Iron Age Mirrors: A Biographical Approach 1407307037, 9781407307039

Citation preview

CONTENTS Preface and Acknowledgements

iv

List of figures

vi

List of tables

ix

Chapter 1: Introduction 1.1 Iron Age mirrors Mirror form Metal composition Location of finds Find context 1.2 Literature review Summary of existing work Dating “Origins” Decoration Summary 1.3 Research aims

1 1 1 1 1 2 3 3 4 5 6 6 7

Chapter 2: The biographical approach 2.1 Object biography 2.2 Archaeology and the biographical approach Use-life analysis Summary 2.3 Problems with existing archaeological applications of the biographical approach Biography doesn’t have to be chronological Object vs. human lives Concepts of time Summary 2.4 Anthropology and biography 2.5 Biography as an understanding of social relationships 2.6 A biography of mirrors Events in the life of a mirror 2.7 Biographical mirrors

8 8 8 8 9 9 9 9 10 10 11 11 12 12 13

Chapter 3: Making mirrors 3.1 Introduction 3.2 Approaches to metalworking 3.3 Making bronze mirrors Mining and smelting bronze Casting the handle Making and decorating a plate Assembling a mirror Summary 3.4 Iron mirrors Location of ores Smelting iron ore Location of smelting sites The Foulness Valley Iron smithing Making an iron mirror Summary 3.5 Mirrors made of a combination of bronze and iron components Summary 3.6 Conclusion

14 14 14 15 15 16 16 17 18 18 19 19 19 20 20 21 21 21 22 22

Chapter 4: Re-examining mirror decoration 4.1 Introduction 4.2 Re-assessment of the existing categorisation of mirror decoration 4.3 Analytical processes Terminology

24 24 24 24 25

i

4.4 Mirror-style motif repertoire Positive motifs Negative motifs 4.5 Design rules Design construction Examples of roundel construction 4.6 Unused space 4.7 Complexity 4.8 Vertical symmetry 4.9 Summary 4.10 Why decorate? The ‘effect’ of decoration Who understood the design rules? The ‘sensory environment’ Visual qualities and context The visual experience Summary 4.11 Why mirrors were decorated

25 25 26 27 27 34 34 35 36 36 38 38 38 39 39 40 40

Chapter 5: Analysis of use/wear/repair 5.1 Introduction 5.2 Results Wear to the terminal ring Stress marks between plate and handle Decoration underlying handle or rim Polishing on the handle Evidence of repair Missing edging 5.3 Interpretation 5.4 Conclusions

41 41 42 42 43 43 43 43 43 43 44

Chapter 6: Mirror form 6.1 Mirror component parts ‘reference’ material culture Mirror handles Mirror plate Mirror motifs identified on other artefacts Bronze-binding Techniques of manufacture Summary 6.2 Mirrors in Use: ‘techniques of the body’ ‘Gazing’ at reflection The mirror as ‘third eye’ Mirrors used to ‘evoke’ light 6.3 Conclusion

45 46 46 46 47 48 48 48 48 49 49 50 50

Chapter 7: The mirror as social and cultural nexus 7.1 Making mirrors 7.2 Mirror decoration 7.3 Analysis of use/wear/repair 7.4 Mirror form 7.5 Mirrors as social and cultural nexus

51 51 51 51 51 52

Chapter 8: Dating audit 8.1 Introduction to dating mirror deposition 8.2 Why is establishing a chronology for mirrors important? 8.3 The chronology of iron mirrors 8.4 Mirror decoration – regional and temporal trends South-eastern mirrors ‘Western’ mirrors Sub-regional groupings Temporal trends 8.5 Interpretation of mirror chronology 8.6 Conclusion

53 53 53 54 55 55 55 55 56 56 57

ii

Chapter 9: Depositional context 9.1 Introduction 9.2 Burial contexts East Yorkshire South-western cist burials containing mirrors Cremation burials from southeast England containing mirrors ‘Western’ mirrors Mirrors from graves outside Britain 9.3 Mirrors from other contexts Mirrors found in settlement contexts Mirrors found in watery contexts Mirrors with little or no context 9.4 Conclusions

58 58 58 58 63 64 69 70 71 71 72 72 72

Chapter 10: Biographical mirrors 10.1 Introduction 10.2 Mirrors as socially powerful objects Mirrors associated with ‘mature’ adults Summary Gendering mirrors: archaeological evidence Summary Who ‘owned’ mirrors Mirrors as an indicator of status Summary 10.3 Patterns in the deposition data Position of mirrors in the grave Summary Prominent locations in the landscape Mirrors deposited in covering 10.4 The social relations objectified when mirrors were deposited East Yorkshire – Wetwang Village Southwest – The Bryher mirror burial South-eastern – Aston mirror burial ‘Western’ – the Portesham mirror burial Settlement context – Holcombe mirror 10.5 Conclusions

74 74 74 74 74 75 76 76 76 77 77 77 78 78 79 80 80 80 81 81 82 82

Chapter 11: Conclusions 11.1 Introduction 11.2 Mirrors and the British Iron Age 11.3 The application of the biographical approach to archaeological data 11.4 Biographical mirrors

83 83 83 84 84

Appendix A: Catalogue of British-type mirrors

85

Appendix B: Chronology of Iron Age mirrors

128

Appendix C: Iron Age objects decorated with mirror motifs

136

Appendix D: Mirror handles

141

Appendix E: Trace element analysis of the metal content of mirrors

146

Appendix F: Mirrors with decorated plates

148

Appendix G: The size of mirrors

149

Appendix H: Visual examination of mirrors in museums

152

References

161

iii

PREFACE & ACKNOWLEDGEMENTS This book is based on research conducted for a PhD thesis funded by the AHRC examining British Iron Age mirrors from a biographical perspective (Joy 2008a). The basic premise behind the thesis and this book is that mirrors were active in the formation, maintenance and renegotiation of Iron Age social relationships. The biographical approach is employed as a methodology to reveal some of these relationships. In the past objects like mirrors, which have been labelled as ‘Celtic Art’, were separated from their archaeological context and examined as art pieces. They were seen to passively reflect past activities or identities rather than being intimately bound-up in the creation of society. In this book Celtic Art is re-integrated into mainstream archaeology (see MacDonald 2007a) by examining the social lives of mirrors. Mirrors are amongst the most well known Iron Age objects. They are of a type which is peculiar to Britain and are significantly different in form from contemporary Greek, Etruscan and Roman mirrors. 58 mirrors are known. They are made of bronze and iron, or sometimes a combination of bronze and iron components. Mirrors comprise a handle and a reflective plate, which is often decorated with intricate and free-flowing designs. Some plates are also rimmed. Mirrors are found throughout Britain; two have been discovered in Ireland and two others are known from the continent. They are most commonly found in graves; but were also deposited in bogs and rarely at settlements. They date to the mid-late Iron Age. Previous studies have concentrated on describing decoration and constructing a stylistic sequence. Mirrors have often been interpreted as high status objects and the property of women and are viewed as passive vanity objects. Decoration is seen as ‘pleasing to the eye’. Building on existing analyses this book uses information collected from recent discoveries to interrogate the validity of these interpretations. Objects do not passively reflect ancient society; they were integral to it (Shanks & Tilley 1987a; 1987b; Thomas 1996; Tilley 1996). With this in mind an alternative interpretation of mirrors is put forward. Using the biographical approach to material culture, which takes account of the fact that like people objects can accumulate histories (Kopytoff 1986; Marshall & Gosden 1999), mirrors are reinterpreted and their biographies charted. The biographical approach has been successfully applied to archaeological studies which are anthropologically or historically situated as the data is rich enough to reconstruct the birth, life and death of an object. However, the application of the biographical approach to prehistoric material culture has been less successful. Where it has been used, conclusions are often restricted to indicating that an object must have had a ‘rich’ and ‘complex’ biography (Whitley 2002; Joy 2009a). This book will form a case study to test the applicability of the biographical approach to prehistoric objects and the application of the biographical approach to prehistoric material culture is evaluated by constructing biographies for Iron Age mirrors. To do this the biographical approach is modified to accommodate the limitations of prehistoric

iv

data. Like people, objects are active agents; objects can employ social agency and social relationships can be structured around them (Gell 1998: 17-19; De Marrais et al. 2004). If biography is seen as the sum of the social relationships that constitute a person or object, then it is possible to re-construct parts of the biographies of objects by pinpointing the archaeological information which will allow us to consider social relations (Joy 2009a). In order to construct prehistoric biographies, detailed knowledge of the object and others like it is required. Consequently, it is necessary to study an object corpus. The need for detailed technological and contextual analyses of Iron Age artefacts has recently been highlighted (Haselgrove et al. 2002). A number of mirrors have been found over the past 35 years and these finds have been well reported. However, no-one has assimilated this new information in to a study of mirrors as a group and used it to test existing interpretations. BOOK STRUCTURE This book is divided into three main sections. In the first section mirrors are introduced (Chapter 1) as is the theoretical methodology (Chapter 2). Chapter 1 explains what mirrors look like, the contexts they are found in and how they have been studied in the past to pinpoint what we do not yet understand about them and what needs further clarification. In Chapter 2 the biographical approach to artefacts is outlined; how it has been used in archaeology and how the approach will be utilised to expand our knowledge of mirrors and the broader Iron Age context by reconstructing the relationships that constitute mirrors and their biographies. In the second section a biographical account of the relationships that were built into mirrors at the time of their manufacture and written into mirrors throughout their social lives is provided. Chapter 3 examines evidence for the production of Iron Age metal artefacts as well as investigating the context of the production of metalwork in ethnographic contexts. The aim is to develop an understanding of the technology of mirror production, the relationships established through their production and the potential future trajectories of the life of a mirror set out at the time of manufacture. In Chapter 4 mirror decoration is examined. The decoration on the plates of some mirrors is studied using an archaeological rather than art historical methodology. Analysis of results will focus on the artefact rather than motifs; how does the decoration relate to the mirror it ornaments? What does the decoration do? Chapter 5 summarises the results of a programme of visual examination of the physical condition of surviving mirrors. Over 30 mirrors were examined for signs of wear, polishing and repair; clues which can indicate how mirrors were used and inform us about their social lives. Chapter 6 examines the form of mirrors. The component parts will be compared to other Iron Age artefacts to establish if any connections can be made. The potential relationships formulated between components are also considered.

you also to Mary Douglas, Andy Jones, Vincent Megaw, Naill Sharples and Ian Stead.

In the third section deposition context is examined. Chapter 8 is the first comprehensive dating audit of all Iron Age mirrors. This chapter pinpoints regional and temporal trends in deposition and situates the analysis within the Iron Age context. In Chapter 9 all of the deposition data is collected. Chapter 10 is an analysis of the results of Chapter 9. Taking account of the information compiled in the analysis of the biographical information written into mirrors, it examines how the relationships formulated at the time of manufacture and throughout their lifetimes, were referred to and manipulated when mirrors were deposited.

Adam Gwilt, JD Hill, Fraser Hunter, John Fenn, Vincent Megaw, Ben Roberts, Paul Sealey and Mansel Spratling allowed me to see articles prior to publication. Mandy Jay kindly gave me permission to publish radiocarbon dates for the Wetwang Slack and Wetwang Village burials. Duncan Garrow allowed me to publish the new date for Garton Station. Many museum curators and employees allowed me to examine their mirrors and gave me help and advice. Thank you: Louise Allen, Evan Chapman, Jodie Deacon, Adam Gwilt, Fraser Hunter, Jim Inglis, Paul Sealey, Jenny Stevens, Brett Thorn, Nick Wickendon and Peter Woodward. Apologies to anyone I have missed.

In Chapter 11 the implications of these findings for wider research and the future of the application of the biographical approach to archaeological research, will be assessed. Sections of Chapters 2, 4 and 10 reproduce or contain text that is similar to that subsequently published elsewhere (Joy 2008b; 2009a; in press); it has been included here for the sake of internal coherency within this study.

Many people helped me with, and allowed me to reproduce, illustrations. Stephen Crummy and Craig Williams created new illustrations as well as compiling and scanning images from the British Museum archive. Particular thanks must go to Mansel Spratling and Paul Sealey. Thanks also to Gil Burleigh, John Creighton, Adam Gwilt, Fraser Hunter, Vincent Megaw and Sally Worrell.

ACKNOWLEDGEMENTS First and foremost I would like to thank my PhD supervisor Yvonne Marshall. Without her constant encouragement, guidance and support this book would never have been written. I would also like to thank Tim Champion for first suggesting mirrors as a research topic and for his role as an advisor on all things Iron Age. Louise Revell provided helpful advice. Chris Gosden and Vedia Isset also gave me very useful feedback after examining my thesis. JD Hill, first as a museum curator, and secondly as a colleague, has provided me with invaluable help and support. I would also like to thank Mansel Spratling for commenting on a draft of this book and our many conversations about mirrors. Andrew Fitzpatrick, Mel Giles and Adam Gwilt provided me with invaluable information and advice throughout. Thank

John Cooper, Babis Gerfalakis, Duncan Garrow, Darren Glazier, Shaun Moyler, Fotini Kofidou, Kristin Oma, Ben Roberts, Fraser Sturt and especially Mike Lally, Ross Thomas and Julian Whitewright have all offered advice, or suggested I read various references, often over a pint. I would also like to thank various colleagues at the British Museum for their help and support: thank you Richard Hobbs, Ralph Jackson, Ian Leins, Sonja Marzinzik, Ben Roberts and especially Jonathan Williams and Jill Cook. Finally, I want to thank Miriam for always being there.

v

LIST OF FIGURES Figure 1.1 Mirror Nomenclature (drawn by author) Figure 1.2 Distribution of bronze, iron and composite mirrors (drawn by author) Figure 1.3 Distribution of mirror finds (drawn by author) Figure 1.4 Mirrors with deposition contexts (drawn by author)

1 3 4 5

Figure 2.1 Re-constructing an object life (drawn by author)

10

Figure 3.1 Processes in the construction of the Portesham mirror (drawn by author) Figure 3.2 Processes in the construction of the Wetwang Village mirror (drawn by author) Figure 3.3 Processes in the construction of the Balleybogey mirror (drawn by author)

18 22 23

Figure 4.1 A fin (drawn by author) Figure 4.2 A cusp (drawn by author) Figure 4.3 Trumpet motifs (drawn by author) Figure 4.4 Designs within circles (drawn by author) Figure 4.5 The armadillo motif (drawn by author) Figure 4.6 Roundels on Bromham and Great Chesterford (drawn by author) Figure 4.7 Three additional mirror motifs (drawn by author) Figure 4.8 Negative motifs (drawn by author) Figure 4.9 A diagram to explain mirror composition (drawn by author) Figure 4.10 Fox’s outlines of mirror design (drawn by author) Figure 4.11 Six different outlines of mirror design (drawn by author) Figure 4.12 Dividing the design-field into smaller units (drawn by author) Figure 4.13 Design of the Pegsdon mirror (drawn by author) Figure 4.14 The sequence of Aston mirror design (drawn by author) Figure 4.15 Construction of the design of the Birdlip mirror (drawn by author) Figure 4.16 Sequence of design of the Portesham mirror (drawn by author) Figure 4.17 Design of the Old Warden I mirror (drawn by author) Figure 4.18 Desborough roundel (drawn by author) Figure 4.19 Portesham roundel (drawn by author) Figure 4.20 Mirrors viewed in negative (drawn by author)

25 25 25 26 26 27 27 28 29 30 30 30 31 31 32 33 34 35 36 37

Figure 5.1 Common areas of wear/polishing/repair (drawn by author)

41

Figure 6.1 Mirror components refer to other items of later Iron Age material culture (drawn by author) Figure 6.2 Comparison of mirror handle grip and horse bit (drawn by author) Figure 6.3 So-called ‘bean can’ from Wetwang Slack burial (© Trustees of the British Museum)

45 46 47

Figure 8.1 Dating audit of the deposition of mirrors (drawn by author) Figure 8.2 Calibrated radiocarbon dates for East Riding mirrors Figure 8.3 Sections of the decoration on the Latchmere Green and Rivenhall mirrors (drawn by author)

54 54 56

Figure 9.1 The Wetwang Slack burial (drawn by author) Figure 9.2 The Wetwang Village grave (drawn by author) Figure 9.3 The Bryher mirror burial (drawn by author) Figure 9.4 The Aston burial (drawn by author) Figure 9.5 The Chilham Castle burial (drawn by author) Figure 9.6 The Dorton burial (drawn by author) Figure 9.7 The Latchmere Green burial (drawn by author) Figure 9.8 Contents of grave 13, King Harry Lane (© Trustees of the British Museum) Figure 9.9 The Portesham burial (drawn by author)

60 61 64 65 66 66 67 68 69

Figure 10.1 Location of the Chilham Castle burial (drawn by author) Figure 10.2 Location of the Dorton burial (drawn by author) Figure 10.3 Location of Latchmere Green burial (drawn by author)

78 78 79

Figure A1 Arras I (© Trustees of the British Museum) Figure A2 Aston (drawn by author) Figure A3 Balmaclellan (drawn by author, by kind permission of the National Museum of Scotland) Figure A4 Billericay I (drawn by author, by kind permission of Colchester Museum) Figure A5 Billericay II (drawn by author, by kind permission of Colchester Museum) Figure A6 Birdlip (drawn by author, after Jope 2000: Plate 244)

88 89 90 91 92 93

vi

Figure A7 Brecon Beacons (drawn by author, by kind permission National Museum of Wales) 94 Figure A8 Bromham (drawn by author) 95 Figure A9 Bryher (drawn by Stephen Crummy © Trustees of the British Museum) 96 Figure A10 Chettle (drawn by Stephen Crummy © Trustees of the British Museum) 97 Figure A11 Chilham Castle (© Trustees of the British Museum) 98 Figure A12 Colchester I (drawn by author, by kind permission of Colchester Museum) 99 Figure A13 Colchester II (drawn by Stephen Crummy © Trustees of the British Museum) 100 Figure A14 Desborough (© Trustees of the British Museum) 101 Figure A15 Dorton (drawn by author) 102 Figure A16 Garton Slack (drawn by author, after Brewster 1980) 103 Figure A17 Glastonbury (drawn by Mansel Spratling) 104 Figure A18 Great Chesterford (drawn by author) 105 Figure A19a Holcombe (drawn by Philip Compton © Trustees of the British Museum) 106 Figure A19b Holcombe (drawn by Philip Compton © Trustees of the British Museum) 107 Figure A20 Lambay Island (drawn by author, after Rynne 1976) 108 Figure A21 Latchmere Green (kind permission of John Creighton) 109 Figure A22 Llechwedd-du (drawn by Mansel Spratling) 110 Figure A23 Maiden Castle & Portland II (drawn by author) 111 Figure A24a Nijmegen reconstructed (drawn by M. O. Millar © Trustees of the British Museum) 112 Figure A24b Nijmegen (drawn by author, after Lloyd-Morgan 1981a) 113 Figure A25 Old Warden I (drawn by author) 114 Figure A26 Pegsdon (drawn by Garth Denning © North Hertfordshire Archaeological Society) 115 Figure A27 Portesham (drawn by author, kind permission of Andrew Fitzpatrick) 116 Figure A28 Rickling (drawn by Stuart MacNeil, kind permission of Paul Sealey), Rivenhall I (drawn by author) 117 Figure A29 St. Keverne (drawn by author) 118 Figure A30 Stamford Hill I, II & III (drawn by author) 119 Figure A31 Disney (drawn by Stephen Crummy © Trustees of the British Museum) 120 Figure A32 Gibbs (drawn by Stephen Crummy © Trustees of the British Museum) 121 Figure A33 Mayer (drawn by author) 122 Figure A34 ‘Oxfordshire’ (drawn by Donna Watters © North Hertfordshire Archaeological Society) 123 Figure A35 Akenham (© Portable Antiquities Scheme), Bac Mhic Connain (drawn by author, by kind permission National Museum of Scotland), Badingham (© Portable Antiquities Scheme), Balleybogey (drawn by author) 124 Figure A36 Bridport (drawn by Mansel Spratling), Bulbury (drawn by Mansel Spratling), Carlingwark (drawn by author, by kind permission National Museum of Scotland), Compiègne (drawn by author, after Guillaumet & Schönfelder 2001) 125 Figure A37 Ingleton (© Trustees of the British Museum), Jordan Hill (drawn by author), Lochlee (drawn by author), Old Warden II (drawn by author), Portland I (drawn by author) 126 Figure A38 Westerfield (© Portable Antiquities Scheme), Essex/Sussex border (drawn by author, after Mills 2000), Fison Way (drawn by author), Llanwnda (drawn by author), Merlesford? (drawn by author, by kind permission National Museum of Scotland) 127 Figure B1 Calibrated date for Garton Slack (Harwell 1274) Figure B2 Calibrated date for Garton Slack (OxA-17286) Figure B3 Calibrated date for Wetwang Slack Figure B4 Calibrated date for Wetwang Village Figure B5 The Birdlip brooch (drawn by author, after Staelens 1982) Figure B6 The radiocarbon determinations for Bryher Figure B7 Calibrated results of the Bryher grave Figure B8 The Bryher brooch (drawn by Stephen Crummy © Trustees of the British Museum) Figure B9 The Bryher sword (drawn by Stephen Crummy © Trustees of the British Museum) Figure B10 The Chilham Castle brooches (drawn by author) Figure B11 Portesham brooches Figure B12 St. Keverne brooch Figure B13 Simulation calibrated date for Arras I Figure B14 Simulation calibrated date for Chilham Castle Figure B15 Simulation calibrated date for Aston Figure B16 Simulation calibrated date for Dorton Figure B17 Simulation calibrated date for Latchmere Green Figure B18 Simulation calibrated date for King Harry Lane

vii

128 128 128 128 129 130 130 130 130 131 133 133 134 134 134 134 134 134

Figure B19 Simulation calibrated date for Portesham Figure B20 Summary of estimated dates

134 135

Figure C1 Decoration on Asby Scar sword scabbard (© Trustees of the British Museum) Figure C2 Decoration on the Bugthorpe sword scabbard (© Trustees of the British Museum) Figure C3 Miniature shield 1 from Salisbury (© Trustees of the British Museum)

136 137 139

Figure D1 Handle types (drawn by author) Figure D2 Subdivision of bar handles (drawn by author) Figure D3 Varieties of Type III handle (drawn by author) Figure D4 Mirror handles with loop within their grips (drawn by author) Figure D5 Type V handle (drawn by author) Figure D6 The Merlesford & Fison Way handles Figure D7 Relationship between mirror handle type and deposition date (drawn by author)

141 141 142 142 143 143 145

Figure G1 The relationship between plate width and deposition date

151

viii

LIST OF TABLES Table 1.1 A list of known mirrors

2

Table 4.1 Results of analysis of mirror decoration showing % of used space

36

Table 5.1 Summary of the results of analysis of wear/polishing/repair

42

Table 9.1 Mirrors from burial contexts Table 9.2 Mirror burials from East Riding Table 9.3 East Riding vehicle burials Table 9.4 Mirrors found at settlements Table 9.5 Mirrors from watery contexts Table 9.6 Mirrors from poorly recorded contexts

58 59 62 71 72 72

Table 10.1 Age at death of individuals buried with mirrors Table 10.2 Sexing evidence of individuals buried with mirrors

74 75

Table D1 The geographical distribution of handle types Table D2 Date of deposition of selected handles and handle type

144 145

Table E1 The metal composition of ‘British Type’ mirrors Table E2 The metal composition of selected ‘Roman Type’ mirrors

146 147

Table G1 Mirror dimensions Table G2 The relationships between plate width and region Table G3 The relationship between mirror length and region Table G4 The relationship between plate width and deposition date Table G5 The ratio between length of handle and overall length

149 149 150 150 151

ix

CHAPTER 1 INTRODUCTION This first introductory chapter introduces mirrors as a group of objects defines the corpus and presents the main themes of discussion that have dominated the academic literature. It begins by defining what a British Iron Age mirror is: what they look like; and where they were deposited. The main themes of discussion in the literature are then outlined: how are mirrors interpreted; what aspects are researchers most interested in? To conclude the research aims are presented.

Some of the larger mirrors also have rivets and a bronze edging to make the joint more secure. Many of the plates have decoration on one side and a highly-polished reflecting surface, on the other (Appendix F). Decoration varies in design and complexity and usually consists of elaborate swirls and whirly-gigs. The decoration on some mirrors was applied using compasses and scribing tools (Lowery et al. 1971; 1976).

1.1 IRON AGE MIRRORS 58 mirrors, handles, and plates dating from the mid-late Iron Age have been found in Britain in the past 150 years (Table 1.1 & Appendix A). Various catalogues have been produced by Smith (1909), Bulleid & Gray (1911), Dunning (1928: 77-9), Fox (1949; 1958: 84-105), Fox & Hull (1948), Spratling (1970a), Fox & Pollard (1973) and Fitzpatrick (1997a). There are also a number of reports on more recent finds (Brewster 1980; Rook et al. 1982; Farley 1983; Dent 1985a; Fitzpatrick 1997a; Fulford & Creighton 1998; Parfitt 1998; Hill 2001a; Johns 2006; Burleigh & Megaw 2007).

Iron mirrors were made using different techniques to bronze mirrors. Iron Age furnaces were unable to produce sufficient temperatures to melt iron. Consequently, it was not possible to cast iron objects (Cleere 1972). A typical iron mirror has a small circular plate and a bar shaped handle. Some iron mirrors, like Arras I, East Riding, also have bronze mounts. A number of mirrors were also made from a combination of bronze and iron components. For example, the mirror found at Glastonbury Lake Village had a bronze plate and an iron handle (Fitzpatrick 1997a: 67) and the Maiden Castle mirror has an iron plate with a bronze rim (Wheeler 1943).

Since the publication of the last catalogue – the report on the discovery of the Holcombe mirror (Fox & Pollard 1973) – a number of other mirrors have been discovered. Following the list outlined in Fitzpatrick (1997a: 65) these are: Aston, Hertfordshire (Rook et al. 1982); Chilham Castle, Kent (Anon 1995; Parfitt 1998: 12, no. 3), Colchester II (Hyderabad Barracks), Essex (Sealey 1996: 61; 2006), Dorton, Buckinghamshire (Farley 1983), Latchmere Green, Hampshire (Fulford & Creighton 1998); Llanwnda, Pembrokeshire (Boon 1978-80); Rickling, Essex (Sealey 2006), Rivenhall, Essex (Lloyd-Morgan 1993) and Wetwang Slack, East Yorkshire (Dent 1985a: 90). It is necessary also to add to this list the Portesham mirror (Fitzpatrick 1997a) and the recent discoveries from Pegsdon (Hill 2000; Burleigh & Megaw 2007), Bromham (unpublished) and a handle from the Essex/Sussex border (Mills 2000: 33), a metal detectorist find from Oxfordshire (Burleigh & Megaw 2007), Chettle (unpublished), the Brecon Beacons (Davis & Gwilt 2008: Figures 9.12-3) and Bryher in the Isles of Scilly (Johns 2006; no date), as well as the recent discovery of three handles from Suffolk at Akenham, Badingham and Westerfield.

Figure 1.1 Mirror nomenclature Metal Composition The majority of mirrors are made from copper-alloy (Figure 1.2). Where the composition of this copper-alloy has been analysed, it has been found to be bronze, an alloy of copper and tin, with tin typically around 10% (Appendix E). It is unclear whether bronze mirrors were actually more common than iron mirrors, as bronze is less susceptible to corrosion than iron and the archaeological record may therefore be skewed.

Mirror Form Mirrors have two main component parts: a handle and a plate (Figure 1.1). Some also have a bronze edging to protect the edge of the plate. Plates are round and have a highly polished reflective surface on one side. A typical bronze mirror has a round or kidney-shaped plate, manufactured from a flat piece of bronze, between 200 and 300mm in diameter, which has been hammered to a thickness of around 1mm. Handles are made from cast bronze and are usually no longer than 150mm (Appendix G). Handles range in shape and complexity from a single simple loop or bar handle to an enamelled multiple-loop design (Appendix D). The plate was fixed to the handle using a number of different methods. Some plates were wedged into a slot pre-cast into the top of the handle. This slot seems to have been secure enough to hold the smaller plates in place.

Location of Finds Mirrors have predominantly been found in southern England (Figure 1.3). A group of five iron mirrors have also been found in East Riding. Two have been discovered in Wales and five in Scotland, including a cetacean bone handle from North Uist in the Western Isles. Two mirrors have also been discovered in Ireland, a bronze handle

1

Mirror

County

Context

Old Warden II

Bedfordshire

-

Akenham

Suffolk

-

'Oxfordshire'

-

-

Arras I

East Riding

Inhumation

Pegsdon

Bedfordshire

Cremation

Arras II

East Riding

Inhumation

Portesham

Dorset

Inhumation

Aston

Hertfordshire

Cremation

Bac Mhic Connain

Western Isles

Settlement

Portland I (the Grove)

Portland J (the Grove)

Portland K (the Grove)

Badingham

Suffolk

-

Portland II (the Verne)

Dorset

-

Balleybogey

County Antrim

Watery

Rickling

Essex

-

Balmaclellan

Dumfries & Galloway

Watery

Rivenhall I

Essex

Settlement

Bulbury

Dorset

Hoard

Rivenhall II

Essex

-

Billericay I

Essex

Cremation?

St. Keverne

Cornwall

Inhumation

Billericay II

Essex

Cremation?

Stamford Hill I

Devon

Inhumation

Billericay III

Essex

Burial?

Stamford Hill II

Devon

Inhumation

Birdlip

Gloucestershire

Inhumation

Stamford Hill III

Devon

Inhumation

Brecon Beacons

Powys

Cremation

Westerfield

Suffolk

-

Bridport

Dorset

Inhumation

Wetwang Slack

East Riding

Inhumation

Bromham

Bedfordshire

-

Wetwang Village

East Riding

Inhumation

Bryher

Isles of Scilly

Inhumation

Carlingwark

Dumfries & Galloway

Watery

Chettle

Dorset

Burial?

Chilham Castle

Kent

Cremation

Colchester I (Lexden Grange)

Essex

Cremation

Colchester II (Hyderabad Barracks)

Essex

-

Compiègne

France

Watery

Desborough

Northamptonshire

-

Disney

-

-

Dorton

Buckinghamshire

Cremation

'Essex/Sussex border'

Essex/Sussex

-

Garton Slack

East Riding

Inhumation

Gibbs

-

-

Glastonbury

Somerset

Settlement

Great Chesterford

Essex

-

Holcombe

Devon

Settlement

Ingleton

North Yorkshire

-

Jordan Hill

Dorset

-

Lambay Island

Ireland

Inhumation

Latchmere Green

Hampshire

Cremation

Llechwedd-du

Gwynedd

Burial?

Lochlee Crannog

South Ayrshire

Settlement

Maiden Castle

Dorset

Settlement

Mayer

-

-

Nijmegen

Netherlands

Cremation

Old Warden I

Bedfordshire

-

Table 1.1 A list of known mirrors from a bog deposit at Ballymoney, County Antrim and an iron mirror from an inhumation burial discovered on Lambay Island, near Dublin. The Balleymoney mirror seems to be closely related to examples found in southwest Scotland also found in bogs (Raftery 1983; 1984). Two mirrors have been discovered outside the British Isles and Ireland. They are included in this study because their design and decoration suggests that they originate from Britain. One was found in a grave in Nijmegen, the Netherlands, which also contained a glass vessel (Dunning 1928). Many theories have been put forward as to how a British bronze mirror could have been found in a burial in the Netherlands, it may be a result of strong trade links (ibid: 77), or it could have been taken to the Netherlands by a Roman legionary (Fitzpatrick 1997a: 60). One intriguing fact is that during the late 1st-century AD, Nijmegen was a centre of production for Roman mirrors (Lloyd-Morgan 1977b: 338; 1981a). A handle was also discovered in the River Oise in northeast France (Guillaumet & Schönfelder 2001). This handle is very similar to that on a mirror discovered at Dorton, Buckinghamshire (see Farley 1983; Guillaumet & Schönfelder 2001). Find Context Mirrors have predominantly been found in graves (Figure 1.4), from most of the identified regional burial traditions of the later Iron Age in Britain (see Whimster 1981). They have also been found in settlement contexts at Glastonbury Lake Village (Bulleid & Gray 1911; Coles & Minnitt 1995), Maiden Castle (Wheeler 1943), Holcombe (Fox & Pollard 1973) and Rivenhall (Rodwell & Rodwell 1986; Lloyd-Morgan 1993). The object reported as a mirror from Mucking, Essex (Jones & Jones 1975)

2

Figure 1.2 Distribution of bronze, iron and composite mirrors be items of female toilet, passively reflecting the status of their owners. Even though ideas which are commonplace in mainstream archaeology ascribing agency to artefacts (Gell 1998; Dobres & Robb 2000), examining their roles in day-to-day social relationships and questioning simplistic gender assumptions (see Strathern 1988; Butler 1990; Gero & Conkey 1991; Butler 1993) are beginning to penetrate Iron Age research (e.g. Hill 1997) they have yet to be applied to mirrors and these implicit assumptions are only beginning to be questioned (see Fitzpatrick 1984; Johns 2006).

is most probably a knife. Mirrors have been discovered in watery contexts, especially in bogs in southwest Scotland (Murray 1860-2; MacGregor 1976; Cessford 1997), as well as a bog find from Northern Ireland (Jope 1954; Raftery 1983) and a river (Guillaumet & Schönfelder 2001). Unfortunately, many of the mirrors, such as Desborough and Gibbs, have no recorded context. It has been assumed, because these mirrors are relatively complete, that they originated from burials (e.g. Smith 1909). 1.2 LITERATURE REVIEW The traditional interpretation of mirrors in the literature (whether implicitly or explicitly stated) is that they were the property of wealthy Iron Age women and were buried with their owners alongside other worldly goods as a provision for the afterlife (e.g. Fox 1958: 84). Mirrors are assumed to

Summary of Existing Work Following Palk’s review of the available literature on bridle-bits (Palk 1984: 1) the literature available on mirrors can be fitted into three categories: 1. Articles recording individual finds. These can relate to single finds (e.g. Fox 1960), a mirror found

3

Figure 1.3 Distribution of mirror finds

2. 3.

interpretative sections in much of the subsequent literature. Recent reports on individual finds give detailed contextual studies of the find site and associated objects. However, a general format for these reports can be mapped out. In addition to sections on the find spot, artefact reports and information on any human or faunal remains, discussion tends to reflect the concerns expressed by previous protagonists on mirrors. Individual reports are often included describing decoration (e.g. Jope in Fulford & Creighton 1998) and the techniques employed to inscribe it (e.g. Lowery et al. in Farley 1983). With the exception of Fox & Pollard (1973), Fitzpatrick (1997a) and Johns (2006) these reports include no detailed discussion of mirrors as a group. Even when there is discussion it concentrates on dating and ‘origins’.

in a grave (e.g. Farley 1983) or an excavation (e.g. Wheeler 1943). Discussion of one particular aspect of mirrors, such as decoration (e.g. Lowery et al. 1971) or handles (e.g. Fox 1949). Examination of the corpus, or group (e.g. Jope 2000). Many of these works begin by examining one mirror and then broaden the discussion (e.g. Smith 1909).

The literature can also be broadly divided temporally into two: a body of work discussing poorly recorded finds from the 19th and early 20th-centuries (Stillingfleet 1846; JopeRogers 1873; Smith 1909; Dunning 1928; Fox & Hull 1948; Fox 1949; 1958; 1960; Spratling 1970a; Cunliffe 1988; Stead 1979; Guillaumet & Schönfelder 2001); and more detailed reports outlining individual finds made over the past 35 years (Fox & Pollard 1973; Rook et al. 1982; Farley 1983; Dent 1985a; Fitzpatrick 1997a; Fulford & Creighton 1998; Parfitt 1998; Hill 2002b; Johns 2006; Burleigh & Megaw 2007). The quality of the information available to these different discussants led them to ask different questions. The earlier more general discussions (especially by Fox) were dominated by three interrelated subjects: origins, dating and decoration. These discussions set the future trajectories of mirror studies and put in place a series of implicit assumptions underlying many of the

4

Dating Until recently it was only possible to date four mirrors using associated artefacts, such as brooches or pottery: Birdlip, Colchester I (Lexden Grange), Nijmegen and St. Keverne (Spratling 1970a). All others were dated stylistically by comparison to these four (see Fox & Hull 1948; Fox 1949; 1958) through analysis of the decoration on mirror backs (e.g. Fox 1958: 102; Jope 2000: 137-48), or by handle type (Fox 1949; Spratling 1970a; Jope 2000: 137-8). Many early discussions therefore focussed on stylistic dating of the mirrors from less secure contexts or

Figure 1.4 Mirrors with deposition contexts Roman, Etruscan and Greek mirrors. These discussions are often based on speculation rather than hard facts, consequently opinions vary widely. To date only mirrors of Roman type have been found in Iron Age contexts in Britain (see Lloyd-Morgan 1977a; Lloyd-Morgan & Reedie 1984; Lloyd-Morgan 1986; Stead & Rigby 1989; Niblett 1999). No mirrors of Etruscan or Greek design with secure context have been found in the British Isles. Smith (1909: 346) was the first to suggest that British mirrors are likely to have been influenced by mirrors from other cultures. He suggested that the main influence was Etruscan mirrors, which are made from a bronze disc, decorated with human and other figures, with a handle made from metal, bone or ivory (Smith 1909: 346; Carpino 2003). Smith suggested that Etruscan mirrors could have found their way into southern Britain through contact with Greek and Carthaginian traders. Despite the fact that other Etruscan material culture has been found in European Iron Age graves no mirrors have yet been

without provenance. Since the 1970s eleven more potentially datable mirrors have been recovered; four using radiocarbon dating, a number of others can be dated from associated objects (see Chapter 8). Although there have been some general discussions of deposition date (Fox & Pollard 1973; Fitzpatrick 1997a; Johns 2006), and comments on the deposition date of individual finds in excavation reports (e.g. Stead in Parfitt 1998), there has been no systematic dating audit for the corpus as a whole, taking account of all of the datable mirrors, although Sealey (2006) has recently published a dating audit of mirrors from southern England. “Origins” The possible origins of British mirrors have been discussed at length in the literature (Smith 1909: 346; Dunning 1928: 71-5; Fox 1958: 94; Spratling 1970a: 15; Fitzpatrick 1997a: 65-7; Jope 2000). Although all commentators agree that ‘British’ mirrors are the work of communities in the British Isles, their design and form have been seen to reference

5

discovered (Echt 1999: 207). Dunning (1928), Fitzpatrick (1997a) and Echt (1999) have suggested a Greek origin for British-type mirrors, based on evidence derived from three mirrors found in the Rhine/Marne area of continental Europe: Reinheim, Saarland, Germany (Keller 1965; Echt 1999), Hochheim, Hessen, Germany (Schwabe 1933; Wurm 1972) and ‘La Motte Saint-Valentin’, Courcelles-enMontagne, France (Déchelette 1913). Dunning (1928) saw the discovery of the mirror from Courcelles-en-Montagne as especially significant, suggesting that it could have been a Greek import of the 5th-century BC, which may have reached northern France via the Greek colony of Massilia (Marseilles). It has now been demonstrated that the Hochheim, Reinheim and Courcelles-en-Montagne mirrors were all produced in the regions they were deposited in (Wurm 1972: 114; Echt 1999). However, Fitzpatrick (1997a) has suggested that the design of these European mirrors could have been inspired by Greek-type mirrors and draws on the physical resemblance of the three European mirrors to Greek mirrors to suggest the transmission of the idea of a mirror to East Riding from Greek mirrors. Although the continental mirrors are made from bronze they have circular plates. The mirror from Courcelles-en-Montagne also has a bar handle which is very similar to the iron mirrors found in East Riding (see Wurm 1972: Figs. 3 & 4). All three European mirrors were found in graves, dating to the 4th – 3rd centuries BC and are roughly contemporary with the East Riding mirrors (Schwabe 1933; Wurm 1972; Echt 1999).

is one very short episode of mirror deposition in Britain, perhaps lasting as little as 200 years. If we examine the evidence available to us it is possible to dismiss the first hypothesis. The problem with suggesting that the design of some of the later south-eastern mirrors was inspired by Roman-type mirrors is that this model implies the idea of making mirrors occurred twice in Iron Age Britain: once in East Yorkshire in the 4th – 3rd centuries BC and a second time in southeast England in the 1st-century BC (Fitzpatrick 1997a: 65-7). Fitzpatrick (ibid) has recently argued against Roman origins for the mirror series. He cites three major flaws in the argument. In addition to the problem that Roman inspiration implies that mirror manufacture was invented twice, he also cites chronological problems with this hypothesis. Whilst British mirrors from southeast England look to date to the middle of the 1st- century BC, the mirrors that LloydMorgan cites as being a major influence are contemporary, or later in date. Roman mirrors of the type put forward by Lloyd-Morgan only became popular in the Augustan period. The third major problem stems from the comparative rarity of finds of Roman-type mirrors in Iron Age Britain (see Lloyd-Morgan 1977a; 1986; LloydMorgan & Reedie 1984; Stead & Rigby 1989; Niblett 1999). I would also like to add a fourth criticism as contemporary Roman mirrors found in Britain are decorated differently and made from bronze with a higher tin content. This means that their appearance was markedly different to British mirrors.

The suggestion that some of the later mirrors from southern England were influenced by Roman mirrors has been put forward by Lloyd-Morgan (1991: 32). Fox (1958) also suggested that the design of the later decorated bronze mirrors was influenced by Imperial Rome. According to Fox, mirrors tend to be found in or close to the major ports of the southeast and southwest of Britain. He goes on to say that “it surely was that the women in the higher ranks of British society who were sophisticated enough to commission these costly things from the court craftsmen were those living in areas most open to the influence of Roman manners and customs” (ibid: 94).

The other two hypotheses suggested seem more likely explanations. The idea of a reflective metal plate and handle was transmitted from Greece via Europe to East Yorkshire. To establish whether the mirror form was in circulation for over 500 years, or more like 200 years, a thorough dating audit of the corpus is required (see Chapter 8). Decoration Decorated mirrors have been appropriated as ‘Celtic Art’ objects, consequently many discussions of mirrors have been dominated by the inscribed decoration on mirror backs, most notably comments by Fox (1958: 84-98) and Jope (2000: 137-48). The discussions of Fox and Jope are very descriptive and examine the decoration on the backs of individual mirrors in detail. Both are preoccupied with stylistic dating and Fox especially, creates a stylistic sequence of the ‘development’ of mirror design. The result is that mirrors have been separated from the archaeological context and studied as art objects. Decoration is separated from the object and is viewed as an abstract phenomenon devoid of context. A number of assumptions also underlie many of these discussions concerning the context in which mirrors were made: they were produced by ‘craftsmen’ working under the patronage of high-status individuals.

So how can the question of ‘origins’ be assessed? Summarising the literature and taking account of the distribution of mirrors it is possible to put forward three hypotheses for their ‘origin’. 1. Two separate ‘inspirations’ for mirror production – reflected in the different distributions of East Riding and Southern England. East Riding mirrors ‘inspired’ by European mirrors which in turn were influenced by Greek mirrors. The design of some mirrors from Southern England was influenced by Roman mirrors. 2. One inspiration. The idea of a mirror began in East Riding and came from European mirrors. The mirror form was known for around 500 years in Britain, it is just that we only see them where they appear in the archaeological record. 3. One inspiration. Dates more conflated than present evidence suggests. This model suggests that there

Summary A brief summary of mirrors and the existing academic literature has uncovered a number of potential avenues

6

and temporal trends in form (Chapter 6 and Appendix D), size (Appendix G) and depositional context (Chapter 9). Decoration is also re-examined using an archaeological rather than art historic methodology (Chapter 4). In doing this the aim is to re-incorporate the study of decoration into the archaeological arena and illustrate that decoration cannot be fully understood if it is abstracted from the archaeological context and the artefact it ornaments.

for research and it is clear that the full possibilities of the available data have not been fully realised. The need for detailed technological and contextual analyses of Iron Age material culture has recently been highlighted (Haselgrove et al. 2002). Most previous work (except for recent studies on individual finds e.g. Fitzpatrick 1997a; Parfitt 1998) has largely ignored the social and cultural context and treated mirrors as ‘Celtic Art’ objects, concentrating on stylistic comparisons, analysis of their ‘artistic qualities’ and ignoring their use and significance to the people who made and used them. In the light of a large number of recent finds there is a clear need for detailed technological and contextual analysis of the corpus. The dominant interpretation that mirrors were the property of high-status Iron Age women and were buried with their owners as a symbol of their status and as a provision for the afterlife should also be re-examined based on the new contextual evidence available.

This information is utilised to construct biographies for mirrors. The approach to object biographies is set out in Chapter 2. By using the biographical approach to material culture, it aims to uncover the layers of meaning built up during the life history of individual mirrors. It is argued that if we are to uncover the full complexities of why they were made, how they were used and why they were deposited in particular ways, we need to understand their biographies. Objects are active agents acting to produce and reformulate social relationships (Strathern 1988; Gell 1998). In order to construct these biographies the contexts in which mirrors were made (Chapter 3) and deposited (Chapter 9) are examined, referring to wider patterns occurring in the Iron Age.

1.3 RESEARCH AIMS This book re-examines the corpus of Iron Age mirrors. Building upon existing work and incorporating contextual data from recent excavation reports, It compiles a traditional study of a group of objects comprising: a full catalogue (Appendix A); examination of chronological (Chapter 8) and geographical distribution (see above); current physical condition (Chapter 5); and analysis of regional

To achieve these aims the theoretical methodology needs to be developed. This is outlined in the next chapter.

7

CHAPTER 2 THE BIOGRAPHICAL APPROACH Woodward 2002). For example, a number of studies look at how the perception of and meaning of objects can be transformed when they are exchanged between cultures (Gallardo et al. 1999; Peers 1999; Saunders 1999; Seip 1999; Eckardt & Williams 2003). Whitely (2002) illustrated how objects are valued because of their biographies; the histories they acquire through a lifetime confer meanings to other objects and the individual interred with them in a burial. Lamberg-Karlovsky (1993) examined the ‘academic biography’ of vessels from the Near East in the 3rd- millennium BC, outlining the different ways they have been interpreted by generations of archaeologists. Rainbird (1999) even suggested that in the Pacific island of Pohnpei the biography of one set of objects (ceramics) was transferred to another (tombs) through a process of ‘biographical entanglement’. Similarly, Joy (2002) illustrated how an object biography can be transferred to a replica.

2.1 OBJECT BIOGRAPHY The idea that like humans objects can be seen to have life histories was put forward by the anthropologist Igor Kopytoff (1986: 66-8). Kopytoff’s proposal derives from economic anthropology and from theorising the nature of the gift and the commodity (sees Gregory 1982; Appadurai 1986a; Strathern 1988; Mauss 1990; Thomas 1991). There are two main elements to Kopytoff’s biographical approach. The first is the idea that like humans objects can be seen to have life histories. The second is that it is possible to analyse these life histories by following the lives of objects from birth through life to death. An object is produced, or ‘born’; it is involved in a particular set of social relationships during its lifetime; it also ‘dies’ when it is no longer involved in these relationships (Holtorf 1998). Kopytoff established the usefulness of thinking about objects biographically; his biographical approach provides a methodology by which the changing meanings of objects through time and the context in which these actions take place can be examined from a social perspective. He also takes account of the fact that these meanings can ‘change’ and be ‘renegotiated’ throughout an object’s life history (Gosden & Marshall 1999: 169), underlining the fact that an object cannot be fully understood by examining it at a single point in its existence (Kopytoff 1986).

All of these studies share common themes. The majority, like Kopytoff, see exchange and the passing of time as the main mechanisms of biographical change. The exchange and circulation of objects is the primary means by which objects acquire meaning and value; over time these meanings and values accumulate as the object continues to circulate and builds a history. Some studies have also extended the notion of biography to static objects, by constructing biographies of megaliths and monuments (Holtorf 1998; Last 1998; Gillings & Pollard 1999). Through circulation and social usage, objects also act to produce, cement and redefine social relationships (Tilley 1996: 247). The importance of performance in re-emphasising, changing and renegotiating the meanings attached to objects has also been highlighted (Gosden & Marshall 1999).

2.2 ARCHAEOLOGY AND THE BIOGRAPHICAL APPROACH The biographical approach is becoming increasingly popular in archaeology. An issue of World Archaeology was dedicated to biography (Marshall & Gosden 1999) and well known archaeologists have referred to and used the approach in key works (Bradley 1990; Thomas 1996; Tilley 1996). During the last ten years papers on biography have regularly featured at archaeological conferences. The acceptance of the biographical approach into the mainstream archaeological arena is underlined by its prominent inclusion in a general introduction to the discipline (see Gamble 2001).

Use-life Analysis A term which has become synonymous with archaeological biographies is ‘use-life’ and the two concepts have become conflated (York 2002: 79). Although analysis of use-life and damage of objects can be useful tools in constructing the biography of an object, there is a distinction between use-life and biography. Analysis of wear or damage can inform us about certain aspects of an object biography such as the duration, type, or intensity of usage, but it does not inform us of all. As Jones (2002a: 84) stated “the concept of biography, while embracing the insights that a use-life perspective brings to artefact analysis, also encompasses the idea that objects are used to construct and maintain social identities”. Objects can be active within social relationships; how they are made and used is intimately related to the social relationships they act to formulate and maintain. Use-wear analysis can be important in determining the life history of objects, but the technological and functional properties of objects cannot be divorced from their cultural and social significance.

This popularity is in part due to a number of very influential anthropological studies which have shown how in some societies objects can take on the personalities of people, or be seen to have lives that are similar to people (e.g. Munn 1986; Strathern 1988; Battaglia 1990; Mauss 1990; MacKenzie 1991; Hendrickson 1995; Gell 1998; Hoskins 1998). In the light of these studies, archaeologists have accepted that objects are not only a product of society they are integral to it (Thomas 1996: 141). If objects and persons can be seen to share lives that are similar to each other, it should be possible to write the biographies of objects, just as we write the biographies of people (Kopytoff 1986; Thomas 1996; Tilley 1996; Bradley 1998; Jones 2002a). Much archaeological work utilising the biographical approach has centred on Kopytoff’s observation that there are many different ways in which an object can be conceived of as having a biography (e.g. Hamilakis 1999). Or by emphasising that object biographies can uncover new meanings because the meaning of an object can change throughout the course of its life (e.g. MacGregor 1999; Immonen 2002;

The term use-life also acts to reduce the importance of the production process in the life history of an object because from a use-life perspective the object is made to perform

8

a certain function and it will be used until it is no longer useful in performing that function (Bradley 1998: xx). In an additional chapter of the re-printed edition of ‘The Passage of Arms’, Bradley (ibid: xv-xxxii) highlights the importance of the production of objects and how meanings and social significance can be firmly established at this stage of a life history, in effect setting the trajectories of its future life. He refers to recent studies in use-wear and artefact damage (e.g. Taylor 1993; Bridgeford 1997) which illustrate that objects of the same type found in different depositional contexts are likely to have very different use and damage histories. For example, Bridgeford (1997) found that swords from watery contexts were less worn and more elaborately made than those found in graves, suggesting that from their production they were destined not to be used in combat but to have a particular social life with another end in mind. Fontijn (2002: 28) also makes this point and highlights the importance of the smith in determining the ‘biographical possibilities’ (Kopytoff 1986: 66) of metal objects. The decisions of the person or persons who make an object are crucial to its future life (Fontijn 2002).

made to comment upon these ‘rich’ or ‘complex’ biographies. This phenomenon is more pronounced in conference papers where the biographical approach is often inserted as ‘token theory’. Time and again in conference papers we are told that an object would have had a ‘rich’ or ‘complex’ biography, yet no attempt is made to comment on it. It is acceptable merely to refer to the biographical possibilities of an object, illustrating that the presenter is theoretically aware. Despite the fact that an increasing number of studies purport to use the approach, where it has been used to examine prehistoric objects the usefulness of the approach has yet to be clearly demonstrated. Biography doesn’t have to be Chronological It is possible to say more about a prehistoric object than it probably had a ‘rich and complex biography’. As we have seen, prehistorians have readily accepted the idea that objects can have biographies, but how does this improve our knowledge of the past? Kopytoff suggests that objects can be seen to have lives. He also provides a methodology through which we are able to approach those biographies, by following the life of an object from birth through life to death. However, is this methodology necessarily suitable for archaeological data?

An object is invested with meaning at the time of production. It is also invested with agency; an object can be made in such a way that it is able to act within certain social relationships (Strathern 1988: 176). For example, the prows of Kula canoes are decorated so that they dazzle exchange partners into parting with gifts they may not have wished to part with (Gell 1999; Campbell 2002). At the time of production an object is equipped for its future life; that is, the life envisaged for the object at the time of production: the particular social sphere in which the object will operate and the social relationships in which it will act.

The biographical approach is readily conceivable to us because it constructs a linear narrative for the life of an object: a neat life history with a beginning, middle and end. However, biography need not be chronological. We do not have to chart the birth, life and death of an object to present its biography. The biography of a person or an object is not necessarily formed of a consecutive series of events; this is just the way that we choose to present our own biographies in western culture. This is linked to how we perceive notions of self and personhood, as well as time.

Summary A brief overview of archaeological object biographies has shown that the biographical approach has become an accepted part of archaeological discourse. The potential role of the context of production in setting future life trajectories has also been recognised.

Object Vs Human Lives The metaphor of the human lifecycle is reinforced upon us because of the way we view the archaeological evidence. In the archaeological record we have evidence for the beginning and end of the object’s life, but we know little about the life of an object, the bit in between. To construct what is happening in the middle we have to make inferences from what we already know about an object’s birth and death and from the material condition of the object. Conceptually, constructing the biography of an object in this way is problematic because it only allows for one potential life-path for the object under analysis – the path it actually followed (Figure 2.1). In reality as the object is living out its life there is no single direction its life-path can take. At any point in its life history it can take on new meanings and associations, which have the potential profoundly to alter the directions its future life will take. In addition, the lives of some objects are further complicated because they can extend over a series of human lifetimes. It is necessary to be more careful about how we employ the metaphor of the human lifecycle to objects and its use should be justified in terms of the evidence available to us rather than assumed.

2.3 PROBLEMS WITH EXISTING ARCHAEOLOGICAL APPLICATIONS OF THE BIOGRAPHICAL APPROACH The way the biographical approach has been applied to archaeological data can be criticised. Very often the term is used in passing, in the introduction a mention may be made to key anthropological texts and in the conclusion it may be suggested that the object/s under scrutiny must have had ‘rich’ or ‘complex’ biographies (Whitley 2002). With the exception of some biographies written about Neoloithic monuments (Holtorf 1998; Gillings & Pollard 1999) or by some Dutch archaeologists (Gerritsen 1999; Fontijn 2002), most archaeological studies that have attempted to follow the lives of objects from birth through to death have been anthropologically or historically situated (e.g. Spector 1993; Peers 1999; Seip 1999; Joy 2002). Studies specifically examining prehistoric material culture tend to use Kopytoff’s notion of biography metaphorically to present specific arguments about the available data (e.g. Skeates 1995; Immonen 2002). As a result very little attempt is

9

Figure 2.1 Re-Constructing an object life Concepts of Time The concept of time is something which has hitherto been taken for granted by archaeologists (Bradley 2002: 5). Time has been viewed as a fundamental, abstract and unproblematic phenomenon (see Karlsson 2001a: ix); something that is so natural it does not need to be discussed (Karlsson 2001b: 47). Time has been viewed as the archaeologist’s medium (Bradley 2002: 5). Archaeologists have sought to compartmentalise time and count back from the present day to establish when things happened in the past. In the words of Karlsson (2001a: ix-x), time from the perspective of an archaeologist is “… a linear sequence, with clearly divided horizons, in which we as archaeologists can place both the various happenings and the material culture of the past”. Time in this sense is seen as ‘objective’ and ‘quantifiable’ (McGlade 1999: 142). All of us live our lives within this constant flow of time (Karlsson 2001b: 47), so then why didn’t people in the past? Our own understanding of the concept of time has been carried over into our work as archaeologists. The biographical approach is readily conceivable to us because it constructs a linear narrative for the life of an object that we are able to understand; a neat life history with a beginning, middle and end. Our linear concept of time and its stress on a sequence of events occurring one after another is partly a consequence of medieval traditions of oratory, where a long chain of inter-linked events are memorised and then narrated (Rowlands 1993). Stories in a European tradition have a narrative structure to them; they have to be coherent, there must be a logical sequence of events and above all there must be a beginning, middle and end (Damm 2001: 32; Lucas 2005). This idea continues over into our division of the past into the three ages of stone, bronze and iron (Lucas 2005). The stories we construct for past societies must conform to this structure so that we can understand them. The materiality and durability of an object can act as a physical link with the past, present and future (Rowlands 1993: 144) as objects are invested with meanings throughout their lifetimes and accumulate a biography. The links between biography, memory and narrative are clear. Just as we create our own stories, or written narratives; through the physical materiality of the object we create its past, present and a future: we write an object’s biography (Holtorf 20001; 2002a; 2002b; Hegardt 2004). Through the performance of archaeology, we add layers of meaning to an object from the moment it is found, transforming a non-thing into a thing (see Heidegger 1971). We create a life-story for an object, which satisfies our need for a coherent narrative, giving a beginning, middle and end.

10

Our measurement of time on a single scale may be unique, linked to capitalism and the financial value of time. Anthropologists have known for some time that other societies perceive time differently (Gell 1992; Munn 1993; Damm 2001: 30; Lucas 2005: 61-7). Contrary to our linear concept of time, some societies may conceive of time as being more cyclical (Cremo 1999: 38; Artelius 2001: 9), it can be related to natural phenomena, such as the cycles of the moon or the passing of the seasons (Parker-Pearson 1996;1999b; Bradley 2002: 6). Time may also be experienced differently within the same culture, depending on context (Bailey 1993; Artelius 2001: 9; Damm 2001: 30). As Bradley (2002: 6) puts it “the possibilities are limitless”. Time is not an unproblematic, abstract phenomenon; time is a social product, it is a human creation (Gosden 1994: 1), conceptualised and experienced in many different ways between different societies and even within the same society. In the words of Gosden (ibid: 34) “people create time and space through their actions. Time and space, in turn, become part of the structure of habitual action, shaping the nature of reference between actions”. Things from the past help shape identity in and are part of the present. Archaeology is the study of these mute, ‘vanished temporalities’ of the past which constitute a part of what makes up the present (Olivier 1999; 2001: 68-9). By dividing the past up into bite-size digestible chunks that occur in a comprehensible succession, one after the other, archaeologists construct a past which is comprehensible to us. But by restricting ourselves to this linear conception of time we are missing out on the multi-temporal nature of the present and the past (Lucas 2005: 38). Moreland (1999) has demonstrated that the biography of an object need not follow the linear sequence of birth through ageing to death, suggested by Kopytoff (1986) and reinforced by Holtorf (1998). By following the life history of a stone cross, he showed that at each major juncture in its lifetime the meaning and role of the cross was influenced by attitudes and perceptions of the past (Moreland 1999: 198). As these attitudes towards the past changed the cross consequently suffered or prospered and underwent a series of ‘reincarnations’. Similarly when Early Medieval peoples removed Roman objects from the ground and placed them in their own graves (see Eckardt & Williams 2003), they created a new biography for these objects, shaped by social relationships during the Early Medieval period and Early Medieval attitudes to the past. A similar case could be made for the Bronze Age objects included in the Iron Age Netherhampton hoard (see Stead 1998; Hingley 2009). We need to recognise the fact that when we examine prehistoric material culture ‘biographical’ time may not have been ‘biographical’ to prehistoric peoples. Summary The use of the biographical approach in archaeology has largely been restricted by the analogy of the western human lifecycle and this has been reinforced upon us through the way in which we encounter evidence in the archaeological record. This narrative structure seems to make sense to us because it fits in with and satisfies our

own western sense of storytelling. Telling the life history of an object also presents it within our own understanding of linear time, which dominates the discipline of archaeology. We need to broaden our concept of object biography beyond the perspective of the human lifecycle. This has important implications because the use of the term biography has already entered archaeological discourse. 2.4 ANTHROPOLOGY AND BIOGRAPHY According to Thomas (1996: 169) there is a tendency amongst archaeologists to forget that artefacts exist within a particular social world and are part of a complex set of relationships between persons and things. The problem arises because objects, in the context that archaeologists encounter them through their material properties, are still with us whereas the people who made and used them are not. Consequently, rather than being viewed as a ‘concrete manifestation’ of a series of relationships between people, objects, places, and personal histories, archaeological artefacts are often seen as merely a cultural by-product of these relationships. A number of anthropological works rely heavily on the biographical approach and social relationships between people and objects are integral to the biographies they construct (MacKenzie 1991; Hendrickson 1995; Hoskins 1998). MacKenzie (1991) examined the social roles of string bags in creating gendered identities in central New Guinean society. In her study of Guatemalan textiles, Hendrickson (1995) included two biographical chapters on traditional Mayan woven textiles or Traje. The first, examined how the lives of Traje and people are entwined and how identity and social relationships are expressed through Traje at different stages in a person’s life. The second chapter worked within the parameters set by Kopytoff and followed the social lives of Traje from the moment they were woven or born, to their death. Hoskins (1998) identified a group of objects which she referred to as ‘biographical’ which are used as a medium through which people are able to tell their own life story. The idea of telling a life story directly to another person did not exist in Kodi society, at the western end of the Eastern Indonesian island of Sumba. Instead Hoskins found that particular objects often acted as metaphors for the self and that it was not possible to separate the life histories of persons from the life histories of those objects. Some studies have shown how objects made from or closely associated with, human body parts are able to act in this way (e.g. Hoskins 1985; Geary 1986; Hoskins 1989). However, Hoskins extended this agency to other types of object. The biographies that Hoskins uncovered were quite different from western ones. The stories woven around objects often concentrated on particular events or particular social roles performed by people in their lifetimes. Identities of people and objects merged and objects and people were seen to share similar social lives. The way that people defined themselves and their own lives often centred on these events. Stories were also linked to the lives of ancestors and social relationships with other people. People in Kodi society expressed their lives, social roles and relationships through the metaphorical properties of objects. Objects

11

‘become a pivot for reflexivity and introspection, a tool of autobiographic self-discovery, a way of knowing oneself through things’ (Hoskins 1998: 198). As these anthropological studies illustrate, specific objects, or groups of objects, can become ‘biographical’; the lives of people and objects can become so entangled that they are difficult to separate (Hendrickson 1995: Chapter 4; Hoskins 1998: 2). Identifying these objects archaeologically would be problematic. The biographical objects identified by Hendrickson (1995) and Hoskins (1998) are often organic and would not usually survive in archaeological contexts. We can suggest that some objects are more likely to become significant because they are made of durable materials (see Lillios 1999: 242) or if special care and attention was put into their manufacture, or because of their close association with the human body. Objects associated with the human body, such as clothing, are more likely to be viewed biographically because they act to mediate between the physical body and the social world (Hendrickson 1995: 98). The critical factor in identifying ‘biographical’ objects is that they are often (but not always) constructed in such a way that social relations can cluster around them. 2.5 BIOGRAPHY AS AN UNDERSTANDING OF SOCIAL RELATIONSHIPS The biographies of objects constructed by anthropologists and archaeologists look very different. According to Gell (1998: 11) “anthropologists typically view relationships in a ‘biographical’ context… relationships are seen as part of a biographical series entered into at different phases of the lifecycle”. Obviously the constraints of archaeological data limit what we can uncover; archaeologists cannot go back in time. However, part of the reason why archaeologists have failed to specifically examine biographies from a relational perspective is the manner in which biography has been viewed by archaeologists. The ‘ideal’ birth, life and death cannot be recreated because of the limits of archaeological evidence. It therefore seems satisfactory to state that objects must have had ‘rich and complex biographies’ and not specifically comment upon the form of these biographies. There are other ways of defining biography. According to Gell (ibid: 222) “…any one social individual is the sum of their relations (distributed over biographical time and space) with other persons”. This notion underlines much of the anthropological work summarised above. In Melanesian society the body acts as a register of the social relationships that constitute a person: it is a ‘microcosm of relations’ (Strathern 1988: 131). A person’s biography can be seen as the sum of the social relationships that constitute that person. These relationships can be presented in a number of different ways, one of which is to chart events and relationships from birth through life to death. But critically, a ‘biographical career’ may extend beyond biological death: “… a spread of biographical events and memories of events, and a dispersed category of material objects, traces, and leavings, which can be attributed to a person and which, in aggregate, testify to

agency and patienthood during a biographical career which may, indeed, prolong itself long after biological death” (Gell 1998: 222). Gell (ibid: 17- 18) stated that “social agency can be exercised relative to ‘things’ and social agency can be exercised by ‘things’”: objects can act as social agents in particular social situations and they distribute their agency throughout the social milieu.” The agency Gell is referring to is relational (ibid: Table 1). Just like people, objects can act as ‘agents’ in a relationship or be acted upon as ‘patients’. Critically in the context of this study, if objects are ascribed relational agency, like people, the biography of an object can be seen to comprise the sum of the social relationships that constitute the object. The biographical approach has greater potential to inform us about material culture, if we utilise it from this relational perspective (see also Gamble 2007: Chapter 5; Knappett 2005: 62-63). The life of an object does not necessarily follow a linear pattern from birth through life to death. As we have seen from some of the examples above, after its creation an object can die a number of times as it becomes a part of and leaves different spheres of relationships. It can also have a number of different simultaneous lives which can run concurrently as it acts in different relationship webs. In addition, the biography of some objects is further complicated because they extend over a series of human lifetimes. This means that at each juncture of the life of an object the meaning and role of that object can be influenced by attitudes and perceptions of the past. How an object is interpreted can be dependent then upon how its past is perceived and an object can undergo a series of transformations of meaning, or jumps. Instead of constraining ourselves to the metaphor of a western human lifecycle with a beginning, middle and end, we can also envisage object biographies as being non-linear and consisting of a series of connected jumps as the object becomes alive within certain clusters of social relationships and is inactive at other points in time and space. Conceiving an object biography in this way has the advantage of allowing us to pick up on the biography of an object at specific points and particular contexts where the archaeological evidence will allow, and we need not feel that the biography is lacking because we are unable to construct a neat linear life story for it. 2.6 A BIOGRAPHY OF MIRRORS Context is of key importance to the biography of mirrors. The biography of each mirror is different. Through the act of constructing the biography of a mirror – the social relations it is involved in through time – it is possible to approach how its identity has changed, but an identity is also enforced on that object (see Holtorf 2002b; Lucas 2005). This identity is implied through the maintenance of the physical form of the object. By studying mirrors as a group, it is inferred that there is something intrinsic to their material form which makes them like other objects that look similar. This viewpoint is problematic and is inherent to the concept of the corpus; by grouping these objects together as ‘mirrors’ our own perceptions are imposed on these objects. Should objects be grouped together and studied as a coherent whole because they look similar viewed through a

12

western cultural lens? Here we hit a paradox: in order to understand an object like a mirror, we need to compare it to other ‘similar’ objects but we decide which objects are ‘similar’. Throughout this book the view is taken that the material characteristics of an object, its shape and form, can determine the relations structured around it (Strathern 1988: 176; Knappet 2005). The mirror was constituted by these relationships but at the same time the relationships were legitimated and preserved by the material form of the mirror (Battaglia 1990; Hoskins 1998). By examining how objects of the same shape and form that we call ‘mirrors’ were made, used and deposited, it is hoped that greater insight into their lives will be gained. The biographies of individual objects are unique (see Holtorf 2002b) and biographical analysis is confined to the scale of the individual object. However, it is necessary to study a corpus of objects to set individual objects in a broader context. Events in the Life of a Mirror The biography of an object is the sum of the social relationships that constitute it. Throughout, this book, examines the artefact in process by identifying some of the social relations that constituted mirrors and identifies particular points in their lives that can be examined given the available archaeological evidence. Returning to the ideas of Kopytoff, he suggested a range of ‘biographical possibilities’ for a thing. Objects, like humans, can be conceived as having lived a ‘well-lived life’ (Kopytoff 1986: 66). There are also important and meaningful events within the life of an object (ibid: 67). Like people, objects operate in a particular sphere and in general they live their lives within that sphere. However, sometimes an object is taken out of that sphere ‘to be either sacrilised or isolated or cast out’ (ibid 1986: 89). A mirror acted within specific social spheres set out at the time of production. An idealised or expected life for a mirror would have been mapped out at this time. The majority of mirrors lived their lives within the social spheres they were made to operate within. However, some may have been taken out of these spheres and as a result their lives deviated from the ‘expected’ path. Three main types of event in the life of a mirror can be suggested: 1. Through everyday usage social relationships were acted out and reinforced. 2. Special events that occur within a given social sphere. 3. When an object is taken out of its social sphere. We can detect these different events in the biography of a mirror by employing different methods to interpret the archaeological data: 1. We can examine everyday usage through analysis of use-wear, damage and polishing. We can also postulate the kinds of relationships mirrors are likely to be active in creating, maintaining

2.

3.

and renegotiating by examining mirror form and decoration, which as we have seen could have determined the type of relationships that were likely to be structured around a mirror. Special events within the life of a mirror are difficult to examine. The only two such events we can examine with certainty are production and deposition. By examining a corpus of material we can detect when an object has deviated from the ‘norm’.

During the biography of an object some moments were more ‘important’ than others. For example, it is probable that when a mirror acted within a series of social relationships deliberately stage-set for a specific purpose these relationships were important in the life of that mirror. The ‘birth’ of an artefact is the moment when meanings are set through the performance of making and an ‘expected’ life trajectory is mapped out (see above). Evidence for the production of metal artefacts has been found on many Iron Age sites. The social death of an artefact is also a particular moment of significance as the object is involved in a social performance, the material remains of which are often visible to us archaeologically. This does not take away the importance of everyday actions and repetitive usage, which act to reinforce meanings. Use-life can be seen to be the result of repetitive actions where meanings are reinforced through habitual usage. There may also be times when an object acts outside of its particular social sphere. We can detect this if a mirror is deposited in an unusual context.

cal attention on the object and ascribes agency to objects through the relationships that constitute it. In a prehistoric context we cannot reconstruct a birth, life and death for an object from the available evidence. We can, however, examine the relationships that constitute the biographies of prehistoric objects. The biographies of mirrors are examined from two perspectives: 1. Biographical account of what is written into mirrors 2. Contextual analysis of deposition context These two analyses form the next two sections of this book. The events and relationships formulated during the process of making mirrors, the meanings and relationships attached to them during their lifetimes, the people they were associated with, their origin (mythical or real), their design; and signs of wear and repair all create an ‘aura’ or patina (Benjamin 1973) (physical signs of age or a patina of abstracted meanings), a life.

These moments in an object life can only be considered within the context of the category of objects being examined and the wider social context in which the objects acted out their social lives. In order to position mirrors within their particular regional and temporal contexts a comprehensive dating audit of the corpus will also be conducted. Object biography is therefore necessarily object and context specific.

The first section takes the form of a biographical account of what is written into mirrors. From their manufacture mirrors were afforded with the potential to act in particular social relationships (Knappet 2005). Clues about the ways in which mirrors were used, or their performance in social relations are also written into the artefacts through signs of use, wear and repair. In Chapter 3 the relationships formulated when mirrors were manufactured are investigated. In Chapter 4 mirror decoration is examined, in particular the relationships manifested when a mirror was decorated. In Chapter 5 the evidence for the social usage of mirrors is scrutinised through an examination of use, wear and repair. Finally, in Chapter 6 mirror form and the potential ‘built’ into them to act in multiple relationship spheres is examined.

2.7 BIOGRAPHICAL MIRRORS The argument put forward in this book is that when we think about objects biographically it is not necessary to constrain our thinking to the metaphor of the human lifecycle, with a birth and a death and a linear sequence of events, or a life, in between. By doing so we are limiting the information we can gather when constructing object biographies. The biography of an object is the sum of the social relationships that constitute it. Biography does not have to be chronological in its understanding of the social relationships that mutually constitute person and object. When we think of the biography of a mirror, it is not possible to separate it from social relationships. Thinking about objects biographically enables us to see the relationships between people, objects, places and personal histories that constitute the object. The biographical approach focuses archaeologi-

The second section of this book is more conventional and examines deposition context. As outlined above, meanings for objects are not fixed and they can change and are actively renegotiated throughout a lifetime. Although mirrors may have been made to act within specific social relations, there is potential that meanings attached at the time of manufacture were manipulated and altered during the object life. A dating audit is first conducted (Chapter 8) to set mirrors within their wider social and cultural context. Individual deposition contexts are then detailed (Chapter 9). The results of this analysis are summarised in Chapter 10. This summary takes account of the biographical information written into mirrors and examines how the relationships formulated at the time of mirror manufacture and throughout their lifetimes were referred to and manipulated when mirrors were deposited.

13

CHAPTER 3 MAKING MIRRORS 3.1 INTRODUCTION Metal artefacts are important items of social expression (Childs & Killick 1993: 331) and are commonly used to mark major changes in a person’s life-cycle (Sofaer Derevenski 2000). Essential to the formulation of meanings and associations attached to objects are the social relationships formulated in the production process (Hoffman & Dobres 1999: 211). The making of any artefact, is critical in setting the trajectories of its future life (Bradley 1998: xx). In this chapter the social and symbolic significance of the production of mirrors is examined. Building upon existing technological studies of production, the potential social relationships formed during the act of mirror making are identified. The overall aim is to chart production stage-bystage, questioning what these techniques and processes can tell us about the people who made and used mirrors. 3.2 APPROACHES TO METALWORKING Discussion of the production of Iron Age metalwork has been limited as studies have concentrated on finished artefacts (particularly decoration) and depositional context. A number of studies have examined the social and symbolic significance of metalwork deposition (e.g. Fitzpatrick 1984; Hunter 1997; Hingley 1990; 2005; 2006). Hingley (1997), strongly influenced by the association in African ironworking between iron production, fertility and regeneration (see Herbert 1993), examined the potential social and symbolic significance of the production of iron. Stead (1985), Megaw and Megaw (1990) and Hunter (2001) have examined the construction of individual artefacts in detail. The lack of more discussions on the potential social significance of Iron Age metalwork production is perhaps due to the fact that, although metal artefacts are relatively common finds, traces of metalworking or the metalworkers themselves are comparatively scarce. Some metalwork, identified and studied as ‘Celtic Art’ has also been separated from the archaeological record (Hill 1995a: 53-4), causing a ‘methodological split’ from discussions that occupy more mainstream archaeology (MacDonald 2007a: 333). Three questions dominate the existing literature on the production of Iron Age metalwork: 1. Were metalworkers itinerants or did they operate from workshops? 2. Were they autonomous or working under the control of a patron? 3. What social position did they occupy? Metalworkers are seen to have been organised into craft schools and workshops (Aldhouse-Green 1996: 38). Some researchers have even attempted to identify the work of individual ‘craftsmen’ or ‘workshops’ by singling out similarities between artefacts (e.g. Spratling 1970b: 188; Jope 1971; Cunliffe 1991: 456, 470). A preoccupation with these issues has acted to stifle debate. These questions originate largely from a fixation with ‘Celtic’ society (see Champion 1987: 99) in discussions of Iron Age metalworking and ‘Celtic Art’ (Fitzpatrick 1996: 243), information derived from Classical and Irish texts (e.g. Gillies 1981) and influential discussions of metalworking and metalworkers in the archaeological literature (e.g. Childe 1930: Chapter 1). In light of the extensive literature questioning the use of the

14

terms ‘Celt’, or ‘Celtic’ (e.g. Chapman 1992; Dietler 1994; Collis 1996; 1997; James 1999; Collis 2003) and problems associated with interpreting and contextualising Classical (Fitzpatrick 1996: 244; Webster 1997) and Irish texts (see Hutton 1991), structuring accounts of metalworking on descriptions of ‘Celtic’ society from Classical and Irish literature no longer seems defensible (Dungworth 1997b). Metalworkers are universally ascribed a certain status somewhere between the ‘elite’ and the ‘farmers’ (see James 1993: diagram p. 53). This status is inferred from/ onto metalwork (especially decorated metalwork), which as a category of object has been given a status above other items of material culture by researchers. Hill (2006; 2007) has recently questioned the social pyramid model that has generally been applied to Iron Age society (see also Karl 2008). He argues that many different types of society would have existed at any one time and that there is likely to have been a wide variety in the ways in which society was organised through time and across different regions. Iron Age metalwork production has also been analysed from an industrial, technological perspective (Foster 1995: 49; Hingley 1997: 9). In these studies technology is often separated from society and metalwork is judged on how closely it fits our own industrial ideals of economy of skill and efficiency (Hingley 1997: 9). However, it should not be assumed that Iron Age metalworking should necessarily ‘makes sense’ to us (Dungworth 1997b). Hill (1989: 19-21; 1992: 59-61; 1995b: 5) has argued that the Iron Age was fundamentally different from the present day. What makes sense to us from a modern industrial perspective would have probably been nonsensical to an Iron Age metalworker. Ethnographic examples (e.g. David et al. 1989; Herbert 1993; Rowlands & Warnier 1993) illustrate that it is not possible to separate technology from the society in which it is located. Although there may be technological constraints to metalworking, the parameters of science are sufficiently wide to allow different societies to meet their own cultural goals and needs (Lahiri 1995: 116). Technology cannot be defined as a series of scientific procedures which exist independently and outside of society (Pfaffenberger 1988; 1992a; 1992b; Ingold 1993; Dobres & Hoffman 1994; Dobres 2000); it is embedded within the wider social context (Barber 2003: 133-4) and consists of a pattern of repeatable behaviours (Keller & Keller 1996) which are inherently social (Pfaffenberger 1988: 241). Many of the trends outlined for the analysis of Iron Age metalwork are echoed in work on mirrors. Existing work on their production is technologically based and has examined the techniques of decoration (e.g. Lowery et al. 1971; 1976; Lowery & Savage 1976) and trace element analysis of the composition of bronze mirrors (Northover 1993). Discussions of the context of production have been restricted to questioning the artistic abilities of the people who made them and their manufacture in ‘workshops’ by particular ‘schools’ of ‘artists’ or

‘craftsmen’ (e.g. Fox 1958: 84-105; Jope 2000: 137-46). However, there has been little discussion of what the terms ‘workshop’, ‘craftsmen’, or ‘artist’ actually mean in this context. The people who made mirrors are also described using different and confusing terminology. In a well-known discussion, Fox (1958: 84-105) uses 8 different terms to describe the people who made mirrors: · Bronze-smuth · Metalworker · Artist-craftsman · Craftsman · Celtic artist · Bronze-worker · Mirror-maker · Artist Terms such as ‘artist’ or ‘craftsman’ are interchanged freely even though they have very different implications. The key restraints to moving forward then are some of the underlying assumptions behind discussions of metalwork production which must be challenged and alternative interpretations put forward: 1. There is a need for greater consistency in the ways that metalworkers are described. 2. Metalworkers who made objects that have been identified as ‘Celtic Art’ should not be labelled ‘artists’. 3. Discussions of metal objects traditionally defined and studied as ‘Celtic Art’ should be resituated into mainstream archaeological discourse (MacDonald 2007a: 333). The approach set out in this book is to broaden the discussion of metalwork production beyond questions associated with the status, independence and mobility of metalworkers by examining the potential social relationships involved in the production of metal objects. By breaking down mirror production into a series of steps, a kind of chaîne opératoire of mirror making will be produced. By charting the sequence of technical actions involved in transforming rock into artefact, the aim is to reveal some of the interconnected social relationships which reside between the physical actions of mirror making and the wider cultural framework (see Dobres 1999: 125-6). The next sections of this chapter build upon the existing literature on production and examine some of the potential social relationships accrued in the process of mirror making. Because iron and bronze mirrors were made using different methods, the production of iron and bronze mirrors is examined separately. The significance of mirrors made from both iron and bronze components is also questioned. 3.3 MAKING BRONZE MIRRORS Bronze mirrors are made from a number of component parts (see Figure 1.1): a polished sheet plate, often decorated on one side; a cast handle; and sometimes a plate edging. Mining and Smelting Bronze The first stage in the production of any metal artefact is the extraction and smelting of ores. Mirrors could have been made of metal from specific ore sources from significant

15

locations (see Bradley & Edmonds 1993). Alternatively, artefact production could have been separated from ore extraction and smelting by distance from the source, or through re-cycling. The dominant copper-alloy used throughout the British Iron Age was bronze (Dungworth 1996: 410; 1997a: 49), which is made from copper and tin. Important impurities found in bronze include lead, zinc and arsenic. Iron Age bronze contains very few of these impurities and there is little variation in the composition of bronze artefacts found across Britain (Dungworth 1997b). Copper and tin are found in the highland zones of Britain (Northover 1984b: 126), predominantly in the west and southwest. Extraction of ores in mines probably continued throughout the period even though evidence for mining is limited (Northover 1995: 287; Dungworth 1997a: 48). There is little evidence for the smelting of bronze on Iron Age sites. The presence of ingots of copper, tin and lead, and the relative scarcity and geographical marginality of the ores suggest that like primary ore processing, copper smelting may have taken place close to where the ores were extracted. Bronze may also have been imported from the Continent in the form of ready made artefacts and parts of artefacts (Northover 1984b: 126). Crucible fragments and tiny droplets of bronze indicate melting and casting; a floor surface at Mine Howe in the Orkney Isles, for example, contained large numbers of small crucibles, scraps of bronze and tiny bronze balls (Card et al. 2005: 326). This site is almost unique in having plentiful evidence of metalwork. The metalworking appears to have been highly organised. Within one hut there is evidence for smelting and melting of bronze as well as hammer scale from ironworking (Collard et al. 2006: 12-14). There is also evidence that there were specially made shelves in the hut to accommodate charcoal and clay for making moulds or repairing furnaces (ibid: 12). The arrangement of materials and activities in this hut resembles the organisation of modern day workshops, as outlined by Keller and Keller (1996), and suggests that specialised metalworkers were living and working on the site. There is some evidence that sheet-bronze and cast-bronze artefacts were made on different types of site (Northover 1988: 226-7). Sheet bronze-working evidence has been found on hillforts. At Maiden Castle excavators discovered concentrations of bronze sheet, rivets and studs, possibly associated with the manufacture of cauldrons (Northover 1991a: 161). Similar evidence has been found at Danebury (Northover 1991b: 408). Tools and moulds have been discovered at South Cadbury (Spratling 1970b: 190) and Breiddin (Northover 1995: 290) and a sheetmould was uncovered at Dinorben (Northover 1988: 227). The casting debris of bronze horse-gear has been found on unfortified sites at Gussage-All-Saints (Spratling 1979) and Wheelsby Avenue (Foster 1995). As Northover (1988: 229) cautions, this evidence is limited to a small number of sites and should not be assumed to be the dominant pattern for the whole Iron Age.

This evidence does not necessarily indicate that mirror handles and plates were made on different sites. However, it may suggest that the production of cast-bronze and sheetbronze artefacts was conceptualised differently by some peoples and for some of the mirrors at least, the handle and plate may have been made by different people (Northover 1993: 35). There is also no reason why objects made from metal should have been a privileged category of object in Iron Age Britain. Mirrors are composite objects requiring a number of different techniques to make the respective components. Indeed some of the techniques used in metalworking are paralleled in the making of objects made from other materials. For example, the techniques used to decorate mirrors are similar to those used to punch decoration into leather. Trace element analysis (TEA) of impurity patterns in the bronze used to make different components could help identify whether mirror parts were made by the same people in the same location. TEA has been criticised as a method for matching potential ore sources to artefacts because ore sources are heterogeneous in nature and selective recycling of scrap can blur impurity patterns (see Dungworth 1997b). However, by comparing impurity patterns of the different components of a mirror it may be possible to suggest different parts were made from different metal sources, even if it is not possible to identify the origin of these sources. Northover (1993) has shown that the handle and plate of the Rivenhall mirror contain significantly different trace elements, implying that the individual parts of the mirror were made from different pieces of metal, obtained from different ores and smelted in different ways. This difference between the material properties of the handle and plate of the Rivenhall mirror may not be surprising if we conclude that they were made from recycled metal already in circulation (ibid: 35). However, when we look at the pattern of impurities from other mirrors (Farley 1983: Table 1; Northover 1993: Table 6; Fitzpatrick 1997a: Table 1) (see Appendix E) it can be seen that differences between the impurity patterns in the bronze plates and handles are maintained (Northover 1993: 35). It seems likely that the handle and plate were intended to have similar alloy compositions (Bayley in Fitzpatrick 1997a: 62). The heterogeneity in tin contents between the plates and handles would only have caused a relatively small difference in colour between the two component parts. There may, however, have been a difference in the surface texture of the plates and handles, even after polishing (Northover 1993: 35). This evidence may imply that different supplies of metal were utilised for the manufacture of mirror handles and plates (ibid). The bronze used to make British mirrors is not significantly different from that used to make other Iron Age bronze artefacts (Dungworth 1997b), but it is very different from the metal used to make Roman mirrors (Farley 1983: Table 1). Roman mirrors have a higher tin content (Appendix E), giving better reflection. They also contain more lead. Roman mirrors have been found deposited on sites roughly contemporary with the deposition of some later mirrors (e.g. King Harry Lane) and it is unclear why some of the later mirrors did not include more tin to improve the quality of reflection. The use of a particular type of bronze to make

16

mirrors, rather than a high-tin bronze, may have had wider social and symbolic significance. 10% tin bronze may have been favoured because it was a particular colour (see Hosler 1988; 1994; 1995) or because it produced a specific quality of reflection (see Saunders 1988; 1999; 2001; 2002). Casting the Handle Direct evidence for bronze-casting is limited. Crucibles have been found on a number of sites as have the fragmented remains of casting moulds for horse-fittings and brooches. At Glastonbury Lake Village there is sufficient evidence to suggest that copper-alloy items were melted and cast on the site and the remains of crucibles of various types and dross have been found (Mortimer 1995: 140-1). At Cadbury Castle, Somerset, copper-alloy slag and droplets were found as well as ‘scrap’ pieces of bronze, indicating that bronze was being melted and cast on the site (Barrett et al. 2000: 294-5). The two most widely known sites where large quantities of casting debris have been discovered are Gussage-All-Saints (Spratling 1979) and Wheelsby Avenue (Foster 1995). However, reading too much into information from these ‘type sites’ could be misleading. For example, much of the discussion of bronze-casting has centred-on the finds from Gussage-All-Saints, but how typical was this find? A skull fragment was discovered within the bronze-working debris in pit 209 (Wainwright 1979: 24) and metalworking evidence from the rest of the site suggests the contents of the pit were not typical of activity at Gussage (Wainwright & Spratling 1973: 124). The remains at Gussage-All-Saints could therefore be the result of specific depositional practice rather than representing the cleared away debris of ‘typical’ activity on the site. Mirror handles were made using the lost-wax technique. Using this technique, the handle was first modelled in wax. The wax model was covered in fine clay, leaving a small vent. After being carefully dried the clay covered wax model was baked until the wax melted and could be poured away through the vent, leaving behind a void in the mould with the exact form of the modelled wax. The next stage involved melting bronze in a crucible and pouring it into the mould cavity. Lead could be added to the bronze to increase fluidity, as it was for the Aston and Colchester mirror handles (Northover 1993: Table 6). The mould was left to cool and was broken-up to retrieve the mirror handle. The handle was then filed and polished (Stead 1996: 12). Importantly in the lost-wax process the mould is destroyed, meaning that each handle is a unique artefact. As yet no moulds have been identified as being used to make mirror handles. No remains of the production of mirrors have been found on any site. Even the mirrors that have been found on settlement sites were deliberately deposited there and are not the result of discard during the production process. Making and Decorating a Plate Producing sheet bronze to make a mirror plate would have been a time consuming and laborious task. A cast ingot of bronze was hammered flat to a thickness of

between 1-2mm (ibid: 9). Mechanical properties of the type of bronze used mean that the action of hammering the plate alone would have produced a reasonably reflective surface (John Fenn pers.comm). The plate was then trimmed to shape and polished. For the majority of the bronze mirrors the newly made plate was then decorated. Examination of silicon moulds made of the decorated mirror backs (Lowery et al. 1971; 1976; Lowery & Savage 1976) has uncovered the techniques and sequence of events of applying decoration. For example, it is possible to determine whether an engraver or tracer has been used as an engraver cuts metal away, whereas a tracer pushes the metal aside to make a groove (Lowery et al. 1971: 173). The process of applying decoration is a different stage in mirror making from the construction of the design. The process of constructing mirror design will be described in detail in Chapter 4. Analysis of about ten of the decorated mirrors suggests that a wide variety of techniques and tools were used to apply decoration. The first stage in the process was to mark out the design onto the plate. On some plates this was done using a pair of compasses. The consistency and precision of the arcs of decoration on the Holcombe mirror show that the decoration could not have been marked out by hand (Lowery et al. 1976: 115). Some of the scratches on the surface of the plate could also have been made by a pair of compasses (Stead 1996: 12) and the decoration was probably mapped out entirely using compass arcs (Lowery & Savage 1976: Figure 2). The lines were either marked directly onto the mirror surface or possibly onto a thin layer of wax or grease applied to the mirror back (Stead 1996: 12). The design was then engraved into the surface of the plate using an engraver held in the hand, much like a pencil. A tracer may also have been used to chase the design. This was held in one hand and driven into the surface of the plate with a hammer held in the other hand. The mirror plates were decorated before the component parts were assembled (Lowery et al. 1976: 100; in Farley 1983: 284; Lowery & Savage in Fitzpatrick 1997a: 62). The mirrors from Dorton, Great Chesterford and Portesham have small areas where the decoration has been obscured by the attachment of a handle or rim (Chapter 5). The Birdlip mirror has an area where engraving or chasing techniques had been practised on the surface of the plate (Fox 1958: Pl. 58a; Jope 2000: Pl. 245e) which was later hidden from view by the attachment of the handle (Stead 1996: 12). The discovery of the hidden decoration on the Birdlip mirror raises a further question: why practice the decoration on the surface of the plate rather than on a piece of scrap bronze? The characteristics of bronze can vary greatly depending on the exact composition of the metal and the state of workhardening. It was therefore necessary for the person engraving the decoration on the Birdlip mirror to adjust to the exact pressure to be applied to the metal being engraved (Lowery et al. 1976: 110). The investment of time needed to decorate the back of a mirror varied greatly depending upon the design and the type of hatching used. For example, it was much more time consuming to produce the intricate basket-work of the Desborough mirror than the Old Warden mirror which

17

could have been engraved in one sitting (ibid: 113-4). As has already been suggested, a mirror design was not necessarily created by the same person who engraved the lines. The engraving of decoration is never ‘perfect’ and close examination reveals areas where the engraved line has strayed from guidelines, these include ‘run-ons’ which are characteristic of this type of metalworking. For example, on the Portesham mirror the person engraving the line has strayed away from the compass drawn guidelines (Lowery & Savage 1976: Pl. 2a). The outline of the decoration on the Birdlip mirror has been worked over twice. The second engraving meticulously followed the inaccuracies of the first (Lowery et al. 1976: 109). A number of authors have judged the mirrors from a contemporary perspective and have suggested that while they are generally competently manufactured, the standard of design and execution is more variable (Megaw & Megaw 2001: 211). Here mirrors have been assessed using our own criteria of skilled craftsmen based on accuracy and economy of skill (Lowery & Savage in Fitzpatrick 1997a: 63-4). However, how do we know that the types of skill we value today were important in the Iron Age? How do we know that the elements of design we find attractive are the same as those valued by the people who made and used these artefacts (Megaw 1985: 164; Megaw & Megaw 1995: 345; MacDonald 2007a: 334)? The performative acts associated with marking sheet metal may have been far more important than how the decoration is viewed through a western cultural lens on the finished artefact. The makers of mirrors were not interested in how it looked to us, rather the skilled performance of making a mirror within Iron Age society would have provided a stage setting to create, reaffirm and maintain identities and relationships (see Dobres 1995: 27-8). Skill does not reside in the artefact; it is a ‘property of people’ and their relations with artefacts (Ingold 1999: x). It has been suggested that skilled people like metalworkers occupied a distinctive position within an Iron Age social hierarchy (see James 1993: 53). However, it is possible to question what status actually means in an Iron Age context (see Hill 2006). Metalworking knowledge would have made metalworkers different from other people and this difference was maintained and emphasised through the performance of making metal artefacts. However, as Herbert (1993) has shown for African ironworkers, how this difference was perceived may have varied greatly through time and from region to region. Assembling a Mirror Once the mirror back was decorated, the front of the plate was planished, to remove indentations and bends caused by engraving or chasing. The plate was then ground and polished on both sides (Lowery et al. 1971: 175-7). After the mirror handle was cast and the plate produced and decorated the two components were attached to one another. Methods of attachment varied. The component parts of the smaller mirrors were attached by wedging the plate into a slot pre-cast into the handle. As well as the

slot, the handles of the larger mirrors were secured to the plate using rivets or pins. The addition of a bronze rim or edge binding around the plate acted to secure the attachment further and to protect the plate edge. It is likely that final polishing and finishing took place after the mirror had been brought together as one artefact. Analysis of the plate of the Portesham mirror at the juncture of the plate and handle has revealed a surface close to, but not in its final, polished condition (Lowery & Savage in Fitzpatrick 1997a: 64). Summary As a summary to this section the potential social relationships that were formulated when a bronze mirror was made will be outlined using the example of the Portesham mirror. The Portesham mirror has three component parts: a cast handle, decorated plate and edging. Figure 3.1 shows the different production processes required to make the Portesham mirror. It also divides these production steps into stages (the grey boxes numbered 1-7). The collection and processing of ores (1) and the smelting of metal (2) were the first two stages. As outlined above, these could have been very far removed from the final artefact in both distance and time; however, they could have had major implications for the meaning of a mirror. The metal the mirror was made from could have originally been used to make other artefacts (3) before it was used to make a mirror. Bronze may also have been obtained through exchange and trade (4) before it was used to make the different component parts. The three components of a complex mirror like Portesham were made separately. The handle (5) and the plate (6) required different skills to make and may have been made mirror by different people, possibly in different locations. The decoration on the plate could also have been designed

Figure 3.1 Processes in construction of Portesham mirror

18

and inscribed (7) by different individuals (Chapter 4). At each of these different stages a number of individuals could have been involved in the making of a bronze mirror and they provide a stage setting for social interaction. The manner in which the component parts come together to form a well ordered and coherent artefact suggests that even though mirrors may have been made by several individuals, a particular person or group was responsible for coordinating their making: they wanted the mirror and caused it to be made by formulating specific social relationships centred around the production stages highlighted above. As will be shown in later chapters, there is no reason to suppose that mirrors were made for the exclusive use of a single individual. They could have been made by the community for the community to perform specific social roles. By decorating a mirror and adding a rim to it, the complexity of these social interactions are increased as an extra component is added and potentially more individuals become involved to decorate the mirror. 3.4 IRON MIRRORS Iron mirrors were made using very different techniques to bronze mirrors (cf. Salter & Ehrenreich 1984: 146) this is because of the different properties of iron, especially its higher melting point. Iron Age furnaces were not able to produce a sufficient temperature to melt iron. Consequently, there are no cast iron Iron Age objects and iron mirrors are made from wrought iron instead. Producing wrought iron is a complex process with more than one stage (Hingley 1997: 11). First, iron was separated from the ore, or smelted. Then the metal was forged to hammer out impurities and smithed, a term used to describe the process of shaping the metal to produce the final artefact.

Location of Ores In contrast to copper and tin, iron ores are widely distributed throughout the British Isles. In southern England iron ores were even collected from the surface of fields after ploughing. Bog irons are relatively plentiful and have been used in the production of iron from the earliest times (Tylecote & Clough 1983). These are often found in locations marginal to settlement, in areas of high rainfall. It is possible that some significance was drawn from the marginality of these locations. Iron ore was probably collected on a seasonal basis. It is unclear if access was restricted to particular communities, or community members, or what were the social mechanisms of control. Smelting Iron Ore Pure iron melts at a temperature of 1534•C, whereas pure copper melts at 1083•C (Salter & Ehrenreich 1984: 146; Ehrenreich 1985: 21). The furnaces of the Iron Age were unable to produce sufficiently high temperatures to melt iron. The most efficient furnaces at the time were probably only able to maintain a temperature of 1100-1300•C (Cleere 1972: 8). Iron remained solid throughout the smelting process and the separation of the metal from impurities, or slag (which melts at around 1200•C), was often poor. Consequently, iron from the period is characterised by the inclusion of slag. At Danebury slag inclusions made up an average of 5% of final artefacts, causing a detrimental effect on mechanical properties (Salter & Ehrenreich 1984: 146). As a result of these problems only rich iron ores were utilised and metal had to be forged to hammer out impurities (ibid). Even after the introduction of more efficient slag-tapping furnaces in the 1st-century BC (ibid), the whole process of smelting iron would have been very time consuming and labour intensive. Salter & Ehrenreich (1984: 147-9) suggest that to produce 1kg of iron it would require: 20kg of iron ore; 90kg of charcoal (which was produced using wood from managed coppiced woodland and required 630kg of wood to produce the required weight in charcoal); and 40kg of clay to make the furnace. This obviously restricts the sites where iron could be smelted to areas with a supply of iron rich ore and sufficient woodland. A note of caution should be expressed for this generally techno-centric, evolutionary outline of iron smelting. For example, until recently it was thought that ‘traditional’ methods could produce only one type of iron – wrought – which is produced in low temperatures. The end product is a bloom high in impurities which have to be hammered out. However, in some parts of Africa they were able to produce steel by placing the tuyères deeper inside the furnace to heat up the air inside them and thus achieve higher furnace temperatures (Kense & Okoro 1993: 450). It is probable that steel was produced on a number of separate occasions in Iron Age Britain (J.D. Hill pers.comm). It is illustrative of the kinds of relationships that existed between metalworkers that the methods used to make steel were not common knowledge and were not necessarily passed on to others, hence they were probably rediscovered a number of times throughout the Iron Age period. Traditional interpretations of furnace technology (e.g. Cleere 1972) have also charted a linear progression from simple bowl furnaces to

19

Roman blast furnaces, however it has been shown that bowl furnaces are only effective in producing metal from high quality ores and that furnace type is more likely to be dictated by the type of ore being smelted (Crew 1991: 150). Once the iron had been smelted and the impurities hammered out the final artefact was made by a process of heating and hammering, called smithing. Location of Smelting Sites The biggest clue as to where smelting sites in Britain were located comes from the production of slag, the waste products from smelting, which were probably not removed far from the location of smelting. It is difficult to work through evidence for ironworking on Iron Age sites because of the lack of distinction between smelting and smithing in site reports and confusions in terminology (Crew 1991: 150). One of the key problems in the identification of smithing and smelting sites is that it is extremely difficult to distinguish between smithing and smelting slag (ibid: 151). Reflecting this, the following brief overview of evidence for ferrous metalworking will include evidence for both smelting and smithing activities. A fuller account of iron smithing will also be included below. A number of sites in Wales have produced evidence for ferrous metalworking. Evidence for metalworking has been found in the ramparts of Llwyn Bryn-Dinas hillfort, Clwyd (Musson et al. 1992: 277). At Bryn y Castell hillfort, evidence for smelting and smithing has been found dating to the late Iron Age and the late 2nd/3rd centuries AD (Crew 1991: 155). At Crawcwellt, south of Trawsfynydd, evidence of smelting was found specifically at an isolated hut on the fringes of the settlement (ibid: 157). The excavators found the remains of 15 furnaces at the site, although it is not certain to what stage the iron was processed after smelting (ibid: 14). Crew (1989) debated the nature of the iron smelting activities and hypothesised whether the excavated material represents a relatively intensive period of smelting or smaller scale activities over a period of time. Given the longevity of the site, the latter explanation is perhaps most likely. The site has also produced evidence for a ‘dump’ of ore and another of roasted ore near the entrance of the enclosure (ibid: fig. 2). At Bryn y Castell hillfort a number of metalworking features were found dating from the early 1st-century BC – late 1st-century AD (Crew 1987: 98). A well preserved smelting furnace was discovered just outside of the north rampart and just inside the terminal of the north entrance a dump of iron ore was also found. There is also the ‘suggestion’ of a smithy in the north corner of the fort. Crew (1991: 157) stated that the location of these Welsh sites is significant because they are close to areas containing bog iron with very low agricultural potential. In the rest of the country evidence for ferrous metalworking is patchier. At Brooklands, near Weybridge, two main

concentrations of ironworking were found in the east and west of the site (Cleere 1977: 15). The scale of ironworking was small (ibid: 22) but smelting appears to have occurred in the western area and forging in the eastern area (ibid: 19). This physical separation of smithing and smelting has been noted on other sites (ibid: 20). Evidence of metalworking is widespread at Wakerley, with a concentration of smelting furnaces outside the southern corner of the main enclosure (Jackson et al. 1978: 151). There was also evidence of forging and smithing. A purpose built hut at Round Pound, Kester, Devon contained a furnace and an area for forging (Fox 1954: 97). Two bowl furnaces for smelting iron were also uncovered at West Brandon, near Durham (Jobey 1962: 19-21). Iron slags and smithing hearths have been found at Glastonbury Lake Village suggesting that both smithing and smelting took place on the site (Mortimer & Starley 1995: 139). The quantity of material found – especially evidence for smelting – is insufficient to indicate that metalworking was an important activity (ibid: 140). Two working hollows and fragments from the 6th-century BC were found at the Ashville Trading Estate in Oxfordshire (Cleere 1978: 38). There is evidence of bronze and iron working and it is possible that some of the tools found on the site were made there. Metalworking was on a relatively small-scale, probably for local consumption (ibid: 90). Based on fragments of a tuyereè, mould and hearth lining, dating to the early 1st-century AD, Spratling (1981: 13-14) suggests that metalworking was an important activity at Stanwick, Yorkshire. Iron slag was also found at Gussage All Saints in Dorset, as well as the base of an iron smelting furnace (Fell 1988: 73). The Foulness Valley One area where extensive evidence for iron smelting has been uncovered is the Foulness Valley in the East Riding (see Halkon 2008). Smelting in this area has important implications for the study of mirrors because it is on the Wolds to the west that five iron mirrors have been found in graves. Iron ore is not found on the high, well-drained chalk land of the Wolds. However, further east in the low-lying, poorly drained, valley of the river Foulness ‘bog iron’ occurs commonly. It is probably here that the iron used to make the objects found in the barrow burials of the Wolds (including the mirrors) originated (Dent 1998: 17), as there is also extensive evidence for iron smelting and smithing in the area. In a recent field walking survey of the valley (Halkon & Millett 1999) iron slag was discovered on fortynine sites and the majority of these dated to the Iron Age. For example, at Welham Bridge a slagheap exposed by peat shrinkage was excavated in 1985 and radiocarbon dated to between the 6th and 3rd centuries BC. The material from Welham Bridge probably represents the waste from approximately 35 different smelts (ibid: 75-94). Where sites have been identified they occur on the edges of the valley’s creek system and seem to be marginal to the main areas of settlement in the region, which aerial photographs have identified on the higher, drier ridges (ibid: 95). This location places iron smelting close to the source of raw material – bog iron – and may also have facilitated the easy transport of a bulky material by boat (Millett & McGrail 1987). It also places metalworking in a marginal location, on the

20

edge of settlements and at the junction of contact/trade with other communities. Environmental evidence suggests that the valley was made up of managed woodland (probably providing the wood to produce the charcoal required for iron smelting) with small areas of arable land and some areas for livestock (Halkon & Millett 1999: 221). The production of iron ore was on a small scale over a long period of time. This does not mean, however, that the quantity of iron ore produced was not large. At Welham Bridge total iron production has been calculated to be around 540kg, or the equivalent of over 800 currency bars (ibid: 95). The people living in this area probably practised a mixed farming economy, managing woodland for charcoal production and smelting iron ore. It is likely that some of this iron ore was used in the Wolds, making the Foulness Valley critical to our understanding of the Iron Age in eastern Yorkshire (ibid). To sum up, there is limited evidence for smelting sites in the Iron Age and it may have been a specialised activity occurring on a small number of sites, or in specialised areas such as the Foulness Valley, or some parts of Wales. Smelted iron may then have been exchanged to the locality where it was smithed (see Hingley 1990). The smelting of iron intensified throughout the Iron Age. Hingley (1997: 10) has noted that by the later Iron Age smelting was occurring, on or close to settlement sites rather than in isolated localities. Iron Smithing The smithing of iron to produce the final artefact or to repair/modify artefacts has proved to be difficult to locate because a domestic hearth was adequate in order to generate sufficient temperatures to work iron. An inability to melt iron and cast it in moulds necessitated the development of a smithing technology to hammer out impurities and shape the metal. This required the production of blacksmithing implements like hammers, fullers, tongs, files, chisels and anvils (Ehrenreich 1985: 94). The presence of smithing tools may indicate smithing activities. For example, at Garton Slack a pair of iron tongs and a set of blacksmith’s tools were found in a pit used to store grain, dated to the 2nd-century BC (Brewster 1976: 112). This not only provides evidence of blacksmithing in the Wolds, it also suggests a link between iron smithing, agriculture, and perhaps fertility (see Herbert 1993). It also suggests that if metal objects were not made in settlements like Garton Slack, they certainly could have been repaired there. Metalworking tongs were also found in a grave in the Rudston cemetery (Stead 1991c: 79) and at Llyn Cerrig Bach, Anglesey (Fox 1946). A number of iron objects, dating to the 3rd and 2nd centuries BC were found in the ramparts at the Breiddin hillfort, including a knife, a punch and some tongs (Saunders 1991: 141). Iron metalworking tools have also been found at Glastonbury Lake Village, Hod Hill, Hunsbury, Llanmelin and Meare Lake Village (Spratling 1970b: 190). At Santon in Norfolk two pairs of tongs, an iron hammerhead, two bronze formers and a bronze anvil were placed in a hoard alongside Roman legionary armour dating to the mid 1st-cen-

tury AD. In the Polden Hills hoard, Somerset, there was a small bronze hammerhead (Spratling 1970b). The presence of hammer-scale suggests the occurrence of this stage in the making of iron artefacts. This has been detected at Gussage-All-Saints (Spratling 1979: 127; Fell 1988: 73) and dated to the 10th-century BC, at Hartshill Copse, West Berkshire (Collard et al. 2006). In many parts of Africa the smithing stage in the production of metal artefacts is highly ritualised. There are rituals associated with smithing especially related to the establishment of a new forge or the making of a new hammer and anvil. Making the first hammer and anvil are sometimes seen as a rite of passage for a new smith (Herbert 1993: 98-9). Smithing, like smelting, is seen as a predominantly male activity, subject to many of the same taboos as smelting. For example, women are often, but not always, excluded from forges, especially during pregnancy or menstruation (ibid: 107). Smithing is generally a more public activity than smelting. Smithies are also associated with fertility and they are seen to help barren women conceive. Some smiths also take control over circumcision rituals (ibid: 108). The archaeological evidence from the Iron Age also suggests a separation of smelting and smithing activities with smelting occurring close to the origin of iron ores and smithing taking place on the fringes of settlements (see Hingley 1997; 2005; 2006). This trend changes towards the end of the Iron Age, as smelting and smithing took place in, or close to some sites, although there seems to be a separation of smelting and smithing activities on an intra-site level. Making an Iron Mirror The archaeological evidence indicates that smelting was undertaken by selected groups of people who practised smelting on a relatively small scale over a period of time, fitting metal smelting into the agricultural calendar (see Halkon & Millett 1999). As in some parts of Africa, the people who smelted iron may have been specialist endogamous groups. Smelted metal was then passed between peoples to a different location, sometimes in the form of iron ‘currency bars’, where it was made into artefacts by smiths. The process of making and repairing artefacts is likely to have taken place in or close to the boundaries of settlements. These activities required a different level of skill and it is likely that each settlement possessed someone who was able to repair artefacts. A smaller number were able to make artefacts from scratch. Five iron mirrors have been found in East Yorkshire (Stillingfleet 1846; Brewster 1980; Dent 1985a; Hill 2001a) and another from Lambay Island, Dublin (Raftery 1994; Rynne 1976) in addition to a number of other mirrors with iron plates. The iron used to make the East Yorkshire mirrors was probably sourced and smelted in the Foulness Valley. The isolated and ‘different’ nature of the Foulness Valley from the Wolds may have added to the significance of the metal. The metal could have then been smithed in the Wolds. For example, as already mentioned, at Garton Slack a pair of iron tongs and a set of blacksmith’s tools were

21

found in a pit used to store grain, dated to the 2nd-century BC (Brewster 1976: 112). Blacksmith tongs have also been found in a grave in the Rudston cemetery (Stead 1991c: 79). Making an iron mirror was a prolonged and highly skilled process. A piece of iron was repeatedly heated and hammered to produce the circular plate. An iron bar for the handle was also produced in a similar fashion. The majority of the iron mirrors have a ring at each end of the handle. A hole was punched into each end of the bar handle to produce the end rings. The plate was then polished to produce a reflective surface before the plate and the handle were welded together. Unlike the bronze mirrors iron mirrors were not decorated, although some had bronze mounts. The plate and handle also required similar skills to produce and could have been made by one individual. Much of the significance of iron mirrors, especially from East Yorkshire, derived from the origin of the iron used to make them, rather than the relationships acted out through their manufacture. Summary As a summary to this section the potential social relationships formulated when making an iron mirror will be examined. As is illustrated in Figure 3.2, which shows the stages in the production of the Wetwang Village mirror, making an iron mirror is far less complicated than the processes involved in making a bronze one. The first group of steps involved collecting iron ore and smelting it (1), in the case of Wetwang Village this could have taken place in an area like the Foulness Valley that is geographically distinct from the Yorkshire Wolds where the mirror was probably made. In order to obtain iron to make iron objects like the mirror a system of relationships was in place between the communities of the Wolds and the Foulness Valley (2). The third stage in the production of the Wetwang Village mirror (3) involved smithing the metal into the final artefact. Making an iron mirror involved fewer individuals and the formulation of far fewer relationships than a bronze mirror. 3.5 MIRRORS MADE OF A COMBINATION OF BRONZE AND IRON COMPONENTS Some mirrors were made from both iron and bronze component parts. The mirror from Glastonbury Lake Village has a bronze plate and originally an iron handle (Fitzpatrick 1997a: 67), the Balleybogey Bog bronze mirror handle originally had an iron plate (Jope 1954). The Maiden Castle mirror had an iron plate and a bronze rim (Wheeler 1943). It has generally been assumed that bronze and ironworking were separated because they involved different skills. The metals may have also been conceived of differently (Gosden & Hill 2008: 8). At Cadbury Castle, Somerset, evidence for both bronze working and iron working has been found (Barrett et al. 2000). Both ferrous and non-ferrous metalworking debris have been found at Glastonbury Lake Village (Mortimer 1995: 138). Importantly, evidence for bronze and ferrous metalworking was found on the same floor level, indicat-

Figure 3.2 Processes in the construction of the Wetwang Village mirror ing that the two procedures took place contemporaneously, at Glastonbury Lake Village (Coles & Minnitt 1995: 142) and Gussage. Very few metalworking tools were found but the presence of sheet-bronze and dome-headed rivets suggests that sheet bronze was also being worked (ibid: 143). Two iron currency bars were also found suggesting that, although iron may not necessarily have been smelted on the site it was certainly forged (ibid). At Mine Howe, Orkney, bronze was cast and iron worked in the same hut (Collard et al. 2006). This implies that the working of bronze and iron may have been separated in some cases, while they operated together in others. The social significance of making a mirror from both iron and bronze component parts should be discussed. The use of different techniques and materials to create a single artefact could reference a whole series of different Iron Age artefacts in the material form of the mirror (see Chapter 6). The combination of different materials could also have been utilised to emphasise particular colours, such as the pale yellow colour of bronze which may have had social and symbolic significance (see Herbert 1984; Hosler 1988; 1994; 1995; Jones 2002b; Jones & MacGregor 2002a; 2002b; Saunders 2002; Scarre 2002). Summary The production of a mirror made of both bronze and iron components involved a very complex series of different relationships. These are illustrated through the example of the Balleybogey Bog mirror. Figure 3.3 shows the possible stages in the construction of the Balleybogey Bog mirror, which comprises a cast bronze handle, and originally, an iron plate. Contrary to the impression created by

22

some commentators, composite mirrors would have required the creation of a whole series of different social relationships in order to produce them. In many cases these relationships were more complex than those formulated in the manufacture of bronze mirrors. Two metals (iron and bronze) were sourced from different locations. The two component parts were processed in different ways (1, 2, 3, 4 & 6). The skills of bronze casting (5) and iron smithing (7) were also required to make the mirror implying that the two components of the Balleybogey Bog mirror may not necessarily have been made by the same individual. 3.6 CONCLUSION To conclude the number, variety and complexity of different social relationships formulated when a mirror was made will be highlighted. Thinking about mirror making from a biographical perspective has facilitated the division of mirror production into a series of stages, each of which involved social interaction. Stages in production differed depending upon the material/s a mirror was made from and the number of component parts; a decorated, rimmed mirror required more production steps than a plain un-rimmed mirror. Similarly, a mirror with an iron plate and a bronze handle required wide ranging technological skills to construct. The ease with which bronze can be recycled, the presence of ingots of copper, tin and lead on later Iron Age sites, and the relative scarcity and geographical marginality of

Figure 3.3 Processes in the construction of the Balleybogey mirror ores, suggests that smelting of the ores was separated from the actual production of bronze artefacts like mirrors. A sequence of events can be seen in the production of bronze mirrors. The handle was cast, the mirror plate was hammered flat and cut to shape and then decorated before the handle was attached and a rim formed around the edge of the plate. Bronze mirrors are special because they comprise both cast and decorated sheet components. It may not be possible to suggest that the handle and plate were made in different places by different people or groups of people; however, the consistent differences in trace elements between mirror handles and plates suggests that either different supply arrangements existed for cast and sheet copper-alloy artefacts and/or that the metalworkers were deliberately selecting metal with distinct physical properties to make the mirror handles and plates. Iron mirrors were made using different techniques. Much of their significance arises from the origin of the iron used to make them.

23

Composite mirrors neatly combine the different materials and methods of artefact production into one material form. All three types of mirror express different meanings through the context of manufacture, methods of constrution, and the different relationships and stages which brought about their making. They are all complicated artefacts requiring the creation of a number of relationships in order to make them. It is these relationships that were critical to the formulation of the meanings and associations these objects would carry into their future lives. However, these relationships were not only meaningful in relation to the future lives of the objects being made, through the making of a complex artefact like a mirror, a shield or a sword for example, the person or group of people making that object were demonstrating their ability to coordinate and create the required nexus of relationships; they were recreating their own social world.

CHAPTER 4 RE-EXAMINING MIRROR DECORATION 4.1 INTRODUCTION In the previous chapter a whole series of social relationships formulated when a mirror was made were revealed. Mirror manufacture can be divided into a series of stages and each of these stages act as a potential setting for social interaction. One identified stage was the decoration of bronze mirrors (stages 6 and 7 identified in Figure 3.2). The plates of 28 of the known 58 mirrors are decorated (see Appendix F). Decoration is inscribed or chased on the opposite side to the reflective surface of bronze mirrors and individual designs vary widely (Lowery et al. 1971; 1976; Lowery & Savage 1976). Designs were created using a distinct set of motifs (see Joy 2008b: Appendix 2), filled with baskethatching. In addition to the decoration on mirror backs the term ‘mirror style’ is also used to describe other objects decorated with the same hatched motifs (Appendix C), including some sword scabbards and items of horse-harness (see Fox 1946: 53; 1958: 84-98; de Navarro 1952: 79). Beyond possible references to nature (e.g. Jope 2000: Pl. 185), the meaning of mirror decoration has never been addressed (see Joy 2008b; Spratling 2008). 4.2 RE-ASSESSMENT OF THE EXISTING CATEGORISATION OF MIRROR DECORATION The mirror-style specifically relating to mirrors, has been discussed before, most notably by Fox (1958: 84-98) and Jope (2000: 137-48). Echoing work on other forms of Celtic Art, previous studies of mirror decoration tend to be descriptive. They attempt to identify regional styles, locations of production and work to create a chronologically significant, evolutionary sequence of mirror design. This stylistic sequence is still influential in dating decorated mirrors (e.g. see discussion by Jope in Fulford & Creighton 1998: 339). The sequence was created by comparing designs based on aesthetic value judgements. It is anchored by the limited number of mirrors found with datable associated artefacts. Mirror design, it has been argued, underwent a necessary, preliminary, developmental stage and a ‘flowering’ epitomised by the patterned backs of mirrors placed later in the sequence, like Desborough (see Jope 2000: 142), as each new design ‘improved’ upon previous designs. Mirrors placed latest in the sequence, like the Nijmegen mirror, are examples of mirror design (which is) past its peak (see Dunning 1928: 212). Mirror decoration as it appears in these studies, can be seen to take on a life of its own; it seems to be self-perpetuating, always evolving and becoming ‘better’ as each new generation of mirror ‘improve’ on the previous one, forming an unbroken sequence of design. A model of ‘art’ and the ‘art-world’ has been imposed on mirror design, where ‘aesthetic taste’ is valued and artworks are compared and assessed on the basis of judgements of aesthetic worth. No criteria or explanation for why one design is valued over another is put forward. No specific criteria of what makes one mirror design ‘better’ than another is given, even though assessments of ‘success’ are often made. Often judgements were based, not just on the design itself, but also the technical proficiency of the inscriber. For the majority of mirrors the process of constructing a design and physically inscribing it onto a mirror plate were probably separate operations. Many mirror designs were laid out using compasses. Designs are far too compli-

24

cated and precise to indicate that they were created as they were inscribed. Planning a design and inscribing it are therefore two different processes and should not be confused. As MacDonald (2007a: 334) has suggested, whilst this kind of approach is extremely valuable to dealers, collectors and museum curators, it does not tell us much about objects like mirrors or the Iron Age. Mirror design, like all Celtic Art, is a product of the particular social context in which it was created and we do not know how each individual design was valued in the Iron Age. We are therefore unable to judge whether one mirror design is ‘better’ than another and it is invalid to place all of the decorated mirrors in a line and chart the rise (and fall) of mirror decoration. The assumptions that underlie the idea of a stylistic sequence enforce a particular kind of interpretation with an implied chronological progression. The constructed sequence of mirror design is a result of the privileged viewpoint from which we are able to examine mirror decoration as archaeologists (see Jones 2001: 339), which has been viewed through a western cultural lens. Iron Age people almost certainly did not have the benefit of seeing a large number of decorated mirrors like we do. Based on the number we find in the archaeological record, it is unlikely that anyone would have encountered a significant number in their lifetime. Different models of metalworking, such as the idea of the workshop or itinerant metalworker, have been uncritically applied to the making of mirrors. For example, Fox’s (1958) analysis uses a model that is similar to the world of Renaissance art portrayed in Benvenuto Cellini’s autobiography (Cellini 1926). It relies on the existence of ‘workshops’ and ‘patrons’; these models for artefact production have not been tested in an Iron Age context. Complicated objects like mirrors could have been made by several individuals, the whole process being coordinated by a particular person or group (Chapter 3). The relationships formulated during their manufacture would have been critical in setting the trajectories of the future lives of mirrors and the way they were decorated would have formed a significant part of this process. 4.3 ANALYTICAL PROCESSES In order to try and remove some of the value judgements associated with the history of the analysis of mirror decoration a series of analytical processes were undertaken, establishing criteria to enable comparison between designs on the different mirrors. These criteria are based on individual design motifs. This was done so that the basic building-blocks used to construct mirror design could be identified, providing a means to compare mirrors and their decoration that is derived from formal analysis and defined specific criteria: 1. The design rules used to create mirror design were identified. 2. The space on the mirror which is not utilised was examined. 3. The complexity of mirror design was analysed.

Terminology Before the first stage of this analytical process is discussed it is necessary to briefly explain how mirror decoration works and introduce some terminology. The surface of the mirror back containing decoration and delimited by the edge of the plate will be referred to as the design-field. The overall design is the distinctive pattern within the designfield. Space within the design-field can be defined on three different levels. At the first level are the positive motifs. These are delimited by hatching. At the second level negative motifs are areas of distinctive and repeated form within the design-field left un-hatched. Negative motifs are created by positive motifs and the edge of the design-field. At the third level any part of the design-field which is not a positive or negative motif is defined as unused space. Unused space describes any un-hatched area in the design-field which does not have a coherent form which does not occur three or more times in the same design, or is not repeated on the design of more than one mirror. This definition allows for the fact that in a symmetrical design, like the decoration on the Nijmegen mirror, negative areas will be repeated through the act of symmetry, rather than deliberately demarcating a space of known and repeated form. 4.4 MIRROR-STYLE MOTIF REPERTOIRE Positive Motifs The decoration on all of the decorated mirrors was examined and the positive motifs, delineated by hatching, were identified. The positive motifs used to create mirror design are limited in number. Analysis of other decorated Iron Age material culture has revealed that none of these motifs are unique to mirrors (see Appendix C). The names used to describe individual motifs are derived from Fox (1949: 49; 1958: 147-8), Jope (2000: 379-85) and MacGregor (1976). Where it was not possible to find a term to describe a particular motif a name has been given (e.g. armadillo). These are entirely arbitrary; they have no meaning beyond being a convenient label.

The cusp is another widely used mirror motif (ibid: 380). Like the fin the cusp is three-sided. The two shorter lines are of equal length and are concave. The longest side is convex, creating an axe-shaped motif (Figure 4.2). The cusp differs from the fin in that the two concave sides of the motif are of equal length. The cusp is often employed in a similar manner to the fin as a ‘building-block’ to attach other motifs onto.

Figure 4.2 A cusp

Figure 4.3 Trumpet motifs The third major motif integral to the design of some of the mirrors, including Gibbs and Colchester I (Lexden Grange), is the trumpet motif (see Fox 1958: Fig. 83). The trumpet motif is found in three main forms in mirror design (Figure 4.3). Two are three-sided and one is foursided. The first (A) composes of two s-shaped sides of equal length and a third side which is concave. The second type (B) of trumpet void has one concave side, another which is convex and the third side is s-shaped. The third type (C) is an extreme form of the second and resembles a bird’s wing. It has four sides. Two are convex, one is s-shaped and the fourth side is almost a straight-line, which can tend towards concave. In addition to these three basic building blocks, a number of other less common positive motifs have been identified.

Figure 4.1 A fin The basic design motif common to the majority of mirrors is a flat, long triangular shape called a ‘fin’ (Jope 2000: 381). The shortest two sides of a fin are different lengths and are inward curving. The longest side is convex (Figure 4.1). The fin is stretched and curved to form the main sections of the positive design on the majority of the decorated mirrors. Sometimes with the aid of compasses, different variations of this basic fin shape were combined to compose the overall design within the design-field.

25

The first major group consist of designs within circles (Figure 4.4). They are generally small in size in comparison to some other positive motifs and are not confined to mirror plates, as they are found on some mirror handles. The simplest of these motifs is a circle filled with baskethatching. The whirligig, a common Celtic Art motif, which consists of three half-moon shapes positioned in such a way that they resemble a ships propeller, can also be contained within a circle. This motif appears on the handle of the Balmaclellan mirror, for example. The circled tricorne is another common motif. These consist

Figure 4.4 Designs within circles of a ‘curved triangle, made-up of three inward-curving arcs’ (Jope 2000: 384) contained within an inscribed circle. There are a number of different types of circled tricorne, dependent upon which sections of the motif (if any) are hatched or filled. Circled tricornes are often employed in association with fins to form keeled-roundels. The armadillo motif, which consists of a smaller circle within a larger circle, is often utilised as an embellishment to a fin or cusp, by placing it at one of the points on these motifs where two sides meet (see bottom half of Figure 4.5). The armadillo motif is constructed by positioning the smaller of the two circles off-centre within the larger circle, its outside edge touching the larger circle. The crescentshape created by doing this is then hatched. A keeled-roundel looks like a hard-boiled egg cut in half and viewed from side-on and is more or less circular with a ‘keel-shaped protuberance’ (Jope 2000: 381). The keeledroundel shape is a recurring motif of mirror design and of the mirror-style. They can be created in a simple way. For example, a circled tricorne can be placed within two fins, as on the Bromham mirror, or a filled-circle as is the case of the Great Chesterford mirror (Figure 4.6). The keeled-roundels on Desborough differ from these as instead of forming a single motif, the keel of the roundel is formed of two

Figure 4.5 The Armadillo motif

26

separate fin motifs. The three other motifs appear less widely, or they are unique to one mirror. The kidney and crescent motifs can be seen on Birdlip and Nijmegen respectively. The adjoined double cusp (ADC) appears on Portesham and Billericay I (Figure 4.7). Negative Motifs Figure 4.8 illustrates all of the identified negative motifs. Negative mirror motifs are often identical in form to the positive ones. Negative tricornes, circles, crescentshapes, keeled-roundels, cusps and fins have been identified. Negative motifs are not only delineated by the positive motifs, they can also be framed within the confines of the plate. The edging surrounding some mirrors often acts to re-emphasis this, operating in a similar way to the edge of the positive motifs. Alongside the fin and the cusp, the most significant negative motif is the trumpet-void, which has also been referred to as the ‘Llyn Cerrig void’ (MacGregor 1976). Fox (1946: 48-53) identified the trumpet-void, which is constructed from three lines, one concave, one convex and a third S-shaped (Parfitt 1995: 93; Stead 1996: 34), when he examined the metalwork from Llyn Cerrig Bach, Anglesey. This motif is often created by the negative space in-between the large roundels of a tri-partite design, as on the Gibbs mirror. The trumpet-void has been identified on many

Figure 4.6 Roundels on Bromham (A) and Great Chesterford (B)

Figure 4.7 Three additional mirror motifs (A - Adjoined Double Cusp, B - Half-moon, C - Kidney) objects including the repouseé decoration on the bronze plaque from Llyn Cerrig Bach (Fox 1946: 48-53, Fig. 25) and the openwork sheet bronze cover for the shield-boss from the Deal grave, Kent (Parfitt 1995: Figs. 23 & 39). In general, where trumpet-voids occur, they are a significant element of the design, but appear alongside other motifs. Parfit (ibid: 93) has suggested that for some objects, like the shield-boss cover from Deal and the Old Warden I mirror, trumpet-voids can be the primary design motifs. 4.5 Design Rules The next stage of the analytical process was to identify how mirror motifs are organised to form characteristic designs: how did they take the identified motifs and come up with a design? Fox (1946: Fig. 31) compared the roundels on the Llyn Cerrig shield boss to the designs of the Billericay I, Desborough and Birdlip mirrors. He showed that negative motifs (especially the trumpet-void) consistently form an integral component of the overall design on mirrors, as well as some other objects like the bronze collar from Balmaclellen decorated with ‘mirror style’ motifs. Fox also illustrated the importance of negative motifs to the overall coherency of designs through the example of the Mayer mirror which he illustrated, in his book Pattern and Purpose (1958: 83, Pl. 56a. A1 & A2, see also Figure 17), like a photographic negative so that the negative motifs were filled black. Analysis of the decorated mirrors has confirmed Fox’s sugges-

27

tion that negative motifs are critical to mirror design. Mirror design is consistently constructed using a small number of positive and negative motifs. Mirror decoration can clearly only be understood in terms of the relationship of its positive and negative elements. Mirrors are decorated so that they conform to a set of design rules. Even though designs can vary greatly all but two of the decorated mirrors conform to these design rules. Mirror design is made up of positive and negative motifs which form a balanced composition in order to keep unused spaces to a minimum. Figure 4.9 outlines how mirror decoration operates. All of the positive motifs are illustrated at the top of the diagram. A number of examples of how these motifs are used to create more complicated designs are shown. As you follow these designs down the diagram they become more complex, showing how positive motifs are arranged to form negative motifs. At the bottom of the diagram all of the negative motifs are illustrated. Before examining how well individual mirrors conform to these design rules, it will first be explained how different mirror designs were created. Design Construction A number of different methods were developed to construct designs to conform to these design rules. Fox (1958:

(see MacGregor 1976: 'grammar of ornament'). The decoration on the Aston and Gibbs mirrors employs a lyreloop to section off the design-field. This has the effect of creating three negative motifs at the edge of the designfield, two trumpet-voids and a cusp at the bottom. The design of the Pegsdon mirror is arranged in a different way (Figure 4.13). Two circles are outlined at the bottom of the mirror and an elongated lyre-loop is positioned above creating four negative fin shapes on the right and left hand outside edges of the plate and a trumpetvoid at the bottom. A recently discovered mirror, possibly from Oxfordshire, also uses the same method of dividing the design-field. Figure 4.14 explains the sequence of design for the Aston mirror. Negative motifs are shaded grey. If a particular stage in the design causes new negative motifs to be created then new negative motifs are shaded dark grey. The first stage in the creation of the design for the Aston mirror was to draw a lyre-loop (A). This framed three large negative spaces, two trumpet-voids and a cusp, on the outside edge of the design-field (B). The addition of two fins framed a negative trumpet-void in the centre of the plate (C). This simple design was then completed with the addition of three circles (D) and fin motifs (F) to create a facing pair of keeled-roundels at the bottom of the design-field (G and H) and three negative crescentshaped areas (F).

Figure 4.8 Negative motifs 97, Fig. 62) identified four basic mirror designs (Figure 4.10). Based on new discoveries, it is possible to add two more (Figure 4.11). A sample of these different designs was taken, working through how they were constructed with the identified design rules in mind. It was found that an initial framework could be created which divided the design-field into more readily comprehensible areas. These were then sub-divided until a series of round areas – called roundels by Fox and Jope – had been defined. A roundel in the context of this chapter refers to any area of decoration in the design-field which is arranged in a circular or near-circular space. A discussion of how some roundels were created using mirror motifs will follow this section. The simplest way to divide the design-field into smaller units, which has been used on the St. Keverne mirror, was to draw two equal-sized circles along the centre line of a circular plate. This had the effect of creating two large negative cusp motifs above and below the circles (see Figure 4.12 – all negative motifs highlighted in grey), and two smaller circular areas, in which motifs were inscribed. When three circles are placed in a circular plate as with the design of the Mayer mirror, three negative areas are created, two fins and a cusp. Stamford Hill I also uses a similar method of dividing the plate area (see Fox 1958: Pl. 56b). The majority of mirrors use a ‘lyre-loop’ to divide the plate

28

Some designs employ a ‘lyre-loop with flanking coils’ to section off the design-field (see MacGregor 1976: 'grammer of ornament'). Examples include Bromham, Chilham Castle, Desborough, Dorton, Holcombe and Portesham. Two examples have been selected to explain these more complicated designs: the Birdlip and Portesham mirrors. Figure 4.15 details the sequence of design for the Birdlip mirror. Unless otherwise stated, in the diagram negative motifs are shaded light grey. Where a new negative motif is created in the sequence it is initially shaded dark grey. The first stage in producing the design was to draw the outline of a lyre-loop with flanking coils (A). The shape of the plate allows the positioning of a lyre of this type within the confines of the mirror plate. The lyre delineates five negative motifs towards the edge of the plate (B), four fins at the top and centre and a cusp at the bottom edge. Five other areas are delineated in the middle of the plate including three large tear-shaped and two smaller circular areas (B). Once the lyre was put in the rest of the plate area was divided off using varying methods. The tear-shaped areas were sub-divided using smaller lyre-loops (C) creating more negative motifs (D), further dividing off space on the mirror surface. With the addition of four positive fin motifs (E), the outline of two circles and a trumpet motif are created in the two bottom tear-shaped areas (F). In the third tear-shaped area a further lyre-loop is placed in the bottom half (G). The addition of two positive trumpet motifs (H) creates a negative trumpet-void (I). With the addition of four more positive fin motifs (J) a further two negative trumpetvoids are created (K). At stage K we can see that the basic design of the mirror has been mapped out. The dark grey

Figure 4.9 A diagram to explain mirror composition

29

Figure 4.10 Fox’s (1958: fig. 62) outlines of mirror design combination of fin and armadillo positive motifs (L, M and N) (explained further below for the example of the Desborough mirror). At stage N the design of the mirror looks complete as it broadly conforms to the design rules. However, one extra layer of complexity was added. The two trumpets in the top half of the bottom two tear-shaped areas were further sub-divided, creating two fin-shaped negative motifs (O, P and Q). There is a possibility that the person who produced the design on the Birdlip mirror and understood the design rules so perfectly may not have been the same person who inscribed the decoration. The areas highlighted in dark grey at stage Q represent areas of unused-space. With the exception of the areas at the top of the mirror, the other

Figure 4.11 Six different outlines of mirror design areas highlighted at L show where the design is incomplete. As outlined above (and will be explained in more detail below) the use of lyres has the effect of creating circular areas, or roundels, in mirror design. We can see five such areas on the Birdlip mirror (K). From this stage onwards the particular decisions associated with the design are less easy to explain. In the case of Birdlip, roundels are filled with a

30

Figure 4.12 Dividing the design-field into smaller units

Figure 4.13 Design of the Pegsdon mirror areas of unused-space are created because the inscribed lines of the positive motifs do not touch one another. If these lines were touching they would frame perfectly formed negative motifs. In the design then, you can see the intentions of the ‘designer’ but the person who inscribed the decoration did not understand the importance of framing negative motifs within the design. By definition that person

Figure 4.14 The sequence of Aston mirror design

31

cannot be the same individual that conceived the design. The sequence in which the Holcombe mirror was decorated uncovered by Lowery et al (1976) does not correspond to the way in which I suggest the designs on the kidney-shaped mirrors were created. This suggests a separation in the design process and inscribing the decoration onto the plate. Savage & Lowery suggest that the

Figure 4.15 Construction of the design of the Birdlip mirror

32

Figure 4.16 Sequence of design of the Portesham mirror back design process was ‘organic’ because of the steps that were taken to inscribe the decoration on the Holcombe mirror plate. Designs on some mirrors were so complicated they must have been planned fully before they were inscribed; the preparation of the design and inscribing the design were two different processes. The decoration on the Holcombe mirror has not been fully completed. The outline of the design was inscribed but hatching of motifs is unfinished. One interpretation of this evidence could be that the decoration outline was inscribed by one person and hatched by another. This raises the intriguing possibility that mirror plates were made and an outline for the decoration marked out. The plates were then passed on to be finished elsewhere by someone else. This may explain, for example, why there is a consistency in the overall pattern of some mirror decoration but variation in hatching and the form of mirror handles. Figure 4.16 shows the sequence of design for Portesham. Like Birdlip the framework for the design is a lyre with flanking coils (A). This marks out four fins and a cusp at the fringes of the plate (B). Lyre-loops were placed inside the three tear-drop shaped areas created by the placement of the original lyre (C). These lyre-like shapes frame two fin motifs and two circles within each tear-drop shaped area (C). The addition of two fins within each tear-drop area (D)

33

further reduced the surface decorated. With the clever addition of a further circle (E) and two more fins (F) a negative trumpet-void motif and a negative stemmedpelta within each tear-drop-shaped area were created (G). The design is further complicated with the addition of smaller circles within the existing circles in the design (H) to create negative crescent motifs (I) and two circular areas within these circles (J, K and L). The design of Old Warden I and Great Chesterford work differently. They are not organised into an overall framework of a lyre-loop or within circles, instead designs appear to be random. In order to determine how these more diffused designs were created the negative motifs on Old Warden I were ranked in terms of their coherency (see Figure 4.17). The most coherent motifs on Old Warden I are located on the left-hand-side of the plate circling around to the top. The next most coherent motifs are located on the right-hand edge of the plate (motifs filled black (D)) and the least coherent are located in the middle (motifs filled black (E)). This suggests the decoration was applied in a spiral shape starting at the left-hand edge working its way around the rim of the plate and finishing in the centre (B). The idea that a spiral was used as the framework for the design may not look so unlikely if we examine the modern example of Matisse’s Snail (1953). Matisse constructed a collage by overlaying

Figure 4.17 Design of the Old Warden I mirror pieces of paper of different colours and shapes over the outline of a snail. Examples of Roundel Construction At each stage in the construction of mirror design, the design-field is divided into ever smaller, more readily comprehensible units, which obey the specified design rules. This process continued until a series of small circular units, or roundels, were isolated. How roundels were designed varies greatly from mirror to mirror and the exact reasons for placing particular motifs in certain positions within a circle will remain in the minds of the people who created designs. If we take an example from the Desborough mirror we can see how a fin and armadillo motif can be combined to fill a circular area (see stage ‘3’ on Figure 4.18). An armadillo shape is added to a fin and an additional fin is joined to it end-on-end. The fins are then curved-around to form a simple roundel. The design of Desborough was further complicated with the addition of two more fin shapes and a circled tricorne, forming a keeled-roundel (see stage ‘4’ on). A roundel can be created by combining identified mirror motifs in specific ways as on the Portesham mirror (see Figure 4.19). If two cusps are put together end-on-end they form a circle with two empty circles inside it; or an adjoined double cusp (ADC) (see stage ‘1’ Figure 4.19). A larger circle was drawn around the ADC, with the bottom edge of the ADC touching the bottom of the circle to create a simple

34

roundel. Two roundels of this type, one smaller than the other, can be encompassed within two fins creating the oval-shaped roundels found on Portesham (see stages ‘4’ and ‘5’ Figure 4.19). 4.6 UNUSED SPACE Appreciation of both positive and negative components is of key importance to mirror design. It is this balance between positive and negative components that Fox and Jope judged when they compared the decoration of different mirrors. Those designs that conform best to the design rules, outlined here, are those that are seen as the most ‘successful’ by Jope and Fox. However, the design rules aren’t explicitly defined by Fox and Jope; instead their definition of design success is based on their own aesthetic tastes and judgement rather than rigorous analytical methods. Moving away from the idea of ‘success,’ criteria to compare the designs of individual mirrors were established and unused space quantified. As can be seen in Table 4.1 the method of quantifying unused-space involved counting the number of positive and negative motifs on each mirror in addition to counting areas of unused-space. All of these areas were then totalled and unused-space calculated as a percentage of the total area of the design-field (represented by the total number of all demarcated spaces on the mirror surface: negative motifs + positive motifs + unused-space). Whilst this does not give an exact representation of the percentage area of the design-field occupied by unused-space, this gives a more

Figure 4.18 Desborough roundel accurate representation of how balanced a composition is. This measure allows designs to be compared to one another on an equal footing. It also helps identify which mirrors do not conform to the identified pattern of a balanced composition. By distinguishing how positive and negative mirror motifs are employed to create the overall design, it is possible to identify the depth of the designer’s understanding of the principles of design composition. Where negative motifs are not present, the positive motifs have clearly been copied without a true understanding of the composition rules of mirror-design. The results of this analysis indicate that the decoration on three mirrors most successfully fulfil these criteria: Aston, Desborough and Pegsdon. On the Aston mirror 100% of the mirror plate surface is composed of interlocked positive and negative motifs leaving no unused-space. On the Desbourough and Pegsdon mirrors it is 99%. The majority of the other mirror-designs fall between 70% and 90% of the surface of the mirror plate being filled with demonstrable and consistent positive and negative motifs. 4.7 COMPLEXITY The fourth and final stage of the analytical process is complexity. Complexity is defined by the number of individual motifs in each design: the most complex designs contain the most motifs. The design of the Aston mirror contains only twenty or so motifs. The designs on the Desborough and Pegsdon mirrors, on the other hand, are decorated with more than 200 separate, identifiable motifs with very few areas of unused-space. Simplicity of form does not necessarily imply simplicity of meaning (see Le Bon 1995). However, it would have required a high level of skill to

35

produce a complex mirror-design containing many individual motifs, which obey the specified design rules. The two mirror-designs that do not conform to the dominant pattern are from Bryher and Nijmegen. The Bryher mirror is probably the earliest decorated mirror (Johns 2006). Although it is similar in size and shape to the mirror from St. Keverne the positive motifs that are distinguishable on its plate are very different from those present on the designs of the other decorated mirrors. Therefore the decoration on Bryher bears no resemblances to any of the other decorated mirrors and it must be assumed that it was decorated outside of the particular context in which the majority of mirrors were decorated. The Nijmegen mirror is interesting because it was found in a Roman grave in The Netherlands (Dunning 1928) that is dated to after AD 100 (Isings 1957; Lloyd-Morgan 1981; Fitzpatrick 1997a; Price & Cottam 1998); at least 25 years after most of the other mirrors were deposited. It is possible that the Nijmegen mirror was made and decorated by people separated by time from the context of production of the majority of the decorated mirrors. We should see the decoration on the back of Nijmegen as an example where positive motifs seen on other mirrors have been borrowed and applied to a mirror back without a full appreciation of mirror decoration as a whole. Figure 4.20 shows how the design of the Desborough and Aston mirrors consists of a balance of positive and negative motifs. In contrast although the design on the Nijmegen mirror uses ‘mirror-style’ positive motifs, the negative spaces are largely ignored in the layout of the design. The result is an

Mirror

Number of Positive Motifs

Num ber of Negative Motifs

Unused Space

% of Mirror Surface Used

Aston

11

10

0

100

Billericay I

15

9

2

92

Birdlip

64

77

7

95

Bromham

30

16

2

96

Bryher

2

0

-

-

Bulbury

-

-

-

-

Chettle

44

23

14

83

Chilham Castle

6

5

2

85

Colchester I

14

9

0

100

Colchester II

-

-

-

-

Desborough

130

94

3

99

Disney

7

-

-

-

Dorton

16

13

3

91

Gibbs

21

17

6

86

Great Chesterford

40

24

8

89

Holcombe

82

68

10

94

Latchmere Green

41

26

5

93

Mayer

19

48

4

94

Nijmegen

38

11

35

58

Old Warden I

30

13

3

93

‘Oxford’

84

84

7

96

Pegsdon

104

109

1

99

Portesham

40

40

0

100

Portland II

3

-

-

-

Rickling

-

-

-

-

Rivenhall I

5

5

1

91

St. Keverne

10

12

10

69

Stamford Hill I

22

10

9

78

Figure 4.19 Portesham roundel tastes for symmetry had an effect on the ‘fold-over’ decoration of mirrors. He also suggested that despite the Roman influence of symmetrical designs, every now and then decoration on Iron Age objects “…slip[ped] back into an asymmetry more agreeable to their ethos” . Mirror designs are never perfectly symmetrical. A design that at first looks symmetrical is revealed as being very slightly asymmetric. Mirror design may reference the symmetry of contemporary Roman designs, but designs are also free-flowing and ever so slightly asymmetrical, or the line of symmetry is altered. When asymmetry is utilised in mirror-design it is not a slip, but a very deliberate act. Multiple meanings can be uncovered, for example, the fold-over symmetry of mirror pattern may have been a reference to the reflective properties of the plate; each half of the decoration is a mirror image of the other. Not all mirrors designs possess vertical symmetry. Equally, the Nijmegen mirror has vertical symmetry, but the overall design does not obey the design rules of mirror decoration. Therefore vertical symmetry may have been important to the designs of some mirrors; however this was secondary to the importance of the overall design.

Table 4.1 Results of analysis of mirror decoration showing % of used space unbalanced composition and the creation of extensive unused-spaces in the design-field. 4.8 VERTICAL SYMMETRY As has often been discussed before, the overall design on some mirrors is arranged so that it has vertical symmetry; each half of the decoration is a mirror image of the other. The pattern of some mirrors, decorated with a more ‘diffused’ design like Great Chesterford, is arranged so that the line of symmetry is at a slight angle (Fox 1958: 114-5). Fold-over symmetry can also be seen on a number of other Iron Age artefacts, such as the Battersea shield (Fox 1958; Stead 1985: 27). Fox (1958: 94) has suggested that Roman

36

4.9 SUMMARY Using a series of analytical techniques, evidence has been provided to suggest that mirror decoration is consistently constructed using a small number of positive motifs. These positive motifs are organised so that they frame negative motifs within the design-field. The composition is structured in such a way that unused-space is minimised. Design complexity can be increased if more motifs are added. A mirror decorated without equal reference to negative spaces was designed by a person operating outside of the particular production context in which the majority of decorated mirrors were made, with incomplete knowledge or understanding of the principles of design composition. The majority of mirrors are decorated with designs that conform to two of the analytical criteria, but not all. For example, the design on the Aston mirror minimises unused-space and balances positive and negative components but it is not complex. The design on the Birdlip mirror is complex and balances negative and positive components but it also has some areas of unusedspace. The two mirrors that conform best to all of these

Figure 4.20 Mirrors viewed in negative (a. Nijmegen, b. ‘Mayer’, c. Aston, d. Desborough) criteria are Pegsdon and Desborough. This result confirms Fox’s and Jope’s assessments of mirror decoration based on their own aesthetic judgements but crucially does not reinforce their stylistic sequence. Some mirrors that conform best to mirror design rules, such as Aston or Pegsdon, date

37

to the 1st-century BC and should therefore appear very early in the stylistic sequences created by Fox or Jope. The process of decorating a mirror can be divided into two main tasks: creating the design, and inscribing it. These two tasks were not necessarily performed by the

same person. The rules that mirror design follows were not necessarily fully understood by the person who inscribed the design as has been shown through the example of the Birdlip mirror. A whole series of complicated relationships can be envisaged concerned with how to decorate and the process of decorating a mirror. These relationships and the design of mirror decoration would have had a significant effect on the meaning of a mirror and the future trajectories of its life. 4.10 WHY DECORATE? THE ‘EFFECT’ OF DECORATION This chapter will end with a consideration of the question why were mirrors decorated? Beyond possible references to nature (e.g. Stead & Hughes 1997: 17; Jope 2000: Pl. 185) the motivations behind decorating a mirror have never been questioned. Mirror decoration has been interpreted as passive and merely pleasing to the eye. As was demonstrated in Chapter 3, we cannot disengage decoration from the object and the particular social contexts in which they were made. Mirrors were decorated for specific reasons, these were intimately related to the material properties of the object, its role within society and the particular social qualities ascribed to decoration, which are socially embedded (see Campbell 2002; Coote 1992; Gell 1999). Mirrors were not decorated to ‘fill an unoccupied space’ (see Fox 1958; Jope 2000); they were decorated for specific reasons within a specific social and cultural context. As Gosden and Hill (2008: 8) indicate, it is not possible to investigate motivations or meaning, but what we can do is look at the social and sensory effects of decorated objects; we can examine mirrors in action. Looking specifically at intricately carved Trobriand canoe prows, Gell (1999) argued that art was a special form of technology; artworks display such technological virtuosity that people find it difficult to comprehend. Particularly skilled carvers were seen to have a connection with cosmological forces and acted as a conduit through which these forces were endowed onto canoe prows. Decoration of canoes played a critical role in Trobriand relationships by deceiving exchange partners into giving up their valuables cheaply (Gosden 2001a: 164; Thomas 2001: 3). According to this argument, complicated and incomprehensible mirror designs could have acted to ‘enchant’ the Iron Age observer. Applying Gell’s theories of art to sword-scabbard decoration, Giles (2008) linked it to the effectiveness of the sword and the prowess in battle of its owner. Similarly, decorated horse-harnesses could be argued to add to the power of the chariot and its riders. Here decoration is intimately related to the efficacy of the object and directly affects the social performance of human actors. Who Understood the Design Rules? An important question relating to mirror design is: who in society understood it? Critics of Gell (e.g. Campbell 2001; Gosden 2001b) have pointed out that people producing ‘art’ are not doing it merely to master a particular technology; there is also engagement with an audience. ‘Artworks’ were produced for an audience (Campbell 2001: 122, 134). Not everyone in a particular society necessarily has detailed

38

knowledge of the exact meanings behind particular designs but assessments are culturally and socially specific and grounded in tradition. Even the craftsperson/s decorating an object may not understand the true significance of the designs they are producing. The inscribed lines on the Birdlip mirror frame unused-spaces, where slightly different positioning of the lines would have framed negative motifs. This suggests that although the person who created the design was fully aware of the design rules, the person who inscribed the design was not. The fact that the design was inscribed twice and that the second inscription meticulously followed the errors of the first inscription underlines this point. Campbell (2002: 60) has shown that the craftsmen who produce carvings placed on outrigger canoes in the Trobriand Islands, Papua New Guinea, are not necessarily the people who possess the knowledge of the full significance of what each individual element of a carving means. What is important is that someone in that society understands the significance of designs and that designs are executed in the ‘correct’ way (see Laing & Laing 1992: 129). There are clear temporal and regional trends in the ways in which mirror design is constructed. Within these regional and temporal contexts there are examples of designs which follow design rules better than others. The obvious examples are the Aston and Pegsdon mirrors in the southeastern group and the Desborough mirror in the ‘western’ group. We cannot be certain how design rules were interpreted but the consistency in which designs follow these rules, and the fact that some mirrors follow them better than others, suggests that mirror decoration had significant meaning and that knowledge of these meanings was not homogenously distributed. The intention here is not to reveal an Iron Age aesthetic. As already argued, we need to move away from the idea that mirror decoration can be assessed through a western cultural lens. However, within an Iron Age context, decoration was judged: certain qualities in design, or the clarity of the inscribed lines, made one design more able to perform specific tasks. As already stated, Iron Age people did not have the benefit of seeing a large number of decorated mirrors like we do. It follows that if judgements were made they were set against the context of more general principles. As outlined in the previous chapter, a number of individuals may also have been involved in the manufacture of mirrors (see also Sharples 2008: 206). Processes such as casting the handle, hammering the plate, and composing and engraving the design, provide further arenas for objects to be judged or understood by different audiences. The ‘Sensory Environment’ Gosden (2001a) has argued that aesthetic judgements are in part derived from the sensory impact of objects. Each culture develops its own ‘sensory environment’, where some senses or sensual effects are valued above others (ibid: 166). We do not know what the Iron Age ‘sensory environment’ was like. It is probable that it varied greatly from place-to-place and through time. Through artefacts such as the Deskford Carnyx (Hunter 2001) we get an idea of a rich sonic environment. However, this is only a

tantalising glimpse as objects like carnyxes are likely to have been used in very specific contexts. Early Medieval Irish texts hint at a rich oral tradition. However, the primary sensory environment we encounter through the archaeological record is visual, through Celtic Art. Colour is an important visual element with clear social and cultural correlates (Herbert 1984; Hosler 1994; 1995; Jones 2002b; Jones & MacGregor 2002a; Jones & MacGregor 2002b; Saunders 2002; Scarre 2002). Recent research has shown how important the colour of objects was at the famous site of Snettisham, Norfolk. The colour of torcs found in various hoards at the site was manipulated through surface enrichment of the metal (Duncan Hook & Nigel Meeks pers.comm). Torcs and other items from Snettisham vary greatly in colour from silver to deep gold and copperyred. Some hoards at the site contained a mixture of different coloured torcs (see Stead 1991e). This evidence indicates that some metalworkers knew how to manipulate the colour of metals and objects of different colours were selected for inclusion in the same hoard. Fitzpatrick (2007: 344-5) has recently examined the importance of colour in the context of middle Iron Age British display shields. Like many decorated objects made using comparable alloys they were originally a bright, shining, golden, yellow colour. Although mirrors were probably made a few hundred years after display shields they are made from similar alloys and were probably a very similar colour when polished. The maintenance of the particular alloy that was used to make the majority of bronze objects during the British Iron Age is therefore striking. The material properties of this metal make it tough and resistant to scratching, they also produce a distinctive colour. This colour may have had long lasting significance. Shields and swords, as well as some brooches, were adorned with red coral and glass (Champion 1976). Fitzpatrick (2007: 345) has suggested that the colour red in this context may have been associated with blood and therefore aggression. He juxtaposed the use of the colour red to adorn martial equipment with blue-bead necklaces, which were worn by women. Some mirror handles, such as Holcombe, were ornamented with red glass. At this time horse-harness equipment was also adorned with red glass. The juxtaposition of martial/male/red and female/blue is less clear in this context. However, what is apparent is that red and the golden bronze colour had a clear (and probably changing) cultural significance over a long period and the colour red was sometimes incorporated into the decoration of Celtic Art objects. Visual qualities and context Wells (2008: 129) set out to examine ‘how the visual qualities of things were used to elicit responses in people’ through contextual analysis. Rather than searching for meanings, Wells argues that the placement of ‘visual objects’ in graves, ritual deposits and the landscape can provide clues about what they were used for and how people reacted to them (ibid). It has long been suggested that water was culturally and socially significant in Iron Age Britain and Europe and that activities associated with water may have been intimately connected to spiritual beliefs (Ilkjaer

39

& Lonstrup 1982; Fitzpatrick 1984; Stead 1984; Green 1994; Webster 1995; Field & Parker-Pearson 2003). Weapons and other objects were deposited in east-flowing rivers in England (Fitzpatrick 1984; Bradley 1990). Objects and even people were deposited in bogs in England, Ireland, Scotland and other parts of Europe, particularly Denmark (see Asingh 2007; Joy 2009b). Specific sites were also the focus for deposition in water over an extended period (Fox 1946; Green 1991; Webster 1995; Parker-Pearson 2000; MacDonald 2007b). Many, although by no means all, of the objects deposited in watery contexts were made from shiny metal; it is possible therefore to draw analogies with the reflective properties of water and these metal objects specifically selected for deposition in watery contexts. Mirror deposition can also be argued to be related to water (Giles & Joy 2007). This association will be addressed further in Chapter 10. The visual experience A common element of the visual effect of Iron Age objects made from sheet-bronze is their reflective properties. The bronze used to make them was capable of maintaining a good reflection. This is a prerequisite for an object like a mirror, but for display shields it is less important unless reflection, or capturing light, held social or symbolic significance. The golden yellow colour of the bronze also affected the qualities of reflection, perhaps referencing the colour of the sun. Spratling (2008) has recently questioned the visibility of Celtic Art. Many designs given great prominence in classic works on Celtic Art (e.g. Jacobstahl 1944; Jope 2000), would have only been visible close-up. Only those people given privileged access to decorated objects could fully appreciate the intricacies of design. Others may have caught glimpses of decoration from a distance. The experience of a decorated mirror is very different depending on whether it was viewed from close up or far away. Close up the quality of inscription is important. From further away this is less visible. Using a reconstruction, if viewed from a distance light reflecting off a decorated mirror is distorted by the inscribed pattern, rather like ripples on the surface of a pool of water. By inscribing certain areas and leaving other areas blank this effect is enhanced. Effects also alter depending on the quality of light. The flickering of firelight plays further tricks. It should be emphasised that Celtic Art formed only one element of the sensory environment. Judging from the surviving material evidence, the majority of people in Iron Age Britain must have encountered these objects rarely, if at all. They lived their lives on farms, or in small villages. They used often very drab pots to prepare and serve their food (see Evans 1989). We can only guess at other important aspects of the visual ‘sensory environment’, such as dress and personal appearance (but see Hill 1997; Jundi & Hill 1997). However, based on the evidence available to us, Celtic Art really does standout.

Summary The main evidence for the Iron Age sensory environment that has survived archaeologically is visual: Celtic Art. The colour of bronze objects was significant. In addition to colour, reflection of light, or shininess, was also probably important. Many objects, especially those made of sheet bronze, were highly polished, capturing and reflecting light. The composition of the alloy used to make these objects made surfaces tough and resistant to scratching. Reflections were also distorted by inscribed or raised decoration. 4.11 WHY MIRRORS WERE DECORATED The people who commissioned, cast the handle, hammered the plate, and composed and engraved designs on mirrors were implicated in the final form of the artefact and how it was perceived by different audiences. In terms of social relations, decorating a mirror increases the number of production steps, potentially increasing the number of individuals involved in mirror manufacture. It also accentuates the object’s visual effect. This is extended to makers and users. By decorating a mirror with intricate designs only those with privileged access to the object were able to fully appreciate the design. Others were only able to catch glimpses of decoration. At a distance the effect of pattern on reflection becomes more important, particularly hatching. By decorating a mirror the reflective surface on one side of the plate was broken up. This was a conscious decision. The pattern, quality, depth of engraving, and choice of hatching, affected how reflection was altered. Light and reflection were directly manipulated through decoration. The yellow colour and shininess of the metal mirrors were made from were socially significant. One basis for ‘judging’ mirror decoration could have been how well a particular pattern was perceived to manipulate the reflection of light. How light was manipulated could affect how well a mirror was able to perform the tasks and act in the particular social realm it was made for.

40

Both sides of a mirror are intimately bound; mirror decoration explicitly refers to the reflective surface of a mirror through positive and negative imagery. A number of explanations for this can be suggested. Decoration could have acted to make a mirror better able to capture and reflect light by combining positive and negative elements into one pattern. Mirror decoration in this instance acts to reemphasise and reinforce the primary function and most powerful property of a mirror: its ability to reflect light. Decoration on one side of the plate could also have acted to ‘protect’ onlookers from the reflection. Interlocking positive and negative imagery in a complex pattern could have acted to ‘capture’ light, counteracting the reflective qualities of the plate, leaving only one side with a visible reflection. Neither of these interpretations is mutually exclusive. By blocking off reflective properties on one side of the plate, reflection on the other side could be seen to be more powerful. We could therefore see mirror decoration not as the passive, decorative back to the reflective side of a mirror, but intimately involved in capturing and manipulating reflected light. When the meaning of abstract Celtic Art has been considered in the past it has most often been from the perspective that particular motifs must have stood for something (Aldhouse-Green 1996; Jope 2000: 114-5; Fitzpatrick 2007: 351-2). The methods employed to study Celtic Art motifs have isolated them from the objects they ornament and the social context of the object. This is one of the reasons why previous researchers have been so keen to view Celtic Art motifs as being representative of something else. In the case of mirrors, the decoration cannot be abstracted from the artefact. Mirror decoration is integral to the efficacy of the mirror it ornaments. The decorated and reflective faces of the mirror plate metaphorically refer to one another and act to reinforce the primary function of a mirror: to reflect light. The motifs used to decorate mirrors are therefore not representational of anything other than ‘mirrorness’.

CHAPTER 5 ANALYSIS OF USE/WEAR/REPAIR 5.1 INTRODUCTION To examine how they were used in life mirrors were visually examined looking for signs of wear, polishing and repair. The term ‘use’ relates to physical indications which may indicate how mirrors may have engaged in the

Figure 5.1 Common areas of wear/polishing/repair

41

performance of social relations. The aim was to be as comprehensive as possible but there were a number of constraints in terms of availability and access to mirrors. For example, a large number could not be examined because they are now lost (e.g. Arras II), destroyed (e.g.

Wear to the Terminal Ring

Stress Marks between Plate and Handle

Decoration underlying Plate or Rim

Polishing on the Handle

Evidence of Repair

X

X

Mirror

Bronze/Iron

Arras

Iron

Aston

Bronze

X

Balleybogey

Bronze Handle

X

X

Billericay I

Bronze

X

X

Birdlip

Bronze

X

Bridport

Bronze

X

Bromham

Bronze

X

Carlingwark

Bronze

X

Chettle

Bronze

Colchester I

Bronze

Colchester II

Bronze

Desborough

Bronze

X

Dorton

Bronze

X

Holcombe

Bronze

X

Ingleton

Bronze

X

Latchmere Green

Bronze

Llechwedd-du

Bronze

Old Warden I

Bronze

Pegsdon

Bronze

Portesham

Bronze

Rivenhall I

Bronze

X

St. Keverne

Bronze

X

X

X

X

X

Parts of Rim Missing

X

?

X X X

X X X X X

X ?

X X X

X

Table 5.1 Summary of the results of analysis of wear/ polishing/repair three mirrors from Stamford Hill, Plymouth) or in private hands. Some others, especially those made of iron, are also in such poor condition that examination was not thought worthwhile (e.g. Wetwang Village, Lambay Island). Eventually a list of mirrors was compiled and a schedule arranged with museum curators to examine them. Research was undertaken from 2004-7. The aim was to examine mirrors found from a wide variety of contexts. However, inevitably the mirrors that were eventually analysed were more likely to have been deposited in specific contexts, particularly graves, as they are more likely to survive in good condition to the present. Of the 32 mirrors examined 18 were recovered from graves, 2 from watery contexts, 6 from settlement contexts and 6 finds have no recorded context. Each component of the mirror was examined in detail, measured and photographed. Particular attention was paid to areas of wear and polishing and the exact position of these areas was noted and sketched. Evidence of repair was also noted, as well as any evidence that could suggest how components of the mirror were made and the sequence of construction (see Chapter 3). Following an extensive literature search, three were highlighted as being particularly susceptible to wear, repair and

42

stress: the terminal loop of the handle; the junction between plate and handle; and the area of the mirror plate immediately adjacent to this join. Special attention was paid to these areas. The next sections of this chapter will provide a concise summary of the results of this analysis. A detailed review of each mirror is included in Appendix H. 5.2 RESULTS 22 of the 33 mirrors examined showed signs of wear, polishing or repair. A summary of the results of this analysis can be seen in Figure 5.1 and Table 5.1. It indicates that there were a number of common areas of wear/polishing/repair. Wear to the Terminal Ring 12 mirrors had signs of wear on the terminal ring of the mirror handle. A further mirror handle showed possible indications of wear. This represents over a third of the mirrors examined and suggests a possible common form of usage. The wear was most frequent on the bottom part of the ring on the inside edge and was often restricted to

one side of the handle. A number of the terminal rings also showed signs of flattening and thinning. There was a range in the level of wear and some handles were extremely worn and thinned, such as the handle of Billericay I. To achieve this level of wear significant abrasion must have occurred over an extended period of time. Stress Marks between Plate and Handle Two mirrors, Desborough and Bromham, have visible stress lines on the surface of the plate in the area where the handle and plate join. This indicates that, as would be expected, some mirrors were held by the handle when in use. It is interesting to compare these two plates with the Birdlip plate, which shows no sign of damage in this area. The level of damage to this area of the plate may be a useful index of how long and how intensively a mirror was used. At least ten mirrors have been discovered which are broken along this axis, also adding weight to the interpretation that mirrors were held by the handle, thus causing stress on the join between handle and plate. In fact when a mirror breaks along this line of tension it very often represents the final stage in its life and a number were deposited in the ground, in this broken state. However, it does not necessarily represent the end of the life of a mirror: Bryher probably broke along this line of tension and was repaired in antiquity by removing the lower section of the plate and reattaching the handle (Johns 2006). The decoration on the bottom of the Chettle mirror has been worn away. It was buried without a handle and the wear may be a result of it continuing to be used after the handle broke off (A. P. Fitzpatrick pers.comm). Decoration Underlying Handle or Rim Some mirrors have decoration obscured by the rim or by the arms of the handle. This demonstrates that the plate was decorated before the handle and the rim were fixed to the mirror. This seems like an obvious point as it would have been most practical to decorate the plate without the obstruction provided by the plate and handle. However, the fact that only very small areas of decoration are obscured shows that there was a degree of planning and coordination involved in the construction of the mirror and the making of the various components; the person decorating the plate knew the form of the handle before they undertook the decoration of the plate. This fact is underlined by the discovery of ‘test’ decoration underneath the handle of the Birdlip mirror. Polishing on the Handle One of the most interesting discoveries was of areas of polishing on the handle on nine mirrors. The areas of polishing were located on one side of the handle at the points that protruded furthest away from the main body with the objects lying horizontally. The most common areas of polishing were the nodal points which help link the loops of the grip of a handle and also the handle arms. The polishing indicates that these mirrors had been lying flat predominantly on one side.

43

Evidence of Repair Six of the mirrors examined showed signs of repair. Some other repairs are reported in the literature, but not examined in my analysis (e.g. Johns 2006). This evidence shows that some mirrors were in circulation for a time before deposition. Most of the repairs were on the mirror handle or at the junction between the handle and plate. This further illustrates the main weak point in the design of mirrors and provides more evidence that they were held by the handle in usage. Missing Edging Two of the rimmed mirrors had parts of the edging missing. It is unclear whether this loss occurred in antiquity, or was the result of post-depositional damage. If ancient, it shows that the edging had served one of its purposes – to protect the vulnerable edge of the plate. It was probably very difficult to repair the edging; if one part of it broke it would have been necessary to cut a section away and replace it in whole or part. This may show that small amounts of damage were thought insignificant. 5.3 INTERPRETATION The results of visual examination of mirrors are highly significant. A large proportion of the mirrors examined show signs of wear and polishing, illustrating that many were ‘used’ before they were discarded. 12 had wear on the terminal loop of the handle. Two had visible stress marks on the plate close to the point where the handle was joined to it. A further six showed signs of repair. Mirrors were not just made for show or for deposition; they were actually used to perform particular tasks before they were deposited. Clues as to how mirrors were used are indicated by the lines of stress on the plates of the Desborough and Bromham mirrors, as well as the 10 or so mirrors which are fractured at this point of weakness. These wear patterns suggest that mirrors were held by the handle and that during usage the plate was unsupported, putting stress on the area where the plate and handle join. As has been suggested a means of measuring how intensively a mirror was used could involve an assessment of stress in this area: the more signs of stress at the junction of the plate and the handle the more intensively the mirror was used. The terminal loop of 12 of the mirror handles examined showed signs of wear. Fox (in Fox & Pollard 1973) has suggested that the decoration on some mirrors makes more sense if it is viewed upside down. She hypothesised that when they were not in use mirrors were suspended upside down and used the idea of suspension to explain the noted wear on the terminal loop of the Holcombe mirror. However, the observed wear does not appear to have been caused by rope or fibre used in the suspension of a mirror because it does not have a polished surface. Instead it is characterised by abrasions and striations and some of the extreme wear may have been caused by metal rubbing on metal. It is notable that a number of the mirror handles, such as Portesham and Pegsdon, were found

with brooches attached to their terminal loops (Chapter 9). A number of mirrors were also wrapped with some sort of organic covering, such as textile or a leather pouch, so it is possible that the striations on the terminal loop of some mirrors was caused by a brooch, or some other kind of fastener, which was used to secure a protective covering around the mirror. The polishing noted on the nodal points of one side of nine of the mirror handles examined could result from museum display. The majority of the mirrors with polishing in these areas were 19th-century finds and have been on museum display for over a century. Mirrors are most often displayed lying flat and this could explain why polishing was most often found on only one side of a mirror handle. In the light of the evidence from the wear on the terminal rings of the handles, another explanation for the polishing could be that this was caused by repeated rubbing on the covering of the mirror, but it would be hard to explain why this was only prominent on one side. Finally, the huge variety within the group should be stressed. Mirrors originally had a handle and plate. However, having examined over half of the known mirrors, it is clear just how much they vary in form and size of the

44

handle, decoration, size and shape of the plate, combinations of materials and the presence or absence of edging. 5.4 CONCLUSIONS The key conclusion to be drawn from this chapter is that mirrors performed in social relations before they were deposited: signs of wear indicate social usage. Visual examination of over thirty of the mirrors has shown that they were not just made for show. Mirrors were also clearly valued in some way as some were repaired to prolong their social lives. This implies that mirrors were not just made for deposition; they were made to act in and formulate social relations. Individual mirrors did not necessarily act in the same way, or for the same purposes, but they clearly had meaning and significance that extended beyond production and deposition. As this analysis shows, the period and intensity of usage varied. Importantly, mirrors would have been active in social relationships in the period between production, when many of the meanings tied to individual mirrors were set, and deposition.

CHAPTER 6 MIRROR FORM A mirror is formed of a highly polished, reflective plate of circular or roughly circular shape and a handle. Some mirrors are also edged. Some researchers have argued that the form of Iron Age mirrors references Etruscan, Greek or Roman mirrors (Chapter 1). However, in this chapter it is argued that the form of British mirrors is integral to their meaning in the particular context in which they were made and used. Tied to the ideas developed in Chapter 2, the specific form of mirrors and the shape of the different components are linked to social relationships and are integral to the biography of a mirror: the specific form and shape determines the manner in which relationships can be structured around them (see Strathern 1988: 176). Whether or not the idea of the artefact form originates from elsewhere, the context of usage, the lives led, and the social

Figure 6.1 Mirror components refer to other items of Later Iron Age material culture

45

spheres they operated and acted in, were intimately tied to and dependent upon local, regional and temporal contexts, as well as how it determined the structure of social relationships that could cluster around mirrors. In Chaper 4 it was demonstrated that mirror decoration represents ‘mirrorness’; acting to reinforce the reflective qualities of the mirror plate. In this chapter the notion of ‘mirrorness’ is further explored, it is argued that mirrorness is constituted in two ways, both structured by the physical form of the mirror: 1. Relations formulated through physical form, particularly the component parts. 2. Relationships constituted in use, particularly relations to the human body.

Consequently, this chapter will be in two parts. The first will explore how mirrors, through their physical form, reference other items of Iron Age material culture; affording them with the potential to be interpreted in multiple ways by connecting them to the wider material world. The second section will examine how the form of mirrors structured their relation to the human body. 6.1 MIRROR COMPONENT PARTS ‘REFERENCE’ MATERIAL CULTURE As outlined in Chapter 3 many of the meanings of artefacts are set at the time of production. It is argued that the form of the component parts used to make a mirror, deliberately reference other material objects present during the later Iron Age (Figure 6.1). Mirrors refer to and gather up different elements of a series of types of object forms, recognisable from the later Iron Age, and combine them into a single, multi-component object. Form was set out at the time of production. Meanings and associations were intimately related to the potential future life of a mirror because they determined how relationships could be structured around them. Mirror Handles A typology of mirror handles was created by Fox (1949). This typology was revised by Spratling (1972: 96-8) and is further updated in Appendix D. Mirror handles are separated into five main types. Type I, bar handles, are common on iron mirrors from East Yorkshire. Only one example of Type II ‘shaped’ handles is known from Balmaclellan. Type III ‘looped’ handles are found predominantly on bronze mirrors. Type IV, have loops incorporated within their grips. Like Type II, Type V handles may represent a type which is unique to Scotland and only two examples are known. In respect to this chapter we are interested in Type IV handles. Although mirror handles do not have a loop at both ends, they can also be seen to reference the design of horse-bits. The large terminal loop of mirror handles is similar to the loops at either end of a bit. The middle, looped section of the handle grip is often identical to the looped middle section of a horse-bit (Figure 6.2). The necessity for the mirror handle to be rigid contrasts with the movable parts of the horse-bit and similarities of form would not have been obvious when the horse-bit was being used. However, similarities in manufacture techniques and form could have been maintained in order to underline the linkage between mirror handles and horse-bits. An analogy can also be drawn between holding a mirror in the central part, or grip of the handle, and the way a bit is held in the horse’s mouth: both are clasped. Multi-link horse-bits have been found throughout Britain. They are contemporary with mirrors and date from the 4th-century BC to the end of the Iron Age (Palk 1984: 77-83; 1988), although the majority date to the 1st-century AD (Hutcheson 2004: 34). The mechanical requirement of a mirror handle to be rigid means that the association can only be metaphoric; however the similarity is there and would have been recognisable especially to the makers of metal objects. Techniques used to make common objects may have been adapted to make more unusual ones like

46

Figure 6.2 Comparison of mirror handle grip and horsebit mirrors; the design for one object was borrowed and adapted to make another. Changes in the design of mirror handles and horse-bits can be seen to echo each other. Like some mirror handles, some horse-bits have elaborations to the design inside the terminal loops of the bit (Palk 1984; Jope 2000: Pl. 276). So we can see similarity of form between horse-bits and mirror handles being maintained, with both object forms undergoing transformations. The multi-looped grips of Type IV mirror handles are also very similar in form to the grips of some tankard handles, for example, the tankard handles from Bulbury and Hod Hill, Dorset (Jope 2000: Pl. 227e & 230e). Corcoran (1952) even named one of his groups of tankard handles the ‘mirror group’. Mirror Plate Through the manufacturing techniques used to make mirror plates they reference a whole series of objects made of sheet bronze of similar style, including cauldrons, bowls, display shields (Stead 1985; Fitzpatrick 2007), various shield bosses such as the one found at Llyn Cerrig Bach (Fox 1946), sword scabbards and plaques, originally attached to other objects such as wooden tankards, as well as more enigmatic objects like the ‘bean can’ from the Wetwang Slack mirror burial (Dent 1985a) (Figure 6.3), and the ‘crown’ from the Deal burial (Parfitt 1995). Like mirrors other sheet-bronze objects are often decorated. The majority date to the later Iron Age and are therefore contempory with mirrors. Cauldrons, although they are not commonly found, are known in Britain from the late Bronze Age (Green 1998: 69) and Maiden Castle has been shown to have produced sheet metalwork, specifically cauldrons (Northover 1991a: 161). Swords and

scabbards are also found throughout the British Isles dating from the 4th-century BC – 1st century AD (Stead 2006). Some of these objects have even been found associated with mirrors although rarely so, such as the Wetwang Slack ‘bean-can’ (Dent 1985a), the bronze bowls found with the mirror from Birdlip (Staelens 1982), the bronze cup from the Colchester I (Lexden Grange) mirror burial (Fox & Hull 1948) and a bronze platter found with the Llechwedd-du bach mirror (Fox 1925). The analogy with cauldrons and other sheet bronze objects can be extended further when we look at the use of rivets to secure the attachment between mirror handles and plates, as rivets are integral to the manufacture of sheet bronze artefacts.

Figure 6.3 So-called ‘bean can’ from Wetwang Slack burial (after Dent 1985a) Mirror Motifs Identified on other Artefacts A survey of Iron Age decorated material culture has revealed over 100 objects decorated with one or more ‘mirror-style’ motifs (Appendix C). The survey involved systematic investigation of Fox (1958), Jope (2000), MacGregor (1976), Megaw & Megaw (2001), Parfitt (1995), Spratling (1972), Stead (1996) and recent finds on the Portable Antiquities database (www.findsdatabase.org.uk), looking for the presence, and noting the location on the object, of any of the mirror motifs identified in Chapter 4. The first thing to note from this survey is that relatively few objects are decorated with ‘mirror-style’ motifs (see Appendix C). None of the motifs identified in the survey are unique to mirrors; all the identified motifs can be found on other items of late Iron Age material culture. The motifs most commonly found were the fin and the small circled

47

motifs. The majority of objects are decorated with a single, or a very small number of motifs, although some of the objects like the sword scabbard from Isleham, Cambrideshire, are decorated in a similar way to mirrors with designs of balanced negative and positive components. These decorated objects are nearly all made of bronze. They cross-cut a large number of artefact classes, although usually only a single object, or a small number of objects in each class, was found to have been decorated with these motifs. The tankard from Carrikfergus has a bronze plaque, or apron, which is positioned above the uppermost fixing mount of the handle and is decorated with motifs found on mirrors (Jope 2000: 131, Pl. 229). A spearhead from the river Thames has bronze mounts which are inscribed with mirror motifs (ibid: 200c). A decorated bronze comb from Warwickshire has also recently been discovered. A number of bronze sword-scabbards were decorated with ‘mirror-style’ motifs including, Little Wittenham, Oxfordshire (Stead 2006: No. 86, Fig. 65), Meare Heath, Somerset (ibid: No. 88, Fig. 67), Hunsbury, Northamptonshire (ibid: No. 89, Fig. 68); Amerden Lock, Buckinghamshire (ibid: No 90, Fig. 69), Isleham, Cambridgeshire (ibid: No. 101, Fig. 75), Icklingham, Suffolk (ibid: No. 107, Fig. 78), Wood Burcote, Northamptonshire (ibid: No. 108, Fig. 108), and Bugthorpe, East Riding of Yorkshire (ibid: No. 176, Fig. 92). With the exception of Bugthorpe (Group E) these all belong to Stead’s groups C and D – long swords and scabbards with campanulate (C), or straight mouths (D) (ibid: 149-51). Group C swords and scabbards probably date to the second half of the 2nd-century BC – first half of the 1st-century BC and group D swords were in circulation for two centuries from the end of the 2nd-century BC onwards (ibid: 41, 54). Some articles of horse-harness are decorated with ‘mirror-style’ motifs, especially terret rings, strap unions and the terminals of cheek-pieces and linch-pins (Jope 2000: Pls. 292-4, 296-300). For example, the enamelled and inscribed horse-trappings from the Polden Hills hoard (Brailsford 1975: Pl. 22a&b) are decorated with a number of different mirror-style motifs arranged in complicated designs. A horse-bit from Walthamstow, London is also inscribed with mirror-style motifs (Ward Perkins 1939: Pl. 18). These shared decorative motifs link mirrors to objects made from materials other than bronze as motifs are also found on selected decorated pottery vessels as well as wooden and bone artefacts. Grimes (1952: Figs. 4, 6, 8 & 9) illustrates a number of pottery fragments from Glastonbury and Meare, decorated with mirror motifs (see Bulleid & Gray 1948: Pl. XIII). There is also a wooden bowl from Glastonbury Lake Village (Coles & Minnitt 1995: Fig. 6.11; Jope 2000: Pl. 312), which is decorated with mirror motifs. Fox (1952: 47) has suggested that the techniques of inscribing the triskele motif on bronze copy relief wood-carving. Two ornamented bone combs from

Scotland (Jope 2000: Pl. 305 f-g) were also inscribed with mirror motifs. With the exception of some of the sword-scabbards (e.g. Isleham and Bugthorpe) and the Llyn Cerrig Bach shield boss, many of the artefacts are decorated with single mirror motifs rather than designs in the style of mirror decoration (see Chapter 4). Pottery could also be an exception. A couple of the fragments from Meare Lake Village, illustrated by Grimes (M209 & M212) do hint that some pots were decorated with complex designs comparable to mirrors: “The main south-western series closes with two examples, both from Meare, which can only be a direct reflection of the art of mirrors. It is unfortunate that they are fragments, so that the complete character of the ornament cannot be determined… The potter responsible for this particular vessel must therefore have been familiar with the mirror art at least as represented by such swords as those from Hunsbury and Meare, even if he – or she – was not always equal to the task of copying them worthily” (Grimes 1952: 169). Mirror-style decoration, in most cases, is placed in the position on the object that would be most visible when it was in use (see Fox 1958: 23). However, it is unclear just how visible these motifs were and you would have had to be in close proximity to objects fully to comprehend the form of the motifs (see Giles 2008; Spratling 2008). Motifs are located on the terminal ends of torcs and collars, where it would be most visible to others; as is the case for the decoration on terrets. The decoration on mirror plates also follows this pattern as it is located on the back so that others could see it when someone was looking at the reflective side. It has been suggested that the circled tricorne on the Birdlip mirror was a ‘good-luck charm’ or a ‘maker’s mark’ (Laing & Laing 1992: 115). This could be a conceivable explanation for some of the visible motifs found on horse-harness. Based on arguments set out in Chapter 3 it could also explain why these motifs appear on some mirror handles; if mirror handles were indeed made by different people to those that made mirror plates. However, the explanation of circled-tricornes as a ‘good-luck’ charm or a ‘maker’s mark’ is less satisfactory in the context of decorated mirror plates. On mirror plates the motifs which are highly visible on other objects are incorporated within roundels and are obscured. Although it has been demonstrated that mirror design can be broken down into a series of standardised motifs, the decoration on mirror backs operates on a larger scale. Motifs do not act as solitary elements; they are integral to the design as a whole (see Chapter 4). Decoration on mirrors is not unique because of the use of exclusive motifs. What makes mirror design unique is the way in which mirror-style motifs are combined to produce characteristic, complex designs, as has already been stated. These designs are very rarely found on objects other than mirrors.

48

Bronze-edging If we look at the third component of some mirrors, the bronze strip protecting the edge of mirror plates, further comparisons can be made. Many wooden objects of known and unknown type were bound in bronze-edging and pieces of bronze-edging strip are relatively common finds on Iron Age sites (e.g. Cunliffe 1984: 344; Coles 1987: 68). One type of object which has now been identified as being bound with bronze is a particular type of hide-shaped shield identified by miniatures in the Salisbury hoard (Stead 1991d; Stead 1998) and confirmed by archaeological discoveries (Parfitt 1995; Johns 2006). Binding can also be conceptually related to torcs such as the famous torcs from Snettisham, Norfolk (Burns 1971; Stead 1991e; Fitzpatrick 1992; Hutcheson 2003) and examples from other parts of the country (e.g. Hill et al. 2004). Techniques of Manufacture The mirror plate and handle represent the two techniques of sheet and cast bronze working in the late Iron Age. A number of complicated Iron Age objects encapsulate different metal working skills into their overall form. For example, torcs such as the Snettisham Great Torc, and the newly discovered torc from Newark-on-Trent, comprise neck bands made of twisted strands of hand rolled wire and hollow cast terminals. Some sword-scabbards comprise decorated sheet metalwork and cast chapes. Objects such as the Battersea shield (Stead 1985), Deskford Carnyx (Hunter 2001) and Kirkburn sword (Stead 1991c) are also highly complex objects. However, it is unusual for a single object to possess two such integral components one made from cast bronze and the other made from sheet-bronze. Everything about the various components of the mirrors signals and refers to their ‘Iron Age origin’. Summary Roman type mirrors have been found in southern England and some were in use during the later Iron Age (Stead & Rigby 1989). These were comparatively small, highly tinned and some were convex, working much like today’s compact makeup mirrors. In order to see your own reflection the technology was there in Britain for people to do it using small mirrors, you did not need to have large, ostentatious bronze mirrors with elaborate decoration. It was a very deliberate decision; mirrors were made using techniques familiar from the recent past using bronze of a composition traditionally associated with Iron Age objects (see Chapter 3 and Appendix E). 6.2 MIRRORS IN USE: ‘TECHNIQUES OF THE BODY’ In this next section we examine how the form of a mirror could have structured the relationships formed around it. Of critical importance is the question of how we can interpret the use of the mirror and its relationships to the human body (Hamilakis et al 2002). The introduction of a new artefact form, requires new ‘techniques of the body’ to be learned: it cannot be used ‘in the way it always has’ (Mauss 1973: 71). We encounter objects in active situations. The ‘potentialities’ held by objects have

been termed ‘affordances’, these exist independently of the perceptions of humans but are relational as they are dependent on human actors (see Knappet 2004; 2005: Chapter 3). A mirror has a handle, for example, but the handle could be too small for some people, or too large. Mirrors have a number of affordant properties. They have a distinctive form: a handle and reflective plate. The handle has a grip and was meant to be held in the hand leaving the plate free from obstruction, implying that the plate was meant to be viewed in its entirety. The plate was highly polished and therefore reflective. Mirrors were made to be held in the hand by the handle with the plate uppermost and people looked at the reflective side of the plate. Iconographic evidence shows Greek and Etruscan mirrors, which are of a similar size and form, being held in this way (e.g. Echt 1999: Figs. 63 & 65). Sometimes an attendant may hold the mirror, as was demonstrated in Chapter 5. Many of the mirrors were broken at the junction of the plate and handle during antiquity. This shows that particular stress was put on this area when a mirror was in ‘use’, indicating that one component was being held whilst the other was unsupported, causing stress and eventually breakage at the junction of the handle and plate. Three different interpretations can then be made in terms of how a mirror was used. The first is that the mirror was a device for a single person to look at their own reflection, either held by him/herself or potentially by an attendant. Second, again conducted by a single person, the reflection could have been used so that people could look behind themselves. Third, the mirror was held reflection-side away from the body so that the reflection could be seen by a wider audience. Past interpretations of mirrors have drawn upon their reflective qualities to suggest that they served an ‘everyday’ function as toilet objects (e.g. Fox 1958). More unusually the reflective qualities of the plate have been ascribed prophetic ‘ritual’ qualities and it has been suggested that mirrors were used for divination (see Johns 2006). I argue that there is no reason why these different potential ‘uses’ should be conceptually separated. Bradley (2005) has recently argued that the distinction we make as archaeologists between the ‘everyday’ and ‘ritual’ is artificial. These two different potentialities, structured around material form, could have been equally important in creating and forming relationships during the life of a mirror and are not necessarily mutually exclusive. In some circumstances it could have been used in a very private space, while at other times it could have been involved in more public performances. These uses can be accounted for in the ways other societies have been observed to use mirrors (Schweig 1941; Saunders 1988; Loveday 1990; Izzet 1998; Saunders 2001; Pendergrast 2003; 2007) and the way relationships could have been structured around form. ‘Gazing’ at reflection Mirrors were used to monitor personal appearance in the Etruscan, Greek, Roman and Egyptian worlds (Echt 1999). A single person looking into a mirror, holding it at arm’slength would have been able to see their own reflection: the

49

mirror in this instance is an instrument which allows one to view oneself. At a time when looking at your own reflection would have been very difficult, except perhaps from reflections in pools of water, being able to see how other people viewed you would have been a powerful asset. Having control over an artefact that allowed you to see your reflection would have meant that you had control over how you looked to others. Objects can act as extensions of the human body and become part of them (Tilley 2004: 9). A mirror therefore by becoming an extended part of the human body was powerful because it allowed control over how other people may have perceived you. It also became an active part of who that person was, alongside the tools which allowed people to adjust their appearance like tweezers and ear-scoops (Hill 1997), shears, razors, cosmetic sets and grinders (Jackson 1985; Carr 2000; 2005), and different modes of dress (Jundi & Hill 1997; Sørensen 1997). Physical appearance is a powerful and meaningful means of communicating social identity (Sørensen 1997: 93). By using a mirror to inspect the body, hair or clothing and to monitor appearance as well as gesture or posture, ideals of how the body should look could be reproduced (Foucault 1990; Giles & Joy 2007: 21). It is not being suggested that people did not take an interest in personal appearance before they made mirrors. However, prior to their use people were reliant on others to tell them how they looked. Mirrors facilitated a much more personal engagement with the self and allowed a person to take direct control over how they looked. This is often overlooked in archaeological interpretations which view ‘cosmetic’ objects as passively reflecting a general concern for personal appearance and dress. Objects like mirrors were actively involved in the creation of new social identities (see Olsen 2003) expressed through the way people looked. The act of looking in a mirror, if it was done in a public place also communicated something about the person using it. A person was being seen to be concerned with the way they looked through the acts of altering bodily appearance. Messages were also transmitted to others by the movement of the modified human body. The deposition of mirrors, brooches and tweezers alongside bodies in graves also communicated something about the dead person, and possibly more importantly the dead person’s relatives, to those present at the burial (Parker-Pearson 1999a) (see Chapter 10). The mirror as ‘third eye’ The second way in which a mirror could have been used by a single person would be to look over one’s shoulder. A mirror reflects light and allows you to look behind yourself, extending normal perception and bodily experience, acting as a ‘third eye’. Forward and back, as well as up and down, and left and right, are fundamental to the way in which we experience the world through our own bodies (see Tilley 2004: 4-10). In our own society an opposition is often set up between forward and backwards. You can metaphorically move forward and back-

wards in your career (Tilley 1999; Lackoff & Johnson 2003 (1980)). A forwards step is positive, a backward step is negative. In the Greco-Roman world looking backwards was linked to looking into the future or the past and the reflection from a mirror was used in divination. Catoptromancy is the practice of divination using mirrors (Addey 2007: 32). It was not the most well known or prevailing method of divination in the Greco-Roman world. One method of catoptromancy was to read reflections in a mirror. This was linked to hydromancy which is the practice of divination using reflections off of water. Mirrors were used in association with water by lowering them into it and ‘reading’ reflections (see Schweig 1941; Addey 2007: 33). Mirrors were particularly associated with foretelling death (Addey 2007: 33). This is not an attempt to imply that this is the manner in which mirrors were used in an Iron Age context, however reflections and shiny objects were important to many different cultures (see Saunders 1988; Saunders 2002). “The ancient Egyptians, Indians, Chinese, Mayans, Incas and Aztecs buried their dead with metal or stone reflectors, to hold the soul, ward off evil spirits, or allow the body to check its appearance, before taking the final trip to the after-life” (Pendergrast 2007: 1-2). The potential remains for the reflectivity of mirrors to act to extend the relationships that could be structured around them beyond the realms of ‘normal’ human physical experience. For example, owning a mirror was said to be one of the traits of a ‘magician’ in the Roman world (Janowitz 2001). Mirrors allowed a person using them to look backwards as well as forwards. A person could use a mirror as an extension of their own body. If you are seen to use a mirror other people will be aware that you are in control of that power. The ‘otherworldly’ relationships structured around mirrors could become associated with the person ‘using’ a mirror through physical proximity: the mirror may have become a part of a person as they act out their identity. Again a similar argument could be forwarded for the relationship between a mirror and the person it is buried with, as well as the people who organised the burial (Chapter 10). Mirrors used to ‘evoke’ light There is a third use for mirrors which extends the potential relationships structured around them beyond the person holding the mirror. If the reflective side of the mirror is held facing away from the body, light can be reflected and can

50

dazzle an audience. A round mirror can act to reflect the sun’s light but it also becomes a miniature version of the sun (see Pendergrast 2007: 2). A mirror could also be used to reflect moonlight. A second type of divination in the Greco-Roman world using mirrors was their use to evoke light (Addey 2007: 40). Reflected light was used to illuminate the ‘soul vehicle,’ to evoke the presence of a divine spirit within the soul of the practitioner. Careful excavation has revealed that a number of British Iron Age mirrors were put in the ground wrapped in an organic container, like a bag or leather pouch (see Chapter 9). The mirrors could have been similarly contained in life as in death perhaps to protect the surface of the mirror. Alternatively, if they were used to ‘evoke’ light it may only have been appropriate to do this in certain contexts, at other times the mirror would have to be covered. 6.3 CONCLUSION The form of a mirror afforded a range of possible relationships with the human body. Although its usage can be seen to be restricted to a single individual, relationships are extended as the reflectivity of the plate is manipulated. Using the mirror alongside other objects to alter physical appearance extends the relationships radiating from a mirror to other cosmetic objects. The movement of the altered human body through space also acts to extend relationships to other individuals. The reflective plate allows one to see forwards and backwards, acting to expand human perception. The plate can also be used to harness light. All of these ‘otherworldly’ qualities further act to increase the potential relationships structured around mirror form. By gathering up different elements of a series of types of object forms recognisable from the later Iron Age, mirrors also reference other items of material culture through form, decoration and methods of construction. This is particularly the case for later bronze mirrors which tend to have a binding strip running along the edge of the mirror plate. This binding strip acts to bind all of the elements together and emphasises the singular nature of the object but at the same time continuing references to wider Iron Age material culture. It is possible that people encountering mirrors would have made these associations and linkages with other items of material culture. Different people, depending upon differing levels of knowledge would have made different associations and linkages.

CHAPTER 7 THE MIRROR AS SOCIAL AND CULTURAL NEXUS This short chapter serves as a summary to this section of the book, which provides a biographical account of what was written into mirrors both at the time of their manufacture (Chapters 3, 4 & 6) and also over the course of their social lives (Chapters 5 & 6). It demonstrates how the specific form of mirrors acted to determine the manner in which relationships could be structured around them (see Strathern 1988: 176). It also shows how use-wear analysis can be used to examine social life. Above all, it underlines the value of the information written into artefacts. 7.1 MAKING MIRRORS In Chapter 3 the relationships formulated when mirrors were manufactured were examined. Mirrors are complicated objects and a number of different stages in their manufacture were identified. It was demonstrated that each stage of manufacture provided a potential stage setting for social interaction. Some mirrors were more complicated to make than others. Stages in production differed depending upon the material/s a mirror was made of and the number of component parts. Iron mirrors required the least number of manufacturing stages. Decorated bronze mirrors and those made of composite components were the most complicated to make. A decorated, edged mirror would have required more production steps than a plain un-edged one. Similarly, a mirror with an iron plate and a bronze handle would have required wide ranging technological skills to construct. A large number of manufacturing stages were required, indicating that a number of different individuals were probably involved in the manufacture of composite mirrors. The relationships formulated during mirror manufacture would have been critical in setting the potential future trajectories of their object lives. For example, a mirror could have become intimately associated with the person or group who made it, carrying that link throughout their social lives. Through the making of a complicated artefact like a mirror the person or group responsible also acted to demonstrate the different relationships they were able to formulate and coordinate. By making complicated objects like mirrors (or coordinating their making), people acted to demonstrate the array of relationship webs they had access to and were able to manipulate. 7.2 MIRROR DECORATION In Chapter 4, through a re-examination of mirror decoration it was demonstrated that decoration is not merely pleasing to the eye. Contrary to many previous interpretations of Celtic Art, it was found that individual motifs used to decorate mirrors are not representative of something else. Decoration is critical to the efficacy of mirrors and can only be understood in respect to them. Design consistently adheres to a set of design rules which are maintained over a wide geographical area and a significant time period. Although there is considerable regional and temporal variation in design, mirror decoration is consistently organised so that there is a balance of positive and negative motifs and unused-space is minimised. Mirror decoration is located on the plate on the opposite face to the reflective surface. This location is significant as design can be seen to evoke the reflective properties of the mirror and positive motifs frame and create distorted images of themselves within the design field: negative motifs. The decorated and reflective faces of

51

the plate possess a kind of ‘inner logic’: they metaphorically refer to one another and act to reinforce meaning. Far from being passive, decoration acts to reemphasise and reinforce the primary function and most powerful property of a mirror: its ability to reflect light. Motifs are therefore not representational of anything other than ‘mirrorness’. 7.3 ANALYSIS OF USE/WEAR/REPAIR Chapter 5 summarises the results of analysis of over 30 mirrors which were visually examined for signs of wear, polishing and repair. The aim was to determine how mirrors were ‘used’: how they performed in social relations. The key conclusion to be drawn from this chapter is that signs of wear, polishing and repair were detected on the majority of mirrors examined, indicating social usage. Mirrors were also clearly valued as some were repaired to prolong their social lives. Mirrors were therefore not just made for deposition; they were made to act in and formulate social relations. Individual mirrors did not necessarily act in the same way or for the same purposes and analysis demonstrates that the period and intensity of usage varied. However, similarities in wear patterns can indicate some of the ways mirrors were used. Many mirrors are broken or damaged at the junction between the handle and the plate, indicating stress in this area and suggesting that mirrors were held by the handle, while the plate was unsupported. Many of the terminal loops of mirror handles were considerably worn indicating that these loops may have been used to secure a covering around the mirror, demonstrating that for much of their lives they were covered up. 7.4 MIRROR FORM Chapter 6 specifically examined the form of mirrors to consider how this could determine the kinds of relationships that were structured around them. It was found that mirrors are made so that their different components reference other items of Iron Age material culture through their form, decoration and the methods in which they have been constructed. Mirrors gather up different elements of a series of types of object forms recognisable from the later Iron Age. This is particularly the case for bronze mirrors which tend to have a binding strip running along the edge of the plate. This binding strip acts to bind all of the elements together and emphasises the singular nature of the object but at the same time continues to reference wider Iron Age material culture. It could be argued that by constructing mirrors in this way, Iron Age origin is reemphasised, possibly expressing an idealised sense of continuity with the past, present and future and a ‘timeless’ quality to their manufacture. Form also affected how mirrors were used in relation to the human body, restricting usage to a single individual. The reflectivity of the plate could also be manipulated. Using the mirror alongside other objects to alter physical appearance extended the relationships radiating from a mirror to other cosmetic objects. Movement of the altered human body through space also acted to extend relationships to further individuals. The reflective plate allowed one to see forwards and backwards, acting to expand

human perception. The plate could also be used to harness light. All of these ‘otherworldly’ qualities further increased the potential relationships structured around the form of a mirror. 7.5 MIRRORS AS SOCIAL AND CULTURAL NEXUS A biographical account of what is written into mirrors has revealed that they were made in such a way that multiple and often multifaceted relationships could cluster around them. More complex mirrors had the potential to accumulate more complex networks and referents. For example, relationships were formulated through the different stages of mirror manufacture. Complicated mirrors required more production stages and therefore broader relationship webs were formulated in order to make them. Biographical information was also built into the artefact at the time of production. Mirrors are complicated objects made from a number of component parts, which are brought together as a single artefact. The effect of this construction is to provide a number of interfaces around which relationships can be structured. The more complicated mirrors, with edges and decorated plates, were provided with more interfaces. Duff (1975; 1983) highlighted the manner in which different elements of an artefact – form, function, decoration – inform each other. According to Duff (1983: 48), it is not sufficient to describe the form of an object or the decoration as this does not account for meaning. Meaning for Duff resides in the interface between the different elements that makeup what we understand as an artefact (ibid: 66); in the metaphors they construct and the relationships they articulate (Marshall 2000: 226). Through their inner

52

logic and the interplay between separate components mirrors are constructed and intended to be inherently meaningful and powerful objects. The separate components, especially the decorated, edged mirrors, act to collect together referents to contemporary Iron Age artefacts through their form, methods of construction and decoration. Each component also refers to other components and the artefact as a whole, facilitating the primary function of a mirror which is to reflect light. The handle allows the mirror to be held so that the reflection is not hindered. Binding, or decoration around the edge, acts to contain the reflection. Mirror decoration refers to and reemphasises the reflective qualities of the plate and expresses ‘mirrorness’. The property of reflection could also have been manipulated in multiple ways. Visual examination of mirrors has revealed that they were ‘used’ and they performed in social relations. Commonalities in wear, repair, and polishing suggest that some mirrors shared life trajectories. However, there was significant variation in the type and intensity of usage, indicating that although mirrors were provided with the potential to act within multiple relationships, they were not all ‘used’ in the same way. The next section of this book examines the different deposition contexts mirrors have been found in. It analyses the date of deposition (Chapter 8) and outlines individual deposition contexts in detail (Chapter 9). The contextual information is then interpreted using the biographical information written into mirrors summarised in this chapter (Chapter 10).

CHAPTER 8 DATING AUDIT Building upon the conclusions from the first section of this book, summarised in Chapter 7, the next three chapters examine mirror deposition context. The aim of this short chapter is to establish when mirrors were deposited. In Chapter 9 the different depositional contexts mirrors have been uncovered in are outlined. In Chapter 10 these different contexts are interpreted, building upon the biographical account of what was written into mirrors before they were deposited. 8.1 INTRODUCTION TO DATING MIRROR DEPOSITION When Franks (1865: 72) said of the British mirrors that they are “…probably not more ancient than the introduction of coinage into Britain from 200 to 100 BC, and not much later than the close of the first century after Christ…”, he set a conceptual timeframe for the date of their deposition. This timeframe was readily accepted by Smith (1909: 341) and has been implicitly assumed in all subsequent analyses of the mirror corpus. However, until recently there was little or no hard evidence to back it up. This chapter will interrogate the dating of the entire corpus and assess how far our knowledge of dating has progressed in the past 150 years. Dating of mirrors is currently reliant on dating of associated objects, especially brooches and pottery. Where there are no associated objects, or they are not easily datable, mirrors have been dated using stylistic analysis of decoration (e.g. Fox 1958: 102; Jope 2000: 137-48) and handle typologies (Fox 1949; Spratling 1970a; Jope 2000: 137-8). Until recently only four mirrors with datable associated artefacts had been recovered: Birdlip, Colchester, Nijmegen and St. Keverne (Spratling 1970a). All others were dated stylistically by comparison to these four (see Fox & Hull 1948; Fox 1949; 1958). Since the 1970s thirteen more potentially datable mirrors have been recovered. Human remains discovered with four of these (Bryher, Garton Slack, Wetwang Slack and Wetwang Village) have been radio-carbon dated, and the remaining nine (Aston, Brecon, Chilham Castle, Chettle, Dorton, Holcombe, Latchmere Green, Pegsdon and Portesham) have been dated using associated artefacts. Dating using associated artefacts and stylistic comparisons has caused a number of problems: 1. Although brooch and some pottery chronologies are relatively well understood for southeast England during the 1st-centuries BC/AD, chronologies are far less precise for other parts of Britain and for earlier periods (see Haselgrove et al. 2002: 2-7 for discussion of current dating issues). Chronologies in these other regions are heavily reliant on dating objects of decorated metalwork. 2. Typologies still dominate present interpretations of Iron Age chronology, causing difficulties when matching-up the dating of different categories of material culture, for example pottery and brooches (Haselgrove et al. 2002: 3). This has caused particular problems when dating the Dorton and Bryher mirrors.

53

Stylistic analysis is subjective, relying on perceived ‘development’ of styles through time. The mirror corpus was initially dated stylistically in relation to only four mirrors found with datable associated objects. Consequently dates tend to be clustered in groups. This has possibly given a false impression of distinct time periods of usage and deposition patterns, where an interpretation of continuous usage may be equally valid. The reports published prior to the revision of the continental dating of La Tène D1 and D2 (see Haselgrove 1997) may also have dated some mirrors too late. 8.2 WHY IS ESTABLISHING A CHRONOLOGY FOR MIRRORS IMPORTANT? It is important to establish when mirrors were deposited so that deposition can be related to the wider social and cultural context. If we know when they were put in the ground, we may be able to identify regional and chronological patterns, and relate these to wider changes occurring in the archaeological record, comparing them to other items of material culture, placing ourselves in a better position to determine the kinds of social relationships that constituted mirrors. Mirrors are a highly recognisable and important group of later Iron Age material culture. There is a significant existing literature discussing their dating on stylistic and typological grounds (e.g. Franks 1865; Smith 1909; Dunning 1928; Fox 1949; 1958; Spratling 1970a; 1972; Fitzpatrick 1997a; Jope 2000). As a result, mirror chronology is thought to be well understood. However, interrogation of the literature has revealed that deep-grained past typological and stylistic assumptions still dominate discussion of chronology. It is necessary to re-examine the dating of mirror deposition so that these inaccuracies are not passed on to wider Iron Age research. For example, because mirror dating is thought to be well understood their existence in particular contexts is often used in discussions of the dating of sites, often resulting in a site being assigned a later date than would otherwise have been the case because of the presence of a mirror. Mirrors in burial contexts are also often found in close association with both pottery and diagnostic metal artefacts such as brooches. A greater understanding of chronology could help reconcile current discrepancies between pottery and metalwork chronologies for the later Iron Age (cf. Haselgrove et al. 2002: 6). We are now in a very good position to re-examine the date of mirror deposition. As outlined above, a number of new mirrors have been discovered over the past 35 years: some have been radiocarbon dated, while others were found with datable associated artefacts. Discussions of the date of deposition for these newly discovered mirrors have been reported elsewhere; however, with the exception of Fox (in Fox & Pollard 1973), Fitzpatrick (1997a), Johns (2006: Table 15) and Burleigh & Megaw (2007) reports concentrate on individual mirrors and pay little attention to the mirror corpus. Sealey (2006) has recently recognised the potential provided by new finds and has conducted a comprehensive dating review of mirrors.

Figure 8.1 Dating audit of the deposition of mirrors Building on Sealey’s work, the evidence for the East Riding mirrors has also been re-examined (see Appendix B). Based upon new evidence, revised dates are also provided for mirrors such as Latchmere Green. 8.3 THE CHRONOLOGY OF IRON AGE MIRRORS A detailed interrogation of deposition date, alongside illustrations of the associated artefacts that have been used as dating evidence, is outlined in Appendix B. Figure 8.1 is a summary of the results of this analysis. From this overview we can see that over the past thirty years or so, our knowledge of the deposition date of mirrors has increased dramatically. A number of trends can be noted. First, the mirrors from East Riding are clearly the earliest with the radiocarbon dates from Wetwang Slack and Wetwang Village dating the mirror burials from around 400-150 BC. The original radiocarbon date from Garton Slack was an anomaly. The new radiocarbon date of 380200 cal BC (95% probability) (see Garrow et al forthcoming) confirms this date range for the East Riding mirrors (Figure 8.2).

54

Figure 8.2 Calibrated radiocarbon dates for East Riding mirrors The two mirrors from Cornwall were both found with similar brooches. The radiocarbon dates for Bryher, when combined with the dating evidence from the metal objects found in the grave, could suggest a late 2nd-century BC date, or a date in the first half of the 1st-century BC (Hill in Johns 2006). Due to similarity in the size and shape of the mirrors and the brooches found at Bryher and St. Keverne, it is reasonable to suggest a similar date for the St. Keverne mirror. The majority of mirrors found in cremation burials in southeast England all date to La Tène D2 (c. 75–25 BC) (Haselgrove 1997), for example, the Aston and Chilham

Castle mirrors and the Billericay III mirror handle. The Latchmere Green burial probably took place much earlier than the date proposed by the excavators, placing it within this time frame. The brooch found with the Pegsdon mirror also dates its deposition to this period (Burleigh & Megaw 2007). Although the bronze cup and pin are interesting, for dating purposes the mica dusted imported flagons are the best dating evidence for the Colchester I mirror (Fitzpatrick 1997a: 67), indicating a date of deposition of around 15 BC or just before (Sealey 2006: 17). The Dorton amphorae, although worn, could also date the Dorton burial to approximately 15 BC (ibid). Finally, where we have firm dating evidence for the larger, kidney-shaped, edged mirrors, found outside of the ‘core’ of southeast England, they appear to have been deposited in the ground during the 1st-century AD. The majority were deposited in the second half of this century. The pottery from Holcombe and the metal objects from Portesham both point to a date at around the time of the Roman Conquest in Southern England, or slightly later. The dating of the recently discovered Chettle mirror to the last quarter of the 1st-century AD (ibid) adds weight to this interpretation. The plate is worn and the handle was lost in antiquity, so it was probably in circulation for an extended period. The Chettle mirror plate is kidney-shaped and the pattern of decoration is similar to that on the Birdlip mirror. The Birdlip burial could have taken place anytime from the late 1st-century BC – AD 60/70. Although, based on the distribution of Birdlip brooches and their general absence from the south-eastern ‘core’, they could date towards the middle of the 1st-century AD, mirroring Colchester derivatives from southeast England. The Nijmegen mirror may have been curated a long time before it was deposited (Lloyd-Morgan 1981: 114) in the late 1st-century or the early 2nd-century AD, but fits into the general pattern of deposition of the large decorated mirrors. The idiosyncratic Brecon mirror, which has an iron plate with a decorated copper-alloy ‘collar’, also dates to the last quarter of the 1st-century AD (Sealey 2006: 17; Davies & Gwilt 2008), but should not be seen as representative of the rest of the mirror corpus. 8.4 MIRROR DECORATION – REGIONAL AND TEMPORAL TRENDS The next stage of analysis is to describe how mirror-design changed through time and to identify regional or temporal patterns in the use of certain motifs. Two broad chronologically significant groups, where mirror design is similar can be identified: southeast England (75-15 BC); and ‘Western’ (1st-century AD). The regional grouping based on decoration is confirmed by contextual evidence (Chapter 9) and other physical characteristics (Appendix G). South-Eastern Mirrors South-eastern mirrors are found in cremation burials, often alongside pottery and brooches (Rook et al. 1982; Fulford & Creighton 1998; Parfitt 1998). They are relatively small in size and have round plates (see Fox & Pollard 1973). Decoration is characterised by a filled/empty circle, or a circled tricorne, within a keeled-roundel (see the Bromham and Aston roundels in Figure 4.6). These motifs are com-

55

mon to all south-eastern mirrors with decorated backs, even those like Old Warden I or Great Chesterford, with ‘diffused’ designs. Aston, Billericay I, Dorton, Old Warden I, Rivenhall I and Latchmere Green all have keeledroundels with empty circles. Bromham, Chilham Castle, Colchester I and Great Chesterford have roundels, or keeled-roundels, containing filled circles or circled-tricornes. The unprovenanced Mayer and Gibbs mirrors are both decorated with keeled-roundels, are small in size and have round plates, suggesting that they belong in the south-eastern group. Western’ Mirrors The ‘western’ mirror group has long been identified (see Fox 1958; Fox & Pollard 1973; Fitzpatrick 1997a). Geographical distribution is confined to the fringes of the south-eastern corner of England. The ‘western’ mirror group are seen to share a similar design, common motifs (Fox & Pollard 1973: Fig. 7) and were made using similar techniques. Fox (in Fox & Pollard 1973: 32-7) discussed the ‘western mirror school’ of mirrors in the context of the Holcombe mirror and defines the ‘western’ mirrors as follows: · Mirrors of large size, probably designed for suspension. · Handles of multi-loop type, some decorated with red enamel and/or with studs of copper cuprite. · Engraving or chasing in regular ‘three-stroke basketry’ and hatching. · Designs of fold-over symmetry, some using compass construction. · Motifs, crescent-rings and peltae The unifying motif is the armadillo (see Figure 4.5) or ‘crescent motif’ identified by Fitzpatrick (1997a: 60). This motif is found on the Birdlip, Desborough, Holcombe, Mount Batten I and Nijmegen mirrors as well as the fragment from The Verne, Portland. The armadillo motif is often combined with fins and cusps to form roundels (see Figure 4.5). The only mirror identified in the group, which is not decorated with this motif, is the mirror from Portesham. Portesham is decorated with an ADC motif (see Figure 4.7). The Billericay mirror also has the same motif, located in the centre of the plate. The Birdlip mirror handle has an ADC in the terminal loop of its handle. Sub-Regional Groupings Within these broad regional groups a number of smaller groups of mirrors sharing similar decoration can be identified. The most obvious are the mirrors with a more ‘diffused’ design, including the Old Warden I and Great Chesterford mirrors. These also have similar handles. The Latchmere Green and Rivenhall I mirrors also have similar decoration with negative motifs incorporated within positive fin motifs (Figure 8.3). As do the newly discovered mirrors from Pegsdon and ‘Oxfordshire.’ The only decorated mirrors excluded from my two main regional ‘groups’ are both from Cornwall: St. Keverne

Figure 8.3 Sections of the decoration on the Latchmere Green and Rivenhall mirrors and Bryher. These two mirrors look to be related due to their geographical proximity and a similarity in size and form, as well as, the distinctive triangular motifs circumnavigating the edge of both plates. The mirrors were also placed in stone-lined cist graves at approximately the same time (Chapter 9). The dating of these burials suggests that they are the earliest decorated mirrors (see above). However, as has already been suggested, despite all of these similarities, decoration is quite different. The decoration on the St. Keverne mirror shares many motifs with other mirrors, especially the south-eastern group, which are probably slightly later. Shared motifs include circled-tricornes and keeled-roundels. Some attention is also paid to negative motifs within the roundels (see above). The major difference between St. Keverne and south-eastern mirrors is the bi-roundel design-field of St. Keverne. The decoration on the Bryher mirror is very different. It contains a unique motif which consists of a circle within a circle demarcated by at least 11 lines in between them. It is also unclear how the overall design was set-out. Temporal Trends Tied to these major groupings a number of trends and changes can also be identified in the ways in which mirrors are decorated through time. The positive motifs on some of the earlier south-eastern mirrors are very difficult to identify. Indeed, on some like Chilham Castle, Dorton and the Disney mirror, the positive decoration is in the form of very narrow hatched lines, which on the surface do not resemble any motif. These lines are still broadly fin-shaped, with the possible exception of the Chilham Castle mirror. On mirrors, where the positive motifs are less well defined, their primary function is to demarcate negative motifs, which form the majority of used space on the plate surface. As noted above, there is a greater concern with clearly defining each individual positive motif on the designs of the later ‘western mirrors’. The negative motifs in these designs also tend to be much smaller in size. Designs are highly complex with more individual positive motifs in each. For example, on south-eastern mirrors keeled-roundels form single mo

56

tifs. The keeled-roundels on the Desborough mirror on the other hand are clearly made-up from two fin-shaped motifs. Keeping the outline of the shorter concave edges of these fins also allowed an additional space to be demarcated, which was filled with a circled tricorne. 8.5 INTERPRETATION OF MIRROR CHRONOLOGY There has been some general discussion of mirror deposition date (Fox & Pollard 1973; Fitzpatrick 1997a; Johns 2006) and comments on the deposition date of individual mirrors in excavation reports (e.g. Stead in Parfitt 1998). Mirror chronology has also been discussed in two recent reports (Sealey 2006; Burleigh & Megaw 2007). These concentrate on the decorated mirrors from southern Britain. Sealey (2006: 16) is concerned with tracking the spread of mirrors across southern Britain, looking for the ‘origins’ of the large decorated mirrors deposited in the 1st-century AD. Sealey suggested that decorated mirrors originated in southeast England and were ‘diffused’ to the west when they went out of fashion in the southeast. The two mirrors from Cornwall are seen as ‘outliers’ from the south-eastern mirror-making tradition. Burleigh & Megaw (2007), on the other hand, question the idea that there are two distinct ‘groups’ of decorated mirrors: southeast and western. The results of my analysis of mirror deposition date clearly distinguish four chronologically and temporally significant episodes of mirror deposition: 1. Iron mirrors in East Yorkshire in 400 – 150 BC 2. Decorated bronze mirrors in Cornwall 125 – 80 BC 3. Decorated bronze mirrors in southeast England 75 – 15 BC 4. Large decorated bronze mirrors outside of the south-eastern ‘core’ AD 1 - 100

These all relate to periods and regions when burials are most visible in the archaeological record. This reveals a bias in the dating of mirrors: all datable mirrors, except Holcombe, which comes from a well recorded excavation (Pollard 1974), are from graves. The mirror from Maiden Castle, Dorset was found in Wheeler’s (1943) ‘Site E’ and is likely to date to some time before the mid-1st century BC (Fitzpatrick 1997a: 67). The excavators of the Rivenhall site, Essex (Rodwell & Rodwell 1986: 19), were unable to determine whether the two mirrors from Rivenhall were deposited before, or after, the Roman Conquest. The bronze mirror plate from Glastonbury Lake Village was found amongst the palisading at the southern margin of mound 30 at the site (Bulleid & Gray 1911: 223). Three La Tène III brooches were also found in mound 30, although these are identified with a question mark by the excavators (Coles & Minnitt 1995: 61, 177). A re-examination of the dating of the site (ibid: 174-9) suggests that it was occupied from 250 – 50 BC. Fitzpatrick (1997a: 67) suggested that the mirror may date to the mid 1st-century BC. Although dating evidence is limited, the mirrors deposited in Wales and Scotland are likely to be later in date, being deposited after the Roman conquest of southern England. In addition to the problems associated with dating mirrors found on settlement sites, they are often different from those placed in burials. Many are composite. The bronze plate of the Glastonbury Lake Village mirror had an iron handle (Bulleid & Gray 1911: 223) and the Maiden Castle mirror has an iron plate with a bronze rim (Wheeler 1943). These mirrors therefore cannot be said to be representative of those found in graves. However, as other finds also indicate, they show us that not all mirrors were decorated, or deposited in graves. This may suggest that mirrors that were deposited in graves were somehow different and that ‘difference’ could have been signalled through mirror decoration (Chapter 4), or the materials used to make mirror component parts (Chapter 3). The results of this analysis of mirror chronology indicate that mirrors were first deposited in East Riding. These were made of iron and were undecorated. The next mirrors to be deposited came from Cornwall and then, marginally later, in southeast England. These were made of bronze and the plates were decorated. It is probable, though not certain, that mirrors continued to be circulated in parts of Britain even though there could be a gap of up to 100 years between deposition of iron mirrors in graves in East Yorkshire and decorated bronze mirrors in Cornwall and southeast England. These were probably disposed of in archaeologically invisible ways. The few mirrors found on settlement sites indicate that some in circulation were significantly different to the decorated bronze mirrors from graves. Unlike Sealey

57

(2006: 16), it is not argued here that the Cornish mirrors were ‘outliers’; they represent mirrors that were probably in circulation throughout southern Britain, which happened to be deposited in graves. A more salient question than searching for ‘origins’, or identifying ‘outliers’, is to ask why mirrors were decorated, or to question why they were put in graves. The deposition of large decorated bronze mirrors in Britain during the mid-late 1st-century AD represents a slightly different phenomenon. We can be reasonably confident that they did not continue to be used in southeast England in the 1st-century AD. With the well recorded increase in the quantity of material found in the archaeological record at this time (see Hill 1997), it seems improbable that if mirrors were in common usage, they would not also be found more frequently. Stead (in Stead & Rigby 1989: 103) suggested that British-type mirrors were replaced by Roman-type ones in this region during the 1st-century AD. However, outside of the King Harry Lane cemetery, the only other Roman-type mirror to be found in a British Iron Age context was found at Hayling Island, Sussex (Sealey 2006: 16). A tubular ring of sheet copper-alloy from Cadbury Castle, Somerset could also represent the remains of a possible Romantype mirror (Barrett et al 2000: 197). The pattern identified for the abandonment of mirrors in south-east England in the 1st-century AD and continued usage in ‘western’ areas past the Roman conquest is a real phenomenon. This blueprint is also echoed in the deposition of other types of material culture. Tankards are found in well-known south-eastern cremation burials at Welwyn and Aylesford dating from 50-25 BC. They are also found in contexts dating to the mid 1st-century AD outside southeast England at Camerton (Jackson 1990) and the Severn Sisters hoard (Davies & Spratling 1976). Horseharness fittings have also been found in hoards dating to the 1st-century AD at Polden Hills (Brailsford 1975) and Stanwick (Spratling 1981). 8.6 CONCLUSION Mirrors were deposited within a period of up to 500 years (400 BC – AD 100). A number of regionally and temporally significant episodes of deposition have been identified. However, it is probable that mirrors remained in circulation throughout this period but were disposed of in archaeologically invisible ways. Analysis of decoration, size and form has revealed that mirrors changed over time and between regions. They became larger over time (Appendix G) and the techniques and materials used to make them altered between regions.

CHAPTER 9 DEPOSITIONAL CONTEXT 9.1 INTRODUCTION Until recently it has not been possible to examine and interrogate the depositional context of mirrors beyond basic interpretations because many of the finds were made in the 19th-century and came from poorly recorded excavations or were chance discoveries. However, subsequent archaeological examination of the findspot of a number of recent metal detector finds, especially from southeast England, has greatly increased the quantity and quality of information available. Although there are a number of very detailed excavation reports (Fox & Pollard 1973; Brewster 1980; Rook et al. 1982; Farley 1983; Fitzpatrick 1997a; Fulford & Creighton 1998; Parfitt 1998; Johns 2006), none has yet synthesised these reports, examined broader trends of mirror deposition, and questioned what these new discoveries can tell us about the use and significance of mirrors. This chapter systematically examines the different depositional contexts mirrors have been found in: burial, settlement, and watery. Most mirrors have been found in graves. Although mirrors are not common finds, they have been found in all the well known regionally distinctive burial traditions of the mid-to-late Iron Age (see Whimster 1981) (Table 9.1). Mirrors have been found in stone-lined cist graves from Cornwall and the Scilly Isles (Jope-Rogers 1873; Thomas 1975; Whimster 1977); inhumations under square barrows in the East Riding (Stead 1991a; 1991c); earlier cremation burials in southeast England (Birchall 1965; Stead 1976a); and more localised burial traditions from western England: the south Dorset or ‘Durotrigian’ burial rite (Woodward 1993) and inhumation burials from Gloucestershire (Staelens 1982). A mirror has even been found in a grave outside Britain at Nijmegen, The Netherlands (Dunning 1928). Mirrors have also been found in settlement contexts at Maiden Castle (Wheeler 1943), Rivenhall (Rodwell & Rodwell 1986; Lloyd-Morgan 1993; Northover 1993), Glastonbury Lake Village (Bulleid & Gray 1911; Coles & Minnitt 1995) and Holcombe (Fox & Pollard 1973; Pollard 1974). A small number were deposited in watery contexts, in bogs in southwest Scotland (MacGregor 1976) and Northern Ireland (Jope 1954) and in the River Oise, France (Guillaumet & Schönfelder 2001). 9.2 BURIAL CONTEXTS The following discussion of mirrors deposited in burials will be structured around specific identified regional burial traditions of the mid-to-late Iron Age in Britain. Each mirror will be discussed in turn and related to its particular regional context (Table 9.1). East Yorkshire The archaeology of Eastern Yorkshire is very distinctive. From the middle Iron Age major changes occurred in the archaeological record, characterised by the creation of square barrow inhumation burials, long-distance boundaries (known locally as ‘dykes’) and ladder/droveway settlements, which came into existence gradually over a number of generations (Giles 2007: 239). Through time there was increasing sub-division of the landscape with track-ways cutting across it, facilitating travel across boundaries. ‘Ladder’ settlements grew up (e.g. Melton (Bishop 1999)),

58

Mirror

Country/Region

Context

Arras I

East Riding

Inhumation

Arras II

East Riding

Inhumation

Aston

Southeast England

Cremation

Billericay I

Southeast England

Cremation?

Billericay II

Southeast England

Cremation?

Billericay III

Southeast England

Burial?

Birdlip

Southwest England

Inhumation

Brecon Beacons

Wales

Cremation

Bridport

Southern England

Inhumation

Bryher

Southwest England

Inhumation

Chettle

Southern England

Cremation?

Chilham Castle

Southeast Englannd

Cremation

Colchester I

Southeast England

Cremation

Dorton

Southeast England

Cremation

Garton Slack

East Riding

Inhumation

Lambay Island

Ireland

Inhumation

Latchmere Green

Southern England

Cremation

Llechwedd-du

Wales

Burial?

Nijmegen

The Netherlands

Cremation

Old Warden I

Southeast England

Cremation?

Old Warden II

Southeast England

Cremation?

Pegsdon

Southeast England

Cremation

Portesham

Southern England

Inhumation

Portland I

Southern England

Inhumation?

Portland II

Southern England

Inhumation?

St. Keverne

Southwest England

Inhumation

Stamford Hill I

Southwest England

Inhumation

Stamford Hill II

Southwest England

Inhumation

Stamford Hill III

Southwest England

Inhumation

Wetwang Slack

East Riding

Inhumation

Wetwang Village

East Riding

Inhumation

Table 9.1 Mirrors from burial contexts orientated with existing boundaries, consisting of a series of house plots alongside a road or track-way (Stoertz 1997: 41) that could be up to 1km in length and a few hundred metres wide. These settlements are often found in close association with cemeteries. When the switch was made from the practice of excarnation (see Carr & Knüsel 1997), which was probably dominant in the region from the late Bronze Age to the early Iron Age, to the burial of individuals under square mounds of earth, the impact must have been great. As Bevan (1999a: 135) points out “this would have had a massive impact on those individuals who saw the first square barrows in their familiar landscape”. To later generations it would have been perceived as an integral part of everyday life,

Description of Grave Occupant

Name/ Location

Sex

Vehicle

Crouched

Grave Orientation

Arras I

Female?

Yes

Extended

NorthSouth

Iron mirror, vehicle and fittings, whip

Arras II

N/A

No

N/A

N/A

Iron mirror

Garton Slack

Female

No

Flexed

NorthSouth

Aged 25-30

Lying or right side, head to North

Iron mirror, chalk blocks

Wetwang Slack

Female

Yes

Flexed

NorthSouth

Young adult

Left side, head to North

Vehicle and fittings. Iron pin, iron mirror and bronze container

Wetwang Village

Female

Yes

Crouched

NorthSouth

Aged 35-45

Head pointing South

Vehicle and fittings. Iron morror. Blue glass beads

Position of Skeleton

Grave Goods

Was the Mirror Contained?

Pig Bones

Yes

Possibly in box or container

Yes

Yes

Mirror possibly contained in a bag

Yes

Table 9.2 Mirror burials from East Riding monumentalising the burial places of individual ancestors (ibid) and perhaps acting to legitimate rights to occupy land (Giles 2007: 240). The inhumation burial rite of East Riding is the most distinctive and most archaeologically visible in the British Iron Age. Burials date to between the mid 5th and 1st centuries BC (Bevan 1997: 186; 1999a: 132; 1999c) and can be found singularly, in small groups (e.g. Cowlam North Humberside (Stead 1986) and in large cemeteries such as Burton Fleming (Stead 1976b), Rudston (ibid) and Scorborough (Stead 1975). Over half the square barrows are grouped into cemeteries (Bevan 1999a: 135). The placement of these cemeteries was often closely associated with pre-existing boundaries, track-ways and watercourses (Stoertz 1997: 39; Bevan 1999a; 1999b). Typically a burial consists of a square-plan ditch enclosing a low earth mound or barrow and a central grave. Barrows can vary in width from 3 to 15 metres across (Bevan 1999a: 134). Although square-barrows in Dorset, Hertfordshire and the Lunan Water Valley in Scotland have been found to contain burials, in Eastern Yorkshire they occur in much larger numbers and in a unified area (ibid: 133-5). In addition to the burial of human remains there is evidence for the separate ritual burial of complete animals such as sheep, pigs, ox and deer, especially immature animals (Ramm 1978: 17). The positioning of grave-goods within burials is highly structured and gendered. As Bevan (1999a: 134) states “pots and animal bones are generally placed by the head and chest of female burials and the pelvis and feet of male burials”. In his survey Iron Age Cemeteries in East Yorkshire, Stead (1991c) describes three broad types of inhumation in the region, all under square barrows. ‘Type A’, the ‘normal rite’ is found at Rudston, Burton Fleming (Manby 1980: 43) and Wetwang Slack. It consisted of a crouched body, head at the north end of the grave, lying of its left

59

side, facing east, sometimes in a coffin. Brooches are the most common grave-goods; graves also contain bracelets, beads and joints of meat, usually the upper-left foreleg of a sheep. The meat was sometimes placed inside a pot (Stead 1991c: 179). ‘Type B’ burials are flexed or fully extended, orientated with the head at the east or west end of the grave. The most common grave goods are swords, spearheads, tools, knives and spindle-whorls, as well as the head and forelimbs of a pig (ibid). Type B inhumations have been found at Makeshift, Burton Fleming and Grindale. Like ‘Type A,’ ‘Type C’ burials are also orientated north-south, but the body is usually flexed. Although the sample of Stead’s ‘Type C’ burials is small the grave-goods differ from ‘Type A’ as swords have been found in two of the graves, vehicles in two and also a shield (ibid). Where animal bones are included they are always pig, comprising a half-head and a forelimb. Type C burials have been found at Garton Slack, Wetwang Slack in the smaller clusters of graves to the east and west of the main cemetery, and probably Arras (ibid: 179-80). In the context of this book it is Stead’s ‘Type C’ burials which are most important. As Table 9.2 demonstrates, four out of the five mirrors found in inhumations in this region come from graves orientated north-south and buried with pig bones (the context of Arras II was not well recorded). A mirror was found underneath the head of a woman in a vehicle burial at Arras cemetery in East Yorkshire (Stillingfleet 1846; Greenwell 1906; Stead 1979). The grave was 12 feet in diameter and the mound above it was 14 feet in diameter. The body was found to be extended, this is unusual for the Arras cemetery as most bodies were interred crouched or flexed (Stead 1979: 37). Behind the head were the foreparts of two pigs. “Behind the back were the iron tires of two wheels laid partly the one over the other, and within each tire were two bronze hoops, those of the corresponding naves, and a circular piece of iron. In

front of the face were two bits laid slightly above the bottom of the grave. A small article of very thin bronze (now lost) was also found. It was in shape like a small round box lid, about 1 inch wide and the same deep. Opposite each other were two thin bronze pins to fasten it on the wood, some of which was still there. It was no doubt, as the workmen suggested, the metal cover of the end of a whip shank” (Greenwell 1906: 284-5). The horse bits are made of iron, coated in bronze. A second mirror was found at the Arras cemetery alongside the remains of a skeleton, under a small barrow (ibid: 294). The context of the find was not well recorded and the mirror is now lost. Another mirror was found at Garton Slack (Brewster 1971; 1976; 1980). The burial was found in one of a group of three barrows. To the east of the barrow where the mirror grave was found were a number of pits containing pottery and carved chalk objects. The barrow containing the mirror was square in plan with fairly deep ditches and three burials in total were found within it, one just off centre to the west and the other two in the ditches. The grave just off centre contained a mirror, a skeleton, some chalk blocks and animal bones. It was oval in shape and excavation of it suggested that it may have once contained a coffin, or there may have been a shallow mortuary chamber at the grave bottom accessed through a rectangular wooden shaft (Brewster 1980: 227). The skeleton found in the grave was of a woman aged about 25-30 years of age “…with a high vaulted brow, long narrow face, and a pointed chin, narrow eyes and a broad prominent nose” (ibid: 228). It was flexed lying on its right-hand side, facing east, with the head pointing north. “Originally, the right fingers were clasped together on the breast, the left arm extended down the side of the body and the hand resting extended and inverted beneath the mirror” (ibid). The mirror was raised above the grave bottom, resting on the side of the grave about 20° from horizontal. Immediately below the mirror was an air-space (ibid: 228-9). When the body decayed its position changed and the right arm fell forward. On the upper part of the back and on the woman’s breast were two suckling pigs. A number of small chalk blocks were also found in the grave, one may have formed a ‘pillow’, another was trapezoidal and smooth and resembled figurines found on other parts of the site (Brewster 1980: 228; see also Stead 1988 for other examples of chalk figurines). In the southeast corner of the same barrow was a grave containing the remains of a woman aged around 20. She was interred in a crouched position, lying on her left side facing east with her head to the north. She also had a chalk ‘pillow’. This woman had cranial features strikingly similar to the woman buried with the mirror (Brewster 1980: 229). The other burial in the northwest corner of the barrow was that of an infant in a flexed position on its left side facing east with its head to the north of the grave. At Wetwang Slack a mirror was found in an inhumation containing the remains of a vehicle. This burial was found alongside two other burials (Figure 9.1), both containing the remains of vehicles and male skeletons with swords

60

Figure 9.1 The Wetwang Slack burial (after Dent 1985b: Figure 3) (Selkirk 1984; Dent 1985a). The inhumation containing the mirror was located centrally within a square-ditched enclosure. It contained the flexed skeleton of a young adult female on her left-hand-side, head to the north. Her arms were outstretched and between her arms and legs were two forequarters from a pig (Dent 1985a: 88). In the central area of the grave was a rectangular box (possibly the structure of the vehicle) which covered the skeleton and the grave goods not integral to the structure of the vehicle (ibid: 90). A small group of objects was found behind the head and shoulders of the skeleton on the western side of the grave. These objects included the mirror, an iron and gold pin decorated with coral, two horse-bits and a decorated bronze cylinder, or ‘bean can’, with attached chain (ibid) (Figure 6.3). The other two burials were both buried with swords and vehicles. The first was that of a young adult male lying on his right-side head to the north and knees drawn up towards his waist. Over his body were the forequarters of a pig. An iron sword was positioned parallel to the body. To the left were the remains of a shield and on and around the body were the remains of seven spearheads. There is also evidence that the body was covered by the superstructure of the vehicle (ibid: 89-90). The second grave was more disturbed, however, it contained: “…two wheels [that] had been laid side by side on the floor of the grave. The tyres and naves were

iron and the wooden parts of the twelve-spoked wheels were preserved as iron replacement, as a calcareous crust, or as stains in the gravel filling. The axle also survives as a stain, and above both wheels and axle were discolourations in the filling which indicated a rectangular structure, which may have been the superstructure of the vehicle as opposed to a simple coffin or cist. Inside this and over the wheels was a crouched skeleton, probably of a young male adult, lying with head to north and on the left side. Behind the back a line of four bronze terrets and pieces of bronze sheeting indicated the line of the yoke” (Dent 1985b: 360). No spearheads were found in this grave but it did contain a shield and a very similar sword to the other grave. It has been suggested that the central female burial was the primary burial and that there is no reason to indicate that a long period of time elapsed between the three burials (ibid: 361). The mysterious ‘bean-can’ found in close association with the mirror could have been used to contain plants or animal remains (ibid). A final mirror was found in a burial alongside the remains of an adult female, the wooden remains of a vehicle, horse harness and pig bones at Wetwang Village. The grave was discovered in March 2001 at the top of a ridge at the east end of the village of Wetwang, overlooking the area excavated by John Dent in the 1980s (Hill 2001a; 2002a). The grave was surrounded by a square ditch and covered by a low mound (Figure 9.2). Analysis of the skeleton found in the grave suggests that she was between 35 and 45 years old and around 1.75 metres tall. This is both a good age and is very tall for an Iron Age woman (Hill 2001c). Some years before she died the woman also dislocated her right shoulder, which did not heal properly and the woman would probably have been unable to ride the vehicle she was buried with unaided. The edition of Meet the Ancestors entitled ‘Chariot Queen’ (BBC 2001) builds a sequence of what was uncovered in the grave. First it was the horse harnesses. Next came the bronze harness rings with coral inlays. Then the presenter of the programme, Julian Richards, excavated an iron ‘tyre’ from the vehicle, when he uncovered another ‘unexpected lump of iron’ in the middle of the cartwheel, it was found to be a linch pin. Building the tension, Richards exclaims that just when he thought all of the metal objects had been found they uncovered the horsebits. The next day they revealed the vehicle’s axle and a few minutes later the first bones were uncovered. As already indicated, the skeleton was not as expected the remains of a male, but a female. Pig bones were also uncovered on the torso of the skeleton. One of the last things to be uncovered was the iron mirror. The mirror lay across the lower legs of the skeleton. A number of small blue glass beads were also discovered (Hill 2001a: 2). Perhaps these were strung and the string used to secure a bag, traces of which have been discovered in the iron corrosion products of the mirror plate. The pig bones were placed over the woman’s torso, which is similar to other vehicle burials. The body was orientated north-south with the head at the southern end of the grave facing west and lying on her left side. The vehicle

61

was dismantled with the wheels at her feet; the Figure 9.2 The Wetwang Village grave (after Hill 2001c: Figure 1) horse bits, terrets and strap unions lay to the north of the wheels. The horse bits are inlayed with glass and the terrets and strap-unions are decorated with coral (see Stacey 2004). Later analysis of the chariot revealed that it had been repaired, suggesting that it was not specially constructed for the burial. Of the five mirrors three were found in association with vehicles: Arras I, Wetwang Slack, and Wetwang Village. To date fifteen reasonably well-recorded vehicle burials have been found in Eastern Yorkshire (Hill 2001a: 2) (see Table 9.3). Two other vehicles have also recently been discovered at Newbridge in Scotland and Ferry Fryston, West Yorkshire (Carter & Hunter 2001; Brown et al 2007). Many vehicle burials were excavated in the 19thcentury and we do not have good contextual information (Stead 1979). Where we do have reliable data it is clear that the majority fit into Stead’s ‘Type C’ burials. Of the fifteen, nine burials contained the remains of pigs and ten were orientated north-south. With the exception of the burial at Cawthorne Camps, the vehicles were dismantled before they were put in the grave. Apart from Danes Graves, vehicle graves contained the remains of only one individual and with the exception of the King’s Barrow at Arras, and the vehicle burial from Hunmanby, the remains of the horses that presumably powered the vehicles were not interred. Except for the remains of the vehicle itself (including iron tires) and the remains of pigs, vehicle graves commonly contained vehicle and horse harness fittings, most usually consisting of two horse bits, five rein rings and two linch pins. It is possible that these fittings also ‘stood for’ the missing horses. There is

Skeletal Position

Grave Contents

Pig Bo nes

NorthSouth

On back, head NE, arms crossed

Tyres, nave-hoops

Yes

Yes

NorthSouth

Head pointing north, lying on shield

Shield, bronze disc, boar's tusk, deer antler

No

Yes

NorthSouth

No

No

No

Name/ Location

Sex

Mirror

Horse

Vehicle Dismantled?

Grave Orientation

Arras King’s

Male

No

Yes

Yes

Arras Charioteer

Male

No

No

Arras Lady’s

Female

Yes

Cawthorne Camps

Crou ched

Description

Iron mirror, two iron tyres, two horse bits, a whip

On left, head to the north

Iron tyres at western end of grave. Vehicle and horse fittings of bronze, iron and coral

Yes

Facing east

Two iron tyres, two bits a whip and harness fittings

Yes

Yes

Lying on left side, head to the north

Vehicle and horse fittings. Wheels placed on western side of grave

Yes

Yes

Yes

Yes

2 Males

No

No

Yes

NorthSouth

Contrac ted

Garton Slack

Male

No

No

Yes

N/A

Crou ched

Garton Station

Male

No

No

Yes

NorthSouth

Fle xed

No

Yes

Yes

No

No

Yes

Stanwick

No

No

Yes

Iron sword, bronze scabbard, vehicle and horse fittings

WESTWOOD

No

No

Yes

Vehicle and horse fittings

Kirkburn

Male

Yes

Vehicle

Danes Graves

Hunmanby

Body Covered by Vehicle Shell

30, 5 feet 9

Two horse teeth, two tyres, lynch pins, rein rings, one horse bit, a shield NorthSouth

Fle xed

30 years old

On back legs flexing east, head facing east at north end of grave

Vehicle and horse fittings, mail tunic

WETWANG I

Male

No

No

Yes

North/ South

Crou ched

Yo ung adult

On right side, head to north

Iron sword, shield, seven spear heads, vehicle and fittings

Yes

Yes

WETWANG II

Female

Yes

No

Yes

NorthSouth

Fle xed

Yo ung adult

On left side head to the north

Vehicle and horse fittings, iron pin, iron mirror, bronze container

Yes

Yes

WETWANG III

Male

No

No

Yes

NorthSouth

Crou ched

Yo ung adult

On left side head to the north

Vehicle and horse-fittings, sword and shield

WETWANG VILLAGE

Female

Yes

No

Yes

NorthSouth

Crou ched

Mature adult 35-45, 1.75m

Head pointing south facing west on left side

Vehicle and horse-fittings, iron mirror, blueglass beads

Table 9.3 East Riding vehicle burials

62

Yes

Yes

Yes

evidence that some of the vehicles were repaired and were therefore not made solely for the funerary ritual (see Brown et al. 2007). The horse fittings are made of bronze, iron or a combination of the two, and can be highly decorated and inlayed with glass or coral. Of the fifteen burials eight contained males and three contained females. All of the female vehicle burials contained iron mirrors. Could it be that all vehicle burials containing females had to include a mirror? Just as many of the male vehicle burials contain weapons, mirrors could have been intimately tied to the social roles of the women buried with them, or the task to be accomplished by them in another place. Drawing from the available evidence, a mirror can be buried with a person without a vehicle but a woman cannot be buried with a vehicle without a mirror. South-western Cist Burials Containing Mirrors The Bryher, St. Keverne and Stamford Hill mirrors were all found in cist-burials, which are part of an identified regional burial rite (Whimster 1981: 72; Cripps 2004; 2007), which includes isolated burials and small cemeteries (Quinnell 1986: 118) found in Cornwall (Whimster 1977), the Isles of Scilly (Thomas 1975; Ashbee 1979) and two sites in Devon (Johns 2006: 61, Fig 39). The burial rite is characterised by short cist graves lined with stone, dating from the 3rd-century BC – 1st-century AD (Quinnell 1986: 118). The human remains tend to be crouched, placed on their right sides with head to the north (Whimster 1977: 82). Burials sometimes contain metal grave-goods of which mirrors are a known type (Johns 2006: 62). Unfortunately only two of the cemeteries containing cist burials have been excavated in recent times (ibid: 61): Porth Cressa on the Scilly island of St Mary’s, where 10 cist graves were excavated (Ashbee 1979); and the cemetery at Harlyn Bay in north Cornwall (Whimster 1977). At Harlyn Bay 130 cist-graves were excavated dating from the 2nd-century BC – 1st-century AD (ibid). A stone-built structure was also found at the site with the burial of an adult and a child under its foundations. Whimster (ibid: 68-70) suggests that this building could have been associated with the funerary rites performed at the site. Interestingly, Whimster (ibid: 82) has demonstrated that the number of people buried at the Harlyn Bay cemetery could not have supported a sustainable population, this means that only certain individuals living in the area at the time were selected for this particular burial rite and it is therefore not normative. Another contemporary, but as yet, archaeologically invisible burial rite was being employed in the region. The cemetery found at Trethellan Farm, Newquay seems to have been in use for 200-300 years and contained crouched inhumations, the majority of which were orientated northsouth, and dated to the 1st-century BC. Significantly no cist stones were used (Nowakowski 1991) even though cist stones were available nearby (Cripps 2007: 151). Cripps (ibid) has suggested that cist-lined inhumations were a cultural preference in Iron Age Devon and Cornwall. The cist burial rite therefore was not the only way in which Iron Age people were disposing of their dead and this raises important questions about social organisation and the significance of mortuary practice in Iron Age society in Corn-

63

wall, Devon and the Scilly Isles at this time. If as Cripps (ibid: 153) suggests, society was primarily organised at the level of the household, or small social group, we have a situation where some groups of people were making the cultural choice to bury their dead, or certainly some of them, in stone-lined cist graves. Significantly, in the context of this book, mirrors found in this region have only been recovered from cist-lined graves. The deposition of mirrors in these graves is therefore intimately tied to the burial by small household groups of people in cemeteries and the cultural choices and social relationships negotiated at this time. Both the Stamford Hill cemetery (Bate 1865; Cunliffe 1988), located near Mount Batten in Plymouth, and the Trelan Bahow cemetery, near St Keverne (Jope-Rogers 1873) on the Lizard Peninsular, Cornwall were discovered in the 19th-century and are poorly documented. The St Keverne mirror was found in one of a number of stone-lined graves aligned east-west which were uncovered during the construction of a road in 1833 (ibid). The mirror was found in a cist burial containing several glass beads, a bronze brooch with a missing foot, and bronze bracelets (ibid: 267). The burial was reported 40 years after its discovery so details are patchy and many of the grave-goods were lost by the time of publication. Parts of three mirrors were found at the Stamford Hill cemetery; a decorated bronze plate, and two bronze handles, one handle with part of the bronze plate still attached (Franks 1865; Spence-Bate 1871; Clark 1968-72; Cunliffe 1988: Fig. 48-9). All three mirrors were destroyed during World War II. The site was discovered in 1861 during work on the defences of Plymouth and the remains of numerous weather-worn, limestone-lined graves were uncovered (Spence-Bate 1871: 501; Clark 1968-72: 139). No plan of the site was drawn and only partial notes of individual graves were made by Spence-Bate (Cunliffe 1988: 87). One grave contained worn pebbles, fragments of glass and coarse pottery. Another contained iron shears, an iron knife and three bronze rings, whilst a different grave contained pig teeth (ibid). Finally, one grave was said to contain a decorated bronze mirror plate, traces of decomposed bones and a bronze brooch (Spence-Bate 1871: 501). The mirror was placed polished side down in the grave (ibid: 502). It has been suggested that one of the mirror handles was once attached to the decorated mirror plate (Fox 1958: 97). However, since the mirror fragments were destroyed in World War II it is not possible to confirm this hypothesis. There is also no evidence that the plate and handle were found together (Cunliffe 1988: 90). Spence-Bate (1871: 502) explicitly states that the handle without the plate section attached was found on its own and that ‘no plate that belonged to this handle was found’. Perhaps the two components of a single mirror were placed in separate graves. The Bryher mirror was discovered in a stone-lined cist grave (Figure 9.3) by a farmer in 1999, which was later excavated by Charles Johns (2006; no date). In addition to a bronze mirror the grave contained an iron sword and bronze scabbard, a ring from a sword belt, or baldric,

poor preservation of the bones meant that it was not possible to identify the sex of the individual (ibid: 1). An unexcavated cist was discovered 5m southeast of the mirror and sword burial and it has been suggested that the two were contemporary, forming part of a more extensive cemetery on the site (ibid: 40). To the south of the graves, evidence for an Iron Age – Romano-British settlement has been found, including the remains of a stone-walled building and limpet-midden scatters (ibid: 40-4).

Figure 9.3 The Bryher mirror burial (after Johns 2006: Figure 13) bronze fittings for a shield, a bronze brooch and spiral ring and an unidentified tin object (Johns 2006: 15-6). The mirror was found at the north-western end of the grave, handle uppermost pointing northwest. The mirror plate was positioned so that the reflected surface was turned towards the grave occupants’ head. The spiral ring and the tin object were found underneath the mirror. It has been suggested that the mirror was kept in a bag secured by the spiral ring (ibid: 17). The sword was on the western side of the grave, hilt pointing north, and the brooch was located near the feet (ibid: 16). In Iron Age burials containing weapons, swords are often located on the right-hand side of the body. This is presumed to reflect how swords were worn in life (ibid: 17). The position of the brooch in the grave is unusual as brooches are often found at the shoulder, or by the side of the head. The excavator suggested that some form of clothing, with brooch attached was placed at the feet, or a cloak was placed upside down over the body (Hill in Johns 2006: 32). In a grave containing weapons from Deal, Kent a brooch was also found near the left leg (Parfitt 1995: Fig. 3). Stead (in Parfitt 1998: 88) compared the position of the brooch in the Deal grave to the Kirkburn cart burial where a mail shirt was placed upside-down over the corpse (Stead 1991c: 54-6). The practice of inverting things, or turning the clothes of corpses inside out, is commonly noted in the ethnographic literature. By reversing day-to-day practices the realms of the dead and the living are actively separated (Parker-Pearson 1999b: 26). Perhaps the inclusion of a mirror in a grave, with its reflective surface capable of reversing the lived-in world, could have acted along similar lines. The shield was probably broken and placed against the western wall of the cist (Johns 2006: 18). The body was placed in a crouched position within the grave, lying on the right-side, facing west with the head pointing north (ibid: 15). The human remains were from one individual aged 20-25 years at death. Despite attempts at DNA testing the

64

Cremation Burials from Southeast England Containing Mirrors The vast majority of mirrors have been found in cremation burials (or ‘probable’ cremations) from southeast England. One mirror has been found in Hertfordshire at Aston (Rook et al. 1982); one in Buckinghamshire at Dorton (Farley 1983); four in Bedfordshire, two at Old Warden (Dyer 1966; Spratling 1970a), one at Pegsdon (Hill 2000; Burleigh & Megaw 2007) and one at Bromham; one mirror has been found in Kent at Chilham Castle (Parfitt 1998); one in Hampshire at Latchmere Green (Fulford & Creighton 1998); and seven mirrors from Essex at Rickling (Sealey 2006), Great Chesterford (Fox 1960), two at Colchester (Fox & Hull 1948; Sealey 2006) and three at Billericay (Weller 1974; Sealey 1996). These burials form part of a tradition of cremation burial in the southeast of England during the late Iron Age (see Fitzpatrick 1997b: Fig. 115), concentrated in Kent, Essex, Hertfordshire, Bedfordshire and Cambridgeshire (Foster 1986: 179) and consisting of individual burials and burials in small groups, or in larger cemeteries. Several hundred cemeteries and individual cremation burials have been discovered. Most were poorly recorded 19th and early 20th-century finds (Hill et al. 1999: 264). A tradition of cremation burial in the later Iron Age in the southeast of England has long been identified (Birchall 1965; Stead 1967; Hawkes 1968; Stead 1976a). These burials are called ‘Aylesford’ (Stead 1976a) or ‘Aylesford-Swarling’ (Birchall 1965), after two cremation cemeteries excavated in 1886 and 1921 respectively (Cunliffe 1991: 132). Cultural historical interpretations of Aylesford cremation burials viewed them as evidence for Belgic invasions and their settlement in southeast England (e.g. Birchall 1965; Hawkes 1968). Consequently, parallels were sought with burial traditions on the continent across northern France and Belgium. More recent interpretations of the burial rite have looked for explanations as to why people were buried in this way with particular sets of grave goods and what this could have communicated to others about the people buried with these grave goods and those who buried them (Hill et al. 1999). Aylesford cremation burials are characterised by flat graves and contain distinctive wheel-thrown pottery types (Birchall 1965; Stead 1991b: 591). The simplest burials consist of a pile of cremated bones with no accompanying grave goods. In others burials the cremated bones were placed in urns and the grave may also have contained an additional pot. Many Aylesford cremation

burials are grouped in small cemeteries. Larger cemeteries have also been found, especially in association with ‘oppida’ such as the Lexden cemetery at Colchester and the King Harry Lane cemetery at St. Albans (Stead 1991b: 592). In addition to pottery and brooches other objects are found, including razor knives and toilet instruments. Weapons are almost completely absent from Aylesford graves (ibid). Some graves cluster in groups around a central burial and may be contained within a square or rectangular enclosure (e.g. at King Harry Lane). The graves surrounding the central burial are often termed ‘satellites’. Stead (1976a: 412) suggested that a number of burials stand out because of the quantity and quality of the grave goods. He divided these burials into two groups: an earlier ‘Welwyn’ phase dating from 50-10 BC (Hill et al. 1999: 262); and a later ‘Lexden phase’ which include distinctive brooch types and Gallo-Belgic pottery that Welwyn phase burials lack. Lexden burials probably date from c. 15/10 BC – AD 50+ (ibid). The ‘Welwyn Phase’, named after a cremation burial found at Welwyn in Hertfordshire (Stead 1967) is most relevant in the context of this thesis as some mirror burials, most notably Dorton, are said to fit into the group of Welwyn burials and the rest all date to this earlier phase as none contain Gallo-Belgic pottery (see Chapter 8). Stead (ibid: 44) defines this type of burial as comprising a “…cremation in a large rectangular grave without covering mound. There is always a quantity of pottery, including at least one amphora, and usually some imported metal or glass vessels”. What marks Welwyn graves out from more common burials, however, is the quantity of pottery found in the graves. The cremated remains in these burials were not placed in urns but may have been contained in bags. ‘Welwyn-type’ burials tend to be isolated although some are known from cemeteries (Fitzpatrick 1997b: 228). Some Welwyn burials also include fire dogs, bronze-bound caskets and sometimes wooden litters (Foster 1986: 187). Because of the number of grave goods Welwyn type burials have been interpreted as high-status and the burials of ‘chieftains’ (see Fitzpatrick 1997b: 211). The majority of mirror burials from southeast England fall into a ‘second tier’. These contain fewer and different objects such as mirrors, buckets and silver brooches (Hill 2007: 30). Hill (ibid) has argued that a grave containing a decorated mirror, or imported wine amphorae, may not necessarily constitute a ‘rich’ grave. He has questioned the input of labour required to make a mirror, or produce a surplus to obtain imported wine. The mirrors found at Aston, Chilham Castle, Dorton and Latchmere Green were excavated relatively recently, providing rich contextual data. The plate of the Aston mirror was found on the surface of a field in 1979. In 1980 the field was investigated, the mirror handle was recovered, and a cremation burial was discovered nearby (Rook et al. 1982). The cremation burial consisted of a shallow pit, containing two pottery vessels located on the western side of the grave (Figure 9.4). Cremated human bone was found in the southeastern corner, with pieces of cranium ringing the southeastern wall of the grave. It has been suggested that the mirror was positioned in the north-eastern quarter of the

65

Figure 9.4 The Aston burial (after Rook et al. 1982: Figures 3-4) grave. The damage to the mirror caused by ploughing indicates that it was placed nearly vertically within the grave with the plate uppermost (ibid: 19). The crematedhuman remains were of an ‘older, but not old,’ human adult. There was no bronze staining on the bones, indicating that they were probably not in contact with the mirror, and the sexually significant bones were missing (Stead in Rook et al 1982: 19-20). S. Stead did, however, suggest that the cremated bone could have been female and she stated that not all of the cremated bone was present in the burial. The two pots from the burial were both handmade and their form is similar to pots found in other cremation burials north of the Thames (Rigby in Rook et al 1982: 23). The burial was isolated and no other graves were found in the immediate vicinity, although a substantial ditch containing post-conquest pottery was found nearby. The grave was also located within 1km of the Roman road from Verulanium via Welwyn to Puckeridge (Rook et al. 1982: 18). It was also placed on the shoulder of a valley, which is now dry for most of the year, but forms a tributary of the River Beane (ibid). The Chilham Castle burial was located on the North Downs of East Kent at 105m OD, almost on the summit of a chalk ridge, only 50m southeast of a track-way called the Pilgrim’s way and overlooking the River Great Stour (Parfitt 1998: 343). The mirror was discovered by a metal detectorist in 1993 and the find-spot was later excavated (ibid). The cremation burial consisted of a small oval pit and in addition to a decorated bronze mirror it contained the fragmented remains of a cinerary urn, a single unworked prehistoric flint flake, two copper-alloy brooches and cremated human bone (Figure 9.5). The grog-tem

brooches form a matched pair and one was found above and the other possibly below or to the side of the mirror. These may have been used to secure some kind of clothcovering around the mirror (ibid: 345). Investigation of the surrounding area suggests that this was an isolated burial (ibid).

Figure 9.5 The Chilham Castle burial (after Parfitt 1998: Figure 2) pered urn, damaged by ploughing, contained the cremated remains of a single human adult, possibly female, aged under 30, and was located on the eastern side of the grave. On the western side of the grave the mirror was placed horizontally with the handle pointing northwest. The two

Figure 9.6 The Dorton burial (after Farley 1983: Figure 3)

66

In 1977 a mirror was found during the excavation of a gas pipeline at Dorton, Buckinghamshire and the remains of a cremation burial were later discovered and excavated (Farley 1983). The burial was located on the northeast side of Dorton Hill, a north/south ridge, 103m OD on a small spur of land facing northeast. The site offers extensive views over the Vale of Aylesbury and is visible for up to 12 miles, although there was no trace of a mound or ditch demarcating the site (ibid: 269). Careful excavation of the grave fill suggested that a short time period elapsed between the grave being dug and filled-in, enough time for it to rain or drizzle (ibid: 271). The burial consisted of a shallow bowl-shaped grave containing three amphorae, two double-handled flagons, a bronze mirror contained within a wooden box alongside cremated bone, an iron band a bi-conical cup and eight pieces of timber (Figure 9.6). The mirror lay horizontally within the wooden box at the northern end of the grave, decorated side down,

with the handle pointing east-northeast. The cremated remains were loose within the box. The interpretation of the iron band is unclear, in other cremation burials in the southeast where two iron bands have been found it has been suggested that they represent the remains of a wooden bucket, however, in combination with the wooden remains found in the grave the excavators suggest that it could have formed part of a bier, or a structure within the grave (ibid: 277). Other ‘Welwyn Type’ burials contained structures, variously interpreted as ‘coffins’, ‘litters’ or ‘couches’, such as the burial from Snailwell, Kent, which contained a wooden couch (Foster 1986: 183). The pottery vessels were located on the northern and western extremities of the grave. Timber was arranged in the southern and eastern portions. Of the three amphorae one possibly originated from Spain, another was of Dressel 1B type, characteristic of the second half of the 1st-century BC and the other was Dressel 2-4, dating from c. 16BC-AD149 (Williams in Farley 1983). All three amphorae show signs of deliberate secondary re-usage. The flagons are unusual. One can be compared to a flagon found at Welwyn Garden City, Hertfordshire. The second is similar to a pair found in a central burial at the King Harry Lane cemetery, St Albans (Farley 1983: 292). Farley (ibid) suggests that these types of vessel probably originated from the Massif Central, France. The carinated cup is grog-tempered and is larger than similar vessels found in the Chilterns. No close parallels have been found in Buckinghamshire, but it is similar to some other fine pottery found in burials in Hertfordshire and Essex (Thompson in Farley 1983). Only a small proportion of the

Figure 9.7 The Latchmere Green burial (after Fulford & Creighton 1998: Figure 2)

67

human remains are represented. No sex or age determination was possible but they comprise the remains of a single adult (Rogers in Farley 1983). The Latchmere Green mirror was discovered by a metal detectorist in 1994 and the find site was later excavated (Fulford & Creighton 1998). The mirror was found to have originated from a shallow pit containing a pedestal jar lying on its side, orientated east-southeast by westnorthwest (Figure 9.7). The jar contained cremated remains, as well as two fragmented iron brooches. No other Iron Age features were found nearby and it was an isolated burial. However, the grave does lie at the southern extremity of a Roman settlement at the junction of Roman roads leading from Silchester, Winchester and Chichester (ibid: 331). The grave was located on the south-facing slope of a low hill 75m OD overlooking several rivers, including a tributary of the Thames. The cremated remains found in the jar included the remains of two individuals, an adult and a young child, as well as animal bone. The adult was aged over thirty and the infant less than five years (Firth in Fulford & Creighton 1998: 338). The presence of the cremated remains of two individuals in one vessel, usually consisting of a younger infant or sub-adult and an older adult, is not common but is well known (ibid). For example, at Westhampnett, West Sussex around 2% of the cremations contained two individuals (McKinley in Fitzpatrick 1997: 69-70). The animal bone represents the cremated remains of the right side of a pig. This is unusual in the context of ‘Welwyn type’ burials but is known from King Harry Lane and Westhampnett (Clark in Fulford & Creighton 1998: 338). The mirror is assumed to have been placed over the top of the pedestal jar containing the cremated remains. Details of the rest of the mirrors probably originating from cremation burials in south-eastern England are less clear, either because no context has been recorded or they are finds from the 19th-century. Looking at the mirrors from Bedfordshire, the Pegsdon mirror was discovered by a metal detectorist on 12th November 2000. Also uncovered was a silver brooch and thirty pieces of pottery from a pedestalled urn, a flat based jar and other vessels (Hill 2000: 15-6; Burleigh & Megaw 2007). The find spot occupies a prominent location in the landscape at around 130m OD. The surrounding area was a site of human occupation and activity from the Neolithic – Roman period and beyond. Examination of the find spot revealed four more sherds from the same pot and one fragment of calcined bone, probably human, indicating that the mirror very probably originated from a disturbed grave (Burleigh & Megaw 2007). It is not possible to determine if the cremated remains were interred in the vessel or on the floor of the grave, in an organic container, or uncontained. The mirror was found in contact with the chalk base of the pit. The silver brooch was one of a pair. They were once attached with a chain, which is now missing. It is not known if the second brooch was originally interred in the grave, or was detached at some point prior to deposition. Silver brooches are rare finds in Iron Age Britain and only about 10 are known, including a pair of

silver brooches from Great Chesterford, Essex (Hill 2000: 16). The Bromham mirror was also a metal detectorist find. It was discovered in a ploughed field during Easter in 2005 and was very fragmented. No other objects were found but it is assumed that the mirror also originated from a cremation burial destroyed by deep ploughing. Two other mirrors have been reported from Bedfordshire, both from Old Warden. Old Warden I is a complete bronze mirror with a decorated plate and it was discovered sometime in the mid-1850’s by labourers during the construction of a rail tunnel (Spratling 1970a: 9). It is reported that large pieces of amphorae and Roman coins were also discovered, these were later sold and the mirror was considered ‘worthless’ (Way 1869: 71). The finds perhaps indicate that the mirror originated from a grave (Fox 1923: 99; Spratling 1970a: 9). A second mirror from Old Warden has also been reported. It comprised a handle and a plate fragment but is now lost (Dyer 1966: 55-6). The mirror was found along with a ‘group’ of Iron Age objects on Quint’s Hill, an earthwork north of Old Warden church, sometime around 1845 (Fox 1923: 98; Dyer 1966: 55). The objects were described as ‘two hoops of iron’ and urns of large size with long handles (amphora?) and two inverted cordoned vases containing ashes (Dyer 1966: 55). They possibly originate from a cremation burial. No details of the depositional context of the Great Chesterford mirror are available, except for the suggestion by Fox (1960: 207) that the mirror is likely to have originated from a grave. Farley (1983: 288) provides more information on the mirror which had hung on the back of a tractor for six months after it was discovered during ploughing. In the immediate area of the find site only a quarter stater has been found (ibid). There is evidence of cremation burials in the area such as earthenware vessels and the fragmentary remains of a brooch, although the exact location of these graves is uncertain (Fox 1923: 98). One of these burials is said to date to Stead’s ‘earlier’ phase of south-eastern cremation burials (Stead 1976a; Farley 1983: 288). Of the Colchester mirrors one fragmentary mirror was found at Colchester in 1974 and was a chance find from Hyderabad Barracks (Colchester II) when contractors were digging the foundations for a new building (Sealey 1996: 61; 2006). The find spot was on flat ground 32m OD. No other Iron Age finds were found nearby. The mirror from Lexden Grange (Colchester I) in Colchester was found in a burial in 1904 (Fox & Hull 1948: 135). This was a single burial separated both from the cemetery at Lexden and the site of Camulodunum (Hawkes & Hull 1947). In addition to a decorated bronze mirror, the burial contained a small bronze cup, part of a bronze pin, a pedestalled urn, a narrow-mouthed cordoned vessel, two cordoned bowls and two tall flagons (Fox & Hull 1948: 135-7). The three mirrors found at Billericay all look to have originated from burials. Billericay I & II comprise a bronze handle and a fragmentary, decorated bronze plate and a smaller, plain circular mirror with a looped handle. They were found along with a collection of British and Roman pottery in 1860 and probably originate from cremation burials (Smith

68

Figure 9.8 Contents of grave 13, King Harry Lane (after Stead & Rigby 1989: Figure 91) 1909: 338). The third mirror found at Billericay (Billericay III) was found within a burial group spanning four centuries (Weller 1974). In 1996 a fragmentary decorated mirror plate and handle was found by a metal detectorist near Rickling (Sealey 2006). The find spot was relatively high up at 105m OD on part of a low, chalky plateau that dominates the surrounding landscape. No other Iron Age artefacts were recovered from the find spot; however, there is evidence for activity nearby. At Clavering, 1.3 km to the west, Iron Age pottery has been discovered. 5.5 km north-west some Iron Age pits have been found which contained pot and amphorae sherds (ibid: 13). At King Harry Lane (KHL), St. Albans, six mirrors of Roman type were found in a large cremation cemetery, in graves 9, 13, 66, 138, 222 and 325 (Stead & Rigby 1989: 103). These mirrors are smaller and more highly tinned than British mirrors (see Farley 1983: Table 1). The majority of these were deposited in graves long after the deposition of British type mirrors in southeast England. However, Grave 13 is from an early phase at KHL and its deposition could date to the second quarter of the 1st century AD (see Chapter 8). The grave contained the cremated remains of an adult ?male (Figure 9.8), two brooches (one iron the other brass) and three pots: a pedestal cup from northern Gaul; a collared Lagena and a honey pot, both from northern Gaul or Lower Germany (Stead & Rigby 1989: 278).

Figure 9.9 The Portesham burial (after Fitzpatrick 1997a: Figure 2) ‘Western’ Mirrors Geographically the ‘western’ mirrors are found outside the southeast of England; in Dorset, Devon, Gloucestershire, Northamptonshire, and beyond. They are thought to originate mainly from graves. The western mirror school grouping is based purely on similarity of form and deposition date (Chapter 8). A number of other parallels can also be drawn. The mirrors from the group that do have a recorded context often come from ‘rich’ burials at Portesham, Dorset (Fitzpatrick & Woodward 1995; Fitzpatrick 1997a) and Birdlip, Gloucestershire (Bellows 1881; Staelens 1982). The Desborough mirror because of its condition is also assumed to originate from a burial (Smith 1909). The western mirrors are deposited later in general than the southeastern mirrors and date to the middle decades of the 1stcentury AD (see Chapter 8).

It is possible that the remains of the mirror grave were contained within a barrow, which with its position would have been visible for miles (Staelens 1982: 21). A fourth grave was discovered a few days after the initial find. This also contained the remains of a male whose face was covered by a wooden or leather bucket (ibid) and also contained a sword. This burial may be contemporary with the mirror burial and was orientated east-west with the head at the east end. Burials dating to the Bronze Age have also been recovered in the immediate area of the 4 graves and it was probably the site of a cemetery (ibid: 19). In the wider regional context, stone-lined inhumation burials aligned east-west have also been found in Gloucestershire, at Hailes (ibid: 28) and a number of undated examples with a similar orientation have been discovered elsewhere in the county.

The Birdlip mirror was discovered at Barrow Wake in 1879 when a quarryman came upon 3 skeletons in a line (Bellows 1881; Staelens 1982: 19). The graves were lined and covered with slabs of whitewashed limestone. The two extreme burials were of males and contained no grave goods. The middle burial is said to have contained the mirror alongside the remains of a woman. The grave also contained a large bronze bowl, a brooch, bangle, beads, tweezers, bronze rings and a small bronze bowl (Staelens 1982: 21). The bronze bowl was placed over the face of the woman (for pictures see Bellows 1881: Pl. XIII), however the position of the rest of the grave goods was not noted. The position of the grave commands a wide view over the Severn Valley.

The Desborough mirror was discovered in a field during ironstone digging in 1908 (Smith 1909). The find was made on high ground in a field to the west of Desborough, Northamptonshire (ibid: 329). A brooch was also found in the same field. It is speculated because of the complete nature of the mirror that it originated from a burial (ibid: 338).

69

The Portesham mirror was found in what is known as a Durotrigian, or south Dorset, inhumation burial (Fitzpatrick & Woodward 1995; Fitzpatrick 1997a). The grave was sub-rectangular and aligned east-west (Figure 9.9). It contained the flexed, or crouched, remains of a ‘probable’ female aged 26 > 45yrs with her head at the

northeast end of the grave lying on her left side. The grave contained two brooches found at the shoulders, presumably securing some form of dress. Placed over the front of the woman were the remains of a pig and at her waist and behind her back were joints of lamb. The mirror was placed, decorated face upwards, on the woman’s waist or chest, a brooch was pinned through the terminal loop of the handle. Brooches are not uncommon finds in Durotrigian burials and all three date to the time of the Roman conquest. A toilet set, consisting of two pairs of tweezers and an ear scoop, was found at the woman’s waist and may also have been secured to the terminal loop of the mirror handle. A copper-alloy pan, a bowl and a jar were also placed behind the woman’s neck. The carinated bowl is the single most popular grave good found in Durotrigian burials (Fitzpatrick 1997a: 68). The high-shouldered jar is rarer but proportionally they are likely to be found in burials. The grave also contained an iron knife. The copper-alloy pan is of Roman type which is usually accompanied by a bronze strainer, although no remains of a strainer survived in the Portesham grave. The function of the pan and strainer is not known, however, it may have been used to strain drink or in other cooking activities (ibid: 58). Durotrigian, or south Dorset, burials are a recognised burial rite found throughout the modern county of Dorset, except in the north, from both before and after the Roman conquest (Whimster 1981; Fitzpatrick 1997a: 65). The most well-known burial groups have been found at Litton Cheney (Wacher 1958), Whitcombe (Aitken & Aitken 1990) and Maiden Castle (Wheeler 1943; Whimster 1981), as well as at Poundbury (Woodward 1993). In general the dead were buried in a crouched or flexed position on their right sides, with the head orientated east. Grave goods occur in about half of the graves (ibid). Pottery and joints of meat are the most common grave goods but brooches are not uncommon. Joints from pigs are usually placed at the head, sheep placed on the chest, in the hand, at the pelvis or feet. Joints of mutton are found with both sexes, pork joints are usually found with females (Fitzpatrick 1997a). The Portesham grave is one of the richest Durotrigian inhumations. The grave of a 25-30 year old male, containing a brooch, sword, scabbard, baldric, spear, hammer and spindle-whorl was found at Whitcombe (Fitzpatrick 1997a: 68). A burial with a horse and horse-bit was also found at Fordington (Woodward 1993). A mirror handle was found at West Bay, near Bridport, Dorset (Fox 1949; Farrar 1956). The mirror handle was discovered after a cliff fall exposed the remains of a burial containing two individuals, a middle-aged man and an elderly woman. Alongside the mirror handle and skeletons were a pottery vessel and ‘a whole lot of bronze objects’ (Farrar 1956: 90). The jaw bone of the elderly woman was stained by a bronze object, possibly the mirror (ibid), indicating possible association between the mirror and the head. The site itself also produced evidence of late Iron Age and early Roman occupation (ibid: 94). Two mirrors have also been found in graves at Portland. In 1875 a mirror handle was discovered at the Grove in Portland in one of more than 200 cist-graves. The graves were said to have been orien-

70

tated north-south and date from the 1st – 4th-centuries AD (Smith 1909: 336-7; Whimster 1981: 258). Another burial at the Verne in Portland contained a fragmented, decorated mirror plate. In the immediate area a series of inhumations in cists have been uncovered (Whimster 1981: 258). At Bulbury (or Belbury) in Dorset the top half of a mirror handle and a small piece of decorated plate was found in 1881 (Cunnington 1884; Cunliffe 1972). Bulbury is a hillfort sited on a spur of land overlooking the River Sherford which flows into Poole Harbour (Cunnington 1884: 114; Cunliffe 1972: 293). The mirror and a series of other objects including two bronze cast figures with bull’s head and horns, two small bronze ornaments hollow cast, two large bronze rings possibly from a snaffle bit, three smaller bronze rings very corroded, part of a sword hilt, part of a bronze chape from a sword-scabbard, a bronze tankard handle in two fragments, bronze bindings probably from a wooden tankard, a bronze bowl/s in five fragments, another decorated bronze fragment possibly from a bowl, an iron anchor and chain, an iron bar, an iron axe hammer and a sledgehammer, a rectangular shaped block of iron, part of a possible iron fire dog and 8 donut-shaped glass beads possibly part of a necklace, were found in the central western side of the hillfort (Cunnington 1884: 114-9; Cunliffe 1972: 294-304). Cunnington (1884: 114-9) also reports that half a quernstone, fragments of burnt pottery, some large nails, a bronze chain and 2-3 rounded/flat pieces of iron were found. These are now missing and cannot be identified with certainty (Cunliffe 1972: 304). Cunnington (1884) suggested that all of these objects came from a hoard. Cunliffe (1972: 306), on the other hand, has argued that the presence of the glass beads, probably belonging to a necklace, could indicate that the remains are from a grave or graves. He suggests that the items could be separated into two graves: the beads, bronze bowls and mirror could have come from a female grave; and the sword, scabbard, tankard and horse-bit from a male grave (ibid). The rest of the objects are said to be part of a ‘blacksmith’s hoard’, however, if this is the case it is surprising that no human bones were mentioned in the report by Cunnington. An unusual mirror with a decorated bronze collar, consisting an iron plate and bronze handle, was found in the Brecon Beacons in south-central Wales (Davis & Gwilt 2008). The mirror came from a grave, which also contained a Roman lamp, a carinated bowl with a lid, which probably contained cremated bone, Roman tweezers and an ear scoop complete with suspension loop and mini bronze terrets, or rein rings. Mirrors from Graves outside Britain Two mirrors have been found outside of Britain, apparently in burial contexts. A decorated bronze mirror was found in a cremation burial (Grave 29) in the ‘Onder Hees’ cemetery, outside the Roman legionary fortress of Noviomagus at Nijmegen, The Netherlands in 1926 or 1927 (Dunning 1928; Spratling 1972: 556). The grave also contained a blue-green glass vessel of a type common in the Rhineland (Dunning 1928: 69). An iron mirror

large chert blocks were found at the bottom of the pit. Nearby lay the bronze mirror which was placed horizontally, decorated side down at the bottom of the pit (Fox & Pollard 1973: 19-20; Pollard 1974: 69). Organic staining and a dense mat of fibrous root found underneath the mirror suggest that it may have been wrapped in cloth or leather (Fox & Pollard 1973: 19-20; Pollard 1974: 69). The pit was one of the earliest of the pit features on the site because it had been cut into by later, shallower pits (Fox & Pollard 1973: 20). The absence of backfill in the pit suggests that it was not specifically dug with the purpose of depositing the mirror. The pit was also too large for this purpose (Pollard 1974: 70). Pollard (ibid) argues against the pit being a grave as no human remains were found and preservation of bones on the rest of the site was good. Although in the report on the Bryher mirror (Johns 2006: 34) this possibility has once again been raised, it seems unlikely that on a site which contains well preserved infant bone, human bone in the pit was overlooked (J. D. Hill pers.comm).

was also found on Lambay Island, County Dublin (Rynne 1976; Raftery 1984: 209-10). The mirror originated from a group of crouched burials discovered in 1927 when improvements were being made to the harbour (Rynne 1976: 232). Contextual information is vague but in addition to the corroded iron mirror an iron sword, shield-boss, five brooches, two bracelets and a beaded bronze torc were also discovered. The description of the find mention that the ‘shield, sword and ornaments’ were ‘found underneath the iron plate’ (ibid: 234, 238), under an iron disc; the iron disc could have been the iron mirror (Raftery 1994: 2001). Therefore although it has been suggested that these grave goods could have come from two separate graves, Rynne (1976: 242-3) suggests that all of the grave goods were found in the same grave. Mirror

Country/Region

Bac Mhic Connain

Scotland

Glastonbury

Southwest

Holcombe

Southwest

Lochlee Crannog

Scotland

Maiden Castle

Southern

Rivenhall I

Southeast

Rivenhall II

Southeast

The mirrors that have been found in settlement contexts tend to be made from a combination of iron and bronze component parts. At Maiden Castle in Dorset a quarter portion of an iron mirror plate, with a bronze rim, was discovered in ‘Biii layer on site E’ (Wheeler 1943: 272). Site E cuts across the innermost rampart, and a section just inside the defences, in the south-western corner of the site (Wheeler 1936: 272). This area contained a hut floor, postholes and an oven as well as a pit and some burnt loom weights (ibid: Pl. 39). A mirror was found at Glastonbury Lake Village in Somerset in 1898 to the southwest of Mound 30 (Coles & Minnitt 1995: 61). It has a plain bronze plate with a very poorly preserved iron handle (Bulleid & Gray 1911: 220-1). A pair of tweezers was found nearby. A second possible mirror handle was also found in Mound 22 in 1893. This has now been categorised as a belt-fastener by Palk (1988). Although the excavators have suggested that the mirror was deposited as a result of casual loss it seems unlikely. Mound 30 is located at the extreme west (Clarke 1972: 836; Coles & Minnitt 1995: Fig. 8.14). The mirror and tweezers found just south of Mound 30 were on the fringes of the settlement in close association with the water surrounding it. There is no indication of human bone being found but the mirror and tweezers could represent a deliberate deposit standing for a burial.

Table 9.4 Mirrors found at settlements 9.3. MIRRORS FROM OTHER CONTEXTS Mirrors found in Settlement Contexts Mirrors are rarely found on settlement sites perhaps because they are not the sort of objects that are likely to be ‘lost’ (Rodwell & Rodwell 1986: 19; Fitzpatrick 1997a: 67). However, this fact suggests that mirrors found at settlement sites are likely to have been deliberately deposited, which makes settlement finds highly significant (see Table 9.4). At Rivenhall in Essex two mirrors were found in close proximity, possibly only 5m apart, nearby a stream and spring (Rodwell & Rodwell 1986: 19). The first mirror Rivenhall I, a bronze handle and fragmentary decorated plate, was found during field drainage in 1846. The context can only be summarised as unknown as this type of mirror is usually found in a funerary context. The second mirror Rivenhall II, a bronze handle, was a chance find in 1954-5 and is more likely to represent a site find (Fitzpatrick pers.comm). At Holcombe, Devon a decorated bronze mirror was found at the bottom of a pit at an Iron Age settlement site found beneath the remains of a Roman villa (Fox 1972; Fox & Pollard 1973; Pollard 1974). The settlement originally contained 4 circular timber huts, later only two were utilised and these were surrounded by a rectilinear ditch (Fox 1972: 293; Fox & Pollard 1973: 18). The site was first occupied in the first half of the 1stcentury AD (Pollard 1974: 70). The pit containing the mirror lay in the south-western corner of the rectangular enclosure (Fox & Pollard 1973: 19). The pit contained four large, unburnished pottery sherds in the fill and looked to have been filled in one episode (Pollard 1974: 69). Two

The Bac Mhic Connain whale-bone mirror handle (Raftery 1984: Fig. 104) was found in a wheelhouse on the north coast of North Uist, in the Western Isles of Scotland (MacGregor 1976: 141). The excavators of this site left behind sketchy reports (Hallen 1994: 224-5). The wheelhouse was occupied during the late Iron Age but the mirror handle lacks a precise context (Cessford 1997: 102). At Lochlee crannog, Dumfries and Galloway, a bronze mirror handle was uncovered during excavations in 1878. Again there is no accurate context for the mirror handle but the site itself was occupied from the Iron Age to the medieval period (MacGregor 1976: Vol. 2, No. 272; Cessford 1997: 101).

71

Mirror

Country/Region

Balleybogey Bog

Northern Ireland

Balmaclellan

Scotland

Carlingwark

Scotland

Compiègne

France

Table 9.5 Mirrors from watery contexts Mirrors found in Watery Contexts Some mirrors have been discovered directly associated with water (see Table 9.5). The bronze mirror from Balmaclellan, Dumfries and Galloway, discovered during the digging of a drain in 1861, was found alongside some decorated sheet bronze, wrapped in three cloth parcels of fine diamond twill (Murray 1860-2; MacGregor 1976: 159-60, Vol. 2, No. 273). The decorated bronze items include the famous crescent-shaped Balmaclellan plate with incised decoration, portions of triangular-shaped plaques, two ornamental studs and bronze belts (Murray 1860-2: 293). These items of bronze were deliberately selected, carefully wrapped in cloth and deposited in a bog along with a decorated stone quern (ibid: 294). A bronze mirror handle was also found by fishermen in 1866 near Fir Island in Carlingwark Loch, Dumfries and Galloway, along with other metalwork (MacGregor 1976: Vol. 2, No. 268; Cessford 1997: 100). It is possible that this find was associated with one of the crannogs in the loch, however, the association between the location of the deposit and water is clear. A bronze mirror handle with a corroded iron disc was also found in Balleybogey Bog, Balleymoney, County Antrim, in Northern Ireland (Jope 1954: 94; MacGregor 1976: 140-1; Raftery 1983: 198; 1984: 209-10). Finally, sometime in the late 19th-century a bronze mirror handle of British type was found in the River Oise in Compiègne, north of Paris, France (Guillaumet & Schönfelder 2001).

Mirror

Country/Region

Akenham

East Anglia

Badingham

East Anglia

Disney

-

Essex/Sussex Border

Southeast England

Gibbs

-

Ingleton

Northern England

Jordan Hill

Southwest England

Llanwnda

Wales

Llechwedd Du Bach

Wales

Mayer

-

Oxfordshire

Southern England

Westerfield

East Anglia

Table 9.6 Mirrors from poorly recorded contexts ever, Smith (1909: 340-1) suggests that there is no reason why the mirror could not have been found elsewhere. Franks suggested that it came from the River Thames, near Barnes (ibid: 341). Three bronze mirror handles have been uncovered recently by metal detectorists in Suffolk and were reported to the Portable Antiquities Scheme (www.finds.org): Akenham (Martin et al. 1998: 209); Badingham (Anon 2002: 75-6); and Westerfield (Anon 2002: Ref. SF6712; Martin et al. 2002). Details of all three finds are sketchy but they are assumed to be singular finds. An additional mirror handle and plate fragment has been reported from the Essex/ Sussex border (Jane 2000) and a complete, decorated mirror from ‘Oxfordshire’ (Burleigh & Megaw 2007). A number of mirrors with very poor contextual details have been found in Wales. A possible mirror handle was uncovered in 1826 at Llanwnda, 15 miles northeast of St. David’s close to the sea. The unusual form of the handle means that its status as a mirror handle can be questioned; however it is difficult to classify it as something else. It was discovered along with some ‘sepulchral remains’ during works to widen a road (Boon 1978-80). A complete undecorated bronze mirror was discovered in the mid 19th-century on the farm of Llechwedd Du Bach (Fox 1925), near Harlech, Gwynedd, on the western coast. It was found along with a platter of bronze, which judging by the circular stain on the mirror plate probably sat on top of the mirror. The condition of both objects suggests that they probably originate from a grave (Fox 1925: 257).

Mirrors with Little or no Context Unfortunately for a number of the mirrors the context was not well recorded (see Table 9.6). Many of these are fragmentary. A bronze mirror handle from Jordon Hill, Weymouth, Dorset, part of the Durden Collection in the British Museum, was found associated with Roman brooches (Smith 1909: 339). The bronze mirror handle from Ingleton, West Yorkshire was found in 1945 two miles south-southeast of Ingleton near Cold Cotes Beck, the circumstances of the find were not recorded (MacGregor 1976: Vol. 2 - No. 269). Two famous mirrors which appear in many early reports (e.g. Smith 1909: 340) are Gibbs and Disney. Both are located at the British Museum (ibid). No context exists for the Gibbs mirror, although since the bulk of the Gibbs Collection comprises Anglo-Saxon material from Faversham, Kent it remains a possibility that the mirror was recovered from Kent (Smith 1909: 340). The Disney mirror was part of the Disney Collection, collected by John Disney (1779-1857) (ibid: 339). A third mirror bearing the name of a collector is the Mayer mirror (Franks 1865). The bulk of the Mayer Collection came from Bryan Faussett, near Canterbury, how

9.4 CONCLUSIONS Mirrors have been discovered in a range of different contexts: graves; settlement; and watery. Mirrors found in watery contexts follow broader regional trends of depositing metal artefacts in watery environments. The remains of mirrors found on settlement sites also represent deliberate depositional acts. The way in which mirrors are made can influence how they were deposited. Mirrors deposited in

72

graves tend to be made of bronze and are decorated. Mirrors found in settlement contexts tend to be made from iron and bronze components. Mirrors are primarily found in graves. Mirror burials are heterogeneous. Mirrors are found in all of the regionally identified burial contexts of the later British Iron Age and individual mirror burials can only be compared on a regional basis. Four distinctive regional groups of mirrors deposited in graves have been identified based on similarities in size (Appendix G), decoration (Chapter 4) and depositional context: ·

·

·

·

slightly later. Mirrors are often deposited with brooches and pottery. The mirrors are decorated, are generally small in size (Appendix G) and may have ‘looped’ or ‘double loop within the grip’ handles (Appendix D). ‘Western’ – These mirrors were deposited most often in graves during the middle centuries of the 1st-century AD. They are large in size and are distributed in southern England, outside of the ‘south-eastern core’.

When they were placed in graves mirrors were part of a complex set of social relationships materialised and authenticated through the presence and arrangement of the grave contents. Relationships were formulated between the person in the grave, the mourners and other grave goods as well as the position in the landscape of the burial. Past relationships – biographies – were probably also referred to in the positing of the mirror and the arrangement of other objects in the grave (see Whitley 2002). It is the relationships expressed by the placement of mirrors in specific deposition contexts that will be examined in the next chapter.

East Yorkshire – Deposited form 400-150 BC in the inhumations of ‘mature’ women. Mirrors are made of iron, they have small circular plates and ‘bar’ handles (Appendix D). Cornwall & The Isles of Scilly – Two mirrors deposited in stone-lined cist graves dating from 120-80 BC. The bronze mirrors are decorated, small in size and have ‘looped’ handles. South-eastern England – Mirrors deposited in cremation graves dating from c.75 – 25 BC, or

73

CHAPTER 10 BIOGRAPHICAL MIRRORS 10.1 INTRODUCTION So far a biographical account of what was written into mirrors both at the time of their manufacture and during their social lives has been provided. In this chapter the contextual information outlined in Chapter 9 is examined alongside the ‘relational socialities’ (see Jones 2002a: 167) established when a mirror was deposited. The deposition of mirrors represents the end of their Iron Age social lives. As such, deposition context provides a unique opportunity to examine mirrors: it is here that we encounter them in a fully ‘intentional’ context, set out by Iron Age actors (Chapter 2). This chapter will concentrate on analysis of funerary evidence as the majority of mirrors were found in graves. When we encounter mirrors in ‘rich’ later Iron Age graves we should examine the ‘richness’ of the relationships their presence signify and objectify. These were intimately tied to relationships formulated between the person in the grave, the mourners and other grave goods, as well as the position in the landscape of the burial. Just as the human actors (living or dead) can be ascribed agency, part of the agency of grave goods (see Gell 1998; Latour 2005) derives from their biographies: the sum of the social relations that constitute the object. Grave goods act to mediate between mourners and the deceased. They do this in a physical sense in terms of the arrangement of objects within the grave, or the relationship of objects to the dead body. They also achieve this biographically by materialising past relationships which are referred to and manipulated through the placement and positioning of objects like mirrors in graves. Throughout this chapter the location and performative rituals of deposition and how they relate to the mirror and its life-history are considered: how was the biography of a mirror referenced in depositional practice? 10.2 MIRRORS AS SOCIALLY POWERFUL OBJECTS Mirrors have been demonstrated to be pivotal objects forming a social nexus (Chapter 7). Rather than passively reflecting the social agency of the individuals they were buried with, mirrors exercise social agency in grave contexts. Mirrors associated with ‘mature’ adults Recent mortuary studies have shown how different kinds of artefacts can be found in association with individuals of particular ages (e.g. Sørensen 1997; Sofaer Derevenski 2000). Identity can change with age (Lucy 2005: 58). Mortuary remains are the furnishings of death rather than simply echoing social relations in life (see Barrett 1994: 116) but they also represent meaningful social categories and identities (Sørensen 1997: 101). Table 10.1 details evidence for the determination of age at death of individuals buried with mirrors. It shows that with the exception of Latchmere Green, which contained the remains of an adult and a child aged less than five years old, all mirror burials were of ‘mature’ adults. The burial of an adult and child in the same burial has been identified in other cremations from southeast England and is not unique to Latchmere Green (see Chapter 9).

74

Mirror

Inhumation/ Cremation

Age Determination

Aston

Cremation

‘Older but not old’

Bridport

Inhumations

‘Elderley woman’ ‘Middle-aged man’

Bryher

Inhumation

20-25

Chilham Castle

Cremation

Under 30

Garton Slack

Inhumation

20-25

KHL Grave 13

Cremation

‘Adult’

Latchmere Green

Cremation

Adult > 30 Child < 5

Portesham

Inhumation

26 > 45

Wetwang Slack

Inhumation

‘Young adult’

Wetwang Village

Inhumation

35-45

Table 10.1 Age at death of individuals buried with mirrors We need to be careful not to impose our own understanding of age construction onto other societies as age identities are social constructions (Lucy 2005) and we do not know how age identities were constituted in Iron Age society. There are also problems and biases associated with the methods employed to determine age at death using human skeletal remains and data can become skewed. Young individuals at Spittalfields, for example (where ages determined using oesteological techniques were compared to the actual age at death ascertained from name plates and burial registers), were consistently ‘overaged’ and older individuals were ‘under-aged’ (ibid: 489). Despite these reservations it is possible to infer from the data, all of the individuals buried with mirrors were at, or past, the sexually reproductive stage in their lifecycle. The association with reproduction could have been emphasised by the inclusion of a child in the Latchmere Green burial. Identity can be intimately related to sexual reproduction, gender and the ways in which perceptions and expectations change throughout the human life course depending upon ‘biological and social points of view’ (Sofaer Derevenski 1997: 876). The Garton Slack, Wetwang Village and Portesham mirrors were orientated towards the sexual organs at the hip or waist of the deceased (Brewster 1980: 228; Fitzpatrick 1997a; Hill 2001a: 2) and the position of some of the mirrors in the grave could indicate an association between mirrors and sexual reproduction, between mirrors and fertility or possibly infertility. Summary Some mirrors were buried with individuals at a particular stage in the human lifecycle, just past when individuals were active in sexual reproduction. It may have been inappropriate to inter a mirror with an individual from a different age category.

Gendering Mirrors: archaeological evidence Despite a lack of corroborative archaeological evidence there has been a widely held assumption in the literature that mirrors were made for and buried with women (e.g. Fox 1958: 84; Finlay 1973: 86; James 1993: 69; Cunliffe 2004: 78). Assigning material culture to particular sex/ gender categories based on our own cultural stereotypes and preconceived notions of gender and sex is highly problematic. Wider discussions of gender (e.g. Butler 1990; Gero & Conkey 1991; Butler 1993) have yet to penetrate Iron Age research (Hill 2006: 171) and the result is that discussions tend to be based on simplistic ethnocentric and androcentric assumptions (Fitzpatrick 1984: 186). Supposition based on pre-conceived ideas is also self-reinforcing (Diaz-Andreu 2005: 38). The association with women and mirrors is drawn from the classical world (Fitzpatrick 1984: 186). Etruscan and Greek mirrors have been linked to women. For example, there are images on cinerary urns showing female figures holding mirrors (Izzet 1998: 211; Echt 1999: Fig. 65). Some Etruscan mirrors were also inscribed to indicate female ownership (Izzet 1998: 211). Mirror burials are also seen as the female counterpart of ‘male’ burials with weapons (see Cunliffe 1991: 509; Cunliffe 1995: 80; Cunliffe 2004: 78-9). Table 10.2 presents the archaeological evidence for sexing of individuals from mirror burials. Unlike the table detailed in Johns (2006: Table 15) it only outlines evidence which can be substantiated. Mirror burials discovered in the 19thcentury said to have originated from female burials will be discussed below. In his table Johns wrongly states that the Dorton and Latchmere Green mirrors originated from female graves. In both cases the sex of the grave occupants was undetermined. Mirror

Inhumation/ Cremation

Sex Determination

Aston

Cremation

‘Probable’ female

Chilham Castle

Cremation

‘Possible’ female

Garton Slack

Inhumation

Female

KHL Grave 13

Cremation

?Male

Portesham

Inhumation

‘Probably’ female

Wetwang Slack

Inhumation

Female

Wetwang Village

Inhumation

Female

Table 10.2 Sexing evidence of individuals buried with mirrors Of the recently excavated inhumation burials containing mirrors, only 3 have been found associated with anatomically female remains: Garton Slack (Brewster 1980: 228), Wetwang Slack (Dent 1985b) and Wetwang Village. According to the excavation report the human remains from Portesham could only be classified as ‘probably’ female (McKinley in Fitzpatrick 1997b: 54). The human remains from Bryher were poorly preserved and DNA tests were inconclusive (Johns 2006: 1).

75

Cremated bone is difficult to sex (see McKinley 1997: 64-5) and it was only possible to sex two of the recently excavated cremation burials containing mirrors. The Aston mirror was found with cremated bone from a ‘probable’ female, although S. Stead (in Rook et al. 1982: 19-20) does state that all of the sexually significant bones were missing. The Chilham Castle mirror was also found associated with the cremated bones of a ‘possible female’ (Parfitt 1998). The sexing of both these graves proved inconclusive and it should be questioned how much the determination of sex was influenced by the presence of a mirror in the two graves. There are a number of mirrors which may have originated from female burials but the documentary evidence is inconclusive, or the claims in the literature cannot be substantiated because the skeletal material is now lost. The mirror from the Lady’s grave at Arras was said to be the burial of a woman (Greenwell 1906: 284-5). The sex of the skeleton found associated with the second mirror from Arras is unrecorded. Unfortunately both of these mirrors were found during the 19th-century and the human remains are now lost. The mirror handle from West Bay was found in the grave of an elderly woman with staining from a bronze object on her jaw bone, this was assumed to have been caused by the mirror (Farrar 1956: 90). However, the remains of a middle-aged man were also found in association with the mirror and the elderly woman. It is not certain that the bronze staining on the woman’s jawbone derives from the mirror; it could originate from one of a number of other bronze objects which are now lost. The Birdlip mirror was found in one of three graves, two containing males and one containing a female. It has always been assumed that the mirror originated from the female burial (see above) (Bellows 1881; Staelens 1982), however, the documentary evidence of the Birdlip burials is poor and it is possible that the mirror could have been associated with one of the two male skeletons (Johns 2006: 70). The three mirrors from Stamford Hill (Bate 1865) and the mirror from St. Keverne (Jope-Rogers 1873) are all said to originate from female burials. However, all discoveries were in the 19th-century and were poorly recorded. As yet no mirror has been categorically proven to have been found alongside the remains of an anatomically male individual, although a Roman-style mirror was found associated with a ‘probable’ male cremation at the King Harry Lane cemetery, near St. Albans (Stead & Rigby 1989: 103). The association between women and mirrors has recently been questioned because of the discovery of the Bryher mirror in a burial which also contained weapons (see Cunliffe 2004: 79; Johns 2006: 70-1). The existence of a mirror and weapons in the same burial cross-cuts perceived gender divisions as mirror burials are traditionally associated with women and weapon burials with men (see Cunliffe 1991: 509; 1995: 80; 2004: 78-9). The Bryher burial is highly unusual as it contained both a mirror and weapons but it may not be unique. The iron mirror from Lambay Island could also have originated from a grave containing a sword and a

shield (Rynne 1976: 242-3; Raftery 1984: 209-10; Raftery 1994: 200-1). As could the remains from Bulbury if the artefacts are interpreted as originating from a single grave, rather than the two suggested by Cunliffe (1972: 362), although based on the artefacts found, it is more likely that these objects originate from hoards. Summary Outside East Yorkshire, the argument that mirrors were buried exclusively with women is not substantiated by the archaeological evidence. We can only say that some mirrors were associated with females but because of the relatively small sample, we cannot be certain that they were not also buried with males. Who ‘Owned’ Mirrors? One interpretation of a mirror found in a grave could be that it was the property of the person it was buried with (see James 1993: 69). Following this interpretation the mirror is socially passive, serving as a provision for that individual in the ‘afterlife’. Whilst it is not argued here that none of the mirrors ‘belonged’ to the individual they were buried with, the information compiled throughout the course of this book illustrates that from their production, mirrors were afforded the potential to exercise social agency. Examination of wear and usage demonstrates they performed in social relations. It is therefore difficult to accept that the social agency of mirrors was not exercised in deposition contexts. It is also important to question why mirrors were not handed-down to the next generation or kept by someone else (see Hunter 2008). Why was a mirror buried at a particular time with a particular person? Was it out of fashion or past its usefulness? Or was there simply no other suitable person in a community for it to be passed-on to? A number of alternative interpretations can be put forward. For example, as discussed in Chapter 6, through their form and methods of manufacture, mirrors can be seen to echo the wider community and its history. A mirror may well have belonged in some way to the person it was buried with without being ‘owned’ in the sense that we would understand it. ‘Ownership’ may also have been intimately related to the biography of the mirror. As outlined in Chapter 3, a mirror may have been perpetually connected with the social relationships formulated during its production and could have been closely tied to the person/s who caused it to be made. These people may not necessarily have been buried with a mirror. The mirror could have been an emblem of a particular social role enacted (or to be enacted) by the person it was buried with. If it was received as a gift then many different obligations and relationships would have been tied to that gift (Gregory 1982; Weiner 1985; see Appadurai 1986a; 1986b; Mauss 1990; Thomas 1991; Weiner 1992 for discussions on gifts and commodities and inalienable objects). We still know little about how Iron Age society was organised. It has recently been argued that society probably operated at the scale of the household or community rather than the individual (Hill 2006: 175). The mirror burial could have been representative of the community or household, expressed through the form of the mirror and its links to the wider community.

76

Mirrors as an Indicator of Status As mirrors are relatively rare and often impressive objects, well made and sometimes beautifully decorated, they have been interpreted as indicators of status (e.g. Cunliffe 2004: 78). This status has been inferred on the individual buried with a mirror. Grave goods are not always good indicators of status in life (Parker-Pearson 1999a: chapter 4). Funerary rituals do not just reinforce the existing social and political hierarchy, they are a forum in which relationships and identities can be renegotiated. Olivier (1999) has shown how some of the objects in the Hochdorf princely grave were probably specially made for the burial, while others were refurbished prior to burial. Were these actions to reinforce the status of the Hochdorf ‘prince’, or to renegotiate the position of those who buried him? We could also question what exactly does Iron Age ‘status’ mean? Although not necessarily explicitly stated, many discussions are still based on the assumption that society was organised into chiefdoms and ruled by a chiefly elite. We do not know if society was especially hierarchical (Hill 2006: Figure 2) and although many interpretations presume the existence of ‘elites’ no one has clearly defined exactly what they mean by the term. As Hill explained: “Iron Age Europe appears on paper to have been full of elites. Yet rarely is it explained who or what an elite was. The term can encompass anyone from a farmer slightly wealthier than many neighbours, through a feudal overlord or a member of a clearly defined aristocratic class in a complex non-industrial empire, to a Roman Emperor or medieval king…” (ibid: 171-2). Although later Iron Age society in southeast England may well have been dominated by ‘kings’ and their political realms (Creighton 2000; 2006), the evidence for this sphere may be overrepresented in the archaeological record and we know very little about what was happening in the rest of the country. We also don’t know how people valued each part of their own society. What is clear is that the organisation of Iron Age societies varied greatly between regions and over time (see Hingley 1984; Hill 1995a; Bevan 1999c; Hill 2006; 2007; Moore 2007a; 2007b; 2007c). Do mirror burials stand out from others in terms of the quantity of grave goods, the existence of imports or the number of artefact types? All of these criteria have been used as markers of ‘status’ in past research (e.g. Haselgrove 1989). In the context of this book, they are regarded as indicators of ‘difference’. As has been shown in Chapter 9, mirror burials do not form a homogenous group which cross-cut localised burial traditions. Therefore it is not appropriate to question the relative status of mirror burials as a group. Mirror burials can only be compared with other burials from the same identified local burial traditions. If we first examine the mirror burials from East Yorkshire we can see that three of the five burials included the remains of vehicles. In East Yorkshire vehicle

burials certainly stand out in terms of the number of grave goods in each burial. They are relatively rare (see Table 9.3) and the group from Wetwang Slack (Dent 1985b), for example, were buried under unusually large barrows, which were set apart from the main cemetery (Giles & Joy 2007: 16). The second Arras and Garton Station burials, however, contained relatively few grave goods. The St. Keverne and Bryher burials contain more grave goods in comparison to other graves from the Cornish cist grave burial tradition. The Bryher burial in particular contained a broad range of different types of artefact, some of which were very unusual. The south-eastern mirror burials, with the possible exception of Dorton, cannot be counted amongst the graves with the highest number of grave goods from that region. These graves have been identified by Stead (1991b) as ‘Welwyn’ type often containing large quantities of pottery some of which are imports. If the mirror is removed from the equation, the majority of mirror cremation burials contain at most a couple of brooches and a small number of pottery vessels. A large number of south-eastern cremation burials contained this mix of grave goods (see Fitzpatrick 1997b). The majority of mirror burials therefore cannot be regarded as being significantly ‘different’: the only indicator of difference is the presence of a mirror. The Portesham mirror burial does stand out from other ‘Durotrigian’ burials because it contains the most number of grave goods of any burial yet found in south Dorset. A number of the objects also indicate links to wider social networks; the pan in the grave was a Roman import and one of the brooches is unusual because it is made of brass, not bronze. The Birdlip mirror could also have been found with a significant group of objects making it standout from other inhumation burials found in Gloucestershire. From this brief overview we can see that some mirror burials are unusual and different from others within the same regional burial rite. The association between mirrors and vehicles in East Yorkshire could be important, although two mirrors were also found in burials which did not contain vehicles. The mirror in this circumstance is not the only indicator of difference. The St. Keverne and Bryher burials stand out from other Cornish cist burials, as do the Portesham and Birdlip mirrors in south Dorset and Gloucestershire respectively. Interestingly, mirror burials from south-eastern cremation burials do not fit this general pattern. If the mirror is removed from the equation then many of the burials are like a large number containing a few pots and possibly brooches. Here we can see the mirror as the artefact that makes a mirror burial stand out, whereas in other regions the mirror is only one of a large collection of objects, which when combined make the burial stand out. Mirrors can therefore be seen not simply to reflect the status of the individual they were buried with (Giles & Joy 2007: 21) they actively generate the social position of that individual, as well as particular individuals, or groups amongst the mourners present at funerary rituals.

77

Summary Mirrors are most often deposited with mature adults. In East Yorkshire there is a proven association between mirrors and mature women but the evidence is not complete enough to comment upon other regions or possible relationships between men and mirrors. Mirrors are not necessarily indicators of the status of the individuals they are buried with. The relationships formed between mirrors, the people they were interred with, and other grave goods, are far more complex. In a grave containing a mirror we can see actors in a particular relationship in still-life. As Parker-Pearson (1999a: 3) stated ‘the dead do not bury themselves but are treated and disposed of by the living’. Funerals are key social and political events in which identity is actively reinforced and renegotiated. The position of a human body in a grave and the juxtaposition of the body with specific chosen objects form a critical part of the funeral, as does the positioning of the grave in the landscape. When we examine the deposition of mirrors in graves we should question why a mirror was selected to be placed in the grave alongside a particular person? What does this tell us, not only of the individual buried with the mirror, but also of the people who conducted the burial and took part in the funerary rituals at the level of the household and on a wider regional level? Grave goods like mirrors played an active role within funerary relationships. As with all grave goods mirrors were put in graves for multiple reasons and their significance is manifold. 10.3 PATTERNS IN THE DEPOSITION DATA Position of Mirrors in the Grave There is little consistency in the position and orientation of mirrors in graves, perhaps because mirror graves should not be viewed as a homogenous whole but occurred within particular regional contexts, following local and regional burial rites. The majority of mirrors were deposited in the ground horizontally. Both the Chilham Castle and Bryher mirrors were deposited horizontally with their handles pointing northwest (Parfitt 1998: 345; Johns 2006: 16). This particular handle orientation could be significant as these are amongst the earliest decorated mirrors. However, there is little overall correlation across the group. The Dorton mirror handle pointed east, the Portesham and Wetwang Village mirror handles were facing southeast and the Wetwang Slack mirror handle pointed in a northerly direction. Where it has been noted, the majority of mirrors were deposited decorated side down including Bryher, Dorton, Holcombe and one of the mirrors from Stamford Hill (Spence-Bate 1871: 502; Fox & Pollard 1973; Farley 1983; Johns 2006). This was not a universal trend as the Portesham mirror was deposited decorated face upwards (Fitzpatrick 1997a). Mirrors were orientated in the grave in respect to the human body rather than to cardinal points; perhaps indicating strong associations between mirrors and the human body, especially the head. As was highlighted in Chapter

6, mirrors are made to be held in the hand allowing one to gaze into a reflective surface. If we examine the orientation of the reflected surface in relation to the person in the grave, a pattern emerges. For example, in the case of the Portesham mirror although it was deposited decorated side upwards, in relation to the body in the grave, the reflective side of the plate was orientated towards the grave occupant. There is a special association between mirrors and the head. The Bryher mirror was positioned so that the reflective side of the plate faced the grave occupant (Johns 2006: 16). At Wetwang Slack the mirror, a pin, two horse-bits and the ‘bean-can’ were located behind the head of the grave occupant (Dent 1985a: 90). The mirror from the Lady’s Grave at Arras was found below the head of the deceased (Stillingfleet 1846; Greenwell 1906; Stead 1979). The Aston mirror was in an upright position on the opposite side of the grave to cremated human remains. There is a clear association with the cremated bones and especially the head as fragments of human cranium were deliberately selected from the cremated bone and positioned at the edge of the grave facing the mirror. In addition to the head, a number of other mirrors have been found with the reflective surface positioned in relation to the waist or hip. At Garton Slack the mirror was supported in the left hand at the hip (Brewster 1980: 228). At Wetwang Village the mirror lay across the lower legs of the deceased (Hill 2001b: 2). The Portesham mirror was placed near the waist or chest with a toilet kit found in close association. It is known that sets of toilet instruments such as tweezers and ear-scoops, often on a ring or brooch (Hill 1997: 98), were possibly suspended in some fashion at the waist (Fitzpatrick 1997a: 56). In the case of the Portesham mirror it could be that as these items were associated with personal appearance, the mirror was positioned similarly as a symbol of its purpose and significance as a ‘technology of the self’ (see Hill 1997) (see Chapter 6).

105m OD (Figure 10.1), overlooking the River Great Stour and only 50m from a prehistoric trackway (Parfitt 1998: 343). The Dorton mirror burial was located on a ridge of Dorton Hill at 103m OD (Figure 10.2), offering extensive views of the Vale of Aylesbury (Farley 1983: 271). The Latchmere Green mirror burial was located on the south-facing slope of a low hill at 75m OD (Figure 10.3). The burial overlooks several river systems including a tributary of the Thames (Fulford & Creighton 1998: 331). The Pegsdon burial overlooked a small stream (Burleigh & Megaw 2007) and Old Warden II was found on Quint’s Hill (Dyer 1966: 55). Finally, the Birdlip mirror was buried on the side of a hill with extensive views of the Severn Valley (Staelens 1982: 21).

Figure 10.1 Location of the Chilham Castle burial (after Parfitt 1998: Figure 1)

If we consider the mirrors found in cremation burials there is a clear association between cremated human remains and mirrors. The Dorton mirror was contained in a wooden box along with cremated human remains. The reflective surface of the plate was uppermost, orientated towards the bone within the box. The Latchmere Green mirror was placed over the top of an urn containing cremated remains. Summary Mirrors are not orientated towards cardinal points but rather their reflective surface is positioned in the grave with reference to the human body. Two key trends have been identified: · Mirrors are positioned with reference to the human head · Mirrors are positioned at the waist or pelvic area Prominent Locations in the Landscape In southern England some mirror burials were found in prominent locations in the landscape, often on the side of a valley overlooking river systems. The Chilham Castle mirror was found high on the North Downs of East Kent at

78

Figure 10.2 Location of the Dorton burial (after Farley 1983: Figure 1) With the possible exception of Birdlip, none of these burials was marked by a burial mound or tumulus. Expe-

ties of cremated human remains (McKinley in Fitzpatrick 1997b: 227); the question remains what was done with the remaining bone? Many of the cremation burials containing mirrors do not contain the full complement of cremated bone from a whole individual. It seems that a part is enough to be representative of the whole. Although none of the mirrors show signs of burning they could have been used in the ceremonies associated with the cremation. As Fitzpatrick (1997b: 228) has pointed out, both the Maldon Hall Farm cemetery from Essex and the Westhampnett cemetery, West Sussex are located on low hills. He suggested that the topographic setting of these cemeteries would have positioned the dead conceptually above the contemporaneously occupied areas on lower ground. This could also be the case with mirror burials.

Figure 10.3 Location of the Latchmere Green burial (after Fulford & Creighton 1998: Figure 1) rience in the present is dependent on an understanding of the past (Connerton 1989). Iron Age people were aware of the past and past peoples (see Stead 1998; Hingley 2009) and they were actively involved in how they themselves would be remembered in the future (Joyce 2003: 105; Van Dyke 2003: 194). Memories can be formulated and shaped by artefacts (Jones 2001) and the landscape (Schama 1995; Campbell 2006). Jones (2001: 349) observed that the efficacy of objects is not confined to the immediate effect they have on people but also those preserved in memory. The memory of mirrors, and the relationships they objectified, could have been preserved in memory alongside the location in which they were deposited. This social memory (Van Dyke & Alcock 2003) could have been perpetuated and retold as people passed the place of burial along terrestrial or riverine route-ways, the prominent location of the burial serving as a visual reminder. The south-eastern mirrors were interred in cremation burials and perhaps the act of cremation in a visible and prominent spot was of more significance than the burial itself. As McKinely (in Fitzpatrick 1997b: 55) points out the act of cremation, burning of human remains, and burial, are often conflated. The two acts are very different and may have carried very different significance. Throughout the Iron Age we see variable treatment of the dead; the remains of the dead are also processed in a series of stages (Lally 2008). During excarnation practices the body was exposed (Carr & Knüsel 1997) and then different parts, at varying stages of deposition, were selected to be deposited. Exact practices vary regionally and temporally and we only see a small percentage of the total quantity of bone in archaeological contexts. Similarly, after the act of cremation, cremation burial is part of a step-by-step process involving the deliberate selection of cremated bone to be deposited in particular contexts (Fitzpatrick 1997b: 227). At Westhampnett many ‘memorial’ burials contained only small quanti-

79

It is interesting that all of the south-eastern burials containing mirrors seem to come from isolated cremation burials. ‘Welwyn type’ burials tend to be isolated although not exclusively so (ibid). Mirror burials from other regions follow regional burial traditions and are found in cemeteries, for example in the southwest. However, despite the fact that cemeteries are known in the southeast burial tradition (e.g. Ayleford and Swarling) none of the mirrors were found in cemetery contexts. The mirror burials which have been well reported at Chilham, Dorton and Latchmere Green are all individual burials in prominent locals. Perhaps it was not appropriate for mirrors to be buried in cremation cemeteries in south-eastern England. Mirrors Deposited in Covering Some mirrors were deposited wrapped in a covering or contained in a wooden box. This is known in the southern Urals where wrapped mirrors have been discovered in Kurgan graves (Morgunova & Khokhlova 2006: 312). Of the mirrors found in cremation burials in the southeast the Dorton mirror was contained in a wooden box alongside cremated human remains. The Chilham Castle mirror was possibly buried inside a cloth bag bound by two brooches. In East Yorkshire at Wetwang Village the mirror was also probably contained in a bag (Hill 2001a: 2). The mirror from Wetwang Slack had impressions of wood attached to it and the Garton Slack burial was also probably contained in a wooden box (Brewster 1976: 110) as there was evidence for the decay of organic remains underneath the mirror on excavation (Brewster 1980: 228). In the Scilly Isles the Bryher mirror was contained in a textile bag, possibly fastened by a spiral ring (Johns 2006: 18). Finally, in the southwest, the Holcombe mirror was wrapped in an organic covering, possibly a leather pouch. Some other mirrors such as Portesham and Pegsdon were found with a brooch attached to the terminal loop of their handles. Visual examination of other mirrors (see Chapter 5) has revealed wear on the terminal loop of handles which would be consistent with a brooch, or some other metal fastener, being secured to the handle and used to secure an organic covering or wrapping around mirrors. Mirrors were not treated in the same way through time and across the country. The coverings could have acted to

protect the mirror surface. A covering could also have acted as a barrier between the reflective surface and the outside world. We know from Chapter 6 that mirrors were powerful objects. They may have been covered as a form of protection from their potent reflections. Mirrors could have been protected in a whole series of different ways which would not have involved damaging the mirror handle. The patterns of wear on the terminal loops of the handles (Chapter 5) suggest that some mirrors were covered in life. The use of brooches to secure a covering could be significant as mirrors can be seen to have been ‘dressed’. The type of covering used and the way it was then interpreted was different for each mirror. For example, wrapping or covering a mirror as part of a funerary rite could be seen as analogous to the preparation of the body after death, wrapped in a shroud secured with a brooch, or placed in a wooden coffin (Giles & Joy 2007: 22). Where there is evidence for a brooch in the terminal loop of a mirror handle but none for a covering, the idea of linkage may have been important. A recent excavation at Manor Farm, Portesham (Valentin 2003) revealed the grave of a young adult (probably female) who was buried with a pair of linked brooches one of which was very similar to the brooch attached to the terminal loop of the Portesham mirror. Mackreth (2003: 42) has drawn attention to the similarity between the style of the brooches and the characteristic of linkage; the Manor Farm brooch was linked to another brooch and the Portesham brooch was linked to the mirror handle. Mackreth suggests that the shared characteristic of linkage was a ‘shared rite,’ although he doesn’t elaborate further. It could be that just as the linked pair of brooches was significant in its use, fastening together the dress of the young woman buried in a grave at Manor Farm; the linked pairing of the brooch and mirror handle could have carried equal significance for the dressing of the Portesham mirror. 10.4 THE SOCIAL RELATIONS OBJECTIFIED WHEN MIRRORS WERE DEPOSITED Using the information set out throughout this book specific mirror contexts will now be examined from the perspective of the potential social relationships they objectify. The majority of examples will be from burial contexts. This is because the quality of information available for the mirrors found in settlement or watery contexts is very often inadequate for this type of analysis. East Yorkshire – Wetwang Village The Wetwang Village mirror was placed in the grave of a mature woman. It was contained in a bag. Nearby lay a number of small blue glass beads. In the cemetery close-by ‘mature’ women were buried with necklaces strung with slightly larger blue beads (Giles & Joy 2007: 26). We know that East Yorkshire mirrors were buried with older women (Tables 10.1 & 10.2). In this circumstance the blue beads and the mirror can be seen to act in the construction of the identity of the grave occupant and clearly relate to age and gender. The grave also included the remains of a vehicle and horse-harness fittings. The majority of vehicle burials were of men in the prime of life who were sometimes buried

80

with weapons and armour, alongside the vehicle and horse fittings (see Table 9.3). Vehicle burials are often physically separated from other graves in a cemetery and some were marked by larger barrows. These graves were also distinct from others as the bodies in the graves were oriented north-south rather than east-west. Only individuals of a particular age category could be buried with vehicles and this was different for both sexes: mature women; and men in the prime of life. In each case objects act to create these identities. We know from examination of their production context (Chapter 3) iron mirrors from East Yorkshire were the least complicated mirrors to make, requiring the fewest number of production stages. It is interesting to note that these mirrors are often found in the most socially complicated deposition contexts. Grave goods may represent particular social roles to mourners present at the burial. These social roles were reinforced by the physical separation of vehicle burials within cemeteries, which were marked by larger barrows, perhaps to make them easily identifiable to later generations. It does not necessarily follow that these were the graves of ‘high status’ individuals and that they indicate the existence of social hierarchies in East Yorkshire. These particular people could have been selected by the community, based on the particular stage they had reached in their lifecycle or because they performed a particular social role within the community. The performance of burial rituals: the procession of the body on the vehicle; the placement of the artefacts in the ground; and the raising of the barrow, could have been representative of a particular community or household. These people could have been buried with vehicles to perform a particular role for the community in ‘another’ place. The social role performed by mirrors, and the mature women buried with them, could have been related to an association between the reflection on mirror plates and the surface of water. The iron the mirror was made from probably originated from the marshy environment of the Foulness Valley. The Wetwang Village burial was also located close to a small pool or mere. We know that some cemeteries in East Yorkshire were located close to seasonal water courses and there is a clear link between water and death, which may also have been associated with the ‘otherworldly’ properties of the reflective surface of a mirror (see Chapter 6). Southwest – The Bryher Mirror Burial The Bryher burial is unusual because it contained both weapons and a mirror. The sword and scabbard were placed on the right side of the body, which is common in an Iron Age context and is presumed to reflect the side swords were worn in life (Johns 2006: 17). The deceased was not wearing the sword but rather it had been placed over the thigh, supported at both ends. The mirror was wrapped in textile and angled so that the reflective surface was orientated towards the head of the deceased (ibid: 16). The mirror and sword were not only orientated in relation to the human remains in the grave, but also to each other as the sword-hilt rested on the bag containing

the mirror. This is interesting because on one level the sword and mirror were positioned so that they relate to the body in a way that is reminiscent of how the objects were used in life, yet the mirror is covered and they are orientated to and in contact with one another. Through the juxtaposition of the mirror and the sword, the objects are not only related to the individual in the burial, they are also clearly related to one another. The Bryher mirror was already old before it was placed in the ground and it underwent a major repair at some point in the past (see Chapter 5). The sword may also have been of significant age before it was deposited. There is therefore potential for these objects to have acquired extensive histories before they were selected for deposition. The juxtaposition of the sword and mirror could then refer to relations between the sword and mirror at some point in the past. The position of some of the other objects found in the grave is also interesting. An idiosyncratic tin object was found lying on top of the mirror. A brooch was discovered at the feet indicating that clothing was placed upside down over the body or folded and placed at the feet. It is unusual, although not unprecedented (see Chapter 9), to find a brooch in this position in a funerary context. If clothing was positioned upside down over the body, the position of the clothing in the grave may have referenced the reflective qualities of the mirror by inverting manners of dress. The individual buried in the Bryher grave almost becomes obscured by the odd combination of grave goods and their novel arrangement in the grave. South-eastern – Aston Mirror Burial In addition to the mirror, the Aston burial contained human cremated remains and two pots. The pots were both handmade and like many other pots placed in cremation graves north of the Thames. What makes this grave stand out from other south-eastern cremations is the presence of a mirror. The mirror was probably positioned in the northeast quarter of the grave in an upright position with the plate uppermost. Significantly, on the opposite side of the grave to the mirror, pieces of human cranium were placed in the walls of the grave. Here we can see a clear relationship between the human head and the mirror specifically referred to in the arrangement of the contents of the grave. As outlined above, this relationship is expressed in other mirror burials. The form of a mirror constitutes the kinds of relationships that can be constructed around it (see Chapter 6). It is designed to be held in the hand so that the face can be viewed free from obstruction. This means that the connection between the mirror and the human head was particularly strong and all sorts of relationships could have been constructed around this association. In some societies people are buried with mirrors to help them pass into the afterlife (Pendergrast 2007). As mirrors are rarely found in south-eastern cremations this is an unlikely scenario. However, it is clear that the reflective qualities of the mirror were highly significant and critical to the decision processes undertaken by the mourners when they arranged the burial. It could be that the individual in the Aston grave was buried with the mirror to perform a particular role in the afterlife

81

or for the community. The social relations objectified by the mirror could also be representative of the community. The burial was located in a prominent location on the shoulder of a valley close to a major route-way. Although there is no evidence that the grave was marked it was isolated from other burials and areas of settlement. This choice of location was obviously significant. We do not know where the actual cremation of the human remains took place but if it occurred near to the location of the grave then the cremation fire would have been visible over a large area, especially to the people who occupied the valley below. It is also possible that the location of the burial was preserved in the social memory of the people who occupied the landscape. Western’ – the Portesham Mirror In addition to the mirror and crouched human remains the Portesham burial contained five pots, pig and sheep bones, three brooches, a knife, a bronze pan and a toilet set. The objects in the grave can be placed into two groups: objects associated with eating and drinking; and objects associated with personal appearance. The pots from the grave were probably used to prepare and serve food. The remains of a pig were placed over the front of the body and at the waist and behind the back were joints of lamb. Joints of lamb are found in the graves of both sexes in south Dorset, but pork is more commonly associated with female burials. The bronze pan is a Roman import and is thought to be part of a pan and strainer set, used for straining and preparing drink, specifically wine (Fitzpatrick 1997a: 58). Two of the brooches from the grave were found at the shoulders and presumably secured a form of dress or mortuary garb, which no longer survives. The third brooch was linked to the terminal loop of the mirror handle. The mirror was placed near the waist or chest, which is where the toilet set was also found. The toilet set consists of two pairs of tweezers and an ear scoop. Toilet sets are more commonly found in Roman contexts (ibid), however, it has been argued that the occurrence of these objects on later Iron Age sites illustrates increased concern for personal appearance and hygiene (see Hill 1997; Jundi & Hill 1997; Hill 2001b). It is significant that both objects associated with personal appearance and dress, as well as eating and drinking, were placed in the grave. Recent research indicates that people in the later Iron Age and early Roman period signalled social difference and actively renegotiated their identity through their appearance (Hill 1997; Jundi & Hill 1997; Hill 2001b). People also signalled their social difference through the preparation and serving of food and drink, by eating and drinking new types of food and drink and using new types of utensils (Hill 2002b; Cool 2006: chapter 16). In the case of the Portesham burial this is particularly apparent with the inclusion of the Roman

pan, which was probably used in the preparation and mixing of imported wine. The meaning of the inclusion of a mirror and toilet set in the grave also seems clear: these objects acted to extend opportunities to express social difference as they allowed people to appear to others in particular ways which they could control themselves. The mirror allowed personal appearance, dress and gestures to be monitored by the individual, offering individual control over how people appeared to others for the first time (Chapter 6). Our knowledge of the biography of the Portesham mirror adds further complexity to the social relations objectified in the burial. A whole series of relationships were formulated when the Portesham mirror was made (see Chapter 3 & Figure 3.2). The form of ‘western’ mirrors like Portesham can also be seen to represent other items of material culture (Chapter 6). However, in the case of the Portesham mirror the associations set out at the time of production have been manipulated and the trajectory of its life has been altered. In the context in which we encounter the Portesham mirror instead of referring to past items of material culture and an idealised past, this association with the past has been turned on its head and the mirror has been used to reinforce changing social identities and relationships: a symbol of past continuity has been manipulated to reinforce renegotiations of social life and identity. Settlement Context – Holcombe Mirror The Holcombe mirror was found at the bottom of a pit which also contained two large chert blocks. The pit was filled in one episode and some broken fragments of pottery were found in the pit fill. Staining and fibrous growth were found in the soil underneath the mirror, suggesting that it may have been wrapped in organic covering. The bronze mirror is large, decorated and fits into the ‘western’ group. Mirrors of this type are most often found in graves. Wrapping a mirror was related to dress; the mirror in this instance became metaphorically related to a human corpse dressed in funerary garb. The ‘burial’ of the mirror therefore would have stood for a real burial. The pit is too large to have been dug solely for the purpose of depositing the mirror (Pollard 1974: 70) and there is a well recognised tradition of structured deposition of objects and human remains in pits (Hill 1995b). The relationships written into

82

the mirror at the time of its production became significantly altered over the course of its life, causing it to be deposited in a pit rather than a grave. Here we can envisage the connection between the form of a mirror and its reference to an idealised past being reinterpreted (Chapter 6). The pit the mirror was placed in was one of the earliest features on the site (Fox & Pollard 1973: 20). The mirror could have been buried in the pit as a ‘foundation’ deposit. The biography of the mirror and its associations with an ‘idealised’ past being used to ensure the future of a new settlement founded at a time of social upheaval during the middle decades of the 1st-century AD. 10.5 CONCLUSIONS Mirror burials can be examined from the perspective of ‘the complexity of the social relationships they objectify’ (Jones 2002a: 171). Following the approach set out throughout this book, instead of regarding mirrors as indicators of status and value, we can see them as part of a complex set of social relationships materialised and authenticated through their presence in graves or other deposition contexts. The biographies of mirrors – the relationships they objectify – were critical in determining how they ended their social lives. One very important trend can be identified between the mirrors with the most identified production steps outlined in Chapter 3 and the complexity of the deposition contexts they are found in. The least complicated mirrors to make are those made of iron. Three iron mirrors have been found in vehicle burials which objectify all manner of complicated relationships (see above). Some of the most complicated mirrors to make (decorated bronze mirrors found in cremation burials in southeast England, western mirrors like Holcombe and mirrors made from composite components found in settlement contexts) are found in more innocuous contexts, often on their own. Of course there are exceptions to this general pattern but it is tempting to suggest that the relationships objectified by the more complicated mirrors are sufficiently complex to make their deposition in seemingly simple contexts more meaningful. The mirrors invested with less meaning at the time of production, and afforded with less social potential, on the other hand, are more likely to be found in more complex depositional contexts.

CHAPTER 11 CONCLUSIONS 11.1 INTRODUCTION Using a modified version of the biographical approach mirrors were examined from a relational perspective. Relationships written into mirrors at the time of their production and during their social lives have been revealed (see Chapter 7). A biographical account of the relationships formulated when mirrors were deposited was provided (see Chapter 10). In this chapter the implications of these findings for wider research and the future application of the biographical approach to archaeological study are assessed. 11.2 MIRRORS AND THE BRITISH IRON AGE Past interpretations of Iron Age metalwork production have concentrated on technology and often bemoan the general absence of evidence for artefact production. In this study the value of the information written into artefacts has been demonstrated. This information was utilised to examine the relationships formulated when mirrors were manufactured. A number of studies have examined the construction of individual artefacts in detail (e.g. Hunter 2001; Stead 1985). However, this is the first time that the potential relationships formulated during the manufacture of artefacts have been examined systematically. As outlined in Chapters 3 and 4, the organisation of metalworking has often been viewed as self explanatory. Models based on Renaissance artefact production with a metalworker, or ‘artist’, working under a ‘patron’ are often uncritically applied to metalwork production. This book has revealed that multiple and complicated social networks were required to make artefacts like mirrors, indicating that the ‘patron’ model of metalwork production is too simplistic and should be revised. To make objects like mirrors, complicated social networks were formulated involving multiple individuals. Rather than a patron simply communicating power through ‘wealth’ and ‘taste’ and the skill of the metalworker they employed, mirror makers expressed social difference through the relationships they could coordinate. More complicated mirrors required more extensive social networks to manufacture. Further research is necessary to explore the organisation of the production of Iron Age metal artefacts from a relational perspective and question: · · · · · ·

Who initiated the making of complex artefacts? Who was the planner and who was the maker? Did the planner use the object, or were they made for exchange? How do these relationships fit into a wider model of society? Who possessed the knowledge of how complex designs should be constructed? What was the wider significance of this design knowledge and how was control over this knowledge manipulated?

As demonstrated in Chapter 4, mirror designs conform to a set of design rules and mirror decoration has been shown to express ‘mirrorness’ (see also Joy 2008b). This is a major departure from existing Celtic Art studies which have separated art motifs from their artefactual context. They have consequently interpreted art motifs as representative of

83

something else, most often animals or mythical beasts (see Jope 2000). Mirror style motifs are found on other types of objects. With the exception of a very small number of sword scabbards (e.g. Bugthorpe and Isleham), a decorated comb from Warwickshire and perhaps some horse-harness equipment, these mirror-style motifs are not organised in the same way as they are on mirrors. They are not formed into designs with a balance of positive and negative components. Further work is required to examine why these other objects are also decorated with mirror-style motifs. Why in particular were objects decorated with specific motifs? Were there rules which governed how different artefacts should be decorated? For example, the majority of artefacts are decorated with abstract designs. However, vessels and buckets are often ornamented with mounts showing human or animal likenesses. Objects like weaving combs also have distinctive ring and dot decoration (Tuohy 1999). By identifying design rules for specific object types it should be possible to relate these to wider social questions and further examine the efficacy of decoration. Mirrors were made and used in Britain for a restricted period, from 400 BC – AD 100. They have been found in watery and settlement contexts but the majority come from graves. These graves are heterogeneous and follow distinctive regional burial practices. Metal artefacts like mirrors were critical in the formulation of identity. In East Yorkshire mirrors are found in the graves of ‘mature’ women and were active in constructing the identity of these individuals. In the 1st-century BC and 1st-century AD there is evidence that people were expressing their social difference in new and different ways, by eating and drinking new types of food using new types of utensils (Hill 2002b) or actively demonstrating a concern for personal hygiene and appearance by using toilet objects like tweezers or mirrors (Hill 1997; 2001b; Jundi & Hill 1997). This social difference may have involved increasing hierarchisation in southeast England with the emergence of ‘kingship’ (Creighton 2000; 2006). As explained in Chapter 8, the deposition of mirrors in southeast England from 75-15 BC, and outside of southeast England in the middle centuries of the 1st-century AD, echoes wider trends of artefact deposition during this period. These depositional episodes can be related to wider changes in the archaeological record occurring at this time. For example, tankards are found in south-eastern cremation burials dating from 50-25 BC, they were also deposited outside of the southeast in contexts dating from the mid 1st-century AD onwards. Similarly, a number of hoards, especially containing horse-harness equipment, have been found outside of southeast England, dating to the mid 1st-century AD. An alternative approach to Iron Age objects can be suggested. Rather than acting as passive representations of the status of the people who made and owned them, it is possible to examine how the biographies of objects were referred to and manipulated when they were deposited. For example, the deposition of metalwork in the middle

1st-century AD can clearly be related to events surrounding the Roman invasion of AD 43, the immediate aftermath and social changes occurring at this time. Through depositional acts people were actively renegotiating identities. A mirror, which, through the form of its components can be linked to many other artefact forms and an idealised sense of the Iron Age past, could have been manipulated in multiple ways to express these identities. 11.3 THE APPLICATION OF THE BIOGRAPHICAL APPROACH TO ARCHAEOLOGICAL DATA As outlined in Chapter 2, the number of prehistoric archaeological studies utilising the biographical approach has increased over recent years. Examination of these studies reveals a pattern. Often a mention of object biography is made in the introduction of a paper with reference to key works by Kopytoff or Appadurai. The conclusions to these papers then include a statement suggesting that the object/s under scrutiny must have had ‘rich’ or ‘complex’ biographies (Whitley 2002; Joy 2009a). No attempt is made to comment upon the form of these ‘rich’ or ‘complex’ biographies. This type of conclusion is deeply unsatisfactory and the application of the biographical approach in this form provides little, if any, benefit to archaeological research. One reason why researchers have reached this type of conclusion is the limitations of prehistoric evidence did not allow them to construct detailed life histories with a birth, life and death. As our own biographies are structured in this way, these researchers believed that the lives of prehistoric objects should also be in this form. Throughout this book a new methodology has been developed that is adapted to the restrictions in data from prehistoric contexts. Following approaches to artefact biographies laid out by Strathern (1988), Gell (1998) and Hoskins (1998), biography was examined from a relational perspective. The biography of an object is the sum of the relationships that constitute it. Viewed from this perspective, biography does not need to rigorously follow birth, life and death. It is therefore possible to study biographies of prehistoric objects through an examination of the relationships that constitute them (see also Jones 2002a; Knappett 2005; Gamble 2007). In relation to mirrors, the biographical information written into them at the time of their manufacture, the relationships formulated during their making and the potential relationships afforded to mirrors through their artefact form was examined. This is the first time that biographical information has been compiled in this way in an archaeological context. This methodology has great potential to inform us of the biographies of many other prehistoric artefact forms when combined with more conventional biographical analysis. The biographical information written into mirrors during their social lives through an analysis of wear and damage was also examined and a detailed analysis of deposition context conducted. Impor-

84

tantly, following Whitely (2002), account was taken of the biographies of mirrors in examination of depositional context. There are many other complicated Iron Age artefacts which social relationships are likely to have clustered around and that could benefit from biographical analysis. Many of these objects have been labelled as ‘Celtic Art’ or ‘La Tène Art’ by various authors (Allen 1904; Leeds 1933; Jacobstahl 1944; De Navarro 1952; Fox 1958; Stead 1996; Jope 2000; Megaw & Megaw 2001). To the detriment of both Iron Age and Celtic Art studies, the study of Celtic Art has long been separated from the main body of Iron Age research particularly because of the methodologies used to examine them (MacDonald 2007a; Fitzpatrick 2007; Garrow et al 2008). Celtic Art has been examined in its own right, separated from archaeological context. By specifically examining the social relationships that constituted mirrors, a well known group of Celtic Art objects, this study has not only increased our depth of knowledge of these objects, it has re-assimilated their study into wider Iron Age research (see also MacDonald 2007a) and demonstrated the value of detailed artefact analysis. 11.4 BIOGRAPHICAL MIRRORS Mirrors mediated social relationships in later Iron Age Britain (see Gell 1998: 7). Through an analysis of the social relationships that constitute mirrors – their biographies – it has been demonstrated that throughout their lifetime’s mirrors acted to construct, maintain and reformulate social identities. The people who made, or coordinated the making of, complicated artefacts like mirrors actively demonstrated the different relationships they could coordinate through the performance of artefact manufacture. Mirrors were made to be inherently meaningful and powerful objects and were afforded with the potential to act within multiple relational spheres. Much of the meaning and significance of mirrors resides in the interplay between mirror component parts. Mirror components possess a kind of ‘inner logic’ and facilitate the primary function of a mirror: to reflect light. By referencing the physical features of contemporary artefacts, mirror components can also be connected to other artefact forms, providing mirrors with the potential to act within further relationships. Linking mirrors to other categories of contemporary material culture also re-emphasises the origin of mirrors providing a ‘timeless’ quality to their manufacture. Mirrors accumulated relationships as they were exchanged and used and the individual biography of a mirror affected the ways in which they were deposited. To borrow a phrase from Strathern (1988), mirrors are a microcosm of relations in later Iron Age Britain.

APPENDIX A CATALOGUE OF BRITISH-TYPE MIRRORS 1. Akenham, Suffolk Bronze handle fragment, type IV. Metal detector find (PAS Database). No context. Privately owned. Figure A35. 2. Arras I, East Riding Iron mirror with bronze mounts, handle type I. Found in an inhumation burial. British Museum. Figure A1. 3. Arras II, East Riding Iron mirror, handle type I. Found in an inhumation burial. Lost. 4. Aston, Hertfordshire Bronze mirror, decorated plate, handle type IV. Found in a cremation burial. British Museum. Figure A2.

15. Bromham, Bedfordshire Fragmentary bronze handle and plate. Plate decorated. Handle type IV. Metal detector find, context unknown. Bedford Museum. Figure A8. 16. Bryher, Isles of Scilly Bronze mirror with decorated plate. Handle type III. From an inhumation burial. St. Mary’s Museum, St. Mary’s, Isles of Scilly. Figure A9. 17. Bulbury Camp, Dorset Bronze handle and plate fragment with traces of a decoration. Found with a group of objects which look to constitute a hoard. Dorchester Museum. Figure A36.

5. Bac Mhic Connain, North Uist, Outer Hebrides Whale bone handle, type V. Found on a settlement site. National Museum of Scotland, Edinburgh. Figure A35.

18. Carlingwark, Dumfries & Galloway Bronze mirror handle fragment. Handle type I. Found in a bog. National Museum of Scotland, Edinburgh. Figure A36.

6. Badingham, Suffolk Bronze handle, type IV. Metal detector find (PAS Database). No context. Privately owned. Figure A35.

19. Chettle, Dorset Decorated bronze mirror plate. Possible cremation. British Museum. Figure A10.

7. Balleybogey Bog, Balleymoney, Co. Antrim Bronze handle, type I. Once had an iron plate which has corroded. Found in a bog. Ashmolean Museum, Oxford. Figure A35.

20. Chilham Castle, Kent Bronze mirror with decorated plate. Handle type III. Found in a cremation burial. Canterbury Museum. Figure A11.

8. Balmaclellan, Dumfries & Galloway Bronze mirror with decorated mounts. Handle type II. Found in a bog. National Museum of Scotland, Edinburgh. Figure A3.

21. Colchester I (Lexden Grange), Essex Bronze mirror with decorated plate. Handle type IV. Found in a cremation grave. Colchester Museum. Figure A12.

9. Billericay I, Essex Bronze mirror handle and decorated plate fragments. Handle type I. Possibly from a cremation burial. Colchester Museum. Figure A4.

22. Colchester II (Hyderabad Barracks), Essex Bronze handle and plate fragments. Plate decorated. Handle type I. Context unknown. British Museum. Figure A13.

10. Billericay II, Essex Bronze mirror handle and plate fragment. Handle type III. From a cremation burial. Colchester Museum. Figure A5.

23. Compiègne, France Bronze mirror handle found in the River Oise at Compiègne in 1814. Handle type IV. Figure A36.

11. Billericay III, Essex Bronze mirror handle, type unknown. Possibly from a burial. Location unknown.

24. Desborough, Northamptonshire Bronze mirror with decorated plate. Handle type IV. Context unknown. British Museum. Figure A14.

12. Birdlip, Gloucester Bronze mirror with decorated plate. Handle type IV. From an inhumation burial. Gloucester Museum. Figure A6.

25. Dorton, Buckinghamshire Bronze mirror with decorated plate. Handle type IV. From a cremation burial. Buckinghamshire County Museum, Aylesbury. Figure A15.

13. Breacon Beacons Mirror with a bronze handle, decorated bronze collar and an iron plate. From a cremation burial. National Museum of Wales, Cardiff. Figure A7.

26. Garton Slack, East Yorkshire Iron mirror with bronze mounts. Handle type IV. From an inhumation. Hull Museum. Figure A16.

14. Bridport, West Bay, Dorset Bronze mirror handle. Handle type III. From an inhumation burial. Dorchester Museum. Figure A36.

27. Glastonbury, Somerset Bronze mirror plate. Found at site of Glastonbury Lake Village. Glastonbury Museum. Figure A17.

85

28. Great Chesterford, Essex Bronze mirror with decorated plate. Context unknown. Cambridge University Museum of Archaeology and Ethnography, Cambridge. Figure A18.

41. Portesham, Dorset Bronze mirror with decorated plate. Handle type IV. From an inhumation grave. Dorchester Museum. Figure A27.

29. Holcombe, Devon Bronze mirror with decorated plate. Handle type IV. Found at the bottom of a pit on a small settlement site. British Museum. Figure A19.

42. Portland I, the Grove, Dorset Bronze mirror handle. Handle type III. Probably from an inhumation burial. Coburg Museum. Figure A37.

30. Ingleton, Yorkshire Bronze mirror handle with two ox-head mounts. Handle type I. Context unknown. British Museum. Figure A37. 31. Jordan Hill, Weymouth Bronze mirror handle. Handle type IV. Context unknown. British Museum. Figure A37. 32. Lambay Island, Dublin Iron mirror. Handle type I. Inhumation burial. Dublin Museum. Figure A20. 33. Latchmere Green, Hampshire Bronze mirror with decorated plate. Handle type III. From a cremation burial. Museum of the Iron Age, Andover. Figure A21. 34. Llechwedd-ddu, Conwy Bronze mirror, plate undecorated. Handle type IV. Probable grave context. National Museum of Wales, Cardiff. Figure A22.

43. Portland II, the Verne, Dorset Decorated bronze plate fragment. Probably from a burial. Dorchester Museum. Figure A23. 44. Rickling, Essex Bronze handle and plate fragments, plate decorated. Context unknown. Colchester Museum. Figure A28. 45. Rivenhall I, Essex Bronze mirror handle and decorated plate fragment. Handle type III. Found on a settlement site but exact context unknown. Chelmsford Museum. Figure A28. 46. Rivenhall II, Essex Bronze mirror handle. Handle type IV. Found on settlement site but exact context unknown. Lost. 47. St. Keverne, Trelan Bahow, Cornwall Bronze mirror with decorated plate. Handle type III. From an inhumation burial but exact context not well recorded. British Museum. Figure A29.

35. Lochlee Crannog, Tarbolton, South Ayrshire Bronze mirror handle. Handle type V. Context not well recorded. Lost. Figure A37.

48. Stamford Hill I, Plymouth Damaged decorated bronze plate. From an inhumation burial but exact context not well recorded. Destroyed in World War II bombing raid on Plymouth. Figure A30.

36. Maiden Castle, Dorset Iron plate fragment with bronze rim. Settlement context although details not well recorded. Dorchester Museum. Figure A23.

49. Stamford Hill II, Plymouth Bronze mirror handle. Handle type III. From an inhumation burial but exact context not well recorded. Destroyed in World War II bombing raid on Plymouth. Figure A30.

37. Nijmegen, The Netherlands Bronze mirror with decorated plate, handle broken in the centre of the grip. Handle type IV. Found in a cremation grave (grave no 29, excavated 1926-7). Nijmegen Museum. Figure A24.

50. Stamford Hill III, Plymouth Bronze mirror handle. Handle type I. From an inhumation burial but exact context not well recorded. Destroyed in World War II bombing raid on Plymouth. Figure A30.

38. Old Warden I, Bedfordshire Bronze mirror with decorated plate. Handle type IV. Context unknown. Higgins Museum, Bedford. Figure A25. 39. Old Warden II, Bedfordshire Bronze mirror handle with fragment of a bronze plate still attached to the handle. Handle type IV. Context unknown. Lost. Figure A37. 40. Pegsdon, Shillington, Bedfordshire Bronze mirror with decorated plate. Handle type III. From a probable cremation burial. Luton Museum. Figure A26.

86

51. Westerfield, Suffolk Bronze mirror handle fragment. Handle type III. Metal detector find, context unknown. Privately owned. Figure A38. 52. Wetwang Slack, East Riding of Yorkshire Iron mirror. Handle type I. From an inhumation burial. Hull Museum. 53. Wetwang Village, East Riding of Yorkshire Iron mirror. Handle type I. From an inhumation burial. British Museum.

Bronze mirror with decorated plate. Handle type III. Context unknown. Privately owned. Figure A34.

UNPROVENANCED 54. Disney Bronze handle and plate fragments. Plate decorated. Handle type IV. Context unknown. British Museum. Figure A31. 55. Essex/Sussex border Bronze handle with plate fragments attached. Handle type IV. Context unknown. Privately owned. Figure A38. 56. Gibbs Almost complete bronze mirror. Handle type III. Context unknown. British Museum. Figure A32. 57. Mayer Bronze mirror. Handle type III. Context unknown. Merseyside Museums, Liverpool. Figure A33. 58. ‘Oxfordshire’

87

POSSIBLE MIRROR HANDLES Fison Way, Thetford, Suffolk Bottom section of a possible bronze mirror handle. Settlement context. Figure A38. Llanwnda, Pembrokeshire Bronze handle. Context unknown. Lost. Figure A38. Merlesford?, Fifeshire Bottom section of a possible bronze mirror handle. Context unknown. National Museum of Scotland, Edinburgh. Figure A38 Illustrations below are at 2/3 scale unless otherwise stated. All credits given in list of illustrations.

Figure A1. Arras I (scale 1:1)

88

Figure A2. Aston

89

Figure A3. Balmaclellan

90

Figure A4. Billericay I (Scale 1:1)

91

Figure A5. Billericay II (scale 1:1)

92

Figure A6. Birdlip

93

Figure A7. Brecon Beacons

94

Figure A8. Bromham

95

Figure A9. Bryher (scale 1:1)

96

Figure A10. Chettle

97

Figure A11. Chilham Castle (scale 1:1)

98

Figure A12. Colchester I

99

Figure A13. Colchester II (scale 1:1)

100

Figure A14. Desborough

101

Figure A15. Dorton

102

Figure A16. Garton Slack

103

Figure A17. Glastonbury (scale 1:1)

104

Figure A18. Great Chesterford

105

Figure A19a. Holcombe

106

Figure A19b. Holcombe

107

Figure A20. Lambay Island

108

Figure A21. Latchmere Green

109

Figure A22. Llechwedd du

110

A

B

Figure A23. Maiden Castle (A) & Portland II (B)

111

Figure A24a. Nijmegen

112

Figure A24b. Nijmegen

113

Figure A25. Old Warden I

114

Figure A26. Pegsdon

115

Figure A27. Portesham

116

A

B Figure A28. Rickling (A) and Rivenhall I (B) (scale 1:1)

117

Figure A29. St. Keverne (scale 1:1) 118

A

B

C

Figure A30. Stamford Hill I (A), II (B) and III (C)

119

Figure A31. Disney (when first reported and detail of plate at present)

120

Figure A32. Gibbs (when first reported and condition now)

121

Figure A33. Mayer (scale 1:1)

122

Figure A34. ‘Oxfordshire’

123

A B

C

D

Figure A35. Akenham (A), Bac Mhic Connain (B), Badingham (C), Balleybogey (D)

124

C A

B D

Figure A36. Bridport (A), Bulbury (B), Carlingwark (C), Compiègne (D)

125

C

B

A

E D Figure A37. Ingleton (A), Jordan Hill (B), Lochlee (C), Old Warden II (D), Portland I (E)

126

C

A

D

B E Figure A38. Westerfield (A), Essex/Sussex border (B), Fison Way (C), Llanwnda (D), Merlesford? (E)

127

APPENDIX B CHRONOLOGY OF IRON AGE MIRRORS This appendix is a detailed interrogation of mirror deposition date for mirrors found in close association with human remains that have been radiocarbon dated, or artefacts which can be closely dated. Throughout it follows closely the work of Sealey (2006). It was decided not to conduct a radiocarbon dating audit for the mirror corpus as a detailed analysis revealed that the investigation would not have been worthwhile (see below).

than that given by Brewster. The burial of another woman in the ditch of the same barrow as the mirror burial produced a radiocarbon date of 30 b.c. ±90 years (Harwell 1274) (Brewster 1980: 229).

A sample of bone from the burial has recently been re-dated (OxA-17286) to 380-200 cal BC (95.4%) (Garrow et al forthcoming). As this is a date obtained using AMS and fits with the other dated mirror burials from East Yorkshire (see B.1 DATING BASED ON RADIOCARBON DATES below) this new date is accepted and the previous Harwell The radiocarbon calibration curve is flat during the Iron date is viewed as an anomaly. Age, causing problems in calibrating radiocarbon years and a reduction in precision (Aitken 1990: 93). Two de Vries effect perturbations occur during the 1st-millennium BC: 380-200 cal BC. IIIa (170-385 BC) and IIIb (405-780 BC) (Taylor et al. 1996: 663). Mirrors almost definitely date to after the Wetwang Slack, East Yorks period affected by perturbation IIIb. IIIa, however, could coincide with the date of the earliest mirrors. All dates Radiocarbon date from a sample of human bone (Figure B3). detailed below are calibrated using OxCal version 4.1. Garton Slack, East Yorks Radiocarbon date from a sample of human bone (see Figure B1 & B2).

Figure B3. Calibrated date for Wetwang Slack (OxA14113) This burial has recently been radiocarbon dated. The date found was 2227±30 b.p. Figure B1. Calibrated date for Garton Slack (Harwell 1274) 386-203 cal BC. Wetwang Village, East Yorks

Figure B2. Calibrated date for Garton Slack (OxA-17286) A sample of bone gave a date of “90 b.c. ± 70 years” (Harwell 3260) (Brewster 1980: 37, 229, 244). The reliability of this date has been questioned (Fitzpatrick 1997a). It Figure B4. Calibrated date for Wetwang Village (OxAwas not possible to locate a publication of this date other 11993)

128

Radiocarbon date from a sample of human bone. The grave has recently been radiocarbon dated to 2151±21 b.p. 353-111 cal BC. B2. DATING BASED ON ASSOCIATED ARTEFACTS Figure B5. The Birdlip brooch (after Staelens 1982: Figure 2)

Aston, Hertfordshire Two hand-made, grog-tempered pots. The top half of one of the pots has been severely damaged by ploughing and may be a larger version of the other pot, which has a hand-made body and a wheel finished top. The form of these two pots is unusual for cremation burials of south-eastern England. Rigby (in Rook et al 1982: 23) suggests that the pots date to the second half of the 1stcentury BC. This date is based on the proportions of the vessels and on the techniques of manufacture, although Rigby does concede that the tradition of grog-tempered pots with wheel-finished rims and handmade bodies did continue well into the 1st-century AD. Sealey (2006: 17) suggests that the dating of these vessels can be extended backwards as far as c. 75 BC. c. 75 – 1 BC.

A group of approximately twenty brooches similar to the Birdlip brooch (and taking its name) have been found in Britain. Two well stratified examples have been found, but unfortunately, they vary considerably in date. One brooch from Dragonby, Lincolnshire was found associated with pottery dated to the late 1st-century BC – early 1st-century AD (Spratling 1970a: 14); a second from Kingsholme, Gloucestershire was recovered from a pit dated to AD 40-50 (Staelens 1982: 23). Stead (in Stead & Rigby 1989: 103) suggested that although the Birdlip brooch may be typologically late it is probably not as late as Spratling suggests and favoured a 1st-century BC date. Olivier (1996: 232) suggested that this type of brooch can only be dated to the period from the mid 1st-century BC to mid 1st-century AD. It is probable, however, that the Birdlip type brooch, shares a similar date to Colchester derivatives which date from the late 1st-century BC until AD 60 – 70.

Billericay III, Essex

Late 1st-century BC – AD 60/70.

Two brooches

Near Brecon, Powys

A carinated bowl. The better preserved of the two brooches is a Knotenfibel (Weller 1974: 282-3; Farley 1983: 289), dating from c.7525 BC (Sealey 2006: 16). This unusual mirror should not be seen as typical of the corpus as a whole. It was found by a metal detectorist in the early 1990’s along with two mini terrets, a cosmetic set, a c.75-25 BC small quantity of cremated human bone, and a carinated bowl. The carinated bowl probably dates to the late 1stBirdlip, Gloucestershire century AD – early 2nd-century AD. A silver brooch. AD 70 – early 2nd-century AD. There has been much debate over the date of the Birdlip Brooch. For example, Smith (1909: 341-5) thought it was related to the ‘Augenfibel’ type. Green (1949: 190) disagreed and related it to the ‘Aylesford’ which he dated to c. 50/30 – 10 BC. Spratling (1970a: 13) argued that the design of the brooch was influenced by Almgren’s Germanic Group series II, dated to the 1st-century AD and ascribed the Birdlip brooch to AD 20 or later.

Bryher, Scilly Isles Radiocarbon date, a copper-alloy brooch, an iron sword and bronze scabbard. A radiocarbon determination for the burial, based on a human long bone sample, gave a date of ‘200±45 cal BC’

129

Figure B6. The radiocarbon determinations from Bryher Figure B9. The Bryher sword (after Johns 2006: Figure 21) (after Johns No Date: 15) from the 2nd-century BC – early 1st-century BC (Stead in Johns 2006). The fragmented bow brooch is considered to date between 125 and 50 BC (Johns No Date: 14). The brooch is very similar to the one found with the St. Keverne mirror (see below). Despite the fact that the copper-alloy brooch was in a very fragmented state and the main diagFigure B7. Calibrated results of the Bryher grave (after nostic features – the foot and catch plate – are missing, Johns No Date) Johns (No Date: 37) suggested that the brooch is likely to date from 100 to 50 BC as did Hill (in Johns 2006). Late 2nd-century BC – early 1st-century BC based on the C14 date. Chettle, Dorset A green-glass cylindrical-handled bottle. A decorated mirror plate was recovered from a pit also containing: a copper-alloy spouted strainer bowl, a flagon, handled pan and two shallow dishes; a green-glass cylindrical-handled bottle; and a blue-glass ribbed bowl (Sealey Figure B8. The Bryher brooch (after Johns 2006: Figure 2006: 17). Cylindrical bottles of this kind were in use from 23(1)) c. AD 70-100 (Price & Cottam 1998: 191; Sealey 2006: 17). The other objects in the grave could all possibly date to the (Johns No Date: 15). Stead (in Johns 2006: 171) believed Flavian period (Fitzpatrick pers.comm). this date is ‘too early’. According to Johns (2006; No Date) the metalwork found in the grave narrows its date to the first half of the 1st-century BC. The sword and scabbard date c. AD 70-100

130

Chilham Castle, Kent

Leeds (1933: 30) dated the mirror to the late 1st-century BC, based on the pottery vessels found in association with it and used style to support his argument. Fox & Hull (1948: 134) suggested a slightly later date of AD 10-25, based on the pottery vessels and the stylistic sequence. Spratling (1970a: 14) agreed and stated that it is not possible to be more specific than AD 10-43. Fox & Pollard (1973: 27) cited the absence of Gallo-Belgic or Samian vessels to suggest a similar date. However, as Stead (in Stead & Rigby 1989: 103) pointed out, the pottery could be dated earlier. Fitzpatrick (1997a: 67) supported Stead, stating that “…the presence of early mica-dusted Central Gaulish flagons now suggests an earlier date, probably before 15 BC”. Sealey (2006: 17) broadly agrees and suggests a date range for the vessels of c.25 BC – AD 5.

Figure B10. The Chilham Castle brooches (after Parfitt c. 25 BC – AD 5, based on the mica-dusted flagons. 1998: Figures 3 & 4) Dorton, Buckinghamshire A funerary urn and a pair of copper-alloy brooches.

Three amphorae, two flagons and a carinated cup.

The funery urn is a fragment of a ‘Belgic-style grog-tempered jar’ (Macpherson-Grant in Parfitt 1998). However, it is not particularly significant in terms of dating the grave as Macpherson-Grant (ibid: 349) uses the brooches from the grave as dating evidence to reassess the dating of this type of pot (which prior to this find had been dated to the second half of the 1st-century BC, between 40 – 20 BC)!

The amphorae date to c. 16 BC – AD 149 (Williams in Farley 1983). The necks of the three amphorae were removed in the past. Williams (ibid: 291) classified amphorae 1 as Dressel 1B and amphorae 3 as Dressel 2-4 (amphorae 2 is more difficult to classify) and suggested a date for the group late in the 1st-century BC. Because there are signs of secondary usage Williams suggested that the amphorae may even be dated slightly later. It is difficult to find close parallels for the Dorton flagons, making dating tricky (Rigby & Freestone in Farley 1983: 292). The carinated cup is similar to other vessels found in cremation graves in Hertfordshire and Essex and probably dates to the early 1st-century AD (Thompson in Farley 1983).

The brooches found in the grave have never been stylistically subdivided. Stead (in Parfitt 1998: 347) suggested that they have typological links with Nauheim brooches and dates them to around 50 BC. This date is based on the similarity of these brooches with a number found in Kent (Deal and Folkestone), Sussex (Westhampnett) and Essex (Great Chesterford). In the report on the Westhampnett graves, Fitzpatrick (1997b: 203-4) reassesses the date of this type of brooch using information derived from Europe. He suggested that they probably date from c. 70 – 30 BC.

The dating evidence from the Dorton mirror burial suggests that the burial could have taken place any time between the late 1st-century BC and the first decades of the 1st-century AD. Sealey (2006: 17) has cited the absence of GalloBelgic imports in the grave to indicate that the burial took Following Stead, Sealey (2006: 17) has suggested that the place sometime before 15 BC. Chilham Castle brooches can be dated more precisely to around c. 75 – 50 BC. Based on the dating of the two brooches the Chilham Castle mirror is presumed to be one A date late in the first century BC or later based on of the earliest recovered from cremation burials in southeast William’s assessment of the amphorae. England (Stead in Parfitt 1998). It also looks to be one of the ‘crudest’ in terms of the proficiency of manufacture and Holcombe, Devon decoration. This presumed ‘poor quality’ could have unduly influenced the dating of the grave. There is no reason The mirror was found at the bottom of a pit. Dating evifor the extra precision of dating argued by Sealey. dence is derived from the sequence of the site and pottery found elsewhere on the site. 75 – 25 BC. Colchester I, Essex Pottery vessels, a bronze cup and pin.

In the fill of the pit four sherds of unburnished, coarse pottery were found, but these were undiagnostic (Pollard 1974: 69-70). Based on the discovery of a sherd from a Durotrigian ribbed-bowl with dotted decoration, as well as

131

three coarse black-burnished sherds, and a small bead-rim, found at the base of two other pits on the same site, Pollard suggested a date of c. 20-44 AD for the deposition of the mirror in the pit. Sealey (2006: 17) has re-examined the pottery from the site and suggested a date for the deposition of the mirror of c. 55-75 BC.

(ibid: fig 13.2). This is also dated to La Tène D2 by the excavators, a type that was replaced by butt-beakers. Contra Sealey (2006), the dating of the Latchmere Green mirror burial can be revised, 75-25 BC. Nijmegen, Netherlands

c. AD 20 – 75.

A Roman glass vessel.

Latchmere Green, Hampshire A crematory urn and two iron brooches. The brooches are iron and in a fragmentary condition. According to Richards (in Fulford & Creighton 1998) the brooches cannot be dated more specifically than sometime in the second half of the 1st-century BC or early in the 1st-century AD. The brooches are not illustrated in the report, however, Richards (ibid.) does compare them to two ‘type S’ examples found in Phase 2 at King Harry Lane (KHL). A reassessment of the chronology of the KHL cemetery by Mackreth (1994: 49-50) suggests a date range of 15 BC – AD 55 for phases 1 to 3. Richards also compares the Latchmere brooches with two brooches from ‘Period 2’ (c.10 BC – c. AD 40-50) at Silchester (Corney 2000: 237; Fulford & Timby 2000: 13-16), as well as a brooch from Gallows Hill, Norfolk, dated to c. 100-75/25 BC (Mackreth 1991: 124). The pottery vessel was almost complete but in fragments. The fabric of the vessel is identical to some pots found in the pre-Conquest levels at Silchester, dated in the last quarter of the 1st-century BC but more commonly from the first half of the 1st-century AD (Timby in Fulford & Creighton 1998: 235-6). Timby also compares the vessel to finds from Skeleton Green, dating to the late 1st-century BC (see Partridge 1981). Fulford & Creighton (1998: 339) date the deposition date of the grave on stylistic grounds to the early 1st-century AD and set an earliest possible date for it of 15 BC. This earliest possible dating seems to be quite arbitrary and has little foundation other than the stylistic dating of the mirror by Jope (in Fulford & Creighton 1998). As already stated the brooches were not dated more specifically than between the second half of the 1st-century BC – 1st-century AD by Richards and Timby provides an earliest possible date of c. 25 BC for the date of the pottery vessel. Recent work at Hinxton Rings, Cambridgeshire (Hill et al. 1999) enables a re-evaluation of the dating evidence from Latchmere Green. A similar brooch to the two Latchmere Green brooches was found at Hinxton Rings. This brooch has also been compared to KHL type S, and has an internal cord and a four coil spring. However, in contrast to the Latchmere Green brooches the brooch from Hinxton Rings has been designated as being La Tène D2 with a probable mid 1st-century BC date (ibid: 254-5). As further evidence, a plain uncordoned pedestal urn, very similar in form to the Latchmere Green vessel, was also found at Hinxton rings

Dunning (1928: 69) dated the grave, based on the glass vessel found with it, to AD 100. Leeds (1933: 30) also argued for a late date of the mirror on stylistic grounds. Fox & Pollard (1973: 30) agreed with both Dunning and Leeds, stating that the mirror was made sometime after AD 50 (based on stylistic analogies) and was interred some time early in the 2nd-century AD. Spratling (1970a: 14) also broadly agreed with this. Stead (1989: 103), based on the chronology for glass urns outlined by Issings (1957: 82), and a comment from Lloyd-Morgan (1991: 114) about ‘extreme wear’ on the mirror, suggests that the burial probably dated to the late 1st-century AD but the mirror was probably already of some age when it was included in the burial as a grave good. Reassessment of this type of glass vessel only allows them to be dated to the 1st and 2nd centuries AD (Price & Cottam 1998: 139; Sealey 2006: 18). Late 1st-century AD – early 2nd-century AD. Pegsdon, Shillington, Bedfordshire A silver boss-on-bow brooch found attached to the terminal loop of the mirror handle. The brooch is a British type dating to La Tène D2 (c.75BC25BC) (Hill 2000; Burleigh & Megaw 2007). c.75 BC – 25 BC. Portesham, Dorset A copper-alloy pan, a cosmetic set, an iron knife, three copper-alloy brooches, two pottery bowls and a jar. The imported Roman pan and strainer set, and the cosmetic set suggest the burial dates to about the time of the Roman conquest (Fitzpatrick 1997b: 67). AD 40 – 60. Rivenhall I, Essex Impurity patterns of the trace elements in the mirror.

132

Figure B.11 Portesham brooches (after Fitzpatrick 1997a) Sealey (2006: 18) cited evidence provided by Northover (1993) on the trace elements found in the bronze of the Rivenhall mirror to suggest a date for the mirror of c. 50 – 1 BC. However, trace element analysis is not a reliable method of dating objects as selective recycling of scrap can blur impurity patterns (see Dungworth 1997b). No close dating possible. St. Keverne, Trelan Bahow, Cornwall

c.120-80 BC. Stead (in Stead & Rigby 1989: 103) on the other hand, has suggested that close dating of this brooch is impossible because it may or may not have had an open catch-plate meaning that it is not possible to tell for definite whether this brooch is a Nauheim or a Nauheim derivative (Sealey 2006: 18). Probably 120 – 80 BC based on the brooch being Nauheim type. This ties in with the dating of the Bryher mirror, which is similar in size and shape to the St. Keverne mirror and was also found with a Nauheim type brooch. Stamford Hill I, II & III, Mount Batten, Devon No close dating evidence. Supposition based on the dating of the cemetery.

Sealey (2006: 18) has suggested it is possible to tie the dating of the mirrors from Stamford Hill based on the dating of the cemetery as a whole by Cunliffe (1988: 98) to AD 43 – 70. Stylistic dating of the decorated mirror plate could add weight to this argument. However, as the graves A copper-alloy brooch. were not properly recorded when they were uncovered during the early 1860’s, and no datable objects can be In his report on the St. Keverne burials Jope-Rogers (1873: firmly tied to the mirror graves, this dating remains specu272) comments that ‘two portions of fibulae’ were found in lative. the same burial as the mirror. Smith (1909: 330) later published an illustration of this brooch fragment with a ‘bi-lateral spring’ and ‘angular bow’ and described it as a Not closely datable – there is no clear evidence for the date ‘small specimen of La Tène III type, sometimes found in of the mirror burials. Roman contexts’. Subsequent comments on the date of deposition of the St. Keverne mirror have essentially fol- B3. IS A RADIOCARBON DATING AUlowed Smith (e.g. Dunning 1928: 77; Leeds 1933: 44). DIT NECESSARY FOR THE MIRROR Spratling (1970a: 14) elaborated on Smith’s analysis, indicating that the brooch is a ‘Nauheim-derivative’. Based on CORPUS? comments in Hawkes & Hull (1947: 312), he suggested that This section of the dating audit was conducted in order to it can be dated no more accurately than the 1st-century AD. determine whether a full radiocarbon dating audit of the Recent adjustments to the continental chronology founded mirrors would be beneficial in order to determine a greater on tree-rings and modifications of linkages with Mediterra- understanding of the time when mirrors were deposited. nean chronologies (outlined in Haselgrove 1997: 56) mean One important criterion needed to be met: would radiocarthat dating of La Tène D1 and D2 can be put back by up to bon dating provide greater accuracy than a general dating 100 years. The consequence of this for the St. Keverne audit based on current knowledge of brooch and pottery brooch, and therefore the date of deposition of the mirror, typologies? To answer this question mirrors found with is that La Tène D1b Nauheim brooches can now be dated to material potentially datable by accelerated radiocarbon dating were listed. Any mirror found with material which is Figure B.12 St. Keverne brooch (after Johns 2006)

133

Figure B.13 Simulation calibrated date for Arras I

Figure B.17 Simulation calibrated date for Latchmere Green

Figure B.14 Simulation calibrated date for Chilham Castle Figure B.18 Simulation calibrated date for King Harry Lane

Figure B.15 Simulation calibrated date for Aston

Figure B.19 Simulation calibrated date for Portesham now lost or would be unsuitable was excluded. A series of potential dates were then plotted (based on existing knowledge of mirror dating) using OxCal for each of the mirrors listed (see Figures B13-20).

Figure B.16 Simulation calibrated date for Dorton

What these projections show is that even with the highest degree of accuracy currently possible, because of de Vries effect perturbation IIIa, when calibrated the precision of the dates obtained would be no greater than the precision of the dates currently available from brooch and pottery chronologies in the later Iron Age in Britain. For example, in southeast England where brooch and pottery chronologies are most precise – a La Tène D2 brooch found

134

Figure B.20 Summary of estimated dates with a mirror would give a date range c.75 – 25 BC (J.D. Hill pers.comm). This level of precision is at least as good, or better, than would be theoretically possible for a calibrated AMS radiocarbon date. A full scale radiocarbon dating audit of the mirror corpus would therefore not be worthwhile. This does not mean, however, that it would not

135

be valuable to obtain radiocarbon dates for individual mirrors where dating from associated objects is currently imprecise, or that it would not be constructive to re-date some graves.

APPENDIX C IRON AGE OBJECTS DECORATED WITH MIRROR MOTIFS Amerden Sword Scabbard, Amerdon Bucks Decoration on the top of scabbard (Jope 2000: Pl. 208b). Positive motifs – Adjoined Double Cusp A, Fin, Trumpet Void A Hatching Negative Motif – Negative Crescent B, Negative Circle, Negative Cusp A, Negative Trumpet Void C Mount, Appleford, Berkshire Raised Decoration on the mount (Jope 2000: Pl. 294m). Positive Motifs – Fin, Armadillo A Negative Motifs – Fin Asby Scar Scabbard, Great Asby, Cumbria Decoration inscribed on the scabbard (Stead 2006: No. 203). Negative Motifs – Keeled Roundel, Trumpet Void A, Fin

Positive Motifs – Fin, Armadillo B, Cusp A Hatched Negative Motifs – Trumpet Void A, Fin Cambered Disc, R. Bann, Loughan Island, near Coleraine, Ireland Raised decoration on the surface of the disc (Jope 2000: Pl. 188). Negative Motifs – Circle, Cusp A, Trumpet Void B, Fin Terret, Bapchild, Kent Enamelled in the side of the terret (Jope 2000: Pl. 293d). Negative Motifs – Cusp A, Trumpet Void A, Fin Terret, Bolton Museum (unprovenanced) Enamelled on the side of the terret (Jope 2000: Pl. 293f). Positive Motifs – Circle, Fin, Keeled Roundel Negative Motifs – Cusp A, Fin Horncap, Brentford, Middlesex Incised on the top of the cap (Fox 1958: Pl. 4). Positive Motifs – Fin, Armadillo B, Cusp A Negative Motifs – Crescent Shape A, Fin Sword Scabbard, Bugthorpe, Yorkshire Incised on the surface of the scabbard (Jope 2000: Pl. 202). Positive Motif – Fin Hatching Negative Motif – Circle, Trumpet Void A, Fin Pair of Spoons, Burnmouth, Berwickshire Incised on the spoon handle (Jope 2000: Pl. 233a). Positive Motif – Armadillo A Hatching Negative Motifs – Circle, Cusp A, Trumpet Void B Chape Frame, Cambridge Inscribed on the chape frame (Jope 2000) (Figure D1). Positive Motif – Cusp Hatching Enamelled Strap-union, Canterbury, Kent Incised decoration (Jope 2000: Pl. 297c). Positive Motif – Circle, Trumpet Void A, Trumpet Void B Negative Motifs – Keeled Roundel, Fin Spoon, Cardigan, Wales On the spoon handle (Fox 1958: Pl. 70b). Negative Motifs – Keeled Roundel, Trumpet Void A

Figure C1. Decoration on Asby Scar sword scabbard Auchendolly, Terret, Auchendolly, Kirkudbright Incised on the side of the terret (Jope 2000: Pl. 293a). Negative Motifs – Circle, Cusp A, Trumpet Void A, Fin Hanger, Bagendon, Gloucestershire Pattern in the form of voids within the hanger (Jope 2000: Pl. 225m). Negative Motifs – Cusp A, Fin Collar, Balmaclellan, Kirkcudbright Incised on the surface of the collar (Jope 2000: Pl. 252d).

136

Tankard, Carrickfergus, Co. Antrim Incised onto a bronze mount on the wooden tankard (Jope 2000: Pl. 229 g-k) Positive Motifs – Fin, Armadillo B, Trumpet Void A Hatching Negative Motifs – Crescent Shape A, Fin Gold Collar, Clevedon, Somerset Incised on the terminal of the collar (Jope 2000: Pl. 120-1). Positive Motifs – Adjoined Double Cusp C, Armadillo B

Negative Motifs – Tricorne, Crescent Shape A, Fin Openwork Fitting, Cornalarach, Co. Monaghan, Ireland Pattern in the form of voids (Jope 2000: Pl. 195a). Positive Motifs – Armadillo B, Cusp A Sword Scabbard, Cotterdale, Yorkshire Incised on scabbard (MacGregor 1976: no. 143). Positive Motifs – Circle, Cusp A Bronze Shield Mount, Deal, Kent Cut-out of shield mount (Parfitt 1995: Fig. 23). Negative Motifs – Circle, Trumpet Void A Sword Scabbard, Deal, Kent Raised on the sword scabbard (Jope 2000: Pl. 205o). Negative Motifs – Stemmed-Pelta, Trumpet Void A Dowalton Loch, Wigtownshire, Ireland Enamelled ornament (Jope 2000: Pl. 294j). Negative Motifs – Cusp A, Fin Mount, Elmswell, Yorkshire Relief ornament (MacGregor 1976: no. 336). Positive Motif – Fin Negative Motifs – Tricorne, Fin Drinking Vessel, Elveden, Suffolk Ornamental roundel (Jope 2000: Pl. 226 e-f). Positive Motifs – Adjoined Double Cusp A, Trumpet Void C, Keeled Roundel Negative Motifs – Cusp A, Fin Sword Scabbard, Fovant, Wiltshire Incised on the scabbard (Jope 2000: P. 204l). Positive Motifs – Fin, Trumpet Void A Negative Motif – Fin

Figure C2. Decoration on the Bugthorpe sword scabbard Hatching Negative Motifs – Crescent Shape B, Fin

Bone Comb, Ghegan Rock, East Lothian Incised on the comb (MacGregor 1976: no. 274). Positive Motif – Adjoined Double Cusp A

Terret, Colchester, Essex Incised on the side of the terret (Jope 2000: Pl. 292g). Positive Motifs – Circle, Fin, Trumpet Void A Negative Motifs – Keeled Roundel, Fin

Wooden Bowl, Glastonbury, Somerset Carved onto the outside of the bowl (Jope 2000: Pl. 312a). Positive Motifs – Fin, Cusp A Hatching Negative Motif – Cusp A

Terret, Colchester, Essex Incised on side of terret (Jope 2000: pl. 293h). Positive Motif – Keeled Roundel Negative Motifs – Cusp A Linchpin, Colne Fen, Huntingdon, Cambridgeshire Incised decoration (Jope 2000: Pl. 300a). Positive Motifs – Circle, Fin, Armadillo A, Cusp A Negative Motifs – Keeled Roundel, Trumpet Void A Sword Scabbard, Congham, Norfolk Incised on the scabbard (Stead 2006: no. 103). Positive Motifs – Fin, Cusp A, Hatching

137

Tankard, Glastonbury, Somerset Ornamented detail on the back of the tankard handle (Jope 2000: Pl. 230j). Positive Motif – Circled Tricorne A, Fin Hatching Chape, Henley, Oxfordshire Relief decoration (Jope 2000: Pl. 208h). Positive Motif – Fin Negative Fins – Crescent Shape A, Stemmed Pelta, Circle, Trumpet Void A

Negative Motifs – Circle, Trumpet Void A

Hanger, Hunsbury, Northamptonshire Pattern cut away from hanger (Jope 2000: Pl.225o). Negative Motifs – Crescent Shape A, Circle, Cusp A Scabbard, Hunsbury, Northamptonshire On the surface of the scabbard (Jope 2000: Pl. 204). Positive Motifs – Circled Tricorne B, Hooked Fin A, Fin Hatching Negative Motifs – Crescent Shape A, Trumpet Void A, Fin

Mount, Llyn Cerrig Bach, Anglesey Relief work on the surface of the mount (Jope 2000: Pl. 184). Positive Motifs – Circle, Keeled Roundel Negative Motifs – Crescent Shape B, Stemmed Pelta, Trumpet Void B Shield Boss, Llyn Cerrig Bach, Anglesey Front of the shield boss (Jope 2000: Pl. 90). Positive Motifs – Armadillo B, Trumpet Void C Negative Motifs – Crescent Shape A, Circle, Trumpet Void A, Trumpet Void B

Suspension Loop, Icklingham, Suffolk (Jope 2000: 210e). Positive Motifs – Adjoined Double Cusp A, Fin Gold Torcs, Ipswich, Suffolk On the terminals of the torcs (Jope 2000: Pl. 116-7). Positive Motif – Adjoined Double Cusp A Negative Motifs – Cusp A, Trumpet Void A, Trumpet Void B, Fin Bronze Disc, ‘Ireland’ Relief on the surface of the disc (Jope 2000: Pl. 192). Negative Motifs – Circle, Cusp, Fin Spoon, ‘Ireland’ On the handle of the spoon (Jope 2000: Pl. 185b). Positive Motif – Keeled Roundel Negative Motif – Crescent Shape A, Circle, Trumpet Void C Sword Scabbard, Isleham, Cambridgeshire On the scabbard (Jope 2000: Pl. 206 h-k). Positive Motifs – Circle, Fin, Armadillo B, Cusp A, Trumpet Void A, Trumpet Void B, Trumpet Void C Hatching Negative Motifs – Negative Cusp A, Trumpet Void A, Trumpet Void C, Negative Fin Linchpin, King’s Langley, Hertfordshire At the end of the linchpin (Fox 1958: Pl. 52a). Positive Motifs – Circle, Fin, Trumpet Void B Negative Motif – Fin

Pottery Fragment P153, Meare Lake Village East, Somerset On surface of pot (Bulleid & Gray 1948: Pl. IX). Positive Motifs – Fin, Armadillo A Pottery Fragment P212, Meare Lake Village East, Somerset On surface of pot (Bulleid & Gray 1948: Pl. XIII). Positive Motifs – Circled Tricorne A, Fin Hatched Antler Spindle Whorl, Meare, Somerset Carved (Jope 2000: Pl. 3124m). Positive Motifs – Circle, Fin, Kidney Hatching Sword Scabbard, Meare, Somerset On the sword scabbard (Jope 2000: Pl. 207 a-e). Positive Motifs – Circled Tricorne B, Adjoined Double Cusp A, Hooked Fin A, Fin, Cusp A Hatching Negative Motifs – Tricorne, Cusp A, Trumpet Void B, Trumpet Void C Meyrick Helmet On the neck guard (MacGregor 1976: No. 189) Negative Motifs – Cusp B

Linchpin, Kirkburn, Yorkshire On the terminal of the linchpin (Jope 2000: Pl. 127L). Negative Motifs – Trumpet Void B, Trumpet Void C

Gold Torc, Needwood Forest, Staffordshire On the torc terminals (Fox 1958: Pl. 25). Positive Motif – Armadillo B Hatching

Terret, Lakenheath, Suffolk Incised on the side of the terret (Jope 2000: Pl. 292d). Positive Motifs – Fin, Keeled Roundel Negative Motifs – Circle, Fin

Norton Enamelled Mount, London Enamelled on the mount (Jope 2000: 297d). Positive Motifs – Circle, Fin, Cusp A

Bone Comb, Langbank, Renfrewshire Incised on the comb (Jope 2000: Pl. 305g). Positive Motif – Circle, Negative Motif – Tricorne, Fin

Openwork Roundel Pattern cut from roundel (Jope 2000: Pl. 96h). Positive Motifs – Circle, Fin, Cusp A, Trumpet Void B, Kidney

Sword Scabbard, Little Wittenham, Thames, Oxfordshire At the top of the scabbard (Jope 2000: Pl. 205m). Positive Motifs – Fin, Armadillo B

River

138

Terret, Owmby, Lincolnshire Enamelled on the side of the terret (Jope 2000: Pl. 293a). Positive Motifs – Fin, Cusp A, Circle Negative Circle – KeeledRoundel

Terret, Pentyrch, Glamorgan Incised on the side of the terret (Jope 2000: P. 290a). Positive Motif – Circle, Fin, Half-Moon Cheek piece A, Polden Hills, Somerset Enamelled on the cheek piece (Jope 2000: Pl. 296f). Positive Motif – Circled Tricorne D, Triangles Cheek piece B, Polden Hills, Somerset Enamelled on the cheek piece (Jope 2000: Pl. 296g). Positive Motif – Circled Tricorne C, Cusp Enamelled Horse Brooch A, Polden Hills, Somerset Enamelled decoration (Jope 2000: Pl. 298d). Positive Motifs – Circle, Fin, Armadillo B Negative Motifs – Keeled Roundel, Cusp A, Fin Enamelled Horse Brooch B, Polden Hills, Somerset Enamelled decoration (Jope 2000: Pl. 299b). Positive Motifs – Fin, Armadillo A Hatching Strap union, Polden Hills, Somerset Enamelled decoration (Jope 2000: 297b) Positive Motifs – Circle, Fin Negative Motif – Cusp A

Figure C3. Miniature shield 1 from Salisbury Strap union, Santon, Norfolk Enamelled on mount (Jope 2000: Pl. 296a). Positive Motifs – Circle, Fin, Trumpet Void A Negative Motif – Cusp A Strap union, Santon, Norfolk Enamelled decoration (Jope 2000: Pl. 297a). Negative Motif – Cusp A

Terret, Richborough, Kent Enamelled on the side of the terret (Jope 2000: Pl. 293e). Positive Motifs – Circle, Keeled Roundel Negative Motifs – Trumpet Void A, Fin

Scabbard, River Witham On front and back of scabbard (Jope 2000: Pl. 53d). Positive Motifs – Circled Tricorne B, Circled Roundel B, Keeled Roundel Negative Motifs – Trumpet Void A, Trumpet Void B

Horse-bit, Ringstead, Norfolk At junction of loop and bar (Fox 1958: Pl. 28). Positive Motifs – Circled Tricorne B, Half-moon Miniature Shield 1 – Salisbury, Wiltshire On the surface of the shield (Stead 1991d: Figure 12) (Figure D5). Positive Motif – Fin Hatching Negative Motifs – Stemmed Pelta, Trumpet Void A, Fin Miniature Shield 2 – Salisbury, Wiltshire On the surface of the shield (Stead 1991d: Figure 12). Negative Motif – Cusp A, Trumpet Void C

Bone Comb, Seacliffe, East Lothian Incised on the comb (Jope 2000: Pl. 305f). Positive Motif – Adjoined Double Cusp A Negative Motif – Fin Gold Arm Ring, Snettisham, Norfolk Relief ornament on the surface of the arm ring (Jope 2000: Pl. 111). Positive Motif – Adjoined Double Cusp A Negative Motif – Trumpet Void A, Trumpet Void B Pottery Fragment, South Cadbury, Somerset Incised on the pot (Jope 2000: Pl. 317e). Positive Motif – Circled Tricorne B Negative Motif – Trumpet Void C

Miniature Shield 4, Salisbury, Wiltshire On the surface of the shield (Stead 1991d: Figure 13). Positive Motif – Fin Hatching Negative Motif – Circle

Openwork Roundel, South Cadbury, Somerset Ornament cut into the roundel (Jope 2000: Pl. 208f). Negative Motif – Cusp A

Miniature Shield 5, Salisbury, Wiltshire On the surface of the shield (Stead 1991d: Figure 13). Positive Motif – Trumpet Void A Hatching

Shaft Finial, probably Suffolk Incised decoration (Jope 2000: Pl. 303e). Positive Motif – Circle, Fin

139

Terret, probably Suffolk Incised on the surface of the terret (Jope 2000: Pl. 288j). Positive Motifs – Circled Tricorne C, Fin, Keeled Roundel Negative Motifs – Cusp A, Fin Shield Boss, Tal-y-Llyn, Merioneth Front of shield boss (Jope 2000: Pl. 96). Positive Motif – Fin, Armadillo B Hatching Negative Motif – Stemmed Pelta

Bronze Comb, Warwickshire Incised on the comb (Portable Antiquities Scheme). Positive Motif – Fin, Armadillo B, Cusp A, Hatching Negative Motif – Cusp A, Trumpet Void A Terret, Weybread, Suffolk Enamelled on the side of the terret (Jope 2000: Pl. 292f). Positive Motifs – Fin, Armadillo B, Keeled Roundel Negative Motifs – Cusp A, Fin

Spearhead, River Thames, London Bronze mount on an iron spearhead (Jope 2000: Pl. 200c). Positive Motif – Adjoined Double Cusp A, Fin Hatching Negative Motif – Circle, Fin Sword Handle, Thorpe, near Bridlington, Yorkshire On the sword handle (Jope 2000: Pl. 213a-c). Positive Motif – Circled Tricorne B

Strap union, Westhall, Suffolk Incised on the mount (Jope 2000: Pl. 296c). Positive Motifs – Circle, Fin, Cusp A Negative Motifs – Cusp A, Fin Terret, Westhall, Suffolk Enamelled on the side of the terret (Jope 2000: Pl. 292a-b). Positive Motifs – Circle, Fin, Armadillo B Negative Motifs – Cusp A, Fin Scabbard, River Witham, Lincolnshire Front and back of scabbard (Fox 1958: Pl. 10e) Positive Motifs – Circled Tricorne B, Circled Roundel B, Crescent Shape A Negative Motif – Fin

Hanger, Thor’s Cave, Staffordshire Pattern cut away from hanger (Jope 2000: Pl. 225q). Negative Motif – Trumpet Void B Horse-bit, Ulceby, Lincolnshire On the outside edge of the bit (Fox 1958: Pl. 24). Positive Motifs – Armadillo B, Trumpet Void B Hatching

Cheek piece, Wooden Eckford, Roxburghshire Incised on the cheek piece (MacGregor 1976: No. 45). Positive Motifs – Circle, Cusp Strap-union, Unprovenanced Incised on the surface of the mount (Fox 1958: Pl. 52d). Positive Motifs – Circle, Fin Negative Motifs – Tear Drop, Fin

Hanger, Verulamium, Hertfordshire Pattern cut away from hanger (Jope 2000: Pl. 225l). Negative Motifs – Circle, Trumpet Void B

140

APPENDIX D MIRROR HANDLES D1. TYPOLOGY OF MIRROR HANDLES A typology of mirror handles was created by Fox (1949; 1958: 98-102, Figure 67a). He highlighted three main types of handle (Figure D1): bar (Type I), shaped (Type II) and looped (Type III). Looped handles were subdivided into simple (Type IIIA) and complex (Type IIIB).

Figure D2. Subdivision of bar handles Type IB: Bar handles with a Ring at each end The rest of the bar handles have a second loop at the top which had a slot cut into it to accommodate the plate. Sometimes, as in the case of Ingleton, this join was secured with a rivet. In addition to Ingleton the other examples of this type are Balleybogey, Garton Slack, Wetwang Village and Wetwang Slack. The handle of the unusual Brecon Beacons mirror is also of this type. Arras II could have had two loops (Fox 1949: Figure 1), although this cannot be confirmed as the handle is now lost. The Lambay Island mirror has a bar handle but it is too corroded to determine how many loops the handle has. Carlingwark is broken at the attachment to the plate so we cannot tell whether it had another loop or not.

Figure D1. Handle types (after Fox 1949)

Type II: ‘Shaped’ Handles Fox (1949: 28; 1958: 99) identified only one shaped handle – the handle of the Balmaclellan mirror. No other mirrors discovered since Fox’s publications have ‘shaped’ handles. Cessford (1997: 107, Figure 5) identified shaped handles as a particular type of Pictish mirror symbol, suggesting that other mirrors with handles of this type were in circulation but have not survived.

Type I ‘Bar’ Handles Bar handles have a thin shaft which is often circular in profile; the shaft is sometimes moulded and terminates with a loop. It is possible to subdivide bar handles into two groups; bar handles with only a terminal ring and bar handles with a ring at each end (Figure D2).

Type III: Loop Handles Fox (1949: 29) divided looped handles into ‘simple’ and ‘complex’. He then further subdivided ‘simple’ loop handles into two groups based on the presence (Group IIIA(i)) or absence (Group IIIA(ii)) of a boss at the join of the loop in the centre of the grip.

Type IA: Bar handles with single Ring Stamford Hill III was originally attached to a plate with a bar which had a slot cut into it to accommodate the plate; the join was secured with rivets. The simple shaft of the Arras I mirror handle ends with a bronze mount which has a slot in it to accommodate the plate.

Fox’s ‘Simple’ looped handle should represent a group in themselves, called ‘Type III Loop handles' here. Looped handles form a continual line. The out bent arms curve inward towards each other and meet in the centre of the grip. The lines then diverge and converge again to create a loop which is very often teardrop shaped. Fox’s (1949: 29) subdivision based on the presence or absence of a boss is maintained. However, the bossed handles can be further subdivided based on the presence or absence of an elaboration above the arms. Like the Portland handle this very often takes the form of a loop (Figure D3).

Colchester II has a bar handle which is octagonal in profile. The handle attaches to the plate with two arms which have a slot which the plate is wedged into. The terminal of the handle is broken so we do not know if it ended with a loop.

141

Type IIIC: Loop without Boss In this category Fox (1949: 32) placed the handle of Rivenhall I and Billericay II. No more handles can be placed in this group.

Figure D3. Varieties of Type III handle Type IIIA: Loop with Boss Fox (1949: 29-30) identified a number of handles in this group: Rivenhall I, Billericay I, Mayer, Bridport, St. Keverne and Gibbs (he also classified Portland and Stamford Hill II in this group which are reclassified in Group IIIB). Since Fox’s publication three more handles have been discovered which can be placed in this group: Bryher, Chilham Castle, and Latchmere Green. Type IIIB: Loop with Boss and Elaboration This group describes looped handles with an addition or elaboration above the arms of the handle; often this describes a simple loop ‘held’ within the handle arms. Examples of this type of handle include Portland and the handles of the newly discovered Pegsden and Oxford mirrors. Stamford Hill II has a different kind of embellishment to the handle arms; contained within an arch above the arms are two circles and four cusps in relief. A fragmentary metal detector find from Westerfield, Suffolk probably also represents a handle of this type.

Figure D4. Mirror handles with loops within their grips (after Fox 1958: Figure 66)

142

Type IV: Double Loop within the Grip Fox (1949: 32-5) identified a further group of handles with loops within the grip. The vast majority of these mirrors have two loops within the grip (Figure D4). The Holcombe mirror handle is highly unusual in that it has three loops. Often these handles also have a large terminal loop. The loops within the grip contain tear-shaped or circular voids. These handles were identified by Fox as Group IIIB. I have assigned them their own group because most of the mirror handles fall into this category and there is a lot of variation within this type. A number of similarities between particular handles in this group can be identified. Disney and Jordan Hill are two handles of this type which do not appear to have had terminal loops. The Disney handle in particular is interesting because it consists of two opposed simple loops and appears to directly reference Type III handles. Like Stamford Hill III (Type I) many of the arms of this type of mirror link directly into the edging to which they are attached by rivets (examples include Birdlip, Desborough, Holcombe and Nijmegen). The majority, however, (like Type III mirrors) have slots in the arms into which the plate was wedged (Aston, Bromham, Colchester I, Great Chesterford, Llechwedd-ddu, Old Warden I and Portesham). The form of the arms of this type of handle varies widely and many don’t even have arms; the juncture of the plate

and handle can also vary in form. Often this part of the handle is decorated with glass or may contain voids shaped like motifs used to decorate plates. The arms of the Aston handle are short, simple bars which are u-shaped in profile. The Bromham handle does not have arms; instead it has a flat base with two joined semi circles above it. Within each semi circle is a half-moon shaped void. The Portesham handle has a further loop at the juncture of the plate and handle. This loop is approximately the same size as the two loops which form the grip of the handle. These grip loops are unusually large. A handle which also has large grip loops is Lllechwedd-ddu. The arms of this handle are unusually wide. They form an extended cusp shape and contain two opposing fin-shaped voids. A number of mirrors have arms which curve up to end in a circle. The majority of handles of this particular form also have a circle above the centre of the arms. These circles are often decorated with glass. The arms of the handle of Colchester I end in circles. The handles of the Dorton and Great Chesterford mirrors as well as the handle from Compiègne and a metal detector find from the Essex/Sussex border (Mills 2000: 33) all have arms that end in circles and a central circle above the centre of the arms. A recent metal detector find from Great Tey, Essex could represent one of the circular elaborations on a handle of this type. The handle of Old Warden I also has circles incorporated within the arms of the handle that were probably inlaid with red glass. The arms of this handle are very unusual and have been compared to flowers or a bird’s head (Jope 2000). A further group of mirror handles within this type have semicircular shaped embellishments placed above the centre of the arms of the handle. These elaborations often have raised decoration, are inscribed with decoration, contain motif-shaped voids and are ornamented with red glass. The Birdlip mirror has three voids outlined by raised decoration. The plate, which is visible within the voids, is inscribed with decoration and ornamented with six small circles of red glass. The handle of the Desborough mirror by comparison is quite simple, and the elaborations above the handle arms comprise raised edges that outline a tear-shaped void. This area on the Holcombe mirror is ornamented with raised decoration and two circles of red glass on each side of the handle. While on the Nijmegen mirror the top of the handle also comprises a mix of raised decoration, voids and circles of red glass. Old Warden II is unusual in that the loops within the grips of the handle each contain three small voids rather than one large one. The juncture between the handle and the plate of this mirror contains four voids. Like a number of the other handles Old Warden II also has further elaborations which are incorporated within the terminal loop. These take the form of a circle, concentric and smaller than the terminal loop. Old Warden II has two circular-shaped elaborations incorporated within the terminal loop of the handle. Each of these circles contain a further three circular voids. Examples, include Birdlip, Desborough and Old Warden I. The handle of the Holcombe mirror has two facing keeled roundels incorporated within the terminal loop. This frames a

143

Figures D5 & D6. Type V handle and the Merlesford & Fison Way Handles further series of voids within the terminal loop of the handle: two circles; a cusp; and two very small fins. A number of recent metal detected finds probably also belong to this group, although it is impossible to be certain because these finds are fragmentary. These include two handles from Suffolk (Akenham and Badingham). Types V Handles In addition to these four identified types of handle, essentially formulated by Fox, a further group has been identified. Like Type II handles these handles are not as numerous as the other three main types. There are only two examples of Type V handle (Figure D5). These handles are characterised by a split shank or grip. They both have a terminal loop. The Lochlee handle is made of bronze and the Bac Mhic Connain handle is unique in that it is made of whale bone. Both of these handles have a slot to accommodate the plate. The contexts in which these handles were discovered are not well recorded. Like the Type II handle from Balmaclellan this type of handle appears as a Pictish symbol (Cessford 1997: 107, Figure 5) and it may be that they represent a later handle form peculiar to Scotland. Possible Mirror Handles There are two metal handles which look like they could belong to Type I: the example from Scotland, possibly from Merlsford; and Fison Way, near Thetford (Figure D6). Both of these handles have a small terminal ring and a moulded grip section. However, the grip section is very short and if they were mirror handles the rest of the handles have not survived. The handle of one of the buckets from Alkham, Kent includes a section which is similar in form to these two handles. Although these handles are unlikely to represent the remains of a bucket it is impossible to exclude the possibility that they could represent the remains of a handle for a type of object other than a mirror. It has been claimed that a metal object found on the surface of mound xxii at Glastonbury Lake Village in 1893 represents the remains of a mirror handle (Bulleid & Gray 1911: Pl. XLII). If this was a mirror handle its form would be unique. However, it has also been suggested that this handle was a piece of horse-harness

Mirror Handle

Handle Type

Region

Stamford Hill III

IV

Southwest England

Disney

IV

?

St. Keverne

IIIA

Southwest England

Gibbs

IIIA

?

Nijmegen

IV

The Netherlands

Mayer

IIIA

?

Brecon Beacons

IB

Wales

Birdlip

IV

‘Western’

Llechwedd-du

IV

Wales

Desborough

IV

‘Western’

Holcombe

IV

‘Western’

Portesham

IV

‘Western’

Akenham

IV

East Anglia

Badingham

IV

East Anglia

Westerfield

IIIB

East Anglia

Arras I

IA

East Yorkshire

Arras II

I?

East Yorkshire

Garton Slack

IB

East Yorkshire

D2. WHAT DOES THIS TYPOLOGY MEAN?

Wetwang Slack

IB

East Yorkshire

Compiègne

IV

France

Lambay Island

I?

Ireland

Ingleton

IB

Northern England

Baleybogey

IB

Northern Ireland

Bac Mhic Connain

V

Scotland

Balmaclellan

II

Scotland

Carlingwark

I

Scotlan

Fox (1949; 1958) related handle type to the date and region of production. Type I bar handles were seen to be characteristic of the iron mirrors from East Yorkshire as well as later mirrors from northern Britain (Fox 1949: 25-8). The Balmaclellan handle (Type II) was seen to be derived from Roman patera (Fox 1949: 28). Type IIIA looped handles were seen to be characteristic of earlier or less well made decorated mirrors from Southern England (ibid: 30-2) and later ‘western’ mirrors, with kidneyshaped plates had Type IIIB handles.

Lochlee

V

Scotland

Portland

IIIB

South England

Bridport

IIIA

South England

Jordan Hill

IV

South England

Bromham

IIIA

Southeast England

Aston

IV

Southeast England

Billericay I

IIIA

Southeast England

Billericay II

IIIC

Southeast England

Colchester I

IV

Southeast England

Colchester II

IA

Southeast England

Dorton

IV

Southeast England

Essex/Sussex

IV

Southeast England

Great Chesterford

IV

Southeast England

Latchmere Green

IIIA

Southeast England

Old Warden I

IV

Southeast England

Old Warden II

IV

Southeast England

Oxford

IIIB

Southeast England

Pegsden

IIIB

Southeast England

Rivenhall I

IIIC

Southeast England

Chilham Castle

IIIA

Southeast England

Bryher

IIIA

Southwest England

Stamford Hill II

IIIB

Southwest England

Table D1. The geographical distribution of handle types equipment. Palk (1988) has argued quite convincingly, however, that this object is not a piece of horse-harness or a mirror handle but actually a belt-fastener. The possible handle from Llanwnda, Pembrokeshire has a triangular connection to the plate.

These assumptions were tested in the light of all of the new finds made in the 50 years since Fox’s paper was written and recent developments in our understanding of the deposition date of mirrors. Mirror Handle and Region Patterns between handle type and region (Table D1) were first examined. We can only test where mirrors were finally deposited and the possibility that mirrors were exchanged throughout their lifetimes cannot be excluded. A number of patterns can be detected in Table D1. The first is to confirm Fox’s observation that the East Yorkshire mirrors tend to have Type I bar shaped handles. The majority have Type IB handles with a loop at either end. One possible explanation for this is that these handles were made of iron and a bar shape would have been much easier to produce when forging. This possibility is underlined by the fact that the Lambay Island iron handle is also bar shaped. As already outlined above all of the examples (albeit only three examples) of Type II and Type V handles are from Scotland. These unusual handles probably represent later types in use in Scotland well after the Roman conquest of southern Britain.

144

Mirror Handle

Handle Type

Date

Wetwang Slack

IB

400-200 BC

Wetwang Village

IB

400-150 BC

Garton Slack

IB

400-200 BC

Bryher

IIIA

125-80 BC

St. Keverne

IIIA

125-80 BC

Aston

IV

75-25 BC

Pegsdon

III B

75-25 BC

Chilham Castle

IIIA

75-25 BC

Latchmere Green

IIIA

75-25 BC

Colchester I

IV

25 BC - AD 5

Dorton

IV

25 BC - 0

Birdlip

IV

AD 1 - 60

Holcombe

IV

AD 25 - 60

Portesham

IV

AD 40 - 60

Brecon Beacons

IB

AD 75 - 100

Nijmegen

IV

AD 100 - 200

Relationship between Deposition Date and Handle Type 5 4 3 2 Handle Type

1 0

-400

-300

-200

-100

0

100

200

Deposition Date

Figure D7. Relationship between mirror handle and deposition date elaboration above the arms which is often ornamented with raised decoration and red glass.

Table D2. Date of deposition of selected handles and handle type The distribution of Type III and Type IV handles is less clear-cut. Type IV handles were not restricted to the ‘western’ mirrors. Some are found in southeast England, East Anglia and one was even discovered in France. Similarly Type III handles are not restricted to the south-eastern and Cornish mirrors. More similarities can be uncovered between mirror handles if we look closely at common features. For example, a group of Type IV mirror handles with circular elaborations on their arms can be identified. The larger ‘western’ mirror handles (with the notable exception of Portesham) do not have arms, instead the arms are incorporated into the rim around the edge of the mirror plates. These handles also have a semi-circular shaped

145

If we examine the relationship between handle type and deposition date (see Table D2 and Figure D7) we can see that there is some correlation. This I believe is entirely due to the process of typology – Fox created a typology with a ‘development’ of mirror handles in mind. This is entirely arbitrary as mirrors with some of the most complicated designs like Pegsdon have simple loop handles. Similarly, Portesham which would be included in Fox’s ‘western’ grouping does not have the characteristic handle ornamented with red glass and raised decoration. Summary Mirror handles can be grouped and these groupings can be seen to have some chronological and regional significance. However, there are many exceptions to any pattern which can be suggested. Handle type is therefore not a good indicator of deposition date or location. What is more interesting are the similarities that can be drawn between particular handles such as the handles with circular elaborations from southeast England or the handles that are ornamented by raised decoration and red glassenamel.

APPENDIX E TRACE ELEMENT ANALYSIS OF THE METAL CONTENT OF MIRRORS Mirror

Plate/ Handle

% Copper

% Tin

% Lead

% Silver

% Iron

% Arsenic

% Antimnoy

% Bismuth

% Nickel

% Zinc

Aston

Plate

87.5

11.6

0.1

0.05

0.08

0.1

0.07

0.01

0.05

0.01

Handle

85.5

12.8

1.1

0.085

0.15

0.15

0.35

n.d.

0.08

n.d.

% Cobalt

Billericay I

Handle

11 .75

0.13

0.03

0.09

0.02

0.02

-

0.04

-

-

Billericay II

Handle

8.75

0.49

0.08

0.07

0.19

0.11

0.01

0.1

-

0.05

Colchester I

Plate

12 .52

0.12

0.06

0.03

0.01

0.05

n.d.

0.06

n.d.

Handle

6.71

5.78

0.05

0.01

0.31

0.21

n.d.

0.10

n.d.

Dorton

Plate

90

9

0.09

0.07

0.08

0.08

0.06

0.006

0.08

0.005

Handle

87

8.6

0.87

0.14

0.36

0.2

0.35

0.006

0.06

0.01

Plate

84

10.2

0.25

0.049

0. 042

n.s.

0.04

n.d.

n.s.

0.011

Handle

87.6

10.4

0.48

0.084

0. 041

n.s.

0.17

0.01

n.s.

0.007

Glastonbury

Plate

84.2

12.7

0.24

0.052

0. 067

n.s.

0.17

n.d.

n.s.

0.008

Holcombe

Plate

85.8

11.2

0.12

0.18

0.61

n.s.

0.57

0.01

n.s.

0.22

Handle

88.3

9.6

0.05

0.04

0.25

n.s.

0.07

0.03

n.s.

0.29

Plate

88.5

11.5

Handle

87.6

12.4

Plate

88.6

11.2

Edge

86.1

13.3

Edge

86.2

12.9

Handle

86.4

12.7

Gibbs

Latchmere Green

Portesham

Rivenhall

St. Keverne

0.2 0.7 0.6

0.2

Plate

10 .47

1.17

0.05

0.1

0.75

0.08

0.06

0.06

-

0.06

Handle

8.86

0.36

0.04

0.32

0.47

0.23

0.02

0.1

-

0.02

Rivet

10 .63

0.58

0.05

0.1

0.55

0.16

0.02

0.07

tr

0.02

Plate

85.7

7.9

0.059

0.061

0.57

n.s.

0.08

0.005

n.s.

0.008

Handle

86.9

9.6

0.63

0.12

0. 031

n.s.

0.74

n.d.

n.s.

0.001

Table E1. The metal composition of ‘British Type’ mirrors (n.d. = no data, n.s. = not sought, tr = trace). The composition of the metal of a number of the mirrors has been analysed. This research has been conducted in an attempt to source the origins of the metal used to make mirrors and locate potential regions of production (Northover 1993) (see discussion in Chapter 3). The data are presented in the form of a table (Table E1). As a means of comparison a table of analysis of the metal composition of Roman type mirrors is also presented (Table E2). The data for these tables have been collated from a number of sources (Farley 1983: Table 1; Stead & Rigby 1989: Table 59; Northover 1993: Tables 6,7 & 8; Fitzpatrick 1997a: Table 1; Fulford & Creighton 1998: 333). Presented data and terminology follow the original data as far as possible. Where multiple samples have been taken

146

from one area of the mirror only the mean results have been presented. Interpretation of the Data There is a remarkable consistency in the type of metal used to make British type mirrors. This consistency in the composition of bronze is echoed throughout the British Iron Age (see Northover 1984a; 1984b; 1987; 1991a; Dungworth 1997a). Zinc is only detected in small quantities, indicating that Roman bronze working traditions had not been incorporated into metal working practices when British type mirrors were made (Dungworth 1996; Dungworth 1997b), confirming the dating evidence set out in Chapter 8. Copper content of the mirror plates and handles is roughly between 85-90%. The tin content of mir

Mirror

Plate/ Handle

% Copper

% Tin

% Lead

% Silver

% Iron

% Arsenic

% Antimony

% Bismuth

% Nickel

% Zinc

Colleyweston

Plate

68

22.5

9.5

Plate

69.31

25.65

4.96

n.m.

Trace

n.m.

trace

n.m.

n.m.

n.m.

Grave 9

Plate

65.2

19.6

13.6

0.09

1.62

0.26

0.08

0.26

Grave 13

Plate

66.4

19.8

9.9

0.07

0.6

0.6

0.07

Grave 48

Plate

69

23

6.4

Grave 138

Plate

64.5

17.7

14.5

0.04

1.1

0.6

0.06

Grave 222

Plate

67.5

21.5

8.3

0.1

0.15

0.07

0.3

0.06

Grave 325

Plate

59.4

19.4

16.3

0.06

0.9

0.08

0.1

0.9

Rosenaubergie bei

Plate

69.24

24.16

6.6

Plate

71.89

20.44

6.52

n.m.

0.05

n.m.

0.93

n.m.

-

n.m.

Handle

69.20

23.08

7.35

n.m.

0.14

n.m.

0.18

n.m.

none

n.m.

(Northamptonshire)

Cologne

King Harry Lane

0.2

0.25

Ungsburg

Upper Lombardy

Table E2. The metal composition of selected ‘Roman Type’ mirrors (n.m. = no mention). ror plates is around 10-13%. Tin content varies more for handles (between 7-13%); this is largely due to the presence of lead in some of the handles, presumably to facilitate casting. If we compare the composition of the metal used to make British type mirrors to Roman type mirrors we can see that the latter contain a much higher proportion of tin.

147

Often Roman type mirrors contain between 20-25% tin and 65-70% copper. This composition would have produced a much more ‘life-like’ reflection than British type mirrors but Roman type mirrors would also have been susceptible to brittle fracture.

APPENDIX F MIRRORS WITH DECORATED PLATES Aston, Hertfordshire (Rook et al. 1982) Billericay I, Essex (Fox 1958) Birdlip, Gloucestershire (Bellows 1881; Staelens 1982) Bromham, Bedfordshire Bryher, Isles of Scilly (Johns 2006) Bulbury, Dorset (Cunnington 1884; Cunliffe 1972) Chilham Castle, Kent (Parfitt 1998) Colchester I (Lexden Grange) (Fox 1958) Colchester II (Hyderabad Barracks) (Sealey 2006) Chettle, Dorset Desborough, Northamptonshire (Smith 1909) Disney (Fox 1958) Dorton, Buckinghamshire (Farley 1983) Gibbs (Fox 1958) Great Chesterford, Essex (Fox 1960)

Holcombe, Devon (Fox 1972; Fox & Pollard 1973) Latchmere Green, Hampshire (Fulford & Creighton 1998) Mayer (Fox 1958) Nijmegen, The Netherlands (Dunning 1928) Old Warden I, Bedfordshire (Spratling 1970a) ‘Oxfordshire’ (Burleigh & Megaw 2007) Pegsdon, Bedfordshire (Burleigh & Megaw 2007) Portesham, Dorset (Fitzpatrick 1997a) Portland II, The Verne, Dorset (Fox & Pollard 1973) Rickling, Essex (Sealey 2006) Rivenhall I, Essex (Rodwell & Rodwell 1986; Lloyd-Morgan 1993) St Keverne, Cornwall (Jope-Rogers 1873) Stamford Hill I, Devon (Spence-Bate 1871; Cunliffe 1988)

148

APPENDIX G THE SIZE OF MIRRORS Maximum Plate Width

Handle Length (mm)

Arras I

175

Aston

193

Mirror

Stamford Hill I

Overall Length (mm)

216

Stamford Hill II

102

157

Stamford Hill III

152

159

St. Keverne

Bac Mhic Connain

78

Westerfield

Balleybogey

121

Wetwang Slack

Balmaclellan

209

128

Billericay I

165

105

Billericay II

123

62

183

Birdlip

269

139

387

Brecon Beacons

201

Bridport Bryher

152

Bulbury (plate fragment)

49

Carlingwark

321

78

220

40 140

165

Table G1. Mirror dimensions Mirror

Maximum Plate Width (mm)

Region

110

Nijemegen

295

Netherlands

115

Birdlip

269

Western

81

Holcolmbe

260

Western

Desborough

258

Western

Stamford Hill I

216

Southwest

Dorton

210

Southeast

Portesham

210

Western

Balmaclellan

209

Scotland

Llechwedd du

208

Wales

98

Chilham Castle

136

72

Colchester I

178?

151

Colchester II

165

64

Compiègne

154

188

170

135

Desborough

258

152

350

Great Chesterford

206

Southeast

Dorton

210

127

302

Brecon Beacons

201

Wales

Disney

95

84

Pegsdon

198.5

Southeast

Garton*

156

148

264

Aston

198

Southeast

Gibbs

147

155

213

Old Warden I

197

Southeast

Glastonbury

143

44

‘Oxfordshire’

184

Southeast

Great Chesterford

206

299

Colchester I

178

Southeast

Holcombe

260

371

Arras I

175

Southeast

183

Ingleton

158

Latchmere Green

170

Southeast

Jordan Hill

83

Garton Slack

156

East Yorkshire

St. Keverne

154

Southwest

Bryher

152

Southwest

Glastonbury

143

Southwest

Wetwang Slack

140

East Yorkshire

Chilham Castle

136

Southeast

Latchmere Green

170

Llanwnda Llechwedd-du

88

263

144 208

Lochlee Crannog

133

298

86

Maiden Castle

62

Mayer

166

224

Nijmegen

295

325

Old Warden I

197

Old Warden II*

83

Table G2. The relationship between plate width and region

283

The data presented in Table G1 is an amalgamation of my own measurements and measurements from published reports (particularly Spratling 1972). The dimensions of the mirrors* are approximate and are derived from scale drawings they are only intended for comparison.

193?

‘Oxfordshire’

184

271.5

Pegsdon

198.5

Portesham

210

Portland II

146

Rickling

61

44

Rivenhall I

77

81

131

305 297

Past analysis of mirror size The size and shape of mirrors was examined by Fox (1973: 32) to illustrate an association between the size of a mirror and the location of deposition. She found that the

155

149

Mirror

Overall Length (mm)

Region

Mirror

Maximum Plate Width (mm)

Deposition Date

Birdlip

387

Western

Nijmegen

325

AD 100

Holcombe

371

Western

Birdlip

269

AD 1-60

Desborough

350

Western

Holcombe

260

AD 40-60

Nijmegen

325

Netherlands

Dorton

210

25-15 BC

Balmaclellan

321

Scotland

Portesham

210

AD 40-60

Pegsdon

305

South East

Brecon Beacons

201

AD 75-100

Dorton

302

South East

Pegsdon

198.5

75-25 BC

Great Chesterford

299

South East

Aston

193

75-25 BC

Llechwedd-du

298

Wales

Latchmere Green

170

75-25 BC

Portesham

297

Western

Garton*

156

300-100 BC

Old Warden I

283

South East

St. Keverne

154

125-80 BC

'Oxfordshire'

271.5

South East

Bryher

152

125-50 BC

Garton*

264

East Riding

Wetwang Slack

140

300-100 BC

Latchmere Green

263

South East

Chilham Castle

136

75-25 BC

St. Keverne

220

South West

Chilham Castle

188

South East

Table G4. The relationship between plate width and deposition date

Table G3. The relationship between mirror length and region plates of the decorated ‘western’ mirrors are larger than the decorated mirror plates from the southeast. Rook et al (1982: 31-2, Figure 9) plotted the approximate area of decorated mirror plates from South Britain in square centimetres against the distances of their find-spots from the Thames estuary, measured from Shoeburyness. These measurements were used alongside handle type, and the design and techniques of decoration, to draw technical relationships between mirrors and identify four regions where decorated mirrors were made: the London Basin; the East Anglian ridge with the Chilterns; the Jurassic belt; and the Southwest. Spratling (2008) sought to draw aesthetic significance from the size, shape and proportions of mirrors. He is concerned with aesthetics and the proportion of mirrors, particularly the relationship between overall length and plate diameter. Spratling has revealed that the ratio between the handle length and the plate diameter is consistent between some mirrors and has related these proportions to wider aesthetic tastes in the Iron Age and beyond. Interpretation of the Results In order to determine if the new data collated in this book would confirm past analyses a table of the basic dimensions of as many of the mirrors as possible was drawn up. Tables G2 and G3 sort all of the mirrors from largest to smallest in terms of the maximum plate width and overall length. Each mirror is assigned a broad region. If we examine the maximum width of the plate we can see that the five mirrors with

150

the widest plates are all from the western region except for Nijmegen, which shares many similarities with the western mirrors. There is then a large group of mirrors from the southeast with plates between 160-200mm in width. The mirrors from the southwest and East Yorkshire have the smallest plates. If we look at the overall length of the mirrors we can see the same pattern. The East Yorkshire mirror appears slightly higher up in this table because of the length of the handle. Chilham Castle, which is from the south-eastern region, is the shortest mirror and has the smallest plate. The small size of the mirror as well as the ‘poor’ standard of decoration have been used to argue that the Chilham Castle mirror is one of the earliest of the south-eastern mirrors (Parfitt 1998). It is argued here that although it may be early in date, it is just a small mirror within the context of the south-eastern group. To see if there is any chronological significance to the size of mirrors deposition date was charted against the size of the mirror plate (Table G4 & Figure G1). Examination of this data shows that in general mirrors with the largest plates are later in date and there seems to be a general trend for mirror plates to become larger over time. However, within each regional grouping there are variations. For example, the Dorton and Great Chesterford mirrors have much larger plates than the other mirrors from the south-eastern group. Size of plate can be no more therefore than an indication of the date of a mirror. Within different regional traditions of mirror making it was usual to make a plate an approximate size. Variation in size occurred within each regional grouping. Table G5 is intended to examine the findings of Spratling, who has

Mirror

Maximum Plate Width (mm)

Overall Length (mm)

Ratio of Plate Width/ length

Holcombe

260

371

0.70

Desborough

258

350

0.74

Balmaclellan

209

321

0.65

Pegsdon

198.5

305

0.65

Dorton

210

302

0.69

Portesham

210

297

0.71

Llechwedd-du

208

298

0.70

Old Warden I

197

283

0.70

'Oxfordshire'

184

271.5

0.68

Garton*

156

264

0.59

Latchmere Green

170

263

0.65

Great Chesterford

163

299

0.69

St. Keverne

154

220

0.70

Chilham Castle

136

188

0.72

Relationship between deposition date and maximum plate width (mm) 150 100

AD

50 0 -50 0

50

100

150

200

250

300

350

-100

BC

-150 -200 -250

Length (mm)

Figure G1. The relationship between plate width and deposition date than relating the dimensions of mirrors to universal notions of aesthetic taste, these patterns in the data relate to the particular traditions of making metal objects in Iron Age Britain, ways of measuring, making and laying out decoration which were handed down from master to apprentice.

Table G5. The ratio between length of handle and overall length suggested that various ratios between the length and plate diameter can be identified, including 7:10. The data do seem to corroborate Spratling’s pattern of consistent ratios between plate diameter and overall length. However, Spratling may be expecting too much from the data. Rather

151

Summary In summary, mirrors became larger in size over time. Size can also be used as an approximate indicator of the region in which mirrors were deposited, especially those from the southeast and the western mirrors. It can also be used as a rough indicator of the date when a mirror was deposited.

APPENDIX H VISUAL EXAMINATION OF MIRRORS IN MUSEUMS MIRRORS IN THE BRITISH MUSEUM 25/02/ 2004 Arras, East Yorkshire (P&E 1877.10-16.8) – The handle is iron and of bar shape (Type I) (see Appendix D). It has two bronze mounts, one at the top and the other at the bottom of the mirror handle. The handle terminates in a small iron loop. The two bronze mounts are interesting as both are decorated with tiny punch marks. On the terminal mount this pattern marks out the form of a tricorne. In the uppermost mount the pattern is more suggestive of a fin-shaped motif. Polishing/Repair – A section of the upper mount was damaged in antiquity, and a portion of it was displaced. A replacement section exactly fitting the form of the lost area was then slotted in and fixed to the mount using two small rivets. In the central section of the opposite face of this mount there is also a polished area. – The plate is circular and made of iron. It is smaller in proportion to the handle than other mirrors. Because of the resin used to conserve the plate very little information could be gleaned from the iron sections. – The plate is wedged into a slot in the top of the circular mount at the top of the handle. This attachment is further secured with two rivets. Aston Handle & Plate (P&E 1981, 05-04.1 & 1979, 10-02.1) – The handle is complete and is made of cast bronze. At some point it became detached from the plate, traces of which remain in the slot cast into the top of the handle. The handle is Type IV and is very heavy. It has a large terminal ring. The bottom half of the grip is made up of an upside-down tear-drop shape. The upper portion of the grip and arms, comprise an alpha-shaped loop. The handle can be compared to other handles in the group, such as Colchester I (Lexden Grange); however the design of the Aston handle is simpler. – The handle shows signs of wear on the lower inside section of the terminal ring. There is also some damage which is post-depositional. The handle has a slight bend to it. It also has a long scrape-mark running down the back – possibly attributable to ploughing which caused the handle and plate to become detached. – The plate shows signs of considerable post-depositional damage. It is circular in shape and decorated on one side. The decoration is hatched in an unusual and irregular manner. The decorated side is demarcated by a line circumnavigating the plate edge. The line is comprised of a series of inscribed arrow shapes. The decoration has been inscribed very deeply; in some areas it almost penetrates the full thickness of the metal. Something of the ‘original’ surface of the mirror plate can be seen on the reflective side, which appears to be golden in colour in some places.

152

– The handle and plate were originally attached by means of a slot cast in the arms of the handle. The plate was then wedged into this slot. No other alternative methods of securing the plate to the handle are visible. Colchester II (Hyderabad Barracks) (P&E 1986,05-01.1) – The handle is cast bronze. Only the top half and grip survive, the terminal loop is lost. The break looks relatively fresh indicating that if the loop was lost in antiquity the mirror was not curated for a significant period before it was deposited. Alternatively it could have become detached post-deposition; however, there is no indication of the terminal ring being recovered. The handle is enigmatic. It is a bar handle (Type I) but it is unlike any other known example. Starting from the bottom and working upwards, the grip is octagonal in cross-section rather than circular or oval. The grip tapers at the bottom down to the point of fracture. A decorative boss demarcates the beginning of tapering. It comprises two concentric circles which circumnavigate the grip. These are linked by vertical incised lines. The main body of the grip also tapers in the middle. The handle has two arms where it joins the plate. Each arm has two concave arches on the bottom and a convex arch on the top. The shape of the arms is accentuated by an incised arm which follows their outline 1-2mm from the edge of the handle. In between the arms at the top of the handle is a circular decorative element. This is incised with decoration on both sides of the mirror. The circular area is incised with a tricorne. The tricorne is outlined by an armadillo motif (Figure 4.5). – The handle grip shows some signs of wear at the edges, perhaps indicating the mirror had indeed been held by the handle. – The plate is severely corroded and very badly damaged. Only around 50% survives. On one side of the plate traces of decoration remain, although it is impossible to make out the pattern. Decoration is only visible at the bottom left-hand and top right-hand edges of the plate and towards the middle-right. It appears to be more extensive than is illustrated in Sealey (2006: Figure 3). The decoration at the bottom of the plate appears to extend below the arms of the handle, indicating that the plate was decorated before the handle was attached. The sections of decoration that are visible show an unusual technique of incising the outline and filling of motifs. The outline is drawn using what can best be described as rapid back and forth strokes giving the impression of a coherent single line from a distance. The motifs are not filled with basket-hatching; instead motifs are filled with inscribed lines which give the impression of fish-scales. This method of inscribing and filling motifs is paralleled on the Rickling mirror, also from Essex, Old Warden I and Chettle. The plate has no edge binding, although there is an inscribed line circumnavigating the plate a few millimetres in from the edge.

On the undecorated side the mirror looks to have been highly polished. Even today it is a deep gold colour and it still catches the light in certain places. – The handle is attached to the plate by wedging the plate edge into a groove at the top of the handle. There is no sign of any rivets and the attachment appears to be secure. Desborough Mirror, Northamptonshire (P&E 1924,01-09.1) – The handle is cast bronze and of a complex loop design (Type IV). There are no close comparisons although a handle fragment found recently at Badingham, Suffolk shows similarities. – Decorated side-up the handle shows some areas of polishing. The handle also shows signs of pitting, as parts of the alloy in the handle have moved to the surface through time. There are clear signs of wear on the bottom inside-section of the terminal loop. The loop shows signs of flattening and thinning and viewed decorated sideup the loop is much thinner on the right-hand-side than the left. – The plate is kidney-shaped and is decorated with an intricate pattern on one side. There are some areas of pitting but otherwise the plate is in excellent condition. On the undecorated side, traces of the original yellow colour of the plate are still visible. The decoration is stunningly well inscribed and hatched. Even so it is possible to see that the decoration was first marked out by scratching a fine line which was later traced over because in places the inscribed line deviates from the trace-line. – On the undecorated side of the plate there are visible stress lines and slight cracking in a semi-circular line running from each arm of the handle. The potential for breakage at this stress point, had the mirror continued in usage, would have been quite high. Pictures of the Nijmegen mirror also illustrate similar more pronounced stress lines and cracking. – The mirror also has a bronze edging to the plate. The edging is approximately the same width all the way around and it has a fine indented line running along the inside edge. It is attached to both arms of the handle by two rivets. The joining edge of the arm and the rim has been lapped to secure the joint further. Especially at the top of the mirror, the decoration is obscured slightly by the edging – indicating that the plate was decorated before the edging was attached. – The handle is secured to the plate by the means of a slot in the top section of the handle which the plate is wedged into. The joint is made secure with the use of four rivets, two in each arm of the handle, which also act to anchor the edging.

153

Holcombe Mirror, Devon (P&E 1971.04-01.1) – The handle is cast bronze, large and heavy. It is a complex-loop design (Type IV) but is unusual as the grip is made up of three circular elements. Other Type IV handles (e.g. Desborough) comprise only two elements in the grip section. The handle has been described in detail elsewhere (Fox & Pollard 1973). The top section of the handle, above the arms, was once decorated with four circular pieces of glass-enamel, only one now survives. – The plate is kidney-shaped and is decorated on one side. The design is very complex but is difficult to make out in some places and the pattern is unfinished, with large areas remaining unfilled. Some areas where the original pattern was marked out are still visible and there are some examples of mistakes in inscribing the decoration, such as evidence of ‘run-ons’. The plate is ever so slightly concave and the edges of the plate sit up when it is reflective side-up. It is unclear as to whether this was deliberate to aid in the reflection. It could have been caused by the weight of earth on the mirror plate when it was placed at the bottom of a pit. – The mirror plate has a bronze edging; this is attached to the arms of the handle by means of rivets. A small section of the rim is missing on the bottom lefthand-side (reflective side up) of the mirror, in the position where the rim has been riveted to one of the arms of the handle. – The handle and plate are attached by means of a slot in the top section of the handle; this join is further secured with the addition of two rivets (one on each arm) which also act to anchor the bronze edging. Trelan Bahow, St. Keverne, Cornwall (P&E 1873.1011.1) – The handle is cast bronze in the shape of a simple loop with a boss – Type IIIA. A similar handle can be seen on the Mayer mirror. The arms at the top of the handle have a groove to accommodate attachment to the plate. These arms curve inwards. At the point at which they meet, there are two raised rings aligned horizontally. The handle ends in a small tear-shaped loop. – The handle looks to have been repaired, possibly to rectify a casting flaw. There is a crack in the middle of the right-hand-side of the terminal loop (plate decoration side up). This has been filled with a small circular piece of bronze. – The plate is circular and decorated on one side. The outline of the decoration looks to have been inscribed with a much finer tool than the hatched lines. The decoration is in a by-partite design, comprising two touching circular areas aligned horizontally along the centreline of the plate. The plate margin is fringed with triangles which have been filled with hatching. This is similar to the visible decoration around the edge of the Bryher mirror. The edge of the plate shows signs of corrosion, especially

to the lower right-hand-side (decorated side up). The decoration on the lower edge of the plate is obscured by the arms of the handle. – The handle and plate are attached to one another by means of a slot pre-cast into both arms of the handle. The plate was wedged into this slot and the attachment secured using two rivets, one on each arm. Gibbs, no provenance (Gibbs Collection 1296,’70) – The handle of the Gibbs mirror is cast bronze and is Type IIIA – loop with boss. The terminal ring of the handle is missing, presumably broken off in antiquity. Unfortunately the techniques used to conserve the mirror have removed much of the surface. As a result, it was not possible to undertake wear analysis on the handle. – The plate is bronze. The techniques used to conserve the mirror have removed any signs of decoration on the plate, which has been previously recorded (e.g. Jope 2000: Pl. 250). – The handle and plate are attached to one another by means of a slot in both arms of the handle. The plate was wedged into this slot and secured by rivets. Disney, no provenance (1889,02-01.4&5) Handle – The handle is cast bronze and has a highly unusual form. It is of Type IV with multi-loops in the grip. It is remarkably fresh and shows no signs of wear. – The plate fragment is made of cast bronze and is decorated on one side. Like the handle the decoration is also unusual. It is generally well inscribed but the hatching is irregular. – The handle and plate were attached to one another by means of a slot in both arms of the handle. The plate was wedged into this slot and the join was secured by a rivet. Ingleton, Yorkshire (1945,11-03.1) – The handle is cast bronze. It is a Type I bar handle with a large ring at each end. On the sides of the top ring are two cast bronze ox heads. This feature is unique to this handle. The grip of the handle is moulded and it has a large terminal ring. – The handle shows some wear on the terminal ring. The handle has been repaired. The top ring has been re-secured onto the grip. This join was secured with a rivet. – The plate was attached to the handle by means of a slot in the handle. The join was secured by a rivet. Because of the position of the cow’s heads the shape of the bottom of the plate would have been unusual in order for it to fit into the slot in the top of the

154

handle. Remains of the plate still present in the slot appear to be iron. Jordan Hill, Dorset (1892,0901.1717) – The handle is cast bronze and is Type IV. The grip comprises a bar handle with two small circular voids. It has some simple moulding on one side and is plain on the other. The arms are simple, short and unusually narrow. Both contain a single rivet hole. – The handle is in good condition and shows few signs of wear. One of the arms is broken at the end. The fracture is very close to one of the rivet holes. – The plate was attached to the handle by two rivets, one on the end of each arm. ASHMOLEAN MUSEUM, OXFORD (MARCH 2005) Balleybogey, Balleymoney, Co. Antrim – The handle is made from cast bronze and is categorised as a Type I bar handle. The terminal loop of the handle is circular and decorated. The grip section is marked by three groups of closely spaced raised concentric rings. The top tear-drop shaped section of the handle is also decorated. An inscription on this section of the handle reads ‘Balleybogey bog, Balleymoney 1866’. The top section of the handle has a slot in the top of it to accommodate a plate. It also has a small circular hole at the very top of the handle, which would originally have accommodated a rivet to secure the join of the handle and plate. The slot section of the handle contains remains of corroded iron, suggesting that the plate was iron and that it corroded in the wet conditions of Balleybogey Bog. Viewed in cross-section, at the bottom of the slot in the handle is a circular rivet, or bar of bronze. It is possible that this may represent the remains of a bronze rim. – There are a number of areas of wear and scratching on the terminal loop especially on the inside surface of the ring. This area is unsmooth and there are a series of striations clearly visible on the surface. Further up the ring on both sides are areas consisting of rough scratches on the surface of the handle and also areas of flattening. There is also an area of the terminal ring towards the top that has been chipped away. Oddly there is more damage on one side of the handle than the other. CHELMSFORD MUSEUM 28/03/06 Rivenhall Mirror I, Essex – The handle is cast bronze and is a simple-loop design (Type IIIC). There are two rivet holes, one on each of the arms of the handle and a slot in each arm to accommodate the plate.

– There is a bar of metal running along the top of the handle. This is separate from the plate. – The plate is bronze and is in two fragments. The smaller of the two fragments is decorated on one side with clearly inscribed and hatched lines. The larger section of plate is also decorated although it is less clear. The hatching is irregular and some of the inscribed circles are poorly executed. The decoration is very deeply inscribed. This larger plate fragment has obviously been restored and the decorated side is covered with some kind of backing paper. Where the surface of the reflective side of the mirror is visible it is very smooth. – The handle is attached to the plate via the slot in both arms of the handle. The join was secured by three rivets and strengthened by a bar.

them to accommodate the plate, some of which is still in situ. This section of plate is decorated on one side and shows a run-on error in the inscribing of the design. – The terminal ring of the handle shows signs of extreme wear, especially at the bottom of the ring and the left-hand edge (decoration side up). The bottom section of the grip also shows signs of wear. – In addition to the section of plate still attached to the handle, there are three bronze plate fragments, all of which are decorated on one side. The standard of inscribing the outline of the decoration is very high. It is not inscribed deeply but it is clearly inscribed. The hatching is irregular. As with most of the decorated mirrors the decoration sometimes strays off line and there are some run-ons. The hatching also extends past the outline in some places.

COLCHESTER MUSEUM 29/03/06 Colchester I (Lexden Grange) – The handle is cast bronze and is of a complex loop design (Type IV). It has a large circular-shaped terminal loop. A grip formed of two opposing tear-shaped areas and the arms of the handle both curve around to encompass a circular area. Each arm and circle has a slot to accommodate the plate. A section of plate still remains in situ. However, this slot has been made too deep; as a result the slot is visible on the underside of each arm, meaning that the edge of the plate is also visible. This can also be seen on the Rickling mirror (see below). – There are signs of polishing on one side of the handle just before the point when the arms branch out. Plate Fragments – The bronze plate is now in four fragments. One small section of the plate and rim shows that the mirror was polished before the rim was attached as there is an area visible on the plate where some of the rim is missing. The second plate fragment is slightly larger and also has a section of the rim still attached to it. This section has some very faint areas of decoration extending underneath the rim – also illustrating that the mirror was decorated before the rim was put in place. A third small section of plate is also decorated on one side. The fourth fragment is much larger and shows faint traces of decoration and gives an indication of the overall pattern of the design. The motifs are hatched using zigzag lines rather than basketweave. – The handle and plate look to have been attached simply by wedging the plate into the slot in the top of the handle. Billericay I, Essex – The handle is cast bronze and is in two pieces. It is broken in half roughly at the mid-point of the grip. The handle is of bar handle type (Type I). The bottom section of the handle consists of the bottom half of the grip and the terminal ring, the uppermost part, the lower section of the grip and the arms of the handle. The arms have a slot in

155

– The handle and plate are attached by wedging the plate into a pre-cast slot in the handle. The join is secured by two rivets, one on each arm of the handle. Billericay II, Essex – The handle is cast bronze and is of bar type (Type I). It is very simply formed and has a tear-drop shaped terminal loop and a plain bar handle. There is a slot in the top to accommodate the plate and a fragment still remains in situ. The handle shows very few signs of wear. It is also one of the smallest of the known mirror handles. – A large fragment of the undecorated bronze plate survives. – The handle and plate are attached by means of a slot in the arms of the handle which the plate has been slotted into. This joint is secured with two rivets, one on each arm. The rivets are a slightly different colour to the handle, indicating that metal of a different composition may have been used to make the rivets. Rickling, Essex (now in the Saffron Walden Museum) – The handle is cast bronze and fragmentary. Only one of the arms and a small section of the grip survive. The remains of the handle are insufficient to be certain of type but the grip contains multiple loops (Type IV). A section of bronze plate, decorated on one side, remains attached to the handle. This section of plate is bent, probably due to ploughing damage. – Two other sections of bronze plate survive; one very small undecorated piece and a larger section, which is inscribed with decoration on one side. The decoration is unusual. The outline of the pattern is inscribed with zigzagged, rather than straight, lines.

– The attachment of the handle to the plate is unusual. Rather like the handle of Colchester I, the slot in the surviving arm of the handle has been cut too deep and is visible on the underside of the arm. A section of plate extends out from this gap and looks to have been hammered flat to secure the plate in place, rather like the hammering of a rivet head. This suggests that the plate had two wedge shaped protrusions which would have slotted into the gaps in the arms to secure the join. NATIONAL MUSEUM OF SCOTLAND 04/04/06 Unknown Provenance – This artefact is potentially the terminal loop and bottom section of the grip of a mirror handle. The loop looks quite small and the handle is slightly curved in profile. It could be from a simple bar mirror handle (Type I) (but see Appendix E). It is very difficult to identify signs of wear on the terminal loop as much of it is severely corroded. Where the surface of the handle is visible it appears to be relatively fresh and unworn. – It could be a mirror handle in the form of similar examples from East Yorkshire, which have very small looped-terminals. Even if the handle had only three looped sections to the grip it would still be around 70-80mm long, which is short for a mirror handle but is comparable to the lengths of the Rivenhall I and Billericay II handles. This would be an unusual mirror handle but it could be seen to replicate wrought iron forms. – The terminal loop is very small and other than Fison Way there are no close parallels. The handle is also curved in profile, which although could be as a result of damage could also indicate that it may have been a handle for another object form. Bac Mhic Connain – The Type V handle is made of whale bone and is of an unusual form only paralleled by the cast bronze Lochlee mirror handle. The handle is marked ‘GNB 61’ and ‘11-21-68’. It has a circular terminal loop. The grip is comprised of a V-shape. The arms form a cross-bar to the V. In the top of the handle there is a slot, presumably cut to accommodate a plate. There is no sign of any metal corrosion in the slot which would indicate what type of metal was used to manufacture the plate (if indeed it was ever made). There is however, a slight notch, or step, to the V-shaped profile of the notch, perhaps indicating that something may have sat in the notch at some point in the past, causing the wear. On the other hand, the handle is relatively unworn (except some polishing around the inner circle of the terminal loop) and there is no rivet hole in the top of the handle which would have acted to secure the joining of the plate. It could be that the mirror was never finished or that the handle was meant only to imitate other mirror handles. Carlingwark

– Bronze mirror handle of ‘bar’ form (Type I). The handle has an oval-shaped terminal ring. At

156

the top of the terminal ring is a label – ‘DW 77’. The grip comprises of five bands of alternating flat and convex profile. The arms no longer survive. The break at the top of the grip cuts across at an angle and looks relatively fresh. If the handle was broken in antiquity the time between breakage and deposition would have been relatively short. It is unclear if the break was deliberate, or accidental, however it would have been relatively easy to repair so there must have been a good reason why this wasn’t done. This part of the handle is very similar to the Ingleton handle where the bar joins to the top ring. It therefore could originally have attached to a plate by a ring which is now lost. The handle is also similar to the bar-shaped handle from Mount Batten, Plymouth, so an alternative form of attachment to a plate could have been via a pair of arms. Based on observations of other mirrors and a similarity with the Ingleton handle, it is suggested that the handle is more likely to have attached to a plate by a ring rather than arms. – The terminal loop shows some signs of wear on the lower section to one side. Overall, however, the handle is in very good condition. There are no signs of repair. Balmaclellan – The mirror is enigmatic. There are no other direct comparisons although similar shaped mirrors are depicted on some Pictish stonework. The handle is made from cast bronze, is scoop-shaped (Type II) and resembles the handle of a Roman patera. At the bottom of the handle three crescent-shaped wedges have been cut out of the metal to produce a whirligig pattern. At least five concentric circles, outlined by closely spaced incised dots, circumnavigate the whirligig. There are six rivet holes in the grip of the handle which secure two panels (only one of which survives) that perhaps acted to strengthen the handle. The edge of the handle is surrounded by edging which would have acted to further extend rigidity. The edging terminates at the junction of the handle and plate. There are no visible traces of any fixing of this rim in addition to the curve of the handle. The edging would have been further secured by the overlay of a second rim circumnavigating the plate. – Both plate and handle are cut from one piece of sheet bronze. The plate is plain and is edged. The edging overlays the plate surrounding the edge of the handle and is secured by two rivets one on either side opposite the transition of the handle and plate. Much of the edging is missing from one half of the mirror. – Three additional panels survive out of a probable four. One has already been described and probably acted to strengthen the handle. The other missing panel would have probably been attached to the other side of the handle. These two panels were attached using five rivets (already described). In addition a further 2 panels were also present positioned above the junction of the handle and plate, one on either side of the mirror. Both of these survive, one in situ and the other loose. They were

probably soldered to the mirror surface. Both panels have raised decoration. The sheet metal is extremely thin and the decoration was made using a former. This type of decoration is unique in the context of a mirror. The whirligig on the handle fits in with motifs on other mirrors, although the method of cutting out the shape rather than inscribing is also unique. The decoration on the plaque still uses familiar motifs such as trumpets and fins but it also includes flowerlike elements as well as crescent motifs. The overall composition of the panels is not exactly symmetrical and is framed by a series of punched bosses around the edge. This echoes the mirror itself which is also bound at the edge. AYLESBURY MUSEUM 11/04/06 Dorton, Buckinghamshire EXAMINATION OF SCIENTIFIC ELECTROTYPES IN ADDITION TO VISUAL ASSESSMENT OF THE MIRROR Handle – The handle is made from cast bronze it is of multi-loop design (Type IV). The terminal of the handle is a ring. The grip is made up of two opposed tear-drop shaped sections with a raised ring in the middle. The arms extend out and curve upwards terminating in circles. There is also a circular nodule in the middle of the two arms. The handle is paralleled by the Great Chesterford mirror and the mirror handle from the Oist at Compiègne, France. Each arm of the handle ends in a crescent, a feature unique to the Dorton mirror. Wear/Repair/Damage – The terminal loop of the handle is worn, in the bottom section. It has been repaired in antiquity. At some point in the past the terminal loop became detached from the main body of the handle. So it was either reattached or a new one was made and attached to the bottom of the handle. Some attempt was made to disguise the repair by moulding some decoration onto the handle. In what is likely to be post-depositional damage, the handle is also bent upwards. When the mirror is lying flat, the terminal ring of the handle is approximately 30 mm higher than the point at which it attaches to the plate. Plate – The plate is made from sheet bronze and is circular in shape. It is decorated on one side. An inscribed line runs just inside the edge of the plate. The plate is in a very fragmentary state and would have required much restoration work. The execution of the decoration is good although the outline wanders in places and the hatching sometimes stray outside the outlines. The hatching is irregular; in areas where the motif is narrow the hatching is replaced by a series of perpendicular lines. Attachment of Handle to Plate – The plate is wedged into a slot in the top of the handle the join is secured by three rivets, one in each circular area at the top of the mirror handle. Each of the voids in the middle of the three circular areas would have presumably originally been filled with enamel.

157

DORCHESTER MUSEUM 26/04/06 Portesham Mirror, Dorset – The handle is cast bronze and of multi-loop design (Type IV). Two larger loops make up the top and bottom sections of the handle and the grip is comprised of two smaller rings with linking bands. The arms of the handle are formed of three shaped bands equally spaced along the body of the arm. There is a casting flaw at the very top of the terminal ring, which has been filled. The handle is not straight: the rings do not match up exactly to create a vertical handle. – There are wear marks on the bottom section of the terminal loop. This section is noticeably thinner than the rest of the loop. There are also signs of polishing on each of the nodes linking the loops of the handle if the mirror is decorated side up. The other side of these linking nodes is decorated with a series of closely-spaced perpendicular horizontal lines. There are also corrosion marks on the lower section of the second bottom loop of the handle. – The plate is made of sheet bronze. It is kidneyshaped and decorated on one side. The decoration shows some areas of run-ons. The decoration also extends under the edging at the very top of the mirror and at the bottom close to the arms of the handle. There are two interesting points to note on the area of the plate visible within the top loop of the handle. First a small circle with a tail is inscribed in this area. This is highly unusual and is certainly unique among mirrors. The second point of interest is that the bottom surface of the plate is notched. This is similar to notching on the bottom of the Birdlip mirror and suggests that the plate may have originally been prepared for a different handle. – The plate is wedged into a slot in the top ring and arms of the handle. The join is secured by two rivets located in the top/middle sections of the upper ring. Maiden Castle, Dorset – The iron plate fragment is extremely corroded and has also been covered in resin so very few details can be gleaned from the plate fragment. – The mirror has a copper-alloy edge. The edging is u-shaped in profile. A capping device is utilised to secure two pieces of the binding to the rim, a unique feature on mirrors. Bridport, West Bay, Dorset Handle – Bronze mirror handle of bar type (Type I). The handle is corroded and in poor overall condition. The terminal of the handle is a tear-drop shaped loop. The bar is plain except for a raised circular ring in the centre. The arms of the handle are also plain. There is a slot in the top of each arm to accommodate the plate which is missing. Each arm also contains a hole for the rivets used to secure the join between the handle and plate. One of the rivets is

still in situ. As with the Rickling mirror the slots in the arms extend all the way through, meaning that a small void is visible on the underside of each arm. The slots in the arms of the mirror are open extremely widely. This is probably caused by corrosion of the surviving bronze rivet. Wear – There are some signs of wear to the innermost bottom edge of the terminal ring. The extent of wear is difficult to make out due to the general poor condition of the handle. The Verne, Portland – Large section of bronze mirror plate, decorated on one side, including a section of the edge of the mirror. Due to the way the mirror has been preserved it is very difficult to make out the decoration or any signs of wear. The areas of decoration it is possible to make out are hatched very finely and regularly. There is a thinning of the mirror plate at the edge, perhaps indicating that it was originally edged. On the opposite side of the mirror it is labelled ‘1923-3.8’. MUSEUM OF THE IRON AGE, ANDOVER 05/05/ 06 Latchmere Green, Hampshire – The handle is made of cast bronze. It is in a fragmentary condition and five pieces survive. The handle is of a simple-loop design with boss (Type IIIA). It has a tear-drop shaped loop at the terminal end of the handle and inside this bottom section is a second circular terminal loop. The grip of the handle has two concentric rings spaced approximately 10mm apart. The top of the handle ends in another circular loop with arms extending from it. The arms of the handle have slots in them to accommodate the mirror plate. Although the handle is broken the bronze is in excellent condition, there are some small areas of corrosion but the handle looks to be in mint condition with no signs of wear. The top right arm of the handle (decorated side up) is broken into two pieces along the line of the slot. This is very interesting as it affords a view inside the slot. The slot extends 17 mm into the handle. – The plate is made of sheet bronze and is in two fragments, one main section and a second smaller fragment. It is decorated on one side. The engraving of the decoration is very fine and is executed excellently with very fine levels of detail. The hatching is deliberate but not uniform. The top left-hand side of the handle is still attached to the main section of the plate. The decoration resembles the Rivenhall mirror in that decorative motifs are included within the lines of larger motifs making up the outline of the decoration (Figure 8.2). Whereas the design of the Rivenhall mirror does this quite simply, the design of the Latchmere Green mirror goes a step further and divides up the area within outline motifs in very complex ways. Another important detail at the bottom of the plate, in the left-hand corner near the handle, is a line just inside the edge of the plate. This could be another example of the lined edge found on the Pegsdon mirror. It is possible that these lines at the edges of apparently un-edged mirrors are meant to perform the same

158

function as the edging – a conceptual separation of the plate from the outside world. – The handle is attached to the plate by wedging the plate into a slot in the handle. This join was secured with a rivet positioned in the centre of the top of the top loop of the handle. NATIONAL MUSEUM OF WALES, CARDIFF 17/05/06 Llechwedd-du, Gwynedd – The handle is cast bronze and is of complexloop form (Type IV). It has a circular terminal ring. The grip is made up of two smaller rings, joined to each other, the terminal ring and the arms by vertically aligned ovalshaped rings. The arms of the handle are very unusual and the design causes the handle to be exceptionally heavy. The arms form a cusp shape with two fin-shaped voids. This means that the plate surface is visible through these voids. A section of plate remains in situ in the slot of the handle illustrating the line of breakage. The cross section of the slot of the handle, also showing a piece of the plate, is informative. The slot is U-shaped in cross section and is too wide to grip the plate along the whole of the line of the slot. To resolve this some kind of packing, possibly extra sheet-metal, has been positioned within the slot to make the join more secure. As with the Rickling mirror that slot extends too far down the handle arms, meaning that voids are visible on the underside of each arm. – There are some signs of thinning and wear on the bottom half of the terminal ring. This may be due to actual wear but could also be explained by differential corrosion of a potential casting flaw. There are some areas of polishing on the vertical oval rings of the handle. There is also a worn or polished area on the underside of one of the arms. This could be as a result of a possible repair or may have been caused by heavy filing. – The plate is made of sheet bronze and is undecorated. The plate has U-shaped edging. On one side of the plate, in the top right-hand corner, there is a large area of discoloration in the patina of the bronze. This could have been caused by the bronze bowl which was found with the mirror, which may have been sitting on top of the plate when it was deposited in the ground. – The join of the handle to the plate is secured by wedging the plate into the slot in the top of the handle. This join is secured by the addition of the edging. Brecon Grave Group (97.7 H) – The handle is made of cast copper-alloy (Type I). The design is unusual; the closest parallel is probably the handle from Balleybogey Bog. The terminal of the handle is a large triangular shape it has two small voids in it, one on either side of the loop. The grip is very small and is made up of a single ring-shaped protrusion. The

top of the handle is also triangular shaped with a void in the centre. – The plate is made up of three components: a central iron plate; a copper-alloy collar or band surrounding the iron plate; and a copper-alloy edging. Very little remains of the iron plate which has corroded. The copper-alloy collar is inscribed on both sides with unusual decoration, which is barely visible. – The handle is secured to the copper-alloy collar by means of a slot in the top of the handle. BEDFORD MUSEUM 15/06/06 Bromham Mirror, Bedfordshire – The mirror handle is made of cast bronze and is of complex loop design (Type IV). The terminal loop of the handle is missing. The grip is composed of two opposing tear-shaped loops. The underside of the arms is flat. The uppermost section is comprised of two convex curves linked in the centre. The handle arms contain two halfmoon shaped voids, through which the plate is visible. When the handle was discovered it was in two separate pieces broken along the join of the arms to the grip. The two parts have now been rejoined. The arms have a slot to accommodate the plate. At the top of the convex curve of each side of the top section of the handle is a rivet. – There is clear polishing/wear on the horizontal ring (polished side-up) linking the two tearshaped sections of the handle grip, which has been flattened. – The plate is in a very fragmentary state. It is made of sheet bronze and is decorated on one side. The decoration on the whole is inscribed well. However, the circles within the design do not link up well, which perhaps indicates the design was inscribed rather than chased. There are also further mistakes and areas where the design does not link up. The area immediately surrounding the handle has a different colour to the patina of the metal. This may indicate that the mirror handle had been replaced in antiquity, suggesting that a period of time elapsed between use and deposition. – There are signs of stress and fracturing on the plate around the area of the handle indicating that the mirror was held by the handle and used in life. – The plate and handle were attached to one another by wedging the handle into a slot in the top of the handle. This join was further secured by two rivets. Old Warden I, Bedfordshire – The handle is made of cast bronze and is of complex loop design (Type IV). The terminal loop of the handle is a circular ring. At the top of the central loop of the terminal ring, directly in line with the grip, there is a smaller circular shaped area, which would originally have been

159

enamelled. The grip is composed of two opposing tearshaped areas joined in the centre by a very thin horizontal ring. The arms of the handle are formed of two semicircular curving areas. A small fin shape motif is cut out of the underside of each of these semi-circular areas. Within each of these semi-circles is a large circular area, probably originally containing glass-enamel. On the undecorated side of the mirror this circular area is decorated by an inscribed tricorne motif. This is similar to the handle of the Birdlip mirror which is decorated in the same position with an ADC motif. – One side of the top tear-drop shaped ring of the handle grip is cracked, possibly because of a casting flaw that was never repaired. The handle looks to be in very good condition with no signs of wear or polishing. – The plate is made from sheet bronze and has been heavily restored. It is surrounded by edge binding which is attached to the mirror in a series of segments. The mirror plate is decorated on one side. The outlines of the decoration are very deeply inscribed, the hatching is shallower. These lines show through on some areas of the reflective side of the plate. The outline of the decoration deviates from the guidelines (still visible in places) in some areas. The decoration also extends underneath the rim in some places, indicating that the plate was decorated before the rim was secured. – The handle is joined to the plate by means of a slot in the top of the handle and two rivets, visible in the two circular areas of the arms of the handle. These rivets would have originally been hidden behind the glass-enamel. GLOUCESTER MUSEUM 19/06/06 Birdlip, Gloucestershire – The handle is cast bronze and of complex loop design (Type IV). It has been described in detail elsewhere; however, of note is the large and substantial size of the handle and the red glass-enamel decoration which is still in place. Like the Old Warden I handle there is a second circular area within the circular void of the terminal ring. This circular area is decorated by an ADC; the voids within this motif are decorated with red glassenamel. The positive cusps are not hatched; instead they are stippled or punched. – There is clear evidence that the circular section within the terminal ring of the handle is either a replacement or has been repaired. There is a clear mark on the top of the terminal ring where some of the metal in the ring has been cut away and the join of the circular area has been precisely fitted into the gap in the metal. It is also possible that this circular area was a later addition to the handle, indicating that a period of time elapsed between the manufacture and deposition of the mirror. The metal of the terminal ring is noticeably thinner at the bottom of the handle. There is also an area of

polishing in one of the vertical rings that joins the loops of the grip. – The plate is made of sheet bronze and is decorated on one side. The outline of the decoration is very deeply inscribed and the line is quite broad in places indicating that the outline has been re-inscribed. The quality of hatching is very fine and it follows a regular pattern. Some areas of decoration are obscured by the rim, indicating that it was fixed onto the mirror after the plate was decorated. The plate is not circular but is kidney-shaped. At the centre of the plate is a punch mark. It is unclear if this damage is representative of a deliberate action in antiquity or whether it occurred when the mirror was rediscovered. However, bowls found in the same burial group also have holes punched in them. The undecorated side of the plate still retains a good surface. There are no stress marks close to the join of the handle that indicate that the mirror has been held by the handle for long periods. There is a slight tear in the left-hand edge of the plate. – The edging is made of bronze and would originally have surrounded the full extent of the edge of the plate. The edging is now partially missing on the left-hand side (decorated side up). The area of the plate exposed by the missing edging shows a series of striations on the surface of the plate which would have been hidden. Perhaps these were made to secure the join of the edging to the plate surface. It could also have served to mark out the position of the edging, when the plate was being decorated to highlight the limits of the design-field. – The handle is fixed to the plate by wedging the plate into a slot in the handle. The edging acts to secure this join. The edging is attached to both arms of the handle, giving the overall impression of a seamless join. The join between each arm and each end of the edging is made by a lap joint. This is secured by two rivets on each join.

160

LUTON MUSEUM 18/01/07 Pegsdon, Shillington, Bedfordshire – The handle is made of cast bronze. The handle is of Type IIIB – loop with boss and elaboration. It is very similar in form to the handle from Oxfordshire and Portland. The ring above the arms is not integral to the rest of the handle. It is formed of one piece of metal joined at the bottom, which secured to the plate with three rivets. In the void of the ring the visible area of the mirror plate is inscribed with a tricorne. This tricorne is inscribed into both sides of the plate. – The handle shows no signs of wear and is remarkably fresh, although there are some signs of polishing on the central boss. There is a casting flaw on the bottom section of the terminal ring. – The plate is made of sheet bronze and is decorated on one side. The decoration is very finely inscribed. There are no obvious compass marks or guidelines; however, some deviations in the line of the decoration are characteristic of the use of a compass to mark out the design. The outline of the decoration is more deeply inscribed than the hatching. There is a series of small inscribed lines at right angles to the edge of the plate on both sides of the mirror. In the centre of the reflective side of the plate are two parallel scratch marks presumably made by a spade when the mirror was discovered. – The handle is joined to the plate by means of a slot in the top of the handle. This join is secured by two rivets, one visible near the end of each arm.

REFERENCES Anon. 1995. Catalogue of Antiquities sale LN5398. Lon don: Sotheby’s. Anon. 2002. Portable Antiquities Annual Report 2000-1. London. Addey, C. 2007. Mirrors and divination: catoptromancy, oracles and earth goddesses in antiquity. In: M. Anderson (ed.), The book of the mirror: an interdisciplinary collection exploring the cultural story of the mirror, pp. 32-46. Newcastle: Cambridge Scholars Publishing. Aitken, G. M. and G. N. Aitken. 1990. Excavations at Whitcombe, 1965-1967. Proceedings of the Dorset Natural History and Archaeological Society 112: 57-94. Aitken, M. J. 1990. Science-based dating in archaeology. London & New York: Longman. Aldhouse-Green, M. 1996. Celtic art: reading the messages. London: The Everyman Art Library. Allen, J. R. 1904. Celtic art: in Pagan and Christian times. London: Methuen & Co. Appadurai, A. 1986a. Introduction: commodities and the politics of value. In: A. Appadurai (ed.), The social life of things: commodities in cultural perspective, pp. 3-63. Cambridge: Cambridge University Press. Appadurai, A., (ed.) 1986b. The social life of things: commodities in cultural perspective. Cambridge: Cambridge University Press. Artelius, T. 2001. The framework of pagan identity: "time" in the southern Scandinavian Late Iron Age. In: H. Karlsson (ed.), It's about time. The concept of time in archaeology, pp. 1-27. Goteborg: Bricoleur Press. Ashbee, P. 1979. The Porth Cressa cist-graves, St. Mary's, Isles of Scilly: a postscript. Cornish Archaeology 19: 61-80. Asingh, P. 2007. The bog people. In P. Asingh and N. Lynnerup, (eds), Grauballe Man: an Iron Age bog body revisited, pp. 290-323. Moesgaard: Jutland Archaeological Society. Bailey, D. W. 1993. Chronotypic tension in Bulgarian Prehistory: 6500-3500 BC. World Archaeology 25(2): 204-222. Barber, M. 2003. Bronze and the Bronze Age: metalwork and society in Britain c.2500 BC - 800 BC. Stroud: Tempus Publishing Ltd. Barrett, J. 1994. Fragments from antiquity: an archaeology of social life in Britain, 2900-1200 BC. Oxford: Blackwell Publishers. Barrett, J. C., Downes, J. MacDonald, P., Northover, J. P., O'Conner, B., Salter, C. J. and Turner, L. 2000. The metalworking evidence. In: J. Barrett, P. W. M. Freeman and A. Woodward (eds), Cadbury Castle, Somerset: the later prehistoric and early historic period, pp. 291-301. London: English Heritage Archaeological Report 20. Bate, C. S. 1865. On the discovery of a Romano-British cemetery near Plymouth. Archaeologia 40(2): 200-9. Battaglia, D. 1990. On the bones of the serpent: person, memory and mortality in Sabarl Island society. Chicago: Chicago University Press.

161

BBC. 2001. Chariot Queen. Meet the Ancestors. Great Britain, BBC: 60 mins. Bellows, J. 1881. On some bronze and other articles found near Birdlip. Transactions of the Bristol and Gloucestershire Archaeological Society 5: 137-41. Benjamin, W. 1973. The work of art in the age of mechanical reproduction. In: W. Benjamin (ed.), Illuminations, pp. 219-53: Fontana. Bevan, B. 1997. Bounding the landscape. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age societies, pp. 181-191. Oxford: Oxford Monograph 71. Bevan, B. 1999a. Land-life-death-regeneration: interpreting a middle Iron Age landscape in eastern Yorkshire. In: B. Bevan (ed.), Northern Exposure: interpretative devolution and the Iron Ages in Britain, pp. 123-148. Leicester: Leicestershire Archaeology Monographs 4. Bevan, B. 1999b. The landscape context of the Iron-Age square-barrow burials, East Yorkshire. In: J. Downes and T. Pollard (eds), The loved body's corruption: archaeological contributions to the study of human mortality. Glasgow: Cruithine Press. Bevan, B. (ed.) 1999c. Northern Exposure: interpretative devolution and the Iron Age in Britain. Leicester: Leicester Archaeology Monographs. Birchall, A. 1965. The Aylesford-Swarling Culture. Proceedings of the Prehistoric Society 31: 241367. Bishop, M. C. 1999. An Iron Age and Romano-British 'Ladder' settlement at Melton, East Yorkshire. Yorkshire Archaeological Journal 71: 23-63. Boon, G. C. 1978-80. A neglected late-Celtic mirrorhandle from Llanwnda near Fishguard. Bulletin Board of Celtic Studies 28: 743-4. Bradley, R. 1990. The passage of arms: and archaeological analysis of prehistoric hoards and votive deposits. Cambridge: Cambridge University Press. Bradley, R. 1998. The passage of arms: an archaeological analysis of prehistoric hoard and votive deposits. 2nd. Oxford: Oxbow Books. Bradley, R. 2002. The past in prehistoric societies. London & New York: Routledge. Bradley, R. 2005. Ritual and domestic life in Prehistoric Europe. London & New York: Routledge. Bradley, R. and M. Edmonds. 1993. Interpreting the Axe Trade. Cambridge: Cambridge University Press. Brailsford, J. W. 1975. The Polden Hill Hoard, Somerset. Proceedings of the Prehistoric Society 41: 22234. Brewster, T. C. M. 1971. The Garton Slack chariot burial, East Yorkshire. Antiquity 45: 289-92. Brewster, T. C. M. 1976. Garton Slack. Current Archaeology 5: 104-114. Brewster, T. C. M. 1980. The excavation of Garton and Wetwang slacks. Wintringham, Yorks: East Riding Archaeological Research Committee: Prehistoric Excavation Reports No 2.

Bridgeford, S. D. 1997. Mightier than the pen? (an edgewise look at Irish Bronze Age swords). In: J. Carman (ed.), Material Harm: archaeological studies of war and violence, pp. 95-115. Glasgow: Cruithne Press. Brown, F., C. Howard-Davis, M. Brennand, A. Boyle, T. Evans, S. O'Connor, A. Spence, R. Heawood and Lupton, A. (eds). 2007. The Archaeology of the A1 (M) Darrington to Dishforth DBFO road scheme. Lancaster: Oxford Archaeology North. Bulleid, A. and H. Gray 1911. The Glastonbury Lake Village: a full description of the excavations and the relics discovered, 1892-1907. Glastonbury: The Glastonbury Antiquarian Society. Bulleid, A. and H. Gray 1948. The Mere Lake Village: a full description of the excavations and the relics from the eastern half of the west village, 1910-1933, Vol I. Taunton: Hammett & Company Ltd. Burleigh, G. and J. V. S. Megaw. 2007. A Late Iron Age Mirror Burial at Pegsdon, Shillington, Bedfordshire: an interim account. Antiquaries Journal 87: 109-40. Burns, J. E. 1971. Additional torcs from Snettisham, Norfolk. Proceedings of the Prehistoric Society 37(1): 228-9. Butler, J. 1990. Gender trouble: feminism and subversion of identity. London & New York: Routledge. Butler, J. 1993. Bodies that matter: on the discursive limits of 'sex'. London & New York: Routledge. Campbell, M. 2006. Memory and monumentality in the Rorotongan landscape. Antiquity 80: 102-117. Campbell, S. F. 2001. The captivating agency of art: many ways of seeing. In: C. Pinney and N. Thomas (eds), Beyond aesthetics: art and technologies of enchantment, pp. 117-36. Oxford: Berg. Campbell, S. F. 2002. The art of Kula. Oxford: Berg. Card, N., J. Downes, Gibson, J. and Sharman, P. 2005. Religion and metal-working at Mine Howe, Orkney. Current Archaeology 199: 322-7. Carpino, A. A. 2003. Discs of splendour: the relief mirrors of the Etruscans. Madison: The University of Wisconsin Press. Carr, G. 2000. 'Romanisation' and the body. In: G. Davies, A. Gardner & K. Lockyear (eds), TRAC 2000. Proceedings of the Tenth Annual Theoretical Roman Archaeology Conference, London 2000, pp. 112-24. Oxford: Oxbow Books. Carr, G. 2005. Woad, tattooing and identity in later Iron Age and early Roman Britain. Oxford Journal of Archaeology 24(3): 273-92. Carr, G. and C. Knüsel 1997. The ritual framework of the early and middle Iron Ages of central southern Britain. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age societies, pp. 167-173. Oxford: Oxbow Monographs 71. Carter, C. and F. Hunter. 2001. The Newbridge chariot. Current Archaeology 178: 413-5. Cellini, B. 1926. The life of Benvenuto Cellini: a Florentine artist. London: Dent. Cessford, C. 1997. The Pictish mirror symbol and archaeological evidence for mirrors in Scotland. Oxford Journal of Archaeology 16(1): 99-119.

162

Champion, S. 1976. Coral in Europe: commerce and Celtic ornament. In: P. M. Duval and C. F. C. Hawkes (eds), Celtic art in ancient Europe: five protohistoric centuries, IVe et IIIe siècles avant notre ère. Paris: Librairie Droz. Champion, T. C. 1987. The European Iron Age: assessing the state of the art. Scottish Archaeological Review 4(2): 98-107. Chapman, M. 1992. The Celts. New York: St. Martin's Press. Childe, V. G. 1930. The Bronze Age. Cambridge: Cambridge University Press. Childs, S. T. and D. J. Killick 1993. Indigenous African metallurgy: nature and culture. Annual Review of Anthropology 22: 317-37. Clark, P. J. 1968-72. The Neolithic, Bronze and Iron Age, and Romano-British finds from Mount Batten, Plymouth 1832-1939. Proceedings of the Devon Archaeological Exploration Society 26-30: 137-61. Clarke, D. L. 1972. A provisional model of an Iron Age society and its settlement system. In: D. L. Clarke (ed.), Models in archaeology, pp. 80169. London: Methuen & Co Ltd. Cleere, H. F. 1972. The classification of early iron-smelting furnaces. The Antiquaries Journal 52: 8-23. Cleere, H., F. 1977. Comments on the iron-working activities. In: R. Hanworth and D. J. Tomalin (eds), Brooklands, Weybridge: the excavation of an Iron Age and Medieval site, 1964-5 and 197071, pp. 19-22. Guildford: Surrey Archaeological Society. Cleere, H. 1978. The slag and crucible fragments. In: M. Parrington (ed.), The excavation of an Iron Age settlement, Bronze Age ring ditches and Roman features at Ashville Trading Estate Abingdon (Oxfordshire) 1974-6, pp. 88-90. London: Oxford Archaeological Unit and the Council for British Archaeology. Coles, J. 1987. Meare Village East: the excavations of A. Bulleid and H. St. George Gray, 1932-1956. Hertford: Somerset Levels Papers Number 13. Coles, J. and S. Minnitt 1995. Industrious and fairly civilised. The Glastonbury lake village. Exeter & Taunton: Somerset levels project & Somerset county council museums service. Collard, M., Darvill, T. and Watts, M. 2006. Ironworking in the Bronze Age? Evidence from a 10th century BC settlement at Hartshill Copse, Upper Bucklebury, West Berkshire. Proceedings of the Prehistoric Society 72: 367-421. Collis, J. 1996. Celts and politics. In: J. Graves-Brown, Jones, S. and Gamble, C. (eds), Cultural identity and archaeology: the construction of European identities, pp. 167-78. London and New York: Routledge. Collis, J. 1997. Celtic myths. Antiquity 71: 195-201. Collis, J. 2003. The Celts, origins, myths and inventions. Stroud: Tempus Publishing Ltd. Connerton, P. 1989. How societies remember. Cambridge: Cambridge University Press.

Cool, H. E. M. 2006. Eating and drinking on Roman Britain. Cambridge: Cambridge University Press. Coote, J. 1992. Marvels of everyday vision: the anthropology of aesthetics and the cattle-keeping Nilotes. In: J. Coote and A. Shelton (eds), Anthropology, art and aesthetics, pp. 245-73. Oxford: Clarendon Press. Corcoran, J. X. W. P. 1952. Tankards and tankard handles of the British early Iron Age. Proceedings of the Prehistoric Society 18: 85-102. Corney, M. 2000. The brooches. In: M. Fulford and J. Timby (eds), Late Iron Age and Roman Silchester. Excavations on the site of the Forum-basilica, 1977 1980–86. London: Britannia Monograph Series 15. Creighton, J. 2000. Coins and power in Late Iron Age Britain. Cambridge: Cambridge University Press. Creighton, J. 2006. Britannia: the creation of a Roman province. London: Routledge. Cremo, A. 1999. Puranic time and the archaeological record. In: T. Murray (ed.), Time and archaeology, pp. 38-48. London: Routledge. Crew, P. 1987. Bryn y Castell hillfort - a late prehistoric ironworking settlement in northwest Wales. In: B. G. Scott and H. Cleere (eds), The crafts of the blacksmith: essays presented to R.F. Tylecote at the 1984 symposium of the UISPP Comité pour la sidérurgie ancienne, Belfast, 16th - 21 September 1984, pp. 91-100. Belfast: UISPP Comité pour la Sidérurgie Ancienne and the Ulster Museum. Crew, P. 1989. Excavations at Crawcwellt West, Merioneth 1986-1989: a late prehistoric upland ironworking settlement. Archaeology in Wales 29: 11-16. Crew, P. 1991. Late Iron Age and Roman iron production in Northwest Wales. In: B. Burnham and J. Davies (eds), pp. 150-60. Lampeter: St. David's University College. Cripps, L. 2004. Studying settlement and society: the Iron Age of Cornwall and Devon. Cripps, L. 2007. Re-situating the Later Iron Age in Cornwall and Devon: new perspectives from the settlement record. In: C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and beyond, pp. 140-155. Oxford: Oxbow Books. Cunliffe, B. 1972. The late Iron Age metalwork from Bulbury, Dorset. The Antiquaries Journal 52: 293308. Cunliffe, B. 1984. Danebury an Iron Age hillfort in Hampshire. Volume 2 the excavations, 1969-1978: the finds. London: Council for British Archaeology Research Report 52. Cunliffe, B. 1988. Mount Batten Plymouth: a prehistoric and Roman fort. Oxford: Oxford University Committee for Archaeology Monograph No 26. Cunliffe, B. 1991. Iron Age Communities in Britain: an account of England, Scotland, and Wales from the seventh century BC until the Roman conquest. 3rd. London: Routledge and Kegan Paul. Cunliffe, B. 1995. Iron Age Britain. London: B. T. Batsford Ltd/English Heritage. Cunliffe, B. 2004. Iron Age Britain. 2nd. London: B. T. Batsford.

163

Cunnington, E. 1884. On a hoard of bronze, Iron, and other objects found in Belbury camp. Archaeologia 48: 115-20. Damm, C. 2001. Another time, another history? Or how to write the history of the Bugakhwe in Botswana. In: H. Karlsson (ed.), It's about time. The concept of time in archaelogy, pp. 29-43. Goteborg: Bricoleur Press. David, N., Heimann, R., Killick, D. J. and Wayman, M. 1989. Between bloomery and blast furnace: Mafa iron-smelting technology in North Cameroon. African Archaeological Review 7: 183208. Davies, J. L. and M. Spratling. 1976. The Severn Sisters hoard: a centenary study. In: G. C. Boon and J. M. Lewis (eds), Welsh Antiquity. Cardiff. Davies, M. and A. Gwilt. 2008. Material, style and identity in first century metalwork, with particular reference to the Severn Sisters Hoard. In: D. Garrow, C. Gosden and J. D. Hill (eds), Rethinking Celtic Art, pp. 146-84. Oxford: Oxbow Books. Déchelette, J. 1913. La collection millon: antiquités préhistoriques et Gallo-Romaines Paris: Librairie Paul Geuthner. De Marrais, E., Gosden, C. and Renfrew, C. (eds). 2004. Rethinking materiality: the engagement of mind with the material world. Cambridge: MacDonald Institute for Archaeological Research. De Navarro, J. M. 1952. The Celts in Britain and their art. In: M. D. Knowles (ed.), The Heritage of early Britain, pp. 56-82. London: G. Bell & Sons Ltd. Dent, J. 1985a. Three cart burials from Wetwang, Yorkshire. Antiquity 59: 85-92. Dent, J. 1985b. Wetwang: a third chariot. Current Archaeology 95: 360-1. Dent, J. S. 1998. The Yorkshire Wolds in late prehistory and the emergence of an Iron Age society. In: P. Halkon (ed.), Further light on the Parisi: recent research in Iron Age and Roman East Yorkshire, pp. 4-13. Hull: East Riding Archaeological Research Trust. Diaz-Andreu, M. 2005. Gender identity. In: M. DiazAndreu, S. Lucy, S. Babic and D. N. Edwards (eds), The archaeology of identity: approaches to gender, age, status, ethnicity and religion, pp. 13-42. London: Routledge. Dietler, M. 1994. "Our ancestors the Gauls": archaeology, ethnic nationalism, and the manipulation of Celtic identity in modern Europe. American Anthropologist 96(3): 584-605. Dobres, M. A. 1995. Gender and prehistoric technology: on the social agency of technical strategies. World Archaeology 27(1): 25-49. Dobres, M. A. 1999. Technology's links and chaines: the processual unfolding of technique and technician. In: M. A. Dobres and C. R. Hoffman (eds), The social dynamics of technology: practice, politics and world views, pp. 124-46. Washington: Smithsonian Institution Press.

Dobres, M. A. 2000. Technology and Social Agency: Outlining a Practice Framework for Archaeology. Oxford: Blackwell. Dobres, M. A. and C. R. Hoffman 1994. Social agency and the dynamics of prehistoric technology. Journal of Archaeological Method and Theory 1(3): 211-58. Dobres, M. A. and J. E. Robb. 2000. (eds). Agency in archaeology. London & New York: Routledge. Duff, W. 1975. Images stone B.C.: thirty centuries of Northwest coast Indian sculpture. Seattle: University of Washington Press. Duff, W. 1983. The world is as sharp as a knife: meaning in northern northwest coast art. In: R. L. Carlson (ed.), Indian art traditions of the northwest coast, pp. 47-66. Burnaby: Archaeology Press. Dungworth, D. 1996. The production of copper alloys in Iron Age Britain. Proceedings of the Prehistoric Society 62: 399-42. Dungworth, D. 1997a. Copper metallurgy in Iron Age Britain: some recent research. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age Societies, pp. 46-50. Oxford: Oxbow Monograph 71. Dungworth, D. 1997b. Iron Age and Roman copper alloys from northern Britain. Internet Archaeology 2. Dunning, G. C. 1928. An engraved bronze mirror from Nijmegen, Holland; with a note on the origin and distribution of the type. Archaeological Journal 85: 69-79. Dyer, J. 1966. A second Iron-Age mirror-handle from Old Warden. Bedfordshire Archaeological Journal 3: 55-56. Echt, R. 1999. Das Fürstinnengrab von Reinheim. Studienzurkulturgeschichte der Früh-Tène-Zect. Bonn: Dr. Rudolf Habelt GMBH. Eckardt, H. and H. Williams 2003. Objects without a past? The use of Roman objects in early Anglo-Saxon graves. In: H. Williams (ed.), Archaeologies of remembrance, pp. 141-70: Plenum Publishers. Ehrenreich, R. M. 1985. Trade, technology and the ironworking community in the Iron Age of southern Britain. Oxford: BAR British Series. Evans, C. 1989. Perishables and worldly goods – artefact decoration and classification in the light of wetlands research. Oxford Journal of Archaeology 8(2): 179-201. Farley, M. 1983. A Mirror Burial at Dorton, Buckinghamshire. Proceedings of the Prehistoric Society 49: 269-302. Farrar, R. A. H. 1956. A Celtic burial with mirror-handle at West Bay, near Bridport. Proceedings of the Dorset Natural History and Archaeological Society 76: 90-4. Fell, V. 1988. Iron Age metalworking tools from Gussage All Saints, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 110: 7376. Field, N. and Parker-Pearson, M. 2003. Fiskerton an Iron Age timber causeway with Iron Age and Roman votive offerings: the 1981 excavations. Oxford: Oxbow Books. Finlay, I. 1973. Celtic art: an introduction. London: Faber & Faber.

164

Fitzpatrick, A. P. 1984. The deposition of La Tène Iron Age metalwork in watery contexts in southern England. In: B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in central southern Britain, pp. 178-90. Oxford: University of Oxford Committee for Archaeology. Fitzpatrick, A. P. 1992. The Snettisham, Norfolk, hoards of Iron Age Torcs: sacred or profane? Antiquity 66: 243-6. Fitzpatrick, A. P. 1996. "Celtic" Iron Age Europe: the theoretical basis. In: J. Graves-Brown, S. Jones and C. Gamble (eds), Cultural Identity and archaeology: the construction of European identities, pp. 238-55. London and New York: Routledge. Fitzpatrick, A. P. 1997a. A 1st-century AD 'Durotrigian' inhumation burial with a decorated Iron Age mirror from Portesham, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 118: 51-70. Fitzpatrick, A. P. 1997b. Archaeological excavations on the route of the A27 Westhampnett bypass, West Sussex, 1992. Volume 2: the cemeteries. Salisbury: Wessex Archaeology Report No. 12. Fitzpatrick, A. P. 2007. Dancing with dragons: fantastic animals in the earlier Celtic art of Iron Age Britain. In: C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond, pp. 339-57. Oxford: Oxbow Books. Fitzpatrick, A. P. and P. J. Woodward 1995. Fieldwork and excavation at Portesham. Proceedings of the Dorset Natural History and Archaeological Society 117: 133. Fontijn, D. R. 2002. Sacrificial landscapes: cultural biographies of persons, objects and 'natural' places in the Bronze Age of the southern Netherlands, c. 2300-600BC. Analecta Prahistorical Leidensia 33/34. Foster, J. 1986. The Lexden tumulus: a reappraisal of an Iron Age burial from Colchester, Essex. Oxford: British Archaeological Report. Foster, J. 1995. Metalworking in the British Iron Age: The evidence from Wheelsby Avenue, Grimsby. In: B. Raftery (ed.), Sites and sights of the Iron Age, pp. 49-60. Oxford: Oxbow Monograph 56. Foucault 1990. The history of sexuality. Volume 3: the care of the self. London: Penguin. Fox, A. 1954. Celtic fields and farms on Dartmoor, in the light of recent excavations at Kestor. Proceedings of the Prehistoric Society 20: 87-102. Fox, A. 1972. The Holcombe Mirror. Antiquity 46: 29396. Fox, A. and S. Pollard 1973. A decorated bronze mirror from an Iron Age settlement at Holcombe, near Uplyme, Devon. Antiquaries Journal LIII: 2141. Fox, C. 1923. The archaeology of the Cambridge region. Cambridge: Cambridge University Press. Fox, C. 1925. A late Celtic bronze mirror from Wales. Antiquaries Journal 5: 254-57.

Fox, C. 1946. A find of the early Iron Age from Llyn Cerig Bach, Anglesey. Cardiff: The National Museum of Wales. Fox, C. 1949. Celtic mirror handles in Britain. Archaeologia Cambensis 100: 24-44. Fox, C. 1952. Triskeles, palmettes and horse-brooches. Proceedings of the Prehistoric Society 18: 47-54. Fox, C. 1958. Pattern and Purpose: a survey of early Celtic art in Britain. Cardiff: The National Museum of Wales. Fox, C. 1960. A Celtic Mirror from Great Chesterford. Antiquity XXXIV: 207-10. Fox, C. and M. R. Hull 1948. The incised ornament on the Celtic mirror from Colchester, Essex. The Antiquaries Journal 28: 123-37. Franks, A. W. 1865. Note on the mirrors, &c. discovered in a cemetery near Plymouth. Archaeologia 40(2): 510. Fulford, M. and J. Creighton 1998. A Late Iron Age mirror burial from Latchmere Green, near Silchester, Hampshire. Proceedings of the Prehistoric Society 64: 331-42. Fulford, M. and J. Timby 2000. Late Iron Age and Roman Silchester. Excavations on the site of a Forumbasilica, 1977 1980-6. London: Britannia Monograph Series. Gallardo, F., V. Castro and Miranda, P. 1999. Riders on the storm: rock art in the Atacama Desert (northern Chile). World Archaeology 31(2): 225-42. Gamble, C. 2001. Archaeology: the basics. London & New York: Routledge. Gamble, C. 2007. Origins and revolutions: human identity in earliest prehistory. Cambridge: Cambridge University Press. Garrow, D., C. Gosden and J. D. Hill (eds). 2008. Rethinking Celtic Art. Oxford: Oxbow Books. Garrow, D. and C. Gosden. in press [due 2010]. Dating Celtic Art: a major radiocarbon dating programme of Iron Age and Early Roman metalwork in Britain. Archaeological Journal 166. Geary, P. 1986. Sacred commodities: the circulation of medieval relics. In: A. Appadurai (ed.), The social life of things: commodities in cultural perspective, pp. 169-91. Cambridge: Cambridge University Press. Gell, A. 1992. The anthropology of time: cultural constructions of temporal maps and images. Oxford: Berg. Gell, A. 1998. Art and agency: an anthropological theory. Oxford: Clarendon Press. Gell, A. 1999. The technology of enchantment and the enchantment of technology. In: E. Hirsch (ed.), The art of anthropology: essays and diagrams, pp. 159-86. London: The Athlone Press. Gero, J. M. and M. W. Conkey, (eds). 1991. Engendering archaeology: women and prehistory. Oxford: Blackwell Publishers Ltd. Gerritsen, F. 1999. To build and to abandon: the cultural biography of late prehistoric houses and farmsteads in the southern Netherlands. Archaeological Dialogues 6(2): 79-97.

165

Giles, M. 2007. Good fences make good neighbours? Exploring the ladder enclosures of Late Iron Age East Yorkshire. In: C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and beyond, pp. 235-49. Oxford: Oxbow Books. Giles, M. 2008. Seeing red: the aesthetics of martial objects in the British and Irish Iron Age. In D. Garrow, D, C. Gosden and J. D. Hill (eds) Rethinking Celtic Art, pp. 59-77. Oxford: Oxbow Books. Giles, M. and J. P. Joy 2007. Mirrors in the British Iron Age. In: M. Anderson (ed.), The book of the mirror: an interdisciplinary collection exploring the cultural story of the mirror, pp. 16-31. Newcastle: Cambridge Scholars Publishing. Gillies, W. 1981. The craftsman in early Celtic literature. Scottish Archaeological Forum 11: 70-85. Gillings, M. and J. Pollard 1999. Non-portable stone artefacts and contexts of meaning: the tale of grey whether (www.museums.ncl.ac.uk/ Avebury/stone4.htm). World Archaeology 31(2): 179-93. Gosden, C. 1994. Social being and time. Oxford: Blackwell Publishers. Gosden, C. 2001a. Making sense: archaeology and aesthetics. World Archaeology 33(2): 163-7. Gosden, C. (ed.) 2001. Archaeology and aesthetics. World Archaeology. London & New York: Routledge. Gosden, C. and J. D. Hill. 2008. Introduction: re-integrating ‘Celtic’ Art. In: D. Garrow, C. Gosden and J. D. Hill (eds), Rethinking Celtic Art, pp. 1-14. Oxford: Oxbow Books. Gosden, C. and Y. Marshall 1999. The cultural biography of objects. World Archaeology 31(2): 169-78. Green, C. 1949. The Birdlip early Iron Age burials: a review. Proceedings of the Prehistoric Society 15: 188-90. Green, F. J. 1994. Early Iron Age stream deposits at La Sagesse, Romsey, Hampshire. In: A. P. Fitzpatrick and E. L. Morris (eds), The Iron Age in Wessex: recent work, pp. 49-52. Salisbury: Association Francaise d'Etude de l'Ange du fer/ Trust for Wessex Archaeology. Green, M. J. 1998. Vessels of death: sacred cauldrons in archaeology and myth. The Antiquaries Journal 78: 63-84. Green, S. 1991. Metalwork from Llyn Cerrig Bach. In: V. Kruta, O. H. Frey, B. Raferty and M. Szabo (eds), The Celts, pp. 609. London: Thames & Hudson. Greenwell, W. 1906. Early Iron Age burials in Yorkshire. Archaeologia LX: 251-324. Gregory, C. 1982. Gifts and commodities. London: Academic Press. Grimes, W. F. 1952. The La Tène art style in British early Iron Age pottery. Proceedings of the Prehistoric Society 18: 160-75. Guillaumet, J. P. and M. Schönfelder 2001. Un manche de miroir de type britannique provenant de Compiègne; un nouveau témoignage des cont

acts à travers la Manche. Antiquités Nationales 33: 125-8. Halkon, P. 2008. Archaeology and environment in a changing East Yorkshire landscape: the Foulness Valley c. 800 BC to c. AD 400. Oxford: BAR British Series 472. Halkon, P. and M. Millett, (eds). 1999. Rural settlement and industry: studies in the Iron Age and Roman archaeology of lowland East Yorkshire. Leeds: Yorkshire Archaeological Report No 4. Hallen, Y. 1994. The use of bone and antler at Foshigarry and Bac Mhic Connain, two Iron Age sites on North Uist, Western Isles. Proceedings of the Society of Antiquaries Scotland 124: 189-231. Hamilakis, Y. 1999. Stories from exile: fragments from the cultural biography of the Parthenon (or 'Elgin') marbles. World Archaeology 31(2): 303-20. Hamilakis, Y. Pluciennik, M. and Tarlow, S. (eds). 2002. Thinking through the body: archaeologies of corporeality. London: Kluwer Academic/Plenum Publishers. Haselgrove, C. 1989. The later Iron Age in southern Britain and beyond. In: M. Todd (ed.), Research on Roman Britain 1960-89, pp. 1-18. London: Britannia Monograph. 11. Haselgrove, C. 1997. Iron Age brooch deposition and chronology. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age societies, pp. 51-72. Oxford: Oxbow Monographs. Haselgrove, C., Armit, I., Champion, T. C., Creighton, J., Gwilt, A., Hill, J. D., Hunter, F. & Woodward, A. 2002. Understanding the British Iron Age: an agenda for action. Salisbury: Trust for Wessex archaeology Ltd. Hawkes, C. F. C. 1968. New thoughts on the Belgae. Antiquity 42: 6-16. Hawkes, C. F. C. and M. R. Hull 1947. Camulodunum: first report on the excavations at Colchester 19301939. Oxford: Charles Batey for the Society of Antiquaries London. Hegardt, J. 2004. The archaeology of the voice. In: H. Karlsson (ed.), Swedish Archaeologists on Ethics, pp. 63-70. Lindome: Bricoleur. Heidegger, M. 1971. The thing. In: T. A. Hofstadter (ed.), Poetry, language, thought, pp. 165-80. New York: Harper & Row. Hendrickson, C. 1995. Weaving identities: construction of dress and self in a highland Guatemala town. Austin: University of Texas Press. Herbert, E. W. 1984. Red gold of Africa: copper in precolonial history and culture. Madison: University of Wisconsin Press. Herbert, E. W. 1993. Iron, gender and power: rituals of transformation in African societies. Bloomington: Indiana University Press. Hill, J. D. 1989. Re-thinking the Iron Age. Scottish Archaeological Review 6: 16-24. Hill, J. D. 1992. Can we recognise a different European past? A contrastive archaeology of later prehistoric settlements in southern England. Journal of European Archaeology 1: 57-75.

166

Hill, J. D. 1995a. The Pre-Roman Iron Age in Britain and Ireland (ca. 800 BC to AD 100). Journal of World Prehistory 9(1): 47-98. The best summary of the Iron Age I have found to date. Hill, J. D. 1995b. Ritual and rubbish in the Iron Age of Wessex: a study on the formation of a specific archaeological record. Oxford: BAR British Series. Hill, J. D. 1997. 'The end of one kind of body and the beginning of another kind of body'? Toilet instruments and 'Romanization' in southern England during the first century AD. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age Societies, pp. 96-107. Oxford: Oxbow Monograph 71. Hill, J. D. 2000. Shillington, Bedfordshire: Iron Age silver brooch, bronze mirror and pottery fragments (P&EE 79). In, Treasure Annual Report 2000, pp. 15-16. London: Department of Culture, Media and Sport. Hill, J. D. 2001a. A new cart/chariot burial from Wetwang, East Yorkshire. Past 38: 2-3. Hill, J. D. 2001b. Romanisation, gender and class: recent approaches to identity in Britain and their possible consequences. In: S. James and M. Millett (eds), Britons and Romans: advancing an archaeological agenda, pp. 12-18. York: Council for British Archaeology Research Reports 125. Hill, J. D. 2001c. Tour - The Wetwang Chariot Burial. www.thebritishmuseum.ac.uk/compass. Hill, J. D. 2002a. Wetwang chariot burial. Current Archaeology 15 (178): 410-12. Hill, J. D. 2002b. Pottery and the expression of society, economy and culture. In: A. Woodward and J. D. Hill (eds), Prehistoric Britain: the ceramic basis, pp. 75-84. Oxford: Oxbow Books. Hill, J. D. 2006. Are we any closer to understanding how later Iron Age societies worked (or did not work)? In: C. Haselgrove (ed.), Celtes et Gaulois, l'archéologie les mutations de la fin de l'âge du Fer, Actes de la table ronde de Cambridge, 7-8 juillet 2005, pp. 169-79. Glux-enGlenne: Bibracte, Centre archéologique européen 12/4. Hill, J. D. 2007. The dynamics of social change in Later Iron Age eastern and south-eastern England c. 300 - AD 43. In: C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and beyond, pp. 16-40. Oxford: Oxbow Books. Hill, J. D., Evans, C. and Alexander, M. 1999. The Hinxton Rings - a Late Iron Age cemetery at Hinxton, Cambridgeshire, with a reconsideration of northern Aylesford-Swarling distributions. Proceedings of the Prehistoric Society 65: 243-73. Hill, J. D., Spence A. J., La Niece, S. & Worrell, S. 2004. The Winchester hoard: a find of unique Iron Age gold jewellery from southern England. The Antiquaries Journal 84: 1-22. Hingley, R. 1984. Towards social analysis in archaeology: Celtic society in the Iron Age of the upper Thames valley (400-0 BC). In: B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in central

southern Britain, pp. 12-45. Oxford: University of Oxford, committee for archaeology Monograph No. 2. Hingley, R. 1990. Iron Age 'currency bars': the archaeological and social context. Archaeological Journal 147: 91-117. Hingley, R. 1997. Iron, iron working and regeneration: a study of the symbolic meaning of metalworking in Iron Age Britain. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age Societies, pp. 918. Oxford: Oxbow Monographs 71. Hingley, R. 2005. Iron Age 'currency bars' in Britain: items of exchange in liminal contexts? In: C. Haselgrove and D. Wigg-Wolf (eds), Iron Age coinage and ritual practice, pp. 183-205. Mainz am Rhein: Phillipp von Zaberg. Studien zu Fundmunzen der Antike. Hingley, R. 2006. The deposition of iron objects in Britain during the later Prehistoric and Roman periods: contextual analysis and the significance of iron. Britannia XXXVII: 213-57. Hingley, 2009. Esoteric knowledge? Ancient bronze artefacts from Iron Age contexts. Proceedings of the Prehistoric Society 75: 143-66. Hoffman, C. R. and M. A. Dobres 1999. Conclusion: making material culture, making cultural material. In: M. A. Dobres and C. R. Hoffman (eds), The social dynamics of technology: practice, politics and world views, pp. 209-22. Washington: Smithsonian Institution Press. Holtorf, C. J. 1998. The life-histories of megaliths in Mecklenburg-Vorpommern (Germany). World Archaeology 30(1): 23-38. Holtorf, C. J. 2000-1. Monumental past: the life-histories of megalithic monuments in Mecklenburg-Vorpommern (Germany). from http:// citdpress.utsc.utoronto.ca/holtorf/. Holtorf, C. J. 2002a. Excavations at Monte da Igreja near Evora (Portugal). From the life-history of a monument to re-uses of ancient objects. Journal of Iberian Archaeology 4: 177-201. Holtorf, C. J. 2002b. Notes on the life history of a pot sherd. Journal of Material Culture 7(1): 49-71. Hoskins, J. 1985. A life history from both sides: the changing poetics of personal experience. Journal of Anthropological Research 41(2): 31-51. Hoskins, J. 1989. On loosing and getting a head: warfare, alliance and exchange in a changing Sumba 18881988. American Ethnologist 16(3): 419-40. Hoskins, J. 1998. Biographical objects: how things tell the stories of people's lives. London & New York: Routledge. Hosler, D. 1988. Ancient west Mexican metallurgy: south and Central American origins and west Mexican transformations. American Anthropologist 90(4): 832-55. Hosler, D. 1994. The sounds and colours of power: the sacred metallurgical technology of ancient west Mexico. Cambridge, MA: MIT Press. Hosler, D. 1995. Sound, colour and meaning in the metallurgy of ancient west Mexico. World Archaeology 27: 100-15.

167

Hunter, F. 1997. Iron Age hoarding in Scotland and northern England. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age Societies, pp. 108-34. Oxford: Oxbow Monograph 71. Hunter, F. 2001. The Carnyx in Iron Age Europe. The Antiquaries Journal 81: 77-108. Hunter, F. 2008. Celtic art in Roman Britain. In D. Garrow, C. Gosden and J. D. Hill (eds) Rethinking Celtic Art, pp. 129-45. Oxford: Oxbow Books. Hutcheson, N. 2003. Material culture in the landscape: a new approach to the Snettisham hoards. In: J. Humphrey (ed.), Re-searching the Iron Age, pp. 87-97. Leicester: Leicester Archaeological Monographs. Hutcheson, N. 2004. Later Iron Age Norfolk: metalwork, landscape and society. Oxford: BAR British Series 361. Hutton, R. 1991. The pagan religions of the ancient British Isles: their nature and legacy. Oxford: Blackwell. Ilkjaer, J. and J. Lonstrup 1982. Interpretation of the great votive deposits of Iron Age weapons. Danish Archaeology 1: 95-103. Immonen, V. 2002. Functional ladles or ceremonial cutlery? A cultural biography of prehistoric wooden spoons from Finland. Acta Borealia 1: 27-47. Ingold, T. 1993. The Reindeerman's lasso. In: P. Lemonnier (ed.), Technological choices: transformation in material cultures since the Neolithic, pp. 108-25. London Routledge. Ingold, T. 1999. Foreword. In: M. A. Dobres and C. R. Hoffman (eds), The social dynamics of technology: practice, politics and world views, pp. viixi. Washington: Smithsonian Institution Press. Issings, C. 1957. Roman glass from dated finds. Groningen: Archaeologica Traiectina. Izzet, V. 1998. Holding a mirror to Etruscan gender. In: R. D. Whitehouse (ed.), Gender and Italian archaeology: challenging the stereotypes, pp. 209-27. London: Accordia specialist studies on Italy volume 7. Jackson, D. A., Pacitto, A. L., Tylecote, R. F. and Ambrose, T. M. 1978. The iron-working features. In: J. Jackson and T. M. Ambrose (eds), Excavations at Wakerley, pp. 151-66. Britannia 9. Jackson, R. 1985. Cosmetic sets from late Iron Age and Roman Britain. Britannia 16: 165-92. Jackson, R. 1990. Camerton: the Late Iron Age and Early Roman Metalwork. London: British Museum Press. Jacobsthal, P. 1944. Early Celtic Art. Oxford: Clarendon. James, S. 1993. Exploring the world of the Celts. London: Thames & Hudson. James, S. 1999. The Atlantic Celts: ancient people or modern invention? London: British Museum Press. Janowitz, N. 2001. Magic in the Roman world: Pagans, Jews and Christians. London & New York: Routledge.

Jobey, G. 1962. An Iron Age homestead at West Brandon, Durham. Archaeologia Aeliana 40: 1-34. Johns, C. 2006. Iron Age sword and mirror cist burial from Bryher, Isles of Scilly. Cornish Archaeology 41-2: 1-80. Johns, C. No Date. Iron Age burial with sword and mirror from Bryher. St. Mary's: Isles of Scilly Museum Publication. Jones, A. 2001. Drawn from memory: the archaeology of aesthetics and the aesthetics of archaeology in Earlier Bronze Age Britain and the present. World Archaeology 33(2): 334-56. Jones, A. 2002a. Archaeological theory and scientific practice. Cambridge: Cambridge University Press. Jones, A. 2002b. A biography of colour: colour, material histories and personhood in the early Bronze Age of Britain and Ireland. In: A. Jones and G. MacGregor (eds), Colouring the past: the significance of colour in archaeological research, pp. 159-74. Oxford: Berg. Jones, A. and G. MacGregor, (eds). 2002a. Colouring the past: the significance of colour in archaeological research. Oxford: Berg. Jones, A. and G. MacGregor. 2002b. Introduction: wonderful things - colour studies in archaeology from Munsell to materiality. In: A. Jones and G. MacGregor (eds), Colouring the past: the significance of colour in archaeological research, pp. 1-21. Oxford: Berg. Jones, M. U. and W. T. Jones. 1975. The crop-mark sites at Mucking, Essex, England. In: R. Bruce-Mitford (ed.), Recent archaeological excavations in Europe, pp. 131-84. London: Routledge & Kegan Paul. Jope-Rogers, J. 1873. Romano-British, or late Celtic, remains at Trelan Bahaw, St. Keverne, Cornwall. Antiquaries Journal 30: 267-73. Jope, E. M. 1954. The Keshcarrigan bowl and a bronze mirror-handle from Ballymoney. Ulster Journal of Archaeology 17: 92-6. Jope, E. M. 1971. The Waldalgesheim Master. In: J. Boardman, M. A. Brown and T. G. E. Powell (eds), The European community in later prehistory. Studies in honour of C. F. C. Hawkes, pp. 167-80. London: Routledge & Kegan Paul. Jope, E. M. 2000. Early Celtic art in the British Isles. Oxford: Clarendon Press. Joy, J. P. 2002. Biography of a medal: people and the things they value. In: J. Schofield, W. G. Johnson and C. M. Beck (eds), Materiel Culture: the archaeology of twentieth-century conflict, pp. 132-42. London & New York: Routledge. Joy, J. 2008a. Reflections on the Iron Age: biographies of Iron Age mirrors. Unpublished PhD thesis, University of Southampton. Joy, J. 2008b. Reflections on Celtic Art: a re-examination of mirror decoration. In D. Garrow, C. Gosden and J. D. Hill (eds), Rethinking Celtic Art, pp. 78-99. Oxford: Oxbow Books. Joy, J. 2009a. Reinvigorating object biography: reproducing the drama of object lives. World Archaeology 41(4): 566–582.

168

Joy, J. 2009b. Lindow Man. London: British Museum Press. Joy, J. in press. Exploring status and identity in later Iron Age Britain: re-interpreting mirror burials. In: L. Amada and T. Moore (eds), Western Europe in the first millennium BC: crossing the divide. Oxford: Oxford Archaeological Press. Joyce, R. A. 2003. Concrete memories: fragments of the past in the Classic Maya present (500-1000 AD). In: R. M. Van Dyke and S. E. Alcock (eds), Archaeologies of memory, pp. 104-25. Oxford: Blackwell. Jundi, S. and J. D. Hill 1997. Brooches and identities in First Century AD Britain: more than meets the eye? In: C. Forcey, J. Hawthorne and R. Witchen (eds), TRAC 97. Proceedings of the seventh annual Theoretical Roman Archaeology Conference, Nottingham 1997, pp. 125-37. Oxford: Oxbow Books. Karl, R. 2008. Random coincidences? Or: The return of the Celtic to Iron Age Britain. Proceedings of the Prehistoric Society 74: 69-78. Karlsson, H. 2001a. Preface and introduction. In: H. Karlsson (ed.), It's about time. The concept of time in archaeology, pp. ix-x. Goteborg: Bricoleur Press. Karlsson, H. 2001b. Time for an archaeological 'Timeout'? In: H. Karlsson (ed.), It's about time. The concept of time in archaeology, pp. 45-59. Goteborg: Bricoleur Press. Keller, J. 1965. Das Keltische Fürstengrab von Reinheim I. Ausgrabungsbericht und catalog der Funde bearbeitet von Joseph Keller, Verlag des Römisch. Bonn: Germanisten zentralmuseums Mainz. Keller, C. M. and J. D. Keller 1996. Cognition and tool use: the blacksmith at work. Cambridge: Cambridge University Press. Kense, F. J. and J. A. Okoro 1993. Changing perspective on traditional iron production in West Africa. In: T. Shaw, P. Sinclair, B. Andah and A. Okpoko (eds), The archaeology of Africa: food, metals and towns, pp. 449-58. London & New York: Routledge. Knappett, C. 2004. The affordances of things: a postGibsonian perspective on the relationality of mind and matter. In: E. DeMarrais, C, Gosden and C, Renfrew (eds), Rethinking materiality: the engagement of mind with the material world, 43-51. Cambridge: MacDonald Institute for Archaeological Research. Knappett, C. 2005. Thinking through material culture: an interdisciplinary perspective. Philadelphia: University of Pennsylvania Press. Kopytoff, I. 1986. The cultural biography of things: commoditization as process. In: A. Appadurai (ed.), The social life of things: commodities in cultural perspective, pp. 64-91. Cambridge: Cambridge University Press. Lackoff, G. and M. Johnson 2003 (1980). Metaphors we live by. Chicago: University of Chicago Press.

Lahiri, N. 1995. Indian metal and metal-related artefacts as cultural signifiers: an ethnographic perspective. World Archaeology 27(1): 116-32. Laing, L. and J. Laing 1992. Art of the Celts. London: Thames & Hudson. Lally, M. 2008. Bodies of difference in Iron Age southern England. In: O. Davis, N. Sharples and K. Waddington (eds), Changing perspectives in the first millennium BC, pp. 119-38. Oxford: Oxbow Books. Lamberg-Karlovsky, C. C. 1993. The biography of an object: the intercultural style vessels of the Third Millennium BC. In: W. D. Kingery (ed.), History from things: essays on material culture, pp. 27092. London: Smithsonian Institution Press. Last, J. 1998. Books of life: biography and memory in a Bronze Age barrow. Oxford Journal of Archaeology 17(1): 43-53. Latour, B. 2005. Reassembling the social: an introduction to actor-network theory. Oxford: Oxford University Press. Le Bon, L. 1995. Ancient ship graffiti: symbol and context. In: O. Crumlin-Pedersen and B. M. Thye (eds), The ship as symbol in prehistoric and Medieval Scandinavia. Papers from and international research seminar at the Danish National Museum, Copenhagen, 5th=7th May 1994, pp. 172-9. Copenhagen: PNM Publicationsfrom the National Museum Studies in Archaeology and History, Vol. 1. Leeds, E. T. 1933. Celtic Ornament in the British Isles down to A.D. 700. Oxford. Lillios, K. T. 1999. Objects of memory: the ethnography and archaeology of heirlooms. Journal of Archaeological Method and Theory 6(3): 235-62. Lloyd-Morgan, G. 1977a. Mirrors in Roman Britain. In: J. Munby and M. Henig (eds), Roman life and art in Britain, pp. 231-52. Oxford: BAR Reports 41(ii). Lloyd-Morgan, G. 1977b. Two Roman mirrors from Corbridge. Britannia 8: 335-8. Lloyd-Morgan, G. 1981a. The mirrors, including a description of the Roman mirrors found in the Netherlands, in other Dutch museums. Nijmegen: Ministry of Culture, Recreation and Social Welfare. Lloyd-Morgan, G. 1981b. Mirror handle. In: A. Gregory (ed.), Excavations in Thetford, 1980-1982, Fison Way, pp. 132. Dereham: East Anglian Archaeological Report. 53. Lloyd-Morgan, G. 1986. A lead mirror frame of the Roman period from Caerwent. Archaeologia Cambensis 135: 201-03. Lloyd-Morgan, G. 1991. Mirror handle. In: T. Gregory (ed.), Excavations in Thetford, 1980-1982, Fison Way, pp. 132. Dereham: East Anglian Archaeology Report 53. Lloyd-Morgan, G. 1993. The Celtic mirrors. In: W. J. Rodwell and K. A. Rodwell (eds), Rivenhall: investigations of a villa, church and village, 19501977. Volume 2 - specialist studies and index to volumes 1 and 2, pp. 29-33. London: CBA Research Report 80.

169

Lloyd-Morgan, G. and K. Reedie 1984. A new hand mirror from Kent. Archaeologia Cantiana 101: 355-57. Loveday, H. 1990. Chinese bronzes. Oxford: Ashmolean Museum. Lowery, P. and R. Savage 1976. Celtic design with compasses as seen on the Holcombe Mirror. In: P. M. Duval and C. F. C. Hawkes (eds), Celtic art in ancient Europe five protohistoric centuries: proceedings of the Colloquy held in 1972 at the Oxford Maison Francaise, pp. 219-31. London: Seminar Press. Lowery, P., Savage, R. and Wilkins, R. L. 1971. Scriber, graver, scorper, tracer: notes on experiments in bronzeworking technique. Proceedings of the Prehistoric Society 37: 167-82. Lowery, P. R., Savage, R. D. L and Wilkins, R. L. 1976. A technical study of the designs on the British mirror series. Archaeologia 105: 99-126. Lucas, G. 2005. The archaeology of time. London & New York: London. Lucy, S. 2005. The archaeology of age. In: M. DiazAndreu, S. Lucy, S. Babic and D. N. Edwards (eds), The Archaeology of identity: approaches to gender, age, status, ethnicity and religion, pp. 43-66. London: Routledge. MacDonald, P. 2007a. Perspectives on insular La Tène art. In: C. Haselgrove and T. Moore (eds), The later Iron Age in Britain and Beyond, pp. 32938. Oxford: Oxbow Books. MacDonald, P. 2007b. Llyn Cerrig Bach: a study of copper-alloy artefacts from the insular La Tène assemblage. Cardiff: University of Wales Press. MacGregor, G. 1999. Making sense of the past in the present: a sensory analysis of carved stone balls. World Archaeology 31(2): 258-71. MacGregor, M. 1976. Early Celtic art in north Britain: a study of decorative metalwork from the third century B.C. to the third century A.D. Leicester: Leicester University Press. MacKenzie, M. A. 1991. Androgynous objects. Amsterdam: Harwood Academic Publishers. MacKreth, D. 1991. Brooches of copper-alloy and iron. In: A. Gregory (ed.), Excavations in Thetford 1980-82, Fison Way. Dereham: East Anglian Archaeology 1. MacKreth, D. 1994. Late La Tène brooch. In: R. Thorpe & J. Sharman, an Iron Age and Romano-British enclosure system at Normanton-le-Heath, Leicestershire. Transactions of the Leicestershire Archaeological and Historical Society 68: 49-50. MacKreth, D. 2003. The linked brooches. In: J. Valentin, Manor Farm, Portesham, Dorset: excavations on a multi-period religious and settlement site. Proceedings of the Dorset Natural History and Archaeological Society 125: 41-3. Manby, T. G. 1980. Excavation of barrows at Grindale and Boynton, East Yorkshire, 1972. The Yorkshire Archaeological Journal 52: 19-47. Marshall, Y. 2000. Reading images stone b.c. World Archaeology 32(2): 222-235.

Marshall, Y. and C. Gosden, (eds). 1999. The cultural biography of objects. World Archaeology. London & New York: Routledge. Martin, E., Pendelton, C., Plouviez, J. and Geake, H. 2002. Archaeology in Suffolk 2001. Proceedings of the Suffolk Institute of Archaeology and History 40(2): 201-33. Martin, E., Pendelton, C. and Wreathall, D. 1998. Archaeology in Suffolk 1997. Proceedings of the Suffolk Institute of Archaeology and History 39(2): 20945. Mauss, M. 1973. Techniques of the body. Economy and Society 2: 70-88. Mauss, M. 1990. The gift: the form and reason for exchange in archaic societies. London: Routledge. McGlade, J. 1999. The times of history: archaeology, narrative and non-linear causality. In: T. Murray (ed.), Time and archaeology, pp. 139-63. London & New York: Routledge. McKinley, J. I. 1997. The cremated human bone from burial and cremation-related contexts. In: A. Fitzpatrick, Archaeological excavations on the route of the A27 Westhampnett Bypass, West Sussex, 1992. Volume 2: the cremations, pp. 55-73. Salisbury: Wessex Archaeology Report No. 12. Megaw, J. V. S. 1985. Meditations on a Celtic hobby-horse: notes towards a social archaeology of Iron Age art. In: T. C. Champion and J. V. S. Megaw (eds), Settlement and society: aspects of west European prehistory in the first Millennium B.C., pp. 161-91. Leicester: University of Leicester Press Megaw, J. V. S. and M. R. Megaw. 1990. The Basse-Yutz find: masterpieces of Celtic art: the 1927 discovery in the British Museum. London: Society of Antiquaries. Megaw, M. R. and J. V. S. Megaw. 1995. The nature and function of Celtic Art. In: M. J. Green (ed.), The Celtic world, pp. 345-75. London & New York: Routledge. Megaw, M. R. and J. V. S. Megaw. 2001. Celtic art: from its beginnings to the Book of Kells. 2nd Edition. London: Thames & Hudson. Millett, M. and S. McGrail 1987. The archaeology of the Hasholme logboat. Archaeological Journal 144: 69-155. Mills, N. 2000. Celic and Roman artefacts. Derbyshire: Buxton Press. Moore, T. 2007a. The early to later Iron Age transition in the Severn-Cotswolds: enclosing the household? In: C. Haselgrove and R. Pope (eds), The earlier Iron Age in Britain and the near Continent, pp. 259-78. Oxford: Oxbow Press. Moore, T. 2007b. Life on the edge? Exchange, community, and identity in the later Iron Age of the SevernCotswolds. In: C. Haselgrove and T. Moore (eds), The later Iron Age in Britain and beyond, pp. 41-61. Oxford: Oxbow Books. Moore, T. 2007c. Perceiving communities: exchange, landscapes and social networks in the later Iron Age of western Britain. Oxford Journal of Archaeology 26(1): 79-102.

170

Moreland, J. 1999. The world(s) of the cross. World Archaeology 21(2): 194-213. Morgunova, N. L. and O. S. Khokhlova 2006. Kurgans and nomads: new investigations of mound burials in the southern Urals. Antiquity 80(2): 30317. Mortimer, C. 1995. Non-ferrous metalwork and metalworking debris. In: J. Coles and S. Minnitt (eds), Industrious and fairly civilised. The Glastonbury Lake Village, pp. 140-2. Taunton: Somerset Levels Project and Somerset County Council Museums Service. Mortimer, C. and D. Starley 1995. Ferrous and non-ferrous metalworking. In: J. Coles and S. Minnitt (eds), Industrious and fairly civilised. The Glastonbury Lake Village, pp. 138. Taunton: Somerset Levels Project and Somerset County Council Museums Service. Munn, N. D. 1986. The fame of Gawa: a symbolic study of value transformation in a Massim (Papua New Guinea) society. Cambridge: Cambridge University Press. Munn, N. D. 1993. The cultural anthropology of time: a critical essay. Annual Review of Anthropology 21: 93-123. Murray, G. 1860-2. Balmaclellan. Proceedings of the Society of Antiquaries for Scotland 4: 293-5. Musson, C. R., Britnell, W. J., Northover, J. P. and Salter, C. J. 1992. Excavations and metal-working at Llwyn Bryn-dinas hillfort, Llangedwyn, Clwyd. Proceedings of the Prehistoric Society 58: 26583. Niblett, R. 1999. The excavation of a ceremonial site at Folly Lane, Verulanium. London: Britannia Monograph Series No 14. Northover, J. P. 1984a. Analysis of the bronze metalwork. In: B. Cunliffe (ed.), Danebury: an Iron Age hillfort in Hampshire. Volume 2: The excavations, 1969-1978: the finds, pp. 430-3. London: Council for British Archaeology Research Report. 52. Northover, J. P. 1984b. Iron Age bronze metallurgy in central southern England. In: B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in central southern Britain, pp. 430-3. Oxford: University of Oxford: Committee for Archaeology. Northover, J. P. 1987. Non-ferrous metallurgy. In: B. Cunliffe (ed.), Hengistbury Head Dorset. Volume 1: the prehistoric and Roman settlement, 3500 BC - AD 500, pp. 186-96. Oxford: Oxford University committee for archaeology monograph No.13. Northover, J. P. 1988. Copper, tin, silver and gold in the Iron Age. In: C. J. Salter and R. M. Ehrenreich (eds), Science and Archaeology Glasgow 1987: proceedings of a conference on the application of scientific techniques to archaeology Glasgow, Sep 1987, Part i, pp. 223-33. Oxford: BAR British Series 196 (i). Northover, J. P. 1991a. Non-ferrous metalwork and metallurgy. In: N. M. Sharples (ed.), Maiden Castle: excavations and field survey 1985-6, pp.

156-62. London: English Heritage Archaeological Report No. 19. Northover, J. P. 1991b. Non-ferrous metalwork and metallurgy. In: B. Cunliffe and C. Poole (eds), Danebury: an Iron Age hillfort in Hampshire, Volume 5: the excavations, 1978-1988, pp. 407-12. London: Council for British Archaeology Research Report. 73. Northover, J. P. 1993. Analysis of Celtic mirror I from Rivenhall. In: W. J. Rodwell and K. A. Rodwell (eds), Rivenhall: investigations of a villa, church and village, 1950-1977. Volume 2 - specialist studies and index to volumes 1 and 2, pp. 33-5. London: CBA Research Report 80. Northover, J. P. 1995. The technology of metalwork: bronze and gold. In: M. J. Green (ed.), The Celtic world, pp. 285-309. London & New York: Routledge. Nowakowski, J. A. 1991. Trethallan Farm, Newquay: the excavation of a lowland Bronze Age settlement and Iron Age cemetery. Cornish Archaeology 30: 5-242. Olivier, A. C. H. 1996. Brooches of silver, copper alloy and iron from Dragonby. In: J. May, Dragonby: report on excavations at an Iron Age and Romano-British settlement in North Lincolnshire, pp. 231-64. Oxford: Oxbow Books. Olivier, L. 1999. The Hochdorf 'princely' grave and the question of the nature of archaeological funerary assemblages. In: T. Murray (ed.), Time and archaeology, pp. 109-38. London & New York: Routledge. Olivier, L. 2001. Duration, memory and the nature of the archaeological record. In: H. Karlsson (ed.), It's about time. The concept of time in archaeology, pp. 61-70. Goteborg: Bricoleur Press. Olsen, B. 2003. Material culture after text: re-membering things. Norwegian Archaeological Review 36(2): 87-104. Palk, N. A. 1984. Iron Age bridle-bits from Britain. Edinburgh: University of Edinburgh Department of Archaeology. Palk, N. A. 1988. Metal horse harness of the British and Irish Iron Ages. Faculty of Anthropology and Geography. University of Oxford. Parfitt, K. 1995. Iron Age burials from Mill Hill, Deal. London: British Museum Press. Parfitt, K. 1998. A late Iron Age burial from Chilham Castle, near Canterbury, Kent. Proceedings of the Prehistoric Society 64: 343-51. Parker-Pearson, M. 1996. Food, fertility and front doors in the first Millennium BC. In: T. C. Champion and J. R. Collis (eds), The Iron Age in Britain and Ireland: recent trends, pp. 117-32. Sheffield: Sheffield Academic Press. Parker-Pearson, M. 1999a. The archaeology of death and burial. Stroud: Sutton Publishers. Parker-Pearson, M. 1999b. Food, sex and death: cosmologies in the British Iron Age with particular reference to East Yorkshire. Cambridge Archaeological Journal 9(1): 43-69.

171

Parker-Pearson, M. 2000. Great Sites: Llyn Cerrig Bach. British Archaeology 53: 8-11. Partridge, C. R. 1981. Skelton Green: a late Iron Age and Romano-British site. London: Britannia Monograph 2. Peers, L. 1999. Many tender ties: the shifting contexts and meanings of the S Black bag. World Archaeology 31(2): 288-302. Pendergrast, M. 2003. Mirror mirror: a history of the human love affair with reflection. New York: Basic Books. Pendergrast, M. 2007. Mirror mirror: a historical & psychological overview. In: M. Anderson (ed.), The book of the mirror: an interdisciplinary collection exploring the cultural story of the mirror, pp. 1-15. Newcastle: Cambridge Scholars Publishing. Pfaffenberger, B. 1988. Fetishised objects and humanised nature: towards an anthropology of technology. Man 23: 236-52. Pfaffenberger, B. 1992a. Social anthropology of technology. Annual Review of Anthropology 21: 491516. Pfaffenberger, B. 1992b. Technological dramas. Science, Technology & Human Values 17(3): 282-312. Pollard, S. 1974. A late Iron Age settlement and a Romano-British villa at Holcombe near Uplyme, Devon. Devon Archaeological Society 32: 59161. Price, A. J. and S. Cottam 1998. Romano-British Glass Vessels: A Handbook. London: CBA Practical Handbook in Archaeology 14. Quinnell, H. 1986. Cornwall during the Iron Age and the Roman period. Cornish Archaeology 25: 111-34. Rainbird, P. 1999. Entangled biographies: western Pacific ceramics and the tombs of Pohnpei. World Archaeology 31(2): 214-44. Ramm 1978. The Parisi. London: Gerald Duckworth and co Ltd. Raftery, B. 1983. A catalogue of Irish Iron Age antiquities. Marburg: Veroffentlinchung Des Vergeschichtlichen Seminars - Sonderband 1. Raftery, B. 1984. La Tene in Ireland: problems of origin and chronology. Marburg: Veroffentlinchung Des Vergeschichtlichen Seminars - Sonderband 2. Raftery, B. 1994. Pagan Celtic Ireland: the enigma of the Irish Iron Age. London: Thames & Hudson. Rodwell, W. J. and K. A. Rodwell 1986. Rivenhall: Investigations of a villa, church and village, 1950-1977. Chelmsford: CBA Research Report 55. Chelmsford Archaeological Trust Report 4. Rook, A., Lowery, P. R. Savage, R. D. A. and Wilkins, R. L. 1982. An Iron Age bronze mirror from Aston, Hertfordshire. Antiquaries Journal LXII: 18-34. Rowlands, M. 1993. The role of memory in the transmission of culture. World Archaeology 25(2): 14151. Rowlands, M. J. and J. P. Warnier 1993. The magical production of iron in the Cameroon Grassfields. In: T. Shaw, P. Sinclair, B. W. Andah and I. A.

Okopoko (eds), The archaeology of Africa: food, metals and towns, pp. 512-50. London and New York. Rynne, E. 1976. The La Tène and Roman finds from Lambay, county Dublin: a re-assessment. Proceedings of the Royal Irish Academy 76C: 231-43. Salter, C. J. and R. M. Ehrenreich 1984. Iron Age iron metallurgy in central southern Britain. In: B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in central southern Britain, pp. 146-61. Oxford: University of Oxford: committee for archaeology. Saunders, C. 1991. Iron objects. In: C. R. Musson (ed.), The Breiddin hillfort: a later prehistoric settlement in the Welsh Marches, pp. 141-7. London: Council for British Archaeology Research Report 76. Saunders, N. J. 1988. Anthropological reflections on archaeological mirrors. In: N. J. Saunders and O. de Montmollin (eds), Recent studies in Pre-Columbian archaeology, pp. 1-39. Oxford: BAR International series. 421(i). Saunders, N. J. 1999. Biographies of brilliance: pearls, transformations of matter and being, c. AD 1492. World Archaeology 31(2): 243-57. Saunders, N. J. 2001. A dark light: reflections on obsidian in Mesoamerica. World Archaeology 33(2): 22036. Saunders, N. J. 2002. The colours of light: materiality and chromatic cultures of the Americas. In: A. Jones and G. MacGregor (eds), Colouring the past: the significance of colour in archaeological research, pp. 209-26. Oxford: Berg. Scarre, C. 2002. Epilogue: colour and materiality in prehistoric society. In: A. Jones and G. MacGregor (eds), Colouring the past: the significance of colour in archaeological research, pp. 227-42. Oxford: Berg. Schama, S. 1995. Landscape and memory. London: Harper Collins. Schwabe, O. 1933. Der keltische spiegel von Hochheim. Mannus 25: 89-107. Schweig, B. 1941. Mirrors. Antiquity 15: 257-68. Sealey, P. R. 1996. The Iron Age of Essex. In: O. Bedwin (ed.), The archaeology of Essex. Proceedings of the Writtle conference, pp. 46-68. Chelmsford: Essex County Council. Sealey, P. R. 2006. Two decorated Iron Age mirror finds from Essex. In: P. Ottaway (ed.), A Victory Celebration: papers on the archaeology of Colchester and Late Iron Age-Roman Britain presented to Philip Crummy, pp. 11-18. Colchester: Friends of Colchester Archaeological Trust. Seip, L. P. 1999. Transformations of meaning: the life history of the Naxalk mask. World Archaeology 31(2): 272-87. Selkirk, A. 1984. Two chariot burials at Wetwang Slack. Current Archaeology 93: 302-6. Shanks, M. and C. Tilley 1987a. Re-constructing archaeology: theory and practice. 2. London & New York: Routledge.

172

Shanks, M. and C. Tilley 1987b. Social theory and archaeology. Cambridge: Polity Press. Sharples, N. 2008. Comment I. Contextualising Iron Age art. In: D. Garrow, C. Gosden and J. D. Hill (eds), Rethinking Celtic Art, pp. 203-13. Oxford: Oxbow Books. Skeates, R. 1995. Animate objects: a biography of prehistoric 'axe-amulets' in the central Mediterranean region. Proceedings of the Prehistoric Society 61: 279-301. Smith, R. A. 1909. On a Late-Celtic mirror found at Desborough, Northants, and other mirrors. Archaeologia 61(B): 329-46. Sofaer Derevenski, J. 1997. Age and gender at the site if Tiszapolgar-Basatanya, Hungary. Antiquity 71: 875-9. Sofaer Derevenski, J. 2000. Rings of life: the role of early metalwork in mediating the gendered life course. World Archaeology 31(3): 389-406. Sørensen, M. L. S. 1997. Reading dress: the construction of social categories and identities in Bronze Age Europe. Journal of European Archaeology 5(1): 93-114. Spector, J. D. 1993. What this awl means: feminist archaeology at a Watipeton Dakota village. St Paul: Minnesota Historical Society Press. Spence-Bate, C. 1871. On the discovery of a RomanoBritish cemetery near Plymouth. Archaeologia 40(2): 500-9. Spratling, M. 1970a. The late pre-Roman Iron Age bronze mirror from Old Warden. Bedfordshire Archaeological Journal 5: 9-16. Spratling, M. 1970b. The smiths of South Cadbury. Current Archaeology 2(18): 188-91. Spratling, M. 1972. Southern British decorated bronzes of the late pre-Roman Iron Age. Unpublished PhD thesis. Institute of Archaeology, University of London. Spratling, M. 1979. The debris of metal working. In: G. J. Wainwright (ed.), Gussage All Saints. An Iron Age settlement in Dorset, pp. 125-49. London: Department of Environment Archaeological Reports 10 Spratling, M. 1981. Metalwork at the Stanwick Oppidum: some new evidence. The Yorkshire Archaeological Journal 53: 13-16. Spratling, M. 2008. On the aesthetics of the Ancient Britons. In: D. Garrow, C. Gosden and J. D. Hill (eds), Rethinking Celtic Art, pp. 185-202. Oxford: Oxbow Books. Stacey, R. 2004. Evidence for the use of birch-bark tar from Iron Age Britain. Past 47: 1-2. Staelens, Y. J. E. 1982. The Birdlip Cemetery. Transactions of the Bristol and Gloucestershire Archaeological Society 100: 19-31. Stead, I. M. 1967. A La Tene III burial at Welwyn Garden City. Archaeologia 101: 1-62. Stead, I. M. 1975. The La Tène cemetery at Scorborough, East Riding. East Riding Archaeologist 2: 1-19. Stead, I. M. 1976a. The earliest burials of the Aylesford Culture. In: G. d. G. Sieveking, I. M. Longworth and K. E. Wilson (eds), Problems in eco-

nomic and social archaeology, pp. 401-16. London: Duckworth. Stead, I. M. 1976b. La Tène burials between Burton Fleming and Rudston, North Humberside. The Antiquaries Journal 56: 217-26. Stead, I. M. 1979. The Arras Culture. York: The Yorkshire Philosophical Society. Stead, I. M. 1984. Celtic Dragons from the River Thames. Antiquaries Journal 64: 269-79. Stead, I. M. 1985. The Battersea shield. London: British Museum Press. Stead, I. M. 1986. A group of Iron Age barrows at Cowlam, North Humberside. The Yorkshire Archaeological Journal 58: 5-15. Stead, I. M. 1988. Chalk figurines of the Parisi. The Antiquaries Journal 68: 9-29. Stead, I. M. 1991a. The Arras Culture. In: V. Kruta, O. H. Frey, B. Raftery and Szabo, M. (eds), The Celts, pp. 587-90. London: Thames & Hudson. Stead, I. M. 1991b. The Belgae in Britain: the Aylesford Culture. In: V. Kruta, O. H. Frey, B. Raftery and Szabo, M. (eds), The Celts, pp. 591-5. London: Thames & Hudson. Stead, I. M. 1991c. Iron Age cemeteries in East Yorkshire. London: Historic Buildings and Monuments Commission for England Archaeological Report 22. Stead, I. M. 1991d. Many more Iron Age shields from Britain. The Antiquaries Journal 71: 1-35. Stead, I. M. 1991e. The Snettisham Treasure: excavations in 1990. Antiquity 65: 447-65. Stead, I. M. 1996. Celtic Art in Britain before the Roman Conquest. 2nd. London: British Museum Press. Stead, I. M. 1998. The Salisbury hoard. Stroud: Tempus Publishing Ltd. Stead, I. M. 2006. British swords and scabbards. London: British Museum Press. Stead, I. M. and V. Rigby 1989. Verulamium: the King Harry Lane site. London: English Heritage Archaeological Report 12. Stead, I. and Hughes, K. 1997. Early Celtic designs. London: British Museum Press. Stillingfleet, E. W. 1846. Account of the opening of some barrows on the Wolds of Yorkshire. Archaeological Institute of Great Britain and Ireland York Vol: 26-32. Stoertz, C. 1997. Ancient landscapes of the Yorkshire Wolds: aerial photographs transcription and analysis. Swindon: Royal Commission on the Historical Monuments of England. Strathern, M. 1988. The gender of the gift. Berkeley: University of California Press. Taylor, R. E., Stuiver, M. and Reimer, P. J. 1996. Development and extension of the calibration of the radiocarbon time scale: archaeological applications. Quaternary Science Reviews 15: 655-68. Taylor, R. J. 1993. Hoards of the Bronze Age in southern Britain: analysis and interpretation. Oxford: British Archaeological Reports, British Series. Thomas, C. 1975. Recent fieldwork in the Isles of Scilly. Cornish Archaeology 14: 87-95. Thomas, J. 1996. Time, culture and identity: an interpretive archaeology. London & New York: Routledge.

173

Thomas, N. 1991. Entangled objects: exchange, material culture, and colonialism in the Pacific. Cambridge, Mas: Harvard University Press. Thomas, N. 2001. Introduction. In: C. Pinney & N. Thomas (eds), Beyond aesthetics: art and the technology of enchantment, pp. 1-12. Oxford & New York: Berg. Tilley, C. 1996. An ethnography of the Neolithic: early prehistoric societies in southern Scandinavia. Cambridge: Cambridge University Press. Tilley, C. 1999. Metaphor and material culture. Oxford: Blackwell Publishers Ltd. Tilley, C. 2004. The materiality of stone: explorations in landscape phenomenology. Oxford: Berg. Tuohy, T. 1999. Prehistoric combs of antler and bone. Oxford: British Archaeology Report, British series 285. Tylecote, R. F. and R. E. Clough 1983. Recent bog iron ore analysis and the smelting of pyrite nodules. Offa 40: 115-8. Valentin, J. 2003. Manor Farm, Portesham, Dorset: excavations on a multi-period religious and settlement site. Proceedings of the Dorset Natural History and Archaeological Society 125: 23-69. Van Dyke, R. M. 2003. Memory and the construction of Chacoan society. In: R. M. Van Dyke and S. E. Alcock (eds), Archaeologies of Memory, pp. 180-200. Oxford: Blackwell. Van Dyke, R. M. and S. E. Alcock, (eds). 2003. Archaeologies of memory. Oxford: Blackwell. Wacher, J. S. 1958. Litton Cheney Excavations, 1956. Proceedings of the Dorset Natural History and Archaeological Society 80: 160-77. Wainwright, G. J. 1979. Gussage All Saints: an Iron Age settlement in Dorset. London: Department of the Environment Archaeological Report 10. Wainwright, G. J. and M. Spratling 1973. The Iron Age settlement of Gussage All Saints. Antiquity 47: 109-30. Ward-Perkins, J. B. 1939. Iron Age metal horses' bits of the British Isles. Proceedings of the Prehistoric Society 5: 173-92. Way, A. 1869. Notices of certain Bronze relics of a particular type, assigned to the late Celtic period. Archaeology Journal 26: 70-3. Webster, J. 1995. Translation and subjection: interpretation and the Celtic Gods. In: J. D. Hill & C. G. Cumberpatch (eds), Different Iron Ages: studies on the Iron Age in temperate Europe, pp. 175183. Oxford: BAR International Series. Webster, J. 1997. Text expectations: the archaeology of 'Celtic' ritual wells and shafts. In: A. Gwilt and C. Haselgrove (eds), Re-constructing Iron Age societies, pp. 134-44. Oxford: Oxford Monograph 71. Weiner, A. B. 1985. Inalienable wealth. American Ethnologist 12: 210-27. Weiner, A. B. 1992. Inalienable possessions: the paradox of keeping-whil-giving. Oxford: University of California Press.

Weller, S. G. P. 1974. A late fourth-century cremation from Billericay, Essex. The Antiquaries Journal 54: 282-3. Wells, P. S. 2008. Image and response in Early Europe. London: Duckworth. Wheeler, R. E. M. 1936. The excavation of Maiden Castle, Dorset. Second interim report. Antiquaries Journal 16(3): 265-83. Wheeler, R. E. M. 1943. Maiden Castle, Dorset. Oxford: Oxford University Press. Whimster, R. P. 1977. Harlyn Bay reconsidered: the excavations of 1900-1905 in the light of recent work. Cornish Archaeology 16: 61-88. Whimster, R. P. 1981. Burial practices in Iron Age Britain. A discussion and gazetteer of the evidence c.700 B.C.- A.D. 43. Oxford: BAR British Series 90. Whitley, J. 2002. Objects with attitude: biographical facts and fallacies in the study of Late Bronze Age and

174

early Iron Age warrior graves. Cambridge Archaeological Journal 12(2): 217-32. Woodward, A. 1993. Discussion. In: D. E. Farwell and T. I. Molleson (eds), Poundbury Volume 2: the cemeteries, pp. 215-39. Dorchester: Dorset Natural History and Archaeological Society Monograph Series Number 11. Woodward, A. 2002. Beads and beakers: heirlooms and relics in the Early Bronze Age. Antiquity 76: 1040-7. Wurm, K. 1972. Eine stilkritische Untersuchung uber den fruhkeltischen Bronzespiegel von Hochheim am Main (Main-Taunus-Kreis). Fundber aus Hessen 12: 230-51. York, J. 2002. The life cycle of Bronze Age metalwork from the Thames. Oxford Journal of Archaeology 21(1): 77-92.