The Excavation of The Old Castle at Caerlaverock

Situated some nine miles south-southeast of Dumfries in southwest Scotland, Caerlaverock Castle is among the most impressive in Britain boasting spectacular views over the Solway Firth to the distant Lakeland fells.

It is a site of considerable antiquity encompassing an area from Wardlaw hill down to the shore of the Solway/Nith estuary. There appears to be little surviving written record about the area so to a large extent we must be guided by the archaeologists. Within this area, in addition to the substantial ruin of the castle, are the remains of an Iron Age Hill Fort, a Roman Fort, a Roman Fortlet, an ancient harbour (possibly Roman in origin, silted up and overgrown with vegetation), an earlier medieval castle, and additional features still to be investigated which may/may not be part of or linked to those already identified. In 1998 Historic Scotland commissioned an archaeological dig to investigate a grassy earthen mound with traces of stone walls, surrounded by a moat, in an area of wooded marshland approximately 200 metres to the south-east of the present castle. The brief was to provide information about this mound and the possible harbour to its south-west so that it could be 'set out' and interpreted for visitors. The dig lasted from mid-June until 11th of September but due to adverse weather conditions was only 80% complete. Historic Scotland considered that the results merited funding for a further four week period and the work was completed between June 9th and July 6th 1999.

The excavation was carried out by a team of archaeologists from the Scottish Urban Archaeological Trust(now known as SUAT) led by Martin Brann, assisted by Ray Cachart, with Adrian Cox responsible for finds interpretation, recording, and conservation.

Outline of map produced from the Ordnance Survey Get-a-map service. Image reproduced with kind permission of Ordnance Survey and Multi - Media Mapping. Detail on map added by hand.

This 'dig' was not the first investigation of the site. It is believed that excavations took place here in the 1860's - no records have been found but evidence remained in the shape of numerous overgrown collapsed trenches and spoil heaps spread across the site. The Royal Commission for Ancient and Historic Monuments spent some days surveying the site in the 1920's in the course of which they re-opened some of these trenches. Historic Scotland did an exploratory dig in the moat in 1978 during which they uncovered a piece of timber (possibly part of the original bridge/drawbridge across the moat) which was dated by dendochronology to the year 1227. Both of these later surveys were used to some extent in planning the 1998 excavation, the fieldwork for which commenced in February with trial trenching and auguring in the flat bottomed basin of the old harbour.

I became involved about three weeks into the actual dig, when I was asked if I would assist with conducting visitor tours of the site and the cleaning and cataloguing of 'finds'. The following account of the dig is my attempt to make sense of what was uncovered based on what I saw, was told by the professionals, and have read. It has no official standing.
Please note that the photographs selected to illustrate the various features were taken at different stages of the dig and are not necessarily in chronological order.

Based on the earlier 1920 and 1978 work it was suggested that the dig might uncover the remains of a 13th century castle built about 1220 and abandoned some 50 years later. It would be a fairly basic structure, built on a mound, consisting of a rectangular outer wall with four corner towers (one of which may have given access across the moat to the harbour) enclosing living quarters built against the south-east wall. The main entrance probably at the centre of the north-west wall.
As the dig progressed it became obvious that this was no 'basic structure' but a very complicated site showing evidence of different phases of building work in both timber and stone (which are very difficult to interpret and date) with its use continuing in step with the occupation of the present castle. It was soon evident that possibly the major reason why the castle was vacated in favour of the later (present) castle was subsidence caused by attempting to build a substantial stone structure on a poor clay site without adequate foundation.

This in turn raised the question as to why this site was chosen in the first place when there was a better one so close to hand - there is no definite answer to this although a theory can be advanced (more on this later). When the building became unstable and forced the move to the new castle (with as much of the stone as possible being salvaged for the new building) the site was used as a workplace - there is evidence of possibly two kilns (the second replacing the first, most likely used for drying grain) and substantial debris from the smithing of iron (and possibly the smelting of ore). Although when you have temperatures high enough to work iron, copper alloy working is perfectly possible also .

The work in the summer of 1998 revealed that the stone castle appeared to have developed through four distinct phases of building work and just a suggestion that there may also have been timber structures . The four phases of building in stone are colour coded on the plan:-

Plan of site resulting from 1998 excavation.

PHASE 1 The erection of a rectangular stone building of at least two stories (access to the second storey being made by an external stone staircase) on the south-east side of the mound . This first stone building is the only one on the site to have any foundations, being built upon a substantial raft of stones (some 0.75 metre deep) bound with clay, and this area of the site is the least affected by subsidence. The building had walls approximately one metre thick and internal dimensions approx' 9.5m x 6.5m, the height of the building is uncertain although one could attempt a calculation based on an assumption of the size of step used in the outer stone staircase.

PHASE 2 The building of the stone curtain wall enclosing the Phase 1 buildings. It appears that the main entrance to the castle was at the centre of the north-west wall, certainly it was in the later stages of occupancy. I think that from the plan of the curtain wall the main entrance was originally in the south corner prior to the erection of the additional internal buildings and tower. Others have suggested that this was only the builders following the shape of the mound. I discount this because it was the digging of the moat which created the shape of the mound and the moat was probably dug after building the curtain wall.

PHASE 3 The construction of buildings flanking the original two storey block and a hall stretching along most of the length of the south-west wall. The thickness of the walls of these later buildings is considerably less than that of the original, nowhere being greater than approximately 0.75m thick. Within the hall area was evidence of a substantial hearth in one corner (hearth 1) also one of a presumed pair of post pads which would have carried upright wooden posts supporting the roof (see plan). Other post pads were found, the most intriguing being a row of three in a line outside the south-east range of buildings. These ran parallel to the stone staircase, suggesting that they may have supported the roof of a timber arcade running along the front of the building.

PHASE 4 The final stage of building was the three corner towers, not four as originally expected. These three towers, from their form, appear to have been built more as braces to shore up the curtain wall rather than as additional defensive features. The reason for three rather than four towers (most likely) being that the fouth corner, adjacent to the first building on its stone raft, was still stable and did not require reinforcement. This work obviously was not a success as the subsidence, particularly on the south-western side, continued to such an extent as to eventually force the move to a new site.

Also uncovered in this first dig, as well as the remains of the later kilns and metal working activities, were two rows of post holes, within the line of the curtain wall, either side of the northern tower. It is unclear what purpose these served, they were not deep enough to be from a pallisade (in the military sense) although they may have been from a lighter temporary protective structure, also (short of dating evidence) we don't know whether they predate, are contemporary with, or postdate the curtain wall. There has been a suggestion that they were intended to reinforce the base of the wall and lock it to the clay; I think this is unlikely as there is insufficient space between the posts to allow stonework to bind, consequently the rows of posts would have split and weakened the base of the wall.

• Substantial quantities of broken pottery mainly 13th century but a few pieces identified as late 12th century.
• Assorted metalworking debris plus the remains of a few metal objects in particular a very nice small swivel ring.
• A small decorated whetstone in very good condition.

• A number of worked stones including the upper part of a hand quern (used for grinding grain) and a nice socketed stone(below right) which would have held the foot of an upright wooden post.This later stone is very similar to one uncovered in Birrens Roman Fort during excavations in 1896.

• Quantities of animal bone.
• A few small fragments of decorated glass which were provisionally identified as part of a 16th century decorated wine glass.

The follow on dig in 1999 was intended to complete the stripping of the top of the mound ready for conservation and to carry out further investigation of the bridge area in the moat. Unfortunately the records for the 1978 dig had been lost and it was later discovered (from some photographs which turned up) that our excavation line had replicated that of 1978 and consequently nothing was found apart from re-finding the piece of bridge timber found in that dig. A tentative exploration to one side uncovered some substantial pieces of broken pottery and probing across the moat (with a metal rod) seemed to indicate a couple of solid areas (remains of bridge supports?) but due to lack of time no further investigation was possible.

The completion of the stripping of the mound was more rewarding.

In the Hall area traces of a quite substantial timber building, or possible range of timber buildings, was uncovered. This predated the hall as the wall line of the timber building ran at right angles underneath the inner stone wall of the hall. A second hearth area (hearth 2) was also uncovered in the hall. Clearing of the courtyard revealed substantial areas of the original rough cobbled surface-in particular a length of well cobbled pathway which appeared to have run from the gate to one end of the double storey stone building. It was suggested that this path led to the foot of the stone staircase but this could not have been so as stairs ascending from this point could not have given access to the upper storey. No source of water such as a well or cistern was discovered although a slight depression towards the centre of the courtyard looked as if it would have rewarded further investigation had time permitted.

The two 'hearths' (areas of ground showing evidence of heat and burning rather than built fireplaces) were sampled for archeomagnetic dating. Their position suggested to me that they were not connected with the hall (or earlier timber building) within which they were found. Initially I thought they were possibly the remains of demolition bonfires used to clear material from the site but the lack of evidence in the form of ash or charcoal tends to rule this out. The archeomagnetic dating of the 'hearths' gave a possible date range of:

• Hearth 1: 1250 - 1310 or 1400 - 1440
• Hearth 2: 1250 - 1275

There may well have been more buildings on the site particularly either side of the gate but this area was so confused by the 'industrial phase' it was impossible to tell.

• Quantities of broken pottery of the same types as previously.
• More metalworking debris again plus the remains of a few unidentified metal objects.
• A small lead object tentatively identified as the handle for a seal.

• Some worked stones including one which had been used as a sharpening or grinding stone.
• The piece of timber previously uncovered in Historic Scotland's 1978 dig in the moat.
• A half coin thought to be a cut silver penny from the 13th century.

I will now give my theory on the development and demise of the old castle based on the information I have so far been able to gather.

It is has been suggested by a number of eminnent historians that the old harbour was the site of "the Roman sea base in Nithsdale with its guardian fort on Wardlaw Hill" and that the castle site may have been the location of the harbour strongpoint garrisoned from the fort. So far there is no archaeological evidence to substantiate the location of a Roman strongpoint on the old castle site and I think that we must look elsewhere in the immediate locality for its position if indeed it existed.
All the evidence available suggests that the site is at earliest late twelth century but more positively early thirteenth century.

The lands of Conheath and Caerlaverock were granted to the Abbey of Holm Cultram sometime between 1153 and 1165, the most likely year being 1157, but were returned to the Scottish Crown sometime before 1174. The Phase 1 two storey building being the only one on site with foundations suggests that its builder was different to that of the rest of the site. I suggest that this two storey building was erected by the monks of Holm Cultram as part of their Grange (dictionary definition "a granary") at Caerlaverock with possibly ancillary timber buildings. The siting adjacent to the harbour made a convenient location to assemble and hold goods for shipping across the Solway to the Abbey harbour of Skinburness (almost directly opposite on the Cumbrian shore). I think it unlikely that the monks would have had a moat around the buildings or any kind of fortification.

Sometime later Caerlaverock was acquired by John de Maccuswell (Maxwell),most likely around 1200 although no charter survives according to The Book of Caerlaverock. At a talk on the dig given by Doreen Grove, of Historic Scotland, she stated "the date of the Royal Charter was 1204"(I have still to verify the grounds for this statement). According to 'Historic Scotland' "Sir John came to Caerlaverock in 1220 and started to build his castle". This suggests that John de Maccuswell was the person responsible for the fortification of the site, which I can accept, but I have a problem with the suggestion that the castle was started in 1220 and vacated for a new castle in 1270. In my view this timescale is not realistic. I think it more likely that after gaining the estate around 1200 John de Maccuswell, requiring a secure base for use when visiting and to house his Steward, decided to modify the existing harbour buildings so Phase 2 - the digging of the moat and erection of the stone curtain wall to enclose the original stone and timber buildings. The fact that for this and later building work no foundations were laid may possibly be explained by the stonemasons seeing a substantial stone building already standing on the site, and not being able to see the solid foundation beneath it, would think that foundation work was not necessary.

John de Maccuswell was knighted in 1221 and it was possibly this increase in status which infuenced the decision that more and better accomodation was needed at Caerlaverock so giving us Phase 3.

One of the many visitors to the dig in 1998 - a geologist by profession I believe - stated that around fifty years was the timescale to expect problems if building a substantial structure without foundations on a clay site. So how does this fit my timing?

Fifty years on from phase 2 - the first with no proper foundation - would take us to around 1250 for Phase 4 the attempt to stabilise the curtain wall with three bracing corner towers. Fifty years on from phase 3 - more building without foundation - takes us to 1271 just one year past 1270 when it was realised that the structural problems were insurmountable and so the move to a new site.

The positive datings that I have knowledge of from the achaeological excavations are:
The fragments of glass found in the '98 dig, provisionally identified as part of a 16th century wineglass, on further specialist examination were found to have arabic script on them and have now been positively identified as part of a Syrian(or Egyptian) fine glass beaker dating from the late 13th century. This is a very interesting find for Caerlaverock and one wonders how the beaker arrived there. Was it a souvenir obtained during foreign travel or the result of trading between Caerlaverock and foreign lands?
The supposed bridge timber uncovered in the moat with a dating of 1227 - this ties in reasonably well with the phase 3 modifications to the site and the resiting of the main entrance. It is an interesting coincidence that this date of 1227 and the date for the first bridge in the new castle (1277) are both seven years later than the dates given for the commencement of building work.
The coin found has been identified as a cut silver half penny of William the Lion struck between 1205 and 1230. Its loss can be assigned to the period c1210-1250. In 1250 there was a re-coinage and coins of this kind would apparently have held no value, hence it was very probably lost before 1250. - Going by where the coin was found would suggest that it was lost during the phase 3 activities around 1221 when the timber buildings were replaced by the stone built hall.
The initial tentative dating of the pottery suggested the period from late 12th century (1157?) to 13th century. More accurate dates should become available from the detailed analysis of the pottery and other artefacts.

• The Transactions of the Dumfries & Galloway Natural History and Antiquarian Society.
• The Register and Records of Holm Cultram by Grainger and Collingwood.
• Publications of Historic Scotland.
• The Book of Caerlaverock.(The Maxwell/Herries family history compiled at the end of the 19th century)

The 'Head' shown at the top of the page is a figure that was found in the moat of the present Castle and is thought to have come from the first castle. It is on display in the Visitor Centre at Caerlaverock.

I am particularly grateful to Adrian Cox( adrian@wildcat4.freeserve.co.uk) for all the help he has given me and in being a very tolerant expert upon whom I have inflicted (tried out) my theories. Adrian devotes a considerable amount of his own time to 'The Young Archaeologist's Club' and is developing a special website for them which could be useful for anyone new to archaeology. The address is http://vzone.virgin.net/adrian.artefacts

It will be interesting to see if the official report on the dig provides any kind of verification of my theory.

If any one reading this has any information relating to Caerlaverock that can add to the knowledge I have already then I would be delighted to hear it. I can be contacted by e-mail at the address shown on the Home page.