use of fulling in Britain after the Roman
conquest in AD 43 (Wild 1967), however there are almost no archaeological remains to
indicate whether this technique was present prior to this date. Because of this, fulling will not
be included further in this thesis.
Embellishing
Rare finds from northwest Europe have demonstrated a wide range of embellishment used on
clothing since the Bronze Age (Barber 1991; Grömer 2018). These finds are typically limited
to funerary contexts, which raises questions surrounding the social meaning of these textiles
45
(i.e.: ‘were these clothes solely for funerals or were these the best clothes that belonged to the
deceased?’ ‘Could these be examples of common clothes or were they indicative of particular
social status or rank?’ and so on). However, in regards to the production of these textiles, the
context of the dress worn by the deceased does not particularly matter. What matters is that
these forms of embellishment were known and used; the knowledge and skill required for
these techniques did exist. How widespread this knowledge was, whether it came into fashion
and later disappeared, whether or not it was limited to a few areas or craft workers - these are
the issues at hand.
One form of embellishment is the use of metal threads in the fabric, either woven or sewn
in. Several examples of gold threads come from three Bronze Age Age cremation burials in
Vösendorf (Austria) (Talaa 1991: fig. 33),  a gold hoard in Óbuda (Hungary) (Barth
1988/89), and the hillfort of Várvölgy (Hungary) (Müller 2012). This technique was also
known in the eastern Mediterranean and the Near East: Assyrians and Babylonians made
fabrics with gold threads and gold appliqueé (Gleba 2008), tunics with gold sequins were
excavated from Tutankamun’s burial chamber (Vogelsang-Eastwood 1999). The main
difference between the gold threads used in these areas is the method in which it was created
and placed in the fabric. Textile fragments from the Iron Age site of Grafenbühel were
narrow strips of gold inserted to the weave as supplemental weft threads (Banck-Burgess
1999: fig. 10). In comparison, the contemporary Graeco-Roman world mainly used threads of
natural fibre wrapped with gold threads (Gleba 2008).
Embroidery is another method of embellishment, generally considered an ‘applied
decoration’ along with beads, sequins, and appliqueé (Grömer 2018). Two prehistoric sites
stand out as examples of embroidery: Molina di Ledro, a Bronze Age pile-driven settlement
in Northern Italy, and Pfäffikon-Irgenhausen, a Bronze Age settlement in Switzerland.
Molina di Ledro has several examples of sewing techniques, including a stitch similar to the
modern running stitch, on a fine tabby cloth with an open weave (Bazzanella and Mayr 2009:
58- 9). In comparison, the Pfäffikon-Irgenhausen textile is more complicated, with patterns of
checkered borders, triangles, and stripes on a fine plain white linen tabby (Rast-Eicher 2012:
381; Rast-Eicher and Reinhard 1998: 288). These designs are reminiscent of the rare
depictions of clothed figures, such as the clay figurine  from Kličevac in Serbia. Made at the
same time as the Pfäffikon-Irgenhausen textile (c.1500-1400 BC) (Müller-Karpe 1980: tab.
326), this is believed to depict a female wearing a wide skirt-like cloth, decorated with
46
triangles along the belt seam and a checkered border. Recently, the technique used on the
Pfäffikon-Irgenhausen textile has been disputed as ‘real’ embroidery (Banck-Burgess 2017;
Igel 2016, 2017). Closer inspection of the fragments revealed that the decorative threads were
added during the weaving process, with the threads running over the regular tabby weave in
various directions (vertically, horizontally, and diagonally) before looping through the
warp/weft to hold it in place (Ingel 2016, 2017). Ingel (2016) conducted several weaving
experiments to better understand this embellishment method and the appearance of the final
product. These tests revealed a bright and complex embellishment method that was
dependent on the skill and knowledge of weavers to produce, rather than another craft worker
at a separate stage of the textile chaîne opératoire.
In addition to the aforementioned fragments, Denmark’s oak-log burials have revealed
embroidered designs from female burials. Blouses with heavily-embroidered neck and sleeve
hems appear in oak-coffins from several sites, including Skrydstrup (Broholm and Hald
1940: 91-4), Melhøj and Emmedsbo Mark (Bender Jørgensen 1986; Fossøy 2012: 49-50),
and Flintbek in Schleswig-Holstein, north Germany (Ehlers 1998: 422-5, tab. 29). Less
embellished designs appear in male burials as well, as demonstrated in four burials from
Denmark (Fossøy 2012: 50-3) and on a bog body from Emmer-Erfscheidenveen, the
Netherlands (Comis 2003: 194-6). A rare late Iron Age find from Nové Zamký, Slovakia,
demonstrates that the practice of embroidery continued into the Iron Age; the embroidery
patterns on this piece copied Iron Age tastes and stood out as bright red yarn against the
white linen it was sewn on (Belanová-Štolcová 2005; Grömer 2016b). These various finds
from across Bronze- and Iron Age Europe indicate a well-established and complex tradition
of embroidery-style embellishment, the production of which may have depended solely on
the weaver and their skill.
Textile Workers and Organisation
The study of textile tools is an excellent method of gathering information on the possible
forms of textile produced in the past, yet it does not provide the full context of textiles in the
communities that produced them. Close analysis of the data collected from textile tools
cannot inform archaeologists of the skill of the textile workers or whether spinning or
weaving traditions changed over time. However, work on textile fragments and the
deposition of textile tools across Europe has provided crucial information on where the
47
textile producers lived, how they worked, and what roles they may have filled in the
communities around them. The methods and conclusions of such work can inform studies on
British ancient textiles, provided caution is taken not to draw sweeping comparisons that are
ultimately unfounded.
Direct evidence of skill
The largest corpus of evidence on Bronze and Iron Age textiles in Europe comes from the
salt mines of Hallstatt in Austria. Salt and clay within the mine preserved not only organic
textiles but also wooden support beams, which have allowed archaeologists to
dendrochronology- date the various shafts and create a reasonably precise timeline for the
opening and operation of each mine area (Harris et al. 2008: 22). By 2010, over 700
individual textile pieces were excavated from Hallstatt (c.800 - 400 B.C.) and La Tène (c.400
B.C. - 40 A.D.) Iron Age contexts within these mines, the majority of which are wool
(Grömer 2013: 54-5). Examinations of these textiles revealed major differences in the quality
of the textiles through the various periods of mining, differences that indicated the use-life of
these textiles and the skills of the textile producers themselves (Grömer 2013: 56).
Several factors constitute textile quality: how the raw material was prepared; the length,
fineness, and diameter of fibres; the yarn’s diameter, direction, and angle of twist, and
orientation of fibres within the yarn; and the thread count and type of weave (Hammarlund
2004, 2005: 106; Grömer 2013: 56). By analysing and comparing such characteristics,
researchers demonstrated that raw materials from the Hallstatt period were more intensively
worked than earlier textiles and were finer, with threads averaging 0.3-0.6 mm in diameter
compared to the Bronze Age 1.2-2 mm diameters (Grömer 2013: 56; van Strydonck and
Grömer 2013: 193-4). The change appears to occur, at least in part, due to the change in cloth
use within the mines: the Bronze Age miners used woollen sacks to carry out the mined salt
and switched to hide and leather bags during the Hallstatt period (Grömer 2013: 56). As the
hard-wearing sacks made for mining became unneeded, rags became more useful and so the
roles of textiles in the mine changed. Small pieces of finer, higher quality textiles that were
no longer needed or suitable elsewhere were reused in the mines for various purposes (to tie
things together, as hand- or tool-wraps, to clean, etc.) (Harris et al. 2008: 24; Grömer 2013:
57). In addition to this, the increase in thread fineness shown in these textiles reflects changes
in textile tools outside the mines; smaller, lighter spindle whorls developed during this period
as well, indicating a change in the culture towards finer, lighter threads (Grömer 2013: 57).
48
There is also a greater variety of spindle whorl shapes during this period, most frequently
discoid, conical, and biconical-hollow topped forms made from stone or clay (Grömer 2012:
54).
Textiles from the Hallstatt period demonstrate complex weaving practices, with a far
greater proportion of twills compared to tabby weaves than seen in earlier assemblages and
higher thread counts (Grömer 2013: figs. 17 and 20), and cloth patterned in stripes and
checks (van Strydonck and Grömer 2013: 194). A wider range of dyes were used, like
madder, woad, or weld (Hofmann-de Keijzer et al. 2005), which were applied in various
ways (dyeing the fleece, the yarns, or the finished cloth) (Grömer 2013: 36). There was also
more experimentation with twist patterns in these weaves; by using threads with alternating
twist directions, weavers could create different visual effects in the finished textiles, adding
to the complexity of the final fabric (Grömer 2013: 57-8). The impressive increase in quality
and complexity of these textiles is evidence that more time and skill was dedicated to
producing such fabrics (Grömer 2005: figs 14–15, 2012; van Strydonck and Grömer 2013:
194) and does imply specialist production of textiles (Andersson 2003; Grömer 2012: 58-9,
2013: 98).
Hints of possible textile workshops, or centres of weaving, come from sites where rows of
loom weights have been found, marking out where a loom stood in use when it was
destroyed. Kleinklein in Austria produced a row of loom weights three meters long,
indicating a very large loom once stood there (Dobiat 1990); Smolenice-Molpir in Slovakia
held a similar find, identified as a specialist workshop (Belanová-Štolcová 2012).
Unfortunately, the lack of textile tools and workspaces from the Hallstatt area means it is
impossible to determine whether such specialist textile work relied on exchanging resources
(raw materials, dyestuffs, etc.) or finished products across a local or regional area, or whether
the textile workers were able to concentrate on their work or were required to take part in
other communal tasks as well  (Grömer 2013: 98).
In comparison, the later La Tène textiles from Dürrnberg salt mines and elsewhere in
Europe demonstrate that fabrics become less varied over time, although maintaining the high
thread counts and fine thread diameters (Belanová-Štolcová 2005, 2012; Grömer 2012: 60).
Of around 600 fragments excavated from the site, 25% are made with plant fibres, 60% are
tabby weave, and spin patterning is extremely rare (Grömer 2012: 46). The range of twill
types are also fewer compared to the Hallstatt period textiles discussed above, although
49
textiles still make use of similar dyes that were used in the Hallstatt mines (woad, madder,
weld, etc.), with stripes comprising the majority of coloured patterns (Grömer 2012: 46).
Some of these differences are reflected in the La Tène textile tools as well; spindle whorls
from this period average 10-30 g, far heavier than earlier spindle whorls, and these tools are
very limited in shape (Grömer 2012: 54). These whorls tend to be globular in shape and are
produced from pottery shards, a form rarely (if ever) seen in this area before this period
(Grömer 2012: 54). Using pottery shards provides spinners with a simple way of making
many whorls that are very similar in size and weight, which potentially increases the number
of spinners working in a population as well as the likelihood of these extra spinners
producing more standardised threads for weaving.
This standardisation of textile tools corresponds with the simplification of the weaving
demonstrated in the fabrics; simple, striped patterns, tabby weaves, and less complex forms
of twill that dominate La Tène textiles are easier to learn and quicker to produce (Grömer
2012: 60; von Kurzynski 1996: 35). This could suggest that textiles were produced at a larger
scale for trading purposes (Andersson 2003), although such production relies on textile
workers being able to focus on their work without interruption. Grömer believes that the
evidence from Dürrnberg supports this interpretation of specialist textile production (2012:
60-61), since the economy created by the salt mines was able to support a wide range of other
craft workshops in the nearby settlement of Ramsautal, including butchery, pottery,
glassworks, metal, and carpentry workshops as well as more general trading activities
(Stöllner 2005: 171). But is specialist production only limited to workshops as opposed to
weaving in domestic contexts? And are specialist producers defined by the quantity that they
produce, even if it is less unique, or the quality of the final product?
Craft specialisation
The term ‘specialist’ is commonly limited to a portion of the overall population who are, with
sufficient time, resources, and skill, able to produce objects of a higher quality and in larger
quantities than required in a standard household (Costin 1991; DeRoche 1995; Sahlins 1972).
To be capable of this, these craftspeople, or craft specialists, must focus more time and
energy on making their product than other activities, such as agricultural work or other
household activities (Costin 1991; Rice 1981; Tatje and Naroll 1973). Because of such
concentration on an individual craft, specialisation is commonly restricted to a small number
of areas or people rather than universally practiced, as these specialist workers must be
50
supported by a wider community to get food and housing, as well as other necessities.
Specialisation may occur in a craft that has a long and complicated production process, or
that requires special training to get right (David and Kramer 2001: 311-315). Most of the
early methods of identifying specialisation were developed for crafts which produce a lot of
production debris that can survive in the archaeological record (i.e.: metalworking, flint
working, and ceramic production) (Costin 1991: 19; Hagstrum 1985; Peacock 1982; Rice
1984). Unfortunately, the debris from textile production rarely survives deposition, which
forces archaeologists to adapt elements of these methods for use in the context of textile
production (Costin 1991: 19; DeRoche 1995 37-39; Dimova 2016: 654).
The current consensus among archaeologists is there are varying degrees of specialisation
in the archaeological record (see van der Leeuw 1977; Peacock 1982; Costin 1991).
However, the variety of specialisation depends on the period and society studied, as well as
the evidence available for analysis. Costin (1991) defines eight distinct forms of
specialisation, ranging from individual specialisation (single people or households
independently producing for locals) and increasing in complexity, scale, and production
intensity up to retainer workshops (large-scale operation employing full-time artisans in a
segregated, highly specialised setting or facility, producing only for elite patrons or a
governing institution). Costin also introduced the term corvée to describe individual craft
specialists working for patrons or governmental institutions, like palaces or ministries. The
evidence of the type of society found in the British Iron Age supports only 4 of Costin’s
possible modes of production (see table 2, below). Likewise, van der Leeuw (1977)
developed 6 similar modes of production, 4 of which reflect the evidence from the British
Iron Age (see table 2 below). In both cases, the typologies emphasize that an increase in
specialisation is matched by an increase in the amount of material produced, the complexity
of the craft processes (van der Leeuw 1984: 720-721), and the specific function of the people
and tools involved in these processes (van der Leeuw 1984; Costin 1991).
51
Table 2. Showing the evidence required by the various specialisation modes as set out by Costin (1991) and van
der Leeuw (1977)
Costin (1991)
Individuals 1. Local materials
2. Low number of tools
3. Few households with debris evidence
Dispersed workshops 1. Local and neighbouring regional materials
2. Spaced out over region
3. Few households in a community
Community specialisation 1. Local and neighbouring regional materials
2. Tools and debris for/from different stages divided between
communities
3. Division of labour
4. Most households and open areas throughout community
Nucleated workshops 1. Local to foreign materials
2. Several structures dedicated to craft production
3. Clusters of such structures
4. Labour division
van der Leeuw
(1977)
Household industry 1. Group distribution
2. Wide range of local materials
3. Few tools
4. Few uses for products
Individual industry 1. Local distribution
2. Wide range of local materials
3. Few tools
4. Many uses for products
Workshop industry 1. Several settlement distribution
2. Neighbouring region materials
3. Larger/more complex tools
4. Division of labour
5. Few uses for products
Village industry 1. Regional distribution
2. Neighbouring region materials
3. Larger/more complex tools
4. Division of labour
5. Few uses for products
The largest forms of craft work is the large-scale workshop- and industry production for
trade, which include Costin’s nucleated workshop and van der Leuuw’s workshop/village
industry definitions. Large-scale workshops are associated with quick production of
standardised and relatively simple goods that are sent directly to trade or the market; this
relies on a large and solid economy with a high demand for the goods being made and are
typically controlled by the social elite or government. Evidence of such large-scale and
controlled textile production is found in Mycenaean Palaces (c.14th and 13th centuries BC)
and Crete, where Linear B tablets describe flocks of sheep numbering around 100,000 under
the control of the palace (Grömer 2016aa: 256; Killen 1985: 108). While no palatial sites or
similar institutions occur in north-west Europe, the previously discussed Dürrnberg textiles
52
do indicate the type of simpler, standardised textile production typical of large-scale,
controlled workshops and its craft specialists. This is not to say that the Dürrnberg textiles are
poor quality. Dürrnberg’s textiles have comparable thread diameters and thread count per cm
to the Hallstatt assemblage; the differences are that tabby is easier and faster to weave,
checked fabric requires changing between the dyed and plain weft at regular intervals (also
slowing down the weaving process), and preparing to weave using alternating spin directions
also takes longer than setting up a plain tabby weave (Grömer 2016a: 257). The weavers at
Dürrnberg are deliberately choosing to produce more, faster; those at Hallstatt are choosing
to focus on quality rather than speed.
In Iron Age Britain, there is no evidence from any craft for the social stratification needed
to support such large-scale workshop modes of production (see chap. 2, Settlement, Society,
and Economy section for a discussion of social structure and settlement patterns in the British
Iron Age), such as centralised governmental bodies, institutions, or a high-ranking elite class
(Costin 1991: 8-9). Additional evidence of these modes from other forms of craftwork
(metalwork, pottery production, etc.) in the Minoan and Greek Bronze Age cultures includes,
but is not limited to: large middens of craft debris, special building complexes for craft
production connected to palaces and governmental administration centres, and larger than
average structures for large-scale production that would require many people to run and
maintain (i.e.: kilns, furnaces, quarries, etc.) (Haskall 1997: 103-5; Schoep 2010: 71-73). In
comparison, Iron Age Britain has no such middens, nor any palaces or government centres
with attached buildings for workshops, and larger than average buildings that are excavated
appear to be used for social gatherings rather than craft activities (see chap. 3). What we do
have, however, are the tools used to produce the fabrics: loom weights, spindle whorls,
needles, bobbins, and weaving combs being the most common (Barber 1991; DeRoche 1995:
37; DeRoche 2012). Spindle whorls and loom weights are not generally made of organic
materials and survive Britain’s temperate conditions reasonably well. They are also
well-recorded in modern archaeological literature as small finds, which means their exact
location in the site and its chronological sequence are known. On the other hand, weaving
combs and needles are often made of bone and antler, which is poorly preserved in certain
soil types. While these tools may be well-preserved in chalk soils, they are not in clays or
soils that cover granite, which may severely impact any widespread analysis of tools and
textile activities. Fortunately, this preservation issue can be balanced by conducting a
large-scale comparison of textile tools from a number of settlement types within a region
53
(and thus sites with roughly similar preservation conditions), identifying which tools are
preserved and which are not, and factoring this information into the analysis of the wider
area.
As an example of this comparative potential, Sellwood (1984: 238-39) noted that more
loom weights were discovered in Wessex hillforts than the neighbouring contemporary
settlements, indicating that the majority of weaving occurred within these fortified sites.
Sellwood suggested that this demonstrated the presence of specialist weavers in the hillforts
who were supplied pre-spun threads from the surrounding satellite settlements in exchange
for finished fabrics from the hillfort weavers (1984: 238-39). This interpretation has been
criticised for presenting a model of organised site specialisation in the region when it
required more evidence and analysis from further excavations and other areas (Marchant
1989). However, other studies have also demonstrated a possible division in spinning and
weaving activities between caves and open sites in the British Iron Age and Roman period
based on the number of loom weights, weaving combs, and spindle whorls found from these
sites (Bradley 1978: 65-67). Unfortunately, it is difficult to match such evidence of
differential specialisation to the modes of specialisation presented by Costin and van der
Leeuw because the forms of evidence required for different modes often overlap.
Although Mediterranean societies operated under a clearly defined hierarchy, these sites
have demonstrated that textile production both changed over time in these areas and operated
outside of domestic contexts. In Iberia, excavations at the 6th-5th century B.C. country
estates La Mata and Cancho Roano revealed a large quantity of textile tools from both sites
(76 spindle whorls and 15 loom weights, and 366 spindle whorls and 145 loom weights
respectively) (Marín-Aguilera 2019: 244). In comparison, the smaller contemporary
farmsteads surrounding these estates produced no textile tools at all (Rodríguez et al. 2004;
Mayoral et al. 2011). Furthermore, a third of all spindle whorls at Cancho Roano were made
using molds to ensure these tools have exactly the same size, shape, and weight; this
standardisation, combined with the fact that sheep bones were only found at Cancho Roano,
strongly suggests the presence of a textile workshop (Marín-Aguilera 2019: 245). Clearly,
textile workers here were limited to producing one or two types of thread in large quantities,
large enough to necessitate specialist production of spindle whorls by potters
(Marín-Aguilera 2019: 245). This is a rare instance where closer analysis of textile tools
directly demonstrated not only the scale and location of production but also the degree of
specialisation present in textile production.
54
Making spindle whorls from molds is not seen in the British Iron Age, although it may be
possible to determine levels of specialisation where there are a large number of whorls made
to the same/similar weights. Another equally interesting discovery from Cancho Roano was
where textile production took place. Commonly, it is believed that spinning and weaving was
performed in domestic contexts and out of the public eye (Budin 2013: 8–11; Burke 2016);
however, all the textile tools of Cancho Roano are clustered in hallways between rooms or
the main courtyard (Marín-Aguilera 2019: 246). This pattern has also appeared in other
European sites (Andersson Strand and Nosch 2015; Gleba and Mannering 2012) as well as
Glastonbury Lake Village, a British Iron Age excavated in the early 1900s. The location of
the majority of textile tools (including spindle whorls, loom weights, and weaving combs)
changed from roundhouse contexts to outdoor open areas during the Late and Final phase of
the settlement (c.150-50 B.C.) (Cole and Minnett 1995). The presence of tools in these more
open spaces may indicate a shift from private to public, cooperative production where several
spinners or weavers work together on the same activity. This widespread trend towards such
public craft production corresponds with the shift from complex and fine textiles that were
most likely made by a few highly skilled individuals to more standardised, less unique and
complicated textiles, most likely intended for a large number of textile workers to produce in
greater quantities.
While evidence of similar workshops is difficult to find in Britain and north-west Europe,
central Europe has produced some remarkably high quality textiles consistent with those
produced in the specialist workshops of the Villanovan and Etruscan cultures (Grömer 2016a:
252-253). In the previous Bronze Age, central European fabrics were made of simple weaves
with few patterns or dyed pieces; in comparison those from the Iron Age have a wide range
of weaves (tabby, twill, diamond, herringbone, etc.), multiple spin patterns, and yarns with
imported dyes (Grömer 2016a: 252-253). This represents not only the individual skills of
those producing such fabrics, but also communities able to produce the resources to obtain
the imported dyes and support the craftspeople both during production and while they
developed the skills needed to make such textiles (Grömer 2016a: 254). Grömer (2016:
254-5) argues that this evidence from Hallstatt and elsewhere shows the first period of
specialist textile production in Europe outside of the Mediterranean, supported by Hallstatt
nobility. Unfortunately, this same claim cannot be made in Britain as it depends entirely on
surviving textile fragments, which are sorely lacking in the British archaeological record.
Furthermore, it is impossible to determine whether textile specialists in the Hallstatt area
55
were completely exempt from other everyday activities or duties within their communities
and settlements (Grömer 2016a: 255). Without the type of literary evidence found in the
Mediterranean, the position of textile workers within their own communities is almost
impossible to ascertain.
Individual industry (also known as household industry) is the most likely form of specialist
textile production in most smaller farmsteads and villages across Iron Age Britain. This
production method results in goods surplus to the requirements of the household that can be
traded or exchanged elsewhere on a part-time basis, with the specialist splitting their time and
energy between the craft and other jobs (Grömer 2016a: 248). Nonetheless, this extra
production would still be valuable as a means of supplying the household with items it
needed but could not create for itself. Evidence of bartering (the exchange of goods of similar
value) can be found in sites across Europe and as coinage was not introduced to Britain and
north-west Europe until the Iron Age - and not used as a means of exchange until the late
Iron Age/early Roman period - textiles of decent quality could be a valuable resource for
small communities or individual households to barter with (Grömer 2016a: 248-250).
Conclusion
Overall, the evidence from across Iron Age continental Europe demonstrates a highly
developed and complex craft. By the early Iron Age, the textile chaîne opératoire for both
plant and woollen fibres was complex and time-consuming enough to require multiple tools,
lots of time, training, and personal experience to produce fabrics of decent quality. Compared
to the Bronze Age, more textiles had high thread counts and fine diameter threads and
incorporated bright colours. Some areas with highly valuable goods (such as the Hallstatt or
Dürrenburg salt mines) were able to access dyes that had to be moved over great distances
(e.g. lime bast, insect dyes, etc.). Even more common materials like flax were traded across
large distances, possibly because the textile workers who processed it were particularly
skilled or as part of a high-status exchange/gift. In addition to this, textile workers were
introduced to finer quality wool, developed new weaving patterns, and began to use old tools
for previously unrelated uses. Wool and twill weaves grew in popularity, as demonstrated by
the wide array of woollen checked and striped twills from Hallstatt, and spindle whorls
replaced the splicing of bast fibres, speeding up the thread-making process. These pieces of
evidence paint a picture of flourishing, complex, and highly skilled textile cultures and
56
workers in the early Iron Age across northwest Europe, experimenting with a relatively new
material and new spinning and weaving techniques.
By the late Iron Age, the various textile cultures of northwest Europe appear to have
adapted to these innovations and settled into accepted practices of spinning and weaving,
clearly reflected in the diminished number of twills and spin patterning compared to the
previous Hallstatt period. This decrease in experimentation and complexity is not necessarily
reflective of a decrease in skill; the threads and yarns produced in this period remain as fine
as in the early Iron Age and the textiles of late Iron Age Hallstatt and Dürrenburg have
similar thread counts, despite lacking the same variety of twills and dyed fabrics.
Additionally, this standardisation in textile production is reflected in the tools used during the
late Iron Age; spindle whorls of pottery sherds become more common across Europe (as seen
in sites like Cancho Roano, Dürrenburg, and Danebury), which could indicate that textile
workers produced whorls of similar diameters and weights themselves instead of working
with potters. While some argue that such standardisation among textile tools demonstrates a
decrease in general skill, I would argue that it is more reflective of a higher general demand
for fine fabrics. In addition to this, such tool standardisation corresponds to increased
centralised production in many areas, particularly in the Mediterranean. Unfortunately, the
lack of direct textile evidence in Britain has largely stymied research on the forms of textiles
produced, as well as standardisation and organisation of textile production. On the other
hand, recently developed methods of analysing textile tools may provide a new avenue of
research for British assemblages.
On the subject of the organisation of textile workers, production in early Iron Age Britain
depended on the individual settlements, the people within it, and the economic connections
such a site had (Grömer 2013: 53). In the late Iron Age, this changed to larger and more
communal textile production, although it is only in the Mediterranean that we can find
definite signs of textile workshops controlled by social elites. Nevertheless, it is reasonable to
expect that this pattern continues in British Iron Age sites, given the evidence from
Glastonbury Lake Village and Britain’s close contacts with Europe during this time. It may
not be possible to map the location of tools in all the sites used within this study due to the
imprecise recording methods used in some of the excavations (see chapter 4 for further
details), which will make it impossible to determine whether the shift from private to public
textile production occurred at these settlements. However, closer analysis of the
57
standardisation (or lack thereof) of textile tools used at these sites will provide some
indication as to the range of threads and fabrics such tools made, possibly demonstrating
whether the trend towards more homogenous textile production continues in the southern
British Iron Age. These methods of analysis must take into account what type of settlements
produced these tools in order to highlight possible centres of production and forms of
specialisation. An appropriate understanding of the variety and complexity of settlements and
the economies these fueled is just as important as the knowledge of a textile tool’s place in
the chaîne opératoire.
Society in the Iron Age
Defining development in Iron Age Britain and Europe
The Iron Age is a period of expansion and development across Britain, Ireland, and western
Europe. Changes in society can be traced in the material culture of local groups, new forms
of settlement, and differences in crafting, trading, and exchange networks (Bradley et al.
2016: 216). In this chapter I will discuss the general character of the Iron Age across Britain
and western Europe: how society was structured, the crafts and networks used to support the
population, how these crafts were organised, and the various types of settlement in Europe.
This will then be followed by a closer look at the settlement types used in Britain during this
period, from which were excavated the materials used in this thesis.
The Iron Age across western Europe is largely characterised by increased settlement
density, the development of larger settlements and urban (or proto-urban) centres that
stemmed from changes in society during the Bronze Age (Bradley et al. 2016;
Fernández-Götz 2018). These changes partially rely on the development of iron working and
iron tools and a rise in population; while bronze was still a prestigious metal, iron tools were
sturdier, long lasting, and more difficult to blunt, which allowed farmers to plough larger
plots and harvest more food. This change is reflected in the clearing of land and an increase
in the variety of crops grown during the Iron Age. A wider range of wheats, barleys, legumes
are discovered in early Iron Age contexts than in Bronze Age contexts from sites across
Britain and western Europe (Bradley et al. 2016). Additionally, pollen analysis indicates
larger areas of land were deforested across much of north-west Europe (e.g. Blancquaert et
al. 2012: 236-8; Dark 2006); as well as this, marsh- and up-land areas already seasonally
58
used during the Bronze Age saw the construction of high levées during the sixth-fifth
centuries B.C., suitable for more permanent settlement (Bradley et al. 2016: 216). This
pattern is not universal; parts of northern and western Britain and Ireland do not display the
increase in settlement density or the clearing of land, nor are there many artefacts in the
settlements that have been found that can accurately date these sites. Ireland in particular
suffers from a dearth of evidence, both of structural remains and artefacts, from c.800-400
B.C. (Bradley et al. 2016: 216-7). However, the area of southern Britain covered in this thesis
follows the general pattern of development in north-west Europe, so the difficulties in
determining Irish and northern British Iron Age evidence will not be a subject of discussion
here.
During the early Iron Age, a new form of settlement began to develop; previously, Bronze
Age settlements had been small and scattered about the landscape, focused mainly on
localised craft production and agriculture, with only a few regions conducting long distance
trade (Bradley et al. 2016; Fernández-Götz 2018). In the early Iron Age (c.600-400 B.C.),
large and fortified sites began to be constructed; this initial development of large centres was
followed by a period of collapse, when the population returned to a more scattered settlement
pattern (Bradley et al. 2016; Collis 2014; Fernández-Götz 2018). It is not until the late Iron
Age (c. 300-100 B.C.) that large agglomerations and fortified settlements, like the oppida,
appear again in the archaeological record. Typically, archaeologists have relied on
comparisons with the more developed Mediterranean cultures to understand the roles of, and
social structures present within, these centres (Collis 2014; Fernández-Götz 2018). This
Mediterranean model is also commonly used to determine whether a site can be labelled as
an ‘urban’ centre; academics studying the complex societies found in the Mediterranean
developed a checklist of evidence required for a settlement to be deemed ‘urban’, which was
often used on the oppidas and hillforts in north-west Europe and Britain. This checklist (after
Kolb 1984 and Hänsel 2005) included such items as:
● Settlement size ● Topographical closeness of satellite
settlements
● Density of structures ● Population count (above 1000)
● Variability of structures ● Economic diversity
● Long distance contacts ● Role as regional centres
● Intensive craft production ● Administrative and political
separation
59
● Authority and leadership ● Protection and fortification
If the cultures in Britain, Gaul, and Germany were copying those in Greece, Italy, and
Spain, then there should be little to no difference between the two areas, yet archaeologists
quickly realised that these points rarely matched the evidence found outside the
Mediterranean (Collis 2010; Fernández-Götz et al. 2014; Osborne 2005). Furthermore, some
of the features (e.g. administrative and political separation) are near impossible to determine
in cultures that lack the literary evidence of the Mediterranean (Fernández-Götz et al. 2014:
6-7). The belief that the Mediterranean was a cultural centre that was copied (poorly) by the
more ‘barbaric’ northern cultures was deeply entrenched in archaeological theory, even with
the mounting evidence against this model (Fernández-Götz et al. 2014; Fernández-Götz
2018). The common consensus now is that the urban development and expansion in
north-western Europe and Britain occurred in parallel (as reflected by the movement or
transmission of people, artefacts, and ideas), rather than that the long-standing connections
and networks with the Mediterranean influenced north-west Europe (Collis 2014;
Fernández-Götz 2018). This allows the ongoing investigation of these societies and
settlements to remain aware of contemporary influences and interactions from the
Mediterranean with the north-western Europeans (and vice versa). Furthermore, it allows the
definition of the roles and structure of settlements, as well as the societies that inhabited
these, to be based first and foremost on the evidence from these sites rather than a checklist
from other cultures and civilisations. With this in mind, what forms of society and settlement
existed across Iron Age Britain and north-western Europe?
Iron Age society and organisation
In discussing crafts in a society, it is important to first determine whether this society can
support craft workers in any capacity. The population of this society and the resources
available to it must be large enough to allow some of its members to focus their time, energy,
and skills on their craft work, even for a short length of time. Once the population and
resource size is estimated, it is then possible to determine how a society was organised, how
many craft workers it could support, and how its craft production was organised. It is
important to stress that both methods discussed here can only provide a rough population
count and that these techniques should always be considered against the archaeological
evidence available, rather than be favoured over it.
60
There are two main methods of determining population sizes in ancient settlements. The
first measures the density-of-habitation coefficient, based on either the number of people per
square area (acres, hectares, etc.) or the number of people per household and multiplied per
dwelling (Zorn 1994: 32). The second measures the quantity and availability of natural
resources in an area, such as:
● Water: annual rainfall, number of
springs and natural water sources
nearby, etc.
● Food: number and quality of arable
fields, yield of various crops,
consumption of animals, etc.
● Natural resource exploitation:
hunting opportunities and yield,
foraging opportunities and yield,
marine resources and yield
● Consumption: the amount of food
and water required for human and
animal population
● Storage capabilities: amount of
water storage, food storage, animal
stockade size, herd size, etc.
orn 1994: 33)
This method is complicated and frequently relies on ethnographic evidence that may not be
best suited to interpret the Iron Age evidence. Unfortunately, this method is not suitable for
this thesis, in part due to time constraints and in part because of the lack of environmental
data in a significant portion of the excavation reports, particularly those written prior to the
1960s.
The density-of-habitation method has seen several critiques over the years (see Hopkins
1985: 153-7 and Frick 1985: 141-59 for further details) but there are two key issues in
applying this method to Iron Age sites in the study area of this thesis. Firstly, the
area-estimate system was developed for pre-Modern settlements in the Middle East that
varied very little in population size and density (Broshi and Gophna 1984: 41-42; Stone
1987: 2-3; Zorn 1994: 32); Iron Age settlements have periods of expansion and decline,
during which the population and number of active dwellings fluctuates. This is not accounted
for in the area model, rendering it unsuitable for such sites. Secondly, some Iron Age sites
(particularly fortified settlements) have large areas that appear to be purposefully
uninhabited, possibly as stock enclosures. So while estimates based on site area may provide
a fast and simple way to estimate populations, the risk of over-estimating a site’s size and
importance is too high to employ this method. On the other hand, dwelling-based estimates
61
can provide room for the fluctuations in population that are common in Iron Age sites
throughout Britain and north-west Europe.
Unfortunately, household size is tricky to determine. Archaeologists employing this
technique often rely on ethnographic analogies, which can be difficult to find, and these
estimates rarely take into account that households in the past often had several generations
under one roof (Zorn 1994: 32-33). Another way of determining a household’s size is to
estimate the size of an average family, with most estimates ranging between 3.5 and 8 (Zorn
1994: 33). Despite these issues, the flexibility of the dwelling-based estimate makes it best
suited to working on Iron Age settlements and therefore, this is the main method I will be
using in this thesis. Given that extended families can occupy the same domestic space, I
believe it appropriate to place the average family size at 7 individuals: two parents, three
children, and two extended family members. The minimum population size will be estimated
in the methodology chapter below.
Settlement in the Iron Age
Settlement starts with individual households, which inhabited roundhouses during the Iron
Age. Roundhouses were typically constructed with wattle-and-daub walls and a deep-pitched
roof and were the only form of dwelling in Britain, Ireland, and northern France (particularly
Brittany) (Fernández-Götz et al. 2014; Harding 2009; Webley 2018). Separate, smaller
buildings or structures could also belong to a household, such as granaries, storage pits,
single posts, and so on (Webley 2018). The main roundhouses rarely had internal divisions
built into the structure, although it is possible that different sections of the internal area had
distinct roles and were treated differently by its inhabitants. Excavations at the Late Bronze
Age/Early Iron Age sites of Longbridge Deverill, Wiltshire (Chadwick Hawkes 1994),
Pimperne, Dorset (Harding et al. 1993), and Dunston Park, Berkshire (Fitzpatrick 1994;
Fitzpatrick et al.1995) have demonstrated remarkably clear clusters of artefacts (pottery,
struck flint, etc.) that are restricted to the southern half of the buildings, close to the entrances
(Webley 2007: 127-8). The excavators suggested this showed the delineation of the
roundhouse into a sleeping area in the northern half and an area for domestic activities in the
southern half (Chadwick Hawkes 1994: 67), although they did also note that the deposition of
artefacts could have occurred when the roundhouse was abandoned and not when it was in
regular use (Fitzpatrick et al. 1995: 87).
62
Nevertheless, modern societies that still use roundhouse structures without internal walls or
other physical dividers also separate internal space into areas of activities or by social rank.
The Mongolian ger (yurt) is divided into several areas depending on gender and social status:
the eastern half of the ger is the men’s area, the western half is the area for women and
children, and the northern part of the ger is the location of the altar and the seat of the guest
of honour (National Geographic 2020). Passage around the ger is also strictly ordered; upon
entry everyone walks clockwise around the central fireplace, passing in front of the male
side, then the altar, and finally passing the women to leave (Kemery 2010). Similar social
divisions are recorded from other yurt cultures, such as the nomadic Afghan, Khazak,
Kirghiz, Turkomen, and Uzbek (Kemery 2010). As previously mentioned, the archaeological
evidence from the British early Iron Age does not demonstrate subdivision based on gender
or social standing but on activities. These activities may have included specialist craft work,
such as textile production; however, the household would have to be able to support the craft
worker. Thus, it becomes important to understand the settlement that this household inhabits
and relies upon for its resources.
While roundhouses characterised settlements in Britain and northern France, other areas of
northwest Europe built rectangular houses, often with internal divisions to separate the
household’s living quarters from the area where livestock were housed (Webley 2018). It is
possible that patterns of human activity in domestic spaces would differ in these settlements,
where division between human and animal spaces are clearly visible, from those seen in
roundhouses. On the other hand, the human area of these rectangular structures often lacks
physical divisions, similar to roundhouse interiors.
Settlements across Britain and western Europe ranged from small, individual
home-/farm-steads and enclosed lowland settlements to large fortified settlements, like
hillforts, promontory forts and oppidas. The larger fortified settlements developed in the early
Iron Age and represent a new method of organising society that was not present in the
previous Bronze Age or Neolithic (Cunliffe 2005; Harding 2012). The smaller settlements
often vary in form or size across regions depending on the environment and society they
belong to; in comparison, hillforts are not restricted to purely defensive roles (Harding 2012).
Across Britain and Europe, hillforts of all sizes frequently exist alongside smaller defended
or undefended homesteads and open ‘agglomerations’ (Fernández-Götz et al. 2014: 3;
Forde-Johnston 1976: 12). These smaller satellite settlements could have filled several roles
63
depending on the nature of the hillfort; they could provide food and materials to a fort with
permanent occupation, or be the summer farmstead of a hillfort with seasonal winter
occupation, or they could be the permanent homes while the hillfort provided a refuge in
times of crisis (Cunliffe 2005; Cunliffe et al. 2006). The adaptability of smaller settlements to
provide resources or permanent homes to hillforts presents a challenge to archaeologists; in
order to correctly identify how the society was organised, where centres of production and
areas of supply were located and so on, we must understand the role of all settlements in an
area.
Over the last fifty years, excavations of settlements of all sizes have increasingly focused
on discerning the function and economic activities at these sites, and how these relate to
larger regional patterns of occupation and craft production (Cunliffe 1988; Fasham 1985;
Rady 2010; Wainwright 1979). Most notably, the Danebury Environs Program (Cunliffe and
Poole 2000a-g) compared recent excavations of Danebury hillfort and surrounding
settlements to determine how the presence of this large centre of production affected the
communities around it. These excavations and studies demonstrate that there are broad
patterns to settlement types and uses across the southern regions in Britain, which will be
explained further below.
An estimated six out of ten British hillforts enclose less than 1.2 hectares (3 acres)
(Forde-Johnston 1976: 12; Harding 2012: 9) and are only distinguishable from defended
homesteads by their disproportionately large earthwork fortifications (Harding 2012: 9).
Three out of ten hillforts enclose between 1.2 - 6 hectares (3 - 15 acres) while the remaining
one of ten enclose areas over 6 hectares (15 acres) (Forde-Johnston 1976: 12; Harding 2012:
9). As Harding (2012) points out, it is reasonable to expect the social roles of these hillforts
to change according to their size and complexity, although there is still disagreement
regarding what sort of roles hillforts held around the country (see Cunliffe 1995; Payne et al.
2006; and Oswold et al. 2006). This is one of the primary reasons that hillforts are included
in this study alongside the smaller, potentially undefended, settlements. However, these
hillforts are dwarfed by the ‘oppida’ of central Europe, which could cover several hundred
hectares alone (Fernández-Götz et al. 2014; Fernández-Götz and Krausse 2013).
Few sites are classified as oppida in Britain as they do not reach the large sizes of the
European sites, the expansive occupation, or the multitudes of craft and trade activities that
are used to define oppida sites in Europe (Fernández-Götz et al. 2014; Fernández-Götz and
Krausse 2013). ‘Oppida’ is a Roman term, used to describe large non-Roman tribal
64
settlements that acted as centres of administration, trade, religion, and/or craft. Proving that
these sites held such a role is difficult and much discussed by ancient historians,
archaeologists, and anthropologists (Fernández-Götz et al. 2014: 6; Smith 2007). Previously,
archaeologists have focused on proving a site has urban characteristics (such as monumental
architecture, presence of foreign goods, etc), although this approach has been criticised as
insufficient (Fernández-Götz et al. 2014: 6). Current method of determining a site’s role
follows Smith’s (2007: 4) split method where centres are defined by their ability to affect a
wider hinterland, whether socially (via a large or permanent population) or functionally (as a
location for institutions or activities) (see also Smith 2010).
Clearly, hillforts and oppidas have a potential role as centres of control over regions,
society, or surplus and/or foreign goods, trade, and craft activities. However, British hillforts
vary in character, size, and use across the country - some lack evidence of occupation, some
appear intermittently occupied, and others display high levels of occupation, including
planned storage and residential areas (Cunliffe 2005). Therefore it is vital that this study gets
a more rounded understanding of the settlement uses by comparing data from these sites with
other settlements in each region. This would ensure that the data from each area is presented
and analysed in their local context, which would also allow any differences between regions
to be more easily identified and compared.
South West
The South West is a region with unique character; surrounded on three sides by sea,
containing various mineral resources, fertile valleys, at least six high moorlands (Dartmoor,
Exmoor, Bodmin, Hensbarrow, Carnmenelis, and Penwith), and cut off from the rest of the
country by the Somerset fenlands (Cunliffe 2005: 275). There are several distinct settlement
styles that developed from the late Bronze Age onwards in Devon and Cornwall to take
advantage of the various climates in these regions (Simmons 1970; Silvester 1979; Cunliffe
1991: 41-9).
The high moorlands of Dartmoor, Exmoor, Bodmin, and Penwith are typically
characterised by groups of dry-stone walled huts within an enclosing wall (see figure 15
below), known as pounds (Henderson 2007: 109). Grimspound on Dartmoor (Pattison and
Fletcher 1996: fig.4.6) is a typical example of the pound settlement style. Grimspound’s
65
ovoid wall encloses roughly 1.45 ha. and twenty-four huts; the finds were scarce and
consisted mainly of coarse pottery and broken querns (Henderson 2007: 109). Settlements
like Grimspound are predominantly clustered on the southern area of these moorlands,
frequently near water sources; the current theory is that these enclosed settlements had a
largely pastoral role, with only a little cultivation taking place within the pounds themselves
(Cunliffe 1991: 41).
Similar settlements, known as ‘rounds’, were also constructed across Devon and Cornwall
up to the 1st century BC (Cunliffe 1991: 41-42). Like pounds, most are circular and under 1
hectare in size. These sites differed from pounds in two main ways: firstly, these are not
restricted to moorlands (although most were constructed on upland areas); secondly, these are
enclosed by a single ditch and bank structure rather than a stone wall (Cunliffe 1991: 41-42).
An example of a round included in this study is the Cornish site of Trevisker. Artefacts from
rounds are scarce and are predominately pottery for cooking and storage purposes or querns,
all commonly made on or close to each site with locally sourced materials (ApSimon and
Greenfield 1972). No evidence of craft activities (textile production, metalworking, etc.) or
items of long-distance trade are usually excavated from these sites (Cunliffe 1991: 41-42),
lending weight to the theory that these are largely agricultural settlements that bring in
pottery, metal items, and any other necessities from local trade.
The last small settlement types are single enclosure settlements and unenclosed huts. Most
of the single enclosure sites are generally clustered on the wetter western edges of the
moorlands and are characterised by small stone enclosures attached to individual huts. These
could be for livestock or cultivation, suggesting mixed agricultural practices in these sites,
although no in-depth studies have been conducted in these areas. Finally, the unenclosed huts
are often associated with stone-walled field systems and clustered along the drier eastern
moorlands (Henderson 2007: 111). These field systems generally lie on the most fertile soils
in the area, suggesting their primary use was for arable farming. As Cunliffe (1991: 45)
noted, the field systems in most of these sites could not have supported all of the huts at the
same time, if these buildings had been contemporary domestic spaces; therefore, these
farming sites also represent a long-lasting community. Unfortunately finds are scarce, as with
the majority of these moorland settlements; querns and cooking or storage pottery are the
most common artefacts, suggesting little outside trade in non-organic goods (Henderson
2007: 111-112). This scarcity of evidence makes it difficult to accurately chart the
development of such sites, although a general date of the 1st millenium can be applied from
66
ceramic remains (Henderson 2007: 112). Open settlements included in this study are Carn
Euny (Cornwall), Dainton (Devon), and Allard’s Quarry (Dorset).
Outside of the moorlands, multiple-enclosure settlements are the most common sites, which
developed around 500-400 BC (Cunliffe 2005: 280). Excavations suggest these supported
predominantly pastoral farmers; they have large open spaces within their enclosures, well
suited for corralling livestock and with little to no archaeological features or finds, and are
frequently found on hill slopes or ridges close to water sources and field systems. Some, such
as Milber Down in Devon (see figure 15 below), are defended by multiple ditches and banks
similar to developed hillforts (Johnson and Rose 1982; Cunliffe 2005: 281-3). However, in
both cases their entrances remain relatively simple and uncomplicated, angled to face the
nearest water source. The most common finds from these settlements are utilitarian tools and
objects like cooking and storage pottery, querns, and agricultural tools (Johnson and Rose
1982; Cunliffe 2005: 281-3), providing strong evidence that there were little to no craft
activities at these sites. The only multiple-enclosed site included in this study is the Devon
site Milber Down.
Courtyard settlements are mostly found in the west Cornwall upland areas and are
characterised by their unusual construction. A central, paved ‘courtyard’ with rooms leading
off of it, encased in a thick dry-stone wall (see figure 15 below; Cunliffe 2005: 285-7). While
these courtyards are commonly thought to be open to the sky, Wood (2001) has proposed
they were in fact roofed, which would suggest the ‘courtyards’ were actually a communal
living space with more private rooms attached. Cunliffe (2005: 285) notes that this is
possible, although he questions whether they could have easily acquired the ‘substantial
timbers’ needed for such roofing. These sites are usually unenclosed and found with
associated field systems, such as those preserved at Chysauster (Smith 1996). The few
exceptions include Goldherring and Porthmeor; the former being a single homestead and the
latter enclosing a small community (Guthrie 1969; Hirst 1936; Cunliffe 2005:  287-8). Both
the well excavated sites of Chysauster and Carn Euny have signs of earlier occupation dating
from the Bronze Age, although the courtyard houses at Carn Euny date to about 700 - 600
BC (Cunliffe 2005: 286). This early date is highly unusual, with the common consensus
being that these settlements are a first century BC phenomenon that continue well into the
Roman period, as at Chyauster (Quinnell 1986: 120). Artefacts from these sites are similar to
the assemblages from the multiple-enclosure settlements: large quantities of domestic pottery
67
with few (if any) craft tools or debris and foreign traded goods. Again, there is little evidence
of textile production at these sites.
Promontory forts (also known as cliff castles) are fortified sites found on headlands that jut
out into the sea. Their landward approaches are frequently protected by a series of banks and
ditches, similar to the complex entrances of developed hillforts elsewhere in the country (see
figure 15 below). The similarities between these sites and cliff castles in Brittany has
attracted a lot of attention from the archaeological community over the decades (Cunliffe
2005: 288). Pottery made with similar production techniques and decoration styles of native
Brittany ware have also been found in the Rumps and Castle Dinas, although current
consensus holds that these are the results of cultural contact or movement of individual craft
specialists from northern France to Cornwall rather than an indication of large-scale
immigration from Brittany (Cunliffe 2005). Most excavated cliff castles have produced
domestic waste and the remains of small circular houses (Brooks 1974; Nowakowski 2004),
suggesting at least partial occupation; on the other hand, these sites are usually remote and
exposed to weather fronts and rough seas. This has led some authors to believe that these
settlements were more likely to hold symbolic significance than house large communities or
elite families (Sharp 1992; Herring 1994; Cunliffe 2005: 289). Cliff castles included in this
study are the Rumps and Trevelgue Head (Cornwall).
68
Figure 15. Examples of typical settlements in the South West. A) Trevisker, enclosure. B) Carn Euny, open
settlement. C) Milber Down, multiple enclosed settlement. D) Chyauster, courtyard settlement. E) Trevelgue
Head, cliff castle. After Cunliffe 2005.
Central South
Enclosed and unenclosed settlements are found throughout the Central South area, which
covers Dorset and Hampshire in this work. Their size and form depend largely on the
environment and the community’s needs, although roughly rectangular shapes are common
(Cunliffe 2009). Eldon’s Seat (Dorset) is an example of such an enclosed settlement,
although it is unique for two main reasons. Firstly, the archaeological evidence suggests it
69
was a seasonally occupied settlement and secondly, that its purpose was primarily for
part-time craft specialisation (Cunliffe and Phillipson 1968). From this settlement,
craftspeople produced lathe-turned Kimmeridge shale bracelets en masse for distribution
across the south coast; occasionally, the cores of these lathe-turned bracelets would be turned
into spindle whorls (Harding 2014: 99). However, none of these lathe cores-turned-spindle
whorls are present at Eldon’s Seat itself, instead appearing at sites from other counties along
the south coast; this suggests that these items had more value as gifts or tradable goods.
Banjo enclosures are one of the most common small settlements in the Wessex area,
although they are also found across the country (Cunliffe 2009: 154). These settlements are
characterised by a single ditch and bank defence that surrounds the site and extends into an
elongated entrance, ensuring that the settlement can only be approached from one direction.
It is not unusual for these ditches to connect to the ditches of a wider field system, such as at
Bramdean, which provides some insight into land ownership during this period (see figure 16
below). Banjo enclosures tend to be small, usually containing a single structure or farmstead,
although the uses of these settlements are wide ranging. Some include evidence of rebuilding
or maintainance of roundhouses and 4-post structures that demonstrates long-term occupation
and agricultural activity, such as at Little Waltham, Essex (Cunliffe 2009: 154). Others have a
wide range of finds and craft debris that indicates a high level of craft activity, as seen at
Gussage All Saints (Wainwright 1979). Such a variety of activities and lengths of occupation
reflects the adaptability of this settlement type and its suitability as both farmstead and
possible production centres. Gussage All Saints (Dorset) is the only banjo enclosure in this
study.
Similar to banjo enclosures, ditched enclosures can surround a small farmstead with a small
ditch and bank; however, these lack the unusual entranceway and associated field systems of
the banjo enclosure. These sites typically enclose a large area and contain scattered storage
pits and 4-posted buildings, widely agreed to be granaries (Wainwright 1969). This
settlement type centres around Dorset and is prominent from the early Iron Age through to
the 1st century AD, represented at Tollard Royal and the early phase of Gussage All Saints
(see figure 16 below).
Earthwork-enclosed settlements are typically found across Southern Britain and date to the
Middle Iron Age. Several sites in Hampshire’s Danebury Environs Project have been
particularly well excavated and recorded, including Old Down Farm, Houghton Down,
70
Farley Mount, and Winnell Down (see figure 16 below). Interestingly, the settlement at
Winnell Down had outgrown its initial enclosure by the Middle Iron Age and continued to
grow outside of its original bounds (Cunliffe 2005: 243). Given this continued expansion
combined with the common nature of these settlements, Cunliffe suggests they represent a
particular form of socio-economic structure that differed from areas without
earthwork-enclosed settlements (2005: 244). However, without a large-scale cross-county
comparison of sites and settlement types, it remains unclear what differences in society or
economy could be represented here. Enclosed settlements included in this study are Kennels
Farm, Winnall Down (Hampshire), and Thanet Earth (Kent).
The size, placement, and role of hillforts changes throughout the Iron Age more noticeably
in Wessex than elsewhere in the south of England (Cunliffe 2009; Harding 2012). However
this may be because more hillforts were excavated in this area than any others, making this
region over-represented in archaeological literature compared to others (i.e.: the Danebury
environs project). In the early Iron Age, two types of hillfort are built along the chalk
highlands of north Wessex:
1) Well fortified hillforts, covering 6 hectares or less. These typically produce evidence
of dense interior activity; see Bury Hill (Hampshire) and Chalbury Camp (Dorset).
2) Hilltop enclosures, covering more than 6 hectares (see figure 16). Usually displays
scarce evidence of occupation and has simple single ditch and rampart defences in this
period; see Danebury (Hampshire) and Maiden Castle (Dorset).
Around the mid 4th century BC onwards, the majority of hillforts were abandoned
throughout Wessex and the rest of the country (Cunliffe 2009: 156; Harding 2012). Those
that remained in use grew in size and added multiple lines of defences, complicated
entrances, and occasionally created more than one entrance (Cunliffe 2009: 156). This led to
archaeologists coining the term “developed hillforts” to describe these later sites, although
Harding argues this added an implication that these sites had become centres of community,
economy, and authority for local chiefs or tribes (2012: 203). This theory is difficult to
support with the archaeological evidence, even at a thoroughly excavated site such as
Danebury (Harding 2012: 203).
71
Figure 16. Examples of typical settlements in the central South. A) Bramdean, banjo enclosure. B) Tollard
Royal, ditched enclosures. C) Winnall Down, earthwork enclosed settlement. D) Maiden Castle, hillfort. After
Cunliffe 2005.
On one hand, these ‘developed hillforts’ have the same material culture as the other
surrounding settlements, demonstrating that the same activities occurred at both types of sites
and no elite activities (i.e.: gold-smithing, sword manufacturing, etc.) were confined to any
type of hillforts (Harding 2012: 203). On the other hand, hillforts do contain more non-local
materials than the smaller settlements, suggesting a higher degree of long distance trade
between other British areas or with European sites. Cunliffe (1995: 93) discussed the range of
non-local raw materials (i.e.: currency bars from Northamptonshire/Forest of Dean,
Kimmeridge Shale from Dorset, copper alloy from south-west England/Wales) in Danebury
as the result of the site’s central role in a redistribution-exchange network. Hillforts included
in this study are Hod Hill, Hengistbury Head, Maiden Castle (Dorset), Blackbury Castle,
Mount Batten (Devon), and Danebury (Hampshire).
72
South East
The regions of Sussex and Kent have a limited range of settlement types compared to other
areas along the south coast. Whereas other regions have a wide range of site types that rarely
alter in form over the Iron Age, South East sites are frequently characterised by long-lived
settlements that change their form several times throughout this period (Cunliffe 2005).
Additionally, this region contains notably fewer hillforts than the others in this study,
particularly in Kent. Excavation is complicated, particularly in Sussex where ploughing has
damaged a large number of sites; many smaller sites are found on ridges or hill spurs along
the productive farming region of the South Downs (Bedwin 1984: 46).
The majority of settlements in the South East are palisaded enclosures, enclosed by
rectangular palisades rather than the rounded ditch or ditch-and-bank so common in Wessex
or the South West. Most appear to enclose single farmsteads, although Cunliffe noted that
palisades were not necessarily limited to small sites since some transitioned to settlements
fortified by banks and ditches of hillfort proportions (2005: 216-217). The Caburn (East
Sussex) follows this pattern, with an initial palisaded enclosure followed by a late Iron Age
hillfort (see figure 17 below); the pottery recovered from the palisade period huts and
trenches dating these features to the 7th and 6th century BC (Wilson 1938; Cunliffe 2005:
215). There are no palisaded enclosures included in this study, as the artefacts from the
Caburn could not be tracked down and examined.
The other main form of settlements in the South East appear to be open settlements,
although these are difficult for excavators to find since there are no crop marks to highlight
where they are. Some sites go through a phase of open settlement before being enclosed,
which tends to be early in the site’s history. For example, Slonk Hill’s (West Sussex) open
settlement phase is dated to the 6th and 5th centuries BC and is later enclosed by earthwork
defenses (Cunliffe 2005: 223). In Kent, the only 2 sites of this type are Highstead and
Underdown Lane, Eddington. Cunliffe suggests that this form of settlement may be more
common across the country given the number of examples of open settlements excavated in
the Sussex region (2005: 223). However, there could be other reasons for such a difference;
for example, special forms of settlement could be a deliberate method of establishing or
displaying social identity (i.e.: pounds in Devon and Cornwall). Or they could be a way to
make the most of the local environment (i.e., cliff castles). On the other hand, open
73
settlements may have sprung up to quickly take advantage of a resource or answer an
unexpected need (i.e., providing a stopping point on a new trade route). While it is currently
near impossible to determine whether any of these are the reason for this form of settlement
pattern, I believe it is important to keep these possibilities in mind as we examine the
evidence.
Only two hillforts have been positively identified in Kent, Oldbury and Bigbury (formerly
known as Bigberry). Bigbury is classified as an oppida due to its large size; its original
enclosures were expanded twice with large annexes and it is enclosed by a large ditch and
rampart fortifications (Cunliffe 1991). The finds from Bigbury demonstrate that this was a
significant settlement that relied primarily on agricultural activities (Holman 2005: 24) rather
than as a centre of craft production. Otherwise, there are a number of hilltop enclosures that
have bank and ditch defenses (such as St Thomas Hill, Canterbury, and North Foreland,
Broadstairs), although there is a second possible oppida site at Folkestone. This recently
discovered site, named East Wear Bay, is considered an oppida due to the extent of
occupation and the number of exotic artefacts uncovered from the site (Colsen 2013). These
artefacts demonstrate the high level of European trade conducted at the site and its standing
as an important centre of trade and power in this region. Unfortunately, East Wear Bay was
discovered during rescue excavations prior to development; because only a small area could
be excavated, the size, character, or defenses of the settlement could not be determined
(Colsen 2013).
74
Figure 17. Examples of typical settlements in the South East. A) Caburn, palisaded enclosure. B) Oldbury,
hillfort. After Cunliffe 2005.
Conclusion
The settlement types discussed above cover a wide variety of geographical areas and it is
reasonable to argue that their functions are just as wide-ranging to answer the needs of the
communities that built and occupied them. For example, the features and artefacts uncovered
during excavations clearly demonstrate that it would be misleading for archaeologists to say
that hillforts were built for the same reasons and occupied in the same way across the south
of Britain. Equally, the type of extensive occupation and crafting activities present in the
Cornish cliff castles are either very limited or in some cases completely absent in Kent. Even
the smaller, agricultural homesteads or small villages demonstrate regional differences in
structure, form, and use. The rounds and pounds of the South West have very little evidence
of craft activities (such as debris from smelting or forging), whereas the banjo enclosures of
Dorset and Hampshire frequently produce evidence of on-site production of pottery or metal
items.
The evidence of trade is varied at these sites as well. In the South West, smaller settlements
that focus on agriculture or animal husbandry often lack artefacts from elsewhere in Britain
or further afield. In comparison, fortified sites from this region appear to be centres of
trading, with on-site processing of raw materials (i.e.: smelting ore) into tradable goods and
pottery that is heavily influenced by continental styles. This pattern can, to some extent, be
traced in the Central Southern region of Dorset and Hampshire. Certainly, large fortified
settlements near the coast (like Hengistbury Head or Maiden Castle) share similar finds and
features of continental trading and on-site crafting; however, the smaller settlements
surrounding them do contain more evidence of crafting and regional trade than similar sites
in the South West. Gussage All Saints in Dorset, for example, revealed substantial amounts
of metalworking and ceramic production debris, as well as pottery that had to be brought in
from Hampshire or Wiltshire. This difference between the connections of settlements in the
South West and the Central South could be a reflection of their geography - the South
Western peninsula is relatively isolated, with no counties connecting to it, while Dorset and
Hampshire are surrounded on three sides by other regions. It is not possible to reach the sea
from further north or to travel between the South East and South West without going through
the Central South. Arguably, this would expose the population of this region to a greater
75
number of influences and styles than in the South West, which could explain the difference
between the diversity of artefacts in both regions.
76
3. Methodology
During this study, I gathered the following sets of data: the location, number, and type of
textile tools associated with each settlement phase and the dimensions and weights of two
specific types of tools (spindle whorls and loom weights) from each site (ordered by county).
I used the information from each site’s excavation record to identify the phase each find
belonged to. Unfortunately, early excavations frequently have issues with dating the tools due
to incomplete excavation records, so only the tools with a confirmed connection to the
relevant phase were used here. Sites with five or more tools were selected for this study, as
these would provide a suitably large data pool for comparisons to be made between sites. I
recorded the tool’s dimensions and weights directly from the artefacts. I omitted weaving
combs, needles, and bobbins from the data collection due to time constraints. The
construction of the sites have not been discussed beyond whether or not it had standing
structures on it. The tools have been recorded and labelled as follows: loom weights (LW)
and spindle whorls (SW).
Issues and resolutions
I encountered several issues in collecting and compiling this data. First, not all of the textile
tools survived the archaeological record, with organic tools particularly at risk of
deteriorating. Second, some site reports particularly from the early 20th century contained
little information regarding textile tool numbers and locations. Thirdly, tracking down the
tools from these sites was difficult and occasionally impossible. Finally, those textile tools
that did survive were frequently broken and missing pieces. Because of the varying nature of
these issues, I had to be adaptable in responding to and resolving these various challenges.
The issue of preservation could not be mitigated. Organic tools like bone spindle whorls,
bone needles, antler weaving combs, and so on, are simply unable to survive the acidic soils
present in areas such as Devon and Cornwall. What tools have survived and been excavated,
like spindle whorls and loom weights, cannot be considered a complete and accurate
representation of all the tools used at the respective sites. Organic textile tools did survive in
the chalklands of Dorset and Hampshire with little damage from deposition and in
considerable numbers from particular sites, such as Danebury. However, these have not been
included in this thesis due to the previously mentioned scarcity of the tools from other
77
counties; any work on organic textile tools risks being applied to those counties without any
surviving tools, regardless of the possibility that the weaving cultures and traditions could be
very different in these areas. A more thorough and critical overview of organic weaving tools
in Dorset and Hampshire may indeed provide useful information concerning the weaving
culture and tool choices of weavers in these regions; however, without the chance to compare
and contrast with other regions, such work has no place in this thesis.
The second challenge in compiling my dataset was that sites excavated in the early 20th
century often had vague excavation records that lacked the context, phasing, or exact number
of spindle whorls or loom weights. Many sites had to be excluded from this study due to the
lack of information, and others like Milber Down (Devon), Allard’s Quarry (Dorset),
Cissbury Rings, Slonk Hill, the Caburn, and the Trundle (East and West Sussex) merely
mentioned that ‘a number’ of these tools were excavated, making it impossible to judge the
exact quantity of tools that should be in storage or their state of preservation. This made it
impossible to effectively map the context of each tool at every site, so I was unable to
determine patterns of use across each county according to tool deposition. However, it was
possible to gain more information on the organisation of textile workers and the character of
textile production at each site by comparing the textile tool assemblages and the information
obtained through my analysis of each tool to the excavation reports and remains of other craft
production at these settlements. So while precise excavation records for the textile tools at
each site would have been useful, I was able to determine the nature of textile organisation
through slightly different means.
This leads to the third issue: tracking all of the tools excavated from likely sites was
difficult, since museums frequently lacked complete collections from certain sites. For
example, Hengistbury Head museum only contained 5 spindle whorls and no loom weights
(pers. experience), although many more were excavated in the 1970s (Cunliffe 1978).
Fortunately in this case, Cunliffe provided all the data required for this study via microfiche.
Many other sites excavated in the 1970s and 1980s recorded spindle whorl and loom weight
data in this way, and so I was able to include these in my datasets. Unfortunately, many Iron
Age sites in both East and West Sussex appear to have misplaced the textile tools discovered
at these sites (Lewes Museum pers. corr.). Reports from the 1920s to the 1940s on
excavations at the Caburn, Slonk Hill, and the Trundle noted that large quantities (although
no precise number) of spindle whorls and loom weights were discovered, yet no further
78
information was given. I was unable to track down any spindle whorls or loom weights from
the Caburn or the Trundle, although the numbers of tools from these hillforts apparently
rivalled Danebury and Maiden Castle (Curwen 1927, 1929, 1931). The tools were not present
at the British Museum, nor the local museums of Lewes, Brighton, Chichester, or Fishbourne;
it is possible that some of the tools were given away by Curwen to friends or colleagues (as
antiquarians occasionally did) but there are no records of this either. Since the excavators
considered them unimportant, no information on the size or weights of these tools was noted
and subsequently, all this data is effectively lost. Furthermore, no recent excavations in East
or West Sussex have met the requirements of this study and so both of these counties must be
omitted from this study.
Finally, many artefacts available for analysis were incomplete due to damage obtained
either during use, during deposition and preservation, or during the time on display and/or in
storage. Few loom weights were undamaged, although many of the spindle whorls appeared
unscathed. As damage to the tools affected their dimensions and weights, I was forced to
reconstruct their data based on what I was able to record. I used tools with 50% or more of
their form and material left to calculate the likely data as closely as possible to their original
size and weight. Those tools that had under 50% of their original form or material were
omitted from the study. Therefore, the analysis and conclusions drawn in this thesis must be
viewed as the average or most likely form of textile production and organisation for weaving
techniques that use these tools.
Site selection and study area
The sites used in this thesis were chosen according to four clear parameters. These sites had
to have:
● More than five textile tools per site (more than four in Kent), to create a suitably
large database per site that could both highlight patterns in tool dimensions or
weights and calculate the average measurements for spindle whorls and loom
weights.
● Textile tools that dated to the Iron Age, regardless of the size of the settlement
during this period.
● Clear excavation records, in order to determine the character, population count, and
other craft production that occurred at each site.
79
● Tools that could be measured and recorded in person, in order to collect precise data
and evaluate the conditions of damaged tools before reconstructing the weights and
measurements as close as possible to the original tools.
Because a variety of settlement forms were required for this thesis, I did not restrict my
dataset according to the type of site (fortified, enclosed, etc.). My limit for the number of
sites was a maximum of 5 per county; this allowed me to collect data from various types of
sites and from across each county (see figure 18 below), so that there was an appropriately
large dataset to note similarities and differences between assemblages.
Figure 18. Map displaying the study area and the sites used in this thesis. 1- Carn Euny; 2- Bodrifty; 3- The
Rumps; 4- Trevelgue Head; 5- Trevisker; 6- Mount Batten; 7- Kent’s Cavern; 8- Dainton; 9- Milber Down; 10-
Blackbury Camp; 11- Hengistbury Head; 12- Allard’s Quarry; 13- Hod Hill; 14- Gussage All Saints; 15-
Maiden Castle; 16- Danebury; 17- Zionshill; 18- Easton Lane and Winnall Down; 19- Kennel Farm; 20-
Highstead; 21- Sunset Caravan Park; 22- Thanet Earth; 23- East Wear Bay. Author’s illustration.
The sites chosen for this thesis, an estimated population number for each, and the number
of textile tools found at these sites are as follows:
80
Table 3. Sites in this study with population estimates and number of textile tools
County Site Population Spinning tools Weaving tools
Cornwall Bodrifty 28 - 35 population size 6 spindle whorls No loom weights
Carn Euny 63 population size 14 spindle whorls No loom weights
The Rumps 42 population size 15 spindle whorls 12 loom weights
Trevelgue Head 42 - 63 population size 22 spindle whorls 2 loom weights
Trevisker 21 population size 10 spindle whorls 13 loom weights
Devon Blackbury Camp 7 population size 1 spindle whorl 4 loom weights
Dainton Unclear population size 5 spindle whorls 1 loom weight
Kent’s Cavern Unclear population size 20 spindle whorls No loom weights
Milbern Down 7+ population size 4 spindle whorls 2 loom weights
Mount Batten 21+ population size 4 spindle whorls No loom weights
Dorset Allard’s Quarry 14 - 21 population size 6 spindle whorls 2 loom weights
Gussage All Saints 7 population size 8 spindle whorls 12 loom weights
Hengistbury Head 21+ population size 39 spindle whorls 10 loom weights
Hod Hill 226 population size 35 spindle whorls 11 loom weights
Maiden Castle 161 population size 58 spindle whorls 19 loom weights
County Site Population Spinning tools Weaving tools
Hampshire Danebury 168 population size 28 spindle whorls 62 loom weights
Easton Lane Unclear population size 1 spindle whorl 8 loom weights
Kennel Farm 7+ population size 1 spindle whorl 7 loom weights
Winnall Down 56 population size No spindle whorls 13 loom weights
Zion Hill Farm Unclear population size No spindle whorls 5 loom weights
Kent East Wear Bay 7+ population size 24 spindle whorls 2 loom weights
Highstead 14 population size 2 spindle whorls 2 loom weights
Sunset Caravan 21 population size 5 spindle whorls No loom weights
Thanet Earth 21 population size 1 spindle whorl 5 loom weights
As seen in figure 18 (above), there is a reasonable distribution of sites across the central
counties of Dorset and Hampshire. The other counties of Cornwall, Devon, and Kent have
sites clustered along the coast; this was the result of both Iron Age communities building
close to trading routes and modern rescue excavations being conducted prior to building
work in these coastal towns. It was difficult to find Iron Age settlements in these counties
that lay further inland and fit the criteria of this thesis due to issues with preservation in
acidic soils (such as in Cornwall and Devon), vague accounts and reports of excavations
81
conducted in the early 20th century, and difficulties in tracking down where the textile tools
were stored. A synopsis of each site and its features is provided in Appendices 1, 2, and 3.
Data collection and recording
I used callipers to record the height, width, and depth in millimeters of each tool before using
a high precision digital scale to measure its weight in grams. This information was recorded
on a data sheet that I adapted from artefact recording forms used in excavations (see
Appendix 4). Additionally, I took photographs of each artefact and made notes on the
recording sheets of remarkable features, materials, and visible signs of manufacture. The
information from these sheets was then transferred to spreadsheets and used to determine the
average tool dimension and weight at each site. Scatter charts were made using the diameter
and weight of spindle whorls from each county. I used these charts to identify clusters
(hereafter called ‘groups’) of tools in each county, which demonstrated the most commonly
used tools in these counties. I tabulated this information and used it to determine the most
likely threads these tool groups produced, using the example of Grömer’s experiments
(2005). I repeated this process using the weight and depth of loom weights and tabulated how
well suited these were to weaving the types of threads produced in the respective counties,
using the example from Mårtenson et al (2007, 2009). This allowed me to calculate the most
common tools and most likely textiles that these produced for each county. Then, I repeated
this process for each type of site in this study: fortified, enclosed, open, and other (caves,
etc.) to determine whether these larger trends differed in certain site types.
Data presentation and discussion
Once I had calculated the types of threads and textiles that each site most likely produced, I
mapped out and presented the number of spindle whorls and loom weights at each site as pie
charts. Then, I placed this information onto maps of the various counties and related these
tool distributions to the types of textiles produced at each site. My aim was to determine what
information this could provide about the organisation of textile production - which sites only
produced fine-spun threads and which produced thicker threads; which only wove fine
textiles and which could weave a wider variety.
I also wanted to determine how much land was available for animal herding or crop
cultivation. Therefore, I overlaid the information regarding the number of textile tools at each
site onto maps made with the Historic Environment Record GIS data from each county. I
82
highlighted areas unsuitable for grazing or crops (fenlands, estuaries, ancient woodlands,
etc.), areas suitable for grazing (moorlands, heath, rough ground, etc.), and areas of known
prehistoric fields. As already discussed, access to the raw materials necessary for textile
production has a huge impact on the organisation of production. One site may specialise in
spinning and have little to no evidence of weaving because it has access to the grazing areas
or croplands whereas another site specialises in weaving, possibly because it lacks this access
to raw materials. By including this environmental data, I could better understand why some
sites had limited forms of textile tools compared to others.
Finally, I traced which sites had reused pottery sherds as spindle whorls, since this material
is used to identify a shift to large-quantity production of standardised textiles during the late
Iron Age at various northwestern European sites. Once I determined the number of sherds per
site, I compared this with environmental data of each county, specifically the geological data.
Some sites may rely on pottery sherds because there is no suitable stone in the area that could
be carved into spindle whorls; alternatively, settlements may depend on pottery sherds
despite having access to stone that could be used for spindle whorls. By comparing the data
to the region’s geology, I aimed to gain a better insight into factors that guided the spinner’s
decisions on which material to use for their tools.
83
4. Data
In this chapter, each site is presented in alphabetical order with basic information on the
settlement, its excavation, and its excavation reports. The minimum, maximum, and average
weights and dimensions of the sites’ spindle whorls and loom weights are provided, along
with a summary of other finds at the site that demonstrate craft activities and trade (either
local or long distance).
Cornwall
Bodrifty
Bodrifty is an enclosed settlement in the middle of the Penwith peninsula (NGR SW 44436
35570) with an occupation date of roughly 400-50 B.C. Excavations were conducted by the
West Cornwall Field Club and Dorothy Dudley (1956) from 1950-1954 and revealed that the
site began as an open settlement, probably during the Bronze Age. The main features at
Bodrifty are a stone enclosure wall and nine stone roundhouse structures; five of these are
large houses and four are smaller structures that appear late in the settlement phase. The
earlier, larger structures have wider entrances facing southeast and thicker walls, while the
smaller roundhouses lack hearths, postholes, or gullies and their entrances are narrow and
face southwest. This suggests a population of 35 in the earlier Iron Age phase and 28 in the
later phase. Dudley’s report details the excavation methods well but does not provide much
detail on the contexts of finds. For example: the report describes the spindle whorls by
focusing on the shape and material of each whorl found, yet only a few whorls are given
contexts or locations in the report. Although context numbers were written on each spindle
whorl, it was not possible to determine where each artefact came from.
Six spindle whorls were recorded from the Royal Cornwall Museum archives; all carved
from repurposed sherds or local stone. See Appendix 5, Table 1 for full tabulated data.
Table 4. The minimum, maximum, and average measurements of Bodrifty spindle whorls.
Diameter Depth Weight
Minimum 17 mm 12 mm 5 g
Average 35.8 mm 19.6 mm 24 g
84
Maximum 58 mm 26 mm 42.4 g
These artefacts are dated to the Late Iron Age and only two were found in the same context
- the annexe of structure 43. No loom weights were excavated at this site. The number of
spindle whorls, while small, could have easily supplied each household with one spinner to
provide threads for further work. It is important to remember that the lack of loom weights
does not represent a total lack of weaving, simply the lack of warp-weighted looms. Other
craft production tools are scarce; no metal- or wood-working tools found in the site, although
one hut produced ironworking residue. Subsistence tools are present, with saddle querns
recovered from each of the 9 roundhouses. Pottery in both Iron Age phases was locally
produced and there was no evidence of long-distance trade (Dudley 1956).
Carn Euny
Carn Euny is a courtyard settlement on the Penwith peninsula near the hillfort of Carn Bran
(NGR SW 40250 28839) with an occupation date of roughly 400 B.C.-A.D. 300. Patrica
Christie conducted excavations between 1964 and 1972 covered the fogou and several houses
on the site, focusing on investigating and restoring the fogou and circular underground
chamber. These works also revealed nine earlier hut foundations, which represent an
estimated population of 63 and demonstrated that Carn Euny was an open settlement during
the Iron Age before being enclosed during the Roman period. Finds are scarce due to the
acidic soil conditions and are largely limited to locally made pottery and quernstones, which
makes it difficult to judge any trading links this settlement may have had. While the
excavations were thorough and well-recorded, its reports consist of four articles covering
each season of work (Christie 1965, 1966, 1968, 1976). None of these recorded more than the
shape and materials of each spindle whorl, which made it difficult to trace the context and
season in which the spindle whorls were discovered.
14 spindle whorls were recorded from museum archives; half are made from locally
sourced stone or shale, and half from repurposed shards or baked clay. See Appendix 5, Table
2 for full tabulated data.
Table 5. The minimum, maximum, and average measurements of Carn Euny spindle whorls.
Diameter Depth Weight
85
Minimum 16.4 mm 5.4 mm 9 g
Average 33.79 mm 11.1 mm 19.1 g
Maximum 46 mm 24.4 mm 61 g
Most tools found were scattered around the site with only two found in Hut A. These
whorls were dated to the early phase of the site, approximately 1st century BC. This number
of spindle whorls would provide each household with at least one spinner to produce threads
for further work. No loom weights were discovered, indicating either that weaving took place
elsewhere or that another form of weaving occurred at this site. Other craft production tools
are scarce, in part due to the acidic soil destroying organic remains. No metalworking or
woodworking tools found on the site. Subsistence tools are uncommon, with only mortars
and saddle querns found across the site. No evidence of trade appeared in the archaeological
record, possibly due to the acidic soil conditions destroying potential evidence (Christie
1979).
The Rumps
A fortified settlement located on a promenatory along Cornwall’s north coast (NGR SW
93411 81092) with three lines of defensive ditches and banks separating it from the mainland.
Six hut platforms cluster near the entrance, providing an estimated population of 42 and
suggesting that the rear of the castle was used for arable or pastoral purposes. Evidence of
metalworking is common, with tools and debris found clustered around the various huts.
Woodworking tools are more difficult to find, probably due to the acidic soil conditions.
Unusually for South West sites, some organic material has survived deposition: middens of
shellfish remains demonstrate that the surrounding cliffs and sea were used to supply the
population as an important food source. Subsistence evidence consists of quern stones and
corn-drying ovens, along with large quantities of limpets and animal bones. 60% of all
animal bones are sheep, which may have supplied higher levels of textile production than
smaller settlements (Brooks 1974). The pottery at this site mimics forms and decorations
found in Brittany, implying that close trading connections with the continent existed. Brooks
(1974) conducted excavations over 1965-1967.
86
15 spindle whorls were recorded from museum archives; ten are made from locally sourced
stone or shale, four from repurposed shards, and one from animal bone. See Appendix 5,
Table 3 for full tabulated data.
87
Table 6. The minimum, maximum, and average measurements of the Rumps spindle whorls.
Diameter Depth Weight
Minimum 29 mm 3 mm 6 g
Average 37.67 mm 11.1 mm 20.84 g
Maximum 63.6 mm 19.7 mm 44.4 g
Three tools were found in ditch A and were dated to the second occupation phase of the
site, approximately 1st century BC. This number of spindle whorls appears limited, although
it could still provide each household with at least one spinner.
11 possible loom weights were recorded from museum archives, all made from locally
sourced slate and roughly carved into a discoid shape. See Appendix 5, Table 4 for full
tabulated data.
Table 7. The minimum, maximum, and average measurements of the Rumps loom weights.
Diameter Depth Weight
Minimum 33 mm 3 mm 11 g
Average 95.26 mm 7.9 mm 151.16 g
Maximum 181 mm 16.7 mm 543.6 g
Half of all loom weights were found in R2’s inner ditch extension and shaped in rough
disks from local slate. Such a large number of loom weights found in the same deposit
suggests that these were either owned by the same person or used on the same loom as a tool
set, although only four are of similar weights and dimensions. These artefacts are also dated
to the second occupation phase of the site.
Trevelgue Head
This fortified settlement is located on a spur of land along the north coast of Cornwall (NGR
SW 82572 63040) and is defended by several lines of ditch-and-bank earthwork defences.
Two Bronze Age barrows demonstrate long term occupation in this area. Multiple
88
roundhouse hut platforms were discovered in the innermost enclosure, most dating to the 2nd
century BC. This provides an approximate population of 42-63 people. One roundhouse
measured 14.5 m in diameter and is the largest found in Cornwall; it is believed to have been
a meeting place rather than a dwelling. 200 kg of iron slag, several furnaces, and
metalworking tools were discovered in one area of the site (close to the entrance), indicating
iron smelting was a major craft activity on this site with some copper smelting performed as
well. Pottery found on site indicates a preference for local ceramic production and trade,
suggesting only metal was traded elsewhere. No woodworking tools were found but
subsistence evidence is common, with a number of quern stones, large quantities of animal
bones and burnt grain in midden deposits, and extensive field systems linked to the fort
(Nowakowski et al. 2011). Trevelgue Head was initially excavated in 1939 by CK Croft
Andrews and was interrupted by the outbreak of the Second World War; the excavation
record and notes were not written up and published until 2011. The contexts of various textile
tools were not recorded
22 spindle whorls were recorded from museum archives; 19 are made from locally sourced
stone or shale and three from baked clay. See Appendix 5, Table 5 for full tabulated data.
Table 8. The minimum, maximum, and average measurements of Trevelgue Head spindle whorls
Diameter Depth Weight
Minimum 14 mm 4 mm 9 g
Average 32.04 mm 11.14 mm 19.18 g
Maximum 45.7 mm 21.8 mm 46.8 g
.
The majority of tools were found in houses around the site, rather than in rubbish
depositions. These whorls were dated to the Late Iron Age period, approximately 150 B.C.
Two possible loom weights were recorded from museum archives, all made from locally
sourced slate and roughly carved into a discoid shape. See Appendix 5, Table 6 for full
tabulated data.
Table 9. The minimum, maximum, and average measurements of Trevelgue Head loom weights.
89
Museum number Diameter Depth Weight
133 88.3 mm 36.2 mm 654 g
277 78 mm 7 mm 57 g
Both loom weights were shaped in rough discs and found in House 420 (along with four
spindle whorls), suggesting this was the dwelling and/or workplace of a textile producer.
Whether this producer was a craft specialist may be determined during data analysis. These
artefacts are dated to the Late Iron Age phase of the site.
Trevisker
Trevisker is a Round settlement with an occupation phase of 200 BC - AD 100, located on
gently sloping ground between two river systems along the north Cornish coast (SW 88800
68700) and excavated by ApSimon and Greenfield in 1955-56. The site began as a large
enclosure in the Bronze Age and a smaller, inner enclosure was added during the Iron Age;
the remains of multiple roundhouses from both periods were excavated, as well as a series of
pits, ovens and gullies. Houses 1-3 dated to the Iron Age, although only House 2 produced a
spindle whorl in its occupation layer. This represents an estimated population of 21. House 3
contained several ovens and hearths, possibly demonstrating its use as a cereal processing
area, although little evidence of grain storage was determined. No wood- or metalworking
tools or features were found in the site, although pieces of slag and metal artefacts were
discovered; the excavators determined that smelting occurred near to, but not inside,
Trevisker itself. Evidence of subsistence includes a possible iron sickle, a number of rotary
querns, and the remains of several corn-drying ovens. The excavation was thorough and
mainly focused on establishing the building sequence and ApSimon and Greenfield’s report
was reasonably detailed when discussing the excavation methods and site features, or the
metal finds and debris. Unfortunately, the loom weights and spindle whorls were not detailed
in the report; these were ignored when discussing the economy and subsistence of the site
and no context given for any textile tools discussed in the finds sections (ApSimon and
Greenfield 1972: 368).
90
Ten spindle whorls were recorded from museum archives; eight are made from locally
sourced stone or shale, one from repurposed shards, and one from baked clay. See Appendix
5, Table 7 for full tabulated data.
Table 10. The minimum, maximum, and average measurements of Treviker spindle whorls.
Diameter Depth Weight
Minimum 22.7mm 7mm 6g
Average 32.86mm 10.5mm 19.65g
Maximum 42mm 19mm 46.25g
One spindle whorl was found in House 2, alongside iron slag and other evidence of
ironworking. This whorl was dated to the second phase of the site, approximately 200-100
B.C. The number of spindle whorls found on site is proportionally greater per head of
population than other Cornish sites and could have provided half of the people living here
with spindle whorls.
13 possible loom weights were recorded from museum archives: 11 made from locally
sourced slate and roughly carved into a discoid shape and two baked clay weights, shaped in
the same triangular form typical of Iron Age loom weights. See Appendix 5, Table 8 for full
tabulated data.
Table 11. The minimum, maximum, and average measurements of Trevisker loom weights.
Diameter Depth Weight
Minimum 41.4mm 1.8mm 6g
Average 81.85mm 20.78mm 380.35g
Maximum 154mm 92mm 1372.8g
91
No contexts were provided for these tools and these weights were unable to be dated.
However, this number of loom weights suggests a large amount of weaving occurred at this
site in comparison to other small Iron Age sites in Cornwall.
Devon
Blackbury Camp
Previously known as Blackbury Castle, this fortified site is located on a ridge (NGR SY
18700 92400), with univaliate defences and a developed entranceway (described as a
‘barbican’) with gate-posts. Archaeological evidence shows this site was occupied roughly
300 B.C. - A.D. 100. Young and Richardson conducted limited excavations in 1952-1954
within the enclosure that revealed the remains of one roundhouse along with an oven, a
cooking pit, and several pits and gullies, which suggests a population of seven. Pottery makes
up most of the finds, due to the acidic soil destroying any organic artefacts. The earliest
pottery is similar to the Iron Age B ceramics of Dorsert, Sussex, Hampshire, and Wiltshire
and appears to have influenced the later pottery styles found at the Camp. Metalworking is
evidenced by several fragments of bronze, one of iron, and four whetstones. The excavations
were reasonably thorough, although there was little to find compared to other sites.
Unfortunately there was little detail given on the textile tools recorded; no contexts and only
a brief description on the shape and materials of these items. These textile tools demonstrated
the only other craft activity at the site.
Only one spindle whorl was recorded from the Royal Albert Memorial Museum archives;
made from locally sourced sandstone and carved into shape. Its diameter measured 47 mm;
the depth measured 19 mm; the weight measured 42 g. This whorl was found in the vicinity
of the roundhouse and main entrance to the camp and dated to the Middle Iron Age phase,
approximately 300-100 B.C. See Appendix 6, Table 1 for location details.
Four loom weights were recorded from museum archives, made from locally sourced clay
and roughly shaped into a triangle with a hole through each corner. See Appendix 6, Table 2
for full tabulated data.
Table 12. The minimum, maximum, and average measurements of Blackbury Camp loom weights
92
Diameter Depth Weight
Minimum 48 mm 30 mm 78 g
Average 137.25 mm 96.75 mm 1174.5 g
Maximum 183 mm 162 mm 2040 g
All loom weights found in the same section of rampart 5 and are shaped in the typical
triangular form of Iron Age loom weights. These artefacts are dated to the Middle Iron Age
phase of the site, roughly 300-100 B.C. These limited number of textile tools are suitable for
supplying a single household’s requirements. Other craft production tools are scarce. A small
amount of iron slag found in a cooking pit outside a roundhouse close to the hillfort’s main
entrance indicates at least smelting on site. Close analysis of pottery from this site shows
predominantly local production with some vessels traded from Somerset.  No mention of
subsistence or woodworking tools in the excavation report (Young and Richardson 1954).
Dainton
An unenclosed settlement that lies on Dainton Common (NGR SX 85900 66800) and
includes a possible field system, several hut platforms, and a probable midden in the form of
a concentric bank. This bank contained spindle whorls, pottery, querns, ox and pig teeth, and
charcoal fragments. Crucible fragments were discovered in one pit, but nowhere else on the
site. The site was excavated in 1939 and 1949 by Willis and Rogers, and the evidence was
re-evaluated by Silvester in 1980. In both reports, the textile tools are given little detail and
no context, other than being found in or at the bottom of the midden bank. Both reports
agreed that the main occupation of this site occurred during the Bronze Age, although there
are no structural remains and only evidence of occasional (possibly seasonal) occupation
during the Iron Age (c.400 - 50 B.C.). The exact use of this site is unclear; Willis and Rogers
suggested it functioned as a hunting camp in the Bronze Age and as a periphery to a
metalworking area in the Iron Age. Silvester agreed this was the most likely explanation,
given the lack of Iron Age structures and hearths or cooking areas. It is also possible that this
area functioned in a manner similar to Kent’s Cavern (see below for details), where spinners
would process raw material before transporting the spun threads elsewhere for further use.
93
This possibility was not considered in either report, possibly due to the author’s focus on any
metal items and debris.
Five spindle whorls were recorded from the Royal Albert Memorial Museum archives;
three made from locally sourced stone and carved into shape, remaining two made from local
clay. See Appendix 6, Table 3 for full tabulated data.
Table 13. The minimum, maximum, and average measurements of Dainton spindle whorls
Diameter Depth Weight
Minimum 34.7 mm 18 mm 18 g
Average 50 mm 36.8 mm 53.8 g
Maximum 44 mm 50 mm 39 g
These whorls were dated to the Early Iron Age phase of the site, approximately 600-400
B.C. Due to the lack of structural remains and the evidence of seasonal occupation and site
use, it is not possible to estimate this site’s population. Nor is it possible to determine
whether this number of spindle whorls would appropriately supply the Iron Age population.
Only two loom weights were recorded from the museum archives, made from local shale
and roughly carved into shape. These weights are roughly carved into a disc with a central
hole and can be dated to the Early Iron Age phase of the site. See Appendix 6, Table 4 for full
tabulated data.
Table 14. The measurements of Dainton loom weights
Museum number Diameter Depth Weight
A4431 93.6 mm 44 mm 166 g
A5976 130 mm 20 mm 272 g
Other craft production tools are scarce in the Iron Age. A large number of metalworking
tools and scrap were found from Bronze Age contexts, indicating the presence of highly
capable bronze smiths. No other woodworking or subsistence tools are mentioned in the
94
report. The Iron Age ceramic finds are locally made and provide the most reliable dating
methods from Dainton for the textile tools mentioned here (Silvester 1980).
Kents Cavern
Kent’s Cavern is a large cave system found in the centre of Torquay (NGR SX 93452 64171);
only the entrance to the system was used throughout prehistory and into the Roman period.
Pottery remains have dated the Iron Age occupation of this site to c.700 BC - AD 43. The
entire cave system was excavated between 1856-1880 by William Pengelly using a
groundbreaking method of excavation that has since become the basis for most cave
excavations (Silvester 1986). The centre of human activity lies in only the largest and most
accessible parts of the cave system, which was defined by the ‘black mold’ deposit that
covered this area (Pengelley 1884; Silvester 1986). Unfortunately, no further information on
the nature of this ‘black mold’ exists; only a few samples survived excavation, being
destroyed by Pengelley during further excavation. The current theory is that this black
substance was the result of organic material brought in by humans mixed with ashes and
charcoal from the fires these people lit (Silvester 1986). Quern stones, cooking pottery, and
small areas of burning (possible cooking fires) found on the site, but no archaeological
features that indicated long-term occupation. Weaving combs and high numbers of spindle
whorls but no loom weights suggests the processing of fibres (possibly animal wool) was
carried out here for transportation and weaving elsewhere.
20 spindle whorls were recorded from the Torquay and British Museum archives; six made
from shale, four from local sandstone, one from local clay, and nine from unidentified stone.
See Appendix 3, Table 5 for full tabulated data.
95
Table 15. The minimum, maximum, and average measurements of Kent’s Cavern spindle whorls
Diameter Depth Weight
Minimum 20 mm 8 mm 4.2 g
Average 32.6 mm 15.57 mm 23.69 g
Maximum 42 mm 30 mm 48.4 g
The location and context of these tools is unrecorded in the excavation reports; however,
proximity to Iron Age pottery forms has dated these tools to approximately 500-300 B.C.
Other craft production tools do not appear in the archaeological record. Iron Age pottery
made with local clay was used to date spindle whorls; this pottery is predominantly utilitarian
cooking ware and demonstrates short-term, possibly seasonal, occupation at this site
(Silvester 1986).
Milber Down
This hillfort is located on a hill overlooking the Teign estuary ( NGR SX 88400 69800);
excavations conducted in 1937-1938 revealed multiple ditches and banks, gullies, a corn
drier, and the remains of one (possible) hut. Cottrill led excavations and wrote two interim
reports during work (Cottrill 1937, 1938) before a more comprehensive report was compiled
by Fox, Raleigh Radford, Rogers, and Shorter (1949-1950). The main period of occupation
occurred approximately 100 BC - AD 100, with an estimated population of seven;
unfortunately acidic soil conditions degraded most of the organic materials. Furthermore, the
excavations were limited in scope and mainly focused on the hillfort's defences; little work
was done on the inner enclosure where the domestic activity would have occurred, making a
population estimate difficult. This is a frequent issue in excavations from this period:
excavators were more interested in signs of warfare, metallurgic crafts, and signs of
chieftains or warlords than in how people lived or worked in prehistory. As such, there is
little information given on the textile tools recovered from Milber Down. Several spindle
whorls were found in gullies and ditches, suggesting these were no longer used at time of
deposition. Other finds included a corn drying oven, large amounts of charcoal, iron slag, and
an iron pommel, which suggests some form of iron working on this site.
96
Four spindle whorls were recorded from the Royal Albert Memorial Museum archives;
three made from locally sourced shale and one from local sandstone. See Appendix 6, Table
6 for full tabulated data.
Table 16. The minimum, maximum, and average measurements of Milber Down spindle whorls
Diameter Depth Weight
Minimum 28 mm 12 mm 8 g
Average 31.25 mm 13 mm 17.92 g
Maximum 65 mm 14 mm 64.4 g
Several spindle whorls found in gullies and ditches, suggesting they were no longer used at
time of deposition. This number of spindle whorls could easily supply a single household
with an appropriate amount of material for further work.
Two possible loom weights were recorded from the museum archives, both made from
locally sourced stone and roughly carved into shape. See Appendix 3, Table 7 for full
tabulated data.
Table 17. The measurements of Milber Down loom weights
Museum number Diameter Depth Weight
A5939 65 mm 13 mm 64.4 g
A4342 86 mm 15 mm 181.25 g
These tools have the triangular form typical of Iron Age loom weights and date to the late
phase of the site, roughly 200-100 B.C. The limited number of loom weights suggests that
weaving was a limited activity to supply the residents of this hillfort with their required
textiles, not for large-scale production. The excavated pottery was utilitarian local ware and
no foreign coinage or personal trinkets were discovered. Other finds included a corn drying
oven, large amounts of charcoal, iron slag, and an iron pommel, which suggests some form of
iron working at this site (Fox et al. 1949-1950).
97
Mount Batten
A promontory fort located on a small headland in the Plymouth Sound (NGR SX 48679
53252), Mount Batten has been built on repeatedly in modern times, effectively erasing any
evidence of prehistoric defensive earthworks. A series of rescue excavations were carried out
by Barry Cunliffe in 1983-1985; these were limited in scope and therefore were unable to
determine the size and shape of the Iron Age site. However, study of the bronze tools and
ornaments, as well as the items’ stratigraphical placement, suggests continuous occupation
from the Late Bronze Age into the Middle Iron Age, approximately 900-100 B.C. No
structural remains were discovered, which does not allow for the population size to be
estimated. The excavations (and the publication that followed) were thorough and detailed,
although little information on the scale of textile production could be made given the limits
of the excavations. Cunliffe was able to show that Mount Batten was in decline throughout
the Iron Age until its abandonment in the Mid Iron Age. Personal ornaments were found in
large numbers here, along with bronze ingots and smelting debris. Given the location of the
site (close to the mines of Dartmoor and on the trade route from Cornwall to Brittany), this
suggests Mount Batten was an important manufacturing centre and trading post for at least
the metal trade during the Late Bronze and Early Iron Ages (Cunliffe 1988). Pottery in phase
3 (Early to Mid Iron Age) is mostly local southwestern ware with a few Cornish sherds, with
no pieces of continental pottery.
Four spindle whorls were recorded from the Plymouth Museum archives; all made from
locally sourced sandstone or stone and carved into shape. See Appendix 3, Table 8 for full
tabulated data.
Table 18. The minimum, maximum, and average measurements of Kent’s Cavern spindle whorls
Diameter Depth Weight
Minimum 30 mm 13 mm 26 g
Average 35.9 mm 16.5 mm 38.2 /g
Maximum 45 mm 19 mm 68 g
These whorls were dated to approximately the Second to First millennium B.C. Other craft
production tools are scarce at this site, although personal ornaments and coinage are
98
common. Finds of brooches and Kimmeridge shale bracelets were found in large numbers,
along with bronze ingots and smelting debris. Coins from along the southern coast of Britain
and northern France demonstrate the personal wealth of people on site, as well as the
presence of long distance trade.
Dorset
Allard’s Quarry
An open settlement located on a low ridge of the Blackmoor Vale, overlooking the River
Stour (NGR ST 79491 19881). This site was excavated by C.E. Bean in 1932–9 and Audrey
Williams in 1944-1945 prior to its sale to a quarrying company and subsequent destruction.
Their excavations produced evidence of long-term occupation throughout the Iron Age (c.800
BC - AD 100). The settlement was largest in the Early Iron Age, when it consisted of three
groups of post-holes for rectangular huts, several pairs of postholes, 42 pits, and two burials;
the estimated population numbered 21. During the second phase (Mid Iron Age) occupation
decreased, excavations revealed the remains of two circular huts and 26 pits, providing an
estimated population of 14. In the final phase of the site (A.D. 0-43), there were the remains
of one possible hut, a 4-post structure (likely to be a granary), and 16 pits, providing an
estimated population of seven. Large quantities of pottery were excavated, as well as a
bronze ring and brooch, a Kimmeridge shale bracelet and pedestal base, the textile tools
detailed below, and two Durotriges coins. Although no imported pottery was discovered in
any of the site’s phases, the presence of personal ornaments, coins, and the Kimmeridge shale
bracelet suggests a reasonably prosperous settlement with access to regional trade.
Six spindle whorls were recorded from the Dorchester Museum archives; four made from
locally sourced limestone, one from local clay, and one from human bone. See Appendix 7,
Table 1 for full tabulated data.
Table 19. The minimum, maximum, and average measurements of Allard’s Quarry spindle whorls
Diameter Depth Weight
Minimum 26 mm 11.2 mm 17 g
Average 41.58 mm 19.4 mm 38.7 g
99
Maximum 82.3 mm 27.5 mm 71 g
Four whorls were dated to the Early Iron Age, one to the Middle Iron Age, and one to the
Late Iron Age. The number of tools in the early and late Iron Ages would be enough to
supply at least one spinner in each household, but in the middle Iron Age only one of the
households is likely to have had a spinner. This would suggest that the textile workers here
were able to supply their households but may not have been able to produce a surplus for
further trading or exchange.
Two loom weights were recorded from the museum archives, made from locally sourced
material and roughly shaped. See Appendix 7, Table 2 for full tabulated data.
Table 20. The measurements of Allard’s Quarry loom weights
Museum number Width Depth Weight
Unnumbered 186.7 mm 31 mm 303.3 g
13 213.3 mm 61 mm 1313.3 g
Both have the same triangular form typical of Iron Age loom weights. These artefacts are
dated to the Early Iron Age. Other craft production tools are scarce. No metalworking or
woodworking tools found in the site. Although bone and flint tools appear prominently in the
first two phases (Early to Middle Iron Age), there is no evidence that their manufacture
occured on site. The same can be said for the Kimmeridge shale armlets found in the Early
Iron Age deposits. The presence of these shale artefacts, as well as the pottery evidence
throughout the site, demonstrates that only regional trade occurred at Allard’s Quarry
(Williams 1950).
Gussage All Saints
A banjo enclosure located on the eastern slopes of a ridge overlooking a small valley of a
tributary to the River Allen (NGR ST 99895 10620), with evidence of occupation from 300
B.C. - A.D. 100. This site was the focus of a thorough excavation in 1972 by Wainwright.
The area around the single roundhouse is divided into clear activity zones, with storage pits
clustered to the east of the roundhouse and crafting debris to the south. This roundhouse
100
suggests a population size of seven. Craft activity is represented by ceramic debris from pit
209 (dating to the second phase of settlement, 1st century BC), bronze casting and iron
forging debris (from both early settlement phases), and a large number of spindle whorls
(dated to phase 3 of the settlement) and loom weights (an equal number found in each of the
settlement phases). The agricultural economy of this site is demonstrated by a variety of
carbonised grains from various rubbish deposits. The high number of spinning and weaving
tools from this site suggest that this site probably produced more than required locally;
Wainwright believes this was part of a wider economic trade (1979: 190-1). This is supported
by the evidence that this site traded in querns and pottery from the larger Wessex trading
area.
Nine spindle whorls were recorded from the Dorchester Museum archives; five made from
locally sourced clay, two from local chalk, and one from a reused pottery sherd. See
Appendix 7, Table 3 for full tabulated data.
Table 21. The minimum, maximum, and average measurements of Gussage All Saints spindle whorls
Diameter Depth Weight
Minimum 30 mm 8 mm 9.4 g
Average 35.9 mm 18.9 mm 24.1 g
Maximum 44 mm 27.4 mm 36 g
Two tools found in pit 212 from the north-east of the site. These whorls were dated to the
second phase of the site, approximately 1st century BC. This number of spindle whorls
suggests a higher level of textile production than seen in other small settlements in Dorset;
considered alongside the large number of loom weights, this suggests that this household
produced a surplus of textiles for further trade and exchange to support the population and its
needs.
12 loom weights were recorded from the museum archives, made from locally sourced clay
and chalk, roughly shaped. See Appendix 7, Table 4 for full tabulated data.
Table 22. The minimum, maximum, and average measurements of Gussage All Saints loom weights
101
Width Depth Weight
Minimum 59 mm 52.8 mm 520 g
Average 399.4 mm 72 mm 1753.79 g
Maximum 1807 mm 110 mm 3710 g
Three loom weights were found in feature 379 and two in feature 639. Nine loom weights
in total are shaped in the same triangular form typical of Iron Age loom weights. These
artefacts are equally divided between the 1st and 2nd phases of the site. A large quantity of
metal working debris (including metal-working tools, moulds, crucibles, scrap, and slag)
found in pit 209 dated to the 1st century BC. No woodworking tools from the site, although
quern stones traded from the wider Wessex area were discovered (1979).
Hengistbury Head
A promontory fort located on a spur of land jutting out into the English Channel, just west of
Bournemouth (NGR SZ 17300 90700); the site has evidence of occupation dating from 700
B.C.-A.D. 50. Because of its exposed location, this site is particularly vulnerable to erosion;
by 1970, approximately half of the Iron Age settlement had been lost to the sea, which
prompted Cunliffe’s excavations in that year. This site has evidence of non-continuous
occupation from the Paleolithic period onwards. Pottery finds date the earliest Iron Age
occupation to around 700-400 BC, the exact nature of this period’s settlement is unclear but
probably a scattering of individual farmsteads before the double dyke defences were made,
turning the site into a hill/promontory fort. Two bronze axe finds (one 300m out at sea,
possibly as the result of erosion or being lost from passing boats) hint at this site being a
possible stop-off for significant long-distance trading routes. Few artifacts or materials were
definitely assigned to the Middle Iron Age period (400-100 BC), indicating no or little
occupation. In the Late Iron Age phase (100-0 BC), the settlement became a major
international trading port, with a number of roundhouse features. This suggests a population
of over 21; the exact number will remain unclear due to the erosion of the site. Local,
regional, and European trade is demonstrated by pottery finds (from west England,
north-west France, and locally); metal-working/manufacturing; Kimmeridge shale
importation and manufacturing (waste material and finished artefacts); and limited glass
102
production (crude glass lumps and fragments of finished glass products). Ingots of copper
indicated copper importation, silver-copper ore and ingots demonstrated silver refining, and
bronze and gold working debris also indicated the range of smithing practiced at the site.
Other finds included a large number of coins (local and imported) and a range of luxury items
(jewellery, personal items, etc.), all of which demonstrate the importance of Hengistbury
Head as a settlement and centre of trade and craft production.
15 spindle whorls recorded from the Hengistbury Head Museum archives; all made from
locally sourced material. See Appendix 7, Table 5 for full tabulated data.
103
Table 23. The minimum, maximum, and average measurements of Hengistbury Head spindle whorls
Diameter Depth Weight
Minimum 28 mm 5 mm 5.8 g
Average 42.76 mm 10 mm 21 g
Maximum 59 mm 28 mm 84.2 g
Two tools were found in the southeast area of the excavation. Three whorls date to the
Early Iron Age phase of the site (800-500 BC); two whorls date to the Middle Iron Age phase
(400-100 BC); two whorls date to the Late Iron Age 1 phase (100-50 BC); and other 31 date
to the Late Iron Age 2 phase (50 BC - AD 43). This clearly demonstrates a sharp increase in
the production of textile tools during this final Late Iron Age phase, indicating that textiles
began to be made in larger quantities for trade and exchange, rather than to supply the
community.
Ten loom weights were recorded from the museum archives, made from locally sourced
clay and roughly shaped. See Appendix 7, Table 6 for full tabulated data.
Table 24. The minimum, maximum, and average measurements of Hengistbury Head loom weights
Width Depth Weight
Minimum 60 mm 40 mm 278.75 g
Average 67.3 mm 76 mm 1177.5 g
Maximum 99 mm 96 mm 2574.6 g
Seven loom weights were found in the northern area of the site; six were shaped in the
same triangular form typical of Iron Age loom weights and the rest in an elongated form.
These artefacts are dated to the Late Iron Age 2 phase of the site (100-0 BC), further
supporting the theory that textiles became more economically valuable in this phase. The
presence of textile production as a craft activity for trading purposes is not unreasonable
given the breadth of materials worked at Hengistbury Head and the various stages of
production shown in the archaeological record (Cunliffe 1978).
104
Hod Hill
A multivallate hillfort located on (NGR ST 85500 10700) and dated to roughly 800 B.C. -
A.D. 43. The site covers 22 hectares and has a Roman fort in the northwestern corner.
Multiple excavations have occurred at this hillfort; the first by in 1897, the second by in
1949, and the final excavations by in 1951-2 by Sir Ian Richmand. Unfortunately, much of
the archaeology had been damaged by plough-scars by the time it was excavated. The final
excavation report included information from the previous work at the site as well as the
excavations conducted in the 1950s. As is typical of archaeological reports from this era, the
excavation methods were thorough and well documented but focused almost entirely on the
defensive structures rather than the domestic. As such, most of the information regarding the
layout of the hillfort is derived from observations of the roundhouse remains, rather than
actual excavation. Roundhouse footprints are extensive and often overlap each other, with
associated pits and hearths following this pattern. LIDAR scans have identified many of
these roundhouse footprints; the clearest of which number 97 in total, although these would
not have been occupied at the same time. Dividing this number by the three Iron Age phases
provides a more reasonable estimate of 32 roundhouses occupied in each phase and presents
a population estimate of 224 people. Several trackways are recorded from the absence of
roundhouse circles; these converge at the northeastern entrance, indicating this was the main
thoroughfare. Metalworking has a large presence on site that appears to be centred on a
possible ‘industrial’ compound with several quarry scars/pits in the southeast corner, close to
the thoroughly-excavated entrances. Magnotromic surveys indicated these pits are filled with
slag and other metalworking debris. Several Kimmeridge shale spindle whorls from this site
are repurposed lathe-turned cores that are left over from producing shale bracelets.
Interestingly, there are few weaving tools at this site compared to the 52 spindle whorls
recorded in the excavations report, which suggests that weaving was limited at this
settlement. Large numbers of slingshots and spearheads suggests the hillfort was abandoned
after several battles or skirmishes.
35 spindle whorls were recorded from the Dorchester Museum archives; 25 made from
locally sourced stone, six from local clay, three from reused pottery sherds, and one from
antler material. See Appendix 7, Table 7 for full tabulated data.
Table 25. The minimum, maximum, and average measurements of Hod Hill spindle whorls
105
Diameter Depth Weight
Minimum 20 mm 7 mm 6 g
Average 38.3 mm 18.9 mm 62 g
Maximum 89 mm 40 mm 134 g
Only two tools were given context and of these two, only one of which could be dated to
the Early Iron Age using pottery excavated from the same feature. The three spindle whorls
made from reused pottery sherds were also dated to the Early Iron Age. The uncertainty over
the date of the other spindle whorls makes it more difficult to determine what form of textile
production took place at this site; however, the extensive and crowded occupation of Hod
Hill suggests that it was at least able to supply the population at any phase.
11 loom weights were recorded from the museum archives, ten made from locally sourced
chalk and one from local clay, all roughly shaped. See Appendix 7, Table 8 for full tabulated
data.
Table 26. The minimum, maximum, and average measurements of Hod Hill loom weights
Width Depth Weight
Minimum 105 mm 52.6 mm 650 g
Average 222 mm 98.4 mm 2287.99 g
Maximum 1086 mm 164 mm 3456 g
These are shaped in the same triangular form typical of Iron Age loom weights;
unfortunately, there is little to no mention of the context of these tools. There are notably
fewer weaving tools compared to spindle whorls in Hod Hill, suggesting either that spinning
was the main textile craft activity at the site or that weaving was communal. If weaving was
shared and not a universal household activity, fewer looms would be needed, resulting in
fewer loom weights in the archaeological record. Most of these artefacts can be dated to the
Early Iron Age by analysing other artefacts in the same feature and fill. Other craft
production tools are common, with large quantities of metalworking debris (mostly iron slag)
106
found filling quarry pits in the southeastern corner’s ‘working compound’. No woodworking
tools were excavated, although subsistence tools (such as quern stones) are common across
the site (Richmond 1968).
Maiden Castle
A large, multivaliate hillfort with extensive, complicated entrances, located on a large hilltop
1.6 miles southwest of Dorchester (NGR SY 66900 88400). Several excavations occurred at
the site, one by Wheeler in 1934 and 1937 and another by Sharples in 1985-6. These
excavations demonstrated a long period of dense occupation from roughly 600 B.C.-A.D.
100, as well as a series of developments to the hillfort’s size, shape, and defenses. During the
first Iron Age phase (c.600-400 B.C.), the small Neolithic enclosure was reoccupied and later
extended during the Middle Iron Age to cover the whole of the hilltop. Occupation appears
densest during the Middle Iron Age (c.400-100 B.C.), with 23 houses dated to this phase
across the site, providing a population estimate of 161 people; this became limited to the
southeastern area during the Late Iron Age (c.100 B.C.-A.D. 43). A thorough excavation
revealed large quantities of artefacts, demonstrating little metalworking during the Early Iron
Age. Animal remains indicate large numbers of sheep, possibly to provide fibres for the
textile production occurring at the site. Pottery finds suggest there was on-site ceramic
production during the Early Iron Age that was replaced in the Middle Iron Age by
importation of ceramics made in Poole workshops. This importation of pottery and the
massive increase in personal ornaments and craft production at the site demonstrates the
importance of Maiden Castle as a centre of production and trade in this region during the
Middle Iron Age. The site was largely abandoned during the Late Iron Age with the
remaining occupation clustered around the eastern entrance, outside of which was the
remains of a smithy. The excavation report compiled by Sharples is thorough and detailed,
with contexts, dates, dimensions, and weights of each find provided in the microfiche.
58 spindle whorls were recorded from the Dorchester Museum archives; 57 made from
locally sourced stone and one from bone. See Appendix 7, Table 9 for full tabulated data.
Table 27. The minimum, maximum, and average measurements of Maiden Castle spindle whorls
Diameter Depth Weight
Minimum 28 mm 9.4 mm 7.3 g
107
Average 43.6 mm 19.66 mm 36.27 g
Maximum 71 mm 36 mm 106 g
14 tools found in area F (see fig. 19 below); 13 in area B; eight in area D; seven in area Q;
five in area A and area L respectively; two in area G; and one each in area H and area R. This
distribution is mostly due to deliberate, possibly ritual, deposition of spindle whorls in post
holes and pits (e.g. at area F, an entrance to the hillfort). Regardless of the meaning of these
depositions and the distribution across the interior of the hillfort, the number of these tools
demonstrates that this site was capable of supplying its own population as well as producing
a surplus of textiles for further trade and exchange.
Figure 19. Plan of the areas of excavation at Maiden Castle with areas that produced textile tools highlighted,
from Sharples 1991.
19 loom weights recorded from the museum archives, made from locally sourced stone and
roughly carved into shape. See Appendix 7, Table 10 for full tabulated data.
Table 28. The minimum, maximum, and average measurements of Maiden Castle loom weights
Width Depth Weight
108
Minimum 48 mm 20 mm 40 g
Average 296.5 mm 59.7 mm 877.3 g
Maximum 1868 mm 102 mm 2237.3 g
Five loom weights came from area D (see above); three from both area B and E; two from
area A; and one from areas F, G, L, and Q. Six are shaped in the same triangular form typical
of Iron Age loom weights, although all these tools are dated to the Mid-Late Iron Age. The
large number of spindle whorls and loom weights from area D suggests this was a special
area for textile production, either because a specialist lived there or because there was a
workshop where textile workers could gather. The number of loom weights in comparison to
spindle whorls is very low, possibly because other forms of weaving were present at this site.
18 weaving combs that dated to the mid-late Iron Age were discovered during excavations,
suggesting that other forms of weaving were present at Maiden Castle during the same
periods. Other craft production tools are common. Huge quantities of metalworking debris
(including 20 kg of hammerscale) were recovered across all excavated areas; a house in the
centre of the extended hillfort contained a possible forge, with a potential copper-smithing
workshop in the southwest corner of the extended fort. Two possible crucible pieces were
also found in this area of the site and dated to the Mid-Late Iron Age; considered alongside
the other finds, it appears that metalworking increased in scale during the Late Iron Age. No
woodworking tools have been identified from the hillfort. Subsistence tools are common and
predominantly quern stones, scattered across the site throughout the Iron Age contexts. Trade
with sites along the British south coast and continental Europe is demonstrated in the
Mid-Late Iron Age pottery (predominantly Wareham-Poole Ware with a few Roman
amphorae), a glass bead from the Late Iron Age and Kimmeridge shale bracelets (Sharples
1991).
Hampshire
Danebury
A large and well-excavated hillfort overlooking the Stour valley (NGR SU 32300 37600),
built and occupied roughly 550-100 B.C. and excavated by Barry Cunliffe from 1969-1988.
This was the first rigorous and large-scale excavation of a hillfort, with 57% of the site
109
opened and studied over the years. This in-depth and thorough excavation produced 24
roundhouses, suggesting an estimated population of 196, and a large number of postholes,
pits, hearths, and forges, as well as huge quantities of finds. Artefacts such as iron ingots and
briquetage provided evidence of trade that Cunliffe suggests were exchanged locally for raw
materials like wool or cereals. While no evidence of smelting was found on site, iron and
bronze working debris, materials, and tools show on-site smithing of metal tools. Evidence of
craft activities with shale, bone, basketry, and so on are rare, but there is evidence that textile
production was a significant activity throughout the life of the hillfort. These findings support
the interpretation of Danebury as a centre of production for certain crafts and of trading, both
locally and long-distance. Cunliffe’s report on Danebury was as detailed and meticulous as
the excavations; the textile tools were studied to the same standards as other craft tools or
debris, with each tool dated as precisely as possible and its distribution mapped at each
phase. This allowed Cunliffe and Poole to determine that although the number of textile tools
increased in the Mid Iron Age, there was no clear centre for textile craft activity at any stage
during Danebury’s occupation.
28 spindle whorls were recorded from the Chilcomb House collections; all made from
locally sourced material (21 from chalk and seven from clay). See Appendix 8, Table 1 for
full tabulated data.
Table 29. The minimum, maximum, and average measurements of Danebury spindle whorls
Diameter Depth Weight
Minimum 29 mm 13 mm 11 g
Average 48.59 mm 23.14 mm 61.24 g
Maximum 85 mm 42 mm 122 g
The excavation report noted that while spindle whorls were generally scattered across the
site, in the Middle to Late Iron Age (c.310-50 B.C) those from the southern area are more
often found deposited alongside multiple other textile tools (loom weights, needles, combs,
etc.) compared to those from the northern area in the same period. Spindle whorls from the
early Iron Age (c.600-400 B.C.) do not follow this pattern, possibly indicating that textile
workers at this site developed a perception of a ‘set’ of tools that required several tools from
different stages of the chaîne opératoire. These ‘sets’ of tools could have symbolised the
110
stages of textile work or the various skills that textile workers needed to learn or master. The
number of spindle whorls found at Danebury suggests limited production of threads, enough
to supply the population but possibly not enough to produce a surplus of textiles.
62 loom weights were recorded from the Chilcomb House collections, 41 made from
locally sourced chalk and 21 from local clay, all roughly shaped. See Appendix 8, Table 2 for
full tabulated data.
Table 30. The minimum, maximum, and average measurements of Danebury loom weights
Width Depth Weight
Minimum 63mm 40mm 170g
Average 125.44mm 87.76mm 1369.94g
Maximum 474mm 165mm 2622g
Two pits produced groups of six loom weights each; three pits each produced groups of
three loom weights; and 2 postholes produced groups of loom weights (six from one and
three from another). Roughly half the loom weights (32) are shaped in the same triangular
form typical of Iron Age loom weights, 22 are elongated with a single hole at the top, and
eight are rounded with a single hole through the centre. Loom weights are common
throughout the Iron Age phases, although there is a slightly larger proportion in the Late Iron
Age (310-50 B.C.). This suggests a slight increase in weaving during the late Iron Age, but it
is not enough to determine whether textile production at this site underwent significant
changes (e.g. from supplying the household to producing for trade or exchange). The number
of loom weights at Danebury is significantly larger than spindle whorls, as is the number of
weaving combs discovered from this site. 71 weaving combs were found, 48 of which were
dated to the late Iron Age. This large number of weaving tools indicates that textile
production not only increased in the late Iron Age, but that the spinners of Danebury may not
have been able to supply the weavers alone. This supports Cunliffe and Sharple’s theory that
the smaller satellite settlements provided Danebury with spun threads for weaving, similar to
the Cornish cliff castles previously discussed.
Other craft production tools are common. A large amount of metalworking debris,
materials, and tools for iron and bronze found throughout the site in all Iron Age contexts.
111
There does not appear to be any single workshop area nor evidence of smelting, metal
appears to have been imported as ingots before being melted down for further use. No
woodworking tools found across the site, although subsistence tools are common. Querns are
scattered across the site and throughout Iron Age contexts. Regional and continental trade is
evidenced at Danebury from Kimmeridge shale bracelets, beads of amber and glass, and a
few sherds of foreign pottery (Cunliffe 1994).
Easton Lane
A small settlement surrounded by a ring ditch (NGR SU 49500 31300); occupied
approximately 800-600 B.C. and excavated in 1971 as part of the M3 extension by Peter
Fasham and the Hampshire Field Club and Archaeology Society. Easton Lane is very close to
Winnall Down, which may be why there are no roundhouse remains present here. Although
its exact use in this period is unclear (Fasham 1982: 50), the proximity of Easton Lane to
Winnall Down and the similarities in the artefacts from these sites imply that Easton Lane
acted as a craft/activity area for the population of Winnall Down (Fasham 1984). Three
phases of activity were identified: Neolithic pre-ring ditch activities; the Bronze Age ring
ditch and occupation; and the Iron Age post-ring ditch activities.
One spindle whorl was recorded from the Chilcomb House collections; made from locally
sourced material. The diameter measures 32 mm; depth measures 46 mm; weight measures
65 g. This was found in feature 4560 along with five loom weights to the north of Winnall
Down, and all tools were dated to the Middle Iron Age, approximately 400-100 B.C.
Eight loom weights were recorded from the Chilcomb House collections, six made from
locally sourced chalk and roughly carved into shape. See Appendix 8, Table 4 for full
tabulated data.
Table 31. The minimum, maximum, and average measurements of Easton Lane loom weights
Width Depth Weight
Minimum 73 mm 60 mm 929 g
Average 105 mm 89.5 mm 1216 g
Maximum 186 mm 118 mm 1992 g
112
All were made from local materials and roughly shaped; although only two share the
triangular form typical of Iron Age loom weights. Other finds follow the pattern seen at
Winnall Down: craft production debris and tools are scarce, as are long-distance traded items.
A few crucible pieces were found scattered across the site; no woodworking tools were
found; a few querns were also discovered in the area. No evidence of long-distance trade was
discovered.
Kennel Farm
An enclosed settlement located at (NGR SU 59555 48561); large excavations in 1966 by
Northamptonshire Archaeology dated the occupation of this site to approximately 800-300
B.C. Kennel Farm is situated close to Brighton Hill South (Hatch Warren), another settlement
of the Mid Iron Age to Roman period to the immediate north-east. Half of the Kennel Farm
site was excavated; no house structures were found in the excavated area so a population
estimate is not possible. However, there was a pattern of four distinct groups of storage pits,
suggesting a chronological sequence of storage pits being used then closed (Chapman 2006:
20). The postholes found during excavation rarely formed any structural patterns, leading the
excavators to believe these indicated single or paired posts used for external activities (such
as looms, haystack posts, etc.) (Chapman 2006: 21). The excavation report was detailed and
clear, with excellent plans of complex features that are clearly labelled, and clear locations of
both the features that produced weaving tools and the layers in which these tools were found.
The loom weights were also described in detail; the report on these tools listed the weights
and locations of each find (although not the width, height, or depth of any of these tools) and
discussed the distribution across the site (Chapman 2006: 47-48). As such, this report was an
excellent and invaluable source of information and context for these tools.
One spindle whorl was recorded from the Chilcomb House collections; made from locally
sourced material. The diameter measures 42 mm; depth measures 20 mm; weight measures 3
g. Found in a pit dated to the Middle Iron Age, approximately 400-100 B.C.
Seven loom weights recorded from the Chilcomb House collections, made from locally
sourced clay and roughly shaped. See Appendix 8, Table 6 for full tabulated data.
Table 32. The minimum, maximum, and average measurements of Kennel Farm loom weights
113
Diameter Depth Weight
Minimum 120 mm 58 mm 1100 g
Average 192 mm 78.38 mm 2334.5 g
Maximum 418.25 mm 93 mm 3183 g
Five loom weights in total were found from the same feature, both whole and fragmented
(not included in study). All were made from local clay and shaped in the same triangular
form typical of Iron Age loom weights. These artefacts are dated to the main occupation
phase of the site, 400-100 BC. These weights and the single spindle whorl represent a site
able to supply a single household with textile, but not to produce a surplus for trading or
exchange. Other craft production tools are scarce and only indicate the burning of flint. No
metalworking, woodworking, or subsistence tools or evidence was found at the site. No
evidence of trade discovered (Chapman 2006).
Winnall Down
An enclosed settlement located in (NGR SU 50600 30000), and excavated by Fasham and the
Hampshire Field Club and Archaeological Society in 1976-1977. The occupation of this site
is dated to roughly 600-50 B.C. This site consists of a single enclosure ditch containing eight
roundhouses (providing an estimated population of 54) and ten 4-post structures, although
not all dwellings or storage buildings would have been built or used at the same time. Use of
the enclosure appears very structured, with certain features and activities restricted to distinct
areas of the site. The northwest corner is characterised by 4-post structures and high levels of
loom weight deposition, while the southeast corner was used for crop processing and storage.
Pits 2630 and 2676, close to 4-post structures G and T, contained about 11 and 14 loom
weights respectively (all fragmented to some degree). Fasham noted that this spatial
correlation suggests structures G and T were weaving workshops (1985:). A few pieces of
briquetage and locally sourced querns were excavated, although no evidence of metalworking
were found. Fasham’s report is clear and detailed in the site’s structure and phasing, as well
as the artefacts uncovered and any difficulties in dating these finds. The textile tools are
reported in good detail and the locations of these tools are given clearly, which is a great help
in identifying a potential weaving work area or presence of a weaving specialist.
114
No spindle whorls were discovered in the enclosure, although two were excavated from
features to the north of Winnall Down’s enclosure, an area called Easton Lane (mentioned
above). Nine loom weights were recorded from the Chilcomb House collections are made
from locally sourced stone and clay, roughly carved into shape. See Appendix 8, Table 7 for
full tabulated data.
Table 33. The minimum, maximum, and average measurements of Winnall Down loom weights
Width Depth Weight
Minimum 142 mm 42 mm 1216 g
Average 164.15 mm 89.5 mm 1591 g
Maximum 185 mm 104 mm 3331 g
The loom weights all come from pit contexts and all were shaped in the same triangular
form typical of Iron Age loom weights; these were dated to the Early Iron Age (700-400
B.C.). This remains a small selection of the weaving tools found at Winnall Down, as the
others were too fragmented to record. Altogether, 41 fragments of 11 loom weights were
discovered from one pit 2630 and another 25 fragments of 14 loom weights from pit 2676.
These pits lay close to two 4-post structures, leading the excavators to suggest that these
structures were related to weaving or even semi-permanent upright looms, which could be
used whether the weather permitted (Fasham 1985: 127). This suggests that Winall Down
was able to produce textiles for its own population, although it is unclear whether it was able
to produce a surplus for further trading as so few spindle whorls were recovered. On the other
hand, the lack of spinning tools could indicate that other, smaller settlements in the
surrounding area supplied spun threads to Winnall Down. The number of total loom weights
corresponds to the amount of loom weights found in fortified settlements elsewhere, such as
Hengistbury Head, Maiden Castle, and so on. This, along with the large groups of weights
found in the same features, suggests that similarly large quantities of textiles were woven at
this site.
Other finds indicate that a variety of craft work occurred at Winnall Down: domestic
activities (such as weaving) in the northwest corner; possible clay digging or chalk quarrying
115
in the northeast corner; worked bone objects, fragments of crucible, and a large amount of
pottery were excavated from a few pits in the south/southeast; and a series of structures in the
southwest corner suggest this area was used for crop-processing and storage. No evidence of
long-distance trade was excavated (Fasham 1985).
Zions Hill Farm
A small unenclosed site, possibly a single farmstead, located in (NGR SU 41500 20200) and
excavated in 1995-1996 by Rawlins, Chadwick, Reynier, and Seager Smith as the prelude to
development work at this site. The finds from Zions Hill have dated the site to roughly 100-0
B.C. This site is unusual in that it has no structural evidence, only pits and gullies; this means
no population estimate can be made. No foreign traded artefacts were discovered; the pottery
is limited to Late Iron Age coarseware made from local materials, similar to finds from other
Iron Age sites in the area. Zions Hill appears to be the working area of a small agricultural
settlement/single farmstead nearby. The excavation of this site came after a series of
field-walking and surface collections revealed artefacts from several periods (Neolithic to
Roman) and pointed to several areas of interest, guiding where excavations occurred. These
excavations revealed limited features, allowing for thorough investigation across the site. The
excavation report provides detailed information on the excavation method and the features
found, including the locations of the weaving tools discussed below (Rawling et al 2003). It
also provides a reasonable overview of the loom weights’ materials, shapes, weights, and
state of repair after excavation. Unfortunately, there were no precise measurements of the
width, depth, or height of these tools included in the report.
Five loom weights were recorded from the Chilcomb House collections, made from locally
sourced clay and roughly shaped. No spindle whorls discovered. See Appendix 8, Table 8 for
full tabulated data.
Table 34. The minimum, maximum, and average measurements of Zions Hill Farm loom weights
Width Depth Weight
Minimum 72 mm 77 mm 522 g
Average 100 mm 91.6 mm 1261.3 g
Maximum 161 mm 106 mm 2760 g
116
Three features produced loom weights, these being pits that were clustered in the centre of
the site and not connected to any structure or ditch features. The triangular loom weights
dated to AD 50-100, while the cylindrical loom weights dated to 50-0 B.C. The small number
of loom weights would be enough to supply a single household but it probably would not be
able to produce a surplus of textiles. No other evidence of craft production tools from this
site. Site pottery indicates on-site production, probably at low levels, and there is no evidence
of long-distance trade (Rawlins et al. 2003).
Kent
East Wear Bay, Folkestone
A possible oppida site at the top of the cliffs at East Wear Bay (NGR TR 24000 36600); this
site has been excavated in 1989 and 2010-2011 by the Canterbury Archaeological Trust.
Evidence showed that the main occupation of this site occurred at roughly 500 B.C.-A.D. 43.
Unfortunately, much of this site has been lost to the sea via cliff erosion, so the remains must
be taken as only part of the whole site. Full size of the Iron Age occupation is unknown but
its high status and large-scale craft production have allowed it to be labelled an oppida. The
site was probably at its largest around 150 B.C.-A.D. 43. Quernstone manufacture was
possibly performed as a production line, as several querns in various stages of manufacture
were abandoned by workers. A Roman villa complex was built over the remains of the
oppida in the late 2nd century A.D. Finds include large quantities of British and continental
pottery, including Italian amphorae; 13 Iron Age coins; large amounts of quern debris and
flint debitage; and a significant number of spindle whorls. Keller’s excavation and report
focused on the prehistoric quern production, rather than the main settlement itself. In
comparison, Colsen’s excavation report looked at the occupation of the oppida, its
development, various craft activities at this site, and the evidence of trade links with the
continent and elsewhere in Britain.
29 spindle whorls were recorded from the Canterbury Archaeological Trust archives; all
made from locally sourced stone and clay. See Appendix 9, Table 1 for full tabulated data.
Table 35. The minimum, maximum, and average measurements of East Wear Bay spindle whorls
117
Diameter Depth Weight
Minimum 7.3 mm 5.3 mm 7 g
Average 38.5 mm 12 mm 23 g
Maximum 62 mm 29.5 mm 46 g
Most of these tools were found throughout the site and date to approximately 150 B.C. -
A.D. 43. The number of spindle whorls indicates a large amount of threads were produced on
this site, possibly enough to provide a surplus for trade and exchange.
Two loom weights were recorded from the Trust archives, made from locally sourced clay
and roughly carved into shape.
Table 36. The minimum, maximum, and average measurements of East Wear Bay loom weights
Museum number Width Depth Weight
689 103 mm 132 mm 1340 g
612 180 mm 158 mm 1423.75 g
These tools have the same triangular form typical of Iron Age loom weights and are dated
to the main phase of the site (150 B.C. - A.D. 43). It is strange that so few loom weights were
discovered, considering the large number of spindle whorls excavated; it is possible that more
weights were deposited elsewhere in the site rather than in the domestic contexts that the
spinning tools were found in. Until further excavations are conducted, the weaving
capabilities of East Wear Bay will remain unclear. Other craft production tools are scarce.
Few metal or wood working tools were found, although large quantities of local greenstone
quern fragments were discovered in two outlying trenches. Added to evidence discovered in
earlier excavations, it is clear that this large site specialised in quernstone production that was
traded along the coast, further inland, and across the channel. The discovery of 13 Iron Age
coins, a large amount of foreign pottery, and wine amphorae imported from Italy further
highlights the high status of East Wear Bay and a trading site of significant importance
(Parfitt 2012).
118
Highstead
An enclosed settlement located on marshand near Chislet (NGR TR 21300 66100) and
excavated by Bennett, Couldrey, and Macpherson-Grant in 1975-1977 prior to gravel
extraction. The Iron Age occupation of this site is dated to approximately 800 B.C.-A.D. 43.
In the Late Bronze Age/Early Iron Age, the site had 3 enclosures. Enclosure B70 in the north
of the site was rectangular with defensive palisade post holes and a gatehouse within its
ditch. A hearth and clay-lined pits (storage areas) excavated within B70 indicate occupation;
pottery, copper dross, and vitrified clay finds from the enclosure ditch point towards crafting
use rather than domestic. Enclosure A24 (to the southwest) was oval in shape and domestic,
with two roundhouses (built sequentially) in the centre, suggesting a population of seven at
any one time. The Middle Iron Age (600-400 BC) is characterised by an open settlement;
occupation remains constant, with copper, pottery, and subsistence activities continuing.
Debris and remains of Ironworking and salt production are found during this period, along
with 4-post structures. This is followed by a period of abandonment until 100 BC, when a
series of rectangular enclosures are dug, possibly for livestock management. No evidence of
further occupation is recorded, although pottery evidence suggests occupation in the
surrounding area. The site suffered some ploughing damage prior to excavation. The
excavation report is a lengthy, detailed record of the excavations with a good discussion on
the site’s importance in Kent’s Bronze and Iron Age landscapes, as well as for our current
understanding of these periods. Unfortunately, while a great deal of detail is given to these
discussion and the excavation methods and features, little detail is given on the textile tools
discovered at this site. It is difficult to discover the location and context of these tools, nor are
there discussions of the distribution of such tools. As is the case with many textile tools from
excavations in the past, more time was spent on the ceramic or metallurgical analysis than
these.
Only four textile tools in total were excavated from this site; although this did not match
the parameters for site selection, there were too few Iron Age sites in Kent that produced any
textile tools. As Highstead produced the highest number of textile tools compared to other
sites, it was included in order to have enough data for analysis in Kent.
119
Two spindle whorls recorded from the Canterbury Archaeological Trust archives; all made
from locally sourced material. Both whorls share the same diameter. See Appendix 9, Table 3
for full tabulated data.
Table 37. The measurements of Highstead spindle whorls
Museum number Diameter Depth Weight
41 40 mm 6 mm 45 g
40 40 mm 20 mm 52 g
One tool was found in each of the Early Iron Age enclosures in the southwest of the site,
dated to approximately 800-550 B.C.
Two loom weights recorded from the Trust archives, made from locally sourced clay and
roughly carved into shape. See Appendix 9, Table 4 for full tabulated data.
Table 38. The measurements of Highstead loom weights
Museum number Width Depth Weight
99 70 mm 106 mm 1400 g
70 109 mm 109 mm 1750 g
Both tools are shaped in the same triangular form typical of Iron Age loom weights; one
loom weight was oval in shape and came from a domestic context in the southwest of the
site, where two roundhouses (built sequentially) were in the centre of the excavation area.
These artefacts are dated to the Early Iron Age of the site, c.800-550 B.C. These suggest that
Highstead was able to supply its community with textiles, but not create a surplus for further
trade or exchange. Other craft production tools are scarce, with no metal- or wood-working
tools or debris. Pottery finds consisted of utilitarian cooking ware and there was no evidence
of long-distance trading connections (Bennett et al. 2007).
Sunset Caravan Park
A site of unclear character located in Whitstable, along Kent’s north coast (NGR TR 10300
64600), and found by Canterbury Archaeological Trust as part of a rescue excavation prior to
120
building work in 1998-1999. The Iron Age occupation is dated to roughly 700 B.C.-A.D.
100. Given the confines of this excavation, the extent of this site and whether it was enclosed
or defended in any way is unclear, although it has been posited that the nearby excavations of
Church Lane East (TR 10000 64700) represent a continuation of the Sunset Caravan Park
settlement, as it contains similar features and finds of the same date. Roundhouse features
were found throughout the excavation, demonstrating occupation at this site and allowing for
an estimated population size of over 21. Few metal- and wood-working tools or debris were
found scattered about the site; however, a large number of pottery kilns and pottery debris
suggests an extensive ceramic industry that expanded during the Late Iron Age period.
Remains of briquetage were found across the site, indicating large scale salt production
throughout the Iron Age. Multiple clay and chalk quarries found on site demonstrate
widespread quarrying during the Iron Age, increasing in the Late Iron Age period. Finds of
Early Iron Age Italian amphorae and other continental pottery wares demonstrate that
long-distance trade was an important part of the economy of this site (Allen 2000).
As these excavations were conducted by commercial units prior to building work, the
excavation reports are ‘grey literature’ and difficult to find. Fortunately, a summary of the
work was published by the Canterbury Archaeological Trust (Allen 2000). This presented
little detail on the location of finds, but more information on the breadth of artefacts
excavated as well as the clustering of the settlement’s structures. No map or plan of the site
and its features was included in this summary, which would have benefitted from it.
Five spindle whorls were recorded from the Canterbury Archaeological Trust archives; all
made from locally sourced clay. See Appendix 9, Table 5 for full tabulated data.
Table 39. The minimum, maximum, and average measurements of Sunset Caravan Park spindle whorls
Diameter Depth Weight
Minimum 21 mm 19.5 mm 20 g
Average 32.2 mm 27 mm 31 g
Maximum 39 mm 36 mm 44 g
121
These whorls were dated to the Late Iron Age, approximately 150-0 B.C. No loom weights
were recovered from the site, which suggests that another form of weaving occurred in this
settlement. These spindle whorls suggest that this site was able to supply its community with
textiles, although it could probably not produce a surplus for further trade or exchange.
Thanet Earth
An enclosed settlement located in (NGR TR 28900 66700) and excavated by Canterbury
Archaeological Trust in 2007-2008. This was a large excavation conducted prior to building
work, which revealed features spanning the Neolithic to the Modern period. The Iron Age
occupation of this site was dated to 500 B.C.-A.D. 43; most features from this date were
clustered on plateau 8, including ditches, pits, postholes, ring gullies, and the remains of at
least three roundhouses. This provides an estimated population size of 21. This site also has
domestic features within and beyond its enclosure ditch, suggesting that the population
expanded during the Middle Iron Age. Plateau 8 also produced an unusually high number of
Iron Age spinning and weaving tools, suggesting textile production was a major craft activity
at the site. Other craft production tools are common. A quantity of iron slag was excavated
from various large pits in Plateau 8, demonstrating that smelting was occuring on site. Pottery
production was another large-scale activity at this site, with kilns and a high quantity of
pottery sherds made of local clay excavated from this area as well. There is little to no
mention of wood- or shale-working tools and debris, or quernstones and other agricultural
tools in the excavation reports (Rady 2009, 2010a, 2010b). Evidence of trade at Thanet Earth
is limited to a few potin coins but Rady (2010a) believes that the large number and size of the
storage pits indicates that Thanet Earth was an important site, possibly used as a centre for
various communities to meet or feast. The occupation decreased in the Late Iron Age before
the site was completely abandoned. The excavation of Thanet Earth was well managed and
thorough, despite the size of the site and the complexity of some of the features. Likewise,
the reports of these excavations are well presented and detailed; the locations of the Iron Age
settlement is clear and the textile tools are given context and described in suitable detail.
One spindle whorl was recorded from the Canterbury Archaeological Trust archives; made
from locally sourced clay. The diameter measures 33.5 mm; depth measures 27 mm; and
weight measures 14 g. This whorl was dated to the Middle Iron Age, approximately 500-300
B.C. This is one of four clay spindle whorls from this period excavated at the site, three stone
spindle whorls were also found; although the presence of these tools are recorded in the
122
excavation report, no measurements or weights were noted. These additional tools could not
be located in the archives and therefore were not included in this dataset.
Five loom weights were recorded from the Trust archives, made from locally sourced clay
and roughly carved into shape. See Appendix 9, Table 7 for full tabulated data.
Table 40. The minimum, maximum, and average measurements of Thanet Earth loom weights
Width Depth Weight
Minimum 150 mm 54 mm 291 g
Average 162.45 mm 59.7 mm 847.2 g
Maximum 183.75 mm 70 mm 1177.5 g
All these loom weights were shaped in the same triangular form typical of Iron Age loom
weights and found in Plateau 8. These artefacts are dated to the Middle Iron Age phase of the
site. A total of 31 clay loom weights were excavated from the Plateau 8 area and the
excavation report noted that four of these tools came from context 12365. Unfortunately, the
dimensions and weights of these finds were not recorded (Rady 2010: 88-89). These
additional loom weights could not be located in the Trust’s archives and therefore were not
included in this dataset. However, the high numbers of loom weights excavated at Plateau 8
compared to the other Plateaus with Iron Age features strongly suggests that weaving was a
major activity at this settlement that could produce a surplus of textiles for trade and
exchange, with Plateau 8 being the centre of this activity.
123
5. Discussion
Tool analysis
I compared the data gathered in the previous chapters with the results of Grömer’s
experiments on spindle whorl speed and thread thickness (2005) to determine the most likely
threads and fabric produced by these tools. I analysed the loom weight data using the results
of Mårtenson’s experiments, adapted to the spindle whorl data here. By using the information
in Grömer and Mårtenson’s experiments, I aimed to discover broad patterns of production in
certain types of site and across the various counties. I focused on analysing data clusters
where possible, as these indicate the most common (and likely, the most desired) forms of
tool used and fabrics produced at these sites.
County by county overview - Spindle whorls
Figure 20. The diameter and weight of spindle whorls from all sites, separated by county.
The total number of spindle whorls recorded in each county is as follows:
Cornwall - 67 Hampshire - 31
Devon - 35 Kent - 32
Dorset - 150
124
As demonstrated in figure 20 (above), the majority of spindle whorls from all counties
overlap each other; whorls under 50 g in weight and 50 mm in diameter clearly answered the
needs of most sites in all counties. However, there are a significant number of whorls from
Dorset and Hampshire that increase in both weight and diameter. This combination of heavy
whorls with wide diameters indicates tools that can both turn for longer at a slower pace than
others and may produce thicker threads than in other counties. Such whorls would be
excellent for spinning long plant fibres or plying plant or wool threads together. Since plying
can be practiced anywhere, I would expect there to be a pattern of a few wide, heavy spindle
whorls throughout the counties that would indicate plying took place.
The unusual concentration of such wide and heavy whorls in Dorset and Hampshire does
not correspond to the use of wool alone, but instead to the use of plant fibres (see below for
further discussion). Although difficult to distinguish in fig 20, these heavier and larger whorls
are present in all Dorset sites (one in Allard’s Quarry, six in Hengistbury Head, three in Hod
Hill, and 13 in Maiden Castle) and all but one are found in Danebury in Hampshire, where
they comprise over half of all spindle whorls (18 out of 29 whorls). The predominance of
heavier whorls in Danebury indicates that this site was producing a larger quantity of thicker
thread than any other, possibly in a specialist capacity. There are clusters present in each
county’s dataset, which demonstrate the most common forms of spindle whorl (and by
extension, threads that they produce) across the county. These data clusters are presented
below, in table 40, along with the thread thickness they are most likely to produce and the
speed at which the whorl can turn (based on Grömer 2005).
125
Table 41. Measurements of spindle whorl clusters from all sites and counties, the most likely thread thickness
they would produce, and the general speed of their spin.
County Min.
diameter
(mm)
Max
diameter
(mm)
Min weight
(g)
Max
weight
(g)
Likely thread
diameter
(mm)
Likely
speed of
spin
Likely
warp
tension (g)
Cornwall
Group 1 28 42 5 10 0.2 - 0.4 Very fast 10 - 20
Group 2 27 38 10 20 0.2 - 0.5 Very fast 10 - 25
Devon
Group 1 25 40 15 23 0.4 - 0.7
Very fast -
fast 25 - 30
Group 2 32 37 24 32 0.5 - 1
Fast -
medium 25 - 40
Dorset
Group 1 25 45 2 25 0.2 - 0.7
Very fast -
fast 10 - 35
Group 2 30 50 25 38 0.5 - 1.5
Fast -
medium 25 - 60
Group 3 40 52 45 55 1 - 2
Medium -
slow 40 - 80
Hampshire
27 71 11 111 0.4 - 2
Very fast -
slow 25 - 80
Kent
Group 1 26 42 11 20 0.2 - 0.5
Very fast -
fast 10 - 25
Group 2 32 47 23 24 0.5 - 0.7
Very fast -
fast 25 - 35
Group 3 21 62 27 35 0.6 - 1
Fast -
medium 30 - 40
Group 4 44 46 36 57 0.8 - 2 Fast - slow 40 - 80
In most counties, these clusters can produce a range of thread thicknesses, from very fine
(0.2 - 0.3 mm) to very thick (1.5 - 2 mm); however, Cornwall’s spindle whorls are restricted
to fine threads alone. This remarkable departure from the common dimensions of spindle
whorls used elsewhere could be reflective of the materials available to the spinners in these
areas (see below). The heavier spindle whorls in all counties (except Cornwall) could
demonstrate a larger presence of fibers like flax or nettle by spinners in these areas. Likewise,
the absence of heavy spindle whorls in Cornwall’s archaeological record is a strong
indication that textile production in this county was focused on producing fine threads from
animal fibers alone.
Another point of interest is the overall data from Hampshire; here, the spindle whorls have
a wider variety of weights and diameters, forming multiple smaller clusters than the data
from other counties. There are two possible reasons for such a range of measurements.
Firstly, the sites in Hampshire could have required different forms of textiles, with each site
126
producing fabrics fit for its own purpose. If this were the case, I would expect to find
weaving tools at all sites regardless of size or the presence of other craft tools. Secondly,
these sites could belong to various parts of the Iron Age; if textile production in the British
Iron Age followed the same pattern seen in the European Iron Age, then early sites are more
likely to have spindle whorls with a wide range of weights and dimensions compared to
spindle whorls with a more limited range of measurement. As it is, there are weaving tools at
every Hampshire settlement, suggesting that each settlement in this area produced textiles for
its own use; in sites able to support a community of workers, such as Danebury, textile
workers could spend more time on spinning and weaving surplus textiles, meaning this was a
site of household specialists. Unfortunately, the archaeological evidence remains unclear as
to whether individual spinners and weavers worked together as a workshop or were otherwise
organised as a larger community of specialists, each focused on their own part of the textile
production process.
Loom weights
Figure 21. The weights and depths of loom weights from all sites, separated by county.
127
The total number of loom weights from each county:
Cornwall - 27 Hampshire - 97
Devon - 14 Kent - 9
Dorset - 57
Unlike spindle whorls, loom weights do not have a clear overlap to indicate the range of
weights and dimensions of tools that could produce most of what a site required. Instead,
each county appears to have its own requirements for loom weights, most likely based on the
sort of textiles each area produced (see table 42 below). Any areas of overlap appear to be
largely co-incidental, except for Dorset and Hampshire, where the majority of loom weights
share the same range of weights and dimensions. The most likely reason for this is that most
of the loom weights in these counties come from hillforts, Danebury in Hampshire and
Maiden Castle, Hengistbury Head, and Hod Hill in Dorset. As centres of settlement in these
regions, it is more likely that these were also centres of craft production that produced
surplus textiles for further distribution or trade. Given the range of spindle whorl weights and
dimensions in these counties, as well as the range of loom weights seen above, it is likely that
these sites produced a variety of textiles rather than specialise in larger production of more
standardised textiles. This hypothesis will be examined further below, in a closer look at
fortified site data. It is more difficult to ascertain the form of loom weights in these counties’
smaller settlements, although it is clear that Dorset has a wider variety of loom weights
compared to Hampshire. It is possible that this larger spread of data represents these smaller
sites weaving items to suit individual needs. Kent loom weights form two separate, loose
clusters that have a similar range of weights and dimensions to Dorset and Hampshire,
although this region produced the fewest loom weights, which means any conclusions drawn
must be treated with caution.
Two other counties with intriguing data overlaps are Cornwall and Devon Cornish loom
weights are lighter and thinner than any other county’s loom weights, similar to its spindle
whorls, in part due to the slate that they are made from. This material is the most common
stone found across Cornwall and allows these tools to be easily cut to similar depths and
weights. The Devon loom weights share the similar weights, differing only in depth. The
restriction in the weights of these weaving tools suggests either high specialisation in textile
production (e.g. very fine textiles compared to other areas) or more flexibility in weaving.
With lighter loom weights it would be possible for weavers to more easily alter the tension
placed on the warp threads than with heavier weights, allowing a weaver to produce fine
128
textiles, open weaves, warp-faced fabric, or to work with thick threads. It is only by
comparing the loom weight data with the spindle whorl data from these regions that it is
possible to determine whether these are specialist or more general tools.
Table 42. Measurements of loom weight clusters from all sites and counties, the thread thickness best suited for
weaving
County Min. depth
(mm)
Max depth
(mm)
Min weight
(g)
Max weight
(g)
Threads per loom weight for
following tension
10g 20g 30g 40g 60g 80g
Cornwall
Group 1 2 10 6 59 0 - 6 0 - 3 N/A N/A N/A N/A
Group 2 6 17 108 258
10 -
25
5 -
12 N/A N/A N/A N/A
Devon
Group 1 13 64 60 87 N/A 3 - 4 2 - 2 1 - 2 N/A N/A
Group 2 78 94 154 181
15 -
18 7 - 9 5 - 6 3 - 4 N/A N/A
Dorset
Group 1 21 40 193 379
19 -
37
9 -
18
6 -
12 4 - 9 3 - 6 2 - 4
Group 2 52 102 1526 2082
152 -
208
76 -
104
50 -
69
38 -
52
25 -
34
19 -
26
Hampshire
Group 1 60 100 862 1212 N/A
43 -
60
28 -
40
21 -
30
14 -
20
10 -
15
Group 2 47 77 1251 1516 N/A
62 -
75
41 -
50
31 -
37
20 -
25
15 -
18
Group 3 92 138 1114 1418 N/A
55 -
70
37 -
47
27 -
35
18 -
23
13 -
17
Group 4 60 93 1612 1808 N/A
80 -
90
53 -
60
40 -
45
26 -
30
20 -
22
Group 5 58 90 1803 2061 N/A
90 -
103
60 -
68
45 -
51
30 -
34
22 -
25
Kent
Group 1 56 70 802 1155
80 -
115
40 -
57
26 -
38
20 -
28
13 -
19
13 -
14
Group 2 100 158 1340 1424
134 -
142
67 -
71
44 -
47
33 -
35
22 -
23
16 -
17
In these calculations, N/A is used to denote that no spindle whorls capable of spinning
threads requiring these tensions were present in the corresponding county.
Cornwall loom weight Group 1 is too light to weave with either Cornwall spindle whorl
groups, suggesting these were not designed for use on a loom but could have been weights
for nets instead. Loom weight Group 2 is optimally suited for working with the threads spun
at these sites. Working with the thinnest threads in particular would have resulted in having
many threads on each loom weight; considering how thin these loom weights and the threads
spun at these sites are, this would have allowed the weavers to put many loom weights
129
side-by-side on a single loom, which would have enabled the Cornish settlements to easily
produce warp-faced textiles as well as tightly woven fabrics. Warp-faced textiles have a very
distinct appearance compared to plain tabby, which can be confused with some twill weaves,
but are difficult to achieve with wider (or deeper) loom weights. As mentioned in the tool
analysis chapter, too many loom weights side-by-side can pull the fabric on the loom out of
shape and risk ruining both the strength of the fabric and the appearance of warp-faced
textiles. It is possible that the Iron Age Cornish weavers were able to take advantage of the
slate that they made their loom weights from to become specialist producers of this unusual
and otherwise difficult to produce style of weaving.
Similar to the Cornish loom weight groups, Devon’s loom weight Group 1 is not suited for
either Devon spindle whorl group, although Devon’s Group 2 is suited for working with the
threads spun by both Devon spindle whorl groups. Unlike the Cornish loom weights, these
tools appear to work best with fewer threads attached to each loom weight and as they are
deeper than the Cornish loom weights, this would limit how many could be attached to a
loom before the fabric would be pulled out of shape. Together with the thickness of the
threads spun by the Devon tools, this seems to suggest that weavers were either producing
textiles with an open weave or a tight balanced tabby weave, and few warp-faced fabrics.
Dorset loom weight Group 1 works well with Dorset spindle whorl Group 1 and 2, but not
with Group 3. The depth of these loom weights, combined with the number and thickness of
the threads attached to each weight, suggests these tools were best suited to weaving a tight
balanced tabby with the finer threads and a more open weave with the thicker ones. In
comparison, Dorset loom weight Group 2 is too heavy to work with threads produced by
Dorset spindle whorl Group 1 and most of Group 2; however, it does suit any threads needing
over 60 g of tension (like the rest of Dorset spindle whorl Group 2 and all of Group 3). The
number of threads per loom weight needed to provide 60-80 g of tension compared to the
depth of these tools indicates they could best weave a very tight tabby weave with these thick
threads. They may be able to produce a warp-faced fabric, but the number of loom weights
required would have a combined width more likely to pull the fabric out of shape.
Hampshire loom weight groups 1, 2, and 3 would be unlikely to weave with any threads
needing 20-30 g. The depths of the tools in these groups combined with the thickness of the
threads spun at these sites suggest that they are best suited to weaving a tight, balanced tabby
130
weave. In comparison, Hampshire loom weights Group 4 and 5 are unlikely to be able to
weave any threads requiring below 60 g in tension, yet they need fewer threads per loom
weight. This implies that they would be able to produce a more open weave despite the
thickness of the threads used.
Kent loom weight Group 1 is best suited to weaving with threads requiring over 30 g of
tension; the depth of these tools and the number of threads required suggests they could
produce a tight, balanced tabby weave or possibly a warp-faced fabric. On the other hand,
Kent loom weight Group 2 works best with thick threads needing over 60 g of tension,
although there are fewer threads needed per loom weight. This low number of threads
combined with the depth of these tools and the thickness of the threads spun at their sites
indicates they are best suited to producing a tight balanced tabby or an open weave.
Tools by settlement type - Fortified sites: Spindle whorls
Figure 22. The diameter and weight of spindle whorls from fortified sites.
The total number of spindle whorls from each fortified site is as follows:
The Rumps (Cornwall) - 15 Hengistbury Head (Dorset) - 39
Trevelgue Head (Cornwall) - 22 Hod Hill (Dorset) - 35
Blackberry Castle (Devon) - 1 Maiden Castle (Dorset) - 58
Mount Batten (Devon) - 5 Danebury (Hampshire) - 29
131
Milber Down (Devon) - 5 East Wear Bay (Kent) - 29
The data gathered from fortified sites broadly follows the same patterns as the counties in
general for a simple reason: hillforts and other fortified sites provided most of the data for
this study, particularly from Dorset sites (Maiden Castle and Hengistbury Head in particular).
These sites were subject to more rigorous excavation than other sites, although this alone
cannot explain the disproportionate number of tools recovered from these sites compared to
other types of settlements. The higher quantities of spindle whorls represent larger
populations at these sites than at others; it is reasonable to assume that a farmstead with one
or two families would have a smaller number of textile tools compared to a site occupied by
whole communities.
When the data clusters are split into their individual sites (see table 43, below), it is
possible to see what each site was producing in more detail. In particular, it becomes clear
which sites have broad ranges of spindle whorls and which have clusters of whorls for
specific purposes. Both Cornish sites have two clear, separate clusters that produce similar
thread thicknesses: The Rumps 1 (six of 15 whorls) and Trevelgue Head 1 (11 of 22 whorls)
are best suited to spinning very fine threads at very high speeds. The Rumps 2 (seven of 15
whorls) and  Trevelgue Head 2 (ten of 22 whorls) are best suited for spinning fine to medium
thickness thread, also at very high speeds; both groups are perfect for spinning animal fibres.
Devon sites have fewer tools, possibly due to poor preservation from the acidic soil
conditions, which makes determining patterns and trends more difficult. Milber Down has a
range of lighter whorls (three of five) that match what is produced at Cornish sites, while
Mount Batten’s spindle whorls (four of five) are more suited to producing a wide range of
thread thicknesses. This would allow spinners to work with both animal and plant fibres,
especially as the whorls from this site have more varied rates of spin than the previously
mentioned sites.
Of the Dorset sites, two have clearly defined sets of tools able to produce all ranges of
thread thickness with little overlap between each group. Hod Hill has smaller clusters, with
Hod Hill 1 (nine of 35 spindle whorls) best suited to spinning very fine threads from animal
fibres and Hod Hill 2 (11 of 35 whorls) best suited to spinning fine to medium thickness
threads, also probably suited to animal fibres but capable of spinning plant fibres. Maiden
Castle and Danebury each have 3 clear groups; Maiden Castle 1 (16 of 58 spindle whorls)
132
and Danebury 1 (six of 29 spindle whorls) are both able to spin fine to medium thickness
threads, probably best suited to spinning animal fibres. Maiden Castle 2 (19 of 58 spindle
whorls) is best suited to spinning medium to thicker threads, either with animal or plant
fibres. Maiden Castle 3 (58 of 39 spindle whorls) and Danebury 2 (ten of 29 spindle whorls)
can produce medium to thick threads using both animal and plant fibres without much
difficulty. Danebury 3 (eight of 29) is only able to spin thick threads (using animal fibres),
but could spin plant fibres as well. In comparison, Hengistbury Head (Dorset) and East Wear
Bay (Kent) have no such weight clusters, with an even range of whorl weights in both sites.
Table 43. Measurements of spindle whorl clusters from fortified settlements, the most likely thread thickness
they would produce, and the general speed of their spin.
County Min.
diameter
(mm)
Max
diameter
(mm)
Min weight
(g)
Max weight
(g)
Likely thread
diameter
(mm)
Likely
speed of
spin
Likely
warp
tension (g)
Cornwall
The Rumps
1 28 34 5 7 0.2 - 0.3 Very fast 10
The Rumps
2 32 39 15 20 0.4 - 0.7
Very fast -
fast 25 - 35
Trevelgue
Head 1 29 46 4 9 0.2 - 0.4 Very fast 10 - 20
Trevelgue
Head 2 27 38 11 18 0.4 - 0.6
Very fast -
fast 25 - 30
Devon
Milber
Down 28 36 7 27 0.2 - 0.7
Very fast -
fast 10 - 35
Mount
Batten 30 36 26 36 0.6 - 1.5
Fast -
medium 30 - 60
Dorset
Hengistbury
Head 27 59 6 37 0.2 - 1.5
Very fast -
medium 10 - 60
Hod Hill 1 19 33 6 10 0.2 - 0.5 Very fast 10 - 25
Hod Hill 2 32 47 26 35 0.6 - 1
Fast -
medium 30 - 40
Maiden
Castle 1 27 46 14 24 0.4 - 0.7
Very fast =
Fast 25 - 35
Maiden
Castle 2 32 57 27 42 0.6 - 1
Very fast -
fast 30 - 40
Maiden
Castle 3 32 53 46 68 0.8 - 1.5
Fast -
medium 40 - 60
Hampshire
Danebury 1 29 43 11 27 0.4 - 0.7
Very fast -
fast 24 - 35
Danebury 2 37 57 35 70 0.8 - 1.5
Fast -
medium 40 - 60
Danebury 3 40 66 80 94 1 - 2 Medium - 40 - 80
133
Slow
Kent
East Wear
Bay 25 63 6 30 0.2 - 1
Very fast -
medium 10 - 40
The confinement of spindle whorls to a few groups of limited weights suggests that
spinners at these sites produced large quantities of threads that had to be of similar
thicknesses. Rather than needing spindle whorls of all weights to make threads for any
purpose, spinners at these sites chose whorls of limited weights in order to produce thread for
specific purposes. The sites of The Rumps, Trevelgue Head, Maiden Castle, Danebury, and
Hod Hill are most likely centres of textile production, either for specific forms of textiles
(fabrics made from fine thread, as at The Rumps and Trevelgue Head) or for greater
quantities of fabrics in general (as at Maiden Castle, Danebury, and Hod Hill). This
conclusion can be tested by analysing the loom weights, using the methods developed by the
CCTR (Nosch, Andersson), and comparing those results with the types of threads spun at the
same sites.
Loom weights
Figure 23. The weights and depths of loom weights from fortified sites in all counties.
The number of loom weights from these sites:
The Rumps (Cornwall) - 12 Hod Hill (Dorset) - 12
Trevelgue Head (Cornwall) - 2 Maiden Castle (Dorset) - 19
Blackbury Castle (Devon) - 4 Danebury (Hampshire) - 62
134
Milber Down (Devon) - 2 East Wear Bay (Kent) - 2
Hengistbury Head (Dorset) - 10
As with the spindle whorls, fortified sites provided the majority of loom weights in this
study; however, this may be explained in part by the more modern excavation techniques
used in many of these fortified sites, such as at Danebury. By and large, the Dorset and
Hampshire sites produced most of the loom weights studied here, aside from The Rumps in
Cornwall (see table 44 below). Given the high numbers from these areas, the scarcity of loom
weights in Devon and Kent is worthy of note; this could be reflective of the nature of these
site’s occupation (i.e.: that these are not permanently occupied sites) or demonstrating that
weaving was not commonly conducted at fortified sites in these areas.
Of the Cornish sites, The Rumps loom weights 2 are well suited for weaving all threads
made at the site. The high number of 10 g threads per loom weight suggests possible
production of warp-faced fabrics, while the low number of threads per loom weight that
require 20-30 g of tension are better suited to producing an open, balanced tabby. The
Trevelgue Head loom weights can work with all threads produced at the site, although the
lighter loom weights cannot work with any threads requiring over 10 g of tension and the
heavier weights would struggle to weave with the thinnest threads. Given how few loom
weights were discovered at this site, it is difficult to determine the most likely form of fabric
produced at this site, although it is unlikely that warp-faced fabrics could be produced here.
Likewise, there were too few loom weights from Blackbury Castle in Devon to determine
the most likely type of fabrics woven at the site. However, the lightest loom weights work
well with threads requiring 10 g of tension, while the heavier loom weights work best with
threads needing 60-80 g of tension; therefore it is most likely that this site best produced
tight, balanced or open weaves. In comparison, the Milber Down loom weights are best
suited to threads requiring 10 g of tension, although they can weave with threads needing
20-30 g tension. The high number of threads per loom weight here combined with the depth
of these tools suggests they were best suited to producing a warp-faced or tight, balanced
weave.
135
Table 44. Measurements of fortified sites loom weight clusters and thread thickness best suited for weaving
County Min. depth
(mm)
Max depth
(mm)
Min weight
(g)
Max weight
(g)
Threads per loom weight for
following tension
10g 20g 30g 40g 60g 80g
Cornwall
The Rumps 1 3 6 11 59 1 - 6 0 - 3 0 - 2 N/A N/A N/A
The Rumps 2 6 17 108 236
11 -
23
5 -
11 3 - 7 N/A N/A N/A
Trevelgue
Head 7 36.2 56.85 654
5 -
65
2 -
32
1 -
21 N/A N/A N/A
Devon
Blackbury
Castle 30 162 78 2040
7 -
204
3 -
102
2 -
68
1 -
51
1 -
34
0 -
25
Milber Down 13 15 64.4 181.25
6 -
18 3 - 9 2 - 6 N/A N/A N/A
Dorset
Hengistbury
Head 40 85 1090 1360
109 -
136
54 -
68
36 -
45
27 -
34
18 -
22 N/A
Hod Hill 1 70 100 1625 1807
162 -
180
81 -
90
54 -
60
40 -
45 N/A N/A
Hod Hill 2 117 164 2466.6 2996.25
246 -
299
123 -
149
82 -
99
61 -
74 N/A N/A
Maiden
Castle 1 21 83.3 193 477.7 N/A
9 -
23
6 -
15
4 -
11 3 - 7 N.A
Maiden
Castle 2 59 72 973.75 1255 N/A
48 -
62
32 -
41
24 -
31
16 -
20 N.A
Maiden
Castle 3 76.6 102 1570 1820 N/A
78 -
91
52 -
60
39 -
45
26 -
30 N/A
Hampshire
Danebury 1 86 121 202 518 N/A
10 -
25
6 -
17
5 -
12 3 - 8 2 - 6
Danebury 2 52 77 951 1212 N/A
47 -
60
31 -
40
23 -
30
15 -
20
11 -
15
Danebury 3 92 138 1114 1418 N/A
55 -
70
37 -
47
27 -
35
18 -
23
13 -
17
Danebury 4 59 78 1301 1460 N/A
65 -
73
43 -
48
32 -
36
21 -
24
16 -
18
Kent
East Wear
Bay 132 158 1340 1423.75
134 -
142
67 -
71
44 -
47
33 -
35 N/A N/A
In these calculations, N/A is used to denote that no spindle whorls capable of spinning
threads requiring these tensions were present in the corresponding county.
Of the Dorset sites, Hengistbury Head loom weights can weave threads needing 30 g of
tension, but work best with threads requiring 40-60 g of tension. Combined with the depth of
the tools, the fabric most likely to be produced at this site is an open tabby. Both Hod Hill
136
loom weights 1 and 2 are unlikely to be able to weave any of the threads produced at the site;
either these weights were used with plant fibres or these were not used on a loom at all.
However, Hod Hill 1 may be able to weave with threads requiring 40g of tension; if so, they
would be best suited to producing an open weave fabric. In comparison, Maiden Castle loom
weights 1 is excellently suited to working with any of the threads spun on site and could
produce tight, balanced or open fabrics. However, Maiden Castle loom weights 2 does not
work well with threads requiring 20 g of tension, instead working best with threads of 30-60
g of tension. Likewise, Maiden Castle loom weights 3 can only work with threads needing 60
g tension; both this group and Maiden Castle 2 would best suit weaving open fabrics.
In Hampshire, Danebury Group 1 is best suited to weaving threads with 20-40 g, although
they could work with those needing 60-80 g of tension too; these tools would be able to
produce tight, balanced or open fabrics. Danebury Group 2 works best with threads needing
40-80 g of tension; these can also work with threads of 30 g tension, although not with
anything lighter. This group is able to weave open fabrics with threads requiring around 40 g,
although these are best suited to producing tight balanced fabrics with the thicker threads.
Danebury Group 3 is also best suited to threads requiring 40-80 g of tension and can also
work with threads needing 30 g, but nothing needing lighter tensions. Likewise, Danebury
Group 4 can only work with 40-80 g of tension, not with any other threads produced on site.
Both Danebury 3 and 4 are best suited to weaving open fabrics.
In Kent, East Wear Bay would struggle to work with any of the threads produced on site,
although they would be able to weave with the thicker threads from the site that require 30 g
of tension. These tools would be best suited to weaving open fabrics.
137
Enclosed sites: Spindle whorls
Figure 24. The diameter and weight of spindle whorls from enclosed sites.
Enclosed sites produced a more limited number of textile tools in total:
Bodrifty(Cornwall) - 6 Kennel Farm (Hampshire) - 1
Trevisker (Cornwall) - 10 Highstead (Kent) - 2
Gussage All Saints (Dorset) - 9 Thanet Earth (Kent) - 1
Easton Lane (Hampshire) - 1
Although figure 24 (above) appears to demonstrate clear clusters of spindle whorl weights,
only the Cornish and Dorset sites have enough tools to make a reasonable sample size to
analyse. These few numbers from the Hampshire and Kent sites make identifying patterns or
data clusters unreliable; however, the fact that there are so few spindle whorls from these
areas is worth exploring. It is possible that these sites were not focused on craft production of
any kind. Certainly, site reports from Thanet Earth indicate that farming and possibly stock
breeding was the predominant activity in the area, with little evidence of metal-, leather- or
wood-working. This would strongly suggest that any craft activity was to support the
household alone while the main economic activity was farming and/or animal husbandry. On
the other hand, excavations at Highstead demonstrated that pottery production and copper
smithing occurred as a craft activity, rather than to supply the settlement. It is most likely that
in Kent, small sites such as Thanet Earth and Highstead supplied themselves with textiles,
whereas larger sites with trading contacts (such as East Wear Bay) acted as the main
138
production centres of this craft, processing the raw material and weaving the required fabrics
on site.
Table 45. Measurements of spindle whorl clusters from enclosed settlements, the most likely thread thickness
they would produce, and the general speed of their spin.
County Min.
diameter
(mm)
Max
diameter
(mm)
Min weight
(g)
Max weight
(g)
Likely thread
diameter
(mm)
Likely
speed of
spin
Likely
warp
tension (g)
Cornwall
Bodrifty 33 41 23 26 0.4 - 0.7
Very fast -
fast 25 - 35
Trevisker 30 38 7 14 0.2 - 0.5 Very fast 10 - 25
Dorset
Gussage All
Saints 31 43 24 36 0.4 - 1
Very fast -
fast 25 - 40
Hampshire
Easton Lane 32 32 65 65 0.8 - 1.5
Fast -
medium 40 - 60
Kennel
Farm 42 42 3 3 0.2 Very fast 10
Kent
Highstead 40 40 45 52 0.6 - 1
Fast -
medium 30 - 40
Thanet
Earth 33 33 47 47 0.6 - 0.9 Fast 30 - 40
Despite the fewer numbers of tools, there is a significant data cluster in Trevisker; six out
of the ten spindle whorls in this settlement are between 6-10 g in weight, limiting thread
thickness to 0.2 - 0.5 mm. This corresponds with the pattern from the fortified Cornish sites -
that thread production is largely limited to producing very fine threads compared to other
areas. The other Cornish site, Bodrifty, has a wider and more equal spread of weights (see
table 45 above), producing fine to medium thickness threads. The Dorset site, Gussage All
Saints, has spindle whorls best suited to making a range of fine and thick threads, which
would allow them to produce fabrics for a variety of purposes.
Interestingly, enclosed sites have the narrowest grouping of diameters of spindle whorls
across all counties and settlement types, which suggests that spinners here required tools that
could spin at a slower speed for longer. The shape of spindle whorls at these sites is also
extremely limited, with 24 out of 30 spindle whorls being discoid; there is a wider range of
materials than shapes, with 12 made of stone, nine of fired clay, and nine from reused pottery.
This variation of raw materials can be explained more easily than why spinners in these
settlements opted for discoid whorls. The stone and clay used was readily available and
easily shaped, and the reused sherds occur predominantly in Late Iron Age settlements,
139
which follow a pattern of repurposing broken pottery as whorls more than earlier Iron Age
sites.
Loom weights
Figure 25. The weights and depths of loom weights from enclosed sites in all counties.
The number of loom weights from these sites are as follows:
Trevisker (Cornwall) - 13 Winnall Down (Hampshire) - 13
Gussage All Saints (Dorset) - 13 Zion Hill Farm (Hampshire) - 5
Easton Lane (Hampshire) - 9 Thanet Earth (Kent) - 5
Kennel Farm (Hampshire) - 8 Highstead (Kent) - 2
Again, the tools from Cornwall are the lightest and thinnest of all loom weights, and the
Hampshire tools cover the widest range of weights and depths (see table 46 below).
However, the loom weights from Dorset and Kent have tighter clusters than enclosed (and
other) sites from other counties. No loom weights were found in any of Devon’s enclosed
settlements; Devon overall produced some of the lowest numbers of spinning and weaving
tools in enclosed settlements.
140
Table 46. Measurements of loom weight clusters from enclosed sites compared to the thread diameters best
suited for weaving
County Min. depth
(mm)
Max depth
(mm)
Min weight
(g)
Max weight
(g)
Threads per loom weight for
following tension
10g 20g 30g 40g 60g 80g
Cornwall
Trevisker 1 1.8 9.8 6.1 36 0 - 3 0 - 1 N/A N/A N/A N/A
Trevisker 2 9.3 17.5 115.5 258
11 -
25
5 -
12 N/A N/A N/A N/A
Dorset
Gussage All
Saints 58 92.8 1673.3 2082.6 N/A
83 -
104
55 -
69
41 -
52 N/A N/A
Hampshire
Easton Lane 73 100 929 1196 N/A N/A N/A
23 -
29
15 -
19 N/A
Kennel Farm 64 76 3062 3183
306 -
318 N/A N/A N/A N/A N/A
Winnall
Down 1 63 90 1803 2061
180 -
206
90 -
103
60 -
68
45 -
51
30 -
34
22 -
25
Winnall
Down 2 82 103 2548 2909
254 -
290
127 -
142
84 -
96
63 -
72
42 -
48
34 -
36
Zion Hill
Farm 77 106 522 2760
52 -
276
26 -
138
17 -
92
13 -
69
8 -
46
6 -
34
Kent
Highstead 100 109 1400 1750 N/A N/A
46 -
58
35 -
43 N/A N/A
Thanet Earth 56 70 801.7 1177.5 N/A N/A
26 -
39
20 -
29 N/A N/A
In the Cornish sites, Trevisker loom weights 1 are too light to be able to weave any of the
threads produced on site, implying that these were not used for weaving in the same way that
the Rumps loom weights 1, and were used for another purpose. In comparison, Trevisker
loom weights 2 are excellently suited to weaving all the threads from the site. The number of
10 g warp threads combined with the depth of the loom weights indicates they were best
suited to producing a warp-faced fabric. On the other hand, the lower number of 20 g warp
threads per loom weight suggests production of a more tight, balanced weave.
In Dorset, Gussage All Saints loom weights are not suited for weaving most of the thread
produced on the site; this suggests either that the threads produced here were woven
elsewhere or that these weights were not for looms but for thatch weights or oven weights
instead. While these tools would be able to weave using thread that needs 40 g of tension, it
would be difficult to do so; the fabric produced could be extremely tight, possibly even
141
warp-faced, and may lack the correct tension due to the number of threads attached to each
weight.
In comparison, Hampshire’s Easton Lane loom weights are excellently suited for weaving
all the threads produced at the site; the number of threads per loom weight combined with the
thread thickness and loom weight depths indicates the production of fabrics ranging from
tight to open weave. However, the Kennel Farm loom weights are far too heavy for working
with any thread spun at the site, suggesting these were not used for weaving at all but
possibly as thatch weights instead. Because these artefacts are very unlikely to be loom
weights, these weights will not be discussed further. This matches the work of the Danebury
Environs Project, which suggested that small settlements in Hampshire specialised in
spinning thread that was then sent to the nearby hillfort for weaving. In Winnall Down, loom
weight Group 1 is best suited to weaving thick threads requiring 60-80 g of tension; given the
depth of these tools, the low number of threads required indicates that they are best suited to
weaving open fabrics. In comparison, Winnall Down loom weights 2 are deeper and would
struggle to work with very thick thread needing 80 g tension; if they could work with these
threads, they would also produce an open weave fabric, although it is likely these tools were
not meant for weaving at all. It is difficult to determine whether Zion Hill Farm produced a
specific type of fabric, since there were no clusters of loom weights like there were at the
other Hampshire settlements. This could indicate that weavers here produced a range of
fabrics for different purposes, although it is impossible to be certain. Regardless, the loom
weights from this site could work with threads requiring 20-80 g of tension, with the lighter
loom weights best suited to weaving threads needing 20-60 g of tension and the heavier loom
weights work well with the thicker threads needing 80 g of tension. These tools would best
produce tight, balanced fabrics and open weave respectively.
In Kent, Highstead loom weights can weave with threads needing 40 g of tension but not
with any other threads produced at the site. In comparison, the Thanet Earth loom weights are
excellently suited to weaving with all threads spun at the site. Regardless, both sites are able
to only produce open weave fabrics. Otherwise, Highstead only produced 2 loom weights,
making it impossible to note any clusters or patterns.
142
Open sites: Spindle whorls
Figure 26. The diameter and weight of spindle whorls from open sites.
There are few spindle whorls from open settlements:
Carn Euny (Cornwall) - 14 Allard’s Quarry (Dorset) - 9
Dainton (Devon) - 4 Sunset Caravan Park (Kent) - 5
Because each of the counties are only represented by one site with enough tools, it is only
possible to determine general patterns in each of the sites. Again, Cornwall has tight clusters
of the lightest whorls, whereas Dorset’s settlements have the widest range of spindle whorl
weights, able to produce very fine to very thick threads (see table 47 below). Excavations of
these open settlements demonstrate a wide variation of economies between the counties; little
evidence of craft production was found in the Cornish, Devon, and Dorset sites, with Dainton
lacking any obvious evidence of long-term occupation. In contrast, Kent’s Sunset Park had
large and long-term occupation, huge pottery production, and evidence of local and
Mediterranean trade.
143
Table 47. Measurements of spindle whorl clusters from open settlements, the most likely thread thickness they
would produce, and the general speed of their spin.
County Min.
diameter
(mm)
Max
diameter
(mm)
Min weight
(g)
Max weight
(g)
Likely thread
diameter
(mm)
Likely
speed of
spin
Likely
warp
tension (g)
Cornwall
Carn Euny
Cluster 1 27 46 5 9 0.2 - 0.4 Very fast 10 - 20
Carn Euny
Cluster 2 30 34 13 16 0.3 - 0.4 Very fast 15 - 20
Devon
Dainton 30 34 17 24 0.3 - 0.7
Very fast -
fast 15 - 35
Dorset
Allard’s
Quarry 25 43 10 25 0.3 - 0.7
Very fast -
fast 15 - 35
Kent
Sunset
Caravan
Park 21 39 24 30 0.6 - 1
Very fast -
medium 30 - 40
Cornwall remains the lightest spindle whorls tightly clustered in weight; over half (9 out of
14) are between 5 and 9 g. This light weight would allow for a very fast spin, although the
low diameter of these whorls would lose momentum quickly; characteristics that suit work
with animal fibres rather than plant fibres. In comparison, the majority of spindle whorls
from the other counties lie between 15 and 30 g, best suited to producing a range of fine and
medium thickness threads. However, the Dorset and Kent sites have the option of making
thick threads if necessary.
Loom weights
Only 4 loom weights were recorded from open sites in this study:
Dainton (Devon) - 2 Allard’s Quarry (Dorset) - 2
This is not enough to be able to draw any meaningful conclusions of weaving on open sites,
other than that weaving as an activity was probably minimal in these settlements.
144
Other sites: Spindle whorls
Kents Cavern is a unique site in this study; a seasonally occupied cave site that has no
evidence of permanent occupation or craft activity other than spinning and cooking. No
weaving implements have been found, suggesting that weaving activities were conducted
elsewhere. It is possible that this site was used to process raw wool after shearing (skeins of
spun thread are easier to transport than fleece and take up less space) and if so, this data
could determine what thicknesses of thread were required for wider distribution throughout
the Torbay region.
Figure 27. The diameter and weight of spindle whorls from other sites.
These data clusters can be distinguished by the difference in weight between the whorls,
rather than the difference in diameter. Given that these whorls were designed to spin at
similar speeds and lengths of time, it is most likely that the heavier whorls were not used for
different types of fibre (i.e.: plant fibres instead of wool), but for plying instead of spinning.
As the site itself lacks evidence of long-term occupation and other craft activities, I agree
with the proposal that Kents Cavern was a seasonal site used to process wool into spun
threads before further transportation to various settlements (Bradley 1978; Silvester 1986).
Why Kents Cavern was used to do so will be discussed in the Distribution chapter (see
below).
145
Table 48. Measurements of spindle whorl clusters from other settlements, the most likely thread thickness they
would produce, and the general speed of their spin.
County Min.
diameter
(mm)
Max
diameter
(mm)
Min weight
(g)
Max weight
(g)
Likely thread
diameter (mm)
Likely
speed of
spin
Likely
warp
tension (g)
Devon
Kents
Cavern
Group 1 25 39 14 22 0.4 - 0.7
Very fast
-fast 25 - 35
Kents
Cavern
Group 2 32 40 25 31 0.6 - 1
Very fast -
medium 30 - 40
Almost half of all spindle whorls at Kents Cavern make up Group cluster 1 (nine out of 20),
able to produce fine to medium thickness threads, and a quarter of the spindle whorls (five
out of 20) comprise Group 2, best suited to medium and thick threads. Only two whorls are
under 14 g, while the remaining four are over 32 g in weight, which corresponds to the data
from other Devon sites. Evidently, spinners at Kents Cavern wanted whorls able to spin fine
to medium thickness threads, with the option of producing thicker threads when required.
European textile evidence
As discussed in the textile tool analysis above, the thread diameters of spindle whorls across
the study area varies a great deal according to the materials these whorls are made from and
the needs of the site using these whorls. While experienced spinners could compensate for a
spindle’s weight and produce a range of thread diameters, the data collected in this study still
presents the threads most likely to be produced and allows us to determine the characteristics
that these textiles most likely had. In contrast to the second-hand evidence of textiles that is
available in Britain, the salt mines of Hallstatt and grave finds from across north-west Europe
provide direct data on the Iron Age textiles produced in this area. These allow us to compare
the evidence found here to what was produced in north-west Europe at this time.
The majority of Iron Age textiles from Hallstatt have very fine threads, with most
diameters ranging between 0.2 and 0.3 mm (roughly the same thickness as modern sewing
threads) and not exceeding 0.5 mm in diameter (Grömer 2012: 43). There are very few
examples of textiles with threads of 1 mm or more in diameter (Grömer 2012: 43),
demonstrating that the ability and knowledge to spin thick thread was present but was not
preferred. These fine threads resulted in high thread counts and high quality textiles; threads
146
measuring 0.1-0.2 mm in diameter have been found in textiles with 40 threads per cm,
although almost half of the Iron Age assemblage have a much lower 10-15 threads per cm
(Grömer 2012: 43). Similarly, the Iron Age textiles of northern Germany (c. 500-150 BC)
have numerous fine textiles with thin thread diameters used as grave goods (Banck-Burgess
1999: 34, fig. 6–8). Although the weave pattern itself is usually a plain tabby, the structure of
the fabric is extremely light and open, creating a veil-like effect with thread diameters of 0.2
mm. These fabrics were used as coverings or wraps for grave goods in cremation burials
across Lower Saxony (Schleswig- Holstein) and Mecklenburg (Western Pomerania)
(Banck-Burgess 1999: 34; Grömer 2016: 299). The widespread production and use of such
fabrics during the middle to late Iron Age demonstrates the high level of expertise of the
spinners and weavers, as well as the level of standardisation expected for this form of
burial-wrap across northern Germany. The textile tool assemblages of Germany are difficult
to fully construct, in part due to issues with the preservation of lowland settlements, and also
due to the absence of spindle whorls throughout the Bronze and Iron Ages (Kneisel and
Schaefer-Di Maida 2019). In comparison to the lack of spinning tools, there are a large
number of loom weights and weaving huts throughout Germany (see Kneisel and
Schaefer-Di Maida 2019: 94-96, table 4.2 and 4.3). It is possible that in Iron Age Germany,
spindles and spindle whorls were made from wood, bone, or other organic materials that are
more susceptible to deterioration in the archaeological record.
The textile and textile tool evidence from Iron Age sites around the Mediterranean shows
similar thread diameters and high thread counts to those found in north-west Europe (Gleba
2017; Marín-Aguilera 2019). The majority of textiles from Italian grave sites, dated to c.
600-400 BC, have thread diameters measuring 0.2-0.5 mm and are usually woven with 10-30
threads per cm (Gleba 2017: 1210). Generally, textiles in Italy have similar weave patterns
that have little difference between the way linen and wool textiles are woven, with dyes,
borders, and personal ornaments (e.g. fibula brooches) used to provide decoration. In
comparison, weave patterns across Iron Age Greece depend on the material used to make the
cloth; the majority of balanced tabbies were made with linen and where it is possible to
identify the material used in weft-faced tabbies, the majority were made of wool (Gleba
2017: 1215). In addition to the differences in weaving styles according to the type of fibre,
the thread counts of these woollen Greek textiles are far higher than those in Italy or
northwest Europe, although the thread diameters are similar to those in northwest Europe
(Gleba 2017: 1215). Woollen weft threads often measure only 0.1-0.2 mm in diameter with
147
very little spin on the threads, yet the majority of the weft-faced tabbies number 50 threads
per cm, with some reaching 100 threads per cm (see Gleba 2017: table 2, fig. 10). The
methods of decorating these textiles are also different from the contemporary Italian textiles,
either with patterns derived from weaving different thread diameters into the textile and
altering the twist or twist angle spun into the individual threads (Spantidaki & Moulhérat
2012:192, 194), or by embroidering, dyeing, or adding appliqués to the textile after weaving
(Gleba 2017: 1215; Spantidaki 2014).
The spindle whorls of La Mata and Cancho Roano, two Iron Age sites in Iberia, are not
made of reused pottery sherds but are made in molds for the same reason as pottery sherds
were used elsewhere: to create tools of set weights for a standardised spinning and thread
production (Marín-Aguilera 2019). Despite the similarities between the methods of
production and the use of the finished textiles for trade and exchange at both sites, the La
Mata spindle whorls have a broader range of weights than those at Cancho Roano. La Mata’s
whorls weigh between 4 and 53 g, producing threads and yarns of between 0.3 and 1.5-2 mm
in diameter (Marín-Aguilera 2019: 237). In contrast, the majority of Cancho Roano’s spindle
whorls weigh less than 15 g and measure no more than 15 mm in diameter (Marín-Aguilera
2019: 237). These whorls were best suited to spinning short woollen fibres into threads and
yarns of up to a maximum of 0.5 mm in diameter (Marín-Aguilera 2019: 237). Unfortunately,
no textiles from these sites survive, which means we are unable to know the common thread
count, weaving patterns, or methods of decoration. Therefore, these represent the most likely
threads that were made by these tools.
Comparing the data collected from sites in the South coast study area to the evidence from
Iron Age sites across Europe highlights the number of similarities between textile production
in these areas. Many of the British sites analysed here produce threads of the same or similar
thicknesses as those found in Hallstatt, Italy, La Mata, and Cancho Roano. However, there is
evidence that these British sites required a broader range of thread thicknesses than these
European sites. This could be reflective of regional specialisation in textile production; for
example, the Mediterranean sites in Italy, Spain, and Greece may have specialised in
producing thin threads in answer to the hotter environments in these areas. Some regions,
such as northern Germany, definitely had specialist production of particularly fine textiles,
such as the open weave fabrics found in burials. Unfortunately, the lack of spindle whorls and
textiles from other contexts means we are unable to determine whether thicker threads were
148
also commonly produced or what the average textile used in daily life looked like and how
such textiles were made. This underlines the importance of textile tools in such studies and
analyses; without the tools themselves, we are forced to rely on evidence that rarely survives
outside of specific environments and contexts. While there is a lot of information we are able
to extract from textiles, this lack of evidence also limits our understanding of the variety of
textiles produced and used in these areas even more than textile tools would. Unlike textile
remains, textile tools are able to provide more in-depth information about the modes of
textile production across regions, the importance of textiles in trade and a site’s economy, and
the comparative importance of textile production to other crafts conducted at the same
settlement; issues and points that are discussed in detail in the following section.
149
Tool distribution
South West
By mapping out the distribution of tools across the various settlement types across the south
coast, it is possible to get a clearer picture of the likely organisation of textile production and
producers in each region. However, it is also important to remember that these maps only
represent the complete tools from each site; it is possible that more tools were excavated but
were not complete enough to include in the study. So these maps must be taken as a guide
rather than as definitive proof of organisation.
Figure 28. Map of tool types found across all site types in the South West. 1.: Carn Euny. 2.: Bodrifty.
3.: Trevelgue Head. 4.: Trevisker. 5.: The Rumps. 6.: Mount Batten. 7.: Dainton. 8.: Kent’s Cavern. 9.: Milber
Down. 10.: Blackbury Castle. Author’s illustration.
The distribution of the Cornish site’s textile tools indicates an interesting division between
the spinning and weaving stages of textile production. It appears that this divide focuses more
on the location of the settlements themselves than the type of settlement (see figure 28
150
above). That is, weavers were not restricted to one type of site (i.e.: fortified sites) while
spinning occurred at all others; instead, it appears that spinning occurred on all sites, but
weaving was limited to sites that were either trading centres or close to such centres. This
pattern is similar to those of the Danebury Environs Project, where weaving tools were
clustered in the hillfort while spindle whorls were more common in the smaller settlements
surrounding it (Cunliffe and Poole 1989, 1991). Given that this separation of spinning and
weaving exists, the next step should be to investigate the evidence of further specialisation in
the textiles produced in this county.
Analysis of Cornish textile tools has already shown that Cornish fabrics were consistently
made with very fine threads compared to those from other regions; dyes and embellishments
could have been used to make the finished textiles stand out even more and create a unique
textile tradition. Unfortunately, the acidic nature of the Cornish soil makes it highly unlikely
that enough evidence of dye residues on ceramics or textile fragments with embellishments
would survive to allow any meaningful research into dyeing and embellishing techniques
and/or traditions. On the other hand, isotopic analysis on select textile remains in Europe
have already demonstrated that textiles were traded across large distances (see Chapter 2,
above). A large-scale study could focus on using this method of analysis on as many Bronze
and Iron Age textiles as possible, identifying areas of raw material (and possible textile)
production that could indicate the bare bones of prehistoric textile trade. If Iron Age
Cornwall was producing textiles for trade, it is possible that some of these textiles went to
Europe via the continental trade links already established in this county.
The pattern of textile tool distribution in Devon generally follows that seen in Cornwall,
with weaving tools clustered according to geographical areas rather than settlement type.
This is especially clear in the Torbay region (see fig. 28, nos. 7-9), although more work could
be done in the north of the county to track down sites with sufficient numbers of textile tools
in order to determine if this pattern is repeated around Devon. Both counties demonstrate
weaving cultures that have areas with specialised weavers close to trade hubs that could
produce unusually fine yarns and fabrics. In Cornwall, this is in part due to the prevalence of
slate in the region, which allows easy production of tools with set thicknesses and weights.
While slate is rarer in Devon, Kimmeridge shale is a common material for spindle whorls in
the Torbay region. This imported material was usually used for personal ornaments like
lathe-turned bracelets, although the cores of such bracelets are easily adapted into spinning
tools. The frequent use of Kimmeridge shale spindle whorls in such a restricted area heavily
151
suggests that these tools were more valued than whorls made from other materials. This is
further evidenced by the decorations that cover most of the Torbay spindle whorls. Silvester
(1986) noted that the Torbay region appears to have a significantly different Iron Age textile
culture compared to elsewhere in the country, due to the extensive decorating of spindle
whorls in this area. It is especially notable due to the fact that whorls are usually considered
by archaeologists to be a common, utilitarian tool, easily made and easily discarded.
Tools can be decorated for several reasons: to distinguish one person’s tool from anothers’,
to draw the eye when the tool is used in public, to identify a skilled craft worker, and to
embellish an unusual material used in the tool. Torbay’s spindle whorls are predominantly
made from Kimmeridge shale (a material imported from Dorset) and have a greater level of
standardisation in weight and dimensions than spindle whorls of clay, partly because these
are by-products of carving bracelets of similar sizes. So these whorls provided Torbay
spinners with the tools to produce spun threads with highly standardised diameters, already
allowing this region to stand out from areas without these tools. Furthermore, the discovery
of a great many decorated spindle whorls in areas like Kent’s Cavern, an open area without
private spaces, suggests that spinning was a very public activity for workers in this region.
This would suggest that spinning and spinners were valued in this society and that
decorations on these spindles were a method of displaying the craft worker’s status and their
affiliation with the specialist production in the Torbay region. Given the poor preservation
conditions that dog the other Devon sites, I would argue that the Torbay region is an
important example of the pattern seen in Cornwall continuing across this county as well,
where the stages of textile production are split between geographical areas.
Other sites in Devon did not produce similar evidence for spinning specialisation. Milber
Down (fig. 28, no. 9) and Blackbury Castle (fig. 28, no. 10) produced little evidence for other
craft activities, as well as no foreign or traded artefacts. While this suggests that the
occupants here were self-sufficient to some degree, textile tools easily outnumber the finds
from other crafts and point to a higher level of production than perhaps necessary for small
sites. On the other hand, the excavations at Kent’s Cavern and Dainton produced very little
evidence for occupation during the Iron Age, indicating that these were not permanent
settlements. Silvester (1986) has already suggested that Kent’s Cavern was a seasonally-used
site primarily for spinning fibres prior to transportation, which would have been easier than
moving raw wool or plants for preparation. The evidence from these sites strongly suggests
152
that Devon followed the pattern seen in Cornwall, where raw material preparation and
spinning were conducted in sites separate from weaving and finishing.
Analysis of the tools from Devon sites included in this study have demonstrated that the
textile producers were making a broader range of fabrics than those in Cornwall, although
these tools are still more suited to producing fine yarns and fabrics than those across Dorset,
Hampshire, and Kent. This suggests a mixture of Cornish traditions with techniques from
other counties; such a combination of styles would be reasonable, considering that Devon has
acted (physically and historically) as a buffer between the South West peninsula and the rest
of the country. Furthermore, it is interesting to note the other evidence of occupation
alongside the textile tools excavated at the Devon sites, particularly in the Torbay region. The
promontory fort of Mount Batten produced a large number of high status finds, such as
personal ornaments (brooches and bracelets) and a large number of coins (British and
continental), but almost no evidence of metal working, salt or pottery production, or textile
work. Therefore, I would argue that for this settlement, other forms of economic activity
were more important. Long-distance trade between Cornwall and the continent and
metal-working appear to be the primary concerns at Mount Batten, with any crafts merely
there to sustain the community rather than to provide goods for trade.
153
Central South
Figure 29. Map of tool types found across all site types in the Central South. 1.: Hengistbury Head.
2.: Allard’s Quarry. 3.: Hod Hill. 4.: Gussage All Saints. 5.: Maiden Castle. 6.: Danebury. 7.: Zions Hill Farm.
8.: Easton Lane. 9.: Winnall Down. 10.: Kennel Farm. Author’s illustration
In contrast to the South West region, all sites within Dorset have spinning and weaving tools.
This suggests that each site had some form of weaving, no matter how limited in scope, to
produce the textiles that they needed at each settlement. On the other hand, the larger ratio of
loom weights to spindle whorls in Gussage All Saints stands out from the other Dorset sites
and suggests a higher rate of production at this site compared to the others, possibly due to
the presence of a specialist weaver.
Of the fortified sites, Hod Hill (fig. 29, no. 3) and Maiden Castle (fig. 29, no. 5) have two
notably distinct groups of spindle whorls each, as discussed previously. The groups of Hod
Hill are best suited to producing very fine and fine to medium yarns respectively, while
154
Maiden Castle’s spindle whorls are able to spin fine to medium and medium to thick yarns.
The limited range of yarn diameters at Hod Hill suggests that textile production was more
focused at this site, either because spinners had limited materials to work with or because this
site produced more specialised fine or open weave textiles. In comparison, Maiden Castle has
a wider variety of whorls that are mostly confined to several groups of tools, each with its
own limited range of weights and dimensions. This both allows for the production of a wider
range of yarn diameters and ensures larger scale production of these textiles compared to the
smaller sites in the region. It is possible that Maiden Castle specialised in a
quantity-over-quality approach compared to Hod Hill; having tools of similar weights and
dimensions would make spinning large quantities of similar yarns quick and easy, with
standardised groups of lighter and heavier whorls allowing more people to produce a wider
variety of yarn diameters regardless of the spinner’s level of skill. The loom weights of Hod
Hill and Maiden Castle also reflect the range of fabrics woven from the yarns produced on
their respective sites. Hod Hill’s weights are best suited to a limited range of yarn tensions
compared to Maiden Castle’s wider range of weights, which could weave a broader variety of
textiles.
Interestingly, Hengistbury Head (fig. 29, no. 1) did not demonstrate any clear groups in
tools; both spinning and weaving tools from this site appear to favour production of a wide
range of textiles. This could be explained in several ways:
● Higher level of skill among Hengistbury’s spinners, as experienced spinners would
be more able to compensate for the tools’ weight while producing a fine yarn
● Wider range of fibres used in spinning and weaving that required a greater range of
tool weights, such new fibres could be introduced by the various traders passing
through Hengistbury Head (Cunliffe et al. 1987)
● Local preference or spinning traditions, influenced by the range of cultures and
traditions that passed through such a main trading post (as evidenced by the range of
foreign artefacts excavated from the site)
Given Hengistbury’s position as an easily accessible safe harbour on the main South Coast -
Continental trading route, I believe the second point has most merit, although this could only
be proven by the discovery and examination of actual textiles from the site. Regardless of the
variations in Hengistbury’s textile tools, excavations at each site demonstrate that these large
fortified sites were centres of a number of craft activity, including salt production, bronze and
155
iron smelting and working, pottery firing, and shale and bone carving (Cunliffe et al. 1987).
So the presence of large-scale textile work aimed at producing surplus, tradable goods should
not be surprising but viewed as simply another facet of the range of goods produced at these
trading hubs.
In comparison, Dorset’s smaller settlements demonstrate other textile production practices.
Allard’s Quarry (fig. 29, no. 2) has the smallest number of textile tools and presents a good
example of a site spinning and weaving to meet its own needs. Excavations here produced
few personal ornaments, no evidence of metalworking, only pottery of local make, and no
other craft debris (i.e.: shale cores from lathe-turning) (Williams 1951). This was a site that
relied on agriculture and limited itself to local trade for whatever it could not produce itself.
Given the presence of three large production centres in the region and it would not be
unreasonable to expect Gussage All Saints (fig. 29, no. 4), the other small site of this study, to
follow this pattern. Instead, Gussage All Saints wildly deviates from what could be expected
of it. For one, it is the only site in this region where loom weights outnumber spindle whorls.
Furthermore, the largest number of spindle whorls appear in the late Iron Age settlement
phase, suggesting a sharp increase in textile activity during this phase (Wainright 1979: 185).
This increase coincides with a decrease in metalworking (which peaked in the mid Iron Age)
and could indicate that textiles took the place of metalworking as Gussage All Saints’ main
tradable craft  (Wainright 1979: 185).
This is supported by the animal bone evidence from the site, which showed sheep were the
most common animal at the site during the late Iron Age and their herd structure was best
suited to wool and/or milk production, rather than meat (Wainright 1979: 186-7). Wainright
(1979: 190-1) also notes other evidence (pottery, personal ornaments, etc.) clearly
demonstrated that Gussage was actively participating in Wessex’s ‘specialised trade patterns’,
connecting this small site to the larger craft and trade centres like Maiden Castle. Although
Gussage All Saints appear to be a consumer of pottery and other artefacts, the mid- and late
Iron Age evidence certainly points to it participating in textile and metal trade. The
examination of textile tools in this study demonstrates that the spinners and weavers of
Gussage All Saints could produced a range of yarns and fabrics; this would allow the
craftspeople here to meet almost any requirements, rather than forcing them to specialise in
the finer fabrics of the South West. The heavier weights of these tools would also suit
spinning and weaving wool, which the faunal evidence has already established was produced
at this site. So Gussage All Saints provides an excellent example of a site that was producing
156
its own raw material, spinning and weaving it on site, and trading these finished fabrics on
rather than following the South West model of specialising in one phase of textile production
and passing that material to another specialist for further work.
The example of Gussage All Saints raises the interesting and important question of why
there is such a difference in the organisation of textile production. The fortified sites of the
South West (particularly Cornwall) appear to have more direct control over the means of
production compared to Wessex; for example, where the raw materials were produced, where
stages of textile production occurred, and by extension, control over where the producers
work and possibly what they produce. However, I would argue this difference can largely be
explained by the geography of these regions, as demonstrated in Appendices 10-14. Thanks
to its expansive chalkland, Wessex has a great deal more land suitable for agriculture than the
South West, thereby allowing livestock to graze on lowland, higher-quality pastures and
enabling anyone in these areas to own sheep (see Appendices 12-13). With sheep being so
widespread, spinners and weavers would be able to work in any settlement. By comparison,
both Devon and Cornwall have expansive moorland and other areas of poor soil, limiting
where the best arable crops can be grown and confining livestock to these rough grazing
areas (see Appendices 10-11). With the raw material confined to specific areas, it is more
efficient for spinners to work in these places and for the finished yarns to be transported
elsewhere for weaving.
Compared to the other counties, Hampshire is characterised by an abundance of loom
weights throughout the various settlements included in this study; this is the only area in
which there are sites with loom weights and no other textile tools. The presence of such
weaving tools over so many different sites indicates that each settlement produced textiles for
their own use, instead of restricting weaving to a small area and distributing the finished
fabrics to those outside. It also suggests that weaving was a significant craft activity among
the people of this region, with weavers in each settlement to produce more textiles than other
regions. It is important to remember that this does not mean that the fabrics were of a higher
quality, simply that more would be made overall. Despite the predominance of weaving in
this county, the small number of spindle whorls and other textile tools in this region
compared to the others is puzzling; where were the spinners? How was the raw material
processed for all the weavers? The yarns could be imported but Hampshire has plenty of
space for wool or flax sp could easily produce its own materials. It seems more likely that the
157
tools themselves did not survive, possibly because they were made from organic materials
(wood, bone, etc.) rather than the more standard clay, stone, and ceramics. The weight of
such organic spindle whorls would depend on the density of the materials and could be well
suited to making fine or medium diameter threads for use on warp-weighted looms or other
loom setups. If such whorls were used, this preference would reflect a distinct tradition
among the spinners, similar to the decorated spindle whorls of the Torbay region.
Unfortunately, organic material does not survive well across much of Hampshire; until such
tools are excavated, this study must focus on the materials we have rather than those that are
absent.
In Winnall Down and Zionhill Farm, spindle whorls are completely absent from the
archaeological record, despite evidence of other industries being present. The majority of
loom weights found at Winnall Down came from two pits in the site’s northwest area, close
to two 4-post structures that the excavators believed either housed looms or were the remains
of a pair of upright looms (Fasham 1985: 127).  It should also be noted here that 22 loom
weights in total were excavated from these pits, while only 13 were preserved enough to
include in this study; nonetheless, this is more than needed for low-level self-sufficient
production. The rest of the site has clear zones of activities; the southeast area containing
crucible and pottery fragments, the southwest containing a number of 4-post structures and
oven remains that indicate crop processing and storage, and the northeast area probably used
to quarry chalk to carve into loom weights (Fasham 1985: 128-9).  This production of
weaving tools at Winnall Down combined with the large clusters of loom weights, as well as
the restricted area these tools were found in, suggests specialised production of fabric. As
shown elsewhere, the loom weights themselves were only capable of weaving very thick
yarns or threads that required heavy tension, which supports the theory of specialised textile
production but not the possibility of organic spindle whorls. This evidence demonstrates that
spinning was more likely to have been conducted elsewhere and the finished threads brought
in for weaving.
There is little evidence for other craft activity at Zions Hill Farm during the late Iron Age,
aside from one clay oven bar in a single context feature that suggests some pottery
production during this period (Adam et al. 1996: 5). Otherwise, metalworking evidence (such
as slag) is confined to the early Bronze Age period (Adam et al. 1996: 9). Despite this, ten
loom weights were discovered at the site alongside other small, amorphous baked clay
158
fragments (which are not specified as being possible pieces of loom weights) (Adam et al.
1996: 10). Of these weights, six came from the same pit and were of the typical Iron Age
triangular shape; the other four were from another pit and cylindrical in shape, a form more
commonly associated with the late Bronze Age/early Iron Age rather than the far later
contexts (Adam et al. 1996: 10; Rawlings et al. 2003: 5). In discussing the site, Rawlings
argues that the extent and variety of artefacts (in both the late Bronze Age/early Iron Age and
the late Iron Age) indicates the presence of a farmstead or small settlement nearby, although
it is unclear where (2003: 18). Analysis of the tools themselves demonstrated that these loom
weights are best suited to weaving a variety of fabrics, using fine threads with high thread
counts to thick threads with low thread counts (see Chapter 4).
Easton Lane is an unusual site composed of pits and ditches close to the Winnall Down
settlement; Fasham believed this was a spillover of activity from the main enclosure, since no
roundhouse remains were discovered at this site (Fasham et al. 1985). Certainly, the finds
from this area are similar to those of Winnall Down, although there is little to no evidence at
Easton Lane of activities being restricted to certain areas. Despite the similarities and
closeness of these two sites, the loom weights at Easton Lane are roughly only half the
weight of those at Winnall Down (see Chapter 4). This would allow the weavers at Easton
Lane to weave threads of medium thickness and lighter fabrics, thereby expanding the range
of fabrics possibly produced by the people of Winnall Down. That being said, the fact that
these lighter loom weights were deposited in pits outside the enclosure may indicate that
these tools were used for different styles of weaving (e.g. open weave instead of a tight
weave) or by different weavers. Unfortunately, this is purely speculative, unless further work
on loom weight use-wear can accurately identify differences between weaving styles.
Danebury’s textile tools have already been reviewed as part of Cunliffe’s extensive
excavations (Cunliffe and Poole 1984; 1991). It demonstrated that the number and
distribution of textile tools in the early Iron Age was very different to those of the middle
Iron Age, corresponding to a change in metal working practices (Cunliffe and Poole 1984:
436-439; Cunliffe and Poole 1991: 415-416). During the early period of the site, textile tools
were distributed evenly but thinly across the site and number:
● 62 clay loom weights
● 144 chalk loom weights
● 18 clay spindle whorls
● 53 chalk spindle whorls
● 39 bone/antler weaving combs
159
Clusters of tools are uncommon, although all appear in pits (Cunliffe and Poole 1984: 438).
In comparison, the tools from the later periods of the site number as follows (Cunliffe and
Poole 1991: 354-402):
● 106 clay loom weights
● 32 chalk loom weights
● 15 clay spindle whorls
● 41 chalk spindle whorls
● 33 weaving combs
These tools are mostly confined to the southern section of the site, with some excavated
from a round house and the rest in pits, and with clusters of tools frequently deposited. These
clusters are never of one tool type but a mixture of spinning, weaving, and finishing tools of
various materials (Cunliffe and Poole 1991: 354-402), possibly representing an individual’s
old or unwanted tools rather than lost or broken tools that needed replacing. At the same
time, early metal working at Danebury is characterised by predominantly bronze smelting,
although there is no evidence for large-scale, extensive metal working (Cunliffe and Poole
1984: 436-437). During the later phases of the site, this changes to iron working without any
evidence of smelting, also widespread across the site (Cunliffe and Poole 1984: 436-437).
Given this evidence, the change in Danebury’s textile production could be linked to a
change in textile technology or practices. Such a change is echoed in the massive shift from
chalk loom weights in the early Iron Age (which typically have one perforation and a long,
sub-rectangular shape) to the triangular clay loom weights of the mid to late Iron Age (which
typically have three perforations and are more standardised in their size and shape). The
metric data of these weights gathered by this study demonstrates that the differences between
chalk and clay loom weights also extend to the types of fabric they were best suited to
weaving (see fig. 30 below). The chalk weights are concentrated in a cluster of weights
approximately between 900 to 1400 g and with depths roughly between 50 to 140 mm. On
the other hand, the majority of later clay weights make up two close clusters that cover a
larger range of weights but are more restricted in their depths. The first cluster has weights
between 1000 to 1400 g and depths of 55 to 80 mm; the second weighs between 1600 to
2000 g and measures 60 to 90 mm in depth. The triangular shape of the clay loom weights
allowed for better standardisation than carved chalk could provide, which also resulted in
more control over the range of weights available to the weavers. In the early Iron Age,
Danebury’s weavers prioritised having tools of similar weight over tools of similar
dimensions, since the chalk they used for their tools differed so much in density. However,
160
the change in the late Iron Age to triangular loom weights (with set, easily copied dimensions
and material) allowed the weavers to use heavier weights with less risk of their loom weights
pulling the fabric out of shape because of the weight’s uneven depths. This enabled these
craftspeople to weave a wider range of fabrics than before (see Chapter 2 for details).
Figure 30. Danebury chalk and baked clay loom weights. Author’s illustration.
Furthermore, the concentration of textile tools in the southern area of Danebury, compared
to the consistent presence of metal working across the site, demonstrates the growth of more
standardised textile production capable of spinning and weaving a range of materials for
further trade or distribution (Cunliffe and Poole 1984: 438-439). In the types of textile
produced, the organisation of production, and the ratio of tool types (including those omitted
from this study due to poor preservation), Danebury follows the pattern found at Maiden
Castle and Hengistbury Head in Dorset (Cunliffe and Poole 1984: 438-439). This pattern, of
a fortified site producing large quantities of textiles with tools able to make a range of yarns
and fabrics to fit different requirements, remains distinct from that found in the Cornish
peninsula (discussed above).
As well as their work at Danebury, Cunliffe and Poole also reviewed the distribution of
textile tools in various settlements, both fortified and unfortified, around the hillfort as part of
the Danebury Environs Program (2000a). This work was able to demonstrate that the
161
surrounding settlements followed the same low-level textile production present in Dorset
during this period, where each site produced and processed the raw materials into finished
fabrics, with few textile tools or evidence of other craft activities (Cunliffe and Poole
2000b-g). These sites rarely show specialisation in their tools, preferring those capable of
spinning and weaving a range of yarn and fabric thicknesses and producing little to no
evidence of dedicated textile craftspeople, such as a specific area for textile work. As noted
in the Dorset region, the abundance of good-quality arable land across Hampshire allows for
a mix of arable and pastoral activity throughout the region, which is reflected in the faunal
remains from each settlement. Rather than forcing livestock into upland areas (and thus
limiting access to some of the raw material required for spinning), sheep are present at all
sites included in this study (see Appendix 12) and the Danebury environs program (Cunliffe
and Poole 2000b-g). This easy access to wool helps explain the widespread nature of
spinning and weaving throughout Dorset.
South East
The evidence from the South East is unfortunately very limited as most excavations in this
area are rescue excavations done before building work. This frequently results in small sites
with reports that are difficult to access, various methods of reporting and excavation
(depending on the archaeological unit responsible), and trouble in tracking down finds for
analysis. Despite these challenges, five sites were reviewed in this study; given the limited
extent of the excavations, more material may come to light later that could alter our
understanding of their textile practices. For now, the most recent information from these
areas are presented here.
162
Figure 31. Map of tool types found across all site types in the South East. 1.: East Wear Bay. 2.: Sunset Caravan
Park. 3.: Thanet Earth. 4.: Highstead. Author’s illustration
The site of East Wear Bay (fig. 31, no. 1) has unfortunately been severely damaged over the
years from the erosion of the cliffs on which it was built. In the last decade or so, rescue
excavations revealed what may have been an oppida, a settlement type known in Europe as
major centres of power, trade, and crafting. Excavations of this Iron Age site were hampered
by the modern town of Folkestone, which lies above the Iron Age remains and forced
excavators to work in small trenches in the most vulnerable areas. Considering the evidence
of high levels of spinning and weaving in other such centres (e.g. Danebury, Maiden Castle,
The Rumps, etc.), it is surprising that so few weaving tools were found in this settlement,
especially considering the large number of spindle whorls excavated from East Wear Bay.
The excavation of Sunset Caravan Park (fig. 31, no. 2) at Whitstable only covered a small
area but it revealed an extensive, long lasting settlement with multiple large buildings and
several craft activities. Evidence of clay and flint quarrying, stone working, and pottery
making points to large-scale production (Allen 1999: 5); both the settlement itself and the
scale of these craft activities expanded in the late Iron Age (Allen 1999: 5-6). Alongside this,
finds demonstrate long-distance trade links with Europe, including late Bronze/early Iron
163
Age Italy, and along the rest of the southern coast of Britain (Allen 1999: 5). Despite this,
few textile tools and no weaving tools were recovered at this site. It is possible that this
settlement primarily relied on its status as a centre of stone and pottery production, as well as
a source of European trade, and that other crafts necessary for those living there were
conducted on a small, domestic scale. This would correspond to the evidence from Mount
Batten (see above), yet varies quite a bit from the evidence in Dorset and Hampshire.
Keeping in mind the similarities in the landscapes of Kent and Wessex, I would argue that
Sunset Caravan Park is only part of a wider settlement, with areas dedicated to various craft
activities. The section excavated so far certainly reflects this - there is little indication of the
agricultural activities necessary for sustaining such a large settlement and the main activities
of this area are dedicated to only two crafts (flint working and pottery production). Further
excavations may uncover more textile tools and, possibly, an area dedicated to textile
production. For now, the evidence available to us demonstrates textile production focused on
spinning medium to thick threads, best suited for heavier tensions on a loom that would
produce thick fabrics.
The excavation of Thanet Earth revealed a large settlement during the middle Iron Age (c.
500-300 BC) with evidence of other forms of craft production throughout. Soil samples
showed that the population supported itself by farming various cereals and legumes that were
stored in around 300 pits, as well as sheep, cattle, and pig bones (Rady 2009). Large
quantities of non-ferrous slag demonstrated large craft production of metalworking. The
excavators believed that this site was a local centre for the surrounding settlements, where the
smaller communities could come to trade and mingle (Rady 2009: 8; 2010). Although only
five loom weights and one spindle whorl could be accessed for measuring, 31 complete or
near-complete triangular loom weights and 5 spindle whorls were excavated from the main
Iron Age occupation. The high number of tools at this sort of local centre indicates textile
production focused on creating surplus materials to trade or exchange onwards, not providing
for individual households. Unfortunately, the limited number that were available for study
restricts the information that can be gathered on the types of textile that could be woven at
Thanet Earth.
Occupation at Highstead began during the early Iron Age, approximately 800 BC, as a
series of small enclosed farmsteads. It was not until the middle Iron Age (c. 500-300 BC)
when the settlement expanded into an open site with five roundhouses clustered close
164
together; this was the phase that produced the textile tools included in this study.
Interestingly, two rectangular huts dating to this period were excavated at Highstead; these
structures are more common in northwest Europe and few are found in Britain, so the use of
these buildings are unclear (Bennett et al. 2007). These may have been used as community
storage, guest housing, or houses built by immigrants from the continent; however, no
foreign pottery forms or finds were discovered from the site, which does not support the
latter theory. The other structures of the middle Iron Age occupation of Highstead included
clay-lined pits that were used for food storage and several 4-post structures that may have
been granaries, indicating agricultural production and storage to support the population.
Other finds included iron-working and salt production, which appear to have provided
Highstead with its main tradable goods. Few textile tools were recovered from Highstead,
indicating that textile production was not one of the crafts that helped to support the
community, but instead was focused on providing for the needs of the various households.
Reused pottery spindle whorls
One possible development in late Iron Age standardisation in textiles is the level of
standardisation that developed in textile tools, particularly spindle whorls and loom weights.
Late Iron Age spindle whorls in  north-west Europe are mostly made from pottery shards that
have been reshaped into discs and perforated with the hole for the spindle. By using this
material to create spindle whorls, the level of variation between the weights of each whorl is
significantly reduced (Grömer 2012). Creating a tool assemblage of similar weights and sizes
allows spinners with less experience to produce more uniform threads and yarns suitable for
weaving faster than if they had to learn on a variety of spindle weights (Mårtensson et al.
2006b). Furthermore, these tools can be produced in large quantities quickly and easily,
allowing tools to be replaced faster and new spinners to be provided with tools more
efficiently (Barber 1991; Mårtensson et al. 2006b). The numbers of reused shard spindle
whorls greatly increased in the late Iron Age across north-west Europe, which is a clear
indication of the change in textile practices during this period (Belanová-Štolcová and
Grömer 2010; Grömer 2012; Marín-Aguilera 2019). Interestingly, the shard whorls of the La
Tène period (c. 450-0 BC) Hallstatt salt mines are usually heavier than those from the
previous Hallstatt period, often weighing between 10-30 g (Grömer 2012: 54). This could be
a regional preference of the Hallstatt textile workers that allowed them to produce specific
165
forms of textiles more easily, or it could indicate a wider trend in the tools as part of the
standardisation of textile production.
Figure 32. A map of the geography of Cornwall, with sites highlighted. After Scrivener et al. 1997.
In Cornwall, both sites on the Penwith peninsula (Carn Euny and Bodrifty) have a
significant proportion of reused pottery spindle whorls. Half of Carn Euny’s spindle whorls
(seven out of 14) are reused pottery, with only one weighing under 10 g; the remaining tools
are made from local stone, with only one measuring over 30 g. Almost all of Bodrifty’s
whorls (four out of six) are made from reused sherds, with only one of these weighing over
30 g; the remaining spindle whorls are stone, with only one measuring under 10 g. In
comparison, the majority of spinning tools at the other Cornish sites are made from local
166
stone. The Rumps has ten stone spindle whorls, one bone, and four of reused pottery; all of
the pottery whorls measure between 10 and 30 g, while three stone whorls and the bone
whorl weigh under 10 g, and three other stone whorl weigh over 30 g. At Trevisker, only two
of the ten spindle whorls are made from reused sherds, both weighing within 10-30 g, with
the other eight made from local stone, including one measuring under 10 g and two over 30
g. Finally, Trevelgue Head produced three reused sherd spindle whorls out of 22, with the
other 19 made from local stone. While the reused sherd spindle whorls weighed between 10
and 30 g, two of the stone whorls measured under 10 g and two over 30 g.
It is telling that the majority of reused pottery spindle whorls are concentrated in the two
Cornish sites that specialise in spinning; this suggests that this material was adopted to allow
spinners to more easily produce threads of similar, set thicknesses. However, it should also be
noted that the majority of stone spindle whorls at other Cornish sites still weigh between 10
and 30 g, even though these sites do not have as many reused sherd spindle whorls. The
difference in the assemblage’s material may be explained by the geography of the region (see
fig. 32). The Rumps, Trevelgue Head, and Trevisker are located on deposits of Devonian
shale and sandstone that can be easily quarried, split into regular thicknesses, and carved to
shape. Therefore, these sites would not need to rely on pottery as a material for spindle
whorls, since the Devonian stone has the same qualities that make the pottery so useful
elsewhere. In comparison, Carn Euny and Bodrifty are located on granite deposits that form
the entire Penwith Peninsula; without the Devonian shale and sandstone present in the rest of
Cornwall, these sites had to rely on pottery to easily make spindle whorls of similar thickness
and weights.
167
Figure 33. A map of the geography of Devon, with sites highlighted. After Devon County Council 2020
In contrast to the assemblages in Cornwall, Devon did not have any spindle whorls of
reused pottery at all. Every site in Devon used local clay or stone with only the Kent’s
Cavern and Milber Down assemblages deviating from this pattern by the inclusion of a
number of Kimmeridge shale spindle whorls. None of the Devon sites had spindle whorls
that weighed less than 10 g or more than 30g, so it is possible that there was no need for these
sites to use pottery to produce tools of the correct weight. Each of the settlements included in
this study are located on or near limestone and sandstone beds, which would provide these
sites with suitably dense material that could be locally quarried and easily carved into spindle
whorls (see fig. 33).
168
Figure 34. A map of the geography of Dorset, with sites highlighted. After Thomas 2008.
The majority of sites in Dorset have very few spindle whorls of reused pottery, in part
because these settlements have access to chalk and limestone that is easily quarried and
carved (see fig. 34). Gussage All Saints has two out of nine spindle whorls of pottery, the
others made from local stone and clay; Hod Hill has three out of 29 spindle whorls, two of
which are under 10 g in weight; and Allard’s Quarry and Maiden Castle have no reused
pottery spindle whorls at all. Of the seven spindle whorls that constitute the Allard’s Quarry
assemblage, five are made from local limestone and three of these are over 30 g in weight;
additionally, there is a single clay spindle whorl that is also over 30 g. Maiden Castle’s
assemblage totals 58 spindle whorls, including 49 chalk whorls, of which two are under 10 g
and 31 are over 30 g, and nine stone or clay whorls, of which three are over 30 g. Chalk is
readily available in the local area and easy to carve into shape, which could explain its
169
popularity; furthermore, there is very little variation in the dimensions and weights of these
whorls. In other words, using the chalk still allowed the tool makers of Maiden Castle to
create spindle whorls to set sizes and weights including larger spindles of over 30 g for their
work where chalk was the most suitable material for tools of this weight.
In comparison, Hengistbury Head has the largest number of reused pottery spindle whorls
of all the Dorset sites. 39 of the 40 spindle whorls in its textile tool assemblage are made
from old pottery sherds, with only five under 10 g and five over 30 g in weight. As discussed
elsewhere, Hengistbury Head became a major centre of trade and craft production during the
middle to late Iron Age, which follows the pattern of increased pottery sherd usage found in
Europe. Furthermore, this site had no local access to the chalk or limestone that the other
sites in this region use to create spindle whorls (see fig. 34) so pottery sherds would have
provided the most reliable method of creating more standardised tools.
Figure 35. A map of the geography of Hampshire, with sites highlighted. After Hampshire County Council
2020.
170
There are only two spindle whorls of reused pottery across all the Hampshire sites included
in this study, both of which were found at Danebury. The complete lack of this material in the
Hampshire assemblages may be explained through the geology of the region. Four of the
settlements are located on chalk (see fig. 35); like Dorset, this material is dense enough to
create tools of the correct weights and dimensions for spinning and would make reused
pottery unnecessary for tool use. The only site not located on the chalklands is Zions Hill
Farm; the material at this site is mudstone surrounded by clay, both of which are unsuited to
producing carved spindle whorls. However, Zions Hill did not produce any spindle whorls,
possibly because the textile production in Hampshire was organised differently to the Dorset
region.
Figure 36. A map of the geography of Kent, with sites highlighted. After English Heritage 2011.
In Kent, the oppida site of East Wear Bay contains the largest number of reused pottery
sherd spindle whorls, with 12 out of 24 made from this material. Of these whorls, only one
measured under 10 g and two over 30 g. East Wear Bay is sited on a mixture of sandstone
and mudstone (see fig. 36) that could be well suited for carving into spindle whorls but was
not used for spindle whorl production. This could be because a large industry already existed
that quarried this for quernstones (Keller 1988). This quernstone quarry clearly provided a
valuable economy for East Wear Bay, as stones made at this site can be found across the
South coast and further north (Keller 1988). It is likely that pottery was used in place of the
local stone because it was easily available, easily carved, and a cheaper material in
comparison. Of the other sites, Thanet Earth, East Wear Bay, and Sunset Caravan Park
171
produced no pottery spindle whorls. Like East Wear Bay, Thanet Earth is located on stone
suitable for spindle whorl manufacture. Without the presence of a quernstone industry
making the stone more valuable as a trade commodity, the chalk and sandstone of Thanet
Earth could have made pottery spindle whorls unnecessary.
In comparison, Highstead and Sunset Caravan Park are sited on a mixture of gravel, clay,
silt, and sand (see fig. 36). This material is suitable for pottery production and could provide
pottery for spindle whorl manufacture; however, while one of the two spindle whorls found at
Highstead is made from a reused pottery sherd, Sunset Caravan Park did not produce any.
There was significant evidence of large-scale pottery production at Sunset Caravan Park, yet
the spindle whorls mentioned in the excavation record (Allen 1997) and recorded in this
study were all of fired clay, not reused sherds. It is possible that because of the high level of
pottery production at this site, that any broken pottery was reused as grog by the potters and
was therefore unavailable for the spinners to adapt into whorls. If sherds were unavailable but
clay was common, using and firing clay to make spindle whorls would have been the next
easiest method of making tools, regardless of any potential difficulty in making standardised
tools.
The evidence from Kent and Hampshire presents an intriguing qualification to the theory
that the use of pottery sherds is only indicative of textile standardisation in the late Iron Age;
the use of pottery sherds does not solely depend on the requirements for textiles and textile
production, but also on the other crafts at each settlement and the economic importance of
these activities compared to textile production. This can be demonstrated in two ways. First,
where the local stone suitable for spindle whorl production is already used for an
economically important craft (such as quern production at East Wear Bay), pottery sherds are
the material that spinners are most likely to use instead. The consistent shape and weights of
this material is well-suited to spindle whorl use, it is readily available, easily replaceable, and
can be produced by the spinner without much difficulty. Furthermore, it is not valuable to
others and is not required elsewhere. Second, where there is no suitable stone and pottery
sherds are useful in another economically important craft (such as pottery production at
Sunset Caravan Park), these sherds may not be available to the spinners and may be used in
the other craft’s chaîne opératoire.
This has the potential to highlight the importance of textile production at such sites. For
example, the use of pottery sherds for further pottery production at Sunset Caravan Park
indicates that textile production was not as economically valuable - that is, that textiles did
172
not have the trade importance that pottery had. This supports the theory that textile
production at this site was a domestic craft used to supply the site itself, not to create a
surplus for further trade or exchange. In comparison, Hengistbury Head is located on similar
deposits of clay, gravel, and silt but has an assemblage of spindle whorls almost completely
made from pottery sherds. This site also produced evidence of large-scale pottery production,
so the use of sherds for these spindle whorls suggests that textile production on this site
produced a surplus of items that could be traded or exchanged and possibly held higher value
than the ceramics.
Wider context
It is clear that the evidence and its implications that have been discussed so far present a
unique insight into textile production, craft worker skill and status, and modes of organisation
in the British Iron Age. But what about the wider context? How does this evidence match the
contemporary finds in Europe? And how do these issues fit into our current understanding of
Iron Age textile production, craft workers, and organisation on the continent?
Areas of specialisation and raw material production
An example of the distinct character of the organisation of textile production in Cornwall and
Devon is the distribution of textile tools. As previously discussed, spindle whorls appear in
greater numbers close to areas of rough land best suited for grazing livestock, leaving the
more fertile lowland areas free for farming crops. This resulted in areas that specialised in
one part of the textile chaîne opératoire - spinning - and other areas that were close to centres
of trade that specialised in the other stages - weaving and possibly finishing the textiles. That
pattern was not visible in Dorset, Hampshire, and Kent, which suggested that the abundance
of fertile land allowed for more settlements to keep livestock and fully produce textiles. This
method of linking textile tool distribution to possible agricultural practices in order to
determine areas where different stages of the chaîne opératoire occurred has not been
performed before. Yet it is able to highlight areas of potential division in the organisation of
textile production and specialist textile workers. Further studies across Britain and Europe
using this technique could identify similar areas of potential specialisation and organisation,
providing valuable insight into the organisation of regional communities and the varying
roles of settlements that were previously unavailable.
173
Evidence from Europe is equally varied, depending on not only the geography of the area
but also on the materials available for textile production. For example, textile evidence from
Etruscan Italy demonstrates a wide variety of materials, including linen, hemp, esparto grass,
a range of tree basts, sheep wool, goat hair, and white asbestos (Gleba 2016: 239). Some of
these materials (e.g. the mineral asbestos) depend on limited natural sources, while others
may grow wild across the region (such as esparto grass and the trees used for bast). Flax and
hemp may have been widely farmed but rarely left traces in the archaeological record.
However, there are areas of rough pasture suitable for sheep and goats. In northern Etruria,
some seasonal lowland grazing was available in valleys and around coastal lagoons but most
pasture was confined to upland areas and woodlands (Stoddart 2016: 45). In the southern
region, volcanic activity provided some very fertile upland fields. However, these were
limited to areas with soft volcanic rocks and so grazing was confined to those areas with
harder, untillable stone (Stoddart 2016: 45-46). Other than this change, grazing patterns
remained much the same as those in the northern region, that is, limited to upland areas
(Stoddart 2016: 46). This pattern of restricting wool-bearing animals to upland areas of rough
grazing is similar to that found in Cornwall and Devon, so I would expect to find a similar
divide in the stages of textile production as that seen in Britain’s South West. However, this
method of analysis depends on comparing assemblages from a number of well-excavated
sites. Without a suitably large and thorough excavation, the evidence is too vague to
accurately determine the overall character and organisation of the site, and we risk
misinterpreting the data.
A good example of the risk of misinterpretation can be found in comparing two
contemporary Estrucan sites in the Murlo region: Poggio Civitate and Vescovado di Murlo.
Poggio Civitate lay on higher land to the south of Vescovado and excavations uncovered a
vast quantity of various textile tools from across this site (Gleba 2016: 239-240). The size of
the settlement at Vescovado is currently unclear, as limited rescue excavations have been
conducted in the modern town that covers the site (Tuck et al. 2009). At first glance, there
seems to be a clear difference in the textile tool assemblages from these sites. The majority of
these tools at Poggio Civitate were spindle whorls with only a small proportion of loom
weights found in the site, which suggests that the workers were producing a surplus of
threads and yarns. There are, however, a number of spools from the site, which may have
been used for tablet weaving (Gleba 2016: 240; see also Raeder Knudsen 2012). If these are
indicative of another weaving technique, then this allows the possibility that Poggio Civitate
174
specialised in tablet woven items on a commercial scale (Gleba 2016: 240). On the other
hand, such a high quantity of spindle whorls could suggest that Poggio Civitate was
supplying other sites in the area with spun wool for weaving.
The limited rescue excavations in Vescovado di Murlo appear to support this theory. A 5th
century BC domestic structure was uncovered that revealed 15 loom weights resting on the
structure’s floor, with no other textile tools nearby (Tuck et al. 2009: 228). This is significant
since loom weights at Poggio Civitate were always discovered alongside other textile tools,
including spindle whorls, needles, and spools (Tuck et al. 2009: 228). The excavators
believed that this suggested varying organisations for textile production in the region:
specialist weavers in Vescovado and a larger centre of production in Poggio Civitate,
producing a wider variety of items (Tuck et al. 2009: 228-9). This may be an overestimation,
because while the number of weaving tools in Vescovado is indicative of domestic
production, there is not enough evidence to make definitive statements on whether this
settlement was specialising in one form of textile production. It is possible that further
excavations in Vescovado and the other Etruscan settlements surrounding Poggio Civitate
could show that the textile production in this region was split between the spinners at Poggio
Civitate and weavers in the surrounding settlements, with the woven textiles distributed
before the final finishing stages. However, Poggio Civitate may have been able to produce
enough textiles for its own needs and specialised in producing woollen belts and straps for
further exchange or trading.
Compared to the uncertain Etruscan evidence, Iron Age settlements in Austria and Iberia
show clear patterns of textile tool distribution. Throughout the Iron Age, Austrian sites
demonstrate no noticeable difference in the distribution of textile tools (Grömer 2012, 2013),
which suggests a more equal production of textiles across settlement types, similar to the
evidence from Dorset, Hampshire, and Kent. The only difference between production can be
found in the oppidas of Austria and north-west Europe; these heavily fortified sites cover
many hectares and typically produce a larger quantity of textile tools, sometimes numbering
in the thousands (Grömer 2012; Gleba and Mannering 2012). It is widely agreed that these
large numbers of textile tools produced more than required for the population of the site,
creating a surplus of textiles that could be traded or exchanged for other goods (Gleba and
Mannering 2012). Unfortunately, textile trade has only developed as an area of research over
the last few decades, in part due to the difficulties of finding and tracing organic materials
(Gleba and Mannering 2012; Hofmann de Keijzer et al. 2005). Without the textiles
175
themselves, it is impossible to determine how far textiles, yarns, threads, or even raw
material travelled and limits research to noting possible centres of surplus production. For
non-oppida settlements, it appears that all sites across northwest Europe worked on all stages
of the chaîne opératoire, primarily to produce textiles for their own population (Gleba and
Mannering 2012). While it is possible that traded textiles were used in these various, smaller
sites and settlements, it is not clear how many were brought into a site or what these foreign
fabrics would have been used for.
In comparison, Iberian sites have a clear and noticeable difference in the number and
standardisation of textile tools. Country estates have assemblages of textile tools made to set
sizes and weights using molds, as shown on a number of spindle whorls from Cancho Raono
and La Mata (Marín-Aguilera 2019). Making spindle whorls to the same shape, size, and
weight allowed less experienced spinners to produce a larger quantity of threads and yarn of
the same (or similar) diameter, which in turn allowed the weavers to produce more textiles of
a set thickness (Marín-Aguilera 2019). On the other hand, excavations of the smaller
settlements surrounding Cancho Roano and La Mata produced no textile tools at all
(Marín-Aguilera 2019); only religious centres produced textile tools, which appear to have
made textile for the rites and ceremonies held at these centres (Marín-Aguilera et al. 2019).
This clearly reveals a highly controlled organisation both of tool production and textile work,
as well as of the specialist workers who made the tools and those who made the textiles. The
textiles made at Cancho Roano and La Mata were not primarily meant to supply the
household, but to provide goods that could be traded and exchanged elsewhere and
economically support the household (Marín-Aguilera 2019). The evidence from this region
does not match any of the evidence found in the study area; textile organisation appears to be
dominated by social factors and independent of  analysis of agricultural and grazing areas.
Specialist workers and centres of production
When specialisation is defined as workshop production, it is difficult to definitively identify
areas with specialist workers or centres of production, especially when dealing with textiles
and other crafts that were primarily made in domestic contexts. With the current definitions
of industry and specialism, positive identification of specialists and specialist production is
debatable unless separate, purpose-built workshops are found. Only Winnall Down has
produced structures that could be considered weaving workshops; even the evidence from
176
Kent’s Cavern is debatable, since this was not a man-made structure. Nonetheless, we can
compare the evidence from each county to the definitions of specialist production/industry
made by Costin (1991) and van der Leeuw (1977) (see Table 2, p. 51).
The spindle whorls and loom weights from Cornwall and Devon are made primarily from
local materials, with only a few areas using imported materials from elsewhere along the
south coast. The distribution of tools is strong evidence that different stages of production
were split between communities, yet there are no clear structures dedicated to textile
production in any of the settlements included in this study. Van der Leuuw’s definitions
mainly focus on the distribution of finished products (e.g. the textile itself), which is
impossible to track in Iron Age Britain. However, the pattern found in Cornwall and Devon
fits Costin’s definition of community specialisation almost perfectly. The majority of
households took part in textile production, spinning and weaving were performed in separate
areas, and materials from neighbouring areas were used to create textile tools.
In Dorset, the evidence is more mixed, with all sites except Gussage All Saints having
similar proportions of spindle whorls to loom weights. Only local materials are used for these
tools, there are no sites with only one form of tool, yet the evidence suggests that most
households were involved in textile production. As there is no division of the stages of
production, Costin’s definition of dispersed workshops could be applied to the Dorset
evidence. On the other hand, community specialisation could also apply as most households
appear to be involved in textile production. In addition to the above, the high number of tools
discovered in Dorset’s sites better fits Van der Leeuw’s definition of workshop or village
industry.
Hampshire’s tools distribution shows a clearer pattern than that of Dorset. Like the South
West, some sites appear to be limited to certain activities. While spinning was the only
activity in some South West sites, weaving is the only textile activity in Zion Hill and
Winnall Down. Easton Lane also has a heavier emphasis on weaving, with only a single
spindle whorl from this site. In comparison, Danebury has the highest proportion of spindle
whorls in the county, although weaving tools make up the majority of its textile tool
assemblage. All the textile tools in this county are made from local materials, although few
houses in the smaller settlements have evidence of textile production. This evidence is very
similar to that found in the South East. Kent provides the greatest challenge in terms of
177
defining production, due to the limited evidence from this region. Nonetheless, it is clear that
weaving is not common across all the sites, being limited to Thanet Earth and Highstead
(with only two loom weights found at East Wear Bay). The highest proportion of loom
weights is found at Highstead, a reasonably large open site with a variety of other craft
production activities, and the highest proportion of spindle whorls from East Wear Bay, an
oppida site with strong trade links to the continent. The other sites produced few textile tools
and all tools in this region were made from local materials. This evidence suggests that
dispersed workshops existed in the smaller sites, with community specialisation occurring at
a few settlements with strong trade links or high levels of craft production (Costin 1991).
Alternatively, Van der Leeuw’s definition of individual industry fits the evidence from
smaller settlements, while workshop industry fits the evidence from larger settlements.
As with the variety of modes of specialisation found in the study area, contemporary
evidence from Europe suggests that textile production differed across regions and in response
to other social and economic factors. For example, the Iberian Iron Age sites of Le Mata and
Cancho Roano have large numbers of textile tools, with tools for spinning and weaving made
to precise, standardised weights and dimensions, and textile production occurring both inside
the main structures and outside in open areas (Marín-Aguilera 2019). In the smaller
farmstead settlements surrounding these wealthy sites, no textile production was conducted at
all (Rodríguez et al. 2004; Mayoral et al. 2011). Clearly, the textile production in this region
relied upon dispersed workshops (Costin 1991) or workshop industry (Van der Leeuw 1977).
Marín-Aguilera (2019) suggested two possible explanations for the presence of increased
craft specialisation in individual, wealthy sites: firstly, that these sites produced high-quality
fabrics for use in religious ceremonies (as seen at La Mata) and secondly, that these sites
produced textiles to trade or sell to other local settlements in order to fully or partially
support the household (as at Cancho Roano).
This is a vast departure from the evidence found in this study, where even smaller sites
surrounding larger centres of production had some level of textile production. While there are
some sites in Britain without textile tools (not included in this study), it remains unclear
whether the lack of tools is due to issues in preservation rather than a complete lack of textile
production at these sites. Given the importance of textiles across Europe in this period, it is
unlikely that few sites held complete control over the production and distribution of textiles
for the same economic reasons (e.g. trade or exchange) as Cancho Roano.
178
Instead, I would argue that the structure of society was the most important factor in
determining the organisation of textile production. In Devon and Cornwall, it is clear that
certain areas specialised in spinning and others (commonly the sites with access to trade
routes) in weaving. This strongly suggests that while weaving was focused in a few sites,
finished textiles were then distributed to the smaller settlements, either as payment for
providing spun threads and yarn or in exchange for other goods (e.g. crops, livestock, other
raw materials, etc.). This is very different from other areas along the south coast, where each
small site appears to produce textiles for its own use and larger sites produce textiles for
trade.
While Cancho Roano provides clear evidence for one mode of textile production, evidence
for the organisation of textile production elsewhere in north-west Europe is more difficult to
find. The assemblages at the Hallstatt salt mines, for example, do not include textile tools,
which makes it difficult to determine the organisation of textile production. Elsewhere in
Austria, Hallstatt-period settlements do not have a clear distinction between textile tool
assemblages in large, fortified hilltop sites and smaller, lowland settlements (Grömer 2013:
47; Lauermann 2000; Stöllner 2002a: 110; 1996: 101, taf. 88). Looms over several meters
wide, spindle whorls, needles, shears, and weaving combs are found at different types of
early Iron Age sites in this region (Grömer 2013: 47), which suggests that the various stages
of textile production were not separated between settlement types, similar to the evidence
from Dorset. However, there is no evidence that these large looms were a common feature
across all or most settlements. When considered alongside the diverse weaves, dyes, and spin
patterns found in the Hallstatt textiles; it seems more likely that every community was
producing its own textiles while its textile producers experimented with various techniques.
For example, some communities where most textile workers produced these large textiles,
while another settlement’s textile workers focused on experimenting with weave patterns,
and another site’s textile workers worked on spin direction in weaving. Taking this into
consideration, early Iron Age production in the Hallstatt area appears to be most similar to
individual or household production (Costin 1991; Van der Leeuw 1977).
In comparison, late Iron Age sites in Austria corresponds to a change in textile
standardisation, which suggests a change in the organisation of textile production from
skilled individuals to community production. The variety of textile tools diminishes, which
suggests a shift away from the specialist skilled individuals to more standardised spinning
179
and weaving techniques (Grömer 2013: 47). For example, the settlement of
Dürrnberg-Ramsaukopf (near the Dürrnberg salt mines) has an assemblage of textile tools
similar to those from the early Iron Age; these were found alongside clear evidence of other
craft production based in workshops (e.g. metal working, pottery production, etc.) (Brand
1995; Irlinger 1995). Since this site had workshop production for other crafts, it is not
unreasonable that textile production could also occur in separate structures as well. However,
the distribution of textile tools in the settlement makes it clear that no such workshops were
used for textile production and that this craft was performed in domestic structures (Brand
1995; Irlinger 1995: taf. 76-81). This distribution pattern is strong evidence for community
specialised production similar to Cornwall and Devon, especially when considered alongside
the textiles from the Dürrnberg salt mines, which were less complex and more standardised
than the textiles from Hallstatt (von Kurzynski 1996: 36).
Skill and social status
The discussion of the status of textile workers commonly relies on the presence of textile
tools in graves and overlooks the use of decoration on textile tools. Some Iron Age textile
tools, such as spindle whorls and weaving combs, are heavily decorated across large areas
and various types of settlement, which would suggest that decoration was not a method of
identifying textile workers from centres of production or power. However, these decorations
are often dismissed for occurring on utilitarian items, rather than items of prestige, power, or
ritual importance. I would argue that decoration of everyday, utilitarian items provides more
of an insight into the common use and social importance of these objects and the people who
made or used these. While decorations on prestige goods or ritual artefacts demonstrate
fashionable or valuable items, similar embellishments on otherwise unremarkable tools
present an insight into the choices of craft workers and the value placed upon them, their
skills, and the items they produced on a daily basis.
The textile tools chosen for decoration in British Iron Age are typically spindle whorls and
weaving combs, which are smaller, easy to carry, and often simple to carve decorations into.
The types of decorations used on spindle whorls are similar to those found on contemporary
pottery - lines following the shape of the item, chevrons, diamond patterns, and dots (see fig.
37, below). Decorated spindle whorls are not common in most sites but there are areas where
decoration appears to be the norm for these tools; the Torbay region was one such area that
180
happened to be included in this thesis’ study area. It is interesting that elsewhere in the study
area, spindle whorls with decoration were exceptional. Perhaps these decorated spindle
whorls indicated spinners of particular skill or status? Or possibly these were gifts from
contacts in the Torbay region, strengthening networking between the communities and the
craft workers. Another explanation is that these represent women (or possibly men, if they
conducted spinning) moving to other areas, possibly following family members or moving
for marriage purposes. Closer analysis of the materials that these tools were made from is
required in order to determine if the finished tool originally came from the Torbay region (or
another region with a similar culture of spindle whorl decoration) or the area in which it was
discovered. Without this, it will not be possible to determine if it demonstrates the movement
of people or a display of skill or status.
Figure 37. Two examples of the forms of decoration present on Torbay spindle whorls, from the Torbay
Museum’s collections. After Silvester 1986.
The purpose and meaning of decorated spindle whorls in areas with a tradition of
decoration, such as the Torbay area, is more complex. The Torbay region covers a small area
in South Devon that includes Kent’s Cavern, Milber Down, and Dainton (see fig. 38 below).
This area has larger numbers of decorated spindle whorls than elsewhere in the study area, as
shown below:
Kent’s Cavern: 12 decorated SWs out of 12 Cornish sites: 4 decorated SWs out of 71
Milber Down: 2 decorated SWs out of 6 Dorset sites: 11 decorated SWs out of 149
Dainton: 2 decorated SWs out of 6 Hampshire sites: 4 decorated SWs out of 30
Other Devon sites: 0 decorated out of 7 Kent sites: 0 decorated SWs out of 31
181
The presence of decoration on so many whorls in this area (almost all excavated) indicates
that these were not markers of particular status within the communities here. Spindle whorls
were tools that could be used while on the move - many historical and modern sources show
women spinning while walking, herding livestock, and chatting to friends - could the public
aspect of this craft influence the desire for decoration? A way of displaying your abilities or
status as a spinner for all to see? If so, we might expect to see more decorated whorls
elsewhere along the coast, after all, there is no evidence that spinning was an indoor activity
elsewhere in the study area. As this is not the case, I argue that it is more likely that the
Torbay region was known as an area of particular skill and specialisation in spinning, such
that the spinners decorated their tools as a way to display that they were part of this skilled
set of specialist workers.
Torbay is also the region where Kent’s Cavern is located, a cave site that was almost
certainly used as a seasonal workshop for spinners, cleaning and processing raw wool from
sheep before distributing the threads and yarns for weaving elsewhere. This may have
required significant time to accomplish, depending on the amount of raw material and the
number of spinners available to work it. The number of spindle whorls excavated from
Kent’s Cavern is higher than would be expected for a site that produced material for the local
area. Considering this alongside the significant numbers of decorated spindle whorls from the
Torbay area, I argue that Kent’s Cavern was a place where the spinners of Torbay came
together to process and spin material required for further weaving. The decoration of
Torbay’s spindle whorls marked those who took part in this annual event that connected the
settlements and the spinners in this region in a way unique to other societies along the south
coast. These decorations may not have been to identify particularly skilled craft workers, but
to display a group identity linked with this region’s textile traditions.
182
Figure 38. A map of the Torbay region, with sites included in the study area highlighted. A - Kent’s Cavern; B -
Milber Down; C - Dainton. Author’s illustration.
Decorated spindle whorls have been found in great numbers from early Iron Age (Hallstatt
period) contexts throughout central Europe, particularly in the Hallstatt region (Belanová
Štolcová and Grömer 2010: 12-13). These spindle whorls can be elaborately decorated, as at
Molpír hillfort, Slovakia, which produced over 2200 spindle whorls (Belanová Štolcová
2007: 41-43). These decorations disappear in the late Iron Age when spinners shift to using
whorls made from pottery shards; this change also corresponds to the standardisation of
textile weaves, patterns, and dyes seen in the Hallstatt and Dürrnberg assemblages (Belanová
Štolcová and Grömer 2010: 13). This is further evidence of a shift in the focus and
organisation of textile production from individual specialists to a more community-based
system of production. Unfortunately, little work has been done on the distribution of
decorated and undecorated spindle whorls of the Hallstatt period, so it is difficult to
determine whether these tools have a similar regional distribution pattern to that seen in
Torbay. If so, this could suggest annual community work performed by craft workers in
settlements containing large numbers of decorated tools; if not, it is likely that these
decorated tools had symbolised an individual’s skill or status, rather than their participation
in a larger, regional textile community.
The decoration of British weaving combs is slightly different from the forms found on
spindle whorls. Additionally, the distribution of such tools is more widespread, with
183
examples found in every Iron Age settlement where several of these tools survived
deposition. Instead of having lines incised around the edge of the tool (as the spindle whorls
often do), weaving combs are commonly decorated with chevrons, boxes with diagonal lines,
crosses, and circles with a central dot (see fig. 39 below for examples). Like spindle whorls,
weaving combs appear to be designed for carrying about; at the end of the handle, most of
these combs either have a small hole for a string or thread to go through or a wide butt that a
string or thread can wrap around, in order to create a wrist or belt strap.
The requirement for an attachment, in addition to the decoration found on the majority of
weaving combs, suggests a public function or social aspect to these tools beyond the
utilitarian needs of the craftsperson. However, there are three main factors that distinguish
decorated weaving combs from decorated spindle whorls: first, decorated weaving combs are
found across most sites and regions. There are no unusual concentrations of these tools as
there are with the spindle whorls and weaving combs are only absent from late Iron Age sites
with acidic soil conditions that would quickly destroy these tools. Second, weaving combs
(with or without decoration) are predominantly found in the later Iron Age across Britain
(Barber 1991; Tuohy 2004), while decorated spindle whorls are more commonly limited to
the early and middle Iron Age. Finally, weaving combs are almost always decorated while
spindle whorls are not, despite both tools having a public function. Why should two different
tools that are both seen on a daily basis and connected with an important craft be treated so
differently?
184
Figure 39. Examples of common decoration on Iron Age weaving combs from Danebury. Left: Comb number
3.238; centre: comb number 3.239; right: comb number 3.240. After Cunliffe 1991.
The decrease in decorated spindle whorls may be in part due to the change in the material
of these tools during the late Iron Age from carvable stone and clay to harder pottery that was
reshaped into whorls. This alteration occurred across Iron Age Europe (as seen at Dürrnberg
and Cancho Roano) and is generally considered to reflect a change in the range of thread
diameters used in textile production. It is also possible (at least in Britain) that the increase in
weaving combs and the change to spindle whorls of reused pottery reflect a shift in weaving
technology (Tuohy 1999). Tina Tuohy (1999) proposed that weaving combs were designed
for use on smaller belt looms, rather than as weft-beaters for the larger warp-weighted loom,
as the use-wear marks on the teeth of combs across Britain demonstrated little to no wear on
the outermost sections. If these tools had worked on large looms, we would expect to find
wear marks across all sections of the comb’s teeth; such wear is limited to the inner teeth
instead, which suggests a far smaller section of warp was used. Although there was still a
need for large textiles and the large looms that made these, the increase in weaving combs is
possible evidence for larger production of belts and straps and suggests that weaving swords
185
were in use for larger looms across Britain, despite the lack of evidence for these tools due to
poor preservation. Previously, such belts and straps were tablet woven, but the set-up for this
method takes time to thread and specialist knowledge and training in constructing the pattern
and using the tablets properly. In comparison, weaving combs can be used on a simpler
set-up based on the weaving techniques of larger looms. Rather than twisting warp threads
with tablets, the warp threads can be raised and lowered in the same way as on a large loom
to create belts and straps with tabby or twill weave patterns. This faster method of weaving
such items would have benefitted the textile workers of the late Iron Age, especially as the
evidence strongly suggests that textile work had shifted to faster production, larger quantities,
and more standardised textiles. This evidence matches the developments found in European
textile tool assemblages, although it is difficult to determine whether this change occurred in
response to new demands from textile traders on the continent.
The widespread adoption of decoration on weaving combs compared to the limited practice
of decorating spindle whorls in the early and middle Iron Age could demonstrate the status of
textile workers in everyday society. Experienced spinners could walk and complete other
tasks while using spindle whorls, however weaving cannot be moved around in the same
way. This allows spinning to be a more public activity than weaving, so a spinner may not
need to draw attention to the tool they are using; on the other hand, weavers cannot publicly
display their skill in the same way. If public display of a person’s status as a weaver was
important, it could explain why the new, smaller, and lighter weaving tools that developed in
the late Iron Age became so commonly decorated. If these weaving combs were worn on the
person, it could be a clear symbol of their status as a craft worker while the decoration also
showed their individual tastes. Decorated weaving combs are also found in late Iron Age
contexts in north-west Europe, although little work has been conducted on the distribution
patterns of textile tools in general. It is difficult to determine whether decorated weaving
combs outnumber undecorated ones or how the overall distribution of these tools (both
decorated and plain) relates to other textile tools, both contemporary and earlier in date. Until
such a study has taken place, it is impossible to determine whether the evidence in Britain
corresponds to that in Europe.
Burial contexts provide other sources of evidence in Europe as to the status of textile
workers, and the possible importance of textile work as part of an individual’s identity.
However, burial customs do not accurately represent how the craft work was structured in the
186
living communities. Grömer (2013: 46) argued that textile tools found in Hallstatt graves
with jewellery were used to identify women who were heads of the household, who possibly
had the same roles as women in ancient Greece who were responsible for the production of
standard textiles as well as the more complex and high-quality textiles used for high-status
gifts (2013: 46). However, she also notes that textile tools (including spindle whorls) are
found in ‘poorer’ burials with little to no jewellery; in these cases, Grömer presents the idea
that textile tools were used to symbolise some form of ‘womanhood’ or the individual’s
contribution to society (2013: 46). I do not agree with the claim that the same tool, made
from similar materials in similar shapes, represents such different ideas based solely on other
grave goods. There are no notable differences in settlement structure or building size or
construction that could support the theory of wealthy households where one woman could
exercise the same level of control over textile production as found in ancient Greece. While
Grömer noted that the variety and experimentation in Hallstatt textiles could be used to
represent wealthy or important individuals or families in society (2013: 46), it is unclear how
these textiles were originally used and the status of the original wearers, since these were
found as scraps reused in a mine and not as full textiles. I would argue that it is more likely
that the textile tools found in burials are used to display part of the person’s overall identity
or social status; that this person was a spinner, an important craft worker, as well as a woman.
Evidence for textile standardisation
One form of textile production common to European centres of production in the Iron Age is
the production of highly standardised textiles. As demonstrated by the textile assemblages in
Hallstatt and Dürrnberg, textiles shifted from using a variety of weave patterns (tabbys,
twills, and so on), alternating spin patterns, and a range of dye colours in the early and
middle Iron Age (Grömer 2005, 2013; Hofmann-de Keijzer et al. 2013) to a limited number
of weave patterns (predominantly tabby), few spin patterns, and little to no dyes used to
pattern the fabrics (Stöllner 2005). This change in weaving techniques and textile production
can be found across north-west Europe (Belanová-Štolcová 2005) and is reflected in the
textile tools used by the workers. While the decrease in the complexity of textiles may
demonstrate a decrease in knowledge and skill, this does not necessarily mark a decrease in
specialisation; rather, it marks a change from individual specialists within a community to a
broader form of community specialisation (Grömer 2005, 2013; Stöllner 2005). This
development of large-scale textile production and specialisation matches the form of textile
187
production common throughout southern Europe, as seen at the Iberian sites of Cancho
Roano and La Mata, as well as in sites across Italy and the Aegean (Andersson Strand and
Nosch 2015; Burke 2005; Marín-Aguilera 2019; Marín-Aguilera et al. 2019). So how does
the evidence found in this study compare to the patterns of production found in Europe?
As discussed above, the use of pottery for spindle whorls in British sites along the South
coast depends not only on the geology of the surrounding area, but also on the economic
importance of the materials each site has access to. In areas where pottery production
supplied major tradable goods, broken pottery was reused in as grog by potters; in areas
where stone was a valuable export, it was not available for spinners, even if it was perfectly
suited to their needs. These results clearly show that textile tools do not only reflect the
choices of its own craft workers and the textile needs of its community, but also the larger
importance of textile production as an economic activity compared to other crafts performed
at each site. Because textile tools, production, and workers supplied a community, it is
inherently integrated with the wider community; the needs and comparative importance of
another craft could have a large impact on textile production and producers, and vice versa.
Hengistbury Head’s textile tools are almost completely made of reused pottery sherds, which
would have forced potters to find, gather, and prepare another (possibly less
readily-available) material to use as grog. This clearly demonstrates the importance and value
placed on textile production compared to pottery production in this site. Given the status of
Hangistbury Head as a centre of trade for the South coast, it further suggests textile workers,
not potters, were producing tradable goods and not just supplying their own community.
188
6. Conclusions
This study set out to determine what textile production techniques and modes of organisation
were present across various settlements of the south coast of Britain during the Iron Age, how
these may have differed between site types, whether these differences could identify areas of
specialisation or particularly high levels of skill, and whether textile tool materials or use
could demonstrate specialisation in textile production.
Aims and objectives
The first aim of this thesis was to identify differences in textile production techniques; this
aim was fully met, as the methods of tool analysis were able to shed light on what the textile
tools were most likely used to produce. Furthermore, in areas where preservation allows for
the survival of organic tools (such as weaving combs), the wide scale survey of textile tools
conducted can highlight different weaving techniques and forms across large timeframes. The
difference between weaving forms in an early Iron Age site compared to those from its late
Iron Age phase could be traced via the numbers of weaving combs compared to the number
of loom weights. These could then be compared to the character of the threads produced by
spinners to determine the most likely form of textiles produced in these periods and identify
possible changes in what the communities needed these textiles for. Similarly, this thesis has
shown differences in textile production techniques and traditions across the south coast of
Britain by proving that the weaving techniques in Cornwall and spinning traditions in Devon
(specifically the Torbay area) were extremely different from those in Kent, Hampshire, and
Dorset. Only this closer analysis of the dimensions and weights of various textile tools was
able to identify that Cornwall and Devon were producing far finer textiles - most likely only
with wool - than elsewhere in the study area.
The second aim was to identify differences in the organisation of textile production, which
this thesis was able to achieve in three main ways. Firstly, through the identification of textile
tools that produced unusually fine textiles (as mentioned above). Secondly, by identifying
sites with only one form of textile tool, such as Bodrifty and Carn Euny in Cornwall, as well
as Kent’s Cavern in Devon. Having identifying these sites, it was possible to examine other
possible reasons why these may have a restricted selection of tools compared to other areas,
such as the stages of textile production were split between settlements rather than that tools
189
were destroyed due to poor preservation. The third and final method of identifying
differences in the organisation of textile production was through the identification of unusual
concentrations of tool features, like the spindle whorl decoration present in the Torbay region.
Investigating the distribution and context of these features could determine whether these
were part of a group tradition or identity due to the large numbers of tools displaying these
features, or whether these were specialist tools that were used to distinguish a skilled worker.
The third aim of this thesis was to identify textile specialists and/or skilled textile workers.
While this thesis was able to note specialisation in certain areas (such as Cornwall and
Devon), identifying skilled individuals was more difficult. This was in part due to the nature
of textile production as a craft activity in the British Iron Age. Unlike in Iron Age Iberia, the
Mediterranean, and the Aegean, textile production was a common activity performed on a
household scale in almost all sites across the study area. This constant practice and
production would have resulted in a certain level of skill in the population, which would
make the identification of particularly skilled individuals more difficult. It was easier to note
areas of specialisation, especially where this specialisation resulted in an increased quantity
rather than quality of textiles.
This thesis was able to identify areas of specialist production by contrasting tool
assemblages with contemporary environmental and geological information. This made it
possible to gain closer insights into the availability of raw materials for textile production and
for textile tools, which in turn highlighted areas that specialised in one stage of the chaîne
opératiore, such as Bodrifty and Carn Euny in Cornwall, which only worked on spinning and
not weaving. This technique indicates regions that focus on producing a surplus of textiles for
further trading or exchange. However, it also allows for further considerations of prehistoric
craft to be explored. Questions can be raised as to how production was organised across
regions, as well as how the finished textiles were distributed to these non-weaving
settlements. If the textile workers of one area are specialist spinners and no weaving occurs
there, how does the community get clothing and other essential fabrics?
These communities could trade other goods for such items from the sites that specialise in
weaving, or the woven textiles could be redistributed to these spinning communities in return
for supplying the weavers with the necessary materials. If the supply of textiles in all
communities within the kinship groups in a region depended on trade, this means there was
an economy for textiles at all levels of society and not just for trade outside of these
communities. This would then carry implications of control over textile producers, the stages
190
of textile production, and the raw materials needed for both. On the other hand, if the textiles
were provided in exchange for the time spent on spinning and the spun threads required by
the weavers, then this exchange would suggest a regional community working together to
produce textiles and supporting the textile workers and smaller settlements. This would
further suggest regional kinship systems that could also exchange other craft goods, rather
than only textiles.
Advantages
The advantages of analysing textile tool weights and dimensions are well known, as shown
by the extensive experiments performed by the Centre of Textile Research (Mårtensson et al.
2006a, 2006b, 2007, 2009). Without this previous work on textile tools, it would not be
possible to determine what textiles were most likely made in areas where no textile fragments
existed. This method of tool analysis is able to provide a basic insight into textile production
in prehistory; however, it is only with the addition of distribution analysis that more detail
can be gained about the organisation of textile workers, the choices they made in what tools
they used, the status and contribution of these workers to local or regional economy, and
wider textile traditions across regions and settlement types.
Performing such a wide scale survey had the advantage that it could reveal common
patterns of textile production across various regions; this highlighted the general
characteristics of textiles that were produced during this period in southern Britain. For
example, the most common thread diameter across the various counties and which areas were
better suited for using wool compared to plant fibres. The wide area surveyed was able to
demonstrate that Dorset, Hampshire, and Kent’s most commonly produced thread diameters
were roughly 0.4-0.8 mm in diameter (although some sites were also capable of producing
both fine and thick threads), which could produce a wide variety of textiles but does not
suggest particularly high levels of skill or a small group of specialist spinners. Furthermore,
the comparison of spindle whorl diameters in various sites was able to highlight whether
these tools were best suited to spinning wool or plant fibres (e.g. small diameter compared to
large diameter). When considered on a regional scale, this was able to highlight areas that
were more likely to use wool compared to areas most likely to use both fibres in textile
production.
191
This technique could also identify areas that vary from common patterns of textile
production, for example: Cornwall and Devon. The textile tools in this South West region
(particularly Cornwall) were significantly lighter and thinner than the rest of the south coast,
strongly suggesting that this region specialised in spinning and weaving particularly fine
textiles compared to the rest of the study area. If this study had focused on textile tool data
alone, this would have been the extent of the conclusions drawn from the Cornwall and
Devon assemblages. However, this data was further analysed by examining the distribution of
these textile tools across the study area.
This highlights another advantage of this form of wide scale survey: that it can reveal
common patterns of regional tool distribution across various types of settlement and highlight
areas that differed from this pattern, whether as large as Cornwall or as small as Gussage All
Saints. It is important that tool distribution is examined while keeping in mind the context of
each site; its date, likely population size, other craft activities, and so on. Not all sites with
few textile tools are small settlements of only one or two households; for example, the
settlement of Sunset Caravan Park has large-scale pottery production during the Iron Age.
For this site to have so few textile tools is strong evidence that textiles were being produced
for this settlement’s use only, despite there being similarly large sites that relied on textile
production as well as other craft activities for trade (Gussage All Saints, East Wear Bay, etc.).
Conducting tool distribution analysis in a large survey also has the advantage of
determining why certain sites only have spinning tools, or why some sites have a larger
quantity of reused pottery spindle whorls compared to others. By comparing the distribution
of textile tools to known areas of rough grazing and poor agricultural land, it was possible to
determine areas of specialisation, particularly in regions with limited land suited to farming.
For example, the Cornish sites of Bodrifty and Carn Euny, which only had spindle whorls,
were surrounded by poor agricultural land that was unsuitable for farming but were used for
grazing livestock. It is very likely that these sites relied on trade with other settlements in
Cornwall to supply these communities with much of their food needs in return for meat,
leather, and secondary animal products (horns, dairy, wool, etc.). However, when comparing
this information to the textile tool distribution, it becomes clear that no weaving occurred at
these sites, even though spinning was performed. Only the sites with access to arable
farmland and trade to the continent had loom weights. When considered alongside the
evidence for Cornwall producing unusually fine textiles, I would argue this is strong evidence
192
that textile production was performed by communities and organised on a regional scale, with
each area contributing to achieve the production of a larger quantity of textiles.
I repeated this using information on the geology of the region, in order to specifically
investigate the presence of reused pottery sherds (or lack thereof) in late Iron Age textile tool
assemblages. This demonstrated that by comparing the spindle whorl materials from late Iron
Age British sites to the geology of the area, it is possible to determine whether spinners chose
to use locally available material (such as slate in Cornwall and chalk in Hampshire), when
they chose to use reused pottery (as at Carn Euny, Hengistbury Head, and East Wear Bay),
and when they were unable to use reused pottery although it would have been readily
available (as at Sunset Caravan Park).
The method of combining a wide-scale survey of tools and distribution with environmental
and geological data has clear advantages over simply analysing textile tool data, even if this
data is collected across a large area. These additional techniques provided valuable insight
into the types of fabric both produced and used in different settlements or across different
regions that would otherwise not be possible without finding textile fragments. These
techniques highlighted settlements that produced wide varieties of textiles (like Maiden
Castle, Hengistbury Head, and Gussage All Saints) compared to others (such as the Cornish
sites) that made a more limited selection of textiles. These identified areas that specialised in
one stage of the chaîne opératoire, such as Bodrifty and Carn Euny, and demonstrated why
some settlements relied on reused pottery sherd spindle whorls (like Hengistbury Head and
East Wear Bay) while others (such as Maiden Castle and Danebury) did not, even though
these sites produced similar types of textiles in similarly large quantities. These methods also
highlight potential evidence on which raw materials may have been used in textile production
across various regions and settlements, which also raises questions as to the economy of these
settlements (e.g. to what extent it depended on textile production). The importance of textiles
in trade can also be investigated using this form of survey, especially when contrasted to the
evidence of other craft activities at each site.
Issues
There are challenges in conducting such large surveys; primarily with establishing an
appropriately-sized set of raw data. In order for a suitable survey of textile tools to be
conducted, each site should ideally have ten tools or more. The difficulties of finding Iron
Age sites in Britain with this many textile tools required this thesis to work with sites that had
193
over five (or four, in the case of Highstead) tools in total. This low number was not ideal as it
did not provide a detailed picture of the textile tools in these sites, which further limited the
conclusions that could be drawn from the data regarding quality and quantity of production.
On the other hand, it is just as important to include sites with so few tools in order to develop
a more rounded picture of craft production and economy at all levels of society.
It is essential to have the dimensions and weights of textile tools in order to conduct such
surveys, which leads to another major issue: the lack of textile tool data in archaeological
reports. Most modern reports do not include data on the weights or dimensions of the tools
and up until recently, older reports barely mentioned these tools at all. Sites excavated by
antiquarians or up until the mid 1900s rarely provide precise numbers of tools that were
excavated, describing the amount found as ‘a number of spindle whorls’ or ‘several loom
weights’. The best recourse for such an issue is to track down the tools from each site and
determine how many there are before measuring and weighing these items.
Unfortunately, this also highlights another issue with old collections: these can be misplaced
or lost before the tools could be recorded. This occurred in many of the sites throughout East
and West Sussex, which lead to these counties being omitted entirely from this thesis even
though there was evidence that textile tool assemblages at some hillfort sites rivalled those of
Maiden Castle and Danebury in size (see Curwen 1927, 1929, 1931). Although it is
frustrating to lose such a large collection of data, there is little that can be done with lost
assemblages other than to hope that they eventually come to light again.
Further work
It is clear that the study presented in this thesis can provide a thorough and more nuanced
understanding of textile production, textile workers, Iron Age communities, and Iron Age
economic activities. This work has the potential to provide a new insight into the
complexities of craft production in the Iron Age in Britain and northwest Europe. There are
several avenues that can be explored from here onwards.
Firstly, more work needs to be done elsewhere in Britain and Europe on the analysis of
textile tools and the distribution of these artefacts, with comparisons to the contemporary
local environment and the geology of sites. This will be able to prove whether the distinctive
194
nature of British sites shown in this thesis is common throughout Britain and Ireland, and
whether similar evidence exists in contemporary European settlement. So far, the common
consensus holds that northwest Europe developed a system of producing large quantities of
standardised textiles in the Iron Age. By applying the methods of analysis used in this thesis,
it will be possible to gain further insights into the choices available to textile workers and a
better indication of the contributions these workers made to their communities. Furthermore,
it will be possible to not only identify areas, large or small, where the textile workers deviated
from the common pattern of production or organisation, but also highlight the most likely
reasons for such deviation. This would also allow for further study and discussion of group
identities, social status of skilled individuals, and the textile traditions across large areas.
The study itself could be expanded by including data from other forms of textile tools, such
as weaving combs, needles, and so on. This would enable the study to trace developments in
weaving techniques and traditions throughout the Iron Age and over various settlements or
regions. As such organic tools rarely survive outside of alkali or anaerobic soil conditions,
this dataset should be considered as a possible representation of weaving developments
within the local area of the settlements where these organic tools survive. For example, to
claim that Hampshire’s pattern of organic tool distribution - and the technological, traditional,
and social implications that these carry - can be applied to Cornwall’s equally distinctive
weaving traditions and practices, is impractical, unfounded, and lazy. This does not diminish
the importance of the information that Hampshire’s organic tools provide, but it does
demonstrate the importance of considering organic tool data on a local scale, rather than
regional.
It may also be beneficial to create a database of British textile tools for use in similar
studies. Such a database would need to include the following information for each tool:
● Tool’s special finds/museum
number
● Museum and collection information
● Tool dimensions and weight ● Tool material
● Information on tool decoration ● Preservation condition of the tool
● Individual tool context ● Tool illustration, where possible
● Site name, region, national grid ● Site plan, including map of tool
195
reference, and country locations (where possible)
● Number of site’s domestic
structures/population estimate
● Other craft and subsistence activity at
site
● Sources of site information ● Other textile production evidence at site
Similar databases could be created for other countries and time periods, which would create
a useful resource for textile researchers to compare and contrast the evidence from various
regions or countries. This database would be the first of its kind to provide data on craft tools,
although the challenges such a large project presents are huge, including accessing materials
and gaining permission to share such data from each museum or collection. The scale of a
textile tool database, as well as the practical issues of its construction, make such a project
unlikely for the time being.
Nevertheless, the nature of the information gained by the survey conducted by this thesis is
unprecedented. The combination of Mårtensson et al’s tool analysis with distribution
analysis, compared with environmental and geological data, provides a way to explore the
organisation, production, choices, and status of textile workers in prehistory from a fresh
perspective that is more nuanced and thorough than before.
196
Bibliography
Albarella, U. 2007. 'The end of the Sheep Age: people and animals in the late Iron Age.' In: C.
Haselgrove and T. Moore, The Late Iron Age in Britain and Beyond. Oxford: Oxbow Books,
pp.393-406
Allen, T. 1997. The Whitstable Iron Age Settlement. The Archaeological Evaluation and Results of
Preliminary Excavation at Sunset Caravan Park, Thanet Way, Whitstable, Kent, 1997/61.
Canterbury Archaeological Trust
1999. The Whitstable Iron Age Settlement: Parts 1 and 2. Unpublished Canterbury
Archaeological Trust
Allen, T., and Wilson, J. 2001. Sunset Caravan Park and Church Lane East, Whitstable. Canterbury’s
Archaeology 1998-1999: 10-11
Allison, P. M. 1999. The Archaeology of Household Activities. London: Routledge
Andersson, E. 2003. Tools for Textile Production from Birka and Hedeby. Stockholm:
Riksantikvarieämbetet (Birka Studies 8: Excavations in the Black Earth 1990-1995)
Andersson Strand, E. 1998. 'Textile production in late Iron Age Scania - a methodological approach'.
In L.B. Jorgensen and Ch. Rinaldo (eds.) NESAT 6: . Göteborg: Göteborg University
Department of Archaeology.
2005. 'Engendering Central Places: some aspects of the organisation of textile production
during the Viking Age' In: A. Rast-Eicher and R. Windler (eds.) Archëologische Textilfunde
NESAT IX: Braunwald 18.-20. Mai, 148-153
2008. 'Tools, Textile Production and Society in Viking Age Birka'. In M. Gleba, C. Munkholt
and M.-L. Nosch (Eds), Dressing the Past. Ancient Textiles Series 3. Oxford: Oxbow Books,
pp. 65–85
2011. ‘The basics of textile tools and textile technology: from fibre to fabric.’ In C. Michel and
M.-L. Nosch (eds.) Textile terminologies in the ancient Near East and Mediterranean from
the third to the first millennia BC: Ancient Textile Series. Oxford: Oxbow Books, pp. 10-22
2012. The textile chaîne opératoire: using a multidisciplinary approach to textile archaeology
with a focus on the Ancient Near East. Paléorient 38(1-2): 21-40
Andersson Strand, E. and Nosch, M.-L. 2015. Tools, Textiles, and Contexts: investigating textile
production in the Aegean and Eastern Mediterranean Bronze Age. Oxford: Oxbow
Anderson Strand, E., Frei, K.M., Gleba, M., Mannering, U., Nosch, M.-L., and Skals, I. 2010. Old
Textiles - New Possibilities. European Journal of Archaeology 13 (2): 149-173
Andresen, S.T. and Karg, S. 2011. Retting pit for textile fibre plants at Danish prehistoric sites dated
between 800 B.C. and A.D. 1050. Vegetation History and Archaeobotany 20: 517-526
DOI: 10.1007/s00334-011-0324-0
ApSimon, A.M. and Greenfield, E. 1972. The Excavation of Bronze Age and Iron Age Settlements at
Trevisker, St. Eval, Cornwall. Proceedings of the Prehistoric Society 38: 302-381
Arnold, D. 1985. Ceramic Theory and Cultural Process. Cambridge: Cambridge University Press
Ashby, S. 2009. Combs, Contact and Chronology. Reconsidering Hair Combs in Early-Historic and
Viking-Age Atlantic Scotland. Medieval Archaeology 53: 1-33
197
Banck-Burgess, J. 1999. Hochdorf IV. Die Textilfunde aus dem späthallstattzeit-lichen Fürstengrab
von Eberdingen-Hochdorf (Kreis Ludwigburg) und weitere Grabtextilien aus hallstatt- und
latènezeitlichen Kulturgruppen. Forschungen und Berichte zur Vor- und Frühgeschichte in
Baden-Württemberg 70. Stuttgart: Theiss Verlag
2012. ‘Case Study: The Textiles from the Princely Burial at Eberdingen Hochdorf, Germany’.
In: M. Gleba and U. Mannering (eds.): Textiles and Textile Production in Europe from
Prehistory to AD 400. Ancient Textiles Series Vol. 11, Oxford: Oxbow Books: pp 139–152
Barber, E. J. W. 1992. Prehistoric Textiles: The Development of Cloth in the Neolithic and Bronze Age
with Special Reference to the Aegean. Princeton: Princeton University Press. 3rd printing.
1994. Women’s Work: the first 20,000 years, Women, cloth and society in early times. New
York: Norton.
Barrett, J.C. 2000. ‘A thesis on agency.’ In: M.A. Dobres and J. Robb (eds.), Agency in archaeology.
London: Routledge, pp. 61-68
Barth, F.E. 1988/89. ‘Zum Goldscheibenfund von Óbuda’. Mitteilungen der Anthropologischen
Gesellschaft in Wien 118/119: 155-62
Bazzanella, M., Mayr, A., Moser, L., and Rast-Eicher, A. 2003. Textiles: intrecci e tessuti dalla
preistoria europea. Catalogo della mostra tenutasi a Riva del Garda dal 24 maggio al 19
ottobre 2003. Riva del Garda, Trento: Esperia
Bazzanella, M. and Mayr, A. 2009. I reperti tessili, le fusaiole e I pesi da telaio dalla palafitta di
Molina di Ledro. Trento: Provincia Autonoma di Trento, Sopraintendenza per i beni library,
archivisti e archeologici
Beamer, J. 2017. ‘Long Handled Combs: Determining 'Weaving' Comb Function by Examining
Textile Tool Materials and Construction Techniques.’ 3rd Annual SAAH PGR Conference.
University of Leicester, Leicester, 17 November
2018. ‘Long-handled weaving combs – a rubric for recording.’ LTS: Exploring Textiles and
Textile Working from Prehistory to AD 500. University of Liverpool, Liverpool, 29 October
Belanová-Štolcová, T. 2005. 'The state of research of La Tène Textiles from Slovakia and Moravia.' In
P. Bichler, K. Grömer, R. Hofmann-de Keijzer, A. Kern and H. Reschreiter (eds), Hallstatt
Textiles Technical Analysis, Scientific Investigation and Experiment on Iron Age Textiles.
Oxford: BAR Int. Series 1351, 175–189
2012. 'Slovak and Czech Republic.' In M. Gleba and U. Mannering (eds.) Textiles and Textile
Production in Europe: From Prehistory to AD 400. Oxford: Oxbow Books, 306-333
Belanová-Štolcová, T. and Grömer, K. 2010. 'Loom-weights, Spindles and Textiles - Textile
Production in central Europe from the Bronze Age to the Iron Age.' In E. B. Andersson
Strand, M. Gleba, U. Mannering, Ch. Munkholt, M. Ringgard (eds.), North European
Symposium for Archaeological Textiles: 13.-18. 5. 2008 in Copenhagen. Oxford: Oxbow
Books, 9-20
Bender Jørgensen, L. 1986. Forhistoriske textiler i Skandinavien. Prehistoric Scandinavian Textiles.
Nordiske Fortidsminder Ser. B, vol. 9. Copenhagen: Det Kongelige Nordiske
Oldskriftselskab
1992. North European Textiles: until AD 1000. Aarhus: Aarhus University Press
2012. ‘The Introduction of Sails to Scandinavia: Raw materials, Labour and Land’ In: R. Berge,
M.E. Jasinski, and K. Sognnes (eds.), N-TAG TEN Proceedings of the 10th Nordic TAG
conference at Stiklestad, Norway 2009. BAR International Series 2399. Oxford:
Archaeopress, pp. 173-181
198
Bender Jørgensen, L. and Rast-Eicher, A. 2013. Sheep wool in Bronze Age and Iron Age Europe.
Journal of Archaeological Science 40(2): 1224–1241
2016. Innovations in European Bronze Age textiles. Praehistorische Zeitschrift 91(1): 68-102.
https://doi.org/10.1515/pz-2016-0004
Bennett, P., Couldrey, P., and Macpherson-Grant, N. 2007. Highstead, near Chislet, Kent. Excavations
1975–1977. Canterbury Archaeological Trust Monographs: New Series Vol IV
Bergfjord, C., Mannering, U., Frei, K.M., Gleba, M., Bruselius Scharff, A.. Skals, I., Heinemeier, J.,
Nosch, M.-L., and Holst, B. 2012. Nettle as a distinct Bronze Age textile plant. Scientific
Reports 2, article No. 664, doi: 10.1038/srep00664
Big Dig Diary, Must Farm. 2015. Textiles at Must Farm [online]. Available at:
http://www.mustfarm.com/progress/site-diary-6-textiles [Accessed 14 August 2020]
Blancquaert, G., Leroyer, C., Lorho, T., Malrain, F. and Zech-Matterne, V. 2012. ‘Rythmes de
créations et d’abandons des établissements ruraux du second âge du Fer et interactions
environmentales’. In: F. Bertoncello and F. Braemer (eds.) Variabilités environnementales,
mutations sociales: nature, intensités, échelles et temporalités des changements. Antibes:
Association pour la Promotion et la Diffusion des Connaissances Archéologiques,
pp.233-245
Bradley, R. 1978. The Prehistoric Settlement of Britain. London: Routledge
Bradley, R., Haselgrove, C., Vander Linder, M., and Webley, L. 2016. The Later Prehistory of
North-West Europe. Oxford: Oxbow Books
Brailsford, J., & Richmond, I. 1952. British Museum Excavation at Hod Hill, Dorset. The British
Museum Quarterly 17 (3), 49-50
Brand, C. 1995. Zur eisenzeitlichen Besiedlung des Dürrnberges bei Hallein. Internationale
Archäologie 19, Espelkamp
Brockhaus, F.A. 1908. Brockhaus' Konversations-Lexikon. Berlin: F.A. Brockhaus
Broholm, H.C. and Hald, M. 1940. Costumes of the Bronze Age in Denmark. Copenhagen: Arnold
Busck
Brooks, R.T. 1974. The Excavations of The Rumps Cliff Castle, St Minver, Cornwall. Cornish
Archaeology 13: 5-50
Broshi, M. and Gophna, R. 1984. ‘The Settlements and Population of Palestine During the Early
Bronze Age II-III.’ Bulletins of the American Schools of Oriental Research 253: 41-53
Brumfiel, E. M., and Earle, T. 1987. 'Specialization, Exchange, and Complex Societies: An
Introduction.' In  E. M. Brumfiel and T. Earle (eds.), Specialization, Exchange, and Complex
Societies, 1–9. Cambridge: Cambridge University Press
Budin, S. L. 2013. Intimate Lives of the Ancient Greeks. Santa Barbara, CA: Praeger
Bulleid, A. and Gray, H.ST.G. 1911. The Glastonbury Lake Village, Vol. I. Taunton Castle
Burke, B. 2005. 'Textile Production at Gordion and the Phrygian Economy.' In L. Kealhofer (ed.), The
Archaeology of Midas and the Phrygians. Recent Work at Gordion, 69–81. Philadelphia:
University of Pennsylvania Press
199
2016. 'Beyond Penelope: Women and the Role of Textiles in the Bronze Age Aegean.' In J. M.
Turfa and S. L. Budin (eds.) Women in Antiquity: Real Women across the Ancient World.
New York: Routledge, pp. 635–646
Chadwick Hawkes, S. 1994. Longbridge Deverill Cow Down, Wiltshire, house 3: a major roundhouse
of the Early Iron Age. Oxford Journal of Archaeology 13: 49-69
Chapman, A. 2006. An Iron Age Enclosure at Site A, Kennel Farm, Basingstoke, Hampshire.
Proceedings of the Hampshire Field Club and Archaeological Society 61: 16-62
Childe, V. G. 1950. The Urban Revolution. Town Planning Review 21: 3–17.
doi:10.3828/tpr.21.1.k853061t614q42qh
Christie, P.M. 1965. Carn Euny Excavations: Interim report of 1964 season. Cornish Archaeology 4:
24-29
1966. Carn Euny: A brief note on the 1965 excavations. Cornish Archaeology 5: 17-20
1968. Carn Euny: Fourth Interim Report on the 1967 Season. Cornish Archaeology 7: 33-37
1976. Carn Euny – Interim Report on the Final Seasons 1970 and 1972. Cornish Archaeology
15: 68-72
Christie, P. M. L., Elsdon, S. M., Dimbleby, G. W., Saville, A., Rees, S., Fowler, E. and Williams, D.
F. 1978. The Excavation of an Iron Age Souterrain and Settlement at Carn Euny, Sancreed,
Cornwall. Proceedings of the Prehistoric Society 44: 309-433
Coles, J.M. 1987. Meare Village East. Exeter: Somerset Levels Papers 13
Coles, J., and Minnett, S. 1995. Industrious and Fairly Civilized: the Glastonbury Lake Village.
Somerset County Council and Glastonbury Antiquarian Society
Collingwood, P. 1982. The Techniques of Tablet Weaving. London: Batsford
Collis, J. 2010. Zentralisierung und Urbanisierung in Europa nördlich der Alpen während der
Eisenzeit. In D. Krauss (ed.) “Fürstensitze” und Zentralorte der frühen Kelten.
Abschlusskolloquium des DFG-Schwerpunktprogramms 1171 in Stuttgart, 12.-15. Oktober
2009. Teil I und II. Forschungen und Berichte zur Vor- und Frühgeschichte in
Baden-Württenburg 120, Stuttgart
2014. Urbanisation in temperate Europe in the Iron Age: Mediterranean influence or
indigenous? In M. Fernández-Götz, H. Wendling, K. and Winger (eds.), Paths to
Complexity: Centralisation and Urbanisation in Iron Age Europe, Oxford: Oxbow Books,
pp. 15–23
Colsen, I. 2013. A Town Unearthed: Folkestone Before 1500. Canterbury Archaeological Trust
Comis, S.Y. 2003. ‘Prehistoric Garments from the Netherlands’. In L. Bender Jørgensen, J.
Banck-Burgess, and A. Rast-Eicher (eds.) Textilien aus Archäologie und Geschichte.
Festschrift Klaus Tidow. Neumünster: Wachholz Verlag, pp.193-204
Costin, C. L. 1991. Craft Specialization: Issues in Defining, Documenting, and Explaining the
Organization of Production. Archaeological Method and Theory 3: 1–56
1998
Cottrill. 1937. Interim Excavation Report. Proceedings of the Prehistoric Society 3: 448-9
1938. Interim Excavation Report. Proceedings of the Prehistoric Society 4: 321
Cunliffe, B.W. 1978. Hengistbury Head. London: Elek Books
200
1984a. Danebury: An Iron age hillfort in Hampshire Vol 1 The excavations 1969-1978: the
excavation. London: CBA Research Report 52a
1984b. Danebury: An Iron age hillfort in Hampshire Vol 2 The excavations 1969-1978: the
finds. London: CBA Research Report 52b
1988. Mount Batten Plymouth: a prehistoric and Roman port. Oxford University Committee for
Archaeology Monographs
1991. Iron Age Communities in Britain, 3rd edn, London: Routledge
1994. 'Danebury, Hampshire'. In A.P. Fitzpatrick and E. Morris (eds.) The Iron Age in Wessex:
Recent Work. Wessex Archaeology, pp. 94-97
1995. Danebury: An Iron Age hillfort in Hampshire, VI: A Hillfort Community in Perspective,
York: Council for British Archaeology (Research Report 102)
2005. Iron Age Communities in Britain (4th edition). London: Routledge
2009. Iron Age Communities in Britain: an account of England, Scotland, and Wales from the
seventh century BC until the Roman conquest. 4th edition. London: Routledge
Cunliffe, B., Payne, A. and Corney, M., 2006. Understanding hillforts: have we progressed? Oxford.
Cunliffe, B and Phillipson, D.W. 1968. 'Excavations at Eldon's Seat, Encombe, Dorset.' Proceedings
of the Prehistoric Society 34: 191-237
Cunliffe, B. and Poole, C. 1984. Danebury: An Iron Age Hillfort in Hampshire. Vol 2 The excavations
1969-1978: the finds. CBA Research Report No. 52
1991. Danebury: An Iron Age Hillfort in Hampshire. Vol 5 The excavations 1979-1988: the
finds. CBA Research Report No. 73
2000a. Introduction: The Danebury Environs Programme: The Prehistory of a Wessex
Landscape. Vol 1. English Heritage and Oxford University Committee for Archaeology
Monograph No. 49
2000b. Woodbury and Stockbridge Down, Stockbridge, Hants., 1989. The Danebury Environs
Programme: The Prehistory of a Wessex Landscape. Vol 2, Part 1. English Heritage and
Oxford University Committee for Archaeology Monograph No. 49
2000c. Bury Hill, Upper Clatford, Hants., 1990 The Danebury Environs Programme: The
Prehistory of a Wessex Landscape. Vol 2, Part 2. English Heritage and Oxford University
Committee for Archaeology Monograph No. 49
2000d. Suddern Farm, Middle Wallop, Hants., 1991 and 1996 The Danebury Environs
Programme: The Prehistory of a Wessex Landscape. Vol. 2, part 3. English Heritage and
Oxford University Committee for Archaeology Monograph No. 49
2000e. New Buildings, Longstock, Hants., 1992 and Fiveways, Longstock, Hants., 1996 The
Danebury Environs Programme: The Prehistory of a Wessex Landscape. Vol 2, Part 4.
English Heritage and Oxford University Committee for Archaeology Monograph No. 49
2000f. Nettlebank Copse, Wherwell, Hants., 1993 The Danebury Environs Programme: The
Prehistory of a Wessex Landscape. Vol 2, Part 5. English Heritage and Oxford University
Committee for Archaeology Monograph No. 49
2000g. Haughton Down, Stockbridge, Hants. 1994. The Danebury Environs Programme: The
Prehistory of a Wessex Landscape. Vol 2, Part 6I. English Heritage and Oxford University
Committee for Archaeology Monograph No. 49
Curwen, E.C. 1927. Excavations in the Caburn, Near Lewes. Sussex Archaeological Collections 68:
1-56
1929. Excavations in the Trundle, Goodwood, 1928. Sussex Archaeological Collections 70
1931. Excavations in the Trundle, Second Season 1930, Sussex Archaeological Collections 72
Cutler, J., Nosch, M.-L.B., and Andersson Strand, E. 2013. ‘Textile Production in Quartier Mu’. In:
J.-C. Poursat (ed.), Fouilles exécutées à Malia. Le Quartier Mu V. Vie quotidienne et
techniques au Minoen Moyen II. Études Crétoises 34. Athènes: École française d'Athènes,
pp. 95–118
201
Cybulska, M. and Maik, J. 2007. ‘Archaeological Textiles – A Need for New Methods of Analysis
and Reconstruction’. Fibres and Textiles in Eastern Europe 15(5-6): 185-189
Dark, P. 2006. Climate deterioration and land-use change in the first millennium BC: perspectives
from the BRitish palynological record. Journal of Archaeological Science 33: 1381-95
David, N. and Kramer, C. 2001. Ethnoarchaeology in Action. The South African Archaeological
Bulletin 57(175): 42
DeRoche, C. D. 1995. Textile production in Britain during the first millennium B.C. (Doctoral thesis).
Cambridge. doi.org/10.17863/CAM.37739
DeRoche, D. 2012. ‘England: pre-Roman period.’ In: Gleba M, Mannering (eds), Textiles and textile
production in Europe from prehistory to AD 400. Ancient Textiles Series, vol 11. Oxbow
Books, Oxford: pp. 444–450
Devon County Council. 2020. Devon’s rocks - a geological guide [online]. Available at:
https://www.devon.gov.uk/geology/devons-rocks-a-geological-guide [Accessed 4 September
2020]
Dimova, B. 2016. Textile production in Iron Age Thrace. European Journal of Archaeology 19(4):
652–680. doi: 10.1080/14619571.2016.1164457
Dobiat, C. 1990. Der Burgstallkogel bei Kleinklein. Marburger Studien zur Vor- und Frühgesch 13.
Marburg
Dudley, D, 1956. An Excavation at Bodrifty, Mulfra, near Penzance. Archaeological Journal 113:
1-32
Ehlers, S. 1998. Bronzezeitliche Textilien aus Schleswig-Holstein. Eine technische Analyse und
Funktionsbestimmung. Unpublished doctoral dissertation. Kiel: Christian-Albrechts-
Universität zu Kiel
English Heritage. 2011. Strategic Stone Study: A building stone atlas of Kent [online]. Available at:
https://www.bgs.ac.uk/mineralsuk/buildingStones/StrategicStoneStudy/EH_atlases.html
[Accessed 4 September 2020]
Farke, H. 1993. ‘Textile Reste von Dietzhausen und Schwarza’. In: R. Feustel (ed.), Zur
bronzezeitlichen Hügelgräberkultur in Südthüringen. Alt-Thüringen 27: pp. 53-123
Fasham, P.J. 1982. The Excavation of Four Ring-Ditches in Central Hampshire. Proceedings of the
Hampshire Field Club and Archaeological Society 38: 19-56
1985. The Prehistoric Settlement at Winnall Down, Winchester: excavations of MARC3 site R17
in 1976 and 1977. Monograph 2
Fasham, P. J., Farwell, D. E., Whinney, R. and et 15 alii. 1989. The archaeological site at Easton Lane,
Winchester. Hampshire Field Club Monograph 6:
Fernández-Götz, M. 2018. ‘Urbanization in Iron Age Europe: Trajectories, Patterns, and Social
Dynamics’. Journal of Archaeological Research 26: 117-162
Fernández-Götz, M., and Krausse, D. 2013. Rethinking Early Iron Age urbanisation in central Europe:
The Heuneburg site and its archaeological environment. Antiquity 87: 473–487
202
Fernández-Götz, M., Wendling, H., and Winger, K. 2014. Paths to Complexity: Centralisation and
Urbanisation in Iron Age Europe. Oxford: Oxbow Books
Ferrero, L.J. 2014. An Experimental Comparison of Impressions Made from Replicated Neolithic
Linen and Bronze Age Woolen Textiles on Pottery. EXARC Online Journal 2014(4)
exarc.net/ark:/88735/10177
Fitzpatrick, A. 1994. ‘Outside in: the structure of an Early Iron Age house at Dunston Park,
Thatcham, Berkshire.’ In: A. Fitzpatrick and E. Morris (eds.) The Iron Age in Wessex:
Recent Work, Salisbury, pp. 68-72
Fitzpatrick, A., Barnes, I., and Cleal, R. 1995. ‘An Early Iron Age settlement at Dunston Park,
Thatcham.’ In I. Barnes, W. Boismier, R. Cleal, A. Fitzpatrick and M. Roberts, Early
Settlement in Berkshire. Salisbury, pp. 65-92
Flohr, M. 2013. The World of the Fullo. Work, economy and society in Roman Italy. Oxford Studies on
the Roman Economy 4. Oxford
Forbes, R.J. 1956. Studies in Ancient Technology. Wash-ing, Bleaching, Fulling and Felting. Dyes and
Dyeing, Spinning, Sewing, Basketry and Weaving and Looms. Fabrics and Weavers. The
fibres and fabrics of Antiquity, Volume IV. Leiden
Forde-Johnston, J. 1976. Hillforts of the Iron Age in England and Wales. Liverpool: Liverpool
University Press
Fossøy, S.H. 2012. Rom for variasjon. En studie i bronsealderens tekstilhåndverk. Unpublished
Master’s thesis. Trondheim: Norwegian University of Technology and Science
Fox, A., Raleigh Radford, C. A., Rogers, E. H., Shorter, A. H., 1949 - 1950. Report on the
Excavations at Milber Down, 1937-8. Proceedings of the Devon Archaeological Exploration
Society 4: 27-65
Frick, F.S. 1985. The Formation of the State in Ancient Israel. Decatur, GA: Almond
Gleba, M. 2007. 'Textile Production in Proto-Historic Italy: From Specialists to Workshops.' In C.
Gillis and M. L. Nosch (eds.), Ancient Textiles: Production, Crafts, and Society, 71–76.
Ancient Textiles Series 1. Oxford: Oxbow Books
2008. Textile Production in Pre‐Roman Italy: Archaeological Evidence. Oxford: Ancient
Textiles Series 4
2016. ‘Etruscan Textiles in Context’. In: A Companion to the Etruscans. Blackwell: Wiley, pp.
237-246
Gleba, M. and Mannering, U. 2012. Textiles and Textile Production in Europe: From Prehistory to AD
400. Oxford: Oxbow
Gleba, M. and Harris, S. 2019. The first plant bast fibre technology: identifying splicing in
archaeological textiles. Archaeological and Anthropological Sciences 11(5): 2326-2346
González, P., C. Masvidal, S. Montón, and M. Picazo, eds. 2007. Interpreting Household Practices:
Reflections on the Social and Cultural Roles of Maintenance Activities. Bellaterra:
Universitat Autònoma de Barcelona
Gostenčnik, K. 2013. ‘Textile Production and Trade in Roman Noricum’. In: M. Gleba and J.
Pásztókai-Szeőke (eds.) Making Textiles in Pre-roman and Roman Times. People, Places,
Identities. Ancient Textiles Series Vol. 13. Oxford: Oxbow Books, pp. 60–86
203
Grant, A. 1984. 'Animal husbandry in Wessex and the Thames valley' In: B. Cunliffe and D. Miles
(eds), Aspects of  the Iron Age in Central Southern Britain. Oxford: Oxford University
Committee for Archaeology Monograph 2, pp. 102–119
Greenfield, H.J. 2014. Animal Secondary Products: Archaeological Perspective on Domesticated
Animal Exploitation in the Neolithic and Bronze Age. Oxbow Books Ltd, Oxford
Grömer, K. 2005. 'The Textiles from the Prehistoric Salt-mines at Hallstatt.' In P. Bichler, K. Grömer,
R. Hofmann-de Keijzer, A. Kern and H. Reschreiter (eds), Hallstatt Textiles – Technical
Analysis, Scientific Investigation and Experiments on Iron Age Textiles. BAR IS-1351.
Oxford: pp.17–40
2006. Von Spinnen und Weben, Flechten und Zwirnen. Hin weise zur neolithischen
Textiltechnik an österreichischen Fundstellen. Festschrift für Elisabeth Ruttkay, Archäologie
Österreichs 17(2): 177-192. Wein
2012. 'Austria: Bronze And Iron Ages.' In M. Gleba and U. Mannering (eds.) Textiles and
Textile Production in Europe: From Prehistory to AD 400. Oxford: Oxbow Books, 27-64
2013. 'Tradition, creativity and innovation - The development of textile expertise from the
Bronze Age to the Hallstatt Period.' In K. Grömer, A. Kern, H. Reschreiter, and H.
Rösel-Mautendorfer (eds.), Textiles from Hallstatt. Weaving Culture in Bronze and Iron Age
Salt Mines. Textilien aus Hallstatt. Gewebte Kultur aus dem bronze- und eisenzeitlichen
Salzbergwerk. Budapest: Archaeolingua, 53-98
2016a. The Art of Prehistoric Textile Making. Veröffentlichungen der Prähistorischen Abteilung
(VPA) 5. Vienna: Natural History Museum Vienna
2016b. ‘Colour, Pattern, and Glamour: Textiles in Central Europe 2000-400 BC’. In: J. Ortiz, C.
Alfaro, L Turell, and M.ªJ. Martínez (eds.) Textiles, Basketry, and Dyes in the Ancient
Mediterranean World: Proceedings of the Vth International Symposium on Textiles and
Dyes in the Ancient Mediterranean World (Montserrat, 19-22 March 2014). València:
Universitat De València, pp. 37-44
2018. ‘Textiles: Pattern, Structure, Texture, and Decoration’ In: L. Bender Jørgensen, J. Sofaer,
M.L. Stig Sørensen, (eds.) Creativity in the Bronze Age: Understanding Innovation in
Pottery, Textile, and Metalwork Production. Cambridge: Cambridge University Press, pp.
247-262
Guthrie, A. 1969. Excavation of a Settlement at Goldherring, Sancreed, 1958-1961. Cornish
Archaeological Society 8: 5-39
Hagstrum, M. 1985. Measuring Prehistoric Craft Specialisation: A Test Case in the American
Southwest. Journal of Field Archaeology 12(1): 65-76
Hald, M. 1980. Ancient Danish Textiles from Bogs and Burials. The National Museum of Denmark
Vol. XI
Halstead, P. 1998. Mortality models and milking: problems of uniformitarianism, optimality and
equifinality reconsidered. Anthropozoologica 27: 3–20
Hammarlund, L. 2004. Handicraft knowledge applied to archaeological textiles - Fabric thickness and
density: A method of grouping textiles. Archaeological Textiles Newsletter 39: 7-11
2005. Handicraft Knowledge Applied to Archaeological Textiles. The Nordic Textile Journal
2005: 86-119
Hampshire County Council. 2012. Hampshire Minerals and Waste Plan. Minerals in Hampshire:
Background Study Version 7 [online]. Available at:
https://documents.hants.gov.uk/mineralsandwaste/HMWP100bMineralsinHampshirev7-Sub
missionRevisedFeb2013.pdf [Accessed 4 September 2020]
204
Hänsel, B. 2005. Wege zur Stadt im vorgeschichtlichen Mitteleuropa. In H. Falk (ed.) Wege zur Stadt.
Entwicklung und Formen urbanen Lebens in der alten Welt. Vergleichende Studien zu
Antike und Orient 2, Brennen, pp. 16-195
Harding, D.W., Blake, I., and Reynolds, P. 1993. An Iron Age Settlement in Dorset: Excavation and
Reconstruction. Edinburgh
Harding, D.W. 2012. Iron Age Hillforts in Britain and Beyond. 4th edition. Oxford: Oxford University
Press
2014. The Iron Age in Lowland Britain. London: Routledge
Harris, S.; Rösel-Mautendorfer, H.; Grömer, K.; and Reschreiter, H. 2008. Cloth Cultures in
Prehistoric Europe: the Bronze Age evidence from Hallstatt. Archaeology International 12:
22-26
Haselgrove, C.C. 1982. ‘Wealth, prestige and power: the dynamics of late Iron Age political
centralization in south-east England.’ In C. Renfrew and S. Shennan, (eds.), Ranking,
Resource and Exchange. Cambridge: Cambridge University Press: pp. 79–88
Haskell, H. 1997. 'Trade and Production in 'Mycenaean' Crete'. In C. Gillis, C. Risberg, and B Sjöberg
(eds.) Trade and Production in Promentory Greece, Proceedings of the 4th and 5th
International Workshops, Athens 1994 and 1995. Stockholm: Paul Äströms Förlag.
p.101-111
Henderson, J. 2007. The Atlantic Iron Age: Settlement and Identity in the First Millennium BC.
London: Routledge
Hendon, J. A. 1996. Archaeological Approaches to the Organization of Domestic Labor: Household
Practice and Domestic Relations. Annual Review of Anthropology 25: 45–61.
doi:10.1146/annurev.anthro.25.1.45
Herring, P. 1994. The cliff castles and hillforts of West Penwith in the light of recent work at Maen
Castle and Treryn Dinas. Cornish Archaeology 33: 40-56
Hill, J.D. 2011. 'How Did British Middle and Late Pre-Roman Iron Age Societies Work (if they did)?'
in T.H. Moore and X.L. Armada (eds.) Atlantic Europe in the First Millennium BC:
Crossing the Divide, Oxford: Oxford University Press, 242-268
DOI:10.1093/acprof:osobl/9780199567959.003.0010
Hirth, K. G. 2009. 'Housework and Domestic Craft Production: An Introduction.' In K. G. Hirth (ed.),
Housework: Craft Production and Domestic Economy in Ancient Mesoamerica, 1–12.
Hoboken: Wiley
Hirst, F.C. 1936. Excavations at Porthmeor, Cornwall, 1933, 1934 and 1935. Journal of the Royal
Institute of Cornwall 24: appendix ii, 1–81
Hodder, I. and Hedges, J.W. 1977. ‘Weaving combs: their typology and distribution with some
introductory remarks on date and function’. In J.R. Collis (ed.), The Iron Age in Britain: A
Review. Sheffield: Dept.of Prehistory and Archaeology, pp.17-28
Hoffmann, M. 1974. The Warp Weighted Loom. Oslo
Hofmann-de Keijzer, R.; van Bommel, M.R.; Hartl, A.; Grömer, K.; Rösel-Mautendorfer, H.;
Reschreiter, H.; Kania, K.; Joosten, I.; Ness Proaño Gaibor, A.; Erlach, R.; Lachner, E.;
205
Wandl, M.; de Keijzer, M. 2013. Coloured Hallstatt Textiles: 3500 Year-old Textile and
Dyeing Techniques and their Contemporary Application. In: J. Banck-Burgess and C.
Nübold (eds.) NESAT XI; The North European Symposium for Archaeological Textiles XI;
10-13 May 2011 in Esslingen am Neckar. Leidorf: Verlag Marie Leidorf GmbH, pp. 125-130
Hofmann de Keijzer, R., van Bommel, M. R., and Joosten, I. 2005. Dyestuff and Element Analysis on
Textiles from the Prehistoric Salt-mines of Hallstatt. In P. Bichler, K. Grömer, R.
Hofmann-de Keijzer, A. Kern and H. Reschreiter (eds), Hallstatt Textiles – Technical
Analysis, Scientific Investigation and Experiments on Iron Age Textiles. BAR IS-1351.
Oxford: 55–72
Holman, D. 2005. Iron Age Coinage and Settlement in East Kent. Britannia 36: 1–54.
doi:10.3815/000000005784016892
Hopkins, D.C. 1985. The Highlands of Canaan. Decatur, GA: Almond
Hruby, J. 2013. Crafts, Specialists, and Markets in Mycenaean Greece. The Palace of Nestor, Craft
Production, and Mechanisms for the Transfer of Goods. American Journal of Archaeology
117(3): 423-427
Jacoby, D. 2004. Silk economics and cross-cultural artistic interaction: Byzantium, the Muslim world,
and the Christian west. Dumbarton Oaks Papers 58: 197–240
Johnson, N. and Rose, P. 1982. ‘Defended Settlements in Cornwall - an illustrated discussion.’ In: D.
Miles (ed.), The Romano-British Countryside. Oxford: BAR Brit. Ser. 103: 151-207
Keller, T. 1988. The Evidence for Ancient Quern Production at Folkestone. Kent Archeological
Review 93: 59-68
Kemery, B. 2010. Yurtstory: the history of yurst ancient and modern [online]. Available at:
https://www.yurtinfo.org/yurtstory-the-history-of-yurts-ancient-and-modern [Accessed 4
August 2020]
Kneisel, J. and Schaefer-Di Maida, S. 2019. ‘Four - Loom Weights in Bronze Age Central Europe’.
In: S. Sabatini and S. Bergerbrant (eds.) The Textile Revolution in Bronze Age Europe:
Production, Specialisation, Consumption. Cambridge: Cambridge University Press, pp.
80-116
DOI: https://doi.org/10.1017/9781108656405.004
Knight, M., Ballantyne, R., Robinson Zeki, I., and Gibson, D. 2019. The Must Farm pile-dwelling
settlement. Antiquity 93 (369): 645-663
Kolb, F. 1984. Die Stadt im Altertum. Beck, Munich
Kuijpers. M. 2013. ‘The sound of fire, taste of copper, feel of bronze, and colours of the cast: sensory
aspects of metalworking technology’. In M.-L. S. Sørensen & K. Rebay-Salisbury (eds.),
Embodied knowledge: Historical perspectives on belief and technology. Oxford: Oxbow
Books: pp. 137–150
2014. PhD. Early Bronze Age Metalworking Craftsmanship. Cambridge: University of
Cambridge
Larsson, M. 2013. Cultivation and Processing of Linum Usitatissimum and Camelina Sativa in
southern Scandinavia during the Roman Iron Age. Vegetation History and Archaeobotany
22(6): 509-520
206
Lauermann, E. 2000. Archäologische Forschungen in Michelstetten, NÖ. Archäologie  Österreichs
11(1): 5–30
Laws, K. 1991. ‘The chalk’ In: N.M. Sharples (ed.), Maiden Castle: Excavations and field survey
1985-6. English Heritage: Archaeological Report no 19, pp. 210-214
Le Boutillier, N. 2004. The Baseline Lithogeochemistry of the Metasedimentary Rocks of Cornwall.
Cornwall: Camborne School of Mines.
DOI: 10.13140/RG.2.2.13636.22409
Leuzinger, U. and Rast-Eicher, A. 2011. Flax processing in the Neolithic and Bronze Age
pile-dwelling settlements of eastern Switzerland. Vegetation History and Archaeobotany
20(6): 535-542
Maier, U. and Schlichtherle, H. 2011. Flax cultivation and textile production in Neolithic wetland
settlements on Lake Constance and in Upper Swabia (south-west Germany). Vegetation
History and Archaeobotany 20(6): 567-578
Mannering, U., Possnert, G., Heinemeier, J., and Gleba, M. 2010. Dating Danish textiles and skins
from bog finds by means of 14C AMS. Journal of Archaeological Science 37(2): 261-268
Mannering, U., Gleba, M. and Bloch Hansen, M. 2012. ‘Denmark’. In: M. Gleba and U. Mannering
(eds.) Textiles and Textile Production in Europe from Prehistory to AD 400. Ancient Textiles
Series Vol. 11. Oxford: Oxbow Books: pp. 91–121
Marchant, T. 1989. The evidence for textile production in the Iron Age. Scottish Archaeological
Review 6: 5-11
Marín-Aguilera, B. 2019. Weaving Rural Economies: Textile Production and Societal Complexity in
Iron Age Southwestern Iberia. World Archaeology 51(2): 226-251
Marín-Aguilera, B., Rodríguez-González, E., Celestino, S., & Gleba, M. 2019. Dressing the Sacrifice:
textiles, textile production and the sacrificial economy at Casas del Turuñuelo in
fifth-century BC Iberia. Antiquity 42: 933-953. doi.org/10.15184/aqy.2019.42
Mårtensson, L., Andersson, E., Nosch, M.-L. and Batzer, A. 2006a. Technical Report: Experimental
Archaeology Part 1, 2005-2006. Tools and Textiles - Texts and Contexts Research
Programme. Copenhagen: The Danish National Research Foundation's Centre for Textile
Research (CTR)
2006b. Technical Report: Experimental Archaeology Part 2:2, Whorl or Bead? 2006. Tools and
Textiles - Texts and Contexts Research Programme. Copenhagen: The Danish National
Research Foundation's Centre for Textile Research (CTR)
2007. Technical Report. Experimental Archaeology. Part 3 Loom weights. 2007. Tools and
Textiles - Texts and Contexts Research Programme. Copenhagen: The Danish National
Research Foundation's Centre for Textile Research (CTR)
Mårtensson, L., Nosch, M-L., Andersson Strand, E. 2009. The Shape of Things: Understanding a
loom weight. Oxford Journal of Archaeology 28 (4): 373-389
Mayoral, V., S. Celestino, and S. Walid Sbeinati. 2011. 'Intensive Survey and Protohistoric Settlement
in the Middle Guadiana Basin.' In M. van Leusen, G. Pizziolo, and L. Sarti (eds.), Hidden
207
Landscapes of Mediterranean Europe: Cultural and Methodological Biases in Pre- and
Protohistoric Landscape Studies. Oxford: Archaeopress: pp. 27–34
Médard, F. 2010. L’art du tissage au Néolithique. IVe-IIIe millénaires avant J.-C. en Suisse. Paris:
CNRS Èditions
Michel, C. and Nosch, M.L. (eds). 2010. Textile Terminologies in the Ancient Near East and
Mediterranean from the Third to the First Millennia BC. Ancient Textiles Series, vol 8.
Oxford: Oxbow Books Ltd
Müller, R. 2012. Késő bronzkori, aranyszalaggal díszített textíliák (Spätbronzezeitliche, mit
Goldbändern verzierte Textilien). Archeometriai Műhely 2012/2: 117-22
Müller-Karpe, H. 1980. Handbuch der Vorgeschichte, Band IV, 1– 3 Bronzezeit. Munich: Verlag
National Geographic. 2020. Yurt [online]. Available at
https://www.nationalgeographic.org/encyclopedia/yurt [Accessed 4 August 2020]
Nørgård, A. 1999. Weaving samples of sailcloth on a warp-weighted loom. Experiments carried out at
the Viking Ship Museum in Roskilde, Denmark in 1999. Roskilde: Vikingeskibsmuseet
Nowakowski, J.A. 2004. Revisiting Trevelgue Head -- sixty years on. Cornish Archaeology 39-40:
190-191
Nowakowski, J.A., Quinnell, H. 2011. Trevelgue Head, Cornwall : the importance of CK Croft
Andrew's 1939 excavations for prehistoric and Roman Cornwall. Truro: Cornwall Council
Osborne, R. 2005. Urban sprawl: what is urbanisation and why does it matter? In: R. Osborne and B.
Cunliffe (eds.) Mediterranean Urbanisation 800-600 BC. Oxford: Oxford University Press,
pp. 1-16
Oswold, A., Ainsworth, S., and Pearson, T. 2006. Hillforts: Prehistoric Strongholds of
Northumberland National Park. London: Historic England
Pattison, P. and Fletcher, M. 1996. Grimspound, One Hundred Years On. Proc Devon Arch Soc 52:
21–34
Payne, A., Cunliffe, B., and Corney, M. 2006. The Wessex Hillforts Project: Extensive Survey of
Hillfort Interiors in Central Southern England. London: English Heritage
Peacock, D.P.S. 1982. Pottery in the Roman World: A Ethnoarchaeological Approach. London and
New York: Longman
Pengelly, W. 1869. The Literature of Kent’s Cavern, Part II. Reports and Transactions of the
Devonshire Association 3: 191-482
1878. The Literature of Kent’s Cavern, Part IV. Rep. Trans. Dev. Ass. 10: 141-181
1884. The Literature of Kent’s Cavern, Part V. Rep. Trans. Dev. Ass. 14: 189-343
Peyronel, L. 2004. Gli strumenti di tessitura dall’età del Bronzo all’epoca persiana. Materiali e Studi
Archeologici di Ebla IV. Roma
Pieroni, A. 2005. Gathering Food from the Wild. In G. Prance and M. Nesbitt (eds.) The Cultural
History of Plants. London: Routledge, pp. 27-42
208
Poole, C. 1984. ‘The structural use of daub, clay, and timber.’ In: B. Cunliffe (ed.), Danebury: An Iron
Age Hillfort in Hampshire. Vol. 1: The excavations 1969-1978: the site. Council of British
Archaeology Research Report 52: 110-123
1991a. ‘Objects of baked clay.’ In: B. Cunliffe (ed.), Danebury: An Iron Age Hillfort in
Hampshire Vol 5 The excavations 1979-1988: the finds. CBA Research Report No. 73:
370-382
1991b. ‘The small objects of daub and clay’ In: N.M. Sharples (ed.), Maiden Castle:
Excavations and field survey 1985-6. English Heritage: Archaeological Report no 19, pp.
209-210
Pope, R. E. (2007). ‘Ritual and the roundhouse: a critique of recent ideas on domestic space in later
British prehistory’. In C. C. Haselgrove, & R. E. Pope (Eds.), The Earlier Iron Age in
Britain and the near Continent. Oxford: Oxbow, pp. 204-228
Quinnell, H. 1986. Cornwall during the Iron Age and the Roman period. Cornish Archaeology 25:
111–134
Rady, J. 2010. Excavations at Thanet Earth 2007-2008 Assessment Report Volume 1. Canterbury
Archaeological Trust
Rast-Eicher, A. 2008. Textilien, Wolle, Schafe der Eisenzeit in der Schweiz. Antiqua 44, Basel:
Archäologie Schweiz
2012. ‘Ornamente auf Textilien – Textile Ornamente’. In A. Boschetti-Maradi, A. de Capitani,
S. Hochuli and U. Niffeler (eds.) Form, Zeit und Raum. Grundlagen für eine Geschichte aus
dem Boden. Festschrift für Werner E. Stöckli zu seinem 65. Geburtstag. Antiqua 50. Basel:
Archäologie Schweiz, pp. 225-34
Rast-Eicher, A. and Reinhard, J. 1998. ‘Gewebe und Geflechte’. In: U.S. Hochuli, U. Niffeler and V.
Rychner (eds.) Die Schweiz vom Paläolithikum bis ins Mittelalter, Die Bronzezeit. Basel:
Archäologie Schweiz, pp. 285-90
Rawlins, M., Chadwick, A. M., Reynier, M. J. and Seager Smith, R. H. 2003. Prehistoric and Roman
activity at Zionhill Farm, Chandlers Ford. Proceedings of the Hampshire Field Club &
Archaeological Society 58: 1 - 23
Reese, D.S. 1987. Palaikastro Shells and Bronze Age Purple-Dye Production in the Mediterranean
Basin. Annual of the British School of Archaeology at Athens 82: 201–206
Reurings, S. 2019. ‘Farming in West-Frisia during the Late Neolithic’. The Later European Prehistory
Group’s Third Annual Symposium. University of Cambridge, 4-5 May. Cambridge:
McDonald Institute for Archaeological Research
Reuther, W. and Webber, H.J. 1967. The Citrus Industry. Vol 1. Berkeley: University of California,
Division of Agricultural Sciences
Rice, P.M. 1981. Evolution of Specialised Pottery Production: A Trial Model. Current Anthropology
22(3): 219-240
1984. Pottery Analysis: A Sourcebook. Chicago: University of Chicago Press
Richmond, I. 1968. Hod Hill, Vol 1-2. London: Trustees of the British Museum
Richter, É. 2010. ‘Our threads to the Past: Plaited Motifs as Predecessors of Woven Binding
Structures’. In: E. Andersson Strand, M. Gleba, U. Man-nering, C. Munkholt and M.
Ringgaard (eds.): North European Sympo-sium for Archaeological Textiles X. Oxford:
Oxbow Books, pp. 211–216
209
Rodríguez, A., I. Pavón, and D. Duque. 2004. '‘La Mata’ y su Territorio.' In A. Rodríguez (ed.), El
edificio protohistórico de 'La Mata' (Campanario, Badajoz) y su estudio territorial. Cáceres:
Universidad de Extremadura: pp. 497–571
Ryder, M.L. 1968. The Origin of Spinning. Textile History 1: 73-82
1983. Sheep and Man. London: Duckworth
Sabatini, S., Bergerbrant, S., Brandt, L.Ø., Margaryan, A., and Allentoft, M.E. 2019. Approaching
sheep herds origins and the emergence of the wool economy in continental Europe during
the Bronze Age. Archaeological and Anthropological Sciences 11: 4909–4925
https://doi.org/10.1007/s12520-019-00856-x
Sahlins, M.D. 1972. Stone Age Economics. Chicago: Aldine
Scrivener, R.C.; Highley, D.E.; Cameron, D.G.; Linley, K.A.; and White, R. 1997. Mineral Resource
Information for Development Plans Phase One Cornwall: Resources and constraints. British
Geological Survey Technical Report WF/97/11
Schoep, I. 2010. 'Making Elites: Political Economy and Elite culture(s) in Middle Minoan Crete'. In
D.J. Pullen (ed.) Political Economies of the Aegean Bronze Age, Papers from the Langford
Conference, Florida State University, Tallahassee, 22-24 February 2007. Oxford: Oxbow
Books.
Schierer, I. 1987. Ein Webstuhlbefund aus Gars-Thunau. Rekonstruk-tionsversuch und
Funktionsanalyse. Archaeologia Austriaca 71: 29-88
Sellwood, L. 1984. 'Objects of bone and antler.' In B.W. Cunliffe (ed.) Danebury: An Iron Age Hillfort
in Hampshire, Vol 2. The excavations 1969-78: The finds. London: CBA Res Rep 52,
pp.371-395
Shaffrey, R. 2017. 'A re-investigation of British stone loomweights.' In R. Shaffrey (ed.), Written in
Stone: Papers on the Function, Form, and Provenancing of Prehistoric Stone Objects in
Memory of Fiona Roe. Highfield Press: St Andrews. 229-248
Sharples, N.M. 1991. Maiden Castle: Excavations and field survey 1985-6. English Heritage:
Archaeological Report no 19
1994. 'Maiden Castle, Dorset'. In A.P. Fitzpatrick and E. Morris (eds.) The Iron Age in Wessex:
Recent Work. Wessex Archaeology, pp. 92-94
2005. ‘Life histories and the buildings of the Atlantic Iron Age.’ In: V. Turner, R.A. Nicholson,
S.J. Dockrill and J.M. Bond (eds.), Tall Stories: 2 Millennia of Brochs. Lerwick: Shetland
Amenity Trust, pp. 102-114
2007. Building communities and creating identities in the first millennium BC. In: C.
Haselgrove and R. Pope (eds.). The Earlier Iron Age in Britain and the Near Continent,
Oxford: Oxbow Books, pp. 174-184
Sharp, A. 1992. Treryn Dinas: cliff castles reconsidered. Cornish Archaeology 31: 65-68
Silvester, R. 1979. The relationship of first millenium settlement to the upland areas of the south-west.
Proc Devon Archaeol Soc 37: 176-190
1980. The prehistoric open settlement at Dainton, South Devon. Proceedings of the Dorset
Archaeological Exploration Society 38: 17-48
1986. The later prehistoric and Roman material from Kent’s cavern, Torquay. Proc Devon
Archaeol Soc 44: 9-38
210
Simmons, I.G. 1970. Environment and Early Man on Dartmoor. Proceedings of the Prehistoric
Society 35: 203-219
Singer, Ch.; Holmyard, E.J.; Hall, A.R.; and Williams, T.I. 1962. Storia della tecnologia. Vol. II. Le
civiltà mediterranee e il Medioevo.Torino
Smith, G. 1996. Archaeology and environment of a Bronze Age cairn and prehistoric and
Romano-British field system at Chysauster, Gulval, near Penzance, Cornwall. Proc. Preh.
Soc. 62: 167-220
Smith, M.E. 2007. Form and Meaning in the Earliest Cities: A New Approach to Ancient Urban
Planning. Journal of Planning History 6(1): 3-47
2010. Sprawl, squatters and sustainable cities: Can archaeological data shed light on modern
urban issues? Cambridge Archaeological Journal 20(2): 229–253
Soriga, E. 2017. ‘A Diachronic View on Fulling Technology in the Mediterranean and the Ancient
Near East: Tools, Raw Materials and Natural Resources for the Finishing of Textiles.’ In: S.
Gaspa, C. Michel, and M.-L. Nosch (eds.) Textile Terminologies from the Orient to the
Mediterranean and Europe, 1000 BC to 1000 AD. Lincoln, NE: Zea Books, pp. 24-46
Sørensen, M.L.S. 2007. ‘English and Danish Iron Ages – a Comparison through Houses, Burials and
Hoards’. In C. Haselgrove and R. Pope (eds) The Earlier Iron Age in Britain and the Near
Continent. Oxford: Oxbow, pp. 328-337
Stoddart, S. 2016. ‘Etruscan Italy: Physical Geography and Environment’. In: A Companion to the
Etruscans. Blackwell: Wiley, pp. 43-54
Stöllner, T. 2002. Die  Hallstattzeit und  der Beginn der Latènezeit  im  Inn-Salzach-Raum,
Auswertung. Archäologie in Salzburg 3/I. Salzburg
2005. 'More than old rags – Textiles from the Iron Age Salt-mine at the Dürrnberg.' In P.
Bichler, K. Grömer, R. Hofmann-de Keijzer, A. Kern and H. Reschreiter (eds), Hallstatt
Textiles Technical Analysis, Scientific Investigation and Experiment on Iron Age Textiles.
Oxford: BAR Int. Series 1351, 161–174
Stone, E.C. 1987. Nippur Neighbourhoods. Studies in Ancient Oriental Civilisation 44. Chicago:
Oriental Institute of the University of Chicago
Talaa, D. 1991. Urgeschichtliche Funde aus Vösendorf. Vösendorf
Tatje, T.A. and Naroll, R. 1973. 'Two Measures of Societal Complexity: An Empirical Cross-Cultural
Comparison.' In R. Naroll and R. Cohen (eds.), A Handbook of Method in Cultural
Anthropology. New York: Columbia University Press, pp.766-833.
Thomas, J. 2008. Dorset Stone. Stanbridge: Dovecote Press
Threipland, L.M. 1957. An Excavation in St Magwan-in-Pyder, North Cornwall. Archaeological
Journal 113: 33-81
Tuck, A.; Bauer, J.; Huntsman, T.; Kreindler, K.; Pancaldo, S.; Powell, C.;  and Miller, S. 2009.
Core and Periphery in Inland Etruria: Poggio Civitate and the Etruscan Settlement in
Vescovado di Murlo. Etruscan and Italic Studies 12 (1): 215-240
DOI: https://doi.org/10.1515/etst.2009.12.1.215
Tuohy, T. 1999. Prehistoric Combs of Antler and Bone, Vol. I. Oxford: BAR British Series 285
211
Tykot, R.H., Iovino, M.R., Martinelli, M.C., and Beyer, L. 2006. Ossidiana di Lipari: le fonti, la
distribuzione, la tipologia e le tracce d’usura. Atti del XXXIX Riunione Scientifica
dell’Istituto Italiano di Preistoria e Protostoria: Materie prime e scambi nella preistoria
italiana, Firenze, 25-27 November 2004. Firenze: pp. 592-597
van der Leeuw, S. 1977. 'Towards a Study in the Economics of Pottery Making.' In B.L. Beek, R.W.
Brant, and W. Gruenman van Watteringe (eds), Ex Horreo. Cingvla, 4. Amsterdam: Albert
Egges van Griffen Instituut voor Prae- en Protohistorie, University of Amsterdam, pp. 68-76
1984. ‘Dust to dust: A transformational view of the ceramic cycle.’ In S. Van der Leeuw &
A.C. Pritchard (eds.), The many dimensions of pottery: ceramics in archaeology and
anthropology. Amsterdam: Universiteit van Amsterdam, pp. 705-773
van Strydonck and Grömer, K. 2013. 'Summary.' In K. Grömer, A. Kern, H. Reschreiter, and H.
Rösel-Mautendorfer (eds.), Textiles from Hallstatt. Weaving Culture in Bronze and Iron Age
Salt Mines. Textilien aus Hallstatt. Gewebte Kultur aus dem bronze- und eisenzeitlichen
Salzbergwerk. Budapest: Archaeolingua: pp.
Vogelsang-Eastwood, G. 1999. Tutankhamun’s Wardrobe. Garments from the Tomb of Tutankhamun.
Rotterdam: Barjesteh van Waalwijk van Doorn
von Kurzynski, K. 1996. '... und ihre Hosen nennen sie bracas'. Textilfunde und Textiltechnologie der
Hallstatt- und Latènzeit und ihr Kontext. Internationale Archäologie 22, Leidorf
Wainwright, G.C. 1969. Mount Pleasant a henge near Dorchester, Dorset. Current Archaeology 2:
320-324
1979. Gussage All Saints: An Iron Age Settlement in Dorset. Department of the Environment:
Archaeological Reports No 10
Webley, L. 2007. Using and Abandoning Roundhouses: A Reinterpretation of the evidence from Late
Bronze Age–Early Iron Age Southern England. Oxford Journal of Archaeology 26(2):
127-144
2018. ‘Households and Communities.’ In: C. Haselgrove, P. S. Wells, and K. Rebay-Salisbury
(eds) The Oxford Handbook of the European Iron Age. Oxford: Oxford University Press, pp.
Wheeler, R.E.M. 1943. Maiden Castle, Dorset. London: Society of Antiquaries of London
Wild, J.P. 1967. Soft-Finished Textiles in Roman Britain. The Classical Quarterly 17(1): 133-135
Williams, A. 1950. Excavations at Allards Quarry, Marnhull, Dorset. Proceedings of the Dorset
Natural History and Archaeological Society 72: 20-75
Willis, L., Rogers, E.H. 1951. Dainton Earthworks. Proceedings of the Dorset Archaeological
Exploration Society 4(4): 79-101
Wilson, A. 1938. Excavation at the Caburn. Sussex Archaeological Collections Vol. 80
Wood, J. 2001. A new perspective on west Cornwall courtyard houses. Cornish Archaeology 36:
95-106
Young, A., Richardson, K. 1954. Report on the excavations at Blackbury Castle. Proceedings of the
Devon Archaeological Exploration Society 5 (2-3): 43-67
Zawadzki, S. 2013. Garments of the Gods. Studies on the Textile Industry and the Pantheon of Sippar
according to the Texts from the Ebabbar Archive. Vol. 2: Texts. Orbis Biblicus et Orientalis
260. Fribourg/Göttingen
212
Zorn, J.R. 1994. ‘Estimating the Population Size of Ancient Settlements: Methods, Problems,
Solutions, and a Case Study’ Bulletin of the American Schools of Oriental Research 295:
31-48
213
Appendix 1. South West sites in this study
Cornwall
Bodrifty
Type Enclosed settlement Date of occupation 400 - 50 BC
Grid Reference SW 44436 35570 Date of excavation 1950-1954
Features Stone enclosure wall; 9
roundhouse structures.
Associated activities Domestic occupation;
spinning; food processing;
animal husbandry.
Notes This site began as an open settlement in the middle of the Penwith peninsula and
possibly dates back to the Bronze Age; 5 structures are larger houses and 4 are
smaller. The smaller roundhouses lack hearths, postholes, or gullies and their
entrances are narrow and face southwest. These structures appear late in the
settlement phase, suggesting a change in regional climate (Dudley 1956). The
earlier, larger structures have wider entrances facing southeast and thicker walls.
Finds include querns, locally made pottery, whetstones, and worked stone; there is
little to no evidence of long-distance trade or metal-working on site.
Publications Dudley, D, 1956, An Excavation at Bodrifty, Mulfra, near Penzance.
Archaeological Journal 113: 1-32
Carn Euny
Type Open settlement Date of occupation 400 BC - AD 300
Grid Reference SW 40250 28839 Date of excavation 1962-1974
Features Roundhouses; fogou; later
courtyard houses.
Associated activities Domestic occupation;
spinning; food
preparation; weapon
sharpening; animal
husbandry.
Notes The site is located on the Penwith peninsula near the hillfort of Carn Bran; it
consists of Iron Age roundhouses, a fogou, and later courtyard houses.
Excavations between 1964 and 1972 covered the fogou and several houses on the
site, focusing on investigating and restoring the fogou and circular underground
chamber. These works also revealed 9 hut foundations. Finds are scarce due to the
acidic soil conditions and are largely limited to locally made pottery and
quernstones. This makes it difficult to judge any trading links this settlement may
have had.
Publications Christie, P.M. 1965. Carn Euny Excavations: Interim report of 1964 season.
Cornish Archaeology 4: 24-29
Ibid. 1966. Carn Euny: A brief note on the 1965 excavations. Cornish Archaeology
5: 17-20
Ibid. 1968. Carn Euny: Fourth Interim Report on the 1967 Season. Cornish
Archaeology 7: 33-37
Ibid. 1976. Carn Euny – Interim Report on the Final Seasons 1970 and 1972.
Cornish Archaeology 15: 68-72
214
The Rumps
Type Promontory Fort Date of occupation 200 BC - AD 100
Grid Reference SW 93411 81092 Date of excavation 1965-1967
Features 3 ramparts with ditches
across the narrowest point
of the promenantory;
gatehouse remains;
postholes; 6 hut platforms.
Associated activities Food preparation; spinning
and weaving; pottery
making; use of metal but
no metalworking.
Notes Located on a promenatory along Cornwall’s north coast with 3 lines of defensive
ditch and banks separating it from the mainland. Hut platforms cluster near the
entrance, suggesting that the rear of the castle was used for arable or pastoral
purposes. Surrounding environment used to supply shellfish. Spinning and
(possibly) weaving well represented in excavation finds; sheep are the most
represented animal in the record, suggesting a textile economy of sorts. Pottery
mimics forms and decorations found in Brittany, although no other material culture
from that region has been discovered, implying close trading connections with the
continent but not a migration to Cornwall.
Publications Brooks, R.T. 1974. The Excavations of The Rumps Promontory Fort, St Minver,
Cornwall. Cornish Archaeology 13: 5-50
Trevelgue Head
Type Promontory Fort Date of occupation 500 - 50 BC
Grid Reference SW 82572 63040 Date of excavation 1939
Features Multiple ditch and bank
defences; two Bronze Age
barrows; hut foundations;
post holes; pits; furnaces;
ovens; hearths; middens.
Associated activities Copper and iron smelting
and working; spinning and
weaving.
Notes Located on a spur of land, this site is defended by several lines of ditch-and-bank
earthwork defences. On the innermost enclosure, roundhouse hut platforms were
discovered, most dating to the 2nd century BC. One roundhouse measured 14.5 m
in diameter and is the largest found in Cornwall; it is believed to have been a
meeting place rather than a dwelling. Significant metalworking features and debris
were discovered in one area of the site, indicating iron smelting was a major craft
activity on this site with some copper smelting performed as well. Pottery found
on site indicates a preference for local ceramic production and trade, suggesting
only metal was traded elsewhere.
Publications Nowakowski, J.A. et al., 2011. Trevelgue Head, Cornwall : the importance of CK
Croft Andrew's 1939 excavations for prehistoric and Roman Cornwall. Truro:
Cornwall Council
215
Trevisker
Type Round Date of occupation 200 BC - AD 100
Grid Reference SW 88800 68700 Date of excavation 1955-56
Features Previous Bronze Age huts
and activity; 3 Iron Age
huts; pits; ditches; ovens;
gullies.
Associated activities Possible iron smelting;
spinning; food processing
and cooking.
Notes Located on gently sloping ground between two river systems along the north
Cornish coast. Smaller, inner enclosure built during the Iron Age, inside a larger
Bronze Age enclosure. Houses 1-3 dated to the Iron Age, although only House 2
produced a spindle whorl in its occupation layer. House 3 contained several ovens
and hearths, possibly demonstrating an area for cereal processing. Little evidence
of grain storage available and an iron smelting/working area is near the site,
although the exact location is unknown.
Publications ApSimon, A.M. and Greenfield, E. 1972. The Excavation of Bronze Age and Iron
Age Settlements at Trevisker, St. Eval, Cornwall. Proceedings of the Prehistoric
Society 38: 302-381
Devon
Blackbury Castle
Type Hillfort Date of occupation 300 BC - AD 100
Grid Reference SY 18700 92400 Date of excavation 1952-1954
Features Single ditch and bank
defences, complex
entrance, post holes,
gullies, roundhouse, oven.
Associated activities Possible iron smelting;
spinning; food processing
Notes Previously known as Blackbury Castle. Located on a ridge, with univaliate
defences but a developed entranceway (described as a’barbrican’) with gate-posts.
Limited excavation within the enclosure revealed the remains of one roundhouse,
along with an oven, a cooking pit, and several pits and gullies. Pottery makes up
most of the finds (due to the acidic soil); the earliest pottery is similar to the Iron
Age B ceramics of Dorsert, Sussex, Hampshire, and Wiltshire and appears to have
influenced the later pottery, known as Iron Age A in the report. Metalworking is
evidenced by several fragments of bronze, 1 of iron, and 4 whetstones. The only
other craft activity is demonstrated by 2 spindle whorls and 4 loom weight
fragments.
Publications Young, A., Richardson, K. 1954. Report on the excavations at Blackbury Castle.
Proceedings of the Devon Archaeological Exploration Society 5 (2-3): 43-67
216
Dainton
Type Unenclosed settlement Date of occupation 400 BC - 50 BC
Grid Reference SX 85900 66800 Date of excavation 1939; 1949
Features Ditch and banks, eroded
stone roundhouses,
postholes, possible field
system, gullies.
Associated activities Spinning, pottery
production, food
processing, possible iron
smelting.
Notes Settlement lies on Dainton Common and includes a possible field system, several
hut platforms, and a probable midden in the form of a concentric bank. This bank
contained spindle whorls, pottery, querns, ox and pig teeth, and charcoal
fragments. Crucible fragments discovered in 1 pit, but nowhere else on the site.
Publications Willis, L., Rogers, E.H. 1951. Dainton Earthworks. Proceedings of the Dorset
Archaeological Exploration Society 4(4): 79-101
Silvester, R. 1980. The prehistoric open settlement at Dainton, South Devon.
Proceedings of the Dorset Archaeological Exploration Society 38: 17-48
Kents Cavern
Type Cave Date of occupation 700 BC - AD 43
Grid Reference SX 93452 64171 Date of excavation 1856-1880
Features No discernable features. Associated activities Wool processing;
spinning; food
consumption; food
processing.
Notes Site of human activity lies in only the largest and most accessible parts of the cave
system and was defined by the ‘black mold’ deposit that covered this area. Quern
stones, cooking pottery, and small areas of burning (possible cooking fires) found
on the site, but no archaeological features that indicated long-term occupation.
Weaving combs and high numbers of spindle whorls but no loom weights suggests
the processing of fibres (possibly animal wool) for transportation and weaving
elsewhere.
Publications Pengelly, W. 1869. The Literature of Kent’s Cavern, Part II. Reports and
Transactions of the Devonshire Association 3: 191-482
Pengelly, W. 1878. The Literature of Kent’s Cavern, Part IV. Rep. Trans. Dev. Ass.
10: 141-181
Pengelly, W. 1884. The Literature of Kent’s Cavern, Part V. Rep. Trans. Dev. Ass.
14: 189-343
Silvester, R. 1986. The later prehistoric and Roman material from Kent’s cavern,
Torquay. Proceedings of the Devon Archaeological Society 44: 9-38
Bradley, R. 1978. The Prehistoric Settlement of Britain. London: Routledge
217
Milber Down
Type Multiple-enclosure site Date of occupation 100 BC - AD 100
Grid Reference SX 88400 69800 Date of excavation 1937-1938
Features Multiple ditch and bank
fortifications, gullies, corn
drier, one possible hut.
Associated activities Spinning, food processing,
local trade, possible iron
working.
Notes Site composed of multiple ditches and banks, although no evidence of structures
were discovered inside the inner bank during excavation. Several spindle whorls
found in gullies and ditches, suggesting they were no longer used at time of
deposition. Other finds included a corn drying oven, large amounts of charcoal,
iron slag, and an iron pommel; this suggests some form of iron working on this
site.
Publications Fox, A., Raleigh Radford, C. A., Rogers, E. H., Shorter, A. H., 1949 - 1950.
Report on the Excavations at Milber Down, 1937-8. Proceedings of the Devon
Archaeological Exploration Society 4: 27-65
Cottrill. 1937.  Interim Excavation Report. Proceedings of the Prehistoric Society
3: 448-9
Cottrill. 1938. Interim Excavation Report. Proceedings of the Prehistoric Society
4: 321
Mount Batten
Type Promontory fort Date of occupation 900 - 100 BC
Grid Reference SX 48679 53252 Date of excavation 1983-5
Features Post holes, pits, gullies. Associated activities Bronze working and
smelting, farming, food
processing, spinning.
Notes Located on a promontory in the Plymouth Sound, Mount Batten has been built on
repeatedly in modern times, effectively erasing any evidence of prehistoric
defensive earthworks. Study of the bronze tools and ornaments, as well as their
stratigraphical placement, suggests continuous occupation from the Late Bronze
Age into the Middle Iron Age. Personal ornaments were found in large numbers
here, along with bronze ingots and smelting debris. Given the location of the site
(close to the mines of Dartmoor and on the trade route from Cornwall to Brittany),
this suggests Mount Batten was an important manufacturing centre and trading
post for at least the metal trade during the Late Bronze and Early Iron Ages.
Pottery in phase 3 (Early to Mid Iron Age) is mostly local southwestern ware with
a few Cornish sherds.
Publications Cunliffe, B. 1988. Mount Batten Plymouth: a prehistoric and Roman port. Oxford
University Committee for Archaeology Monographs
218
Appendix 2. Central South sites in this study
Dorset
Allard’s Quarry
Type Open settlement Date of occupation 800 BC - AD 100
Grid Reference ST 79491 19881 Date of excavation 1944-1945
Features Pits, post holes, two
burials, roundhouse,
4-post structure
Associated activities Spinning, weaving; bone
manufacturing; food
processing; trading
Notes 3 site phases in Iron Age; 42 pits belonging to 1st phase, 26 to the 2nd, and 16 to
the 3rd. 3 huts in 1st phase, incomplete due to quarrying activity; 2 huts during 2nd
phase; 1 possible hut in 3rd phase with 4-post structure (likely granary). No
imported pottery in any phases, although the presence of Kimmeridge shale
bracelets suggest coastal trading.
Publications Williams, A. 1950. Excavations at Allards Quarry, Marnhull, Dorset. Proceedings
of the Dorset Natural History and Archaeological Society 72: 20-75
Gussage All Saints
Type Banjo Date of occupation 300 BC - AD 100
Grid Reference ST 99895 10620 Date of excavation 1972
Features Ditch, pits, postholes,
roundhouses, and ovens.
Associated activities Pottery production, iron
forging, iron smelting in
the Roman phase,
spinning, weaving, food
production and processing.
Notes Site is divided into clear activity zones, with storage pits clustered to the east of
the roundhouse and crafting debris to the south. Craft activity represented by
ceramic debris from pit 209 (dating to the second phase of settlement, 1st century
BC), bronze casting and iron forging debris (from both early settlement phases),
and a large number of spindle whorls (dated to phase 3 of the settlement) and loom
weights (an equal number found in each of the 3 phases). The agricultural
economy of this site is demonstrated by a variety of carbonised grains from
various rubbish deposits. The high number of spinning and weaving tools from this
site suggest that this site probably produced more than required locally;
Wainwright believes this was part of a wider economic trade (1979: 190-1). This is
supported by the evidence that this site traded in querns and pottery from the larger
Wessex trading area.
Publications Wainwright, G.C. 1979. Gussage All Saints: An Iron Age Settlement in Dorset.
Department of the Environment: Archaeological Reports No 10
219
Hengistbury Head
Type Promontory fort Date of occupation 700 BC - AD 50
Grid Reference SZ 17300 90700 Date of excavation 1970
Features Hearths, roundhouses,
double dykes defences,
pits, and gullies.
Associated activities Long-distance and local
trade; metal, glass, and
shale manufacturing;
spinning and weaving;
food processing
Notes This site has evidence of non-continuous occupation from the Paleolithic period
onwards. Pottery finds date the earliest Iron Age occupation to around 700-400
BC, the exact nature of this period’s settlement is unclear but probably a scattering
of individual farmsteads before the double dyke defences were made, turning the
site into a hill-/promontory- fort. Two bronze axe finds (one 300m out at sea,
possibly result of erosion or dropped from boats) hint at this site being possible
stop-off for significant long-distance trading routes. Few artifacts or materials
definitely assigned to the middle Iron Age period (400-100 BC), indicating no or
little occupation. In the late Iron Age phase (100-0 BC), the settlement becomes a
major international trading port. Local, regional, and European trade demonstrated
by pottery finds (from west England, north-west France, and locally);
metal-working/manufacturing (ingots of copper indicate copper importation, silver
refining demonstrated with silver-copper ore/ingots, and bronze and gold working
debris); Kimmeridge shale importation and manufacturing (waste material and
finished artefacts); and limited glass production (crud glass lumps and fragments
of finished glass products). Other finds include a large number of coins (local and
imported); and range of luxury items (jewellery, personal items, etc.).
Publications Cunliffe, B. 1978. Hengistbury Head. London: Elek Books
Hod Hill
Type Hillfort Date of occupation 800 B.C. - 43 A.D.
Grid Reference ST 85500 10700 Date of excavation 1897; 1949; 1951-2
Features Ditches and banks;
roundhouses; pits; gullies;
postholes
Associated activities Spinning, weaving,
smelting, smithing,
quarrying
Notes A multivallate hillfort covering 22 ha., built in the early Iron Age, with a Roman
fort in the northwestern corner. Roundhouse footprints cover the site, often
overlapping, with associated pits and hearths. Several trackways are recorded from
the absence of roundhouse circle; these converge at the northeastern entrance,
indicating this was the main thoroughfare. Metalworking has a large presence on
site that appears to be centred on a possible ‘industrial’ compound in the southeast
corner with several quarry scars/pits. Magnotromic surveys indicate these pits are
filled with slag and other metalworking debris. Several Kimmeridge shale spindle
whorls from this site are repurposed lathe-turned cores that are left over from
producing shale bracelets. 52 spindle whorls are recorded in the excavations
report, but only 5 loom weights and 2 weaving combs indicate that weaving was
limited at this site. Large numbers of slingshots and spearheads suggests the
hillfort was abandoned after several battles or skirmishes.
Publications Richmond, I. 1968. Hod Hill, Vol 1-2. London: Trustees of the British Museum
Brailsford, J., & Richmond, I. 1952. British Museum Excavation at Hod Hill,
Dorset. The British Museum Quarterly 17 (3), 49-50
220
Maiden Castle
Type Hillfort Date of occupation 600 BC - AD 100
Grid Reference SY 66900 88400 Date of excavation 1934; 1937; 1985-6
Features Multiple ditch and bank
defences; roundhouses;
burials; gullies; roads;
pits; postholes; ovens
Associated activities Spinning and weaving;
metalworking; food
processing
Notes A large, multivaliate hillfort with extensive, complicated entrances. Initially a
small enclosure built in the 1st Millennium B.C. (phase 5), the site was extended
during the middle Iron Age to cover the whole of the hilltop. Occupation appears
densest during the middle Iron Age (phase 6), spanning across the site, but
becomes limited to the southeastern area during the late Iron Age (phase 7). A
thorough excavation revealed large quantities of artefacts, demonstrating little
metalworking during the early Iron Age. Animal remains indicate large numbers of
sheep, possibly to provide fibres for the textile production occurring at the site.
Pottery finds suggest there was on-site ceramic production during the early Iron
Age that was replaced in the middle Iron Age by importation of ceramics made in
Poole workshops. The site was largely abandoned during the late Iron Age with the
remaining occupation clustered around the eastern entrance, outside of which was
the remains of a smithy.
Publications Shaples, N.M. 1994. Maiden Castle, Dorset. In A.P. Fitzpatrick and E. Morris
(eds.) The Iron Age in Wessex: Recent Work. Wessex Archaeology, pp. 92-94
Sharples, N.M. 1991. Maiden Castle: Excavations and field survey 1985-6.
English Heritage: Archaeological Report no 19
Hampshire
Danebury
Type Hillfort Date of occupation 500 BC - AD 100
Grid Reference SU 32300 37600 Date of excavation
Features Ditch and bank
earthworks; hut platforms;
post holes; pits; hearths;
forges.
Associated activities Bronze and iron working;
spinning and weaving; salt
production; food
production and processing.
Notes A large and well-excavated hillfort with evidence of continuous occupation from
the 6th to the 1st century BC. Artefacts such as iron ingots and briquetage provide
evidence of trade that Cunliffe suggests were traded locally for raw materials like
wool or cereals. While no evidence of smelting was found on site, iron and bronze
working debris, materials, and tools show on-site production of artefacts. Evidence
of craft activities with shale, bone, basketry, and so on are rare, but textile
production was a significant activity throughout the life of the hillfort. All this
evidence supports the interpretation of Danebury as a centre of production for
certain crafts and of trading, both locally and long-distance.
Publications Cunliffe, B. 1994. Danebury, Hampshire. In A.P. Fitzpatrick and E. Morris (eds.)
The Iron Age in Wessex: Recent Work. Wessex Archaeology, pp. 94-97
221
Easton Lane
Type Ring Ditch Date of occupation 800 - 600 BC
Grid Reference SU 49500 31300 Date of excavation 1971
Features Linear depression;
irregular pits and scoops;
Associated activities Agricultural; waste
disposal; spinning and
weaving.
Notes Site was excavated as part of the M3 extension; 3 phases of activity identified
(Phase 1: pre-ring ditch; Phase 2: Bronze Age ring ditch; Phase 3: Iron Age
post-ring ditch). The ring ditch contained several associated cremations and
inhumations but was no longer used in the Iron Age. Exact use in this period is
unclear (Fasham 1982: 50).
Publications Fasham, P.J. 1982. The Excavation of Four Ring-Ditches in Central Hampshire.
Proceedings of the Hampshire Field Club and Archaeological Society 38: 19-56
Kennel Farm
Type Enclosed settlement Date of occupation 800 - 300 BC
Grid Reference SU 59555 48561 Date of excavation 1966
Features Oval enclosure ditch;
4-post structure; pits;
fence-line.
Associated activities Domestic occupation;
spinning and weaving;
storage; animal
husbandry; agriculture.
Notes Close to another settlement of Mid Iron Age - Roman date. Half the site was
excavated; no house structures found in excavated area. Storage pits were
clustered in at least 4 distinct groups, suggesting chronological sequence of storage
pits being used then closed (Chapman 2006: 20). Postholes rarely formed any
structural patterns, suggesting they were used alone or in pairs for external
activities (i.e. looms, haystack posts, etc.) (Chapman 2006: 21)..
Publications Chapman, A. 2006. An Iron Age Enclosure at Site A, Kennel Farm, Basingstoke,
Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society
61: 16-62
222
Winnall Down
Type Enclosed settlement Date of occupation 600 - 50 BC
Grid Reference SU 50600 30000 Date of excavation 1976-7
Features Single enclosure ditch;
roundhouses; 4-post
structures; hearths; pits;
post holes.
Associated activities Food production and
processing; food storage;
weaving.
Notes Site is comprised of a single enclosure ditch containing 8 roundhouses and 10
4-post structures, although not all dwellings or storage buildings would have been
built or used at the same time. Use of the enclosure appears very structured, with
certain features and activities restricted to areas of the site. The northwest corner is
characterised by 4-post structures and high levels of loom weight deposition, while
the southeast corner was used for crop processing and storage. Pits 2630 and 2676,
close to 4-post structures G and T, contained about 11 and 14 loom weights
respectively (all fragmented to some degree). Spatial correlation suggests
structures G and T were weaving workshops. A few pieces of briquetage and
locally sourced querns were found, although no evidence of metalworking were
found.
Publications Fasham, P.J. 1985. The Prehistoric Settlement at Winnall Down, Winchester:
excavations of MARC3 site R17 in 1976 and 1977. Monograph 2
Zions Hill Farm
Type Farmstead? Date of occupation 100 - 0 BC
Grid Reference SU 41500 20200 Date of excavation 1995 - 1996
Features Pits, gullies Associated activities Weaving, farming
Notes An unusual site with no structural evidence, only pits and gullies. 5 loom weights
excavated from late Iron Age/early Roman pit 5004. Pottery includes late Iron Age
coarseware made from local materials, similar to finds from other Iron Age sites in
the area. Quernstones indicate farming and seed processing occurring on site,
although no metalworking tools or debris were recorded. Appears to be the
working area of a small agricultural settlement/single farmstead nearby.
Publications Rawlins, M., Chadwick, A. M., Reynier, M. J. and Seager Smith, R. H. 2003.
Prehistoric and Roman activity at Zionhill Farm, Chandlers Ford. Proceedings of
the Hampshire Field Club & Archaeological Society 58: 1 - 23
223
Appendix 3. South East sites in this study
Kent
East Wear Bay, Folkestone
Type Possible oppida Date of occupation 500 B.C.-A.D. 43
Grid Reference TR 24000 36600 Date of excavation 1924; 1989; 2010-2011
Features Roundhouses, gullies,
4-post structures,
quernstone workshops,
pits, postholes
Associated activities Quern manufacture;
spinning; weaving;
continental and British
trade.
Notes Unfortunately, much of this site has been lost to the sea via cliff erosion, so the
remains must be taken as only part of the whole site. Full size of the Iron Age
occupation is unknown but its high status and large-scale craft production have
allowed it to be labelled an oppida. The site was largest around 150 B.C.-A.D.
43. Quernstone manufacture possibly performed as a production line, as several
querns in various stages of manufacture were abandoned by workers. A Roman
villa complex was built over the remains of the oppida in the late 2nd century
A.D. Finds include large quantities of British and continental pottery, including
Italian amphorae; 13 Iron Age coins; large amounts of quern debris and flint
debitage; and a significant number of spindle whorls. Only a few loom weights
found.
Publications Colsen, I. 2013. A Town Unearthed: Folkestone Before 1500. Canterbury
Archaeological Trust
Parfitt, K . 2012. The Rocky Road to Folkestone: Excavations at Folkestone
Roman Villa 2011. Kent Archaeological Society Newsletter 92: 2-4
2010. Folkestone, East Wear Bay Roman villa: Interim report on
excavations 2010. Kent Archaeological Society Newsletter 88: 14-15
Keller, T. 1988. The Evidence for Ancient Quern Production at Folkestone. Kent
Archeological Review 93: 59-68
224
Highstead
Type Enclosed settlement Date of occupation 800 BC-AD 43
Grid Reference TR 21300 66100 Date of excavation 1975-1977
Features Ditches, post holes, pits,
hearths, ovens,
roundhouses, palisade post
holes, gatehouse post
holes, 4-post structures,
gullies.
Associated activities Copper and iron
metalworking, pottery
production, agriculture,
food processing, food
storage, salt production.
Notes In the Late Bronze Age/Early Iron Age, the site had 3 enclosures. Enclosure B70
in the north of the site was rectangular with defensive palisade post holes and a
gatehouse within its ditch. Hearth and clay-lined pits (storage areas) excavated
within B70 indicate occupation; pottery, copper dross, and vitrified clay finds from
the enclosure ditch point towards crafting use rather than domestic. Enclosure A24
(to the southwest) was oval in shape and domestic, with 2 roundhouses (built
sequentially) in the centre. The Middle Iron Age (600-400 BC) is characterised by
an open settlement; occupation remains constant, with copper, pottery, and
subsistence activities continuing. Debris and remains of Ironworking and salt
production are found during this period, along with 4-post structures. This is
followed by a period of abandonment until 100 BC, when a series of rectangular
enclosures are dug, possibly for livestock management. No evidence of further
occupation is recorded, although pottery evidence suggests occupation in the
surrounding area. Site suffered some ploughing damage prior to excavation.
Publications Bennett, P., Couldrey, P., and Macpherson-Grant, N. 2007. Highstead, near
Chislet, Kent. Excavations 1975–1977. Canterbury Archaeological Trust
Monographs: New Series Vol IV
Sunset Caravan Park
Type Unclear Date of occupation 700 BC - AD 100
Grid Reference TR 09296 64478 Date of excavation
Features Pottery kilns, quarry pits,
roundhouses, pits,
postholes, gullies, ditches
Associated activities Clay and flint quarrying;
metal working; pottery
production; spinning; salt
production; long-distance
trade
Notes Given the confines of this excavation (being a rescue excavation prior to building
work), the nature of this site and whether it was enclosed or defended in any way
is unclear. Roundhouse features have been found throughout the excavation,
demonstrating occupation at this site. The predominant activity on this site appears
to be clay and flint quarrying, which increased during the Late Iron Age. Clay
ovens and debris also indicate high levels of pottery production; there is also
evidence for salt production. Early Iron Age Italy pottery finds include italian and
other continental wares, which indicate long-distance trade at this site.
Publications Allen, T. 1997. The Whitstable Iron Age Settlement. The Archaeological
Evaluation and Results of Preliminary Excavation at Sunset Caravan Park, Thanet
Way, Whitstable, Kent, 1997/61. Canterbury Archaeological Trust
Allen, T. 1999. The Whitstable Iron Age Settlement: Parts 1 and 2. Unpublished
Canterbury Archaeological Trust
Allen, T., and Wilson, J. 2001. Sunset Caravan Park and Church Lane East,
Whitstable. Canterbury’s Archaeology 1998-1999: 10-11
225
Thanet Earth
Type Enclosed settlement Date of occupation 500 BC - AD 43
Grid Reference TR 28900 66700 Date of excavation 2007-8
Features Ditches, pits, postholes,
ring gullies, burials, field
systems
Associated activities Spinning and weaving;
food processing; pottery
production; possible antler
manufacturing
Notes Enclosed settlement with domestic features within and beyond enclosure ditch,
suggesting expansion of occupation during Middle Iron Age. Most features from
this date were clustered on plateau 8 and including ditches, pits, postholes, and
ring gullies. Loom weights recovered from rubbish deposits in pits in this area; the
plateau produced an unusually high number of Iron Age spinning and weaving
tools, suggesting textile production as a major craft activity at the site. The
occupation decreases in the Late Iron Age before finishing completely. There was
little to no evidence of other craft activities.
Publications Rady, J. 2009. Thanet Earth, Monkton. Canterbury’s Archaeology 2007-2008:
11-27
2010a. Thanet Earth, Monkton. Canterbury's Archaeology 2008-09: 1-16
2010b. Excavations at Thanet Earth 2007-2008 Assessment Report Volume
1. Canterbury Archaeological Trust
226
Appendix 4. Textile tool data form
Museum
No.
Tool Site Context Material Shape Height Width/
diameter
Depth Weight Decor Preser-
vation
227
Appendix 5. Cornwall artifact data
Bodrifty
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1 Discoid Reused sherd [1] 58 16 26.6
4 Discoid Reused sherd [6] 33 26 25.4
H1/9 BY154 Discoid Reused sherd (148)[43] 17 15 4.8
9 BY54 Discoid Reused sherd (148)[43] 32 12 23.15
G663 Discoid Reused sherd [7] 41 26 42.4
2 G664 Discoid Reused sherd [5] 34 23 22.15
Carn Euny
Table 2: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
52 Discoid Stone Cut 2 CN 33.5 14 26.45
88 Discoid Stone Cut 1 (ss) Peg 1F 43 24.4 61.15
114 Discoid Slate Hut A sq 6 N side 35.5 7.5 16.05
121 Discoid Baked clay Hut A sq 9 Peg E 30 13 10.35
177 Discoid Stone Peg 4F-4H 32 9 13.45
198 Discoid Baked clay Unrecorded 16.4 7 13.8
301 Discoid Reused sherd TA 1 Wall (1) 40 6.5 16.7
386 Discoid Stone Cut 4 Peg J-3M 28 9 12.25
388 Discoid Reused sherd Cut 5 Peg 4d-4c 46 7 15.8
452 Discoid Reused sherd TA 3 Peg 3TF/O 37.4 7.5 13.75
501 Discoid Reused sherd
TOS (N) E. Hut H
behind wall (4) 37 9.3 12.85
680 Discoid Stone CHI (s) sq 4 - wall on E 34 16.2 27.75
699 Discoid Reused sherd Unrecorded 34.6 5.4 8.8
784 Globular Stone
CH2 Room 2 Wall face
(1) 25.6 20 19.05
228
The Rumps
Table 3: Spindle whorl data
Museum No. Form Material Location
Diameter
(mm) Depth (mm) Weight (g)
1 Discoid Slate R67 R26 A(4) [230] 29 5.5 6
2 Cylindrical Baked clay R2K A(4) [218] 35.5 17 27
3 Cylindrical Stone R65 BRI ditch A 39 18.6 41.5
3a Discoid Stone RI (39) <18> ditch C 29.6 6.3 7.6
5 Domed Bone BR1 (9-10) <24> ditch A 37.6 19.7 9.3
6 Cylindrical Baked clay A RI-M (1) [199] 32.4 15 14.7
7 Discoid Stone A RIN (4) [243] 38 15.3 33.6
9 Discoid Baked clay RIP (4) [280] R67 34 17.4 26.75
11 Discoid Slate A R2G (7) [212] 41 5 15
14 Discoid Slate BRI (9-10) ditch A <23> 63.6 6.7 44.4
15 Discoid Slate RIM/RIN (4) [271] 49 8 26.8
16 Discoid Slate N side of BR2B [42] 39 3 7.6
17 Discoid Baked clay A RIJ (2) <261> 33.4 16 20.6
? Discoid Slate RIM (1) [201] 30 6 17
125 Discoid Slate A R2 (4) inner ditch [91] 34 7 14.7
Table 4: Loom weight
Museum
No. Form Material Location Width (mm) Depth (mm) Weight (g)
10 Discoid Slate AR2B (5) [186] 86 16.7 156
18 Discoid Slate B R2 A bottom of (7) 181 10.4 543.6
?A Discoid Slate R2 inner ditch extension (4) [15] 97 14 108
?B Discoid Slate R2 inner ditch extension (4) [15] 85 14 384.6
45 Discoid Slate R2 inner ditch extension (4) [145] 138 7 236
46A Discoid Slate R2 inner ditch extension (4) [37] 75 6 46.8
46B Discoid Slate R2 inner ditch extension (4) [37] 33 4.3 11
46C Discoid Slate R2 inner ditch extension (4) [37] 34.3 5.3 59.4
44/47A Discoid Slate R2 inner ditch (4) [65] 55.5 3 13.95
44/47B Discoid Slate R2 inner ditch (4) [65] 76.4 6.4 146.4
44/47C Discoid Slate R2 inner ditch (4) [65] 50 3.5 53.4
44/47D Discoid Slate R2 inner ditch (4) [65] 50 4.3 54.8
Trevelgue Head
Table 5: Spindle whorl data
229
Museum
No. Form Material Location
Diameter
(mm)
Depth
(mm) Weight (g)
113 Discoid Stone Tr 63 [452] 33 17.3 22.7
155a Discoid Stone House [420] 34 11.2 18.95
155b Discoid Stone House [420] 31 9.4 19
181 Discoid Stone House [421] 30.6 12.4 12
216b Discoid Slate House [136]/[137] 28.5 11.6 17.65
216c Discoid Stone House [316]/[317] 45.7 7.2 20
216d Discoid Stone House [136]/[137] 30.6 13 17.8
228a Discoid Stone House [159] 39.4 6.5 12.95
228b Discoid Stone House [422] 41 8 17.3
243 Discoid Slate House [421] 38.8 5 9
261 Globular Stone House [422] 14 15.7 31.4
279a Globular Baked clay House [420] 15 21.8 46.8
419 Discoid Stone House [139] 38 15.3 31.45
440 Discoid Slate House [440] 41 4.2 9.2
502 Discoid Stone House [142] 31 17.7 32.25
503 Discoid Stone House [142] 29.2 9 10.75
504 Discoid Slate House [420] 39 5.4 14
505 Discoid Stone Tr 65 31.8 17.7 18.45
90170A Discoid Baked clay Tr 6.11 [420] 14 8 16.6
90175A Discoid Baked clay Tr 612 [422] 33.6 8.2 10
90376A Discoid Stone House 1 100/34X Y49 p/h 31 6.5 15.7
90380 Discoid Stone Tr 6 [420] 34.7 14 18
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
133 Discoid Stone House [420] 88.3 36.2 654
277 Discoid Slate House [420] 78 7 56.85
230
Trevisker
Table 7: Spindle whorl data
Museum No. Form Material Location
Diameter
(mm) Depth (mm) Weight (g)
SW 1 Discoid Stone T/116 42 19 46.25
SW 2 Discoid Baked clay T/49 34.4 12.3 16
SW 4 Discoid Stone T/6 33.5 8.5 15.75
SW 5 Discoid Stone T/170 30.5 11.3 12.6
SW 7 Discoid Stone T/88 30.4 13.4 20.6
SW 8 Discoid Reused sherd T/150 38.4 7.3 13.4
SW 9 Discoid Stone T/115 30.7 8.7 14.45
SW 10 Globular Stone T/8 30.5 8.3 19.4
SW 11 Discoid Stone T/115 22.7 7 6
SW 12 Globular Stone T/244 35.5 9.2 32
Table 8: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
<40> Discoid Baked clay [75] 122 92.1 1372.8
T/A/596 Discoid Baked clay [76] 62.4 61.4 691.7
T/29 Discoid Slate Unrec. 154 14.4 582.1
T/32 Discoid Slate Unrec. 99 9.3 162
T/38 Discoid Slate Unrec. 94.3 17.5 258
T/45 Discoid Slate Unrec. 53.6 6 23.8
T/52(?) Discoid Slate Unrec. 114 20 398
T/65a Discoid Slate Unrec. 79.7 9.3 115.5
T/65b Discoid Slate Unrec. 95 9.4 189.6
T/95 Discoid Slate Unrec. 51.7 9.8 36
T/227 Discoid Slate Unrec. 41.4 1.8 6.1
T/230 Discoid Slate Unrec. 49 13.3 1092
T/237 Discoid Slate Unrec. 48 5.8 17
231
Appendix 6. Devon artifact data
Blackbury Camp
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
12/1956.96 Discoid Sandstone BC, B, trench 2 47 19 42
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
12/1956.107 Triangular Fired clay Site A, rampart 5 183 162 1400
12/1956.108 Triangular Fired clay Site A, rampart 5 129 75 1180
12/1956.109 Triangular Fired clay Site A, rampart 5 171 120 2040
12/1956.110 Triangular Fired clay Site A, rampart 5 48 30 78
Dainton
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A4431 Discoid Sandstone Unrecorded 93.6 44 166
A5977 Discoid Sandstone Unrec. 42 22 39
A5978 Discoid Stone Unrec. 44 18 24
A4480 Discoid Baked Clay Unrec. 35.7 50 18
A4481 Discoid Baked Clay Unrec. 34.7 50 22
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
A5976 Rounded Slate/Shale Unrecorded 130 20 272
232
Kents Cavern
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A3281 Discoid Shale Unrecorded 36 8 10
A5958 Discoid Stone Unrecorded 37 28 48.4
A5950 Discoid Shale 30 25 13 14.7
A5951 Discoid Shale 42 27 14 14
A5952 Discoid Shale 48.4 36 12 16
A5953 Discoid Stone Unrecorded 39 12.5 16.25
A5954 Discoid Shale Unrecorded 37 15 25
A5956 Discoid Shale Unrecorded 33 15 20
A5957 Discoid Stone Unrecorded 42 19 41
A5959 Discoid Sandstone Unrecorded 36 16 30.5
A5960 Discoid Sandstone Unrecorded 35 13 19
A5961 Discoid Sandstone Unrecorded 40 13 26.6
1886.12-29.26 Globular Baked Clay Unrecorded 28 16 38.75
1886.12-29.25 Discoid Stone Unrecorded 34 10 21
1886.12-29.27 Discoid Sandstone Unrecorded 34 21 27
1886.12-29.24 Discoid Stone Unrecorded 28 30 33
1886.12-29.34 Discoid Stone Unrecorded 20 8 4.2
1886.12-29.35 Discoid Stone Unrecorded 28 13 19
1886.12-29.36 Discoid Stone Unrecorded 32 15 27.3
1886.12-29.37 Discoid Stone Unrecorded 25 20 22
Milber Down
Table 6: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A4178 Discoid Sandstone Unrecorded 31 12 14.7
A5935 Discoid Shale Unrec. 30 12 8
A5936 Discoid Shale Unrec. 36 14 28
A5937 Discoid Shale Unrec. 28 14 21
233
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
A4342 Discoid Stone Main camp 86 15 181.25
A5939 Discoid Sandstone Unrecorded 65 13 64.4
Mount Batten
Table 8: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1958.11.86.1X Discoid Sandstone Unrecorded 35.5 17 35.5
1958.11.86.2X Discoid Sandstone Unrec. 45 17.5 68
1958.11.86.3X Discoid Stone Unrec. 33 16 26
258.18 Discoid Sandstone Unrec. 36 13 26
234
Appendix 7. Dorset artifact data
Allard’s Quarry
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
3 Discoid Limestone Pit 55 27.8 24.2 71
4 Globular Human bone Pit 17 40 26.4 20
13 Discoid Fired clay Pit 17 82.3 11.2 57
16 Globular Limestone Pit 8a 26 20.5 15
17 Discoid Limestone Pit 12a 38.8 27.5 49
19 Discoid Limestone Pit 22 38 13 34
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
?b Fragmentary Stone Pit 31 186.7 31 303.3
13 Triangular Fired clay Pit 112 213.3 60 1313.3
Gussage All Saints
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
?c Discoid Chalk 212 (3) 42.5 15 30
2010 Discoid Chalk 212 (3) Phase J 42 15 30
4001 Discoid Fired clay 358 (3) 30.5 26.5 27
4004 Discoid Fired clay U/L (2) 30 8 9.4
4008 Globular Fired clay 446 (6) 33.5 25 31
4030 Discoid Fired clay 396 (4) 31 20.9 24
4031 Discoid Reused sherd 111 (4) 43.9 27.4 18
4033 Discoid Fired clay 528 (5) 37.8 22.4 36
235
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
?a Triangular Chalk 2 (5a) 1201 83 2038
?b Triangular Fired clay 539 (6) 205.7 110 3710
4005 Cylinder Fired clay IQ (3) 59 59 820.8
4009 Triangular Fired clay 415 (8) 128.75 68 520
4012 Cylinder Fired clay 620 (6) 68.5 59 871.5
4014 Triangular Fired clay 377 (5) 233.6 92.8 1673.3
4020 Triangular Fired clay 379 (8) 1807 58 2082.6
4020a Triangular Fired clay 379 (8) 166.8 67.2 1946.25
4021 Triangular Fired clay 379 (8) 194.28 61.8 1774.2
4022 Triangular Fired clay 639 [310J] (5) 401.6 71.5 2516.6
4023 Triangular Fired clay 639 [310J] (6) 254.4 61.1 1842.2
4039 Triangular Fired clay 52 (7) 258 52.8 1250
Hengistbury Head
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
69 Discoid Reused sherd Ph 243/1 44 7 18
122 Globular Fired clay O/179 30 19 14
125 Discoid Reused sherd O/180 49 7 19.5
161 Discoid Reused sherd F42/272 59 7 20.4
183 Discoid Reused sherd O/295 46 9 26.25
211 Discoid Reused sherd O/191 44 9 27.5
223 Discoid Reused sherd Ph 839/1 34 8 10
225 Discoid Reused sherd F270/504 28 7 7.2
235 Discoid Reused sherd F273/514 42 28 60.8
242 Discoid Reused sherd F260/452 31 6 9.2
265 Discoid Reused sherd F399/601 56 14 37
295 Discoid Reused sherd O/689 46 11 30.6
300 Discoid Reused sherd O/696 35 5 8.7
305 Discoid Reused sherd O/696 43 10 23
338 Discoid Reused sherd O/837 44 7 17
W3 Discoid Reused sherd X229 36 8 10
W5 Discoid Reused sherd X227 51 14 38
x60 Discoid Reused sherd B-F 40 7.5 15.9
x61 Discoid Reused sherd B-F 40 8.5 12
Museum No. Form Material Material Diameter (mm) Depth (mm) Weight (g)
x62 Discoid Reused sherd B-F 38 9.5 11.4
x66 Discoid Reused sherd B-F 39 7 9
x67 Discoid Reused sherd B-F 48 13.75 30.8
236
x68 Discoid Reused sherd B-F 45 13.5 28
x69 Discoid Reused sherd B-F 40 15 21.5
x70 Discoid Reused sherd B-F 45.5 9.5 22.2
x71 Discoid Reused sherd B-F 44 9 16
x72 Discoid Reused sherd B-F 55 7.5 18.3
x73 Discoid Reused sherd B-F 43.5 7 16.3
x74 Discoid Reused sherd B-F 43.5 10.5 16
x75 Discoid Reused sherd B-F 31 6.5 5.8
x76 Discoid Reused sherd B-F 41.5 10 17.4
x77 Discoid Reused sherd B-F 42 7 14.2
x78 Discoid Reused sherd B-F 37 8 11.3
x79 Discoid Reused sherd B-F 40 8 8
x80 Discoid Reused sherd B-F 32 10 10.8
x81 Discoid Reused sherd B-F 55 23 84.2
x82 Discoid Reused sherd B-F 49 10.5 26.8
x83 Discoid Reused sherd B-F 48.5 9.5 21.2
x86 Discoid Reused sherd B-F 52 8 24.7
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
x92 Elongated Fired clay B-F 80 85 1102.2
x93 Elongated Fired clay B-F 80 80 1292.2
x94 Triangular Fired clay B-F 60 40 1090.4
x95 Triangular Fired clay B-F 62 40 1360
x96 Elongated Fired clay B-F 92 96 1394.25
x97 Triangular Fired clay B-F 60 40 278.75
x98 Elongated Fired clay B-F 99 90 1725
x263 Triangular Fired clay sf 5 75 55 2574.6
x265 Triangular Fired clay Unknown 60 40 2138.3
x266 Triangular Fired clay sf 4 85 65 2346.6
237
Hod Hill
Table 7: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
92.9-1.1386 Discoid Stone Unrecorded 32 13 7.7
92.9-1.1387 Discoid Baked Clay Unrec. 39 14 31.4
92.9-1.1389 Discoid Stone Unrec. 33 9 8.5
92.9-1.1396 Conical Antler Unrec. 36 24 21
92.9-1.1393 Globular Stone Unrec. 44 32 94.4
92.9-1.1394 Globular Stone Unrec. 43 30 62.1
92.9-1.1395 Globular Stone Unrec. 32 27 34.4
92.9-1.1397 Globular Stone Unrec. 37 22 33.3
92.9-1.1398 Globular Stone Unrec. 26 23 50
92.9-1.1399 Globular Sandstone Unrec. 30 20 23.5
92.9-1.1400 Discoid Sandstone Unrec. 40 19 26.6
92.9-1.1401 Globular Sandstone Unrec. 38 22.3 29.4
92.9-1.1402 Globular Baked Clay Unrec. 37 19 26
92.9-1.1403 Discoid Stone Unrec. 39 15.5 26.6
92.9-1.1404 Discoid Stone Unrec. 44 19 35
92.9-1.1405 Discoid Stone Unrec. 37 14.4 26.6
92.9-1.1406 Discoid Stone Unrec. 44 15.5 33.3
92.9-1.1407 Discoid Stone Unrec. 60 15.5 75.5
92.9-1.1408 Discoid Sandstone Unrec. 45 23.3 46.6
92.9-1.1409 Discoid Sandstone Unrec. 47 18 32.6
92.9-1.1410 Discoid Sandstone Unrec. 35 20 20
92.9-1.1411 Globular Stone Unrec. 32 32 34.7
92.9-1.1414 Discoid Sandstone Unrec. 26 13.75 10
92.9-1.1415 Discoid Sandstone Unrec. 28 13.75 7.5
92.9-1.1416 Discoid Baked Clay Unrec. 22 12.5 6.25
92.9-1.1417 Discoid Baked Clay Unrec. 29 13 9.4
92.9-1.1419 Discoid Stone Unrec. 20 15.7 6.3
92.9-1.1420 Discoid Stone Unrec. 30 17 19
92.9-1.1421 Discoid Baked Clay Unrec. 30 14.7 14.7
92.9-1.1423 Discoid Reused sherd Unrec. 30 8.2 9.4
92.9-1.1424 Discoid Reused sherd Unrec. 35 10 15.5
1960.4-5.120 Globular Baked Clay Hut 56 annexe 46 25 47.3
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1960.4-5.122 Discoid Chalk Palisade cut 89 24 125
1960.4-5.1004 Discoid Reused Sherd Unrec. 32 7 6
1960.4-5.1006 Discoid Chalk Unrec. 74 40 134
Table 8: Loom weight data
238
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1960.4-5.124 Elongated Chalk Pit 15b 160 76.6 3091
1960.4-5.125 Elongated Chalk Unrecorded 171.6 116.6 2466.6
1960.4-5.126 Elongated Chalk Unrec. 148.75 138.75 2996.25
1960.4-5.127 Elongated Chalk Unrec. 112.6 79 1807
1960.4-5.128 Elongated Chalk Pit 15b 108 164 2732
1960.4-5.1845-53 Elongated Chalk Pit 15a 183.5 88 2906
1960.4-5.2685 Elongated Chalk Unrec. 150.6 149.3 3456
1960.4-5.3353 Triangular Baked clay Unrec. 106 66 650
1960.4-5.3428 Elongated Chalk Hut 36a posthole 116 52.6 1526.3
1960.4-5.3429 Elongated Chalk Hut 36a posthole 105 70.5 1708.4
1960.4-5.3430a Elongated Chalk Unrec. 1086.3 80 2491.4
Maiden Castle
Table 9: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
113 Globular Bone R Pit R 34 (1) 39 36 23
.55.4 14 34/204 Discoid Stone Area B misc 38 12 17
.55.4 15 34/206 Discoid Chalk B ditch? (35) 57 21 82
.55.4 16 34/207 Discoid Chalk Area B 1 (3) 71 26 106
.55.4 17 34/208 Discoid Stone Area B 45 14 41
.55.4 18 34/210 Globular Stone Area A pit 8 34 22 24
.55.4 19 34/211 Discoid Chalk Area A 24 (8) 49 25 55.5
.55.4 20 34/212 Discoid Stone Area B pit 6 40 36 41
.55.4 21 34/213 Discoid Chalk Area B 6 (14) 49 21 58
.55.4 22 34/214 Discoid Chalk Area B pit 37 58 26 73.3
.55.4 23 34/215 Discoid Chalk B pit 1 (5) 45 23 54
.55.4 24 34/218 Discoid Chalk A pit 21 35 12 17
.55.4 24 34/220 Discoid Chalk B pit 2 34 13 15
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
.55.4 25 34/221 Discoid Chalk Area A 838 33 16 27
.55.4 27 34/225 Discoid Chalk B pit 24 (7) 39 21 35
.55.4 28 34/226 Discoid Chalk D Pit D (2) or A ditch 10? 39 14 30
.55.4 32 34/230 Discoid Stone B pit 8 48 12 37
.55.4 72 35/264 Globular Chalk B pit 14 ? 32 22 34
.55.4 73 35/311 Discoid Chalk Unrecorded 41.25 17.5 16.25
.55.4 74 35/331 Globular Chalk D pit 12 36.8 23 34.7
.55.4 75 35/360 Discoid Chalk Unrec. 44 17 14
.55.4 79 35/415 Discoid Chalk Unrecorded 44 25 52.6
.55.4 80 35/546 Discoid Chalk Pre Hut D 42 19 38
.55.4 81 35/565 Discoid Chalk Pre Hut D 57.5 17.5 38.75
239
.55.4 83 35/605 Discoid Chalk D2 (175) 28.4 11.5 7.3
.55.4 84 35/611 Discoid Chalk Site ? (363) 35 16 21
.55.4 85 35/622 Discoid Chalk DR (365) 31.6 9.4 8.4
.55.4 86 35/684 Discoid Chalk 1103 52.6 17 48.4
.55.4 87 35/712 Globular Chalk Pit F 20 (6) 45.5 26.6 49
.55.4 88 35/728 Globular Chalk 72 (3) 46 25 55
.55.4 89 35/729 Discoid Chalk Pit F 22 (3) 52 20 36.6
.55.4 90 35/735 Globular Chalk Pit F 22 (6) 43.3 21 37.7
.55.4 92 35/896 Discoid Chalk F 623 37 30 46
.55.4 93 35/902 Discoid Chalk F 1258 64 26 96
.55.4 94 35/912 Discoid Chalk F 549 53.3 20 47.7
.55.4 95 35/915 Discoid Chalk F 1258 32 13 12
.55.4 97 35/1014 Discoid Chalk F 849 57.6 16.4 41
.55.4 99 35/1068 Discoid Chalk F 1296 M2 38 15.7 15
37/1172 Discoid Chalk Q dump 51 21 34.4
55.4 100 35/1176 Discoid Chalk D ditch 19 (2) 44 18 27.3
55.4 101 35/1177 Discoid Chalk D Pit 19 38 20 22
55.4 102 35/1083 Discoid Chalk F (169) 46.3 19 32.6
55.4 122 36/483 Discoid Chalk H (1348) 35 14 19
55.4 124 36/594 Discoid Stone L pit LA (1) 55 21.6 16.6
55.4 127 36/812 Discoid Chalk L p.h. 19 52 18 50
55.4 128 36/827 Discoid Chalk L pit 12 (8) 43 23 46
55.4 129 36/828 Discoid Chalk L pit 12 (10) 52 13 64
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
55.4 131 36/1118 Discoid Stone L p.h. 124 28 15 16
55.4 134 36/1305 Discoid Chalk A 4 51.25 16.25 26.25
55.4 135 36/1420 Discoid Chalk B pit 14 50 28.75 46.25
55.4 136 36/1568 Discoid Chalk G pit H (8a) 45.3 16 13.3
55.4 137 36/1569 Discoid Chalk G (8a) 36 17 22
55.4 142 37/93 Discoid Chalk Q (4) 36 14 15
55.4 143 37/193 Discoid Chalk Q (4) 43.3 23.3 31
55.4 144 37/225 Discoid Chalk Q Hearth B 41 17 21
55.4 146 37/563 Globular Chalk Q (2) 39 27 36
55.4 152 35/1094 Discoid Shale Q pit 42 35 17 30
55.4 154 37/1137 Globular Chalk Hearth B (5) 41 23 50
240
Table 10: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1939.55.22 24 Triangular Fired clay Site E (673) 104.4 83.3 477.7
39.55.4 34/167 Triangular Chalk Hut D B2 82 59 781
39.55.4 34/200 Discoid Chalk A 21 99 34 378.75
55.4 13 34/203 Discoid Chalk B 14 (10) 103 31 379
55.4 29 34/227 Discoid Chalk Area B unstrat 75 30 193
55.4 38 34/259 Triangular Chalk Area A ? 1163 84 1694.7
55.4 40 34/275 Elongated Chalk B pit 17 (6) 85 72 1255
55.4 54 35/18 Rounded Chalk D pit D12 (6) n62 1164 83 1539
55.4 68 35/124 Triangular Chalk Pit 6 (826) 1868 93.3 2237.3
55.4 71 35/145 Elongated Chalk Under hut D 93.6 71.5 979
55.4 82 35/568a Elongated Chalk Pre Hut D 75 75 243
55.4 91 35/739 Discoid Chalk Pit F 22 (7) 115 31.25 216.25
55.4 103 35/1201 Rounded Chalk D refill 48 20 40
5.44 110 35/1361/1 Elongated Chalk E (422) 15-16 88.8 76.6 1570
55.4 113 35/1361/2 Triangular Chalk E (422) 15-16 Hearth A 98 102 1820
55.4 133 36/1250 Elongated Chalk Pit G4 A (2) 110.5 65 1064
55.4 145 37/498 Rounded Chalk P3 95 21 214
55.4 149 37/831 Elongated Chalk Q pit 36 (2) 70 59 973.75
55.4 164 Triangular Chalk L pit L1 96 44 613
241
Appendix 8. Hampshire artifact data
Danebury
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1563 Discoid Fired clay P1333 37 24 39
1643 Discoid Fired clay P1569 51 28 87
1654 Discoid Chalk P1548 30 17 14
1704 Discoid Chalk P7781 47 29 63
1770 Globular Fired clay F7179 47 32 70
1830 Discoid Chalk P1978 79 18 122
1928 Cylindrical Chalk F838 50 25 68
1935 Discoid Chalk P2261 35 15 19
1977 Discoid Chalk P2256 35 16 18
1986 Discoid Chalk P2256 62 23 91
1990 Globular Fired clay P2199 85 28 80
1994 Globular Fired clay F947 38 24 40
2048 Discoid Chalk F920 65 25 104
2143 Discoid Chalk F1257 41 17 45
2190 Discoid Chalk P2346 57 21 57
2408 Globular Fired clay P1127 42 42 84
2409 Discoid Chalk F2507 34 18 18
2427 Discoid Chalk F1562 42 17 28
2466 Discoid Chalk P2531 53 20 58
2478 Conical Chalk Ph 9826 39 23 24
2496 Conical Fired clay P2510 40 22 80
2567 Discoid Chalk F1681 44 24 53
2617 Discoid Chalk P2579 64 23 94
2692 Cylindrical Chalk F1941 41 27 86
2750 Cylindrical Chalk F356 45 31 84
2759 Conical Chalk F2005 41 24 35
2762 Discoid Chalk F1997 70 23 111
2775 Discoid Chalk F1997 29 13 11
242
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
602 Triangular Chalk P530 141 72 1808
1304 Triangular Chalk P474 474 97 2247
1313 Elongated Chalk P1026 178 87 1890
1315 Triangular Chalk P1026 138 146 1540
1316 Triangular Chalk P1026 134 78 1970
1330 Elongated Chalk P1070 121 103 1958
1331 Elongated Chalk P1026 157 157 2160
1332 Triangular Chalk P1026 132 101 2523
1333 Elongated Chalk P1026 138 138 1866
1338 Triangular Chalk Ph 3507 169 78 1369
1339 Triangular Chalk P1030 158 60 1682
1507 Rounded Chalk P1207 121 121 443
1542 Rounded Fired clay P1224 86 86 218
1547 Elongated Chalk P1299 92 92 1114
1550 Triangular Chalk P1334 129 74 1324
1551 Elongated Chalk P1334 94 66 1324
1559 Elongated Chalk P1334 125 59 1335
1560 Triangular Chalk Unknown 137 65 1380
1561 Elongated Chalk P1393 101 101 1418
1562 Elongated Chalk P1393 116 116 1238
1566 Elongated Chalk P1393 109 109 1221
1567 Elongated Chalk P1395 102 102 1208
1568 Elongated Chalk Ph 4612 102 102 1264
1572 Elongated Fired clay P1298 125 64 1212
1573 Triangular Fired clay Ph 4568 98 52 528
1575 Triangular Chalk Ph 4611 120 52 951
1576 Elongated Chalk Ph 4611 75 75 979
1577 Triangular Chalk Ph 4611 125 67 1180
1580 Elongated Chalk P1285 126 47 1251
1586 Elongated Chalk P1285 111 111 1898
1618 Triangular Fired clay Ph 6302 91 79 1612
1632 Rounded Chalk P1456 105 105 1580
1647 Triangular Fired clay Ph 1530 88 73 1736
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1684 Triangular Fired clay Ph 7474 119 71 1104
1686 Triangular Fired clay Ph 7474 147 60 1137
1687 Triangular Fired clay Ph 7474 108 77 1173
1689 Triangular Fired clay Ph 7474 113 78 1068
1690 Triangular Fired clay Ph 7474 123 74 1301
243
1694 Triangular Fired clay Ph 7474 119 72 1357
1707 Triangular Fired clay P1820 81 100 1126
1708 Triangular Fired clay P1687 115 68 1024
1709 Rounded Chalk P1790 113 113 348
1732 Triangular Fired clay P1768 155 76 2184
1737 Triangular Chalk P1727 152 84 2622
1739 Triangular Chalk P1727 165 165 1958
1904 Triangular Fired clay P2110 130 64 1335
1906 Triangular Fired clay P2047 102 62 1140
1923 Triangular Fired clay P1768 147 78 2543
1924 Triangular Fired clay P1768 134 80 1690
1931 Rounded Chalk F851 118 118 518
1940 Rounded Chalk F868 63 75 170
1941 Triangular Fired clay P2259 154 68 1460
1956 Elongated Chalk P2200 149 149 2274
1957 Elongated Chalk P2200 145 145 2257
2000 Elongated Chalk P2200 92 52 747
2001 2005
2008 Elongated Chalk P2200 87 138 1328
2006 Rounded Chalk P2200 94 74 760
2011 Elongated Chalk P2200 137 58 2012
2086 Elongated Chalk P2204 91 91 927
2439 Triangular Fired clay P2534 72 40 600
2489 Rounded Chalk P2509 111 111 202
2497 Triangular Fired clay P2541 123 65 1144
244
Easton Lane
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
282 Cylindrical Fired clay F4560 32 46 65
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
9 Triangular Fired clay Trench AA F317A 186 60 1992
62 Cylindrical Fired clay Trench B 115 102 1338
283 Elongated Chalk F4560 73 73 1196
284 Rounded Chalk F4560 100 100 929
287 Elongated Chalk F4560 78 78 973
288 Elongated Chalk F4560 85 85 1025
292 Rounded Chalk F4567 118 118 1340
294 Triangular Chalk F4560 97 97 1144
Kennel Farm
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
21 Discoid Chalk F448 42 20 3
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1 Triangular Fired clay F22 192 76 3062
2-3 Triangular Fired clay F59.2 140 88 1650
4 Triangular Fired clay F23.9 120 85 2426
6 Triangular Chalk F68.3 133 73 1100
7 Triangular Fired clay F22 130 93 2583
42 Triangular Fired clay F182 170 90 3156
43 Triangular Fired clay F182 172 64 3183
245
Winnall Down
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
37 Triangular Stone F1065 162 42 1591
49 Triangular Fired clay F1631 Fragmented 92 2595
55 Triangular Fired clay F2676 172 89 2548
74 Triangular Chalk F2676 185 72 1968
75 Triangular Chalk F2676 145 83 1697
77 Triangular Chalk F2676 142 74 1942
81 Triangular Chalk F2676 165 86 2012
86 Triangular Fired clay F2676 178 82 2061
152 Triangular Fired clay F3901 159 63 1803
152b Triangular Fired clay F3901 142 90 1894
157 Triangular Fired clay F3901 142 93 1664
244 Triangular Fired clay F2630 183 104 3331
255 Triangular Fired clay F2630 181 82 2846
Zion Hill
Table 8: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
6000 Cylindrical Fired clay F5015 83 99 862
6001 Cylindrical Fired clay F5015 87 84 672.5
6002 Cylindrical Fired clay F5015 72 92 522
6003 Triangular Fired clay F5013 153 106 2760
6005 Triangular Fired clay F5003 109 77 1490
246
Appendix 9. Kent artefact data
East Wear Bay, Folkestone
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
22 Discoid Reused pot 24 43 10.5 23
33 Discoid Sandstone 103 34.7 10 16
86 Discoid Reused pot 207 62 5.3 32
100 Discoid Reused pot 217 47 10.5 24
113 Discoid Reused pot 219 7.3 8.4 16
127 Discoid Stone 216 43 16 29
146 Globular Fired clay 166 35 21 24
179 Discoid Sandstone 85 42 8 20
334 Discoid Sandstone 505 38 8.5 18
428 Discoid Chalk 544 56.5 14 33
623 Discoid Reused pot 546 51 11.5 35
639 Discoid Reused pot 621 16 8 8.75
685 Discoid Reused pot 612 38.5 7.5 13
696 Discoid Sandstone 623 49.5 15 46
721 Discoid Reused pot 612 29.5 5.5 7
853 Globular Fired clay 698 36.5 29.5 24
895 Discoid Sandstone 748 44.7 13.5 32
937 Discoid Reused pot 796 35 8 16
982 Discoid Fired clay 975 29.7 12 11
1057 Discoid Reused pot 215 38 7 16
1510 Conical Fired clay 1517 39 23 28
1522 Discoid Reused pot 1537 57 12 43.75
1547 Globular Fired clay 1546 26 20 13
1639 Discoid Reused pot 1567 25.5 7 28
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
612 Triangular Fired clay 570 180 132 1423.75
689 Triangular Fired clay 689 130 158 1340
247
Highstead
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
40 Conical Fired clay Pit complex B15 40 20 52
41 Discoid Reused sherd Pit B146 40 6 45
Table 4: Loom weight data
Museum No. Tool Type Material Location Width (mm) Depth (mm) Weight (g)
70 Pyramidal Fired clay Enclosure ditch A24 109 109 1750
99 Pyramidal Fired clay Pit B168 70 100 1400
Sunset Caravan Park
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
121 Pyramidal Fired clay TR12 (29) 21 36 28
901 Globular Fired clay 2743 37 29 44
1245 Pyramidal Fired clay 157 32 27 24
1246 Globular Fired clay 2345 32 24 29
1322 Discoid Fired clay 1989 39 19.5 30
Thanet Earth
Table 6: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
47 Conical Fired clay 8800 33.5 27 14
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
22 Triangular Fired clay Context 8643 183.75 60 1177.5
25 Triangular Fired clay Context 8637 150 54 291
50 Triangular Fired clay Context 8795 157.5 70 1033.3
61 Triangular Fired clay Context 8620 155 56 801.7
127 Triangular Fired clay Context 12108 166 58.5 932.5
Appendix 1: Bronze and Iron Age textile finds from England and Scotland
Table 1: Catalogue of Bronze Age textile finds in England, including spin direction, type of weave, thread count, and thread diameter (after Crowfoot 1991:
table 8; DeRoche 2012; Walton Rogers 2002b, 2007).
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
Somerleyton
, Suffolk Burial
Socketed
axe Flax S/S, Z-ply Tabby 12/12 -
Linen fabric found stuffed in the socket of a bronze socketed
axe; matted appearance.
Manton Barrow Textile - - - 8-9/11-12 -
Rylston,
Yorkshire Burial Textile Wool - Tabby 12/10 -
Woollen textile; “wrapped…from head to feet, in oak dug-out
coffin 7 feet 3 inches long, 1 foot 11 inches wide and 1 foot
deep” (Elgee and Elgee 1949: 160)
Bush
Barrow,
Normanton
Burial Axe - Z/Z Tabby 12/11 - Mineralised textile; there appears to be two layers of thesame fabric (Henshall 1950: 133)
Pewit’s
Farm,
Berkshire
Burial Dagger Plant Z-ply/Z-ply Tabby 12-14/8 -
Mineralised textile; two different fabrics superimposed on
each other; textile is preserved on a bronze dagger, possibly
made from a plant fibre (Henshall 1950: 133)
Ridgeway Burial Axe - Z-ply/Z-ply - 16/16 -
Mineralised textile found on a bronze axe; noted as being one
of the finer textile in prehistoric Britain. Henshall notes that
the cloth appears closely woven and fairly even (1950: 133)
Ogbourn St.
Andrew Cremation Urn - S-ply/S-ply Tabby
11-12/
11-12 -
Little information given on the context or preservation of this
find (Henshall 1950: 133)
Durrington,
Wiltshire Cremation Urn - Z-ply/Z-ply Tabby 11-12/8 -
Little information given on the context or preservation of this
find (Henshall 1950: 133)
Must Farm Collapsedstructure Textiles
Over 180 items found, including fibres, textiles, twinning,
and knotted net; 10 loom weights and 8 spindle whorls also
found.
Table 2: Catalogue of Iron Age textile finds in England, including spin direction, type of weave, thread count, and thread diameter (after Crowfoot 1991: table
8; DeRoche 2012; Walton Rogers 2002b, 2007).
Location Context Item Fibre Spin
direction
Weave Thread
count/cm
Thread
diam. (mm)
Notes
Rudston
(makeshift
cemetery),
East
Yorkshire
TA 096692 –
095699
Burial R2 Brooch C2 - ?Z/S - c. 6? - A few traces remaining; clear threads present on the pin.
Burial R7 Brooch H1 - Z/Z - 5-? - Deteriorated remains.
Burial R13 Brooch F6 - Z/S 2/2 twill 4-5/5 1-1.5 A clear patch; the weave is open.
Burial R22 Brooch E1 - Z/? Threads - - Remains of threads and grass
Burial R24 Sword Ba1
- Z/S Tabby 6/4-5 1.5-2 Remains are below the hilt; textile had coarse fibres and aloose, open weave.
- - (Leather) - - Leather with impressions of the sword textile on its face.
Burial R27 Brooch E2 - Z/S Tabby 5/3 1-1.5 Deteriorated; Z-threads were tightly spun; fabric was folded.
Burial R34 Brooch C22 - Z/? Threads - - Threads were lying side-by-side; a possible fringe?
Burial R35 Brooch F1 - Z/? - - - Fabric was a possible border; evidence of a knot remains.
Burial R36 Brooch D1
-
Z/Z
2/2 chevron or
broken
diamond twill
5/5 1.3-1.5 Both sides of the fabric, around the edges, are folded.
Burial R37 Brooch G3 - S and Z ply Threads - - Impression on a possible leather surface.
Burial R38 Pin 2 - Z/Z Tabby 3/4 - Remains are on the ring of the pin.
Burial R39
Brooch A4 - Z/Z ? twill - - Remains include some Z-ply threads.
Pin 4 - Z ? threads - - Pin was wrapped by threads, possibly from a textile.
Burial R40 Brooch G4
- Z/Z 2/2 twill c. 7/6 c. 1 Possible remains of a border; even weave; evidence of a knot
?Z ply Threads - - These are finer threads; possibly weft threads.
Burial R46 Brooch D2 - Z/Z ? twill c. 5/4 - Traces of the surface survived.
Burial R60 Brooch E3 - Z/S 2/2 twill - - Surface was deteriorated.
Burial R69 Brooch D5 - Z/S ? - - Weave was damaged.
Burial R76 Brooch E5 - Z/?S ? threads c. 6/? c. 1 Threads were lying parallel; possibly a cord?
Burial R77 Brooch H2 - Z/Z ? tabby 5/5 - Two threads in one central shed; one thread possibly broken.
Burial R82 Brooch F8 - Z/S 2/2 twill c. 6/6 c. 1-1.5 A detached fragment; the Z threads were finer than others.
Burial R83 Brooch C3 - Z/?S ? - - Surface was deteriorated.
Burial R84 Brooch A2 - Z/? ? - - Surface deteriorated.
Table 2 cont.
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
Rudston
(makeshift
cemetery),
cont.
Burial R91 Brooch C4 - Z/? ? - - Majority was deteriorated, though some clear threadsremained.
Burial R97 Brooch D4 - Z/Z ? threads - c. 1.3 Threads lay side-by-side, suggestive of a fringe/border.
Burial R102 Brooch B1
- Z/Z, S ply ? tabby 4/2 (plied) c. 1.2 (2 forplied) Possible border/end; plied threads were tangled.
Z/Z ? 6/6 - Scattered and damaged fragments.
Burial R140 Brooch E6 - Z/S Threads - 1.3 Threads were lying side-by-side; possibly a fringe/border
Burial R143 Brooch G5 - Z/Z, S 2/2 twill 5-6/6 c. 1.5 Occasional S thread in weft system.
Burial R153 Dagger Bd2 - Z/? ? - - Remains included leather from sheath; textile traces on hilt.
Burial R154 Tongs - Z/Z Tabby 5/3 1.3-1.5 Weave was rather open.
Burial R174 Spearhead
A20
- - - - - Possible leather remains.
Burial R175 Brooch K1 - Z/S Tabby c. 6/8 c. 1.5 Patches on pinhead; possible leather under catch.
Burial R178 Brooch A3 - Z/Z 2/2 twill c. 5/6 c. 1.5 Broken threads found in remains.
Burial R180 Brooch C5 - Z/Z 2/2 twill c. 4-5/6 1.2-1.5 Damaged remains; covers 1 side of brooch and around pin.
Rudston
(Argan
Land),
East Yorks.
TA 096702
Burial R190 Brooch F2 - Z/S 2/2 twill ?/6 1.5-1.75 Loose and damaged remains; coarse Z threads around loop.
Burial R194 Brooch F3 - Z/Z 2/2 twill 7/4-5 c. 0.8-1 Threads are broken; preserved around brooch curve.
Burial R199 Brooch E7 - Z/Z Tabby - - Possibly a starting border; plied threads may be at the edge.
Burial R201 Brooch J1 - Z/Z ? twill - - Surface deteriorated.
Burial R202 Brooch F4 - Z/S 2/2 twill - - Surface distorted and flattened.
Burial R206 Brooch J2 - Z/S 2/2 twill c. 8/8 - Fragments were scattered.
Burton
Fleming, Bell
Slack, East
Yorks.
TA 106715
Burial BF41 Brooch E8 - Z/Z 2/2 twill c. 6/? c. 1 One thread per 2.5 round end.
Burial BF53 Brooch C20 - Z/? ? twill - c. 1.75 Threads woven/wound around pin.
Burial BF56 Brooch A5 - Z/S 2/2 twill - - Surface is damaged and patchy.
Burton
Fleming,
Argham Ln
TA 094703
Burial BF1 Brooch C21 - Z/S 2/2 twill 7/6 1.2-1.3 Folded on brooch head, double layer under pin.
Burial BF2 Brooch C6 - Z/Z 2/2 twill c. 8/8 - Around pin; deteriorated; most threads are damaged.
Burial BF4 Brooch C17 - Z/S ? - - Coarse fibres; deteriorated weave.
Table 2 cont.
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
Burton
Fleming,
Argham Ln
TA 094703
Burial BF5 Brooch C7 - Z/Z 2/2 twill c. 4/5 - Mass of parallel threads wrapped around pin; possible fringe.
Burial BF6 Brooch C8
- Z/S ? - - Remains on pin; deteriorated weave.
- Z-ply Threads - - Lump found under possible bone.
Burial BF7 Brooch C18 - Z/Z ? - - Surface is deteriorated.
Burial BF9 Brooch C9 - Z/Z 2/2 twill 4/4 - Patches of clear remains visible.
Burial BF10 Brooch C1 - Z/Z ? twill 6/5 1-1.2 A few degraded threads preserved around brooch head.
Burial BF11 Brooch C19 - Z/? - - - Mass is deteriorated.
Burial BF13 Brooch C10 - Z/Z 2/2 twill 4/4 - Two damaged threads around the pin.
Burial BF14 Brooch C11 - Z/Z 2/2 twill 6/6 - End/starting border present; weft ends are tied knots.
Burial BF15 Brooch C12 - Z/Z 2/2 chevron orbroken diamond 5/5 1.2-1.5 Reverse is preserved.
Burial BF18 Brooch C13 - Z/Z ? - c. 1.4 Found with traces of possible leather.
Burial BF19 Brooch C14 - Z/Z 2/2 twill 5/5 c. 1 Found with end/start border; broken loops.
Burial BF20 Brooch B2
- S/Z 2/2 twill,broken/damaged 9-11/9-10
(z) 0.7-0.9
(s) 1-1.3
Twill ‘ground weave’; possible mistakes made in weaving;
two 7mm repp stripes integrated in textile.
- Z/S Tabby (repp) 13/14 0.5-0.6 Wide textile; alternating spin directions in 1 thread system;embroidered crossbars preserved; striped effect on twill.
Kelverton,
Essex.
TL 87171782
Cremation
burial Sword blade
Flax/nettle Z/Z Tabby 14-16/? - Plant fibre textile, possible linen or nettle.
Flax/nettle Z/Z Tabby 12/12 - Plant fibre textile, linen or nettle; apparently looselyfolded/wrapped around sword.
Westhawk
Farm, Kent
Cremation
burial Patera Flax Z/Z Tabby
18-20/
12-14 -
Confidently identified as linen, fibres were preserved within
mineralisation.
St. Peter
Port,
Guernsey
Burial - Plant Z/Z Tabby 5/5 - Unidentified bast fibre.
Burial - Plant Z/Z Tabby 12/8 - Unidentified bast fibre, different from other textile’s fibre.
Richmond,
Guernsey Burial Sword blade Plant S/S Tabby 13/11 - Possibly plant-based textile, although this is unconfirmed.
River Nene,
near
Peterborough
Water
deposit Sword blade S/Z 2/2 twill - -
Table 2 cont.
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
Kirkburn,
East Yorks. Burial K5 Iron mail
Wool Z/Z ? twill 9-10/7-8 - Coarser fabric; possibly a tunic or gown.
Wool Z/S S/Z ? twill 12-14/9-10 - Mixed spin directions in both thread systems; tightly spun;possible cloak.
Skipwith
Common,
Yorks.
Barrow - Wool Z/Z Brokendiamond twill 5-6/5-6 -
Evidence of warp thread fringe; identified as wool; matted
remains; 20-30% of fibres are pigmented, suggesting use of
dyes.
Wetwang
slack, East
Yorks.
Burial 6/GE Brooch - Z/Z ? twill - - Deteriorated surface; coarse fibres.
Burial 6/LS Brooch - Z/Z 2/2 twill c. 5-64-5 - Coarse fibres; very open weave; possible cloak or blanket.
Burial 6/NJ Brooch - Z? S? ? twill 3/4 - Coarse fibres.
Helpringham,
Lincolnshire
Pit Briquetage - Z/Z Tabby 4/5 - Impression on clay; uneven weave with loose surface fibres.
Pit Briquetage - Z/S Tabby 6-7/5-6 - Impression on clay; fairly even thread throughout.
Pit Briquetage - ? Tabby 9/6 - Impression on clay; moderately coarse; possibly sacking.
Purberry
Shot - Pot - ? Tabby 7/6 - Impression on the side and base of a pot, dated c.300 BC
Table 3: Catalogue of Bronze Age textile finds in Scotland, including spin direction, type of weave, thread count, and thread diameter.
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
Sheshader,
Lewis Burial
Threads Cattle hair - - - - Cattle hair hanks with plaited horsehair bands attached to either side,with separate wool cord; very similar to finds from contemporary
Danish finds from the Etgved Girl and Skrydstrup Woman (Sheridan
1997: 189-190; Mannering et al. 2012). Dated to 1300-840 BC.Textile Horsehair - 2/2 twill - -
St. Andrews,
Fife Hoard
Textile Wool Z/Z Tabby 6/8 - Three textiles from the same coin hoard, used to plug the top of the
pot; the second textile is possibly hemp and the other plant textile is
not identified, possibly flax or nettle; dated to the Scottish Late
Bronze Age (Gabra Sanders 1994: 34-41).
Textile ?Hemp Z-ply/Z-ply Tabby 6/16 -
Textile Plant Z-ply/Z-ply Tabby 14/12 -
Nydie Mains,
Fife Burial Textile Bast Z-ply/Z Tabby 10-11/8-9 -
Textile found stuffed into the socket of a bronze dagger, possibly to
fit the tang into the handle (Henshall 1964: 197-8; Hedges 1974:
293-5)
Pyotolykes,
Angus Burial Textile Nettle
Z-ply/
Z-ply Tabby 10-11/8-9 -
Textile found stuffed into the socket of a bronze socketed axe; fibres
are positively identified as nettle (Henshall 1964: 197-8; Hedges
1974: 293-5)
Table 4: Catalogue of Iron Age textile finds in Scotland, including spin direction, type of weave, thread count, and thread diameter.
Location Context Item Fibre Spindirection Weave
Thread
count/cm
Thread
diam. (mm) Notes
House of Binns,
West Lothian
Cist
burial Brooch Wool Z/S Tabby 6/6 0.9-1.5
Mineralised textile identified as wool; remains of possible selvedge
with 10/6 threads per cm. Walton Rogers 2016: 109.
Dunbar, East
Lothian
Burial
1040
Brooch SF13 Wool Z/Z ? - - Dated to La Tène D1; coarse woollen textile remains.
Pin SF37 Plant Z/Z ? - - Textile remains on pin shank; unclear whether flax, nettle, or hemp.
Burial
1037
Pin SF12 Wool Z/Z ? 7/5 - Remains on ring-headed pin; coarse woollen textile; probably acloak. Appears to be the same textile as on sword SF9.
Sword SF9 Wool Z/Z ? - - Very deteriorated; coarse woollen textile; remains appear folded overblade; probably a cloak (Walton Rogers 2011: 191-2).
Craigie,
Dundee, Angus Brooch - Z/? 2/2 twill 6/5 - Late Iron Age (Bender Jørgensen 1992: 198).
Leith Hill,
Moredun, Perth Iron work - Z/S Tabby 10/6-7 -
Edged with a tablet-woven band (Bender Jørgensen 1992: 198;
Ritchie and Ritchie 1981: 98).
Balmaclellan,
Kirkcudbright Bog find
Bronze
objects Wool -
2/2
diamon
d twill
15/11 -
Very finely woven fabric; identified as wool; remains are
fragmentary and suggest the cloth was wrapped around the bronze
objects before deposition into the bog (Crowfoot 1947-48: 225-7,
230-31; Wild 1970: table A42; Bender Jørgensen 1992: 198).
Antonine Wall,
Falkirk, Sterling
Roman
coin
hoard
Coins - ? Brokentwill - -
Small fragments, apparently used to plug/cover the hoard’s opening;
alternating yellowish and brown checks suggestive of early tartan
(Crowfoot 1947-8: 227-231; Ritchie and Ritchie 1981: 119-120;
Bender Jørgensen 1992: 198).
Oakbank
Crannog, Loch
Tay
Collapse
d
Crannog
Textile Wool ? 2/1 twill 11-14/9-11 ?
A fine woollen textile, preserved in the loch’s anaerobic conditions;
dated to 6th century BC (Morrison 1985: 24-5, 81-98).
Dun Cuir, Isle
of Barra,
Inverness
Iron
fragment - - Tabby - - (Henshall 1955-6; Bender Jørgensen 1992: 198)
St. Andrew’s
Parish, Orkney Peat bog Hooded cape Wool Z/S
2/2
broken
chevron
twill
8-9/9 -
Evidence of repair and reuse, suggesting the original was larger;
woollen cape; single, heavily spun warp threads; two threads used
for weft; two tablet-woven borders of 2-ply yarn; lower band with
long fringes; dated to AD 250-615 (DeRoche 2012: 438-9; Henshall
1950; Ryder 1968: 12, 139; Gabra-Sanders 2001: 98-99).
Table 5: Bronze Age textile production evidence from Scotland.
Location Context Finds Date Notes
Dun Telve, Glenelg,
Inverness-shire Broch 5 spindle whorls
400 BC–AD
100
5 stone spindle whorls, one with decorated by small punctuations around the edge and radiating lines
on the faces; diameters measure 33–49 mm and the only recorded depth measures 28mm; no
information on the context or location of these finds (Curle 1916: 253-254)
Kebister, Shetland Settlement 12 spindlewhorls
400 BC–AD
400
7 made from steatite; 4 from the same Iron Age feature and 2 from disturbed features; similar to Iron
Age spindle whorls from other Scottish sites (Owen and Lowe 1999: 175-6, 282)
Table 6: Bronze Age textile production evidence from Ireland.
Location Context Finds Date Notes
Killymoon, Co.
Tyrone
Work-plac
e
12 spindle whorls,
polished Neolithic
stone axes reused
as linen smoothers
1000–800 BC
Spindle whorls found with textile fragments at a ‘work-place’, suggesting this may have been a group
activity; use-wear marks on the Neolithic axes suggest that by the 7th century AD (Late Iron Age/Early
Medieval period), these were used for smoothing linen (Wincott Heckett 2007)
Lough Gar, Co.
Limerick Settlement 2 spindle whorls 1000–800 BC
1 stone spindle whorl from occupation layer and 1 possible spindle whorl from a pit, identified by
excavators as a bead (Cleary 2003: 130-1, fig. 17)
Ballyveelish,
Co. Tipperary Settlement Spindle whorl 1000–800 BC 1 stone spindle whorl (Doody 1987: fig. 2:4)
Ballinderry, Co.
Offaly Settlement 3 spindle whorls 1000–800 BC Two bone and one sandstone spindle whorls excavated from occupation layers (Hencken 1942:14-16)
Knocknalappa,
Co. Clare Settlement Spindle whorl 1000–800 BC
Made from the head of a femur and found on the ‘breakwater’; no other details given (Raftery 1942:
57)
Dún Aonghasa,
Aran Islands Fort 2 spindle whorls 1000–800 BC (Wincott Heckett 2012: 431)
Moynagh
Lough, Co.
Meath
Crannog 2 spindle whorls 1000–800 BC (Wincott Heckett 2012: 431)
Site L/M,
Richmond, Co.
Tipperary
Outdoor
cooking
area
2 spindle whorls 1000–800 BC (Wincott Heckett 2007: 32)
Coarhamore,
Co. Kerry
Outdoor
cooking
area
2 spindle whorls 1000–800 BC (Wincott Heckett 2007: 32)
Appendix 2: Bronze and Iron Age textile finds from Northern Europe
Table 1: Bronze Age textile finds from Northern Europe and Scandinavia, including spin direction, type of weave, thread count, and thread diameter.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Sweden
Eldsberga prästgård,
Halland Oak coffin Textile Wool s/? Tabby 4/4
Found alongside several threads, probably from another
fabric (Franzén et al. 2012: Table 17.1)
Dömmerstorp,
Hasslöv, Halland
Urn burial Textile Wool s/z Tabby 4/4 -
Stone cist Textile Wool s/z Tabby 4-5/3 Textile has ‘distinct volume’ with various thread diameters.
Stone cist
Textile Wool z/s Tabby 6/4 -
Textile Wool z/s Tabby 6/4 -
Textile Wool z/s Tabby 6/4 -
Textile Wool z/s Tabby 6/4 -
Textile Wool s/z Tabby 4-5/3 ‘Distinct volume’ noted again
Textile Wool z/s Tabby 6/3-4 Uneven weave noted; thread diameters 1.5-2 mm in bothsystems
Textile Wool z/s Tabby 6/3-4 Thread diameters of 1.5-2 mm in both systems
Hasslöf 13, Hasslöv,
Halland Mound
Textile Wool s/s Tabby 4/4 Textile is disintegrating; thread diameters of 1-2 mm in bothsystems
Textile Wool s/? Tabby 12/2-3 Densely warped, with a possible starting border c.2cm;thread diameters of 1-2 mm
Lugnaro, Hasslöv,
Halland Mound Textile Wool s/? Tabby - Very disintergrated; thread diameters of 1.5-2 mm
Laholm, Halland Urn burial
Textile Wool s/z Tabby 5/4 Very flattened
Textile Wool s/z Tabby 5/4 Almost square tabby, very flattened; thread diameters of1.5-2 mm in both systems
Textile Wool s/z Tabby 5/4 -
Björlvö, Skåne Mound Textile Wool s/z Tabby 5/4 Discoloured; threads are ‘close to each other’
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Sweden
Björlvö, Skåne Mound
Textile Wool - Tabby 5/4 Appears felted
Textile Wool - Tabby 4-5/4 -
Textile Wool s/z Tabby 4-5/4 -
Hammarlöv by,
Skåne Mound
Textile Wool s/z Tabby 5/4 -
Textile Wool s/z Tabby 5/4 Probably the same fabric as the others found in the samecontext; fulled cloth
Textile Wool s,z Tabby - -
Stora Köpinge,
Skåne Oak coffin Textile Wool s/z Tabby 5/4
Finer threads in system I; threads in system II also have ‘more
fullness’
Vestra Grevie,
Mellan-Grevie,
Skåne
Oak coffin Textile Wool s/z Tabby 4-5/3 Distinct volume to cloth, found alongside several pieces ofthread
Oak coffin Textile Wool s/z Tabby 5-6/4 z-twisted thread is thicker and looser than s-twisted thread
Unprovenanced,
Skåne - Textile Wool s/z Tabby 3/3 Plaited starting border with textile; thread diameters of 2 mm
Glädjebacken,
Trelleborg, Skåne Oak coffin
Textile Wool s/s Tabby 5/4 ‘Distinct interstices between the binding threads’
Textile Wool s/s Tabby 5/5 Probably the same fabric as above, but more compressed
Österslöv, Skåne Mound Textile Wool s/z Tabby 5-6/4-5 -
Denmark
Benestad, Aringsås,
Småland Stone cist Textile Plant? s?/z? Tabby 13/12 Type of fibre unconfirmed but probably plant
Borum Eshøj A Inhumation Textile Wool z/s Tabby - Dated to 1351 BC
Borum Eshøj C Inhumation Textile Wool z/s Tabby - Dated c.1500-1300 BC
Trindhøj Inhumation Textile Wool z/s Tabby - Dated to 1347 BC
Nybøl Inhumation Textile Wool z/s Tabby - Dated to 1266 BC
Hvidegaard Stone cist Textile Wool z/s Tabby - Dated c. 1300-1100 BC
Lusehøj Cremation
Textile Wool z/s Tabby - Dated to c.950–750 BC, called a woollen rug.
Textile Nettle z/s Tabby - Described as a fine textile, wrapped around human bones.
Norway Jæren, Rogaland Mound Textile ? z/s Tabby 4-6/4-6 Given the similarities to Danish finds, mostly likely wool
Table 2: Catalogue of La Tène textile finds from Northern Europe and Scandinavia.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Sweden
Fole prästgården,
Gotland Grave - Wool? - - -
-
Hjortmossen, Vg Bog find - Wool S/S 2/2 broken twill 7/8 Warp and weft yarn is overspun; irregularities in weave
Lilla Jored, Bo Burialchamber
- Wool S/Z 2/2 twill 14/12 Two fabrics mixed together
- Wool Z/S 2/2 twill 16/14 -
Bläsnungs, Gotland Grave - Wool S/S 2/2 twill 8-9/7 -
Vallviksudden,
Vilhelmina Settlement - Wool S/Z 2/2 twill 9/8
Long fibres
Horsta, Me Grave - Wool Z/Z
2/2 diamond
twill 14-16/14
Very fine fabric, same as in second grave
Grave - Wool Z/Z 14-16/14 Very fine fabric, same as in first grave
Bodarps by, Skåne Östergötland - Wool? - - - -
Öremölla, Skåne Burialchamber
- Wool Z ply/? Tablet weave 100/20 -
- Wool Z, S - - -
- Wool Z/S Diamond twill 29-30/28-29
-
- Wool? - Tablet weave - Very fine quality
- Wool Z/S Diamond twill 28/28 -
Lilla Vi, Smäland Mound - Wool Z/S 2/2 twill
15-16/
13-14
-
Inhumation - Wool Z ply/Z 2/2 twill 16/14 Thread system I has more fullness
Söderby, Uppland Cairn
- Wool Z/S 2/2 twill 15-16/14 -
- Wool Z/Z 2/2 twill 12/10 -
- Flax? Z/Z Tabby 14-15/14-15
-
Breneberga, Vg Stone setting - Wool Z/Z Threads - Probably 2/2 twill- Wool Z/Z 2/2 twill 12/10 Thicker and longer threads in thread system I
Kinna, Vg Stone setting - Wool Z/Z 2/2 twill 14-16/12 -
Borgby, Vs - - - - - - -
Skedemosse, Öl Sacrifice,deposit - Wool? Z/Z 2/2 twill 8/8
-
Sörby Störlinge, Öl Stone setting - Wool? S/S 2/2 twill 14-16/12 -
Södra Kvinneby, Stone cist - Wool? S/S 2/2 twill 9-10/9-10 Quite coarse
Table 2 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Sweden
Viby Bosgård,
Östergötland
Mound,
cremation
- Wool Z/? Tablet weave? 32/? -
- Wool? Z/? Repp 26-27/10-12
Very compact warp, weft is invisible
Alvastra, Östergötland Inhumation - Wool? Z?/S? 2/2 twill 14-16/14-16
Fulled cloth of very fine quality
Denmark
Borremose Peat bog
Textile Wool S/Z 2/2 twill 5/2 Fragment; crossing weft threads
Textile Wool S/S 2/2 twill 4/5 Blanket; selvedges present, woven on tubular loom;frequent crossing wefts
Krogens Mølle Mose,
Torslev Peat bog
Textile Wool - Tabby 9/6
Finely spun wool; fragment measures 40x32 cm; checked
pattern.
- Tabby 8/7 Part of same textile as above; measuring 35x32 cm.
Textile Wool S/S 2/2 twill 9/8 Very damaged; faint checks remaining.
Textile Wool S/S 2/2 twill 3-4/3 Very fragmentary; remains of plaited selvedges
Textile Wool S/S 2/2 twill 9-11/7-8 Crossing weft threads in shed; tubular-woven selvedges; c.3.68 m in length
Ømark, Durup Peat bog Textile Wool S/S 2/2 twill 3/3 Blanket; crossing wefts occur; fringe on all four sides
Fræer Mose Peat bog Textile Wool S/S 2/2 twill 4/4 Fragmentary; plaited border; crossing weft threads in shed
Ræbild Skovhuse Peat bog Textile Wool S/S 2/2 twill 6/6 Large piece; gores woven in; crossing weft threads in theshed; evidence of mending and pre-deposition wear
Bredmose (Arden) Peat bog
Textile Wool S/S 2/2 twill 3-4/2 Coarse cloth; seam along centre; tabby selvedge on eitherside
Textile Wool S/S 2/2 twill 8/6-7 Large pieces; tubular-woven tabby selvedges; thick woollencord threaded through warp loops at both borders
Textile Wool S/S 2/2 twill 8/8
Scarf; fringe along both short edges; tubular-woven tabby
selvedge; striped lengthways; three stripes across the scarf,
one at each edge and one in middle
Stockholm Peat bog Textile Wool S/S 2/2 twill 6/5 Many crossing weft threads.
Sørgaards Mose,
Daugbjerg Peat bog
Textile Wool S/Z 2/2 twill 4/4 Remains of cloth from shoe remains
Textile Wool S/S 2/2 twill 12/7 Leg wrapper; decorated with stripes; selvedges in tact;
Textile Wool S/S 2/2 twill 12/10 Leg wrapper; selvedges preserved; hone side is hemmed.
Grathe Hede Peat bog Textile Wool S/S 2/2 twill
7/9 Three fragments of the same textile; various shades of
brown and occasional hints of check; very similar to textile
from Krogens Mølle bog.
9/9
8/6
Aalestrup Mose Peat bog Textile Wool Z/Z 2/2 twill 6-7/7-8 Three pieces from the same textile
Table 2 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Denmark
Huldremose Peat bog
Textile Wool S/S 2/2 twill 9/7 Checked skirt
Textile Wool S/S 2/2 twill 8/7 Checked scarf; fringed
Textile Wool S/S 2/2 twill 7/6 Band.
Møgelmose Peat bog Textile Wool Z/S 14/10 -
Haraldskjær Peat bog
Textile Wool Z/Z 2/2 twill 10/7 Largest fragment; crossing weft threads; two closedborders; one tubular selvedge
Textile Wool Z/Z 2/2 twill 10/8 Scrap; finely striped in both thread systems; one hemmededge; one fringed side
Moselund Peat bog Textile Wool Z/S ? twill 14/10 Garment made of several panels; pollen to Late IronAge/Early Middle Ages; woven gores.
Thorsbjerg Peat bog
Textile Wool Z/S 2/2 twill 7/? Missing threads, possibly plant fibre; 2nd to 5th century AD
Textile Wool Z/S 2/2 twill 8/8 One transverse border preserved
Textile Wool Z/S Tablet weaving 11/9 Fragment of a band
Textile Wool Z/S 2/2 twill 11/10 Tablet-woven border woven into selvedge
Textile Wool Z/Z 2/2 twill 13/10 Tablet-woven border
Textile Wool S/S Tabby 6/6 Tablet-woven border with spin pattern warp
Corselitze Peat bog
Textile Wool Z/Z 2/2 twill 12-14/11 Blanket; tablet-woven starting border; 3rd century AD
Textile Wool Z/S-ply Tablet weave ? Six four-thread tablets; plain weave
Textile Wool Z/Z 2/2 twill 8/8 Parts of two selvedges left
Borremose Ditch Textile Wool S/S 2/2 twill 8/7 Very torn; found in ditch of fortified camp; 3rd century BC
Tornebuskehøj Barrow Textile Wool S/S 2/2 twill ? Repp stripe; dyed with woad; first half 1st century AD
Brænde Lydinge Grave
Textile Wool Z/S 2/2 twill 12/14
Remains of tablet-woven border on one edge; second half
1st century AD
Z/S 2/2 twill 15/14 Probably belongs to same textile as above
Textile Wool S/S 2/2 twill ? Piece contains two weaves, twill and repp; threads arepaired in the repp
Nortvig Grave Textile Wool S/S 2/2 twill 11/10 Very small fragments of the same cloth; very fine andloosely woven; 1-2nd century AD
Brokær Cremationburial Textile Wool S,Z/S,Z 2/2 twill 23/19
Mineralised remains; found wrapping cremated remains and
placed in a bronze cauldron; spin patterning; second half
2nd century AD
Sneumgaard
Possible
cremation
burial
Textile Wool S/Z 2/2 twill ?
Scraps of textile recovered from cauldron-shaped hollows;
white and brown wool used; poorly preserved; 3rd century
AD
Table 2 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Denmark
Hjørring Barrow
Textile Wool Z,S/Z,S 2/2 twill 13/15 Two fragments; stripes in one direction; spin pattern; tracesof woad found.
Textile Wool Z,S/Z,S/ Brokenlozenge 22/20
Fragmentary; traces on woad found
Textile Wool Z,S/Z,S Brokenlozenge 15/14
Fragmentary; traces on woad found; coarser than other
fabrics
Donbæk
a Textile Wool S,Z/S,Z 2/2 twill 7/6 -
g Textile Wool S,Z/S,Z 2/2 twill 20/18 -
5805 Textile Wool S,Z/S,Z 2/2 twill 17/16 3rd century AD
Vrangstrup
Ia Textile Wool Z,S/Z,S Brokenlozenge 18/14
4th century AD
IIIa Textile Wool Z/Z,S 2/2 twill 13/11 -
IIIc Textile Wool Z,S/Z,S Brokenlozenge 19/16
-
Jelling - Textile Wool Z,S/Z,S 2/2 twill 9/5 -
Appendix 3: Iron Age textile finds from Central Europe
Table 1: Catalogue of La Tène textile finds from Central Europe.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Austria
Dürrnberg
Salzberwerk,
Hallein
- Textile Wool Z-ply/Z 2/2 twill 11/12 Textile fragments found in saltmines; plied yarn used in warp;
to La Tène A (450–380 BC).- Textile Wool Z-ply/Z Tabby 9/9
Schrattenberg,
Mistelbach - Arm ring Flax? Z/Z Tabby 19/18
Textile found in hollow bronze arm ring; to La Tène B
(380–250 BC).
Switzerland
Andelfingen, Kt.
Zürich
Grave 12 Pointed object - Z/Z Tabby 15/12 Mineralised textile remains on iron; to La Tène Tène B
Grave 21 Ring - Z/Z Tabby 15/12 Preserved on iron; to La Tène B.
Grave 29 Object - Z/Z Tabby 13/15 Mineralised textile remains on iron; to La Tène period.
Arbedo-Castion
e, Kt.Ticino
Grave 32 Belt hook - Z,S/Z 2/2 twill 7/7 Textile remains found on back of iron hook; warp uses variousspin directions; to La Tène A.
Grave 45 Fibula - Z/? 2/2 twill 8/7 Iron fibula, faint textile remains; La Tène B.
Grave 108 Fibula - Z,S/Z 2/2 twill 10/10 Faint textile remains on brooch pin; warp uses various spindirections; La Tène A.
Grave 15 Fibula - Z,S/S 2/2 twill 10/9 Bronze fibula, textile remains under head; warp uses variousspin directions; La Tène A/B.
Grave 140 Belt hook - Z/? 2/2 twill 11/8 Mineralised textile on iron; La Tène A/B.Fibula - Z?/Z? Tabby 6/6 Iron fibula, textile remains under foot; La Tène B
Grave 149 Fibula - Z/Z 2/2 twill 6/6 Iron fibula, textile remains on head; La Tène B
Grave 150 Fibula - Z/Z 2/2 twill 8/6 Iron fibula, textile remains under spiral; La Tène B
Arbedo-Molinaz
zi, Kt. Ticino
- Belt hook - Z/Z 2/2 twill 7/8 Mineralised textile on iron; La Tène A/B
Grave 59 Fibula - Z/S 2/2 twill 6/6 Mineralised textile on iron; La Tène A/BFibula - Z/Z 2/2 twill 6/6 Mineralised textile on iron; La Tène A/B
Grave 60 Fibula - Z/Z 2/2 twill 8/7 Large iron fibula, textile remains on ball; La Tène C? (250–150BC)
Grave 69 Fibula - 6Z,6S/6Z,6S 2/2 twill 11/11
Iron fibula, textile remains on foot; spin direction patterning
also used; La Tène A/B
Grave 70 Ring - 4Z,4S/Z 2/2 twill 9/8 Mineralised textile remains on iron; spin direction patterningused in warp; La Tène A/B
Grave 82 Fibula - Z/Z 2/2 twill 8/6 Iron fibula, textile remains on foot; La Tène A/B
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Switzerland
Arbedo-Molinaz
zi, Kt. Ticino
Grave 87 Fibula - Z,S/Z 2/2 twill 10/10 Mineralised textile on iron; warp uses various spin directions;La Tène A
- Fibula - Z/Z 2/2 twill 8/7 Iron fibula, textile remains under head; La Tène A/B
- Fibula - Z/Z 2/2 twill 8/8 Iron with textile remains; La Tène B
Arbedo-S.,
Paolo, Kt.
Ticino
Grave 1 Ring - - 2/2 twill 8/7 Mineralised textiles on iron; La Tène A
Grave 3 Fibula - Z/Z 2/2 twill 8/7 Mineralised textiles on iron; La Tène A
Kastell,
Bellinzona - Fibula - Z,S/Z 2/2 twill 7/8
Iron fibula, textile remains under pin; warp uses various spin
directions; La Tène A/B
Castione-Berga
mo Grave 14 Fibula - - 2/2 twill 10/9 Mineralised textiles on iron; La Tène A/B
Frauenfeld-Lang
dorf Grave 3 Fib - Z/Z Tabby 6/5 Iron fibula, textile remains on underside; La Tène C
Giubiasco
Grave- 24 Belt hook - Z/Z 2/2 twill 10/8 Faint textile remains on iron; La Tène B
Grave 51 Fibula - Z/S Tabby 6/6 Iron fibula, faint textile remains under head; La Tène B
Grave 61 Fibula - Z/Z 2/2 twill 6/7 Iron fibula, textile remains under spiral; La Tène A/B
Grave 68 Belt buckle - Z,S/Z 2/2 twill 8/8 Mineralised textiles on iron; warp uses various spin directions;La Tène B/C
Grave 78 Fibula - Z/Z 2/2 twill 10/8 Mineralised textiles on iron; La Tène B
Grave 79 Fibula - Z/Z 2/2 twill 7/6 Mineralised textiles under iron fibula head; La Tène B
Grave 109 Fibula - Z/Z 2/2 twill 7/8 Iron fibula, textile remains under head; La Tène C
Grave 110 Object - Z/? - 9/? Mineralised textile over iron object; La Tène A
Grave 122 Ring - Z/? 2/2 twill 10/8 Mineralised textiles on iron ring; La Tène A/B
Grave 145 Fibula - Z/Z 2/2 twill 8/7 Iron fibula, textile remains under head; La Tène C
Grave 147 Fibula - Z/S 2/2 twill 9/7 Iron fibula, textile remains under foot; La Tène B/CFibula - Z/Z 2/2 twill 9/6 Mineralised textiles on most of iron fibula.
Grave 161 Fibula - Z/? 2/2 twill 8/8 Iron fibula, textile remains under foot; La Tène B
Grave 164 Fibula - Z/S 2/2 twill 8/7 Iron fibula, textile remains under foot; La Tène B/C
Grave 177 Fibula - Z/S 2/2 twill 6/5 Mineralised textiles on iron; La Tène B
Grave 194 Fibula - Z/Z Tabby 6/8 Mineralised textiles on iron; La Tène B/CKnife - Z/Z Tabby 14/9 Iron with textile remains.
Grave 211 Fibula - Z,S/Z 2/2 twill 8/6 Iron fibula, textile remains under foot; warp uses various spindirections; La Tène A/B.
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Switzerland Giubiasco
Grave 213 Fibula - Z/Z 2/2 twill 8/5 Mineralised textiles on iron bow; La Tène B.
Grave 216 Fibula - Z,S/Z,S 2/2 twill 9/8 Iron fibula, faint textile remains under foot; warp and weft usevarious spin directions; La Tène B/C
Grave 222 Fibula - Z/Z Tabby 5/5 Iron fibula with textile remains; La Tène D
Grave 227 Fibula - Z/Z Tabby 5/5 Iron fibula, textile remains on bow; La Tène B
Grave 243 Fibula - Z/Z 2/2 twill 6/7 Iron fibula, textile remains under head; La Tène A/B
Grave 305 Fibula - Z/Z 2/2 twill 8/6 Mineralised textiles on iron fibula foot; La Tène B/C
Grave 398 Fibula - 4Z,4S/4Z,4S 2/2 twill 9/8
Iron fibula, textile remains around pin and foot; spin direction
patterning also used; La Tène B
Grave 404 Belt - Z/Z Tabby 6/7 Iron belt? with textile remains; to La Tène period
Grave 444 Fragment - Z/Z Tabby 8/9 Iron fragment with textile remains; La Tène B/C
Fibula - Z/Z Tabby 7/6 Iron with textile remains
Grave 477 Fibula - Z/? Tabby 6/5 Mineralised textile on iron; to La Tène period
Grave 498 Fibula - Z,S/Z 2/2 twill 8/6 Iron with textile remains; warp uses various spin directions;La Tène B.
Grave 501 Fibula - Z/Z 2/2 twill 7/6 Mineralised textile on iron fibula; La Tène A/B
Grave 510 Pendant - Z/Z 2/2 twill 10/6 Mineralised textile on bronze pendant; La Tène A/B
- Fibula - Z/? 2/2 twill 8/6 Iron with textile remains under spiral; La Tène B
Misox - Sword - Z/Z ? twill 5/5 Mineralised textile on iron sword; La Tène B?
Germany
Budenheim - Fragments - Z/? 2/2 twill 10/10 Iron fragments with textile remains; to La Tène period
Schwabsburg - Fibulafragment - Z ply/Z Tabby 11/10
Mineralised textile on side of iron bow fibula fragment; plied
yarn used in warp; La Tène A.
Udenheim - Fragment - Z ply/Z Repp/tablet? 19/9 Iron fragments with textile remains; plied yarn used in warp;La Tène B/D
Ülversheim Grave III Pincer - Z/Z - 7/6 Iron with textile remains; La Tène periodShears - Z/Z Tabby 9/7 Iron with textile remains
Breungen-born Tumulus34, grave 9
Razor - Z/Z Tabby 5/7 Textiles wrapped around iron; to La Tène A
- Z ply/Z Tabby 10/7 Plied yarn used in warp
Ring - Z ply/Z Tabby - Iron with textile fragment; plied yarn used in warp
Hasborn - Fibula - - Tabby 10/10 Iron fragment with textile remains; La Tène B/D
Hoppstädten Tumulus Textile - Z/Z ply Tabby 9/10 Cakes of corroded textiles; plied yarn used in weft; La Tène ARing - Z ply/Z 2/2 twill 9/9 Bronze with textile remains; plied yarn used in warp
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Germany
Trier
Paulinstrasse - Sword - - 2/2 twill 8/8 Iron, textile remains on scabbard; La Tène A
Wallscheid - Bowl Wool Z ply/Z Tabby 17/12 Several textile fragments from bronze bowl; plied yarn used inwarp; La Tène A
Schwenderöd Tumulus 3 Fragment - Z/Z Tabby 11/10 Iron fragment with textile remains; La Tène A
Rickofen Grave 6 Fibula - Z/Z? 2/2 twill? 10/10 Iron, with tiny textile remains; La Tène B2
Appendix 4: Iron Age textiles from Eastern Europe
Table 1: Catalogue of La Tène textile finds from Eastern Europe.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Czech
Republic
Unknown Iron Sword ? Z ply/Z 2/2 twill 7/7 Iron sword and mount with textile remains; plied yarn used inwarp; broadly dated to La Tène period; fibre unidentifiable.
Blučina,
Brno-Venkov,
Moravia
Grave 11 Iron sword Wool Z/Z Tabby 9/6 Textile remains with sword and sheath; La Tène B (380– 250BC).
Grave 13 Iron fragment - Z/Z Tabby 7/5 Textile remains dated to La Tène C (250–150 BC).
Grave 16 Lance - Z ply/Z 2/2 twill 8/8 Plied yarn used in warp; remains dated to La Tène C/D (250–1BC).
Brno-Maloměřice,
Brno-Mèsto,
Moravia
Grave 48 Hollow bronzering Flax Z/Z Tabby 11/6 Hollow ring filled with textile;dated to La Tène period.
Grave 67 Hollow bronzering Flax Z/Z Tabby 14/15 Hollow ring filled with textile;dated to La Tène B/C.
- Iron fragment Wool Z/Z Tabby 5/4 dated to La Tène B.
Bučovice, Vyškov,
Moravia
Grave 14 Bronze rings Flax Z/Z Tabby 15/14 Bronze rings found with iron fragment found with textileremains; dated to La Tène B.
Grave 20
Bronze fibula - Z/Z - - dated to La Tène B.
Bronze ring - Z/Z Tabby? 6/8 Possibly made by the same textile as that which was with thefibular; mineralised textile fragments.
Holubice, Vyškov,
Moravia
Grave 8 Iron ring Wool Z/Z Tabby 9/7 Iron has mineralised textile fragment; dated to La Tène B/C.
Grave 21 Chainfragment Wool? Z/Z Tabby 9/8
Chain made with bronze and iron; several fragments found;
dated to La Tène C.
Grave 35
Sword
fragment - Z/Z Half-basket 10/13 Iron sword fragment; dated to La Tène B.
Iron
sword/knife Wool? Z/Z Tabby 4/4 Mineralised textile fragments.
Grave 50 Swordscabbard Wool? Z/Z Tabby 5/4 dated to La Tène B.
Grave 56
Fibula Flax? Z/Z Tabby 18/15 Iron fibula with mineralised textile; dated to La Tène C.
Fibula Wool? Z/Z ? twill 8/8 Iron fibula with mineralised textile; different textile from otherfibula
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Czech
Republic
Holubice, Vyškov,
Moravia Unknown Chain - Z/Z 2/2 twill 8/7
Fragments of chain made with bronze and iron; dated to La
Tène B.
Křenovice u
Slarkova, Vyškov,
Moravia
Grave 12 Fibula Wool? Z/Z Tabby 6/5 Iron fibula with mineralised textile on back; dated to La TèneB.
Grave 17 Arm ring Flax Z/Z Tabby 9/9 Bronze fragments of arm ring, wrapped in textile; dated to LaTène B.
Postoupky,
Kroměříz,
Moravia
-
Sword Wool S/S Tabby 5/5 Iron sword with mineralised textile; dated to La Tène period.
Slavkov, Vyškov,
Moravia
- Button - Z/Z 2/2 twill 8/8 Iron button with mineralised textile; dated to La Tène B.
- Fragment Flax? Z/Z Tabby 14/15 Iron fragments; dated to La Tène period.
St. Břeclav,
Břeclav, Moravia
- Shears? Wool Z/Z Tabby 5/4 Iron find, possibly shears; dated to La Tène period.
Telnice,
Brno-Venkov,
Moravia
-
Fragment - Z/Z Tabby 14/14 Iron fragment; dated to La Tène period.
Židenice, Moravia
- Fragment Wool? Z/Z Tabby 12/9 Iron fragments; dated to La Tène period.
Fragment - Z/Z Tabby 12/9 Iron fragments; probably the same textile as on otherfragment.
Unknown,
Moravia
- Fragment Wool? Z/Z Tabby 10/7 Iron fragments; dated to La Tène period.
Fragment Flax? Z/Z Tabby 15/16 Iron fragments; dated to La Tène period.
Slovakia
Dolný Peter,
Komárno
Grave 55 Hollow ring Flax? Z/Z Tabby 14/14 Textile found within hollow ring; dated to La Tène period.
Grave 58 Textile Flax Z/Z Tabby 24/14 dated to La Tène period.
Hurbanovo,
Komárno Grave 1/80 Ring - Z/Z Tabby 24/22 dated to La Tène period.
Nitra - Arm ring - Z/Z Tabby 16/9 dated to La Tène period.
Nové Zámky Grave 9/70 Textile Flax Z/Z Tabby 24/13 Unclear whether textile are from the same or separate fabrics;dated to La Tène A/B.
Vělká Maňa,
Vráble Grave 362
Textile Flax Z/Z Tabby 20/16 -
Hollow ring Flax Z/Z Tabby 30/14 -
Textile Flax Z/Z Tabby 30/19 -
Hollow arm
ring Flax Z/Z Tabby 24/13 -
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Hungary
Vělká Maňa,
Vráble Grave 362 Sword Wool? Z/Z Tabby 4/3 Dated to La Tène period.
Kismezö Tumulus Buckle - Z/ZZ Half-basket 12/14 Textile remains found on half of iron ring buckle; toHallstatt/La Tène transition.
Bodroghlom,
Borsod-Abaúj-Za
mplen
Grave 20 Fragments - Z/S, Z? ? twill 10/12 Iron fragments preserving textiles; warp uses various spindirections; dated to La Tène C.
- Textile Wool - - - Piece of felt, dated to La Tène C.
Grave 10 Knife - Z/Z 2/2 twill 2/2 twill Iron knife preserving textile, dated to La Tène C?
Kosd, Nógrád
- Razor Wool? Z/Z Tabby 6/6 Textile fragment preserved on sickle-shaped razor; dated to LaTène D.
Grave 43 Knife Wool Z/Z Tabby 5/5
Textile remains wrapped around iron knife; dated to La Tène
period.
Wool Z/Z Tabby 9/6 Fragment preserved on the back of the iron knife.
- Fibula Flax? Z/Z Tabby 23/17 Textile preserved on fibula,dated to La Tène C.
Grave 54 Knife - - Tabby 22/20 Iron knife preserving textile remains;dated to La Tène B.
- Ring - Z/Z Tabby 16/12 Textiles preserved on bronze ring;dated to La Tène period.
Szentes-Vekerúg,
Csongrád - Disc - Z/ZZ Half-basket 13/20
Textile impressions preserved on the back of bronze disc;dated
to La Tène D1
Csanylelek,
Ujhalasto Grave 44 Ring - Z/Z Tabby Tabby Iron ring preserving textile; to Hallstatt/La Tène transition .
Slovenia
Insel Veglia,
Bescanova Grave 3 Fibula - Z,S/Z 2/2 twill 14/12
Bronze fibula with textile remains preserved around head;
warp uses various spin directions; to Hallstatt/La Tène
transition .
Erdöd, Nirobititz,
Krain - Lance - Z/Z Tabby 11/10 Lance head with textile remains;dated to La Tène period.
Roje, Moräntsch,
Krain Grave 1 Knife - - Tabby 16/14 Faint textile remains on iron knife;dated to La Tène period.
Ukraine
Kharkiv Kurhan - Wool S/S ? 11/15 Reddish-brown colour; threads dyed pre-spinning with madderand alum mordent; found with piece of felt; c. 800-750 BC
Nymphaion,
Crimea
Tomb,
grave 1(b) - Wool
Weft-faced
tabby 14/60-70
Brown in colour, found with multiple textile fragments and a
mess of loose threads; made from very fine wool; c. 5th
century BC
Heyevy Mohyly,
Poltava Kurhan 1
- Plant? Z/Z Tabby 8/8 Threads measure 0.5-0.6mm in diameter; c. 5th century BC
- ? S/S Tabby 14/28 Threads measure 0.7-1.1mm in diameter; to c. 5th century BC.
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Ukraine
Taman, Seven
Brothers, k. IV
- - Wool - Tabby 16/21 Resist-painted; red with black background and red and brownpattern; 20 fragments total; c. 5th century BC.
- - Linen - Tabby 14/8 One fragment; c. 5th century BC.
Taman, Seven
Brothers, k. VI
- - Wool - Tabby 20/28 Resist-painted; black background with polychromatic patterns;50 fragments total; c. 4th century BC.
- - Wool - Tapestry 28/50 Polychrome duck patter on red background; fur fragmentattached; three fragments; c. 4th century BC
Horodyshche,
Kharkiv Burial
Bronze mirror
with iron
handle
- - Tabby 7-8/7-8 Mineralised textile; c. 4th century BC
Pisochin, Kharkiv Kurhan 6 Lamellarcuirass
? Z-ply/Z-ply
Weft-faced
tabby 4/14
Mineralised textile fragments on right side of cuirass; thread
diameter of 1mm; c. 400-350 BC
? ? Tabby 9/9 Mineralised textile on left side of cuirass; thread diameter of1mm.
Hryshkove
cemetery, Kharkiv
Burial 1,
Kurhan 25
Quivers, spear
butt Wool Z/Z Tabby 8/10
Mineralised textile fragments on iron; threads have a hard twist
angle and diameter of 1mm; same textile on each object; c.
5th-4th century BC.
Kerch, Pavlovskij Burial 2 - Wool - Tabby 21/24 Red with polychrome embroidery; 7 fragments total; c. 4thcentury BC
Kerch, Ak-Burun - Plant? - Repp 3/10 Very coarse warp; 20 fragments total; c. 4th century BC
Kerch, Mitridat Burial 40 - Wool - Tapestry 12/33 Green, brown, and yellow colours; 4 fragments total; c. 1stcentury BC
Nogaichik, Crimea Burial 18,kurhan 5
- Silk - Tabby 50/60 Possible sleeve dress; 1 century BC - AD 1st century
- Silk - Tabby 70/70 Very open, gauze weave
Kerch,
Karantinnaya gora Burial 1
- Wool - Tabby 11/12 Brown colour; four fragments total; c. 1st century AD
- Wool - Repp 18/28 Brown colour; single fragment; c. 1st century AD
- Wool - Weft-facedtabby 14/21 Brown colour; single fragment; napped surface
- Wool - Tabby 30/30 Brown colour; single fragment; napped surface; c. 1st centuryAD
Kerch, Mitridat no.
4 - - Wool -
Weft-faced
tabby 23/28
Brown and red stripes, black background; single fragment; c. 1st
century AD
Kerch, Mitridat no.
2 - - Wool - Repp 10/24
Brown, green, yellow stripes on red background; single
fragment; c. 1st century AD
Table 1 cont.
Country Location Context Item Fibretype
Spin
direction Weave
Thread
count/cm Notes
Ukraine
Kerch, Mitridat no.
2
- - Wool - Weft-facedtabby 17/23 Reddish-brown colour; single fragment; c. 1st century AD
- - Wool - ? ? Greenish-brown colour; single fragment; c. 1st century AD
Kerch, Glinishche
- - Wool - Tabby 16/18 Stripes; over 15 fragments; c. 1st century AD
- - Wool - Tapestry 14/35 Stripes with geometric and floral motifs; two fragments
- - Wool - Repp 15/32 Green and brown stripes; napped surface; single fragment; c. 1stcentury AD
- - Wool - Tapestry 11/19 Black and yellow ornament on brown background; fourfragments; c. 1st century AD
- - Wool - Weft-facedtabby 12/20 Brown colour; three fragments; c. 1st century AD
- - Wool - Weft-facedtabby 15/24 Brown colour; single fragment; c. 1st century AD
- - Wool - Weft-facedtabby 15/24 Brown colour; five fragments; c. 1st century AD
- - Wool - Weft-facedtabby 10/15
Light stripes on brown background; two fragments; c. 1st
century AD
- - Wool - Weft-facedtabby 16/23
Dark brown stripes on light background; seven fragments; c. 1st
century AD
Kerch - - Wool -
Weft-faced
tabby 20/28
Green stripes, yellow flower motifs, on red background; four
fragments; c. 1st century AD
- - Wool - Tabby 22/20 Repp border; single fragment; c. 1st century AD
Kerch,
Abdul-Kerim
- - Wool - Tabby 18/20 Light brown colour; single fragment; c. 1st-2nd century AD
- - Wool - Tabby 40/42 Dark brown colour; single fragment; c. 1st-2nd century AD
- - Wool - Tabby ? Reddish-brown colour; two fragments; c. 1st-2nd century AD
- - Wool - Weft-facedtabby 16/20 Light warp and weft; single fragment; c. 1st-2nd century AD
Kerch, Kurgen
no.3 - - Silk -
Compound
tabby/twill 30/36
Gold/yellow colour; diamond pattern; three fragments; c.
1st-2nd century AD
Kerch, Glinishche - - Wool - Weft-facedtabby 20/26 Brown colour; two fragments; c. 1st-2nd century AD
Kerch, Mitridat no.
2 - - Wool - ? ? Greenish-brown colour; single fragment; c. 1st century AD
Kerch, Glinishche - - Wool - Tabby 16/18 Stripes; over 15 fragments; c. 1st century AD

Appendix 5: South West sites in this study
Cornwall - Bodrifty
Type Enclosed settlement Date of occupation 400 - 50 BC
Grid Reference SW 44436 35570 Date of excavation 1950-1954
Features Stone enclosure wall; 9
roundhouse structures.
Associated activities Domestic occupation;
spinning; food processing;
animal husbandry.
Notes This site began as an open settlement in the middle of the Penwith peninsula and
possibly dates back to the Bronze Age; 5 structures are larger houses and 4 are
smaller. The smaller roundhouses lack hearths, postholes, or gullies and their
entrances are narrow and face southwest. These structures appear late in the
settlement phase, suggesting a change in regional climate (Dudley 1956). The
earlier, larger structures have wider entrances facing southeast and thicker walls.
Finds include querns, locally made pottery, whetstones, and worked stone; there is
little to no evidence of long-distance trade or metal-working on site.
Publications Dudley, D, 1956, An Excavation at Bodrifty, Mulfra, near Penzance.
Archaeological Journal 113: 1-32
Carn Euny
Type Open settlement Date of occupation 400 BC - AD 300
Grid Reference SW 40250 28839 Date of excavation 1962-1974
Features Roundhouses; fogou; later
courtyard houses.
Associated activities Domestic occupation;
spinning; food
preparation; weapon
sharpening; animal
husbandry.
Notes The site is located on the Penwith peninsula near the hillfort of Carn Bran; it
consists of Iron Age roundhouses, a fogou, and later courtyard houses.
Excavations between 1964 and 1972 covered the fogou and several houses on the
site, focusing on investigating and restoring the fogou and circular underground
chamber. These works also revealed 9 hut foundations. Finds are scarce due to the
acidic soil conditions and are largely limited to locally made pottery and
quernstones. This makes it difficult to judge any trading links this settlement may
have had.
Publications Christie, P.M. 1965. Carn Euny Excavations: Interim report of 1964 season.
Cornish Archaeology 4: 24-29
Ibid. 1966. Carn Euny: A brief note on the 1965 excavations. Cornish
Archaeology 5: 17-20
Ibid. 1968. Carn Euny: Fourth Interim Report on the 1967 Season. Cornish
Archaeology 7: 33-37
Ibid. 1976. Carn Euny – Interim Report on the Final Seasons 1970 and 1972.
Cornish Archaeology 15: 68-72
The Rumps
Type Promontory Fort Date of occupation 200 BC - AD 100
Grid Reference SW 93411 81092 Date of excavation 1965-1967
Features 3 ramparts with ditches
across the narrowest point
of the promenantory;
gatehouse remains;
postholes; 6 hut platforms.
Associated activities Food preparation; spinning
and weaving; pottery
making; use of metal but
no metalworking.
Notes Located on a promenatory along Cornwall’s north coast with 3 lines of defensive
ditch and banks separating it from the mainland. Hut platforms cluster near the
entrance, suggesting that the rear of the castle was used for arable or pastoral
purposes. Surrounding environment used to supply shellfish. Spinning and
(possibly) weaving well represented in excavation finds; sheep are the most
represented animal in the record, suggesting a textile economy of sorts. Pottery
mimics forms and decorations found in Brittany, although no other material culture
from that region has been discovered, implying close trading connections with the
continent but not a migration to Cornwall.
Publications Brooks, R.T. 1974. The Excavations of The Rumps Promontory Fort, St Minver,
Cornwall. Cornish Archaeology 13: 5-50
Trevelgue Head
Type Promontory Fort Date of occupation 500 - 50 BC
Grid Reference SW 82572 63040 Date of excavation 1939
Features Multiple ditch and bank
defences; two Bronze Age
barrows; hut foundations;
post holes; pits; furnaces;
ovens; hearths; middens.
Associated activities Copper and iron smelting
and working; spinning and
weaving.
Notes Located on a spur of land, this site is defended by several lines of ditch-and-bank
earthwork defences. On the innermost enclosure, roundhouse hut platforms were
discovered, most dating to the 2nd century BC. One roundhouse measured 14.5 m
in diameter and is the largest found in Cornwall; it is believed to have been a
meeting place rather than a dwelling. Significant metalworking features and debris
were discovered in one area of the site, indicating iron smelting was a major craft
activity on this site with some copper smelting performed as well. Pottery found
on site indicates a preference for local ceramic production and trade, suggesting
only metal was traded elsewhere.
Publications Nowakowski, J.A. et al., 2011. Trevelgue Head, Cornwall : the importance of CK
Croft Andrew's 1939 excavations for prehistoric and Roman Cornwall. Truro:
Cornwall Council
Trevisker
Type Round Date of occupation 200 BC - AD 100
Grid Reference SW 88800 68700 Date of excavation 1955-56
Features Previous Bronze Age huts
and activity; 3 Iron Age
huts; pits; ditches; ovens;
gullies.
Associated activities Possible iron smelting;
spinning; food processing
and cooking.
Notes Located on gently sloping ground between two river systems along the north
Cornish coast. Smaller, inner enclosure built during the Iron Age, inside a larger
Bronze Age enclosure. Houses 1-3 dated to the Iron Age, although only House 2
produced a spindle whorl in its occupation layer. House 3 contained several ovens
and hearths, possibly demonstrating an area for cereal processing. Little evidence
of grain storage available and an iron smelting/working area is near the site,
although the exact location is unknown.
Publications ApSimon, A.M. and Greenfield, E. 1972. The Excavation of Bronze Age and Iron
Age Settlements at Trevisker, St. Eval, Cornwall. Proceedings of the Prehistoric
Society 38: 302-381
Devon
Blackbury Castle
Type Hillfort Date of occupation 300 BC - AD 100
Grid Reference SY 18700 92400 Date of excavation 1952-1954
Features Single ditch and bank
defences, complex
entrance, post holes,
gullies, roundhouse, oven.
Associated activities Possible iron smelting;
spinning; food processing
Notes Previously known as Blackbury Castle. Located on a ridge, with univaliate
defences but a developed entranceway (described as a’barbrican’) with gate-posts.
Limited excavation within the enclosure revealed the remains of one roundhouse,
along with an oven, a cooking pit, and several pits and gullies. Pottery makes up
most of the finds (due to the acidic soil); the earliest pottery is similar to the Iron
Age B ceramics of Dorsert, Sussex, Hampshire, and Wiltshire and appears to have
influenced the later pottery, known as Iron Age A in the report. Metalworking is
evidenced by several fragments of bronze, 1 of iron, and 4 whetstones. The only
other craft activity is demonstrated by 2 spindle whorls and 4 loom weight
fragments.
Publications Young, A., Richardson, K. 1954. Report on the excavations at Blackbury Castle.
Proceedings of the Devon Archaeological Exploration Society 5 (2-3): 43-67
Dainton
Type Unenclosed settlement Date of occupation 400 BC - 50 BC
Grid Reference SX 85900 66800 Date of excavation 1939; 1949
Features Ditch and banks, eroded
stone roundhouses,
postholes, possible field
system, gullies.
Associated activities Spinning, pottery
production, food
processing, possible iron
smelting.
Notes Settlement lies on Dainton Common and includes a possible field system, several
hut platforms, and a probable midden in the form of a concentric bank. This bank
contained spindle whorls, pottery, querns, ox and pig teeth, and charcoal
fragments. Crucible fragments discovered in 1 pit, but nowhere else on the site.
Publications Willis, L., Rogers, E.H. 1951. Dainton Earthworks. Proceedings of the Dorset
Archaeological Exploration Society 4(4): 79-101
Silvester, R. 1980. The prehistoric open settlement at Dainton, South Devon.
Proceedings of the Dorset Archaeological Exploration Society 38: 17-48
Kents Cavern
Type Cave Date of occupation 700 BC - AD 43
Grid Reference SX 93452 64171 Date of excavation 1856-1880
Features No discernable features. Associated activities Wool processing;
spinning; food
consumption; food
processing.
Notes Site of human activity lies in only the largest and most accessible parts of the cave
system and was defined by the ‘black mold’ deposit that covered this area. Quern
stones, cooking pottery, and small areas of burning (possible cooking fires) found
on the site, but no archaeological features that indicated long-term occupation.
Weaving combs and high numbers of spindle whorls but no loom weights suggests
the processing of fibres (possibly animal wool) for transportation and weaving
elsewhere.
Publications Pengelly, W. 1869. The Literature of Kent’s Cavern, Part II. Reports and
Transactions of the Devonshire Association 3: 191-482
Pengelly, W. 1878. The Literature of Kent’s Cavern, Part IV. Rep. Trans. Dev. Ass.
10: 141-181
Pengelly, W. 1884. The Literature of Kent’s Cavern, Part V. Rep. Trans. Dev. Ass.
14: 189-343
Silvester, R. 1986. The later prehistoric and Roman material from Kent’s cavern,
Torquay. Proceedings of the Devon Archaeological Society 44: 9-38
Bradley, R. 1978. The Prehistoric Settlement of Britain. London: Routledge
Milber Down
Type Multiple-enclosure site Date of occupation 100 BC - AD 100
Grid Reference SX 88400 69800 Date of excavation 1937-1938
Features Multiple ditch and bank
fortifications, gullies, corn
drier, one possible hut.
Associated activities Spinning, food processing,
local trade, possible iron
working.
Notes Site composed of multiple ditches and banks, although no evidence of structures
were discovered inside the inner bank during excavation. Several spindle whorls
found in gullies and ditches, suggesting they were no longer used at time of
deposition. Other finds included a corn drying oven, large amounts of charcoal,
iron slag, and an iron pommel; this suggests some form of iron working on this
site.
Publications Fox, A., Raleigh Radford, C. A., Rogers, E. H., Shorter, A. H., 1949 - 1950.
Report on the Excavations at Milber Down, 1937-8. Proceedings of the Devon
Archaeological Exploration Society 4: 27-65
Cottrill. 1937.  Interim Excavation Report. Proceedings of the Prehistoric Society
3: 448-9
Cottrill. 1938. Interim Excavation Report. Proceedings of the Prehistoric Society
4: 321
Mount Batten
Type Promontory fort Date of occupation 900 - 100 BC
Grid Reference SX 48679 53252 Date of excavation 1983-5
Features Post holes, pits, gullies. Associated activities Bronze working and
smelting, farming, food
processing, spinning.
Notes Located on a promontory in the Plymouth Sound, Mount Batten has been built on
repeatedly in modern times, effectively erasing any evidence of prehistoric
defensive earthworks. Study of the bronze tools and ornaments, as well as their
stratigraphical placement, suggests continuous occupation from the Late Bronze
Age into the Middle Iron Age. Personal ornaments were found in large numbers
here, along with bronze ingots and smelting debris. Given the location of the site
(close to the mines of Dartmoor and on the trade route from Cornwall to Brittany),
this suggests Mount Batten was an important manufacturing centre and trading
post for at least the metal trade during the Late Bronze and Early Iron Ages.
Pottery in phase 3 (Early to Mid Iron Age) is mostly local southwestern ware with
a few Cornish sherds.
Publications Cunliffe, B. 1988. Mount Batten Plymouth: a prehistoric and Roman port. Oxford
University Committee for Archaeology Monographs

Appendix 6: Central South sites in this study
Dorset - Allard’s Quarry
Type Open settlement Date of occupation 800 BC - AD 100
Grid Reference ST 79491 19881 Date of excavation 1944-1945
Features Pits, post holes, two
burials, roundhouse,
4-post structure
Associated activities Spinning, weaving; bone
manufacturing; food
processing; trading
Notes 3 site phases in Iron Age; 42 pits belonging to 1st phase, 26 to the 2nd, and 16 to
the 3rd. 3 huts in 1st phase, incomplete due to quarrying activity; 2 huts during
2nd phase; 1 possible hut in 3rd phase with 4-post structure (likely granary). No
imported pottery in any phases, although the presence of Kimmeridge shale
bracelets suggest coastal trading.
Publications Williams, A. 1950. Excavations at Allards Quarry, Marnhull, Dorset. Proceedings
of the Dorset Natural History and Archaeological Society 72: 20-75
Gussage All Saints
Type Banjo Date of occupation 300 BC - AD 100
Grid Reference ST 99895 10620 Date of excavation 1972
Features Ditch, pits, postholes,
roundhouses, and ovens.
Associated activities Pottery production, iron
forging, iron smelting in
the Roman phase,
spinning, weaving, food
production and
processing.
Notes Site is divided into clear activity zones, with storage pits clustered to the east of
the roundhouse and crafting debris to the south. Craft activity represented by
ceramic debris from pit 209 (dating to the second phase of settlement, 1st century
BC), bronze casting and iron forging debris (from both early settlement phases),
and a large number of spindle whorls (dated to phase 3 of the settlement) and
loom weights (an equal number found in each of the 3 phases). The agricultural
economy of this site is demonstrated by a variety of carbonised grains from
various rubbish deposits. The high number of spinning and weaving tools from
this site suggest that this site probably produced more than required locally;
Wainwright believes this was part of a wider economic trade (1979: 190-1). This
is supported by the evidence that this site traded in querns and pottery from the
larger Wessex trading area.
Publications Wainwright, G.C. 1979. Gussage All Saints: An Iron Age Settlement in Dorset.
Department of the Environment: Archaeological Reports No 10
Hengistbury Head
Type Promontory fort Date of occupation 700 BC - AD 50
Grid Reference SZ 17300 90700 Date of excavation 1970
Features Hearths, roundhouses,
double dykes defences,
pits, and gullies.
Associated activities Long-distance and local
trade; metal, glass, and
shale manufacturing;
spinning and weaving;
food processing
Notes This site has evidence of non-continuous occupation from the Paleolithic period
onwards. Pottery finds date the earliest Iron Age occupation to around 700-400
BC, the exact nature of this period’s settlement is unclear but probably a scattering
of individual farmsteads before the double dyke defences were made, turning the
site into a hill-/promontory- fort. Two bronze axe finds (one 300m out at sea,
possibly result of erosion or dropped from boats) hint at this site being possible
stop-off for significant long-distance trading routes. Few artifacts or materials
definitely assigned to the middle Iron Age period (400-100 BC), indicating no or
little occupation. In the late Iron Age phase (100-0 BC), the settlement becomes a
major international trading port. Local, regional, and European trade demonstrated
by pottery finds (from west England, north-west France, and locally);
metal-working/manufacturing (ingots of copper indicate copper importation,
silver refining demonstrated with silver-copper ore/ingots, and bronze and gold
working debris); Kimmeridge shale importation and manufacturing (waste
material and finished artefacts); and limited glass production (crud glass lumps
and fragments of finished glass products). Other finds include a large number of
coins (local and imported); and range of luxury items (jewellery, personal items,
etc.).
Publications Cunliffe, B. 1978. Hengistbury Head. London: Elek Books
Hod Hill
Type Hillfort Date of occupation 800 B.C. - 43 A.D.
Grid Reference ST 85500 10700 Date of excavation 1897; 1949; 1951-2
Features Ditches and banks;
roundhouses; pits; gullies;
postholes
Associated activities Spinning, weaving,
smelting, smithing,
quarrying
Notes A multivallate hillfort covering 22 ha., built in the early Iron Age, with a Roman
fort in the northwestern corner. Roundhouse footprints cover the site, often
overlapping, with associated pits and hearths. Several trackways are recorded
from the absence of roundhouse circle; these converge at the northeastern
entrance, indicating this was the main thoroughfare. Metalworking has a large
presence on site that appears to be centred on a possible ‘industrial’ compound in
the southeast corner with several quarry scars/pits. Magnotromic surveys indicate
these pits are filled with slag and other metalworking debris. Several Kimmeridge
shale spindle whorls from this site are repurposed lathe-turned cores that are left
over from producing shale bracelets. 52 spindle whorls are recorded in the
excavations report, but only 5 loom weights and 2 weaving combs indicate that
weaving was limited at this site. Large numbers of slingshots and spearheads
suggests the hillfort was abandoned after several battles or skirmishes.
Publications Richmond, I. 1968. Hod Hill, Vol 1-2. London: Trustees of the British Museum
Brailsford, J., & Richmond, I. 1952. British Museum Excavation at Hod Hill,
Dorset. The British Museum Quarterly 17 (3), 49-50
Maiden Castle
Type Hillfort Date of occupation 600 BC - AD 100
Grid Reference SY 66900 88400 Date of excavation 1934; 1937; 1985-6
Features Multiple ditch and bank
defences; roundhouses;
burials; gullies; roads;
pits; postholes; ovens
Associated activities Spinning and weaving;
metalworking; food
processing
Notes A large, multivaliate hillfort with extensive, complicated entrances. Initially a
small enclosure built in the 1st Millennium B.C. (phase 5), the site was extended
during the middle Iron Age to cover the whole of the hilltop. Occupation appears
densest during the middle Iron Age (phase 6), spanning across the site, but
becomes limited to the southeastern area during the late Iron Age (phase 7). A
thorough excavation revealed large quantities of artefacts, demonstrating little
metalworking during the early Iron Age. Animal remains indicate large numbers
of sheep, possibly to provide fibres for the textile production occurring at the site.
Pottery finds suggest there was on-site ceramic production during the early Iron
Age that was replaced in the middle Iron Age by importation of ceramics made in
Poole workshops. The site was largely abandoned during the late Iron Age with
the remaining occupation clustered around the eastern entrance, outside of which
was the remains of a smithy.
Publications Shaples, N.M. 1994. Maiden Castle, Dorset. In A.P. Fitzpatrick and E. Morris
(eds.) The Iron Age in Wessex: Recent Work. Wessex Archaeology, pp. 92-94
Sharples, N.M. 1991. Maiden Castle: Excavations and field survey 1985-6.
English Heritage: Archaeological Report no 19
Hampshire
Danebury
Type Hillfort Date of occupation 500 BC - AD 100
Grid Reference SU 32300 37600 Date of excavation
Features Ditch and bank
earthworks; hut platforms;
post holes; pits; hearths;
forges.
Associated activities Bronze and iron working;
spinning and weaving; salt
production; food
production and
processing.
Notes A large and well-excavated hillfort with evidence of continuous occupation from
the 6th to the 1st century BC. Artefacts such as iron ingots and briquetage provide
evidence of trade that Cunliffe suggests were traded locally for raw materials like
wool or cereals. While no evidence of smelting was found on site, iron and bronze
working debris, materials, and tools show on-site production of artefacts.
Evidence of craft activities with shale, bone, basketry, and so on are rare, but
textile production was a significant activity throughout the life of the hillfort. All
this evidence supports the interpretation of Danebury as a centre of production for
certain crafts and of trading, both locally and long-distance.
Publications Cunliffe, B. 1994. Danebury, Hampshire. In A.P. Fitzpatrick and E. Morris (eds.)
The Iron Age in Wessex: Recent Work. Wessex Archaeology, pp. 94-97
Easton Lane
Type Enclosed settlement Date of occupation 800 - 600 BC
Grid Reference SU 49500 31300 Date of excavation 1971
Features Linear depression;
irregular pits and scoops;
Associated activities Agricultural; waste
disposal; spinning and
weaving.
Notes Site was excavated as part of the M3 extension; 3 phases of activity identified
(Phase 1: pre-ring ditch; Phase 2: Bronze Age ring ditch; Phase 3: Iron Age
post-ring ditch). The ring ditch contained several associated cremations and
inhumations but was no longer used in the Iron Age. Exact use in this period is
unclear (Fasham 1982: 50).
Publications Fasham, P.J. 1982. The Excavation of Four Ring-Ditches in Central Hampshire.
Proceedings of the Hampshire Field Club and Archaeological Society 38: 19-56
Kennel Farm
Type Enclosed settlement Date of occupation 800 - 300 BC
Grid Reference SU 59555 48561 Date of excavation 1966
Features Oval enclosure ditch;
4-post structure; pits;
fence-line.
Associated activities Domestic occupation;
spinning and weaving;
storage; animal
husbandry; agriculture.
Notes Close to another settlement of Mid Iron Age - Roman date. Half the site was
excavated; no house structures found in excavated area. Storage pits were
clustered in at least 4 distinct groups, suggesting chronological sequence of
storage pits being used then closed (Chapman 2006: 20). Postholes rarely formed
any structural patterns, suggesting they were used alone or in pairs for external
activities (i.e. looms, haystack posts, etc.) (Chapman 2006: 21)..
Publications Chapman, A. 2006. An Iron Age Enclosure at Site A, Kennel Farm, Basingstoke,
Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society
61: 16-62
Winnall Down
Type Enclosed settlement Date of occupation 600 - 50 BC
Grid Reference SU 50600 30000 Date of excavation 1976-7
Features Single enclosure ditch;
roundhouses; 4-post
structures; hearths; pits;
post holes.
Associated activities Food production and
processing; food storage;
weaving.
Notes Site is comprised of a single enclosure ditch containing 8 roundhouses and 10
4-post structures, although not all dwellings or storage buildings would have been
built or used at the same time. Use of the enclosure appears very structured, with
certain features and activities restricted to areas of the site. The northwest corner
is characterised by 4-post structures and high levels of loom weight deposition,
while the southeast corner was used for crop processing and storage. Pits 2630
and 2676, close to 4-post structures G and T, contained about 11 and 14 loom
weights respectively (all fragmented to some degree). Spatial correlation suggests
structures G and T were weaving workshops. A few pieces of briquetage and
locally sourced querns were found, although no evidence of metalworking were
found.
Publications Fasham, P.J. 1985. The Prehistoric Settlement at Winnall Down, Winchester:
excavations of MARC3 site R17 in 1976 and 1977. Monograph 2
Zions Hill Farm
Type Farmstead? Date of occupation 100 - 0 BC
Grid Reference SU 41500 20200 Date of excavation 1995 - 1996
Features Pits, gullies Associated activities Weaving, farming
Notes An unusual site with no structural evidence, only pits and gullies. 5 loom weights
excavated from late Iron Age/early Roman pit 5004. Pottery includes late Iron
Age coarseware made from local materials, similar to finds from other Iron Age
sites in the area. Quernstones indicate farming and seed processing occurring on
site, although no metalworking tools or debris were recorded. Appears to be the
working area of a small agricultural settlement/single farmstead nearby.
Publications Rawlins, M., Chadwick, A. M., Reynier, M. J. and Seager Smith, R. H. 2003.
Prehistoric and Roman activity at Zionhill Farm, Chandlers Ford. Proceedings of
the Hampshire Field Club & Archaeological Society 58: 1 - 23

Appendix 7: South East sites in this study
Kent - East Wear Bay, Folkestone
Type Possible oppida Date of occupation 500 B.C.-A.D. 43
Grid Reference TR 24000 36600 Date of excavation 1924; 1989; 2010-2011
Features Roundhouses, gullies,
4-post structures,
quernstone workshops,
pits, postholes
Associated activities Quern manufacture;
spinning; weaving;
continental and British
trade.
Notes Unfortunately, much of this site has been lost to the sea via cliff erosion, so the
remains must be taken as only part of the whole site. Full size of the Iron Age
occupation is unknown but its high status and large-scale craft production have
allowed it to be labelled an oppida. The site was largest around 150 B.C.-A.D.
43. Quernstone manufacture possibly performed as a production line, as several
querns in various stages of manufacture were abandoned by workers. A Roman
villa complex was built over the remains of the oppida in the late 2nd century
A.D. Finds include large quantities of British and continental pottery, including
Italian amphorae; 13 Iron Age coins; large amounts of quern debris and flint
debitage; and a significant number of spindle whorls. Only a few loom weights
found.
Publications Colsen, I. 2013. A Town Unearthed: Folkestone Before 1500. Canterbury
Archaeological Trust
Parfitt, K . 2012. The Rocky Road to Folkestone: Excavations at Folkestone
Roman Villa 2011. Kent Archaeological Society Newsletter 92: 2-4
2010. Folkestone, East Wear Bay Roman villa: Interim report on
excavations 2010. Kent Archaeological Society Newsletter 88: 14-15
Keller, T. 1988. The Evidence for Ancient Quern Production at Folkestone. Kent
Archeological Review 93: 59-68
Highstead
Type Enclosed settlement Date of occupation 800 BC-AD 43
Grid Reference TR 21300 66100 Date of excavation 1975-1977
Features Ditches, post holes, pits,
hearths, ovens,
roundhouses, palisade
post holes, gatehouse post
holes, 4-post structures,
gullies.
Associated activities Copper and iron
metalworking, pottery
production, agriculture,
food processing, food
storage, salt production.
Notes In the Late Bronze Age/Early Iron Age, the site had 3 enclosures. Enclosure B70
in the north of the site was rectangular with defensive palisade post holes and a
gatehouse within its ditch. Hearth and clay-lined pits (storage areas) excavated
within B70 indicate occupation; pottery, copper dross, and vitrified clay finds
from the enclosure ditch point towards crafting use rather than domestic.
Enclosure A24 (to the southwest) was oval in shape and domestic, with 2
roundhouses (built sequentially) in the centre. The Middle Iron Age (600-400 BC)
is characterised by an open settlement; occupation remains constant, with copper,
pottery, and subsistence activities continuing. Debris and remains of Ironworking
and salt production are found during this period, along with 4-post structures. This
is followed by a period of abandonment until 100 BC, when a series of rectangular
enclosures are dug, possibly for livestock management. No evidence of further
occupation is recorded, although pottery evidence suggests occupation in the
surrounding area. Site suffered some ploughing damage prior to excavation.
Publications Bennett, P., Couldrey, P., and Macpherson-Grant, N. 2007. Highstead, near
Chislet, Kent. Excavations 1975–1977. Canterbury Archaeological Trust
Monographs: New Series Vol IV
Sunset Caravan Park
Type Unclear Date of occupation 700 BC - AD 100
Grid Reference TR 09296 64478 Date of excavation
Features Pottery kilns, quarry pits,
roundhouses, pits,
postholes, gullies, ditches
Associated activities Clay and flint quarrying;
metal working; pottery
production; spinning; salt
production; long-distance
trade
Notes Given the confines of this excavation (being a rescue excavation prior to building
work), the nature of this site and whether it was enclosed or defended in any way
is unclear. Roundhouse features have been found throughout the excavation,
demonstrating occupation at this site. The predominant activity on this site
appears to be clay and flint quarrying, which increased during the Late Iron Age.
Clay ovens and debris also indicate high levels of pottery production; there is also
evidence for salt production. Early Iron Age Italy pottery finds include italian and
other continental wares, which indicate long-distance trade at this site.
Publications Allen, T. 1997. The Whitstable Iron Age Settlement. The Archaeological
Evaluation and Results of Preliminary Excavation at Sunset Caravan Park,
Thanet Way, Whitstable, Kent, 1997/61. Canterbury Archaeological Trust
Allen, T. 1999. The Whitstable Iron Age Settlement: Parts 1 and 2. Unpublished
Canterbury Archaeological Trust
Allen, T., and Wilson, J. 2001. Sunset Caravan Park and Church Lane East,
Whitstable. Canterbury’s Archaeology 1998-1999: 10-11
Thanet Earth
Type Enclosed settlement Date of occupation 500 BC - AD 43
Grid Reference TR 28900 66700 Date of excavation 2007-8
Features Ditches, pits, postholes,
ring gullies, burials, field
systems
Associated activities Spinning and weaving;
food processing; pottery
production; possible antler
manufacturing
Notes Enclosed settlement with domestic features within and beyond enclosure ditch,
suggesting expansion of occupation during Middle Iron Age. Most features from
this date were clustered on plateau 8 and including ditches, pits, postholes, and
ring gullies. Loom weights recovered from rubbish deposits in pits in this area; the
plateau produced an unusually high number of Iron Age spinning and weaving
tools, suggesting textile production as a major craft activity at the site. The
occupation decreases in the Late Iron Age before finishing completely. There was
little to no evidence of other craft activities.
Publications Rady, J. 2009. Thanet Earth, Monkton. Canterbury’s Archaeology 2007-2008:
11-27
2010a. Thanet Earth, Monkton. Canterbury's Archaeology 2008-09: 1-16
2010b. Excavations at Thanet Earth 2007-2008 Assessment Report Volume
1. Canterbury Archaeological Trust

Appendix 8: Textile tool data form
Museum
No.
Tool Site Context Material Shape Height Width/
diameter
Depth Weight Decor Preser-
vation
Appendix 9: Cornwall artifact data
Bodrifty
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1 Discoid Reused sherd [1] 58 16 26.6
4 Discoid Reused sherd [6] 33 26 25.4
H1/9 BY154 Discoid Reused sherd (148)[43] 17 15 4.8
9 BY54 Discoid Reused sherd (148)[43] 32 12 23.15
G663 Discoid Reused sherd [7] 41 26 42.4
2 G664 Discoid Reused sherd [5] 34 23 22.15
Carn Euny
Table 2: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
52 Discoid Stone Cut 2 CN 33.5 14 26.45
88 Discoid Stone Cut 1 (ss) Peg 1F 43 24.4 61.15
114 Discoid Slate Hut A sq 6 N side 35.5 7.5 16.05
121 Discoid Baked clay Hut A sq 9 Peg E 30 13 10.35
177 Discoid Stone Peg 4F-4H 32 9 13.45
198 Discoid Baked clay Unrecorded 16.4 7 13.8
301 Discoid Reused sherd TA 1 Wall (1) 40 6.5 16.7
386 Discoid Stone Cut 4 Peg J-3M 28 9 12.25
388 Discoid Reused sherd Cut 5 Peg 4d-4c 46 7 15.8
452 Discoid Reused sherd TA 3 Peg 3TF/O 37.4 7.5 13.75
501 Discoid Reused sherd
TOS (N) E. Hut H
behind wall (4) 37 9.3 12.85
680 Discoid Stone CHI (s) sq 4 - wall on E 34 16.2 27.75
699 Discoid Reused sherd Unrecorded 34.6 5.4 8.8
784 Globular Stone
CH2 Room 2 Wall face
(1) 25.6 20 19.05
The Rumps
Table 3: Spindle whorl data
Museum No. Form Material Location
Diameter
(mm) Depth (mm) Weight (g)
1 Discoid Slate R67 R26 A(4) [230] 29 5.5 6
2 Cylindrical Baked clay R2K A(4) [218] 35.5 17 27
3 Cylindrical Stone R65 BRI ditch A 39 18.6 41.5
3a Discoid Stone RI (39) <18> ditch C 29.6 6.3 7.6
5 Domed Bone BR1 (9-10) <24> ditch A 37.6 19.7 9.3
6 Cylindrical Baked clay A RI-M (1) [199] 32.4 15 14.7
7 Discoid Stone A RIN (4) [243] 38 15.3 33.6
9 Discoid Baked clay RIP (4) [280] R67 34 17.4 26.75
11 Discoid Slate A R2G (7) [212] 41 5 15
14 Discoid Slate BRI (9-10) ditch A <23> 63.6 6.7 44.4
15 Discoid Slate RIM/RIN (4) [271] 49 8 26.8
16 Discoid Slate N side of BR2B [42] 39 3 7.6
17 Discoid Baked clay A RIJ (2) <261> 33.4 16 20.6
? Discoid Slate RIM (1) [201] 30 6 17
125 Discoid Slate A R2 (4) inner ditch [91] 34 7 14.7
Table 4: Loom weight
Museum
No. Form Material Location Width (mm) Depth (mm) Weight (g)
10 Discoid Slate AR2B (5) [186] 86 16.7 156
18 Discoid Slate B R2 A bottom of (7) 181 10.4 543.6
?A Discoid Slate R2 inner ditch extension (4) [15] 97 14 108
?B Discoid Slate R2 inner ditch extension (4) [15] 85 14 384.6
45 Discoid Slate R2 inner ditch extension (4) [145] 138 7 236
46A Discoid Slate R2 inner ditch extension (4) [37] 75 6 46.8
46B Discoid Slate R2 inner ditch extension (4) [37] 33 4.3 11
46C Discoid Slate R2 inner ditch extension (4) [37] 34.3 5.3 59.4
44/47A Discoid Slate R2 inner ditch (4) [65] 55.5 3 13.95
44/47B Discoid Slate R2 inner ditch (4) [65] 76.4 6.4 146.4
44/47C Discoid Slate R2 inner ditch (4) [65] 50 3.5 53.4
44/47D Discoid Slate R2 inner ditch (4) [65] 50 4.3 54.8
Trevelgue Head
Table 5: Spindle whorl data
Museum
No. Form Material Location
Diameter
(mm)
Depth
(mm) Weight (g)
113 Discoid Stone Tr 63 [452] 33 17.3 22.7
155a Discoid Stone House [420] 34 11.2 18.95
155b Discoid Stone House [420] 31 9.4 19
181 Discoid Stone House [421] 30.6 12.4 12
216b Discoid Slate House [136]/[137] 28.5 11.6 17.65
216c Discoid Stone House [316]/[317] 45.7 7.2 20
216d Discoid Stone House [136]/[137] 30.6 13 17.8
228a Discoid Stone House [159] 39.4 6.5 12.95
228b Discoid Stone House [422] 41 8 17.3
243 Discoid Slate House [421] 38.8 5 9
261 Globular Stone House [422] 14 15.7 31.4
279a Globular Baked clay House [420] 15 21.8 46.8
419 Discoid Stone House [139] 38 15.3 31.45
440 Discoid Slate House [440] 41 4.2 9.2
502 Discoid Stone House [142] 31 17.7 32.25
503 Discoid Stone House [142] 29.2 9 10.75
504 Discoid Slate House [420] 39 5.4 14
505 Discoid Stone Tr 65 31.8 17.7 18.45
90170A Discoid Baked clay Tr 6.11 [420] 14 8 16.6
90175A Discoid Baked clay Tr 612 [422] 33.6 8.2 10
90376A Discoid Stone House 1 100/34X Y49 p/h 31 6.5 15.7
90380 Discoid Stone Tr 6 [420] 34.7 14 18
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
133 Discoid Stone House [420] 88.3 36.2 654
277 Discoid Slate House [420] 78 7 56.85
Trevisker
Table 7: Spindle whorl data
Museum No. Form Material Location
Diameter
(mm) Depth (mm) Weight (g)
SW 1 Discoid Stone T/116 42 19 46.25
SW 2 Discoid Baked clay T/49 34.4 12.3 16
SW 4 Discoid Stone T/6 33.5 8.5 15.75
SW 5 Discoid Stone T/170 30.5 11.3 12.6
SW 7 Discoid Stone T/88 30.4 13.4 20.6
SW 8 Discoid Reused sherd T/150 38.4 7.3 13.4
SW 9 Discoid Stone T/115 30.7 8.7 14.45
SW 10 Globular Stone T/8 30.5 8.3 19.4
SW 11 Discoid Stone T/115 22.7 7 6
SW 12 Globular Stone T/244 35.5 9.2 32
Table 8: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
<40> Discoid Baked clay [75] 122 92.1 1372.8
T/A/596 Discoid Baked clay [76] 62.4 61.4 691.7
T/29 Discoid Slate Unrec. 154 14.4 582.1
T/32 Discoid Slate Unrec. 99 9.3 162
T/38 Discoid Slate Unrec. 94.3 17.5 258
T/45 Discoid Slate Unrec. 53.6 6 23.8
T/52(?) Discoid Slate Unrec. 114 20 398
T/65a Discoid Slate Unrec. 79.7 9.3 115.5
T/65b Discoid Slate Unrec. 95 9.4 189.6
T/95 Discoid Slate Unrec. 51.7 9.8 36
T/227 Discoid Slate Unrec. 41.4 1.8 6.1
T/230 Discoid Slate Unrec. 49 13.3 1092
T/237 Discoid Slate Unrec. 48 5.8 17
Appendix 10: Devon artifact data
Blackbury Camp
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
12/1956.96 Discoid Sandstone BC, B, trench 2 47 19 42
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
12/1956.107 Triangular Fired clay Site A, rampart 5 183 162 1400
12/1956.108 Triangular Fired clay Site A, rampart 5 129 75 1180
12/1956.109 Triangular Fired clay Site A, rampart 5 171 120 2040
12/1956.110 Triangular Fired clay Site A, rampart 5 48 30 78
Dainton
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A4431 Discoid Sandstone Unrecorded 93.6 44 166
A5977 Discoid Sandstone Unrec. 42 22 39
A5978 Discoid Stone Unrec. 44 18 24
A4480 Discoid Baked Clay Unrec. 35.7 50 18
A4481 Discoid Baked Clay Unrec. 34.7 50 22
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
A5976 Rounded Slate/Shale Unrecorded 130 20 272
Kents Cavern
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A3281 Discoid Shale Unrecorded 36 8 10
A5958 Discoid Stone Unrecorded 37 28 48.4
A5950 Discoid Shale 30 25 13 14.7
A5951 Discoid Shale 42 27 14 14
A5952 Discoid Shale 48.4 36 12 16
A5953 Discoid Stone Unrecorded 39 12.5 16.25
A5954 Discoid Shale Unrecorded 37 15 25
A5956 Discoid Shale Unrecorded 33 15 20
A5957 Discoid Stone Unrecorded 42 19 41
A5959 Discoid Sandstone Unrecorded 36 16 30.5
A5960 Discoid Sandstone Unrecorded 35 13 19
A5961 Discoid Sandstone Unrecorded 40 13 26.6
1886.12-29.26 Globular Baked Clay Unrecorded 28 16 38.75
1886.12-29.25 Discoid Stone Unrecorded 34 10 21
1886.12-29.27 Discoid Sandstone Unrecorded 34 21 27
1886.12-29.24 Discoid Stone Unrecorded 28 30 33
1886.12-29.34 Discoid Stone Unrecorded 20 8 4.2
1886.12-29.35 Discoid Stone Unrecorded 28 13 19
1886.12-29.36 Discoid Stone Unrecorded 32 15 27.3
1886.12-29.37 Discoid Stone Unrecorded 25 20 22
Milber Down
Table 6: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
A4178 Discoid Sandstone Unrecorded 31 12 14.7
A5935 Discoid Shale Unrec. 30 12 8
A5936 Discoid Shale Unrec. 36 14 28
A5937 Discoid Shale Unrec. 28 14 21
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
A4342 Discoid Stone Main camp 86 15 181.25
A5939 Discoid Sandstone Unrecorded 65 13 64.4
Mount Batten
Table 8: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1958.11.86.1X Discoid Sandstone Unrecorded 35.5 17 35.5
1958.11.86.2X Discoid Sandstone Unrec. 45 17.5 68
1958.11.86.3X Discoid Stone Unrec. 33 16 26
258.18 Discoid Sandstone Unrec. 36 13 26

Appendix 11: Dorset artifact data
Allard’s Quarry
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
3 Discoid Limestone Pit 55 27.8 24.2 71
4 Globular Human bone Pit 17 40 26.4 20
13 Discoid Fired clay Pit 17 82.3 11.2 57
16 Globular Limestone Pit 8a 26 20.5 15
17 Discoid Limestone Pit 12a 38.8 27.5 49
19 Discoid Limestone Pit 22 38 13 34
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
?b Fragmentary Stone Pit 31 186.7 31 303.3
13 Triangular Fired clay Pit 112 213.3 60 1313.3
Gussage All Saints
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
?c Discoid Chalk 212 (3) 42.5 15 30
2010 Discoid Chalk 212 (3) Phase J 42 15 30
4001 Discoid Fired clay 358 (3) 30.5 26.5 27
4004 Discoid Fired clay U/L (2) 30 8 9.4
4008 Globular Fired clay 446 (6) 33.5 25 31
4030 Discoid Fired clay 396 (4) 31 20.9 24
4031 Discoid Reused sherd 111 (4) 43.9 27.4 18
4033 Discoid Fired clay 528 (5) 37.8 22.4 36
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
?a Triangular Chalk 2 (5a) 1201 83 2038
?b Triangular Fired clay 539 (6) 205.7 110 3710
4005 Cylinder Fired clay IQ (3) 59 59 820.8
4009 Triangular Fired clay 415 (8) 128.75 68 520
4012 Cylinder Fired clay 620 (6) 68.5 59 871.5
4014 Triangular Fired clay 377 (5) 233.6 92.8 1673.3
4020 Triangular Fired clay 379 (8) 1807 58 2082.6
4020a Triangular Fired clay 379 (8) 166.8 67.2 1946.25
4021 Triangular Fired clay 379 (8) 194.28 61.8 1774.2
4022 Triangular Fired clay 639 [310J] (5) 401.6 71.5 2516.6
4023 Triangular Fired clay 639 [310J] (6) 254.4 61.1 1842.2
4039 Triangular Fired clay 52 (7) 258 52.8 1250
Hengistbury Head
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
69 Discoid Reused sherd Ph 243/1 44 7 18
122 Globular Fired clay O/179 30 19 14
125 Discoid Reused sherd O/180 49 7 19.5
161 Discoid Reused sherd F42/272 59 7 20.4
183 Discoid Reused sherd O/295 46 9 26.25
211 Discoid Reused sherd O/191 44 9 27.5
223 Discoid Reused sherd Ph 839/1 34 8 10
225 Discoid Reused sherd F270/504 28 7 7.2
235 Discoid Reused sherd F273/514 42 28 60.8
242 Discoid Reused sherd F260/452 31 6 9.2
265 Discoid Reused sherd F399/601 56 14 37
295 Discoid Reused sherd O/689 46 11 30.6
300 Discoid Reused sherd O/696 35 5 8.7
305 Discoid Reused sherd O/696 43 10 23
338 Discoid Reused sherd O/837 44 7 17
W3 Discoid Reused sherd X229 36 8 10
W5 Discoid Reused sherd X227 51 14 38
x60 Discoid Reused sherd B-F 40 7.5 15.9
x61 Discoid Reused sherd B-F 40 8.5 12
x62 Discoid Reused sherd B-F 38 9.5 11.4
x66 Discoid Reused sherd B-F 39 7 9
x67 Discoid Reused sherd B-F 48 13.75 30.8
Museum No. Form Material Material Diameter (mm) Depth (mm) Weight (g)
x68 Discoid Reused sherd B-F 45 13.5 28
x69 Discoid Reused sherd B-F 40 15 21.5
x70 Discoid Reused sherd B-F 45.5 9.5 22.2
x71 Discoid Reused sherd B-F 44 9 16
x72 Discoid Reused sherd B-F 55 7.5 18.3
x73 Discoid Reused sherd B-F 43.5 7 16.3
x74 Discoid Reused sherd B-F 43.5 10.5 16
x75 Discoid Reused sherd B-F 31 6.5 5.8
x76 Discoid Reused sherd B-F 41.5 10 17.4
x77 Discoid Reused sherd B-F 42 7 14.2
x78 Discoid Reused sherd B-F 37 8 11.3
x79 Discoid Reused sherd B-F 40 8 8
x80 Discoid Reused sherd B-F 32 10 10.8
x81 Discoid Reused sherd B-F 55 23 84.2
x82 Discoid Reused sherd B-F 49 10.5 26.8
x83 Discoid Reused sherd B-F 48.5 9.5 21.2
x86 Discoid Reused sherd B-F 52 8 24.7
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
x92 Elongated Fired clay B-F 80 85 1102.2
x93 Elongated Fired clay B-F 80 80 1292.2
x94 Triangular Fired clay B-F 60 40 1090.4
x95 Triangular Fired clay B-F 62 40 1360
x96 Elongated Fired clay B-F 92 96 1394.25
x97 Triangular Fired clay B-F 60 40 278.75
x98 Elongated Fired clay B-F 99 90 1725
x263 Triangular Fired clay sf 5 75 55 2574.6
x265 Triangular Fired clay Unknown 60 40 2138.3
x266 Triangular Fired clay sf 4 85 65 2346.6
Hod Hill
Table 7: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
92.9-1.1386 Discoid Stone Unrecorded 32 13 7.7
92.9-1.1387 Discoid Baked Clay Unrec. 39 14 31.4
92.9-1.1389 Discoid Stone Unrec. 33 9 8.5
92.9-1.1396 Conical Antler Unrec. 36 24 21
92.9-1.1393 Globular Stone Unrec. 44 32 94.4
92.9-1.1394 Globular Stone Unrec. 43 30 62.1
92.9-1.1395 Globular Stone Unrec. 32 27 34.4
92.9-1.1397 Globular Stone Unrec. 37 22 33.3
92.9-1.1398 Globular Stone Unrec. 26 23 50
92.9-1.1399 Globular Sandstone Unrec. 30 20 23.5
92.9-1.1400 Discoid Sandstone Unrec. 40 19 26.6
92.9-1.1401 Globular Sandstone Unrec. 38 22.3 29.4
92.9-1.1402 Globular Baked Clay Unrec. 37 19 26
92.9-1.1403 Discoid Stone Unrec. 39 15.5 26.6
92.9-1.1404 Discoid Stone Unrec. 44 19 35
92.9-1.1405 Discoid Stone Unrec. 37 14.4 26.6
92.9-1.1406 Discoid Stone Unrec. 44 15.5 33.3
92.9-1.1407 Discoid Stone Unrec. 60 15.5 75.5
92.9-1.1408 Discoid Sandstone Unrec. 45 23.3 46.6
92.9-1.1409 Discoid Sandstone Unrec. 47 18 32.6
92.9-1.1410 Discoid Sandstone Unrec. 35 20 20
92.9-1.1411 Globular Stone Unrec. 32 32 34.7
92.9-1.1414 Discoid Sandstone Unrec. 26 13.75 10
92.9-1.1415 Discoid Sandstone Unrec. 28 13.75 7.5
92.9-1.1416 Discoid Baked Clay Unrec. 22 12.5 6.25
92.9-1.1417 Discoid Baked Clay Unrec. 29 13 9.4
92.9-1.1419 Discoid Stone Unrec. 20 15.7 6.3
92.9-1.1420 Discoid Stone Unrec. 30 17 19
92.9-1.1421 Discoid Baked Clay Unrec. 30 14.7 14.7
92.9-1.1423 Discoid Reused sherd Unrec. 30 8.2 9.4
92.9-1.1424 Discoid Reused sherd Unrec. 35 10 15.5
1960.4-5.120 Globular Baked Clay Hut 56 annexe 46 25 47.3
1960.4-5.122 Discoid Chalk Palisade cut 89 24 125
1960.4-5.1004 Discoid Reused Sherd Unrec. 32 7 6
1960.4-5.1006 Discoid Chalk Unrec. 74 40 134
Table 8: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1960.4-5.124 Elongated Chalk Pit 15b 160 76.6 3091
1960.4-5.125 Elongated Chalk Unrecorded 171.6 116.6 2466.6
1960.4-5.126 Elongated Chalk Unrec. 148.75 138.75 2996.25
1960.4-5.127 Elongated Chalk Unrec. 112.6 79 1807
1960.4-5.128 Elongated Chalk Pit 15b 108 164 2732
1960.4-5.1845-53 Elongated Chalk Pit 15a 183.5 88 2906
1960.4-5.2685 Elongated Chalk Unrec. 150.6 149.3 3456
1960.4-5.3353 Triangular Baked clay Unrec. 106 66 650
1960.4-5.3428 Elongated Chalk Hut 36a posthole 116 52.6 1526.3
1960.4-5.3429 Elongated Chalk Hut 36a posthole 105 70.5 1708.4
1960.4-5.3430a Elongated Chalk Unrec. 1086.3 80 2491.4
Maiden Castle
Table 9: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
113 Globular Bone R Pit R 34 (1) 39 36 23
.55.4 14 34/204 Discoid Stone Area B misc 38 12 17
.55.4 15 34/206 Discoid Chalk B ditch? (35) 57 21 82
.55.4 16 34/207 Discoid Chalk Area B 1 (3) 71 26 106
.55.4 17 34/208 Discoid Stone Area B 45 14 41
.55.4 18 34/210 Globular Stone Area A pit 8 34 22 24
.55.4 19 34/211 Discoid Chalk Area A 24 (8) 49 25 55.5
.55.4 20 34/212 Discoid Stone Area B pit 6 40 36 41
.55.4 21 34/213 Discoid Chalk Area B 6 (14) 49 21 58
.55.4 22 34/214 Discoid Chalk Area B pit 37 58 26 73.3
.55.4 23 34/215 Discoid Chalk B pit 1 (5) 45 23 54
.55.4 24 34/218 Discoid Chalk A pit 21 35 12 17
.55.4 24 34/220 Discoid Chalk B pit 2 34 13 15
.55.4 25 34/221 Discoid Chalk Area A 838 33 16 27
.55.4 27 34/225 Discoid Chalk B pit 24 (7) 39 21 35
.55.4 28 34/226 Discoid Chalk D Pit D (2) or A ditch 10? 39 14 30
.55.4 32 34/230 Discoid Stone B pit 8 48 12 37
.55.4 72 35/264 Globular Chalk B pit 14 ? 32 22 34
.55.4 73 35/311 Discoid Chalk Unrecorded 41.25 17.5 16.25
.55.4 74 35/331 Globular Chalk D pit 12 36.8 23 34.7
.55.4 75 35/360 Discoid Chalk Unrec. 44 17 14
.55.4 79 35/415 Discoid Chalk Unrecorded 44 25 52.6
.55.4 80 35/546 Discoid Chalk Pre Hut D 42 19 38
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
.55.4 81 35/565 Discoid Chalk Pre Hut D 57.5 17.5 38.75
.55.4 83 35/605 Discoid Chalk D2 (175) 28.4 11.5 7.3
.55.4 84 35/611 Discoid Chalk Site ? (363) 35 16 21
.55.4 85 35/622 Discoid Chalk DR (365) 31.6 9.4 8.4
.55.4 86 35/684 Discoid Chalk 1103 52.6 17 48.4
.55.4 87 35/712 Globular Chalk Pit F 20 (6) 45.5 26.6 49
.55.4 88 35/728 Globular Chalk 72 (3) 46 25 55
.55.4 89 35/729 Discoid Chalk Pit F 22 (3) 52 20 36.6
.55.4 90 35/735 Globular Chalk Pit F 22 (6) 43.3 21 37.7
.55.4 92 35/896 Discoid Chalk F 623 37 30 46
.55.4 93 35/902 Discoid Chalk F 1258 64 26 96
.55.4 94 35/912 Discoid Chalk F 549 53.3 20 47.7
.55.4 95 35/915 Discoid Chalk F 1258 32 13 12
.55.4 97 35/1014 Discoid Chalk F 849 57.6 16.4 41
.55.4 99 35/1068 Discoid Chalk F 1296 M2 38 15.7 15
37/1172 Discoid Chalk Q dump 51 21 34.4
55.4 100 35/1176 Discoid Chalk D ditch 19 (2) 44 18 27.3
55.4 101 35/1177 Discoid Chalk D Pit 19 38 20 22
55.4 102 35/1083 Discoid Chalk F (169) 46.3 19 32.6
55.4 122 36/483 Discoid Chalk H (1348) 35 14 19
55.4 124 36/594 Discoid Stone L pit LA (1) 55 21.6 16.6
55.4 127 36/812 Discoid Chalk L p.h. 19 52 18 50
55.4 128 36/827 Discoid Chalk L pit 12 (8) 43 23 46
55.4 129 36/828 Discoid Chalk L pit 12 (10) 52 13 64
55.4 131 36/1118 Discoid Stone L p.h. 124 28 15 16
55.4 134 36/1305 Discoid Chalk A 4 51.25 16.25 26.25
55.4 135 36/1420 Discoid Chalk B pit 14 50 28.75 46.25
55.4 136 36/1568 Discoid Chalk G pit H (8a) 45.3 16 13.3
55.4 137 36/1569 Discoid Chalk G (8a) 36 17 22
55.4 142 37/93 Discoid Chalk Q (4) 36 14 15
55.4 143 37/193 Discoid Chalk Q (4) 43.3 23.3 31
55.4 144 37/225 Discoid Chalk Q Hearth B 41 17 21
55.4 146 37/563 Globular Chalk Q (2) 39 27 36
55.4 152 35/1094 Discoid Shale Q pit 42 35 17 30
55.4 154 37/1137 Globular Chalk Hearth B (5) 41 23 50
Table 10: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1939.55.22 24 Triangular Fired clay Site E (673) 104.4 83.3 477.7
39.55.4 34/167 Triangular Chalk Hut D B2 82 59 781
39.55.4 34/200 Discoid Chalk A 21 99 34 378.75
55.4 13 34/203 Discoid Chalk B 14 (10) 103 31 379
55.4 29 34/227 Discoid Chalk Area B unstrat 75 30 193
55.4 38 34/259 Triangular Chalk Area A ? 1163 84 1694.7
55.4 40 34/275 Elongated Chalk B pit 17 (6) 85 72 1255
55.4 54 35/18 Rounded Chalk D pit D12 (6) n62 1164 83 1539
55.4 68 35/124 Triangular Chalk Pit 6 (826) 1868 93.3 2237.3
55.4 71 35/145 Elongated Chalk Under hut D 93.6 71.5 979
55.4 82 35/568a Elongated Chalk Pre Hut D 75 75 243
55.4 91 35/739 Discoid Chalk Pit F 22 (7) 115 31.25 216.25
55.4 103 35/1201 Rounded Chalk D refill 48 20 40
5.44 110 35/1361/1 Elongated Chalk E (422) 15-16 88.8 76.6 1570
55.4 113 35/1361/2 Triangular Chalk E (422) 15-16 Hearth A 98 102 1820
55.4 133 36/1250 Elongated Chalk Pit G4 A (2) 110.5 65 1064
55.4 145 37/498 Rounded Chalk P3 95 21 214
55.4 149 37/831 Elongated Chalk Q pit 36 (2) 70 59 973.75
55.4 164 Triangular Chalk L pit L1 96 44 613

Appendix 12: Hampshire artifact data
Danebury
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
1563 Discoid Fired clay P1333 37 24 39
1643 Discoid Fired clay P1569 51 28 87
1654 Discoid Chalk P1548 30 17 14
1704 Discoid Chalk P7781 47 29 63
1770 Globular Fired clay F7179 47 32 70
1830 Discoid Chalk P1978 79 18 122
1928 Cylindrical Chalk F838 50 25 68
1935 Discoid Chalk P2261 35 15 19
1977 Discoid Chalk P2256 35 16 18
1986 Discoid Chalk P2256 62 23 91
1990 Globular Fired clay P2199 85 28 80
1994 Globular Fired clay F947 38 24 40
2048 Discoid Chalk F920 65 25 104
2143 Discoid Chalk F1257 41 17 45
2190 Discoid Chalk P2346 57 21 57
2408 Globular Fired clay P1127 42 42 84
2409 Discoid Chalk F2507 34 18 18
2427 Discoid Chalk F1562 42 17 28
2466 Discoid Chalk P2531 53 20 58
2478 Conical Chalk Ph 9826 39 23 24
2496 Conical Fired clay P2510 40 22 80
2567 Discoid Chalk F1681 44 24 53
2617 Discoid Chalk P2579 64 23 94
2692 Cylindrical Chalk F1941 41 27 86
2750 Cylindrical Chalk F356 45 31 84
2759 Conical Chalk F2005 41 24 35
2762 Discoid Chalk F1997 70 23 111
2775 Discoid Chalk F1997 29 13 11
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
602 Triangular Chalk P530 141 72 1808
1304 Triangular Chalk P474 474 97 2247
1313 Elongated Chalk P1026 178 87 1890
1315 Triangular Chalk P1026 138 146 1540
1316 Triangular Chalk P1026 134 78 1970
1330 Elongated Chalk P1070 121 103 1958
1331 Elongated Chalk P1026 157 157 2160
1332 Triangular Chalk P1026 132 101 2523
1333 Elongated Chalk P1026 138 138 1866
1338 Triangular Chalk Ph 3507 169 78 1369
1339 Triangular Chalk P1030 158 60 1682
1507 Rounded Chalk P1207 121 121 443
1542 Rounded Fired clay P1224 86 86 218
1547 Elongated Chalk P1299 92 92 1114
1550 Triangular Chalk P1334 129 74 1324
1551 Elongated Chalk P1334 94 66 1324
1559 Elongated Chalk P1334 125 59 1335
1560 Triangular Chalk Unknown 137 65 1380
1561 Elongated Chalk P1393 101 101 1418
1562 Elongated Chalk P1393 116 116 1238
1566 Elongated Chalk P1393 109 109 1221
1567 Elongated Chalk P1395 102 102 1208
1568 Elongated Chalk Ph 4612 102 102 1264
1572 Elongated Fired clay P1298 125 64 1212
1573 Triangular Fired clay Ph 4568 98 52 528
1575 Triangular Chalk Ph 4611 120 52 951
1576 Elongated Chalk Ph 4611 75 75 979
1577 Triangular Chalk Ph 4611 125 67 1180
1580 Elongated Chalk P1285 126 47 1251
1586 Elongated Chalk P1285 111 111 1898
1618 Triangular Fired clay Ph 6302 91 79 1612
1632 Rounded Chalk P1456 105 105 1580
1647 Triangular Fired clay Ph 1530 88 73 1736
1684 Triangular Fired clay Ph 7474 119 71 1104
1686 Triangular Fired clay Ph 7474 147 60 1137
1687 Triangular Fired clay Ph 7474 108 77 1173
1689 Triangular Fired clay Ph 7474 113 78 1068
1690 Triangular Fired clay Ph 7474 123 74 1301
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1694 Triangular Fired clay Ph 7474 119 72 1357
1707 Triangular Fired clay P1820 81 100 1126
1708 Triangular Fired clay P1687 115 68 1024
1709 Rounded Chalk P1790 113 113 348
1732 Triangular Fired clay P1768 155 76 2184
1737 Triangular Chalk P1727 152 84 2622
1739 Triangular Chalk P1727 165 165 1958
1904 Triangular Fired clay P2110 130 64 1335
1906 Triangular Fired clay P2047 102 62 1140
1923 Triangular Fired clay P1768 147 78 2543
1924 Triangular Fired clay P1768 134 80 1690
1931 Rounded Chalk F851 118 118 518
1940 Rounded Chalk F868 63 75 170
1941 Triangular Fired clay P2259 154 68 1460
1956 Elongated Chalk P2200 149 149 2274
1957 Elongated Chalk P2200 145 145 2257
2000 Elongated Chalk P2200 92 52 747
2001 2005
2008 Elongated Chalk P2200 87 138 1328
2006 Rounded Chalk P2200 94 74 760
2011 Elongated Chalk P2200 137 58 2012
2086 Elongated Chalk P2204 91 91 927
2439 Triangular Fired clay P2534 72 40 600
2489 Rounded Chalk P2509 111 111 202
2497 Triangular Fired clay P2541 123 65 1144
Easton Lane
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
282 Cylindrical Fired clay F4560 32 46 65
Table 4: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
9 Triangular Fired clay Trench AA F317A 186 60 1992
62 Cylindrical Fired clay Trench B 115 102 1338
283 Elongated Chalk F4560 73 73 1196
284 Rounded Chalk F4560 100 100 929
287 Elongated Chalk F4560 78 78 973
288 Elongated Chalk F4560 85 85 1025
292 Rounded Chalk F4567 118 118 1340
294 Triangular Chalk F4560 97 97 1144
Kennel Farm
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
21 Discoid Chalk F448 42 20 3
Table 6: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
1 Triangular Fired clay F22 192 76 3062
2-3 Triangular Fired clay F59.2 140 88 1650
4 Triangular Fired clay F23.9 120 85 2426
6 Triangular Chalk F68.3 133 73 1100
7 Triangular Fired clay F22 130 93 2583
42 Triangular Fired clay F182 170 90 3156
43 Triangular Fired clay F182 172 64 3183
Winnall Down
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
37 Triangular Stone F1065 162 42 1591
49 Triangular Fired clay F1631 Fragmented 92 2595
55 Triangular Fired clay F2676 172 89 2548
74 Triangular Chalk F2676 185 72 1968
75 Triangular Chalk F2676 145 83 1697
77 Triangular Chalk F2676 142 74 1942
81 Triangular Chalk F2676 165 86 2012
86 Triangular Fired clay F2676 178 82 2061
152 Triangular Fired clay F3901 159 63 1803
152b Triangular Fired clay F3901 142 90 1894
157 Triangular Fired clay F3901 142 93 1664
244 Triangular Fired clay F2630 183 104 3331
255 Triangular Fired clay F2630 181 82 2846
Zion Hill
Table 8: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
6000 Cylindrical Fired clay F5015 83 99 862
6001 Cylindrical Fired clay F5015 87 84 672.5
6002 Cylindrical Fired clay F5015 72 92 522
6003 Triangular Fired clay F5013 153 106 2760
6005 Triangular Fired clay F5003 109 77 1490

Appendix 13: Kent artefact data
East Wear Bay, Folkestone
Table 1: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
22 Discoid Reused pot 24 43 10.5 23
33 Discoid Sandstone 103 34.7 10 16
86 Discoid Reused pot 207 62 5.3 32
100 Discoid Reused pot 217 47 10.5 24
113 Discoid Reused pot 219 7.3 8.4 16
127 Discoid Stone 216 43 16 29
146 Globular Fired clay 166 35 21 24
179 Discoid Sandstone 85 42 8 20
334 Discoid Sandstone 505 38 8.5 18
428 Discoid Chalk 544 56.5 14 33
623 Discoid Reused pot 546 51 11.5 35
639 Discoid Reused pot 621 16 8 8.75
685 Discoid Reused pot 612 38.5 7.5 13
696 Discoid Sandstone 623 49.5 15 46
721 Discoid Reused pot 612 29.5 5.5 7
853 Globular Fired clay 698 36.5 29.5 24
895 Discoid Sandstone 748 44.7 13.5 32
937 Discoid Reused pot 796 35 8 16
982 Discoid Fired clay 975 29.7 12 11
1057 Discoid Reused pot 215 38 7 16
1510 Conical Fired clay 1517 39 23 28
1522 Discoid Reused pot 1537 57 12 43.75
1547 Globular Fired clay 1546 26 20 13
1639 Discoid Reused pot 1567 25.5 7 28
Table 2: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
612 Triangular Fired clay 570 180 132 1423.75
689 Triangular Fired clay 689 130 158 1340
Highstead
Table 3: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
40 Conical Fired clay Pit complex B15 40 20 52
41 Discoid Reused sherd Pit B146 40 6 45
Table 4: Loom weight data
Museum No. Tool Type Material Location Width (mm) Depth (mm) Weight (g)
70 Pyramidal Fired clay Enclosure ditch A24 109 109 1750
99 Pyramidal Fired clay Pit B168 70 100 1400
Sunset Caravan Park
Table 5: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
121 Pyramidal Fired clay TR12 (29) 21 36 28
901 Globular Fired clay 2743 37 29 44
1245 Pyramidal Fired clay 157 32 27 24
1246 Globular Fired clay 2345 32 24 29
1322 Discoid Fired clay 1989 39 19.5 30
Thanet Earth
Table 6: Spindle whorl data
Museum No. Form Material Location Diameter (mm) Depth (mm) Weight (g)
47 Conical Fired clay 8800 33.5 27 14
Table 7: Loom weight data
Museum No. Form Material Location Width (mm) Depth (mm) Weight (g)
22 Triangular Fired clay Context 8643 183.75 60 1177.5
25 Triangular Fired clay Context 8637 150 54 291
50 Triangular Fired clay Context 8795 157.5 70 1033.3
61 Triangular Fired clay Context 8620 155 56 801.7
127 Triangular Fired clay Context 12108 166 58.5 932.5
!!
!
!
!
Bodrifty
Trevisker
The Rumps
Carn Euny
Trevelgue Head
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS usercommunity
±
x 4: Cornwall HLC with sites
0 4.5 9 13.5 182.25 Kilometers
Legend
! Sites
Ancient Woodland
Coastal Rough Ground
Dunes
Upland Rough Ground
Water: Natural
Cornwall environmental data
Appendix 10. Cornwall environmental data
270
!!
!
!
!
Dainton
Milber Down
Mount Batten
Kent's Cavern
Blackbury Castle
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS usercommunity
±
0 7 14 21 283.5 Kilometers
Legend
! Sites
Ancient woodland
Dunes
Intertidal
Marsh
Mud
Mud & sand
Outcrop/scree/cliffs
Rough ground
Sand
Water
Watermeadow
Devon environmental dataAppendix 11. Devon environmental data
!!
!
!
!
!
Hod Hill
Maiden Castle
Allard's Quarry
Hengistbury Head
Gussage All Saints
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS usercommunity
±
0 6 12 18 243 Kilometers
Legend
! Sites
Beach
Common
Dunes
Heath
Lake
Mudflats
Rough ground
Sand
Scrub
Seaside
Watermeadows
Dorset environmental dataApp ndix 12. Dorse  environmental data
!!
!
!
!
!
!
Danebury
Winnal Down
Kennel Farm
Easton Down
Zions Hill Farm
Hengistbury Head
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS usercommunity
±
0 6 12 18 243 Kilometers
Legend
! Sites
Coastal wetlands
Common heathland
Enclosed heath and scrub
Marsh and rough grazing
Mud flats
Prehistoric and Roman
Salt marsh
Salterns
Shingle and dunes
Unenclosed heath and scrub
Water meadows
Hampshire environmental dataAppendix 13. Hampshire environmental data
!!
!
!
HighsteadThanet Earth
East Wear Bay
Sunset Caravan Park
Esri, HERE, Garmin, (c) OpenStreetMap contributors, and the GIS usercommunity
±0 6 12 18 243 Kilometers
Legend
! Sites
Prehistoric and Roman
Coastal wetlands
Common Marsh
Dunes
Marsh and rough grazing
Mud flats
Other pre-1810 Woodland
Pre-1810 scarp & steep valley-side woodland
Salt marsh and estuarine resources
Salterns
Shingle
Appendix 14. Kent environmental data