Gudme: Iron Age Settlement and Central Halls

By Palle Østergaard Sørensen

Gudme: The Iron Age Settlement and Central Halls presents, describes and interprets the many finds and structures that have been comprised during the extended excavations at the central parts of the Gudme locality on southeast Funen, Denmark. Head of excavation Palle Østergaard Sørensen extracts, combines, classifies, dates and temporalizes the many finds and houses from the excavations Gudmehallerne, Gudme III, Gudme IV.

Since the 19th century the Gudme area has been known as one of the richest prehistoric localities in Scandinavia, and more than 1,000 roman coins, close to 600 fibulas as well as several small mask and animal figurines form part of the Gudme find assemblage. From AD 200 to AD 600 the site expanded rapidly and covered as much as one square kilometer comprising up to 50 farmsteads. At the individual farms, specialized craftsmanship can be found and the debris and tools of gold- and silversmiths as well as bronze casting is abundant – here, bronze fragments, often from statues, imported from the Roman Empire form a distinct category of finds – bearing witness to a flourishing and innovative world of craftsmanship, as well as an extensive trade network.

During this period the unique and monumental halls, with the largest being 47 m long and 10 m wide, dominated the hilltops east of Gudme lake. Just to the south a smaller building accompanied the large hall, and had been purposely demolished and rebuilt several times at the exact same spot. The unusually large entrances to the two adjoined buildings lead straight from one to the other, thus witnessing a duality of buildings that came to define the aristocratic localities throughout Scandinavia in the following 800 years. Hence, Gudme represents a starting point for a significant type of architectural ideal as well as a first generation of central places.

• Language: English
• Publisher: Oxbow Books
• Release date: Sep 15, 2022
• ISBN: 9781789259087

Palle Østergaard Sørensen

Palle Ø Sørensen is an archaeologist and curator for the museum ROMU on Zealand, Denmark. He was head of the excavations in the Gudme area from 1991–2001. From 1995–1998 he held a research grant from the Danish Research Council, which formed the basis of this book.

Book preview

Queen Margrethe II of Denmark visited the excavation of the Gudme Halls on April 30 in 1993 and Palle Ø. Sørensen gave her a guided tour. The same day she opened the Gudme Exhibition at the Odense Bys Museer. Photo Pia Hammer, Krystal Foto.

This publication was sponsored with help from:

Published in the United Kingdom in 2022 by

OXBOW BOOKS

The Old Music Hall, 106–108 Cowley Road, Oxford, OX4 1JE

and in the United States by

OXBOW BOOKS

1950 Lawrence Road, Havertown, PA 19083

© Oxbow Books and Palle Østergaard Sørensen, Mads Dengsø Jessen and Mads Lou Bendtsen 2022

Hardback Edition: ISBN 978-1-78925-907-0

Digital Edition: ISBN 978-1-78925-908-7 (epub)

A CIP record for this book is available from the British Library

Library of Congress Control Number: 2022940169

All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopying, recording or by any information storage and retrieval system, without permission from the publisher in writing.

Printed in Malta by Melita Press

Typeset in India by Lapiz Digital Services, Chennai.

For a complete list of Oxbow titles, please contact:

UNITED KINGDOM

Oxbow Books

Telephone [phone]

Email: [e-mail]

www.oxbowbooks.com

UNITED STATES OF AMERICA

Oxbow Books

Telephone [phone], Fax [phone]

Email: [e-mail]

www.casemateacademic.com/oxbow

Oxbow Books is part of the Casemate Group

Front cover: Photos by Cille Krause, Roskilde Museum, and Kit Weiss and Lennart Larsen, National Museum of Denmark.

Back cover: Photos by Cille Krause, Roskilde Museum, and Lennart Larsen, National Museum of Denmark. Graphics by Mads Lou Bendtsen, National Museum of Denmark.

Contents

Preface

I. Gudme – an introduction

I.1. Discoveries and excavations at Gudme 1833–1966

I.2. A new focus on Gudme and Lundeborg from 1982

I.2.1. Lundeborg

I.3. The Gudme-Lundeborg Project

I.3.1. Phase I

I.3.2. Phase II

1. The Gudme halls

1.1. Introduction

1.2. The Neolithic

1.3. Late Bronze Age

1.4. The Iron Age settlement

1.4.1. Phase 1

1.4.2. Phase 2

1.4.3. Feature 556

1.4.4. Phase 3

1.4.5. House I – the great hall

1.4.6. Phase 4

1.4.7. Phase 5

1.4.8. Phase 6

1.4.9. Phase 7

1.4.10. Phase 8 – farm 1

1.4.11. Phase 8 – farm 2

1.4.12. Phase 9 – farm 1

1.4.13. Phase 9 – farm 2

1.5. The finds

1.5.1. Objects of gold and silver objects with gilded sheet fragments

1.5.2. Objects of silver

1.5.3. Coins

1.5.4. Objects of bronze

1.5.5. Objects of iron

1.5.6. Ceramic material

1.5.7. Glass

1.5.8. Amber objects

1.5.9. Precious stone

1.5.10. Smithing waste

1.5.11. Antler objects

1.6. The halls of the Iron Age

1.6.1. The Gudme halls

1.6.2. Scandinavian sites with hall buildings

1.6.3. Cult centres and royal residences

1.6.4. Hall buildings

2. Gudme III

2.1. The settlement

2.2. The course of the excavation

2.2.1. Excavation method

2.3. The Late Bronze Age

2.4. Pre-Roman Iron Age

2.5. Settlement in the Late Roman and Germanic Iron Age, c. AD 200–750

2.5.1. Phase 1 and 2 – farm 1

2.5.2. Phase 2 – farm 2

2.5.3. Phase 3 – farm 1

2.5.4. Phase 3 – farm 2

2.5.5. Phase 4 – farm 1

2.5.6. Phase 4 – farm 2

2.5.7. Phase 5 – farm 1

2.5.8. Phase 5 – farm 2

2.5.9. Phase 5 – farm 3

2.5.10. Phase 5 – possible burials at farm 3

2.5.11. Phase 6 – farm 1

2.5.12. Phase 6 – farm 2

2.5.13. Phase 6 – farm 3

2.5.14. Phase 7 – farm 1

2.5.15. Phase 7 – farm 2

2.5.16. Phase 7 – farm 3

2.5.17. Phase 8 – farm 1

2.5.18. Phase 8 – farm 2

2.5.19. Phase 8 – farm 3

2.5.20. Phase 9 – farms 1 and 2

2.5.21. Phase 9 – farm 3

2.5.22. The end of the settlement

2.6. The typological and chronological development of the houses

2.6.1. The longhouse

2.6.2. Small houses

2.7. The finds

2.7.1. Objects of gold and silver

2.7.2. Coins

2.7.3. Bronze objects

2.7.4. Tools, fittings and other objects of iron

2.7.5. The ceramic material (Bronze Age and Pre-Roman Iron Age see 2.3. and 2.4.)

2.7.6. Glass

2.7.7. Amber

2.7.8. Wood

2.7.9. Smithing waste

2.7.10. Quern stone and whetstone

2.8. The hoards from Gudme III

2.8.1. The hoard of Roman siliquae

2.8.2. The hacksilver hoard

2.8.3. The hoard containing gold sheet

2.9. The development of Gudme III – original landscape, wetland and settlement

2.9.1. The development of the farms at Gudme III – production and prosperity

3. Gudme IV

3.1. The settlement

3.2. The excavation

3.3. Pre-Roman Iron Age

3.4. The settlement from the Late Roman Iron Age, Germanic Iron Age and Early Viking Age, c. AD 250–850

3.4.1. Phase 1 – farm 1

3.4.2. Phase 2 – farm 1

3.4.3. Phase 2 – farm 2

3.4.4. Phase 3 – farm 1

3.4.5. Phase 3 – farm 2

3.4.6. Phase 4 – farm 1

3.4.7. Phase 4 – farm 2

3.4.8. Phase 5 – farm 1

3.4.9. Phase 5 – farm 2

3.4.10. Phase 6 – farm 1

3.4.11. Phase 6 – farm 2

3.4.12. Phase 7 – farm 1

3.4.13. Phase 7 – farm 2

3.4.14. Phase 8 – farm 1

3.4.15. Phase 8 – farm 2

3.4.16. Phase 9 – farm 1

3.4.17. Phase 10 – farm 1

3.4.18. Phase 11 – farm 1

3.4.19. Phase 12 – farm 1

3.4.20. The end of the settlement

3.5. The finds from Gudme IV

3.5.1. Objects of gold and silver

3.5.2. Coins

3.5.3. Bronze brooches (with Karsten K. Michaelsen)

3.5.4. Bronze objects

3.5.5. Tools and other iron objects

3.5.6. The ceramic material

3.5.7. Glass

3.5.8. Beads of glass and amber

3.5.9. Antler

3.5.10. Remains from metal smithing

3.5.11. Quern and whetstone

3.6. The coin hoard from Gudme IV

3.6.1. The Viking hoard from Gudme IV east

3.7. The development of the settlement and its relation to the wetlands

3.7.1. The development of the houses at Gudme IV compared with Gudme III

4. The house typology at Gudme and in the rest of the Danish area

4.1. The house typology in the Gudme area

4.1.1. Houses with a straight course of roof-bearing posts

4.1.2. Houses with a curved course of roof-bearing posts

4.1.3. Houses from the rest of the Gudme area

4.1.4. The decline in the size of the longhouses at Gudme in the 5th century – a sign of crisis in society?

4.2. Houses and farmsteads from the rest of the Danish area

4.2.1. Houses on Funen

4.2.2. Houses in East Denmark

4.2.3. East Danish house chronology

4.2.4. Houses in Jutland

4.3. Gudme after 25 years

Abbreviations and general information

Bibliography

Preface

For several centuries, the archaeology of Gudme has held a prominent position in the Iron Age studies of South Scandinavia. Beginning with the private collection of landowner Niels Frederik Berhard Sehested and his interest in archaeology, which was sparked in him when, as a youngster, he oversaw the excavation of the famous Broholm treasure in 1833, this corner of Funen has continued to attract attention up to the present. Consequently, and since the early days, the number of important finds and excavations in and around Gudme and all the way to Lundeborg on the coast of the Great Belt has increased steadily year after year. The many metal detector finds are especially worth noticing. Gudme was already in the early 1980s a famous site and has been so ever since. Magnificent objects just keep being unearthed here.

The many treasure hoards and detector finds from the area provided the incentive for an archaeological investigation of the Gudme settlement. Quickly the demand for skilled excavators rose, and the young, experienced student Palle Østergaard Sørensen filled the qualifications. Little did he know that the seemingly inconspicuous excavations in Gudme would end up taking up major parts of his research life. Since his first work in Gudme, Palle has invested enormously in the registration, evaluation and dating of the multitude of postholes, house plans and finds from the excavations. It has for long periods of time been a relatively ‘lone ride’ for Palle, but he has persisted in deciphering the settlement and the outcome of his tenacious effort is Gudme: Iron Age Settlement and Central Halls.

In essence, this book is the first comprehensive presentation of one of the so-called ‘central places’ of the Scandinavian Iron Age. In it you will find a detailed description of almost every feature and find from the excavations at the Gudme Halls, Gudme III and Gudme IV. The finds material is as impressive as the descriptions, and the result is a publication where a single category of finds can be studied in detail as well as the individual phases of the farms from the different excavations.

The first draft was written between 2004–2006 when Palle was employed at the National Museum. The draft was essentially an enhanced version of a thesis that he wrote during a three-year scholarship from the Danish Research Council, which focused on deciphering the many postholes and pits found in the various Gudme excavations, including their finds content, dating and interrelatedness. Many of the illustrations and house plans you see in the present book formed part of his research, carried out in collaboration with archaeologist Cille Krause. Back then, the digitisation of Danish archaeology was still in its infancy. Nevertheless, digital processes were completely integrated as part of the registration and presentation of the different Gudme settlements. Despite the importance and great potential of the site, the actual funding for the publication of the Gudme manuscript could not be raised, and the draft stayed with the National Museum.

Later, in 2011–2014, the Gudme manuscript was updated as part of the research project ‘Pre-Christian Cult Sites’, funded by the A.P. Møller Foundation. The project was hosted by the National Museum under the management of research professor Lars Jørgensen, and Gudme was one of the four selected key archaeological sites included in the project. Palle again took up his Gudme research and integrated new finds, sites and excavations related to Gudme into the draft for the present book. However, this revised version did not find its way to the public either, as Lars shortly afterwards became ill and tragically passed away in the summer of 2016. Lars’s death came as a shock to the museum and to everyone involved with the project, and it of course paused the finalisation of ‘Pre-Christian Cult Sites’.

In late 2017, I took over the project and decided together with Palle to start raising money for the publication. Thankfully, the Carlsberg Foundation saw the potential in the important material and has generously funded the present publication. Mads Lou Bendtsen came on board via a grant from the Elisabeth Munksgaard Foundation and he has been part of the editorial group ever since. We are extremely grateful for the support and help the foundations have provided, especially during the pandemic. Since early 2019 we have worked on the manuscript and a great many people have helped us along the way – from different museums and institutions to individual colleagues and friends. You know who you are and we are very thankful for all the assistance and positive input we have been met with. Still, a special thanks goes to Patrick Marsden for the translation and Johan Z. Gårdsvoll for helping out when needed. Lastly, Oxbow Books has assumed the publishing of the book. It has been a fruitful collaboration and we hope to have more books follow this one and complement the research of ‘Pre-Christian Cult Sites’.

So here it is, the results of the extensive Gudme excavations are laid out and ready for any researcher or layperson alike to delve into. The contents are substantial, and often quite complicated, but presented openly and at the highest possible level of detail. We hope for it to spark much debate and dialogue, and are confident that Gudme: Iron Age Settlement and Central Halls will form part of Iron Age research for many years to come.

Copenhagen, April 2022

Mads Dengsø Jessen

I

Gudme – an introduction

Poul Otto Nielsen

Gudme on Funen perhaps does not need any introduction at all, given the remarkable discoveries that have been made here since the mid-19th century, which reflect the importance of this area during the Iron Age. But a brief summary of events will lead us to what this book is about, which is the discovery of a magnate’s residence, possibly at a royal level, amongst the largest concentration of farmsteads from the Late Roman and Germanic Iron Age found in Denmark, covering approximately 50 ha and consisting of an estimated c. 60 farms. Gudme was also a production centre for jewellery of bronze, silver and gold, and significantly, a place for the worship of gods, as is reflected by the name Gudme, meaning ‘Home of the Gods’. During the development of this community, we discover who this worship was directed towards, and for the first time, find images depicting deities belonging to the Nordic mythology, a belief that was adjusted to the ideals of a warrior aristocracy.

In this book, Palle Østergaard Sørensen describes the excavations of two agrarian settlements, Gudme III and IV, which developed continuously through the Late Roman and Germanic Iron Age, and the excavation of a complex of impressive ‘halls’. The latter are interpreted as belonging to successive, high-ranking leaders, who ruled this society, and as part of their duties, were responsible for religious affairs.

The excavations at Gudme examined here took place in the last decades of the 20th century. During these years as well as at the beginning of this century, other so-called central sites were discovered in Scandinavia. At some of these, remains have been recorded of large buildings of a comparable type to those found at Gudme, or buildings with a special architectural character or function. In the final part of Chapter 1, Palle Ø. Sørensen compares such sites with Gudme, in an attempt to establish which of them could have been royal seats in the Iron Age.

In this introduction, before focusing upon the results of the excavations presented in the following pages, we will summarise the early research into Gudme, and examine the events that led to the new investigations, starting with an overview of the finds and archaeological remains in the order of their discovery.

I.1. Discoveries and excavations at Gudme 1833–1966

It all began in 1833, with the discovery of the second largest gold hoard from the Iron Age in Denmark, the Broholm Treasure, consisting of 51 golden objects with a total weight of 4153.8 g (Fig. I.1) (Sehested 1878, 199 ff., pl. XLI–XLIV; Mackeprang 1952, 120 f.). This was found during ploughing of a previously uncultivated area called Enemærket, less than 1 km from the east coast of Funen and near the fishing village of Lundeborg (Fig. I.2). Metal detector surveys and excavations in 1983 and 1991 added 26 more gold objects to the hoard, and its total weight increased to 4641.54 g (Michaelsen & Thomsen 1991). It also subsequently became apparent that the Broholm Treasure may have consisted of two separate hoards, one containing male objects, such as heavy gold rings, including two Kolbenringe and scabbard mounts, and another containing female objects, including fibulae, finger rings and gold bracteates. Both of the proposed two hoards dated to the first half of the 6th century AD. The Broholm Treasure was found on dry land at a location where no traces of settlement were detected. It is interpreted as one or two hidden collections of valuables, which were intended for subsequent recovery and use (Michaelsen & Thomsen 1991, 13). This was not the first hoard from the Germanic Iron Age to be discovered in south-east Funen, with the Elsehoved hoard having been found in 1826 (Sehested 1878, 211 ff., pl. XLVI), but it was the first hoard that indicated the presence of high-ranking aristocrats in this part of Denmark.

Fig. I.1. The Broholm Treasure. Photo, Kitt Weiss, the National Museum.

Fig. I.2. Map of the Gudme area. 1) The Broholm Mansion, 2) the Broholm Treasure, 3) Elsehoved and 4) Møllegårdsmarken. Primary research area, Zone 1, and Reference area, Zone 2. Green dot: gold, red dot: settlement, yellow dot: stray find, blue dot: cemetery.

An individual who was very much influenced by the discovery of the Broholm Treasure was the young N.F.B. Sehested (1813–82), who inherited the Broholm estate in 1839. During the 1850s, he began to devote himself to archaeology, recording finds from the area around Broholm, and later conducting excavations of burial mounds, cemeteries and settlements, at the same time setting new standards for excavation and documentation, with the results published in two volumes (Sehested 1878; 1884; cf. Randsborg 1995; Henriksen 2009, vol. 1, 18). He built a museum for his collection of antiquities and also became a pioneer in experimental archaeology (Thrane 1994).

One of Sehested’s excavations was undertaken in 1876–81 at the cemetery at Møllegårdsmarken, where he discovered 381 cremation burials dating to the Roman Iron Age. Excavations were resumed here in 1959 by Erling Albrectsen of Fyns Stiftsmuseum, Odense, who found 1642 more graves, making Møllegårdsmarken the largest Iron Age cemetery in Denmark, which was in continuous use from the late Pre-Roman Iron Age until c. AD 400 (Albrectsen 1968; 1971). It stands out in comparison with cemeteries from the Roman Iron Age in neighbouring areas (Michaelsen 1994; Henriksen 2009, 297 ff.), as well as in the rest of Scandinavia, not only because of its size, but also due to the number of Roman imports found in the graves, especially those from the Late Roman Iron Age. In her 1987 study, Ulla Lund Hansen recorded 299 finds of Roman imports from the Late Roman Iron Age in Denmark, 85 of which (28.4%) are from graves excavated at Møllegårdsmarken. Within the community that used this cemetery, some individuals were of a rank that enabled them to possess luxury items of foreign origin. Analyses of the composition of grave goods demonstrate that social differences increased during the Early–Late Roman Iron Age transition, c. AD 175 (Christoffersen 1987).

Sehested recorded 18 stray finds of gold objects from the area around Broholm, most of these probably dating to the Early Germanic Iron Age (1878, 199 ff.; 1884, 113). These included three ring-knobs of solid gold found on different occasions at Egsmosegård, to the east of the present village of Gudme. They were most likely components of a scattered hoard (Fig. I.3).

Fig. I.3. Three ring-knobs of solid gold from Egsmosegård, Gudme, and one from Albjerg (the small one). Photo, Lennart Larsen, the National Museum.

In 1902, a hoard was found at Lillesø, 800 m south-east of Egsmosegård. This contained another Kolbenring, together with two spiral scabbard mounts and three spiral rings, all of gold and weighing 605 g in total (Jørgensen & Petersen 1998, 211 f.).

The area in which hoards and stray finds of precious metal objects have been found extends from the present village of Gudme in the west to the coast in the east (Fig. I.2). In 1950, a hoard was found during forest clearance in Hesselager Fredskov, near the mouth of the Tange watercourse, to the north of Lundeborg (Voss 1952). It contained a fine, twisted gold chain, which terminates in animal-head mounts. This may have functioned as a belt, to which a golden capsule pendant with inlaid garnets also belonged. Both objects were probably made in the East Roman Empire. The find also included fragments of a late Roman silver dish. In addition, three golden objects were found on different occasions in Hesselager Fredskov, and two identical gold bracteates and a neck ring that was finely decorated with a punched pattern (Jørgensen & Petersen 1998, 218 f.).

I.2. A new focus on Gudme and Lundeborg from 1982

A new era began with the introduction of metal detector surveys at Gudme. In 1982, private detectorists located a new hoard containing gold bracteates at the location that is now known as Gudme II (Fig. I.4), and the find-spot was subsequently investigated by the National Museum (Poulsen 1987; Petersen 1988a; 1990; 1994). This was an important find for more than one reason. It was the first time that a hoard containing objects of precious metals dating to the Germanic Iron Age had been found in a settlement context in Denmark, or more specifically, in a posthole belonging to a building. During the following years of metal detector surveys and excavations, more hoards associated with settlement remains were found at Gudme (Madsen & Thrane 1991; Sørensen 1994b; 2001; 2003).

Fig. I.4. Hoard from Gudme II. Photo, Lennart Larsen, the National Museum.

The numerous metal objects that were now being recovered by metal detector surveys in the areas between the new hoards in the Gudme area meant that there was a steady flow of treasure trove to the National Museum. In order to find an explanation for these significant numbers of precious objects, excavations began in areas with concentrations of finds, such as the site of Gudme I, where 11 Roman solidi from the period AD 340–355 were found scattered on several different occasions, although these most likely originated from a single deposition (Kromann 1987, 65). In the same area, 32 denarii from the 1st and 2nd centuries AD and 10 siliquae of approximately the same date as the solidi were also found. The excavation covering an area of only 700 m², which was undertaken in 1984, revealed the remains of seven house sites, some of which could be dated on a typological basis to the 4th and 5th centuries (Kromann & Petersen 1985; Petersen 1987).

At Gudme III (Stærkærvej), another nine Roman silver coins were recovered by metal detector users, and in connection with a construction project in this area, an excavation began here in 1985. During this excavation, a hoard of 285 Roman siliquae was discovered, and this was the first hoard consisting of this coin type that had been found in Northern Europe, and a major settlement complex dating to between c. AD 200 and 650 was also uncovered (Petersen 1987, 57 ff.; 1994, 32–34; Kromann et al. 1991, 145 f., 154 f.; Sørensen 1994a, 41–44; 2001). It became clear that the hoards and concentrations of metal finds indicated the presence of settlement dating to the Late Roman and Germanic Iron Age. Gudme III is one of the settlement complexes that are described by Palle Ø. Sørensen in this volume.

In 1984, a disturbed hoard was discovered at Gudme V (Stenhøjgård), containing silver ingots and scrap silver, much of which had been cut from Roman tableware (Kromann & Petersen 1985, fig. 7). Together with the siliqua hoard from Gudme III and later finds of cut-up Roman silver from Gudme, this find indicates that before the ‘golden age’ of the Early Germanic Iron Age, a massive amount of silver had already been imported to this location during the Late Roman Period. Why was all this metal accumulated? In a posthole in one of the houses excavated at Gudme V, another hoard was found, consisting of melted pieces of silver and gold, as well as melted lumps of blue glass. It has been proposed that the objects were buried below the floor of the house close to the post and that the metal and glass melted, and ran down into the posthole when the house burnt down (Petersen 1994, 35 ff.). An intact silver rosette fibula inlaid with blue glass was also found at Gudme V (Kromann et al. 1991, fig. 5).

A significant number of crucibles with traces of melted silver and gold, as well as thin strips of gold foil, were found at the same site, demonstrating that the production of fine metal ornaments was undertaken here. The objects found using metal detectors at Gudme I and II include scrap silver and melted metal, indicating that smithing also took place there, although these are single finds and thus undated (Kromann & Petersen 1985, 205; Thrane 1987, 23). It is, nevertheless, apparent that some of the farmsteads at Gudme had specialised craftsmen associated with them and, judging from the finds such as the rosette fibula from Gudme V, the quality of their work was excellent. Gudme was probably a centre for the sophisticated production of ornaments in Nordic style, based on imported metals, in the form of ingots of bronze, silver and gold, and Roman coins or scrap metal from overseas sources on the Continent (Jørgensen 1999; Axboe 2012, 126–128).

I.2.1. Lundeborg

In 1986, a large settlement area was discovered during a rescue excavation on the coast, to the north of the fishing village of Lundeborg (Thomsen 1987; 1993; 1994). A somewhat homogeneous occupation layer, with an average thickness of 0.5 m, extended c. 900 m along the coast, 350 m to the south of the outlet of the watercourse Tange Å (Lundeborg I) and 550 m to the north of it (Lundeborg II). The finds included tools for carpentry and other handicrafts, such as bronze casting, and broken pieces of glass, glass beads, sherds of terra sigillata and 150 denarii. Ingots, crucibles, scrap silver and pieces of gold testify to the presence of gold- and silversmiths. But longhouses were not discovered like at Gudme. The settlement remains that were found were interpreted as small buildings for temporary rather than permanent use. This, combined with the foreign imports amongst the finds, led to the conclusion that Lundeborg was an early trading port, which was active from the Late Roman Iron Age onwards, and was the earliest of its kind in Northern Europe. The connection with Gudme, located only 3 km inland from Lundeborg, was obvious. At both locations, there was a particularly high level of activity during the 3rd and 4th centuries. It has therefore also been suggested that Lundeborg was used as a landing place that was associated with periodic meetings of large numbers of people at Gudme, a royal seat and administrative centre (Ulriksen 1998, 219). Lundeborg I lasted until c. AD 550, whilst the activities at Lundeborg II continued until c. AD 800. Very close to Lundeborg II, hoards containing golden objects have previously been found in Hesselager Fredsskov (Voss 1952) and it was also here that 102 miniature gold foil figures (guldgubber) dating to the Late Germanic Iron Age were found within a small area in 1989–92 (Thomsen 1989; 1990; 1992; 1993, 87 ff.). So the two sites shared similarities to begin with, but Lundeborg II had a longer history and stands out as being a place of special significance.

I.3. The Gudme-Lundeborg Project

The new focus on Gudme was followed up by symposia, with the first of these organised by Henrik Thrane and held in 1984, followed by others (Thrane 1985; 1990; Petersen 1989; Henriksen & Michaelsen 1995). These meetings enabled recent finds and discoveries to be presented, as well as being a forum for discussions about new perspectives within a wider audience.

To coordinate the archaeological investigations that began after the finds of treasure trove were made at Gudme, the Danish Archaeological Board established the Gudme Committee in 1987.¹ Four museums were involved in the area, Fyns Oldtid (the former Fyns Stiftsmuseum) in Odense, the main museum on Funen, where there was a long tradition of research into the Iron Age of the island, and where the grave finds from the Møllegårdsmarken cemetery were kept. Svendborg & Omegns Museum, which was responsible for rescue excavations in south-eastern Funen, including Gudme. Langelands Museum, with responsibility for maritime archaeology. And finally, the National Museum, to which the treasure trove was submitted and which subsequently undertook excavations at some of the find-spots, usually in collaboration with one of the other museums. The Gudme Amateur Archaeology Group participated locally, carrying out surveys and recording finds (Henriksen 1987).

It was decided that future investigations should be based on close cooperation between the museums. On this basis, a plan was drawn up, resulting in the ‘Gudme-Lundeborg Project’. This was realised due to generous support from A.P. Møller og Hustru Chastine Mc-Kinney Møllers Fond til almene Formaal. A preliminary programme, Phase I, was undertaken in 1988–90, and a final Phase II in 1990–93, with these focusing on different aspects. In 1991, an account was given of the results of Phase I, together with the plan for the recently launched Phase II (Nielsen 1994). The following provides a summary of the results of both phases of the project, including references to the articles that have been published between then and the time of writing.

I.3.1. Phase I

During this phase, five aspects were focused on.

Landscape and settlement. Locating and recording Iron Age settlements within the area was an important initial task, which had been started by Sehested and was resumed by Klavs Randsborg in the early 1980s (Randsborg 1982). Surveys were now undertaken by Fyns Oldtid and Svendborg & Omegns Museum, assisted by the Gudme Amateur Archaeology Group. In 1990, members of the latter discovered three new hoards close to the present village of Gudme (Madsen & Thrane 1991). New cemeteries dating to the Late Roman Iron Age were identified at Egelygård and were subsequently excavated (Michaelsen 1989; 1990; 1994). In order to find suitable locations for pollen analyses, test samples were taken from small depressions in the middle of the Gudme settlement area by the Geological Survey of Denmark. More useful samples were subsequently obtained from Gudme Lake (see below).

The population background. Since Albrectsen’s excavations of the central cemetery Møllegårdsmarken in 1959–66, the area had been cultivated. Surveys now indicated that additional graves were being disturbed by ploughing close to the previously excavated area. New excavations were therefore begun by Fyns Oldtid, and these continued during Phase II (see below).

Hoard sites and settlement. An important aspect was investigating the connection between hoards and settlements. Excavations had already taken place at sites such as Gudme I, II and III, with significant results, and further excavations were therefore carried out by the National Museum at Uhrenholtgård and Gudme V (Stenhøjgård) (Petersen 1994; Sørensen 1994a). As previously mentioned, the latter excavation resulted in the discovery of a new hoard, which indicated that metalworking had been undertaken at one of the farmsteads dating to the Late Roman Iron Age. Characterisation of the pottery from the Late Roman and Germanic Iron Age was part of this sub-programme (Stilborg 1994).

The coastal settlement and possible landing places. Even before the discovery of the settlement at Lundeborg, Ole Crumlin-Pedersen (1987, 116 ff.) had suggested that the mouth of the Tange Å was a probable natural harbour that was connected to Gudme. In order to test this hypothesis, geological investigations were undertaken, which demonstrated that coastal ridges blocking the mouth of the Tange Å had existed since the Late Neolithic, thus making access to this watercourse difficult (Binderup 1989a; 1989b). It is therefore more likely that ships landed on the low, sandy shore along the coast.

Cult sites. The proximity of place names with allusions to pre-Christian religion indicated that several sacred sites were previously located within the area, such as Gudbjerg – ‘hill of the gods’, Albjerg – ‘hill at or with the heathen temple’, Galdbjerg – ‘hill of the offerings’ and Gudme itself – ‘home of the gods’ (Sørensen 1985). In the middle of this area was Gudme Lake, which may also have been a sacred site, like many lakes and wetlands containing war booty and/or ritual depositions known from the Iron Age in South Scandinavia at least until c. AD 500, when ritual depositions in wetland areas declined (Fabech 1991; Henriksen 2010, 424). Investigations of this lake were made more difficult by its shallowness, and because its sediments were as thick as c. 15 m. In spite of this, Ole Grøn of Langelands Museum managed to lower a metal detector probe through a tube down into the deep sediments (Grøn 1989). But no signals were produced that indicated offerings of metal objects had been made in the lake. This does not, however, rule out the possibility that offerings of some kind were placed in it. We will have to wait until more advanced methods of remote sensing have become available in order to clarify this. But as the deep sediments in the lake turned out to be suitable for sediment and pollen analyses, we returned to Gudme Lake in phase II.

I.3.2. Phase II

Six aspects were focused on during this phase, some of which were in continuation of Phase I.

The settlement history of the Gudme area c. 100 BC–AD 1000. Two zones were defined, Zone 1 representing the area containing the wealthy settlements dating to c. AD 200–600, whilst Zone 2 was defined as a surrounding reference area, in which cemeteries and settlements apparently reflected the general development on Funen during the 1st millennium AD. The two zones are marked on Figure I.2. Besides providing a foundation for establishing priorities in terms of field surveys and excavations, this division into zones was also intended to enable the future planning of physical protection of a highly important archaeological resource area (Zone I). The most intensive field surveys have taken place in Zone I covering c. 60 ha. A study of the metal objects was carried out by Lars Jørgensen in 1990–92, at that time based on 1500 individual finds dating from before c. AD 200 until 1200, but with a concentration of material between c. AD 325 and 550 (Jørgensen 1995a, 207 ff.). Environmental studies also constituted part of this sub-project, which were based on sedimentation and pollen records of samples from Gudme Lake. These, for example, showed that there was intensification of arable farming during the period AD 400–500 (Rasmussen & Olsen 2009).

Population studies. The excavations at the Møllegårdsmarken cemetery continued, resulting in 257 new graves being found after the excavations had been resumed in 1988, with the total number of graves reaching 2,280 (Henriksen 2009, 380 f.). Sieving of the contents of the cremation burials now provided a more complete picture of what had originally been cremated. Age and sex analyses of a representative sample of the cremated bones from the graves were undertaken by Per Holck of the University of Oslo (Holck 1997). New features were recorded during the excavations, and four structures, three rectangular and one circular, consisting of postholes, were interpreted as remains of mortuary houses or platforms used in connection with funerals (Henriksen 1991, 32 f.; 2009, 79; Madsen & Thrane 1995). One of these was constructed on top of a paved trackway, which therefore must have existed at the same time the cemetery was in use, cutting across its eastern part like a via sacra (Madsen & Thrane 1995, 78 ff.; Thrane 2009; 2013). In 1992, a trial excavation was undertaken at Langå to the north of Gudme, in an attempt to locate the cemetery containing chieftains’ graves from the late Pre-Roman Iron Age found by N.F.B. Sehested and H. Petersen in 1877–87 (Sehested 1878, 172 ff.). One more cremation grave was found, although the investigation was not continued (Henriksen 1993). A total excavation was instead undertaken at Brudager, one of the cemeteries selected within Zone II, 3.5 km to the south-west of Gudme. Compared with the Møllegårdsmarken cemetery, the 168 graves excavated at Brudager represent a local population from a lower social level. It began c. AD 100, and like Møllegårdsmarken ended around AD 400, like many other cemeteries on Funen and elsewhere in Scandinavia, because of a change in burial custom (Henriksen 2009).

The Migration Period settlement at Gudme. In order to uncover remains of the settlement pattern in the Early Germanic Iron Age c. AD 400–550, trial excavations were undertaken in and around the Gudme II site, where some of the high-status finds from this period were concentrated. These demonstrated that the individual farms were placed on higher ground surrounded by wetland areas, and due to the topography, the positions of these farms were stationary. They were rebuilt, but remained at the same sites for centuries. This pattern also applied to most of the area where indications of settlement were detected. A method of locating settlements involved measuring the amount of plant-accessible phosphorous, with the high values coinciding with concentrations of surface finds (Jørgensen et al. 1993). The analyses of pottery manufacture continued as part of this sub-project, subsequently resulting in a separate, methodological study (Stilborg 1997).

The ‘trading port’ at Lundeborg. During submarine investigations in 1990, sherds of Early Iron Age pottery were recovered 200 m from the coast at Lundeborg, close to remains of a stone barrier. These may indicate that there was a shallow ‘harbour basin’, which was protected by a barrier along the coast (Grøn 1995). Other investigations carried out in the ‘lagoon area’ at the mouth of the Tange Å watercourse, intending to establish whether it had functioned as a harbour or a shipyard, produced negative results. The numerous finds from the settlements of Lundeborg I and II were subjected to interdisciplinary analyses and prepared for publication under the direction of Per O. Thomsen, Svendborg & Omegns Museum. So far, a general summary has been published (Thomsen et al. 1993), as well as preliminary publications regarding the iron weapons (Ilkjær 1994), the animal figurines (Hardt 1994), the plant macrofossil remains (Robinson 1994), the animal bones (Hatting 1994) and the coins (Horsnæs 2010, 101 ff.).

External relations. The position of Gudme as a centre for metal production requires studies of the material and comparison with other places of production in Scandinavia and beyond. Part of this work was undertaken by Lars Jørgensen († 2016) and continued by others (see the next Gudme volume). Preliminary studies have been published (Thrane 1987; Jørgensen 1994), and examinations of specific finds types have been presented, such as coins (Kromann 1987; 1991; 1994; Horsnæs 2010, 94 ff.) and gold bracteates (Hauch 1987; 1994; Axboe 1987; 1994a). The status of Gudme as a political and religious centre in relation to other centres, including from a diachronic perspective, has been touched upon a number of times, from when the project began until recently (i.a. Thrane 1991; 1992; Jørgensen 1995a; 1995b; 2001; 2009; Axboe 1994b; 1999; 2012; Näsman 2011; Pesch 2011).

Communication of results. An international conference was held in Svendborg in 1991, at which the results that had been obtained up until then were summarised and some of the papers mentioned above were presented (Nielsen et al. 1994). But it was also intended to communicate the results to a wider audience. There was therefore an exhibition about the finds from Gudme and Lundeborg at respectively Fyns Oldtid, the National Museum and Svendborg and Omegns Museum between 1 May 1993 and 31 October 1994 (Thrane 1993). Queen Margrethe II opened the exhibition at Hollufgård near Odense on 1 May 1993, after which Her Majesty was taken to Gudme, where the great hall had just been uncovered (Sørensen 1994b). This fortunate and unexpected discovery was made as the Gudme-Lundeborg Project came to an end, paving the way for new discoveries that followed afterwards.

Fig. I.5. Map showing all excavations in the Gudme area.

Fig. I.6. Reconstruction of house 1, the Great Hall. Graphics, Ragnar L. Børsheim, Arkikon AS, Copyright, The Avaldsnes Royal Manor Project, Museum of Cultural History, University of Oslo.

Fig. I.7. Aerial view of Gudme towards Lundeborg by the Great Belt. Photo, Mogens Bo Henriksen, Odens Bys Museer.

The Gudme-Lundeborg Project was a successful collaboration between institutions and individuals with a shared interest in working with the promising sites in this area. The period of the project was followed by a number of rescue excavations, three of the most important of which are described in this book – the hall complex and the settlements of Gudme III and IV. There is still much more to be done. In Henrik Thrane’s first written survey (1987), 1,550 finds were recorded from the Gudme I and II areas. Today the database contains information about more than 6,000 finds from the Gudme area. This has huge potential, which has only just begun to be exploited.

Note

1    Members of the Gudme Committee (1991) were Berit Pauly Hansen, the Forest and Nature Agency, Poul Otto Nielsen, the National Museum (chairman), Klavs Randsborg, University of Copenhagen, Per O. Thomsen, Svendborg & Omegns Museum, Henrik Thrane, Fyns Oldtid, Hollufgård, and Mogens Ørsnes. Also contributing to the work of the committee, Ole Crumlin-Pedersen, Mogens Bo Henriksen, Henrik M. Jansen, Karsten Kjær Michaelsen and Peter Vang Petersen.

1

The Gudme halls

1.1. Introduction

In the middle of the 1980s, it was not only new archaeological discoveries that meant that the name Gudme had become familiar to the Danish population, but just as much the sporting results of the local handball club, GOG (Gudme-Oure-Gudbjerg). The women’s first team, in particular, were making a mark in the domestic handball league, at the same time as the Danish women’s national team were progressing at international level. This created great interest in handball throughout Denmark. It brought media attention to the small town in south-east Funen, and most people who were interested in sport soon found out where Gudme was located (Michaelsen & Sørensen 1994, fig. 17). But few people, however, were aware that, as early as the Late Iron Age, Gudme had also been the location for the most advanced centre within the geographical area that is now known as Denmark.

Fig. 1.1. Map showing the excavation areas in the sports arena area at Gudme.

By the beginning of the 1990s, the sports arena in Gudme was too small for the ever-growing numbers of handball spectators. A plan was therefore arrived at to expand the existing sports arena, and establish a new, larger arena. At this point, archaeologists from the National Museum and the two local museums in Odense and Svendborg had already been undertaking excavations in the area for almost a decade, and had obtained a detailed insight into the large Iron Age settlement at the location. The area that was chosen as the site of the new sports arena was a small hill, immediately south of the existing arena (Fig. 1.1). This area had been an overgrown fallow field for many years, and had therefore not been surveyed with metal detectors. Neither had this uncultivated area been investigated during the large-scale trial excavation campaign that was undertaken in 1992. The top of the small hill was around 61 m above sea level. The whole of the north-western part of the hill had previously been removed during the construction of the existing sports arena. There was therefore a steep slope down to the north-west, to a large asphalted parking area near the old sports arena. The most thoroughly investigated area was located immediately west of the sports arena and car park, where an excavation was undertaken under the road at the Stærkærvej/Poppelvej junction during the winter of 1992–93. Here, remains of a few small house sites were found, which should probably be dated to the Late Roman Iron Age (Fig. 1.1, Gudme III north-east, NM 7478/92). Towards the south-west, the terrain slopes steadily down towards the northern part of the large settlement area of Gudme II. This area is, however, divided from the other part of the Gudme II area by an extensive wetland, which today is fully cultivated, but which is shown on the oldest map of the area drawn in 1800 (Fig. 1.2, see also Chapter 2.4.1). Here, two connecting trial trenches were excavated during trial excavations in 1992, in which the remains of the end of an Iron Age house were found (Fig. 1.1). Towards the south-east, this wetland also extends into the area near the sports arena and forms a partial boundary with the north-eastern part of the Gudme II area (Fig. 1.2). During the trial excavations of this part of the Gudme II area in 1992, a single trial trench was excavated along the boundary with the sports arena area (Fig. 1.1). Here, a number of small postholes were found, which were most likely associated with the Iron Age farmstead that was subsequently recorded here (see phase 9 below).

Fig. 1.2. The topography and wetland areas in the Gudme halls area. Dark green: the size of wetland areas as shown on the cadastre of 1800. Light green: the actual extent of wetland areas as observed during the archaeological excavations.

At the beginning of 1993, the somewhat overgrown area where the new sports arena was to be constructed was partially cleared, and an initial, minimal metal detector survey was then undertaken. During this survey, a number of metal objects dating to the Iron Age were found, including eight Roman denarii. This demonstrated that the Iron Age settlement probably also extended into the area where the new arena was to be constructed. At the beginning of March 1993, a single trial trench was placed at the top of the hill. Here, a number of features were found that indicated very intensive settlement at the site. The features found in the trial trench later proved to be part of the large hall building, house I.

Fig. 1.3. The first photographic registration of the great hall (house I). Photo, National Museum of Denmark.

On the basis of the metal detector finds and the features that had been identified in the trial trench, it was decided that the area should be excavated before the construction of the sports arena could begin. The investigation was undertaken in the period 22 March–19 June 1993, in a collaboration between the National Museum and Svendborg & Omegns Museum, as a rescue excavation financed by funds allocated by the Keeper of National Antiquities (Rigsantikvaren).¹ The excavation adopted the method that had been developed during the previous excavations of the Gudme area (Sørensen 1995). In the excavated area, a number of features were found that resembled large pits, measuring over 1 m in diameter. The pits, however, formed two long rows that were partly surrounded by a number of small holes, which must have been postholes (Fig. 1.3).

A number of north–south-orientated ditches from an older ridged fields system covered the entire area (see Fig. 1.6). In order to achieve a better overview, it was necessary to undertake further cleaning up of the area, during which the fill in the ditches were removed. The picture gradually became clearer, and amongst the two rows of large holes, the postholes of a large hall building were identified, and this building measured almost 50 m in length and had eight sets of roof-bearing posts (Fig. 1.4). The building’s westernmost set of roof-bearing posts were found on the dug-away slope down to the car park, but the wall and the gable was not preserved at this end of the building (Fig. 1.4). South of the large hall, a large concentration of features was found, which was identified as the remains of six small, very unusual buildings.

As it was obvious that the area contained a number of very unusual buildings, it was decided after discussions with the developer, that the site of the new sports arena should be moved to the north-east of the existing sports arena. After this, the excavation changed character, and it was no longer an emergency excavation and was therefore not under the same time constraints as before. It was also decided to preserve as much as possible of the features that were excavated, so that at some point in the future, new excavations could be undertaken at the site.

During removal of the soil in the area where the sports arena was now to be erected, a large wetland was found, under a thick layer of clay, which was partly associated with the construction of the first sports arena. This wetland is not shown on the map of 1800 and therefore, by this time, must have already been partly filled up with material (Figs 1.2 and 1.6). No objects dating to the Iron Age were, however, retrieved from the wetland, which was partially filled with turf layers. As part of the investigations at the new construction site, a trial excavation was also undertaken of the area that was to be a new car park. It soon became apparent that this area also contained remains of Iron Age settlement.

Fig. 1.4. The great hall (house I) with excavated roof-bearing posts. Photo, National Museum of Denmark.

After the excavation ended in the spring of 1993, two of the excavated hall buildings were marked on the surface with rows of logs (Sørensen 1994b, fig. 13). These were subsequently replaced with more solid logs, and a viewing tower with information boards was constructed. During the 4 October–15 October 1993 period, a small area just to the north-east of the large Iron Age hall was excavated, and here the remains that were identified included house IX (see phase 8 below). During the period 5 April–6 May 1994, an excavation was carried out of all of the new car park area. This excavation identified the small hall building, house XV, and the small house XI, which was partly located within the excavation area from 1993, was also recognised (see phase 1 below). The investigation also indicated that there was an ordinary farm in the north-east, which had probably been constructed late in the Iron Age (see Fig. 1.51). In 2010, the logs marking the two hall buildings were replaced with small concrete pillars and several new information boards were erected at the site.

Fig. 1.5. Section through two roof-bearing posts of the great hall. Photo, National Museum of Denmark.

Fig. 1.6. Plan of the Gudme halls.

1.2. The Neolithic

The earliest remains of settlement at the site date to the Neolithic period. In the south-western corner of the excavated area was a small pit, feature 397, which contained the remains of two funnel beakers (Fig. 1.7). One vessel has a decoration consisting of vertical beads that run down over the girth and can be dated to Early Neolithic C (3500–3300 BC). At the bottom of the pit, the subsoil was red coloured and right next to this was a burnt red area (feature 395), which was probably the bottom of a fireplace. A few posts that were found near this may be the remains of a house dating to this period.

Fig. 1.7. Plan of pits from the Neolithic and Late Bronze Age.

1.3. Late Bronze Age

Scattered over the excavated area were a number of pits, which can be dated to the latter part of the Late Bronze Age (1000–500 BC) (Fig. 1.7). Most of these pits were concentrated in the area surrounding the hall buildings. Several of the pits were very large and most has only been partially excavated as a result of the prioritisation that was adopted during the excavation. North of house I, a quite large pit, feature 1, was found, and beside this, beneath cultural layer 10, was another pit, feature 128. Substantial quantities of pottery, which here are attributed to cultural layer 10, actually originated from this pit, the uppermost layer of which was excavated before the pit was identified. This pit contained a very large vessel with a high neck and a wide girth. The girth area of the vessel is rough, whilst the upper neck is completely smooth. In this area there is also an applied bead depicting a curved snake figure (Fig. 1.8). The south-eastern part of the large hall, house I, also contained a row of pits dating to the Late Bronze Age. A number of the holes for the wall posts and a few of the roof-bearing posts in house I were dug down into these pits (Fig. 1.7, features 46, 64 and 124). The large pit, feature 347, was placed in such a way that the walls of all the small hall buildings passed through it. The walls of the Iron Age houses could not, however, be seen on the surface at the top of the pit, and only part of the wall slot of house II could be recognised in section (Fig. 1.7). Neither could the wall slot in house II be recognised at the top or in the section of another Bronze Age pit (Fig. 1.7, feature 382). In the south-west was another large pit dating to the Late Bronze Age, feature 399, and some of the postholes found in this area may also date to this period.

Fig. 1.8. Large vessel from the Late Bronze Age with a depiction of a curved snake figure, GN 1076. Drawing, Henning Ørsnæs, National Museum of Denmark.

Apart from the pits already mentioned, a very special type of structure was also found, which can be dated to the Late Bronze Age. This consisted of two pits that were located close to one another. One of the pits was completely filled up with a compact layer of daub fragments with branch impressions, whilst the other pit contained ordinary settlement waste in the form of pottery, flint and charcoal. Three structures of this type were found, one just south-east of house I (Fig. 1.7, features 55 and 56) and two east of house II (Fig. 1.7, features 482 and 483 and features 488 and 489). The two closely placed pits, features 128 and 129, can perhaps also be included in this group. The numerous pieces of daub that were found in one of the pits amongst these features may originate from an oven lining, and the special features might therefore be remains of a freestanding oven (or even furnace) structure of unknown function.

1.4. The Iron Age settlement

1.4.1. Phase 1

After the Late Bronze Age activity, the area was uninhabited for around 800 years. It was not until the transition between the Early and Late Roman Iron Age in the middle of the 2nd century AD, that activity resumed once again at the site. The earliest phase of the Iron Age settlement consists of two small houses, located on the northern part of the small hill. There are also a few pits, which should also be dated to this phase on the basis of their ceramic content.

House XI

During the excavation of the large hall, house I, in 1993, part of a narrow ditch was found, which ran along the inner side of the north-eastern wall of the hall and parallel with it. The surface was cleaned up at both ends of the small ditch, but it was not possible to follow its course further under the ditches from the field system and cultural layer 10. At this point, therefore, the ditch could not be associated with any specific construction (Fig. 1.9). During the continued excavations in 1994, a further stretch of a ditch was identified just outside the old excavation area, north of the large hall (Fig. 1.9). This ditch appeared to constitute the end wall of a small house and could be followed into the old excavation area, which was re-excavated in a small section up to the marked posts from the small entrance in house I (Fig. 1.9). The two ditches or wall slots appear to be from a just over 5 m-wide house without roof-bearing posts (Fig. 1.10). The house is the same size and type as house III from phase 2 (see Fig. 1.18, see house III below). The western house end is not preserved, whilst it is possible that remains of the northern course of the wall slot can be found at the boundary between the two excavation areas from 1993 and 1994. The excavators looked for the northern part of the wall slot under the field boundaries that cut the eastern part of the house, but the slot was not preserved in this area. If the house was completely identical to house III, it should have been possible to find two entrance postholes to the building under the field boundary, but this would have required re-excavation of the 1993 excavation area. House XI probably represents the earliest phase of the settlement at the Gudme halls, and even though no objects were recovered that could be associated with the house, based upon the dating of the other phases, it should be placed in the very early part of the Late Roman Iron Age in the middle of the 2nd century AD.

Fig. 1.9. Plan of house XI and pit A10.

Fig. 1.10. House XI.

House VIII

In the north-westernmost part of the excavation area, where the area had been partially dug away, was a small house, consisting of two sets of roof-bearing posts and a single post from a third set, whose second post had disappeared when part of the hill was removed during the construction of the first sports arena (Fig. 1.11). It cannot therefore be definitely established whether the house had further sets of posts further to the west. The very small roof-bearing posts, however, suggest a small house that probably only had three sets of posts. The house is located across the north-western part of the wall of house I, although there is no stratigraphy between the two houses, they cannot have been in use at the same time (Fig. 1.11). The transverse spans between the two sets of roof-bearing posts are 2.5 m, whilst the longitudinal spans from the east are 3.9 m and 4.75 m (Fig. 1.12). The house was originally interpreted as belonging to the later farm settlement at the site, but the pottery from the postholes of the house and several ¹⁴C dates indicate that it dates to the earliest Iron Age phase at the location.

Fig. 1.11. Plan of house VIII.

Fig. 1.12. House VIII.

Feature 76, one of house VIII’s roof-bearing posts, contained sherds from at least four pottery vessels.² GN 459, several sherds from a small vessel, the upper part of which is decorated with a zone of diagonal lines, which above, at the transition between the neck and the rest of the body, are bounded by horizontal lines (see Pl. 1.1). Cf. Gudme halls feature 426 GN 280 (see below), Møllegårdsmarken burial 1964 (Alb. IV, table 116), per. I.,