Castle Wood at Caerlaverock

Caerlaverock Castle with moat and wood

Caerlaverock Castle with moat and wood Copyright C Mills 2020 IMG_0063.JPG

The Guardian invited Coralie to write about a favourite old wood this week, and while there are many amazing old Scottish woods that I know well, there are fewer that have ready visitor access while retaining an atmosphere of intense wildness, and Caerlaverock Castle Wood was the one that sprung to mind immediately. Dendrochronicle is an admirer of this particular wood through having been invited by HES to undertake a detailed study of the wood to inform their understanding of the two castles’ relationship with the treescape, and to inform new visitor interpretation developments.

Shrew pollard

Shrew pollard- so named as we found a perfect but dead shrew on it – with Hamish Darrah: C Mills copyright 2020

A walkover historic woodland assessment of Castle Wood at Caerlaverock was undertaken for HES’ Cultural Resources Team in March 2018, complemented by historic map research. Castle Wood encompasses both the ‘old’ and ‘new’ castle remains and most of their surrounding earthworks, and forms a sub-rectangular block about one kilometre wide. A second phase of work was undertaken in autumn 2019, this time for HES’ interpretation team, to inform new visitor information, and allowed us to pursue some aspects of this fascinating treescape’s history and archaeology in more detail.   As yet, the only publication of this work is a brief entry in the Discovery & Excavation in Scotland annual round up of archaeological projects which is published by our friends at Archaeology Scotland.

Caerlaverock HWA survey

Caerlaverock -Dendrochronicle HWA mapping drawn by David Connolly – Copyright C Mills 2020

LiDAR survey at Caerlaverock by HES revealed a complex palimpsest of old enclosures and field boundaries beneath the tree canopy, and helped us to sub-divide the wood into compartments for the purposes of the assessment. The objective of the historic woodland investigative work was to enhance the understanding of the Property In Care’s relationship with the surrounding wooded landscape and to unravel the story of the evolution of the trees and woods themselves. The assessment embraced the living biocultural heritage, the landscape archaeology and their relationship with the built heritage.

It is as yet uncertain as to who built the ‘old’ castle around 1220; it was abandoned around 1270 when the ‘new’ castle was built, almost certainly by a Maxwell, first noted as lord of Caerlaverock in 1307. The area is believed to have been wooded when the castles were built, at least locally, based on pollen evidence from the old castle ditch sediments. These construction dates are pretty exact, and distinguishable from each other, because they were provided by dendrochronology – by dating remains of oak timber moat bridges which had survived by waterlogging. That work was undertaken by our friends at the Belfast QUB dendro lab a long time ago, in the 1970s and 1980s, when the excavations were ongoing and before Scotland had its own resident dendrochronologists.

Caerlaverock - wet woods

Caerlaverock – Peter Quelch in the wet and wild woods – Copyright C Mills 2020

So there is pollen evidence for some local scrubby wood around the castles just after the old castle was built but, crucially, as yet there is no pollen work from sediments which accumulated before the old castle was built so we do not know how wooded the environment was before the castles were built. It is Coralie’s view that it was not continuously wooded until very recent times, that it was probably quite an open landscape in the 13th century, with scrubby wood only along burnsides and in the areas that were too wet for agriculture. The oak timber used to build the castle moat bridges was probably brought in from elsewhere, as oak would have needed drier growing conditions than those naturally occurring around the castle sites.

The first castle is beside an old harbour, thought to be contemporary with it, and showing how the shoreline has gradually shifted several hundred metres further south since then. Coastal change is a big part of the complex story of Caerlaverock’s landscape change – and the changes are visible in the archaeological remains, especially as revealed by the Lidar survey, as well as in the natural landscape features.

It is thought that instability and subsidence of the old castle, on its very wet site, led to the decision to build the new triangular castle, only some 50 years later, which still survives in great shape today and is an iconic building that everyone should visit. This remarkable castle has a long complex history which I won’t go into here but is certainly rewarding to read about.  The ‘new’ castle continued to be developed into the early 17th century until a siege of 1640 rendered it a partial ruin.

Caer wetness and decay

Caerlaverock wetness and decay Copyright C Mills 2020 IMG_1843.JPG

Based on the Historic Woodland Assessment evidence, our working hypothesis is that small patches of early planted woodland survive at Caerlaverock, often on old boundary features, almost all oak, and it is possible that some trees area as old as the castles themselves. Without any dendrochronological work on the trees however, their age is informed conjecture, based on historic maps and tree form evidence, and many of the oldest candidates survive as rotted stumps only. This medieval wood-bank style landscape is overlain by a much wider patchwork of  oak plantings, mostly subsequently coppiced, on an extended rectangular enclosure footprint which first appears on General Roy’s map of c1750. However, we believe that it goes back much earlier, and was probably created in the early 1600s. Dendrochronology of some of the rarer old single stem oaks, including pollards, from this phase would allow us to establish dating more securely and even reconstruct pollarding dates and cycles.

Caer west bank oak

Caerlaverock west bank oak coppice Copyright CM 2020

These bank systems, medieval and early post-medieval, studded with old oaks, are interspersed with semi-natural wet woodlands and even some modern forestry plantations within the old enclosures. The historic planting of oak followed the drier upstanding old boundaries while alder-willow dominated wet woodland persisted in the lower wetter areas in between. Medieval ditches and later era drains were always part of the enclosure system, as drainage was crucial to being able to manage this wet landscape. There is plentiful evidence of continued economic investment in this wood after occupation of the new castle ceased in 1640. The majority of oaks were coppiced, the last cut probably being in the early 19th century based on approximate ring counts of a couple of fallen stems.

Caerlaverock Crinoline oak

Caerlaverock Old Castle Bailey bank – skirted oak may be one of the oldest trees – Copyright C Mills 2020

Despite the still-functioning improvement drainage system, much of the woodland is very wet today, making for a rich biodiverse ecosystem and also likely to promote good organic survival of archaeological remains. While coppiced oak is most common, the rarer pollard and maiden forms of oak on some of the boundaries are likely to be rather older, for example on the bank of the bailey to the east of the old castle.  A number of skirted old oaks indicate historic grazing pressure, and 18th century mapping, the earliest to show any detail, shows a wood with a number of open meadow areas within. However, we found no obvious evidence of there being a deer park here. The form of a substantial curved dyke just west of the old castle, on which many oak coppice stools have grown, was more suggestive of a former sea wall than a park dyke.  The sequence of landscape development could be teased out more certainly with further targeted dendrochronological, archaeological and documentary research assisted by the LiDAR and historic map evidence.

Much more detailed reports have been provided to HES and will in due course appear distilled into new visitor information provision, and, we hope, as a publication. In the meantime we recommend exploring Castle Wood yourself, as soon as Covid-19 regulations allow. It is an incredibly rewarding place to visit for those who like their natural heritage as much as their cultural heritage. Visitor information can be found on the Historic Environment Scotland web-site.

Posted in Dendrochronicle news | Tagged , , , , | Comments Off on Castle Wood at Caerlaverock

Exploring the science of dendrochronology


Delighted to have co-authored this new learning resource on Dendrochronology, recently published by Forestry and Land Scotland, and developed as part of Dendrochronicle’s ‘SESOD’ research programme. It serves as an accessible introduction to the archaeological science of tree-ring dating, using real Scottish archaeology case studies, and with fun activities for use in and out of school. The creative vision by Matt Ritchie (FLS archaeologist) shines through, the illustrations he commissioned are wonderful. My co-authors Marcia Cook and Jennifer Thoms did a great job on the activities and case studies respectively. Truly a team effort, with support from FLS, Archaeology Scotland, HES and of course Dendrochronicle.

The resource is free to download from here:

Posted in Dendrochronicle news | Comments Off on Exploring the science of dendrochronology

The dendrochronology of St Giles Kirk tower, Edinburgh

By Coralie Mills & Hamish Darrah

St Giles tower and spire: Photo Copyright Peter Backhouse

Few people are aware of the massive five-storey timber frame hidden within the tower under the iconic crown spire of St Giles Kirk in the heart of Edinburgh’s old town. Analysis of the tree-ring patterns, the science of dendrochronology, has revealed the early date and unexpected source of these timbers. This was undertaken as part of Dendrochronicle’s SESOD (SE Scotland Oak Dendrochronology) research project to develop the regional native oak tree-ring record. We have sampled timbers in several historic buildings across Edinburgh, the Lothians and the Borders, including St Giles. The SESOD project is ongoing and results will be fully published in due course.

Felling dates and St Giles Kirk building history

SESOD St Giles sampling: Photo copyright Coralie Mills

Core samples were taken from the timber frame before lockdown. Two felling dates were established from analysis of the cores, in the winters of AD 1453/54 and 1459/60. This gives some very specific dating for the frame which was previously of unknown date.

The timber frame is part of the modification of an earlier tower at St Giles of which some masonry survives, known to have been present by 1387 and on which storks nested in 1416 according to the Scotichronicon, a medieval history of Scotland. This may be the last recorded instance of storks nesting in Britain, topical this year with the first successful hatching of white stork chicks at the Knepp Estate in southern England.

The new dendro-dates show this tower was greatly altered in the mid-15th century at a time when many other structural changes were being made at St Giles. The new dates infer the completion of the crown spire as being after 1460 and probably by 1467 when the church was granted collegiate status from the pope, although the crown spire we see today was repaired and much altered in 1653.

The surprising source of the timber

St Giles bell frame: Photo copyright Coralie Mills

Even more exciting though was the timber source, revealed through comparison of the St Giles tree-ring data with reference chronologies of known origin. The timbers were brought from one of the last remaining extensive medieval reserves of old growth oak in Scotland, the Royal Forest of Darnaway, in Morayshire, including several from trees over 300 years old when felled. The survival of old growth oak at Darnaway into the mid-15th century is a testament to careful management of the forest by the Earls of Moray over a long period. The St Giles tree-ring data matched extremely closely with data from the late 14th century roof of Randolph’s Hall at Darnaway Castle.

By the mid-fifteenth century most of Scotland’s timber supply had switched from the dwindling native resources to Scandinavia. The majority of dendro-dated historic buildings in Edinburgh and the Lothians have imported Scandinavian timber in them, so this discovery is a rare example of use of scarce, good quality native oak here at this time. Timber would have been transported by sea to Edinburgh, probably through the port of Leith.

Darnaway Forest is known to have supplied timber to other high status Scottish medieval building projects, including Stirling Castle, but this is the most recent example so far detected through dendrochronology and it was not known that Darnaway had supplied St Giles. Shortly after these timbers were felled Darnaway Forest was closed to allow its recovery.

Royal patrons of St Giles Kirk

Mary of Guelders and James II – from the Forman Armorial of 1562

Amongst the patrons of the extensive mid-15th century construction work at St Giles are James II and, following his untimely death in 1460, his widow Mary of Guelders, who was regent of Scotland between 1460 and 1463 for their son, the infant James III. Their arms appear on shields on The King’s Pillar in the choir of St Giles, interpreted as a tribute to the late king by his devout queen.

The Forest of Darnaway was forfeited to the crown after the defeat of Archibald, the Black Douglas Earl of Moray, at the Battle of Arkinholm in 1455, which may explain why timber from Darnaway was used to build the bell-frame in the tower at St Giles, through this royal patronage.

Although the timber was felled when James II was still alive, its supply to St Giles may have occurred when Mary of Guelders was regent, between 1460-1463, after James’ death in the summer of 1460. This question could be investigated further with more sampling as the precise felling dates are from the lower levels of the structure.

Further evidence which supports the ascertained construction date of the tower are the bells from the tower itself. The oldest surviving ‘Ave’ or ‘Vesper’ bell dates to 1452 and the ‘Great Bell’, which was melted down and re-cast in the 19th century, originally carried the date of 1460. The ‘Great Bell’ was cast in Flanders and bore the royal coat of arms of Geulderland which is possibly further evidence of Mary of Guelders’ influence on the works at St Giles.

Final note
So, while we didn’t find what we were looking for, that is timber grown in SE Scotland, we are thrilled by this discovery. It has provided some really valuable new data to extend the native oak tree-ring coverage for NE Scotland, as well as significant new information about the history of St Giles Kirk and the Forest of Darnaway.

Thanks go to HES for the grant which made this research possible, to St Giles Kirk Session for permission to sample, to John Gilbert for information about Darnaway Forest history and to the many people who have helped with the SESOD project.

Posted in Dendrochronicle news | 1 Comment

The Scottish Lumberjills

Just back from the NWDG Woodland History Conference, 2019, on the theme of ‘100 years of state forestry in Scotland’. I presented this poster on the Scottish Lumberjills, in tribute to those amazing women.

Huge thanks to Joanna Foat, author of ‘Lumberjills’ who made many of these images available.

Marjory Stark, Bowmont 1944/5
Posted in Dendrochronicle news | Comments Off on The Scottish Lumberjills

Barhill Wood, Kirkcudbright

Delighted to have the opportunity to investigate the heritage of Barhill Wood, Kirkcudbright, as part of the Can You Dig It archaeology programme being run by Rathmell Archaeology for the Galloway Glens Landscape Partnership (GGLP). A prior walkover assessment helped to inform the public guided walk events, highlights of which can be seen here (with huge thanks to Sam Kelly for videoing):

Old elm stool with stilt roots. Barhill Wood (copyright C Mills 2019)

What follows is the interim summary of our findings, prepared by my colleague Peter Quelch who is a forester and native woodlands advisor. A fuller report will become available through the GGLP website later this year. Most of the map resources mentioned can be found on the National Library of Scotland maps website.

Old map analysis shows that Barhill ridge was wooded at the time of Timothy Pont’s survey c 1590, as depicted in Blaeu’s Atlas of 1654. One assumes these must have been natural origin woods, but not necessarily –they could be late medieval/early modern period plantings, being so close to an important town with port and castle.

Either way Barr Hill seems to be bare at the time of Roy’s survey in c1750, though we know that woods in poor condition or unenclosed may be missed off his maps. We do have a very useful survey plan by Robt Heron of 1790 which does indeed show a new star shaped plantation on Barr Hill, and this is confirmed by maps by Ainslie 1797 and Thomson 1821.

The first edition OS map of 1847 shows these now mature woodlands in great detail, as does the second edition OS map of 1895 which is also available at the 25inch to the mile scale. The woodland’s complex outline changes little over this time, and this is verified by a fascinating early aerial view of the town and wooded hill in 1930.

Then the FC acquired Barhill wood in 1952 and successfully replanted the entire woodland and the bare fields included within it, along with Janet’s Plantation to the north, probably after clearing most of any remaining early planting and seminatural woodland. So most parts of the woodland today have tall, heavily stocked and only lightly thinned crops of the common conifers, mainly Norway Spruce and Japanese Larch mixed with tall and slender beech, sycamore and ash.

However the archaeology of the old strip fields and alternating woodland mosaic is fascinating and yet easy to picture. The neighbouring fields often have that type of mosaic landscape, while the old maps and the early aerial photo back that up. This landscape is particularly interesting to the author who sees a very similar arrangement in the landscape of Mid Argyll. We heard that even today the remaining strip field between the FC and council owned woods is still used for outdoor lambing and prized for that reason by the local farmer.

The woodland in both ownerships is heavily used by local residents and is easily accessed by foot, car and bicycle, and sports both a visitor centre classroom and a squirrel observation hide. Janet’s plantation, at the north end, however has no trails and has more difficult access at present.

The resident red squirrels may present a challenge for future management of the woodland as the pines, larch and spruce trees probably represent a significant food source and shelter for these animals. So any kind of simplistic conversion to predominantly native species woodland would almost certainly worsen the squirrel habitat value of the woodlands. The way forward will probably involve selective thinning and felling, probably being hard on the beech and favouring birch and pine. However with all the various interests and issues of this wood, which is an important asset for Kirkcudbright town, both the drawing up of woodland management plans, and the carrying out of any work sensitively, will be quite a challenge for the community and the woodland owners.

Posted in Uncategorized | 2 Comments

SESOD: South East Scotland Oak Dendrochronology Project

The South East Scotland Oak Dendrochronology (SESOD) project aims to build the first long oak reference chronology for SE Scotland which represents a large geographic gap in native oak tree-ring coverage. This is part of a larger issue, that native timbers are generally under-represented in the national Scottish record compared to more-readily identified imports and this limits the degree to which any further native timbers can be recognised and dendro-dated.

Fortunately, dendrochronological work on deadwood from the Old Oaks of Dalkeith Park (see photo) is available to provide the anchor in time, ie the recent end of the new oak reference chronology to be produced by SESOD, with Dalkeith data spanning AD1592-2010. However, only a few trees are older than 1700, and more data from 1700 and earlier is sought.

The project director, Coralie Mills is requesting the help of those who know of any buildings or structures in SE Scotland with old timber which may potentially meet the objectives of SESOD. Coralie can be contacted at

Dalkeith Oaks sampling

Sampling deadwood, Dalkeith Oaks. Copyright C Mills 2018

Assessment visits will be undertaken to the most likely candidates in 2018-19, the first year of the 3-year project. SESOD will concentrate on locating, sampling and analysing oak timbers from selected historic buildings and structures in the Scottish Borders, the Lothians and Edinburgh, in sites where the timber stands a good chance of pre-dating 1700 and being native rather than imported. This will be easiest to predict in medieval buildings built before the great surge in imported timber to the eastern central belt from around 1450. However, away from the coast and where transport was more difficult, we would expect native timber to continue to be used in the late- and post-medieval periods. This was found to be the case at a townhouse on the High Street in Jedburgh which contained native oak felled in AD1667 (see photo). It was only possible to date it by comparison with oak chronologies from the north of England, where local oak was used well into the post-medieval period. Creation of an oak reference chronology for South East Scotland aims to enhance the ‘date-ability’ of historic oak timbers in this region.

Jedburgh Townhouse AD1667

Jedburgh Townhouse oak floor dendro-dated to AD1667. Copyright C Mills 2018

SESOD will run over three years, from 2018/19, with archaeology grant support from HES, and partnership working with a range of bodies, including Archaeology Scotland and the Forestry Commission Scotland on outreach and educational aspects. It also has the valuable support of the council archaeologists in the region.

This article first appeared in the Archaeology Scotland magazine, Issue 32, Summer 2018.

Posted in Dendrochronicle news, Uncategorized | Tagged , , , , | Comments Off on SESOD: South East Scotland Oak Dendrochronology Project

Meeting of UK Dendrochronologists, Cambridge, Dec 2017

The UK has a small number of dendrochronologists, and the majority of them are in this photo, taken at the recent meeting held at Cambridge University Geography Department with their recently appointed dendro group, Dr Ulf Buentgen (seated in photo) and his team, as hosts. This answers the question of whether we could all fit round one table. We are a rare resource! Those present include dendro-climatologists, dendro-ecologists and dendro-archaeologists. There was a full house for the Scottish contingent of Dr Anne Crone (AOC Archaeology), Dr Rob Wilson (St Andrews Uni) and me (Dr Coralie Mills of Dendrochronicle & St Andrews Uni). The attendees gave short presentations of our work, with time for wider discussion, and the meeting enriched our understanding of what the UK Dendro community as a whole think of as the pressing issues for our very specialist field, with long-term archiving and training of new dendrochronologists being among them. We all agreed that the meeting was valuable and pencilled in a plan to meet in Scotland, perhaps in two years’ time. A short article summarising the meeting has been published in Dendrochronologia, written by Ulf and all the participants. Please e-mail me if you would like a copy – see Contacts page for email address.

Copyright Ulf Buentgen 2017

UK Dendrochronologists’ meeting Cambridge Dec 2017: From L to R:  Neil Loader, Rob Wilson, Coralie Mills, Annemarie Eckes, Martin Bridge,
Ulf Büntgen, Rachael Turton, Tom Melvin, Paul J Krusic, Fredrik C
Ljungqvist, Mary Gagen, Alison Arnold, Alma Piermattei, Robert Howard,
Tim Osborn, David Brown, Roderick Bale, Giles Young, Mike Baillie,
Andrew Martin, Ross Cook, Anne Crone, Iain Robertson, Cathy Tyers, Nigel
Nayling, and Dan Miles. Photo: Copyright Ulf Buentgen 2017




Posted in Dendrochronicle news | Comments Off on Meeting of UK Dendrochronologists, Cambridge, Dec 2017

SCOT2K Native pine buildings dendrochronology article published

Castle Menzies lang garret native pine roof dated 1572. Copyright C Mills 2017.

Delighted to receive this early Christmas present, the publication in Vernacular Architecture of our paper on the SCOT2K dendro-dating of native pine historic buildings in Scotland. Using a combination of ring-width and blue intensity measurements, we have dated 20 pine structures from the 15th to the 19th centuries, and from grand castles to modest cruck cottages. While most are in the Highlands, near the native pinewoods, there is also evidence for long distance transport from the medieval period. The discussion considers not only the dendro-dates for the built heritage record but also the implications for historic woodland management, timber use and trade. The full reference and link to the article is given on the Projects and Publications page.

Posted in Dendrochronicle news | Comments Off on SCOT2K Native pine buildings dendrochronology article published

SCOT2K Pine Dendro – lecture video from ARP Conference

The Society of Antiquaries of Scotland have posted videos of the lectures given earlier this year of the 2017 Archaeological Research in Progress Conference in Edinburgh. This includes my lecture about the historical aspects of the NERC-funded SCOT2K Pine Dendro Project, undertaken at the University of St Andrews. You can find the recording here

Posted in Dendrochronicle news | Comments Off on SCOT2K Pine Dendro – lecture video from ARP Conference

SCOT2K pine dendro talk at Highland Archaeology Conference, 14th Oct 2017

Delighted to be speaking at the 2017 Highland Archaeology Festival Conference in Inverness this October about the SCOT2K project which, of course, concentrated on the dendro-dating of historic native pine buildings in the Scottish Highlands. The presentation is part of a panel session on archaeological dating. There’s a great programme for the conference, and for the festival as a whole. More information is available at at

Highland pine cruck: Copyright C Mills

Highland pine cruck: Copyright C Mills 2017

Posted in Dendrochronicle events, Dendrochronicle news | Comments Off on SCOT2K pine dendro talk at Highland Archaeology Conference, 14th Oct 2017