The Archaeological Landscape of Bute George Geddes and Alex Hale The Archaeological Landscape of Bute George Geddes and Alex Hale The Archaeological Landscape of Bute George Geddes and Alex Hale British Library cataloguing in Publication Data: A CIP catalogue record for this book is available on request from the British Library. Crown Copyright © RCAHMS 2010 ISBN: 978 1 902419 74 9 Book layout by Mitch Cosgrove The maps are based on Ordnance Survey material with the permission of the Ordnance Survey on behalf of Her Majesty’s Stationery Office © Crown Copyright. All rights reserved , OS Licence Number OS 1000 25406. Unauthorised reproduction infringes Crown Copyright and may lead to prosecution or civil proceedings. Except where otherwise stated , all the illustrations and images in this publication are Crown Copyright RCAHMS. All rights reserved. No part of this publication may be reproduced, stored in or introduced into a network retrieval system, or transmitted, in any form, or by any means (electronic, mechanical, photocopying, recording or otherwise) other than for initial download without the prior written permission of RCAHMS. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution. Front cover: An oblique aerial view of the vitrified fort of Dunagoil (at right, 40291) and the fort of Little Dunagoil (at left, 40280). See page 22. DP062659 Front cover inset: St Colmac cross, standing stone with later inscribed cross (40317). See page 15. SC408104 Back cover: A detailed plan from 1780 of Dunagoil vitrified fort (40921) and its surrounding landscape. See page 4. DP075180 © Bute Archive at Mount Stuart Page iv: A view from Glencallum bay at the SE corner of Bute towards the Dunagoil archaeological complex in the distance. Much of the dramatic intervening landscape was cultivated until the 18th century, when the south of the island became a sheep farm. DP062664 Published in 2010 by The Royal Commission on the Ancient and Historical Monuments of Scotland. The Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS) John Sinclair House 16 Bernard Terrace Edinburgh EH8 9NX tel 0131 662 1456 fax 0131 247 4163 www.rcahms.gov.uk Registered Charity SC010240 For over 100 years, the Royal Commission on the Ancient and Historical Monuments of Scotland has been collecting, recording and interpreting information on the archaeological, architectural, industrial and maritime heritage of the nation, creating a unique archive that offers a remarkable insight into the special nature of Scotland’s places. Many millions of items, including photographs, maps, drawings and documents are made widely available to the public via the web, through exhibitions and publications, and at the RCAHMS search room in Edinburgh. Commissioners of the Royal Commission on the Ancient and Historical Monuments of Scotland, 2010 Professor John R Hume OBE BSc ARCST Hon FRIAS FSA Scot (Chairman) Dr Gordon G T Masterton OBE BA MSc PhD DIC CEng FICE FIStructE FIES FRSE MCIWEM FConsE (Vice Chairman) Dr Kate Byrne MA MSc Phd CEng MBCS CITP FRSA Mr Tom Dawson FSA Scot Mr Mark Hopton MBA Dip Arch BArch RIBA ARIAS FSA Scot Dr Jeremy Huggett BA Phd MIFA FSA Scot Professor John Hunter BA PhD MIFA FSA FSA Scot Mr Paul Jardine BA MBA ACMA Ms Jude Quartson-Mochrie BA (Hons) BArch (Hons) Ms Elspeth Reid MA MA DAA Secretary, Chief Executive, Mrs Diana Murray MA FSA FSA Scot MIFA iii iv Contents Foreword vi Site References and More Information vii Introduction 1 The Mesolithic and Neolithic 7 The Early and Middle Bronze Age 13 The Late Bronze Age and Iron Age 21 The Early Historic and Medieval Periods 29 The Pre-Improvement Period 37 The Improvement Period 41 Conclusion 47 Further Reading 48 Index 51 v Foreword The partnership between the Discover Bute Landscape Partnership Scheme (DBLPS) and the Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS), from which this volume emanates, has been a truly rewarding example of successful partnership working between Scottish Government agencies and local people. With the assistance of funding from the Heritage Lottery Fund , the long wished for reassessment of the archaeological record of Bute and the enhancement of that record into a modern, easily accessible resource has become a reality. As significant, the symbiotic knowledge transfer between the RCAHMS Survey and Recording staff and the local Bute residents has been a process that has been educational, inspirational and , most importantly, a whole lot of fun! We are delighted that this volume will stand as a long term legacy for the project, and one that will encourage involvement in, learning about and enthusiasm for the archaeology of Bute for generations to come. Paul Duffy Deputy Project Manager (Archaeology) Discover Bute Landscape Partnership Scheme September 2010 Acknowledgements This booklet was written by G Geddes and A Hale and edited by J Sherriff, JB Stevenson, R Bailey and R Turner. The maps and the survey drawings have been prepared by G Brown and I Parker and the aerial photographs were taken by R Adam and D Cowley and other members of the RCAHMS aerial survey team. Additional photography was provided by D Smart, S Wallace and by RCAHMS field investigators. In addition to those noted above, fieldwork was undertaken principally by P Dixon, A Dutton, A Gannon, L Fisher, J Hepher, A Leith, L McCafferty, P McKeague and M Middleton. Thanks also to O Brookes, M Sutherland and W Toole for production, proof reading and indexing. The Commission is grateful to many individuals who provided assistance throughout the project. Those to be thanked at the Discover Bute Landscape Partnership Scheme (DBLPS) include B Paterson (Scheme Co-ordinator), and the members of the Archaeology Research Project, including P Duffy. Of the members of the local community who participated in the fieldwork (too many to individually mention here), special thanks are extended to A Hannah and J Herriot. Thanks also to S Hothersall from the Association of Certificated Field Archaeologists (ACFA), S MacDonald (Administrator), I McArthur (Buteshire Natural History Society), A Speirs (Curator, Bute Museum), A McLean, archivist at Mount Stuart, and J Turner, who facilitated access to Inchmarnock. For providing other specialist advice, thanks are due to T Cowie and A Sheridan (both National Museums of Scotland) and N Finlay (Dept of Archaeology, University of Glasgow). RCAHMS is also especially grateful to all the landowners, including the Marquess of Bute, Lord Attenborough and Sir Robert Smith, and to all the tenants who willingly allowed access to the sites. vi Site References and More Information Sites and monuments are consistently labelled throughout this publication with their RCAHMS number. This is a unique identifier that will allow the reader to make the connection between sites which are mentioned in the main text or in captions, and sites which are highlighted on maps. The number will also enable readers to obtain further information about a particular site, including bibliographic and archive details, by accessing the RCAHMS sites and monuments database Canmore. This can be found at rcahms.gov.uk. Within the advanced search page, the reader should enter the number in the box marked ‘Canmore ID’ and click on the search button. Readers can also discover more information about sites and monuments on Bute by searching the ScotlandsPlaces website: scotlandsplaces.gov.uk. This offers an alternative way of accessing the site-records held within the RCAHMS database as well as other source material. Each illustration in this publication has a caption which includes, when appropriate, the name of the site, its main classification and its number. Further, a code number at the end of each caption identifies the digital file held in the RCAHMS archive and this should be cited if the illustration is required for reproduction. Plan Conventions vii viii Introduction Introduction This booklet is a result of a partnership project undertaken in 2008–10 between the Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS) and the Discover Bute Landscape Partnership Scheme (DBLPS), the co-ordinators of a four-year programme of improvements to Bute’s rural landscape funded largely, but not solely, by the Heritage Lottery Fund. The short-term aims of the project were to revise the existing RCAHMS archaeological records for Bute, working closely with the local community, and to produce a booklet that summarises the archaeology of Bute. The principal long-term aim is to provide the local community with the information that will allow them to determine priorities and make decisions about the archaeological work they are likely to initiate in the future. This volume contains a commentary on the character of the archaeology of the island , one that has been informed principally by the evidence contained within the revised archaeological record. Although there has never been a systematic archaeological survey of Bute a great deal of work has been undertaken. In recent times the locations of several early prehistoric sites have been identified through the discovery of surface scatters of stone artefacts; documentary research has provided clues regarding pre-Improvement period landscapes; fieldwork has resulted in many previously unrecorded upstanding monuments being reported; and , since 1977, aerial survey has yielded evidence The Isle of Bute, showing Rothesay, other settlements and geographical features. GV004571 RCAHMS staff and local volunteers discuss the archaeological landscape of Scalpsie Bay, where prehistoric forts, burial cairns, an improvement farmstead and what may be a medieval building lie in close proximity. DP084694 of a broad range of cropmarks. Nevertheless, comparatively large areas of the island still have secrets to reveal, and further work in all these disciplines, plus others, such as geophysical survey and excavation, will undoubtedly reveal more evidence in the years to come. The revision of the Bute archaeology records has made some use of documentary evidence. First, second and third editions of the Ordnance Survey 6-inch-to-a-mile map, respectively published in 1 The Archaeological Landscape of Bute 1869, 1897 and 1924, have been referred to, but two 18th century estate surveys, now held within the Bute Archive at Mount Stuart, have been particularly useful. One, by John Foulis in about 1758–9, is roughly contemporary with Roy’s Military Map of 1747–55, but it provides a more accurate and detailed record. The other, compiled about 1780 –2 by Peter May, the Bute Estate factor, records precise details of the land-use within the estate, including the location of farmsteads, smallholdings and single buildings, with notes on acreages and tenancies. It is supported by a ledger that contains an astonishing amount of detail. Extract from the 3rd edition of the 6-inch Ordnance Survey map showing Dunagoil (Buteshire 1924, Sheet CCCXXVII). This depiction of the landscape, which was clearly influenced by Dr J N Marshall’s contemporary survey and excavation, includes a number of archaeological sites that had not been previously shown, some of which have subsequently been lost or reclassified. In contrast, there are only two sites depicted on the earliest edition of the survey of Bute (published in 1869) where their interpretation has been subsequently doubted – forts at Achamor Wood (40393) and Suidhe Chatain (40286). SC1208177 2 Scope of survey The partnership agreement between RCAHMS and DBLPS required the archaeological site records for Bute to be thoroughly revised , with the necessary fieldwork providing opportunities for local volunteers to become involved. The evaluation of the record started as a desk-based exercise late in 2008, and a programme of field survey, which was undertaken in the spring and summer of 2009, incorporated the informal Introduction training of members of the local community in archaeological survey techniques. The overall contribution made by the volunteers was invaluable, not only saving RCAHMS staff a considerable amount of time in an unfamiliar landscape by identifying the locations of sites that would have otherwise proved difficult to find , but in making RCAHMS aware of a significant number of sites that had not previously been reported. Although prospective fieldwork – proactively searching for previously unreported monuments – did not constitute a formal part of the project, the simple act of visiting known sites with local people who had already undertaken a considerable amount of fieldwork resulted in about 120 ‘new’ sites being recorded. This was in addition to the 525 previously recorded Bute archaeological records in the RCAHMS database that were revised and the 260 or so other records that have been created through the accessioning of recent work. One aspect of the existing RCAHMS database that the revision did not cover was the broad range of artefacts that have been found; these form a significant proportion of the archaeological record. However, this topic is currently being addressed by another project within the Landscape Partnership Scheme. Although the project specification stipulated the revision of the archaeological records for Bute, it should be recognised that the distinction between archaeology and architecture, especially when ruined buildings are concerned , can be a fine one. A number of relatively recently abandoned farmsteads were visited – not because of any intrinsic architectural merit (though this may be present), but because map or documentary evidence suggested that earlier remains might have been evident. As in the case of those working farmsteadings that were either visited on the ground or photographed during aerial survey, a note about the place was recorded for the RCAHMS database. Extract from the Survey of the Isle of Bute undertaken in 1758–9 by John Foulis. His survey, which is accompanied by brief texts, recorded both the position and orientation of each building, and the size and land-use of each farm. DP077260 © Bute Archive at Mount Stuart John Patrick Crichton-Stuart, the 3rd Marquess of Bute (1847–1900). Along with his predecessors and descendants, the 3rd Marquess sponsored a number of archaeological projects and funded the restoration of buildings that included St Blane’s church. © Mount Stuart Archive Previous work The present archaeological record owes much to the Earls of Bute, who, since the late 18th century, have either initiated work or readily given permission for others to do so. Perhaps the earliest noted archaeological investigation on Bute was that undertaken at Dunagoil, planned in 1780, where several cairns (40070) were ‘excavated’ soon after by John Norton, chief gardener at Mount Stuart, under the direction of John Blain. The 19th century saw several small excavations being documented and probably a lot more not, and culminated with the publication in the 1890s of The Isle of Bute in the Olden Time by James Hewison, Church of Scotland minister of Rothesay parish. Although Hewison’s volumes reproduced a lot of the rather vague information that appeared in earlier publications, including the Statistical Account, they provide an excellent summary of the archaeology of Bute at that time, some of their illustrations containing information that is now lost. The leading figure in Bute archaeology in the early part of the 20th century was Dr John Marshall, a founder member of the Buteshire Natural History Society (BNHS) and of the Bute Museum. Dr Marshall excavated at Dunagoil from 3 The Archaeological Landscape of Bute A detailed plan and sketch section of Dunagoil vitrified fort (40291) and its surrounding landscape. The plan, dated 1780, and probably produced by either Peter or Alexander May, depicts and describes archaeological features at Dunagoil. It is a remarkable symbol of the early interest and investment of the 3rd Earl and 1st Marquess of Bute in aspects of the island’s heritage. DP075180 © Bute Archive at Mount Stuart 1913, first in the cave (40278) and then in the fort (40291), and undertook what would be called today a watching brief at Townhead gravel quarry (40377), where he recorded details of a Neolithic settlement site. Dr Marshall’s contribution to Bute archaeology was eventually eclipsed by his youngest daughter, Dorothy, who studied under Sir Mortimer Wheeler at the Institute of Archaeology in London and , after her mother’s 4 death in 1949, gained considerable experience working on excavations both at home and abroad. Much of what is now understood about the archaeology of Bute is a consequence of work that she instigated , work that has been warmly appraised by Jack Scott (1992). Dorothy Marshall’s legacy, however, is not just the records of the important excavations she undertook or her reorganisation of Bute Museum, it is also the enthusiasm she engendered among the local community. That passion has persisted since her death in 1992 and to this day others have continued to investigate aspects of Bute’s heritage through research, excavation and survey. Initially encouraged by Dorothy Marshall, and more recently led by Edwina Proudfoot, members of the Society have been recording rural Introduction Hawk’s Nib, cave and midden (40675). A group of volunteers accompanied by RCAHMS staff ascend to explore this cave site. Excavators in the 1930s discovered midden material and evidence for hearths, but no categorical dating evidence. DP073220 Excavations at Eilean Buidhe island dun (40458) in August 1936. One of a number of important archive photographs held by Bute Museum, this image shows Dr J N Marshall (1860 – 1945) in the trench flanked by his daughter Dorothy (1900 –1992) and J Harrison Maxwell, who directed the excavations. The photograph was taken from the south-west showing one of the trenches cut through the rampart; the building in the background is the boat house at Caol Ruadh School. SC1210134 © Bute Museum settlement remains for many years. Since 1991, the Bute Settlement Survey project has systematically added to the corpus of information through deskbased research, walk-over surveys and detailed recording. As a result, the local community has made and continues to make a very valuable contribution to Scottish rural settlement studies. At the start of the revision of the RCAHMS records for Bute a large proportion (about 200 out of 525) of the archaeological records related to settlement sites, mainly farmsteads, which had been reported by members of the Society. The abundant legacy outlined here has been built upon by the partnership project between RCAHMS and DBLPS, which has seen such a significant involvement by the local community, and by the RCAHMS Scotland’s Rural Past project. The revision of the archaeological records will provide a solid platform on which to base future research, and the fieldwork undertaken by the local community in tandem with RCAHMS in 2009–10 will better equip them to undertake that work. The authors of this volume hope that it may help in directing and encouraging their efforts. 5 6 The Mesolithic and Neolithic c12,000BC – c2200BC Mesolithic c12,000BC – c4500BC Evidence from Howburn Farm in South Lanarkshire has recently demonstrated that Scotland was visited as early as 14,000 years ago. To date, no evidence for such early settlement has been found on Bute, where, as far as is known, the first visitors were itinerant Mesolithic foragers (fisher-hunter-gatherers) who populated the west coast of Scotland between the end of the last ice age (about 11,000 years ago) and the beginning of the Neolithic period. Three locations on Bute, all close to the south end of the island , have so far yielded Mesolithic artefacts. In the early 1970s an assemblage of agate tools and waste material was discovered on Plan Farm (40287). Most of the material is of Neolithic or later date, but the collection includes two earlier very small worked blades – microliths. A larger assemblage of about 200 flints (40645) from a raised beach at the north end of Kilchattan Bay also contained microliths as well as late Mesolithic cores and waste. Again, most of this material is of later date. In 2003 fieldwalking on Garrochty Farm, just 1km south-west of the source of the agateworking site, resulted in the discovery of another flint and pitchstone assemblage (268254) with a small Mesolithic component, this time including two platform cores and a core rejuvenation blade. Distribution of Mesolithic and Neolithic sites and artefacts. This map probably reflects to a large degree an original bias in the overall distribution of rock carvings and chambered tombs in the northern third of the island. However, that distribution has certainly been affected by other factors, including intensive agriculture since the medieval period, which may have resulted in the removal of even the most substantial monuments. GV004572 A shell-midden below the chambered cairn at Glecknabae (40371) was partly examined when the tomb was excavated at the beginning of the 20th century. No artefacts were found at that time, but a suggestion that the midden was Mesolithic has not necessarily been disproved by the identification of the flint core found more recently in the eroding edge of the midden as probably Neolithic. The small number of Mesolithic sites so far identified on Bute precludes any in-depth comment on their distribution. However, there is considerable potential for identifying further sites of this date on the island , as fieldwalking exercises in other areas of Scotland have demonstrated that Mesolithic material will invariably be found on prolonged and extensive searches of cultivated ground. The fact that the fieldwalking at Black Park Plantation (123472, 123473) at the south end of Bute yielded only material of Neolithic and Bronze Age date need only be a reflection of the comparatively small area examined. Although arable fields provide the most obvious targets for fieldwalking, other locations on Bute ought to be considered, and work elsewhere in Scotland has shown the potential of examining moorland landscapes. However, the best opportunity for locating relatively undisturbed Mesolithic evidence on Bute and Inchmarnock may be in the numerous caves and rock-shelters that have been identified. Several of these were looked at by Dorothy Marshall, who found no evidence of human activity in most of them, the one exception being at Hawk’s Nib (40675) at the south end of Bute, which contained an undated shell-midden. Evidence for the more recent use of caves and 7 The Archaeological Landscape of Bute 8 The Mesolithic and Neolithic Bicker’s Houses, chambered cairn (40381). Sketch of the chambered long cairn viewed from the south by J K Hewison (1893, 65). In addition to the main burial chamber at the north (far) end of the cairn, this sketch shows what may have been a cist that is no longer visible inserted in the southern part of the mound. The ‘fort’ in the background is probably Barone Hill (40424). SC1163929 rock shelters has been found close by at Dunagoil (40278, 40279) and on Inchmarnock (300171, 300178) see page 20. Neolithic c4500BC – c2200BC Our understanding of how and where the first groups of farmers on Bute lived and worked is hampered by the dearth of identified and accurately dated settlement sites. Consequently, the information gained from the distribution of Kilchattan Bay. Excavations at the quarry in the distance uncovered evidence for a Neolithic settlement (212266). The nearby hamlet once included a thriving brick and tile works (158613) and a parish church (81886), while prehistoric burial cairns stood sentinel on the bay’s edge (40639). The pattern of rectangular fields, established in the Improvement period, rolls over the edge of the old shoreline cliffs. DP057287 funerary, ceremonial and ritual sites, rock art (which cannot easily be included in any of those categories) and a thin scatter of diagnostic artefacts, paints an incomplete picture of events that were spread over at least two millennia. Between 1914 and 1919, and again in 1929, the working of a gravel-pit at Townhead (40377) on the southern edge of Rothesay revealed a range of artefacts, including a polished stone axe and pottery, which indicated the presence of some form of Neolithic settlement. However, little detail is known of this settlement and there is scant evidence of any structure that might have existed here in association with the artefacts and what was probably a hearth. A number of shallow trenches that were revealed in the surface of the gravel were interpreted as probable ‘sleeper beams’, but they are more likely to have been field drains or the bottoms of cultivation furrows. A radiocarbon date for a burnt shell of a hazelnut from the probable hearth produced a date around the middle of the 4th millennium BC. More recently, excavation at another gravel quarry – at Kingarth (212266) towards the south end of Bute – revealed what may be a round-ended 9 The Archaeological Landscape of Bute Michael’s Grave, chambered cairn (39905). The position of even the most significant monuments has been carefully checked with survey grade GPS. This particular cairn was excavated by T H Bryce in 1903, and the burial chamber now sits proud of the surrounding field, divorced from its protective round cairn which has been heavily robbed and largely ploughed out. DP082856 Neolithic building. This structure has been compared to broadly similar buildings elsewhere, particularly in Argyll and Ireland , but it yielded no absolute dating evidence and could not be shown to have had any direct association with any of the other excavated features or with the scatter of flint, quartz and pitchstone artefacts recovered in the vicinity. If the paucity of identified habitation sites limits discussion about the character of Neolithic settlement on Bute, then the situation is significantly improved with regard to the funerary, ritual and ceremonial sites, and , of course, the rock art. There are six chambered cairns on Bute, five in the northern third of the island and one, Bicker’s Houses (40381), in the central part, some 4km south-west of Rothesay. A chambered cairn is a large mound of stones that contains a stone burial chamber which could be opened whenever the 10 remains of the dead needed to be interred. All of the known chambered cairns on Bute have either been ‘explored’ by antiquaries or excavated in more recent times, and they have been thoroughly discussed elsewhere. Audrey Henshall’s work makes it clear that each cairn probably had a long and complex history, in some cases extending into the Early Bronze Age. Three of the chambered cairns, Glecknabae (40371), Hilton (40306) and Michael’s Grave (39905), appear to have been built as round cairns and remained that way throughout their use. In its final form, Glenvoidean (39897) is a trapezoidal cairn that covers two earlier round cairns, and Bicker’s Houses and Carnbaan (40370) are long mounds that may well have developed from earlier round cairns. There is only one radiocarbon assessment for any of the Bute tombs, a burnt deposit from beneath one of the side-slabs of the chamber at Glenvoidean producing a date around the middle of the 4th millennium BC. The bias in the distribution of the tombs in the northern third of the island probably reflects their original range even allowing for the undocumented destruction of some tombs and the possibility that previously unrecorded examples will be identified The Mesolithic and Neolithic in the future. The most recently recognised chambered cairn is situated at Hilton, overlooking Kames Bay. First reported in the early 1970s, it exhibited no evidence of its Neolithic origin until it was excavated. Similar monuments could easily have been swept away in the tide of agricultural improvements; others may yet be recognised. There is also the possibility that other forms of Neolithic burial may also be present on Bute. For instance, there may be tombs in which the mortuary chambers were constructed in timber rather than stone. If this is the case, their identification can only be confirmed by excavation. In this context, mention should be made of a number of apparently manmade mounds that have been examined without yielding any evidence of burial. Investigation of two such mounds (40639), which formerly stood about 40m apart at Kerrytonlia on the north side of Kilchattan Bay, established that each comprised a boulder cairn under a cover of beach sand. A thin layer of charcoal was detected under the cairn within the western mound , but no artefacts were recovered. Whether these cairns represented an episode of early prehistory that the excavators simply failed to recognise will now never be known since both mounds have been destroyed. The most ubiquitous, yet least understood , manifestation of Neolithic settlement is rock art – the carving of symbols on bedrock and on stones ranging in size from small portable boulders to large glacial erratics. There are over 80 recorded examples of rock art on Bute, all of them comprising either simple cups or cups within rings. What stands out about this corpus is that it contains only simple carvings. None is elaborate or complex, and this contrasts markedly with neighbouring Argyll. Although widespread throughout Bute, the distribution of the rock art appears to favour the west side of the island and the main concentration is Scarrel, cup-marked rock (40326). Typical of the rock art on Bute, this example is located in improved pasture, and bears a number of cup marks, highlighted by recent rain. The island of Inchmarnock is visible in the background. DP066517 found in the north. This distribution is probably in part influenced by the likelihood that this category of monument will survive (and be detected) in unimproved land , but it may not be a coincidence that this is also where most of the chambered cairns are found. On this note, none of the Neolithic monuments should be considered in isolation, with each forming an integral part of the wider landscape. The pottery and artefacts that were deposited with the dead in the chambered cairns were the same wares and tools that were being used on the settlements. 11 40070 12 The Early and Middle Bronze Age c 2200BC – c1000BC Three general features that mark the transition from the Neolithic to the Early Bronze Age are the introduction of bronze metalwork, the adoption of different ceramic forms, and the move away from the use of megalithic chambered cairns that were essentially charnel houses used over a long period of time. New rites included the inhumation and cremation of single and multiple burials in pits, cists and pottery vessels, often under specially constructed mounds (cairns and barrows), but also sometimes within earlier monuments such as the chambered cairns, or under conspicuous natural knolls. No doubt some burials were interred without anything to mark the grave and there may well have been other methods of disposing of the dead that have left no trace in the archaeological record. It ought to be emphasised that no Bronze Age houses have been positively identified on Bute and that any discussion of the pattern of settlement is dependent on what is known about other contemporary monuments, burials and artefacts. Excluding for the moment those features directly associated with funerary practices, other monuments within the Bronze Age landscape of Bute include stone circles, stone settings and standing stones. The stone circle at St Colmac Cottages (40338) – the only surviving stone circle on the island – originally comprised nine stones, of which four still stand erect, three survive as stumps and an outlier stands close-by to the southThe distribution of Early Bronze Age sites and artefacts demonstrates that activity was widespread across the whole of Bute at this time. However, the main foci of monuments such as standing stones and stone circles appear to have been in the two low-lying and narrowest points of the island. GV004573 St Colmac Cottages, stone circle (40338). This stone circle is one of two that once stood in what is now a cultivated field. The other, which was situated 25m to the west, is depicted on an early 19th century plan of Kames Estate (RHP 14262), but it was removed before 1814, some of the stones having to be blown up by gunpowder first. This circle survives because by that time it had already been enclosed within a small plantation. GV004721 west. The circle appears to have stood adjacent to another about 25m to the west, of which nothing is now visible. However, it was described by Aiton 13 The Archaeological Landscape of Bute St Colmac Cottages, stone circle. A number of monuments in the valley between Kames and Ettrick Bay can be seen to represent an early prehistoric ritual landscape. The most significant of these, this circle of eight stones, was set within a protective plantation bank around 1800, though very few of the trees now survive. DP064299 in 1816 as ‘a circle of stones, of great size and height; but they are all thrown down and most of them removed. Some of them were so large as to require to be blown with powder’. A late 18th or early 19th century estate plan appears to show an arc of five stones on the west side of the current circle. One of the most interesting features of the site is its position within the local topography. Overlooking Ettrick Bay to the south-west, the surviving stone circle stands within a natural amphitheatre in which there are also three standing stones (40316, 40317, 40360), one of them (40317) bearing a later early Christian cross. Also, the stone circle is overlooked by what is likely to be the roughly contemporary Watch Hill 14 burial cairn (40442) at the south-east end of the bay and a possible cist cemetery (40071, 239544) above Kildavannan at the north-west end. Even if the Bronze Age ritual landscape here has been reduced to a series of individual monuments isolated from each other within modern farmland , the range and density of those sites provides enough evidence to suggest that this location was somehow special, apparently more so than anywhere else on Bute. Further, these monuments appear to represent a major change in focus in ceremony, ritual and burial activity away from what are today the marginal lands in the northern third of the island favoured by the Neolithic tomb builders, to a much more fertile farming landscape only a relatively short distance to the south-east. An earthwork, Cnoc an Rath (40312), which stands on the watershed on the north-east edge of the amphitheatre, has been proposed as a henge. If this is the case then the monument may have served as the main focus for ritual, ceremony and funerary practices throughout the Late The Early and Middle Bronze Age St Colmac cross, standing stone with later inscribed cross (40317). Standing near the site of Colmac’s chapel (40349), now removed, this stone may have originally formed part of a Bronze Age ritual landscape. Recorded in the 1990s as part of a long running RCAHMS survey of medieval carved stones, it forms part of an important corpus from Bute, much of which is now housed in Bute Museum. SC408104 Cnoc an Rath, earthwork (40312). This site, situated on the watershed between the bays of Ettrick and Kames, is defined by a ditch and an inner bank, with a causewayed entrance on the eastsouth-east. First surveyed by RCAHMS in 1982, the date and function of the monument remains elusive. SC1208314 Neolithic and Early Bronze Age. However, there is no dating evidence from the site, which others have suggested is an Iron Age fort, a medieval ringwork, and , most recently, a Norse ‘ting’ site. Elsewhere on the island there is a scatter of standing stones and stone settings which are presumed to be of Early Bronze Age date. Of the latter, the three-stone alignment at Largizean (40257) at the south end of the island is similar to many other examples elsewhere in Britain, and it may be significant that the stones stand only about 350m south-east of the find spot of a hoard of bronze halberds found in the 19th century (see below). Another stone setting, at Blackpark Plantation (40258), stands just 850m east of the hoard site. First referred to as ‘the remains of one of the circles of the Druids’, the evidence provided by various descriptions of the site is at best ambiguous, and there is little compelling evidence to show whether the stones belong to a circle that once comprised more than the three stones that are visible today or some other form of setting. There must also be some doubt about a setting of two stones at St Ninian’s Bay (40426). Measuring only up to 1.4m in height and 1.3m apart, the stones are slighter than might be expected of a Bronze Age monument. 15 The Archaeological Landscape of Bute 16 The Early and Middle Bronze Age James Skene’s sketch and plan of the ‘druidical monument’ at Blackpark (40258) captures the unusual character of the individual stones. The original sketch, dating to 1818, is part of a collection from the the Society of Antiquaries of Scotland, gifted originally by James’ son. We are also fortunate to hold a larger version, reproduced here, redrawn in a sketchbook and probably dating to around 1832. DP084695 There are at least 30 Early Bronze Age burial sites on Bute that have produced evidence of either single burials or cemeteries. There are also seven cairns, the characteristics of which suggest they are probably also of this date. The whole corpus of Bronze Age burials cannot be reported here in detail, but a few sites deserve particular mention. At four sites graves have been inserted into earlier, Neolithic, cairns. At Glecknabae (40371) one of the two chambers contained sherds of a Beaker, possibly indicating re-use of the tomb in the Late Neolithic/Early Bronze Age. Later still, a cist containing an inhumation was inserted into the west edge of the cairn. A vessel that was described as ‘looking like a flowerpot’, which was found and then lost before excavations in the early 20th century, was probably another secondary burial. At Glenvoidean (39897), an Enlarged Food Vessel and a cremation were found in an unusual twotiered cist, and at Hilton (40306) Marshall found an apparently empty cist that had been sunk into the body of the chambered cairn but which had , in turn, been covered by its own discrete mound. At Bicker’s Houses (40381), the possible cist inserted into the mound close to its south end , which was noted as ‘rifled’ by Hewison, can no longer be seen. Fine spacer-plate necklaces of jet or jetlike material have been found accompanying inhumations in cists at Kerrycroy (40722) and at Northpark on Inchmarnock (40427). As high prestige artefacts with a broad distribution across northern and western Britain, they provide striking evidence of how rich some sections of the Early Bronze Age population on Bute were. The Kerrycroy example was found along with a Food Vessel and a corroded bronze object in 1887; the Northpark necklace, which comprised 139 beads Scalpsie, barrow (40232). Centred on the Bronze Age barrow, this view captures the character of Bute’s Improvement landscape. The regular fields, the timber plantations, and the well spaced farmsteads are all products of this period, and the water from the distant hill slopes once helped to drive Rothesay’s mills. DP083916 Recently relocated, the Bronze Age cist at Mount Stuart was first opened in 1887 (40722). Careful, given the standards of the time, the excavations yielded a crouched burial, a Food Vessel, a bronze awl, and a jet necklace, all of which were donated to the National Museum in 1888. © Trustees of the National Museums of Scotland. Licensor www.scran.ac.uk. 000-100-035-704-R recycled from at least four other necklaces, was found in one of three cists discovered in 1960 and originally excavated by Marshall. One of the other cists yielded an old and worn spacer plate from another jet necklace. Recent research has seen the re-excavation of the cist that contained the necklace. A sample of bone yielded a radiocarbon date around 2000BC, in the middle of the Early Bronze Age. Bone isotope analysis, undertaken as part of the Beaker People Project, revealed that the bones were those of a woman who was almost certainly local and who did not eat fish – a surprise considering that she was buried (and therefore presumably lived) on a comparatively small island. In addition to the Northpark cist, only one other, at Cnoc-na-mhanan (239544), has been scientifically dated. Here, one of what may be a cemetery of cists contained a miniature Food Vessel and a cremation that has been dated to between 2000 and 1700BC. Another cemetery, which comprised no less than 17 The Archaeological Landscape of Bute eighteen cists and was found beneath a cairn at Ardbeg (40315), provides a striking contrast to the single burials or those in small groups. It was dug into before 1840 and again in 1858, but nothing now survives either of the cairn itself or of the several ‘sepulchral urns’ that were found in the graves. Edwina Proudfoot has recently reviewed the evidence for Early Bronze Age burials on Bute and , notwithstanding those sites (mainly noted 18 by antiquaries) from which the evidence is either dubious or, at best, ambiguous, the picture presented suggests that probably the whole of the island (and Inchmarnock) was exploited by a farming population whose material culture was no different to that on the mainland to the east and Argyll to the north and west. This distribution is heavily influenced by the discovery of burials mainly through the agencies of arable cultivation, quarrying and construction, but it should be recognised that The Early and Middle Bronze Age A group of three cists were found in 1960 by a farmer at Northpark Inchmarnock (40427). One of the cists contained the remains of a single woman, accompanied by a flint knife and a spacer plate necklace of 139 jet beads, re-strung and photographed by Alison Sheridan, NMS. Re-excavation in 2006 formed part of a significant re-assessment of the discovery by by Bute Museum and NMS Archaeology Department. © Trustees of the National Museums of Scotland. weapons, the precise function of those of Early Bronze Age date is not known. The circumstances surrounding the deposition of the hoard are also unknown, but the find-spot is located in an area of poor natural drainage and it is possible that the halberds were a votive offering within an originally wetland location. activities such as these occur in the very areas that are likely to have most attracted early farming populations. In mitigation, though, an additional pointer to the use of the landscape at this time is provided by those Bronze Age burials in Neolithic cairns that are situated on or just outwith the limit of recent intense arable cultivation. A full consideration of the Bronze Age pottery from Bute is beyond the scope of this booklet and commentary is restricted here to noting that there is a paucity of Beaker pottery – especially outwith the chambered tombs, there is only one confirmed discovery of a cinerary urn (40237), but Food Vessels are comparatively common, with no fewer than eight examples being found. In addition to the badly corroded metal object found in the Kerrycroy cist noted above, there has been only one recorded discovery of Early Bronze Age metalwork on Bute – that of five halberds (40255) found in the 19th century when the farmer at Langalbuinoch was digging a drain. A halberd comprises a long, axe-like, blade attached at right-angles to a wooden or metal shaft. While medieval versions were designed and used as Five Bronze Age halberds (40255) were found in the mid 19th century by a farmer digging a drain on Langalbuinoch farm. Three survive in the National Museum of Scotland, where they have been dated stylistically to between 2300BC and 2000BC. © Trustees of the National Museums of Scotland. Licensor www.scran.ac.uk. 000-190-004-241-R 19 20 The Late Bronze Age and Iron Age c1000BC – cAD600 If the Early Bronze Age on Bute is categorised by a lack of settlement sites in favour of ceremonial, ritual and funerary monuments and the Middle Bronze Age is marked by a total dearth of identified monuments of any description and no artefacts, then the Late Bronze Age (c1000BC – 700BC) at least offers some tangible evidence for settlement. Any of the prehistoric round-houses or burnt mounds so far recorded could date from this time and there is also evidence for metalworking in this period. Prehistoric round-houses take two forms on Bute: hut-circles, where the building is preserved as a low, roughly circular stony bank, and platforms, where all that is now visible of the structure is the artificial terrace, dug into a hillside, on which it was built. Perhaps the best example of one of these platforms is that built within the earthworks at Dun Scalpsie (40254). It is impossible to date a prehistoric round-house from visible evidence alone and any of them could date to a period spanning over a thousand years. None of the ten examples so far recorded on Bute has been excavated. A burnt mound is a large pile of fire-shattered stones, charcoal and soil that is thought to have accumulated as a result of an open-air cooking process. Three burnt mounds have been recorded on Bute and there is one possible example on Inchmarnock. The burnt mound at Mecknock Distribution of sites and artefacts of the Late Bronze Age and the Iron Age. Many of the settlements identified here could have been built and/or occupied within a period lasting up to 1,800 years. The bias in the distribution of forts and duns on the west coast may reflect contemporary politics, though the small fort at Kerrytonlia (40643), visible today as a cropmark, indicates that similar sites existed on the east coast. GV004574 Barr Hill, hut-circle and cultivation remains (124782). This hut circle, with its thick stone wall and east-facing entrance, is typical of the small prehistoric round-houses found all over Scotland. The cultivation adjacent to it is probably medieval or later in date. GV004729 (40267), up to 10m across and 1.4m in height, has been excavated but has not been dated. Elsewhere in Britain and Ireland , while similar mounds have generally been found to date to the Middle and Late Bronze Age, others have been found to be Early Historic or even medieval in date. The Late Bronze Age and the first part of the Iron Age – the centuries spanning the end of the 2nd millennium BC and the beginning of the 1st millennium BC – was a period of great transition across Scotland. It was a time of social upheaval, which saw the widespread abandonment of permanent occupation in upland locations affected by increased rainfall, cooler temperatures and the 21 The Archaeological Landscape of Bute An oblique aerial view of the vitrified fort of Dunagoil (at centre right, 40291) and the fort of Little Dunagoil (at centre left, 40280), taken from the north-west. There is no evidence of fortification on the intervening ridge, but the landscape between the two forts is dotted with the remains of burial cairns, clearance cairns and cultivation remains. DP062659 formation of peat. The resulting pressure on the most fertile and productive land is thought to be one of the reasons why forts and other forms of defensive enclosure began to be constructed. However, whilst there are a comparatively large number of sites that can fairly confidently be assigned to the 1st millennium BC, the settlement pattern on Bute at this time is not at all well understood. None of these sites has produced good dating evidence from reliable contexts and it is likely that the picture is complicated by several sites being re-used in the Early Historic or medieval periods. The vitrified fort on Dunagoil (40291), at the south end of Bute, and the fort on the summit of Barone Hill (40424), overlooking Rothesay, are the 22 only two large forts on the island. Barone Hill is undated , but the partial excavation of Dunagoil nearly a hundred years ago recovered artefacts dating from the Neolithic to the Iron Age, and included evidence for bronze working. A vitrified fort is one that had a high and thick stone wall, internally braced by timbers, which has been burnt to such a high temperature that large portions of its rubble core have been fused into a solid mass. As a technique, the use of timber-lacing cannot be used to date forts, but those which have been dated by other means have been shown to lie within a long period spanning the first millennium BC and the first millennium AD. The excavation account of Dunagoil is vague about the exact nature of the metalworking, but given the close proximity of the site to Little Dunagoil (40280), where socketed axes were being produced in the Late Bronze Age, it seems fair to infer that it, too, was probably occupied in some form at that time. Most of the artefacts found during the excavation of Dunagoil are undiagnostic The Late Bronze Age and Iron Age Map of the archaeological landscape at Dunagoil, including: (A) the vitrified fort (40291); (B) a probable round-house platform; (C) the lengths of wall that partly define the enclosure north-east of and below the fort; (D) the fort on Little Dunagoil (40280); (E) the later buildings and building-platforms (90299), and burial cairns (40070). GV004726 in terms of date, but they include a brooch (dating to the 3rd century BC) that probably provides some indication of when the actual fort was occupied. The results of the excavation at Little Dunagoil fort were inconclusive. Dunagoil and its environs, including the fort on Little Dunagoil, have recently been reconsidered by D Harding of Edinburgh University. His conclusion – that the excavations and artefacts demonstrate that the Dunagoil ‘complex’ attracted settlement from the Neolithic to the Early Historic periods is sound enough, but he has also radically reinterpreted the fort and its relationship to the immediate topography and other features within the landscape. Foremost in this reinterpretation is his proposal that adjacent 23 The Archaeological Landscape of Bute to and contemporary with the timber-laced fort was what he calls a ‘bailey’. This comprised a terrace immediately north-east of and below the fort that was defined largely by topographical features, Little Kilchattan, cropmark enclosure (40644). The transcription of oblique aerial photography has allowed cropmarks across Bute to be accurately mapped and their features measured. In the case of this enclosure, which was first recorded in 1977, its purpose is unknown, but its 4m broad ditch suggests that it may have been for something more significant than stock control. SC1203299 24 which offered excellent natural protection, linked by lengths of timber-laced wall (C on the plan on page 23). The claim that this enclosure was contemporary with the summit fort is based on an observation that there is in situ vitrified bedrock visible in one of these lengths of wall at the northwest end. The presence of vitrified material, wall core rather than bedrock, was confirmed in the present survey. However, the function and date of this wall has yet to be established , and it is difficult to see The Late Bronze Age and Iron Age how a ‘bailey’ may have worked: there is a gap of some 100m between the south-east ‘bailey’ wall and the summit fort, a gap that would have allowed relatively unimpeded access to the lower enclosure from the south. Whether these large forts were built as tribal centres, community gathering places or simply for the purpose of making ostentatious political statements may never be known, but, whatever the reason, large forts probably served a very different purpose to smaller sites, which were most likely built by small farming communities as a response to some perceived threat. The majority of these sites on Bute are relatively small, thick-walled enclosures that have variously been classified as forts and duns. However, the distinction between the two is sometimes a fine one and the classification used by RCAHMS in its recent revision of the Bute archaeological records is based on a definition of forts and duns used in the first Argyll Inventory. This simply states that the internal area of forts was large enough to serve the needs of a small community while duns were capable of accommodating only a single family. This fairly loose definition is almost certainly an over-simplification, but using it means that the enclosures at Kerrytonlia (40643), Little Dunagoil (40280), Clachan Ard (40269) and probably Ardnahoe (40242) – the smallest fort with an internal area measuring 25m by 23m – have been added to the list of forts. The enclosure at Balilone (40378), with an internal area of only 24m by 19m, is on the cusp between fort and dun. Previously described as a fort, the enclosure at Aultmore Burn (39906), which has an internal area of only 19m by 16m, has been reclassified as a dun and takes its place with four others – Dunstrone (40276), Dun Scalpsie (40254), Dun Burgidale (40300) and Castle Cree (40441). The last mentioned site, however, is unusually narrow and rectangular and it may actually be medieval. Few of the forts or duns display any notable architectural features, despite several of them having been investigated by excavation. Exceptions are Dun Scalpsie, where a door-check is still visible within the entrance passage, and Dun Burgidale, where there are almost certainly passages between the walls. As far as the distribution of the forts and duns is concerned , there appears to be a definite The cropmark fort at Kerrytonlia (40643) is one of a number of archaeological sites that were first noted during an aerial sortie undertaken by RCAHMS on 7 July 1977. SC1203301 emphasis on the west coast, though some allowance should be made for sites that have either been destroyed or have yet to be recognised. Late 18th and 19th century local historians made unsupported references to all manner of archaeological monuments, including forts which had already been destroyed. In part, that lost record has been retrieved by aerial survey, which, since 1977, has revealed the cropmarks of a number of sites, in particular broad-ditched enclosures that may represent the earthwork equivalent of the stone-built sites. One of these sites is at Kerrytonlia (40643), on the north side of Kilchattan Bay. Here, a promontory extending from the old sea-cliff has been defended by two broad arcing ditches drawn across its neck, the inner one accompanied by a narrow palisade trench along its inner edge. Whether this last feature was an integral part of a system of defence with the ditches or represents an independent period of enclosure is not known. Its presence, however, is a useful reminder that, irrespective of how well preserved a site may appear on the surface, more ephemeral features will seldom be seen and the visible remains will only tell a fraction of the full story of any particular site. That some sites enjoyed extended periods of use has already been noted at Dunagoil, but 25 The Archaeological Landscape of Bute the same is also true at Little Dunagoil (40280) and probably also at Dun Scalpsie (40254). Little Dunagoil is noted in the next chapter, but Dun Scalpsie, which occupies the summit of a rocky knoll on the north side of Scalpsie Bay, is of considerable interest because it is one of the few sites on Bute where more than one phase of construction can be demonstrated from the visible evidence alone. Until recently, the site was thought to comprise simply a dun with a narrow entrance in its north-west side (A on plan). However, survey by RCAHMS in 2010 has demonstrated the presence of a very substantial stone wall (B) lower down the slope on the south and east. This wall follows the Dun Scalpsie was partly excavated in the 1950s (40254). This plan shows the dun (A) that sits on the summit; the wall of the possible fort (B) below; the timber round-house (C), and the foxhole type shelters (D) that were probably constructed by local defence volunteers (LDV), later the Home Guard, during the Second World War. GV004725 26 crest of a steep slope on the north-east and appears to underlie the wall of the dun on the north. On the west flank there is little trace of the wall, but it may have been destroyed by a ditch that was dug to protect the later dun. The visible evidence is ambiguous and it is not possible to state whether the outer wall represents a fort that occupied the rocky knoll before the dun was placed on the summit or is a contemporary outwork of the dun, providing additional defence on its vulnerable flanks. The results of a trench dug inside the dun in 1959, which revealed two periods of occupation, is not helpful in this respect. Also unhelpful is the disturbance to the summit area caused by the construction during the Second World War of several dug-outs (D). Of particular interest at Dun Scalpsie, however, is the recognition of a stance for a timber roundhouse (C) immediately outside the wall of the dun and within the area of the outer enclosure. The exact context of this building, which is one of only a small number of timber round-houses recorded on the island , is not known. It may be contemporary with either the earliest or later enclosures on the knoll; equally it may represent a period of settlement that has no association with either phase of enclosure. Something similar may be said of what is probably the stance for a round-house on the southwest-facing flank of Dunagoil, some 50m south of the fort (B on plan on page 23). Measuring little more than 5m in overall diameter, the undated structure sits on a very steep, craggy area of ground characterised by short lengths of wall, some incorporating blocks of vitrified stone, which define a series of discrete enclosures. Other forms of settlement that deserve mention are crannogs, and caves or rock-shelters. A crannog is an artificial island built to support a building, usually of timber. There are three crannogs on Bute: one in Dhu Loch (40380) and two in Loch Quien (40233, 40234), all undated. The crannog in Dhu Loch was discovered in 1812, when the water level was low enough for a plan to be drawn of the remains, which included horizontal timber beams and numerous piles. Subsequently the water level in the loch was raised and the site now lies completely submerged. The two crannogs in Loch Quien, which are visible today as wooded islets close to the shore, were both surveyed in 2003. The Late Bronze Age and Iron Age Dhu Loch, crannog (40380). First surveyed in the early 19th century, this crannog was subsequently completely submerged when the loch became a reservoir for Rothesay’s water supply. There are three crannogs on Bute, and a survey in 2003 resulted in the discovery of a rotary quern stone at Quien north (40234), now at Bute Museum. SC1163938 Caves and rock-shelters (shallow cave-like openings at the base of a cliff, often with a wall built across the mouth), have been used in a variety of ways by humans from earliest prehistory to recent times. Two rock-shelters at the south end of Inchmarnock (300171, 300178) have been shown to contain evidence of some form of use in the second half of the 1st millennium BC. Dorothy Marshall considered the cave in the cliff below the fort on Dunagoil (40278) to have been used in the Iron Age as well as the Neolithic. As noted above, several caves and rock-shelters contain no obvious evidence for any human activity, but only a relatively small number on Bute have been examined. With more sites of this type yet to be recorded , caves and rock-shelters may represent one of the best opportunities for recording in situ later prehistoric deposits. 27 28 The Early Historic and Medieval Periods cAD600 – cAD1600 For Bute, it is convenient to consider the Iron Age as ending at the start of the Early Historic period , which for the purposes of this text has been assumed to commence about the beginning of the 7th century AD and end around the 10th century. Vikings are only briefly referred to below, and actual Norse settlement in the 11th to 13th centuries is considered to be medieval, a period accepted here as running to the end of the 16th century. In the Early Historic period Bute found itself sandwiched between the power-bases of Dal Riata in Argyll to the west and Strathclyde to the east. In what must have been extremely turbulent times, a dynamic landscape was first disturbed by the Viking incursions in the 7th and 8th centuries and later, during the 12th century in particular, became heavily influenced by the Kingdom of Man and the Isles. Finally, it was included in the Stewart kingdom. Currently, our level of knowledge about the archaeology of this time is limited by a paucity of securely dated sites and a comparatively modest amount of documentary evidence. Consequently, almost nothing is known of the settlement pattern at this period and the little knowledge available is necessarily biased towards three major sites – the monastic settlements at ‘Cenn Garah’ (Kingarth; 40292), at the south end of Bute, and that dedicated to St Ernan on Inchmarnock (40268), both probably established around the beginning of the 7th century AD, and at Little Dunagoil (40280), where Dorothy Marshall’s excavations in the late 1950s and early Distribution of Early Historic and Medieval sites and artefacts. As with the preceding periods, the record is dominated by high status religious and secular establishments. GV004575 St Blane’s Church (40292). An oblique aerial view of the 12th century church and graveyard, taken from the north-west. The upper and lower burial-grounds, both to the right of the church, are reputed to have been reserved for men and women respectively. What may be a medieval chapel is visible in the lower graveyard. DP066101 1960s yielded artefacts indicating activity in the second half of the 1st millennium AD. The evidence from the excavation of the monastery on Inchmarnock has yielded a wealth of information, detailed in an excellent book (Lowe 2008), but little is known about contemporary secular settlement there. At Kingarth we know the names of two of the 7th century bishops but, despite the late 19th century exploration there at the request of the 3rd Marquess of Bute, we know very little about that establishment’s relationship with either its lay neighbours or the wider landscape. The documentary evidence for the Early Historic period in the upper Firth of Clyde implies that Bute contained a flourishing population, the spiritual 29 The Archaeological Landscape of Bute St Blane’s Church (40292). This early photograph, taken around 1895, of the medieval church shows the nave prior to its restoration, which was undertaken by the architect RW Schulz, and sponsored by the 3rd Marquess of Bute. The photograph was presented to the National Buildings Record (now part of RCAHMS) by Mr A Curtis, whose given address at 14 Grays Inn Square, London, was the former office of Schultz. SC1161664 needs of which were serviced through a network of chapels. The existence of these chapels is not documented , although their presence can surely be assumed to coincide with the occurrence of cross-slabs at Kilmachalmaig (St Colmac), where the carved stone (40317) still stands; at St Mary’s Church, Rothesay (40373); and at an unidentified site near Rothesay Castle (301961), where a broken slab (now in Bute Museum) was found in the early 19th century. The MacAlister Stone (40406), also Rothesay Castle, crossslab (301961). This rectangular cross-slab of sandstone was found in Rothesay Castle in 1816, re-used in two fragments at the foot of the stair in the northwest tower. It remained there until some time after 1903 when it was moved to Bute Museum. SC403493 30 now in Bute Museum, probably also came from a chapel, the site of which is not known. Apart from St Ninian’s Chapel (40437), which has been excavated , there is little evidence that any of the other surviving chapels date earlier than about the 12th century AD. The assumption must be that there was a relatively thriving population that could sustain the infrastructure implied by the presence of a network of religious establishments. Elsewhere in the north and west of Scotland there is evidence that some duns were reoccupied in the Early Historic period and in this context we can note that at least one Norse place-name, Dunburgidale, may be linked to a specific structure of earlier date. However, the majority of Bute place-names are later, many being derived from Scottish Gaelic, and the best evidence for a continuity of settlement is provided by the fort on Little Dunagoil (40280) and its immediate environs. Rothesay High Kirk, cross (40406). This cross-shaft, known as ‘MacAlister’s Stone’, was re-erected in 1886 after lying recumbent for many years on a burial-plot belonging to a family of that name. Tradition associated it with the Ascog area, about 2km south-east of Rothesay, but there were conflicting accounts of its earlier provenance. Hewison tentatively identified it with a stone which formerly stood at Kilwhinleck (NS c.0580 6230). The stone was removed to Bute Museum in 1996. SC403494 The Early Historic and Medieval Periods The fort (40280) and farmstead (90299) on Little Dunagoil, showing the location of: (A) the summit fort; (B) the later addition to its east side; (C) the additional wall on the north side; (D) the cluster of subrectangular building platforms, and (E, F) the buildings (farmstead) partly excavated by Dorothy Marshall around 1960. GV004724 Excavations demonstrated that the fort had a long history stretching from prehistory through into the Early Historic period. However, while the excavations demonstrated the depth of chronology of the site in general, it failed to reach an adequate understanding of the defences and the structural remains within the fort. The resurvey of the site in 1994, part of a reassessment of the Dunagoil complex, had a fresh look at the earthworks without radically altering the previous interpretation of the remains. The most recent survey, undertaken by RCAHMS in 2010, shows that the small thickwalled enclosure on the summit of the rocky knoll (A on plan) was added to on the east by another enclosure (B) that effectively doubled the size of the defendable area. It was probably at this time that a short length of additional wall (C) was added on the north. Perhaps more important is the recognition of a group of subrectangular building platforms (D), which lie below the east end of the fort. These platforms are very different in character from the large 12th/13th century buildings (E, F) previously investigated and are demonstrably earlier than those buildings (90299). The significance of these platforms cannot be overstated as they represent part of the Early Historic occupation of the hill that probably dates to a period after the fort went out of use but before the 12th or 13th century buildings were constructed. This small group of platforms therefore represents the most tangible link between the monastic settlement of Kingarth, if it survived into the 9th and 10th centuries, and its nearest secular neighbour. That link may have continued into the late Norse period in the form of the two large halllike structures which are quite unlike any other buildings so far recorded on Bute. 31 The Archaeological Landscape of Bute The hall-like buildings at Little Dunagoil excepted , the only structures of the medieval period (13th to 16th centuries) that are readily identifiable in the landscape are churches, chapels and castles. The churches and chapels are predominantly found on the west coast, perhaps an indication of the direction from which the inspiration for their foundation derived. At Kingarth, a parish church (St Blane’s) succeeded the monastic settlement (40292) and there was also a parish church (St Mary’s) in Rothesay (40373). Inchmarnock may have had medieval parish status, but it is more likely that Rothesay and Kingarth continued as the only parishes until the creation of North Bute in the 19th century. St Michael’s Chapel and Burial Ground (39902). This aerial view taken from the north-east shows the ruinous medieval chapel set within its circular burial-ground. DP067130 Kilchousland Chapel (40436), which lies within an enclosure formed by a thick bank on the south-west and the face of a relic shoreline cliff on the north-east. The original entrance to the enclosure, which may have served as a burial ground, was probably situated in the broad gap in the wall on the south. GV004723 St Blane’s, cross no. 10 (301996). Free-standing cross of white sandstone, re-used as a grave slab south of the nave of the church. SC403036 32 The Early Historic and Medieval Periods In addition to the medieval parish churches there was a network of chapels, one in Rothesay Castle (301959) and a series of others – St Michael’s (39902), St Ninian’s (40437), Kilchousland (40436), Kilmory (40260) and East St Colmac (40349) – dotted at fairly regular intervals along the western seaboard. A chapel may also have existed at Chapelton (40388), where an alleged font was Rothesay Castle (40395) is the most powerful symbol of Bute’s medieval period. Sitting in a prominent location that dominates the town, it is surrounded by a moat that was redesigned as a garden in the late 18th century. SC800098 once seen, and the possible locations of others, including Acholter (40314), Cnoc-na-Mhanan (40369), Kilwhinleck (302265), Kilbride (40372) and Kilmory Hill (294600), are hinted at either in tradition or in place-name. Medieval castles existed at Rothesay (40395), Kames (40309) and possibly also at Kelspoke (40674), Ascog (40724) and Wester Kames (40298). Buildings associated with the Stewart dynasty include Meikle Kilmory (40374), where there is evidence for what may have been the house of the ‘Sheriff Crowner’ of Bute in the 16th century. In contrast to the majority of the churches and chapels, 33 The Archaeological Landscape of Bute Kames Castle (40309). This tower, one of only a handful of medieval strongholds on Bute, is surrounded by a suite of 19th century buildings, and set within extensive later gardens. DP010303 these secular foci are usually found on the east side of the island , and this may be an indication that by that time the people of Bute were focusing more on the Scottish mainland than anywhere else. Other than castles, towers and religious buildings, evidence for rural settlement on Bute in the medieval period is mainly restricted to the mention of farms and estates in documents. With the exception of two excavated kilns (see below), no rural farm buildings earlier than 18th century in date have been identified on Bute, though recent excavation at Quien (88671) may have uncovered a small medieval building and enclosure. It is clear, however, from the detailed histories published in the 19th century that the majority of the modern farms on Bute have roots in land-holdings that were already in existence in the late medieval period. Exchequer rolls from 1450 include references to many names that are still in use today, but it is difficult to tie these references to 34 particular locations and impossible to link them to specific structures. Nevertheless, there are several locations where there are groups of rectangular buildings, albeit surviving only as grassed-over wall-footings, which are not depicted on detailed estate maps of the mid to late 18th century, the inference being that they were already abandoned by that time and as such could belong to a much earlier period. The problem here, though, is a 13th century building that has been reduced to footings will probably look very similar to one dating to the 15th, 16th or 17th century. These potentially early sites have been recognised in the north of the island in particular – including on the lower flanks of Fly Hill (124772, 124773, 302308) and Edinbeg Hill (124763) and around Split Craig Hill (139140, 294622) in the comparatively remote north-west corner. One of them (124793), which is tucked into a corner of a pasture field well above Stuck farmsteading, comprises two buildings beside a stream. It is often difficult to establish from field evidence alone what the function of buildings that have been reduced to footings has been. Here, however, it is possible to state with some confidence that the building which The Early Historic and Medieval Periods that the building which has a drain running out of it is a byre and that the other structure was a dwelling. That medieval buildings can and do survive within a heavily modified landscape has been demonstrated at St Blane’s (302028) and more recently on Inchmarnock (300170), where corndrying kilns have been excavated. However, medieval kilns take much the sam