Site Report

WoSASPIN 44790 NN00NW 6 0387 0685

What may be a single unroofed shieling-hut is depicted on the 1st edition of the OS 6-inch map (Argyllshire 1874, sheet cxxxii), but it is not shown on the current edition of the OS 1:10000 map (1990).
Information from RCAHMS (AKK) 21 July 1998.

Allt Fearna: The area of the Allt Fearna is depicted as marshy upland moorland. There are two small rectangular structures depicted on the northern bank of the Allt Fearna, which can be identified as at the site of the shieling settlement at NGR NN 0387 0685 (NMRS No. NN00NW 6). The Second Edition OS map of 1900 does not show any structures and the two huts shown on the First Edition map are no longer shown. The area around the Allt Fearna is still depicted as moorland. No structures are shown on the modern 1974 OS map.

The Allt Fearna settlement is located on the west bank of the Allt Fearna at NGR NN 0390 0692 on an old river terrace. The east side of the site is defined by the burn and the west side of the site is formed by a steeply sloping natural bank which rises to a level plateau. All of the shieling huts of the main settlement Sites 1 - 13, with the exception of Site 11, are located on the old river terrace on the fairly level but undulating ground beside the burn. A further group of four huts, Sites 14 – 17, are locate across a small burn a little to the south at NN 0385 0687 The ground cover is predominantly rough grass and bracken. All of the shieling huts and associated structures are overgrown with grass, moss and bracken and the majority of the stone foundations are covered with moss.

The vegetation cover coupled with the collapse and general decay of the structures makes it impossible to determine all of the details of the structures and clearance and limited excavation would be required to enable detailed recording or plans of the buildings. All of the structures have been described, measured, photographed and sketched. The sketches of the structures indicate their basic outline and plan but do not show details of the stones used in the foundations as these were generally obscured by vegetation and as noted clearance would be required to allow for detailed plans to be made.

The site is sheltered by the steep banks of the burn rising up to the east and the west.

The prevailing wind is from the SW. Excellent views up the Allt Fearna valley and to the north to Douglas Water and over the surrounding hills are afforded from the top of the west bank of the burn and in particular from Site 13.

Two sites were excavated at the Allt Fearna intake, AF Site 1 – a shieling hut which was found to overlie a burnt mound and AF Site 18 a shooting butt at NN 0400 0700. Both sites would have been destroyed by the hydroelectricity development and were excavated so that they could be preserved by record.

AF Site 1 Shieling hut. 5.3m ENE-WSW x 6.40m WNW – ESE overall including mound. Concentration of stone collapse measures 1.90m WNW-ESE x 1.70m NE-SW and the open interior space in the centre measures 1.3m WNW-ESE x 0.95m NE-SW. The maximum exterior height is 0.90m and the maximum interior height is 0.40m. This site is a poorly defined sub circular mound with no clear definition of the walls but it seems likely to be a sub-circular or rectangular shieling hut with stone foundation walls. The site is disturbed on its N side from test pit operations and the survey peg. The survey peg on this site was used as the survey station for the archaeological survey of the Allt Fearna settlement. The site is covered with bracken, grass and moss covered stones. The entrance is possibly located on the SW side of the structure. This structure was directly affected by the weir construction and required excavation in advance of the weir construction operations. See Allt Fearna Excavations section below.

AF Site 2 Shieling hut. Rectangular structure aligned NE-SW at 45° with a possible entrance at the south end of the west wall. The walls incorporate a bedrock outcrop in the NW and large boulders on the north side. There is no sign of interior features. It has been eroded on east and south side by the Allt Fearna burn. It is overgrown with bracken, moss and grass and measures 6.30m NE-SW by 2.90m W-E with a maximum height of 1.20m.

AF Site 3 Shieling hut and outbuilding. Two rectangular buildings aligned NW-SE, but they are apparently not connected and two discrete structures rather than two rooms of the same building. The smaller structure may be an animal pen or dairy. Both buildings have stone foundations and are best preserved to the south. They are both overgrown with bracken, moss and grass. The larger structure is aligned at 126° and measures 5.65m SE-NW by 3.00m SW-NE with 0.80m thick walls and stands to a maximum height 0.70m. The smaller outbuilding measures 2.10 SW-NE by 1.70 NW-SE.

AF Site 4 Shieling hut. Sub-rectangular with rounded corners located on a low mound. The building is on a low ridge running W to E that may form an old terrace or riverbank to the burn, this forms quite a prominent position. The structure is sub-rectangular with stone foundations and rounded corners with one visible course of stones. The entrance is in the centre of the NW end and is 0.40m wide. The building is on a platform which could be a natural glacial feature but may have been enhanced. The building is aligned at 136° and measures 4.10m NW-SE by 3.50m SW-NE, it has 0.65m thick walls standing to 0.40m high.

AF Site 5 Shieling hut. A sub-rectangular structure aligned SSE-NNW at 156° measuring 6.60m SSE-NNW by 5.30m NNE-SSW overall, with a partition wall running NNE-SSW forming two rooms measuring 2.40m SSE-NNW by 3.00m NNE-SSW and 2.30m SSE-NNW by 1.00m NNE-SSW respectively. There is a possible doorway through the partition wall at the E side of the building. This structure is cut into a hill slope and it slopes considerably from N to S. There are no well defined wall lines but bedrock is probably incorporated at the NE corner. This site would require excavation in order to define the interior details.

AF Site 6 Sub rectangular structure aligned NNW-SSE at 114° measuring 3.90m NNW-SSE by 3.30m SSW-NNE on a low mound. Poorly defined boulder walls 0.40m thick surviving to 0.30m high with no clear wall faces, there is a possible entrance in the middle of the W wall. The site is covered in bracken moss and grass and is interpreted as a shieling hut possibly with stone foundations and turf walls as there is not enough collapsed stone to suggest an all stone construction. Associated with AF Site 12.

AF Site 7 Rectangular structure oriented NW-SE at 138° measuring 4.35m NW-SE by 4.17m NE-SW on a mound sloping off to the SW, S, SE and E. It has well defined stone walls 0.65m thick surviving to 0.46m high with 4 courses visible with no obvious entrance. The walls are constructed of stones 0.40m by 0.40m by 0.10m while the river boulders are 0.35m by 0.25m by 0.25m indicating that the flatter stones are being selected for use in the walls. The amount of stone tumble at this site suggests that the walls stood up to 1m high before the roof springers or any turf upper walling. Interpreted as a shieling hut.

AF Site 8 Shieling hut. A sub-rectangular structure oriented ENE-WSW at 70° measuring 4.70m ENE-WSW by 3.10m NNW-SSE set very close to the river bank. It has boulder built walls 0.55m thick and 0.50m high that are poorly defined due to stone collapse. There is a possible entrance at the E end of the N wall. There is a possible interior partition wall running NNW-SSE that may have a doorway in the middle of its length but this could be tumbled stone. This site is obscured by stone collapse but is well preserved.

AF Site 9 Well preserved oval structure 1.90m N-S by 1.30m E-W. It has corbelled stone walls standing 1.00m high with 9 courses visible creating a flat bottomed interior space 0.55m across. The walls are covered in moss and lichen. This site is interpreted as a shooting butt with a view to the S. The stones used in its construction were probably robbed from the surrounding shielings or it could be an earlier structure that has been converted into a shooting butt.

AF Site 10 Shieling hut. A square stone structure oriented E-W at 90° measuring 3.30m E-W by 3.20m N-S on a low mound especially on the E side, which is stony suggesting a built up platform, which also extends to the W with walls on 2 sides forming a porch or outhouse. The walls of the main structure are 0.50m thick surviving to 0.45m high externally and 0.30m internally. There is an entrance in the middle of the N wall 0.50m wide. The site is covered in bracken, moss and grass and would require excavation to define the walls.

AF Site 11 A well preserved rectangular structure oriented NNE-SSW at 20° measuring 5.40m NNE-SSW by 3.65m ESE-WNW located on higher ground above the rest of the settlement with good views of Allt Fearna. The walls are constructed of river boulders 0.50m thick surviving to 0.50m high with 3 courses visible. There is an entrance 0.40m wide in the middle of the SE wall. This site is interpreted as either a shieling hut or shepherds hut or bothy that may be of a later date as it is relatively well preserved and is located on prominent higher ground as opposed to the sheltered ground along the banks of the burn. There is a moss covered square stone platform 1.40m NNE-SSW by 1.20m ESEWNW 0.50m high built onto the W wall. This is interpreted as a collapsed shooting butt.

AF Site 12 A small sub rectangular structure oriented NE-SW at 136° measuring 2.30m NE-SW by 1.80m NW-SE to the NE of site 6. It has stone walls 0.7m thick surviving to 0.1m high with only one course visible. The walls are poorly defined at the E and SE sides and moss covered. The site is covered in bracken. Interpreted as a small outbuilding or dairy associated with Site 6.

AF Site 13 Curving bank oriented E-W measuring 11.00m long by 2.00m wide standing 0.70m high located at the top of hill edge. Site covered in moss grass and heather. Interpreted as a shooting butt with clear views to the E, S and W. AF Sites 14 – 17 are located further to the west than Sites AF Sites 1 -13 and were not located at the time of the total station survey. They are located at NGR NN 0385 0692.

AF Site 14 NN 03881 06867 241m Shieling hut. A sub-rectangular structure aligned N-S measuring 6.00m N-S by 4.50m E-W. This site has well preserved stone walls 0.80m thick surviving to 0.50m high with 3 courses visible. There is a possible entrance in the W wall and a possible fireplace lintel at the S end of the structure. There is a possible enclosure on the E side of this site. The walls have tumbled into the interior but this is a substantial and well preserved structure. The walls and interior are overgrown with bracken, moss and grass.

AF Site 15 NN 03881 06851 246m Shieling hut. A sub-rectangular structure aligned E-W measuring 5.50m E-W by 3.5m N-S. This site has badly collapsed and slumped, stone walls that are now over 1.20m thick but were perhaps 0.80m thick originally, walls survive to 0.40m high with 2 courses visible. The location of the entrance is not obvious. There is a possible wall running from the E end along the cliff edge leading to Site 16. Building is overgrown with bracken, grass and moss. This building is probably the one shown on the 1st Edition OS map of this area, while none of the rest of the Allt Fearna sites are marked.

AF Site 16 NN 03900 06853 245m Small sub circular structure measuring 2.50m N-S by 2.50m E-W. This site has walls 1.40m thick standing 1.00m high externally and 0.45m internally that appear to be made mainly of turf but probing of which revealed some stone. There is a possible entrance at the E of the site.

AF Site 17 NN 03878 06876 257m Small sub rectangular structure aligned NE-SW measuring 4.50m NE-SW by 3.00m NW-SE on the top of the knoll to the W of Site 14-16. This site has walls that are 0.70m wide standing 0.30m high that appear to be constructed entirely of turf with no stones obvious.

AF Site 18 A small shooting butt located at NN 0400 0700. This site was excavated. See Allt Fearna Excavations Section below.

Excavations were carried out at Allt Fearna between mid June and October 2006. Approximately 2 weeks were lost to bad weather and the shieling / burnt mound site was prone to flooding from ground water after heavy rain and much time was lost on repetitive cleaning of rain lashed and saturated deposits despite covering the site to protect the archaeological deposits.

In addition to the excavation a watching brief was carried out on the new access track up to the N side of the burn, the bridge crossing and construction and two access tracks to Site AF1, one at the bottom of the slope on the west bank of the burn and one at the top of the hill on the West side of the burn. No archaeology was disturbed or discovered on the lines of the tracks with the exception of a small shooting butt AF Site 18 which was excavated in advance of the bridge and track works.

AF Site 18 Shooting Butt Located at NN 0400 0700

A curvilinear feature aligned approximately N-S covered in moss and bracken, measuring 3.50m SE-NW by 2.60m SW-NE, with walls constructed of boulders surviving up to 0.50m high. Removal of vegetation revealed a rather collapsed curvilinear structure incorporating bedrock especially on the SW side where a large slab c. 1.00m by 1.30m forms the base of the wall. Other rounded and sub angular glacial and river boulders of average size 0.35m by 0.20m by 0.15m form the wall. There is no obvious coursing to the wall which is 0.90m thick with a slight batter and has collapsed into the interior and exterior at the NNE side. This structure is interpreted as a shooting butt. Tree roots have damaged its integrity and there is a general collapse of the upper parts of the wall. Three sondages were excavated by hand to check for any anthropogenic deposits in and around the shooting butt. Sondage 1 was located between the shooting butt and a curve of glacial boulders c. 3.00m to the NE. In this trench there was 0.20m depth of black bracken root mat and rotting vegetation F101 which overlay 0.25m depth of mottled orange grey topsoil F102 (60% sand, 20% clay, 20% silt) that was lying on a pale orange gritty sandy gravel natural subsoil F103. The interface between the topsoil and the natural was heavily iron panned and both deposits had tree roots throughout. Sondage 2 was located just outside the SE wall of the shooting butt to investigate the construction method of this structure. In this trench there was 0.20m depth of black bracken root mat and rotten vegetation F201 overlying 0.10m depth of mottled orange and grey topsoil F202 (60% sand, 20% clay, 20% silt) which lay on a pale orange sandy gritty gravel natural subsoil F203. Sondage 3 was located in the centre of the structure in order to investigate whether any floor deposits survived. In this trench 0.15m depth of black bracken root mat and rotten vegetation F301 overlay 0.20m depth of mottled grey and orange topsoil F302 (60% sand, 20% clay, 20% silt) which lay on a pale orange gritty sandy gravel natural subsoil F303. This structure is a dry stone walled shooting butt constructed in a rough and ready style. It was built directly onto the ground with no foundation trenches although several large glacial boulders are incorporated into the walls on the SW side. The shooting butt is located in a place where any game being driven down the Allt Fearna valley would be forced close to the shooting butt as the gap between the river and the steep sides of the valley narrows in the vicinity of the shooting butt.

Site AF 1 The Sheiling Hut and Burnt Mound. The site initially appeared as a poorly defined sub-circular mound with a small stone built cell at the centre and was covered with bracken and rough upland grass. The site had been disturbed along its NNE side by the excavation of a test pit before the archaeologists arrived on site. The site had been used as a survey station and a grid peg with the value of 204025.53E 706955.15N and an elevation of 224.54m was conveniently located immediately adjacent to the stone cell at the highest point of the mound. This survey point was used as the base station for the archaeological total station survey of the Allt Fearna shielings and also as the base point for the excavations of Site 1. Rather than deal with such long numbers for co-ordinates a site grid was laid out over the excavation area and the grid peg is located at 12.70E 14.07N on the small site grid.

The excavation area measured 10m N-S by 11m E-W, the trench size was extended on two occasions as the excavation progressed to ensure that all archaeological remains were exposed and investigated. The trench was aligned NE – SW at 57° but the NE side of the trench was designated as Site North for ease of description during the excavations. A spoil heap bund was created along the southern boundary of the site between Site AF1 and Sites AF2 and AF3 and orange fencing was erected along this at the end of the excavations to protect the other sites from machine damage during the construction of the weir.

The mound site was covered with a mixture of rough pasture grass turf F001 and bracken F008. The bracken had an extensive root mat that had penetrated the underlying turf and caused considerable bioturbation. Initial excavation at the site was excavation by secateurs to remove the extensive roots. Along the W side of the site on the lower lying flat ground covered only with rough pasture grass and no bracken the topsoil was allocated F025. However, F001 and F025 are the same topsoil / turf horizon.

Once the bracken and topsoil had been removed collapsed and tumbled stone F003 from the central cell was found to overlie a turf mound F002 on the exterior of the stone cell F010 and to have tumbled into the interior of the cell, where it, F003, overlay a rich black organic silt deposit F004.

At this stage of the excavation the site was divided into four quadrants (NE, NW, SE, and SW) so that sondages could be excavated through the turf mound. The reason for excavating small 0.5m wide sondages in a criss-cross across the site was to provide a running section through the site so that we could see how the mound was constructed on either side of the central cell – this would allow us to investigate the construction technique of the turf mound, the stratigraphic relationship of the turf mound and the central stone cell, to define the exterior wall face of the cell and to look for any collapsed roofing material or variation within the mound construction.

The deposits encountered in the four sondages excavated through the turf mound and underlying buried soil and burnt mound were all allocated different context numbers at the time of excavation. However, as excavation progressed and we could determine that the deposits revealed in each of the sondages were actually continuous throughout the site and the same feature they were amalgamated into one context number, usually the first number allocated to the deposit. For example the buried ground surface that underlies the shieling cell and mound is F007 and this number is used throughout for description. F007 is equivalent to F018, F019, F022 and F013. Equally the uppermost deposit of the burnt mound is F011 and it is equivalent to F015, F017, F023 and F047. When these deposits are discussed the equivalent context numbers will be given the first time it is mentioned but thereafter only the all encompassing number will be used. The equivalent context numbers are shown in the site’s stratigraphic matrix. The excavation of the sondages will not be described individually as the abundance of context numbers and repetitive results would not give a clear description of the site and the sondages are an excavation technique to obtain the stratigraphic information.

The Shieling Hut
Once the shieling mound had been cleared of vegetation the loose and tumbled stone F003 that had collapsed from the central cell construction was removed from the interior of the cell and from the surface of the turf mound. There was not a huge quantity of stone suggesting the cell may only have stood another course or two higher than the surviving structure F010. The tumbled stones F003 were assessed to determine if they might represent corbelling stones for a stone beehive roof for the cell but their mixed sizes and shapes indicated that this was not the case.

In the interior of the stone cell F010, the black organic silt F004 that underlay the tumbled stones F003 was found to be decomposed and rotting bracken and vegetation that had been squashed under the fallen stones. Underlying this 0.05m deep deposit, which was riddled with bracken roots, was F006 a 0.10m deep layer of firm greybrown clayey silt with inclusions of tumbled stones and some charcoal flecks. This deposit was interpreted as the collapse and abandonment deposit within the small interior living space. This deposit was subjected to metal detecting, as was the whole site, but no artefacts were found.

Excavation continued in the interior of the cell F010 in the hope that a floor deposit or occupation derived material or surface might be found. However, underlying F006 was F013, a soft dark grey clayey silt with frequent charcoal flecks, which was present throughout the interior. This deposit was only up to 0.05m thick maximum in the centre of the cell and up to 0.20m thick at the edges of the cell against the stone walling F010. This layer F013 is interpreted as the floor level of the shieling hut and it had been worn down in the centre of the space by use. When the stone walls of the cell, F010, were removed later on in the excavation it was found that F013 was in fact the same as the buried ground surface F007 on which the shieling had been built.

At the time of excavation the interior of the cell was excavated as a separate entity from the exterior of the cell with sections left in situ so that a section right through all the deposits of the shieling hut could be obtained. The charcoal flecks in F013 differentiated it from the buried ground surface F007 exposed outside the stone cell but there was no doubt it was the same deposit when the stone cell walls were removed. In the interior of the cell where there had been feet traffic and activity the old ground surface had been worn down and mixed and trampled with the underlying burnt mound deposits. Underlying F013 (=F007) in the cell interior was F014 a very compact mottled dark brown and orange burnt clayey silt with frequent angular fire cracked stones up to 0.15m maximum in size. F014 was far too lumpy with pointed stones in it to have been a comfortable floor for the shieling inhabitants and it extended under the cell walls and beyond the extent of the cell and is in fact part of the burnt mound deposits.

The excavations on the exterior of the cell initially focussed in the four sondages cut through the turf mound F002. F002 was a compact pale grey brown sandy clay deposit with frequent linear orange brown sandier patches that represent the edges of individual turf sods. The turf sods could not be easily detected in plan but were clearer in section and measured c. 0.18m x 0.11m x 0.10m on average. The mound was constructed of individual sods of earth, some of which were obvious, but the sods were not uniform in shape or size. The turf mound extended around the cell on all sides but was deepest on the N and W sides and thinnest on the SE side. The mound varied in depth from 0.15m to 0.45m.

Some stones were present in the turf mound particularly on the W side where some boulders up to 0.35m x 0.30m x 0.25m were present. Continuation of the excavations to the west of the shieling hut revealed a palaeochannel and a very high water table at the W side of the site which led to problems with rising ground water and flooding at the W and SW of the site. A drain was excavated along the W side of the site to try and alleviate this problem. The presence of boulders in the mounds construction on the W side, where earth fast glacial erratics were also present, may have been to strengthen the mound in this area and stop water erosion and improve drainage. The turf mound itself would have provided improved drainage to the central cell as water would have shed down the slope of the mound. The turf mound would also have provided insulation for the stone cell F010.

Around the SW edge of the site a more convincing kerb of boulders was present at the base of the mound. The largest boulders were earthfast glacial erratics and continued excavation revealed that underlying the buried soil F007 there was a concentration of stones in this area associated with the burnt mound. The boulders that suggest a kerb at the SW of the site are in fact either natural or protruding through from underlying deposits with which they are associated.

Elsewhere the turf mound had a few intermittent stones at the base of the turf construction but none of them were convincing as a kerb or retaining wall. It is possible that some of them were placed to stop water erosion of the mound, for example at the NE side but others such as F005 probably represent some limited clearance of boulders from the site in advance of building the shieling hut and mound and were originally associated with the burnt mound. There was evidence of erosion of the turf mound down the steep river bank on the E side of the mound.

No trace of any roofing materials were found on the turf mound and if any deposits that may have indicated thatch had been present it is likely that the intensive bioturbation of the surface of the mound caused by the bracken root had obliterated them.

The turf mound F002 and the stone cell of the hut F010 were built as the same time. Excavation revealed that the turves of F002 had been packed up, against and into the boulders that form the walls of F010 and the stones of F010 appear to have been rammed against the turves of F002. It makes sense that both the stone cell and the mound were built at the same time as both reinforce the structural strength of the other.

The remainder of the turf mound was removed under controlled conditions by a minidigger machine fitted with a toothless ditching bucket to expose the underlying buried ground surface F007.

The stone cell F010 measured 2.80m E-W x 2.60m N-S and stood to a maximum internal height of 0.60m. It was entered from the S side through a 0.60m wide doorway defined by two boulders. There was no clear threshold stone but a single flat stone outside the cell and lying on F007 indicated an exterior threshold, almost like a doormat.

No artefacts were recovered from the shieling mound and bioturbation renders any charcoal samples, which may be disturbed from the underlying deposits when the turves for the mound were cut, unreliable. The internal cell is very small and would really have only accommodated one or two people as a sleeping space. There was no sign of any internal fittings and such a small space would have been uninhabitable had a hearth been present. Although some effort had gone into the construction of the shieling it really is a very small and basic shelter as opposed to spacious accommodation.

The Buried Ground Surface
Underlying the shieling hut was a layer of dark grey gritty clayey silt F007 (= F013, F018, F019, F022 and F027). F007 is interpreted as a buried ground surface and this was confirmed by palaeoenvironmental specialist Dr Steve Lancaster on a site visit. A few areas of black silt were found under the turf mound F002 and on the buried soil F007, these patches of black silt F021 and F024 were interpreted as compressed vegetation trampled and compacted under the turf mound and this was also confirmed on site by Dr Lancaster. A few patches of charcoal in which twigs were clearly visible were located at the SW side of the site. These burnt deposits F028 was sealed under the turf mound F002 and lay on the buried ground surface F007. A concentrated patch of charcoal F029 was associated with stones protruding through F007 which seem to have been utilised to form the edges of a hearth. F029 was roughly circular in plan and 0.50m diameter and up to 0.05m deep. This indicates some activity on the old ground surface that had accumulated over the burnt mound before the shieling hut was built. A radiocarbon date for this small hearth feature F029 would provide a date after which the shieling hut was built. Considering shieling huts and burnt mounds were in use for centuries it would be useful to try and identify when the old ground surface was current.

The buried ground surface F007 varied in depth between 0.05m and 0.15m and was deepest at the SW corner of the site in the wettest area. F007 was found to overlie many of the boulders in the SW corner of the site, i.e. the boulders that looked as if they might be a kerb for the turf shieling mound were actually present before the ground surface on which the shieling mound was built had formed.

The buried soil F007 had accumulated over the burnt mound deposits that underlie it and the reformation of this buried topsoil has probably contaminated the top of the underlying burnt mound deposits with bioturbation. It had also had the effect of compacting the underlying deposits and aiding in the formation of iron pan F016 which was present in occasional patches overlying the burnt mound deposits.

The Burnt Mound
Underneath the buried ancient topsoil F007 were a sequence of burnt charcoal rich deposits containing frequent burnt and fire cracked stones, mineralised burnt soils, patches of scorched natural subsoil and areas of iron pan F016 formed by disturbance of the natural subsoil and precipitation and oxygen.

The burnt mound deposits had first been identified in the sondages cut through the turf shieling mound. As might be expected these burnt deposits were rather messy with concentrations of fire cracked stones or charcoal rich mineralised soil or mixtures of mottled charcoal rich scorched earth. Excavation took the approach of allocating individual context numbers to every identifiable discrete patch of burnt material and the different types of burnt material –e.g. fire cracked stones as opposed to charcoal rich silt. However, we found as excavation progressed that many of the burnt deposits merged into one and contexts initially indentified as discrete incidences of burning were in fact part of the same large spread of burnt mound material. It should also be noted that the burnt mound material would have been moved about by the users of the burnt mound as they cleared space for the next incident of stone heating and water boiling.

Following the removal of the buried ancient topsoil F007 it was noted that this deposit extended to the W of the archaeological deposits in the undisturbed natural soil profile. The buried ground surface where it had not been compressed by the overlying turf mound F002 was different in character being a firm reddish brown silty and gritty slightly sandy clay. This buried soil was allocated F027 and is the same horizon as F007 but appears different due to different environmental conditions.

At the SW area of the site the removal of buried soil F007 exposed a spread of small boulders and cobbles F030. F030 extended over an area of 2m N-S and 3m E-W and the deposit was composed of 40% stone in a matrix of reddish-brown-grey gritty silt and silty sand. Most of the stones were water rounded cobbles and small boulders and measured 0.35m x 0.20m x 0.12m in size with some smaller angular stones of 0.07m x 0.6m x 0.03m. These stones appear to be a deliberate dump of cobbles and boulders into what later turned out to be a shallow depression F108 and they overlie the burnt fire cracked stones F011. This area was quite wet and subject to flooding after rain when the water table rose. These stones are interpreted as the stone pile for the burnt mound. They had obviously been deliberately collected and considering their location to the W of the main hearth area represented by F072 this seems the most likely explanation. Some time was spent investigating these stones for possible settings such as hearths and post pads but it was concluded that they were just a stone dump with no structural elements except some larger flat stones amongst them would have proved useful as stepping stones.

The dump of stones F030 in part overlie the black, charcoal rich, fire cracked stone deposit F011. However, the stratigraphic subtleties of this waterlogged part of the site strongly suggest that F030 the stone dump may have been present throughout the life of the burnt mound. Underlying F030 was F034, grey - black silty clay with frequent fire cracked stone and charcoal inclusions and iron panning; a lot of the charcoal had been squashed under F030 stones. Processing of soil samples from F034 produced a hazelnut shell and a flint flake. F034 also underlies the main hearth deposit F072 and the patchy grey silts may also represent ash from rake-out of the hearth F011. A number of stone concentrations within F034 were picked apart to determine if they might represent stone settings of any kind, including a possible post pad F044, but none of the other stones formed convincing individual features. Small concentrations of charcoal and fire cracked stones within F034 suggest individual dumping incidents into the palaeochannel depression formed by edge F108 which contains F030, F072 and F034.

It seems most likely that the palaeochannel was exploited for stones to use for heating and the hearth represented by F072 was located at the drier E end of this waterlogged stony depression. The stones were readily to hand for adding to the fire and heating and when hearth area was cleared the ash and charcoal was raked out into he wet stony depression. The uppermost hearth deposit F011 had also been raked out into the depression but it was not always possible to distinguish between F072 and F011 particularly in this waterlogged area. After the heating incidents represented by F072 and F011 more stones F030 were dumped into the palaeochannel depression F108 presumably for heating the next time the burnt mound was to be used but there wasn’t a next time. The waterlogged stone filled depression interpreted as a palaeochannel, F108 does not have good cut edges but a diffuse variable edge and an irregular suboval shape. It extends over an area of 2.70m NE-SW by 4.40m NW-SE at the SW corner of the site. At the NW corner of F030 in the boggy ground sealed under topsoil F025 was a discrete deposit 0.10m deep and 0.50m diameter of charcoal rich silty clay F071 in a possible cut F073, which cut natural glacial till F061. This feature was convincing in that it was an obvious concentration of charcoal but this area was so waterlogged and the cut amongst so many stones not entirely convincing. It seems most likely that it is a dump of F011 or F072 material.

The deposit of obvious black burnt mound material was F011 (= F015, F047 and F023) which consisted of a soft burnt mottled black, dark red and grey brown silty clay with frequent inclusions of charcoal and frequent inclusions of fire cracked stones measuring 0.06m x 0.04m x 0.04m on average. This deposit spread over an area of roughly 3m N-S x 3.5m E-W and was centred on the highest point of the site, i.e. the top of the mound. It was 0.10m thick on average but at its deepest point, just to the W of the cooking pit F039, it was up to 0.30m deep.

F011 was most concentrated on the top of the mound, just to the S of where the shieling hut cell had been located and in this area it contained up to 40% fire cracked stones. The edges of F011 were rather diffuse but on removal of F011 the underlying deposit F072 had more clearly defined edges. F072 was differentiated from F011 in that it was more compact and contained even more frequent charcoal. F072 was within a natural boggy and stony depression (filled with F034), a palaeochannel, that is F108. It is likely that F011 was to some degree contained within this depression but it had spread over a larger area and as noted above burnt mound deposits tend to be moved to clear space during site use. F072 is interpreted as the central hearth on which the stones were being heated for the boiling.

The fuel would have been burnt on an open hearth, represented by F072 in the depression F108 with some of the earthfast glacial boulders forming a rough back plate and edges to the hearth. The stones selected for heating, collected in the stone dump F030 which also served a dual purpose of making this boggy area easier to work on. Once the stones had been heated they would have been transported to the stone lined cooking trough F039 and dropped into the water bringing it to the boil. The immersion of the heated stones in the water, especially when it was cold at the start of the boiling, would have caused them to fracture. The fractured stones, so abundant in F011 would then have been fished out of the trough before its next use. It is obvious they were not moved far from the trough. Some stone may have fractured in the hearth before being immersed in the trough and some of the fractured stones may have been heated several times over. If one imagines how the heated stones were utilised and moved about it is not surprising that a mound of material built up.

At Allt Fearna the volume of charcoal indicates this was wood but occasional pockets of burnt soil suggest some small amount of peat may have been used, or perhaps turf sods were used around the hearts or for helping to transport heated stones and got burnt in the process.

The stones selected for heating would have been easily acquired from the stream bed and also from the stone dump F030. Excavations at the SW and W side of the site indicated a palaeochannel with a high water table. Although this was a boggy area because of the high water table and the old stream channel there was also an easily available cache of stones for heating available without having to go down the steep bank into the stream.

F011 also overlay F014 which was first recognised under the floor of the cell in the shieling. F014 (= F012 and F020) is compact orange gritty clayey silt with inclusions of angular stones. This layer is interpreted as burning directly on to the natural subsoil F009, which has been trampled into the natural subsoil surface.

The F011 burnt fire cracked stone and charcoal material to the north of the cooking pit F039 overlay a sequence of deposits – F031 on F046 on F097 – that appear to represent upcast soil. F031 was a deposit of pale brown and orange sandy silt with small cobble inclusions (not burnt) which extended over an area of 2.4m N-S x 0.8m E-W. F031 overlay F046 a 0.04m deep deposit of pale grey fine sand interpreted as redeposited natural which extended over an area of 1m NNE-SSW by 0.60m NNWSSE. This is interpreted as upcast from the excavation of post pit F037. F047 in turn overlay a layer of mottled grey and brown silty sand with occasional charcoal inclusions F097. This deposit, F097 extended over an area of 1.30m N-S by 1.60m EW and it overlay the burnt, heat affected subsoil horizon represented by F014, F012 and F020 which overlies the natural subsoil F009. These upcast and redeposited subsoil deposits located to the W and N of the cooking pit appear to represent both the upcast from topsoil clearance before the excavation of cooking pit F039 (represented by F031) and from the excavation of a post pit (represented by F046). They overlie a spread of upcast and redeposited natural F097 which lies on the natural subsoil F009. Further excavation revealed that these upcast and redeposited natural soils were in fact lying up against a scarp cut into the natural mound. It appears that F097 and F031 represents clearance of the ancient topsoil down to the subsoil level before the cooking pit F039 was excavated.

F011 also overlay a deposit of black sand and silt F038 which represents the abandonment fill of the cooking pit F039. A sondage excavated through the spread of black sand and silt F038, which contained a hazelnut shell, revealed that it was the fill of a pit F039, which had a further two cuts to the SE, F043 and F052 which appeared to have held edging stones or planks that have since rotted or have been removed. These cuts have in fact been made by the removal of the edging stones rather than being the cuts to hold the edging stones. Half sectioning pit F039 showed that there were three separate fills; the charcoal rich sand and silt F038, burnt and heat shattered stones F093 and a compact primary fill of dark brown sand and silt that was rich in charcoal F064. One in situ edging stone F076 was exposed on the NW side of the pit and slots for two more removed stones were discovered. Removal of the second half of the pit fill revealed more lining stones that formed two phases of use. During the relining of the pit it was made smaller – reducing it size from 1.50m x 1.10m to c. 0.70m x 0.60m and c. 0.30m deep- with a large sandstone block F074 being placed on the basal fill F064 of the cooking pit. A possible reason for the relining of the pit is that material F058 was being washed into the pit from the NE, the large sandstone block would have helped to stabilise this edge. The basal fill F064 is the compacted material that built up when the pit was in use. Burnt stones F093 that were sitting directly upon F064 made up 40-50% of the fill and are probably the in situ heated stones from the final use of the pit for cooking. It is of note that these were predominantly sandstone which retains heat well. The black sand and silt F038 around the burnt stones F093 is a dump of waste material made after the pit fell out of use which has subsequently washed around the burnt stones. It was after this dump into the pit that the edging stones represented by F060 and F063 were removed.

The sequence of events in the construction of the cooking pit is that the main pit F039 was cut and then the cut F077 for the lining stones was made and the edging stones F076 were inserted and backfilled with F075. The pit was then used and deposit F064 built up in the base. The cooking pit was then relined F074 and made smaller. After which silt F058 continued to wash down the slope and build up around the new edging stones F074. After the final use of the pit for cooking the heated stones were left in place. Once the pit was no longer in use F038 was dumped into it and some of the edging stones were removed leaving the cuts F052 and F043.

To the north of the cooking pit a further sequence of burnt soils were present sealed under the buried ancient topsoil F007. These were F048 silty sand with occasional charcoal flecks extending over an area 0.40m x 0.20m and only 0.02m thick, perhaps slopewash. This in turn overlay a charcoal rich layer of burnt soil and ash F047 which was 0.04m thick and extended over 1.8m N-S x 1.0m E-W. F047 and F048 represent the same phase of activity as F011 although there was no physical relationship between them. Removal of F047 exposed two discrete patches of mottled clayey silt with occasional inclusions of fire cracked stones and charcoal patches F065 (=F066) and F067 (=F068). F065 was 0.06m deep and extended over an area of 1.15m N-S x 0.70m E-W and F067 extended over an area 1.15m E-W x 0.90m N-S and was 0.08m deep. At first it was though these might be individual pit fills but excavation revealed that they were in fact two discrete layers of burnt rake-out material from the burnt mound and perhaps were tipped out of the cooking pit F039. Soil samples from F065 produced a single small pit sherd of prehistoric coarse ware and a hazelnut shell. These deposits had been burrowed into by small animals and a number of possible stake holes were investigated – F069 / F070, F079 / F080, F083 / F084 and F085 / F086 and also a small slot F082 which, like the stakeholes, was sealed under F047 and cut into F065. F082 was 0.19m long and 0.09m wide and 0.06m deep and filled with pale grey brown sandy silty clay with occasional charcoal flecks. It may represent the location of a plank or stone setting. While some of the possible stakeholes were more likely to be burrows the presence of the small slot F082 and the most convincing stakeholes F070, F080 and F085, and also F049 / F050 a stake hole sealed under F007 and cutting natural F009, it is possible to suggest a small shelter or windbreak may have been constructed to the immediate N of cooking pit F039.

It would be logical for the ancient topsoil to have been removed down to the subsoil before the cooking pit was excavated. Although the cooking pit had a stone lining around the edges the base was not lined and it was cut directly into the clay subsoil which would have helped hold the water in the pit. Digging the pit into topsoil would have increased the problems of water seepage out of the pit and the area around the pit would have been muddy, which would have interfered with the cooking and general comfort and ease of using the cooking pit. Indeed the W side of the cooking pit F039 actually cuts the redeposited subsoil layer F031 which did cause problems with slumping and slopewash into the pit. There is no sign of any buried soil under the burnt mound deposits and the rake-out and discarded hearth material to the N and S of the cooking pit lies directly on the natural subsoil.

It therefore seems likely that the natural mound selected by the burnt mound people was stripped of any topsoil down to subsoil level before the burnt mound operations began. This sequence of redeposited subsoil layers also indicates that the substantial post pit F037 was created before the cooking pit was excavated. This is of considerable significance as it indicates that there was a structure of the site that was associated with the burnt mound.

Post pit F037 is a fairly substantial stone lined post hole located at the N of the burnt mound. It was sealed under buried ground surface F007 and cut in to the natural subsoils F045 and F009. The pit has been dug against an earth fast glacial boulder, F096, which has formed the S edge of the pit. The pit is 0.95m long N-S, 0.40m wide and 0.23m deep. It is filled with F036 black to grey brown friable clayey silt with occasional charcoal concentrations. There was also a small stone slab in the base, with some charcoal underneath it, which is part of the lining of the pit. It is possible that stones F005 might have been disturbed packing or lining from this pit. Burnt mound material F012 (= F014 and F020 isolated deposits on natural / scorched earth) may be associated with the use of this pit but there is no clear stratigraphic link. There is an outside possibility that this feature may be a second smaller cooking pit but it is located at the extremity of the burnt mound material at the S end of the site and although the burnt mound deposits have probably been moved around this feature seems rather isolated from the focus of burning. Apart from the charcoal in the fill there is no sign of scorching or burnt stones in the fill

In addition to post hole F037 another post hole F033 was located close by on the W side of the site, outwith the main concentration of burnt mound material but still in an area of scorched earth. The fill F032 was also sealed under the buried soil F007 and cut into the natural subsoil F009. F032 was yellow brown silty sand with grey brown sand lenses. The cut F033 was circular and 0.34m diameter and 0.14m deep.

One other complex feature was located at the SE corner of the site, F055 and F056 was the uppermost fills of this complex feature and the complex, which includes Features F057, F089, F090, F094, F095, F098, F099, F100, F101, F102, F104, F105, F106 and F109 will be described under the general feature number of F056.

The SE corner of the site was a steep slope down to the burn. The subsoil, underlying buried ground surface F007 was quite different in this area and was an orange brown loose gritty peagrit gravel which was between 0.10m and 0.15m deep. It was given context F026 to distinguish it from the F009 subsoil of compact gritty orange sand with occasional pockets of white sand and pockets with higher clay ratios and more cobbles and pebbles in places. This natural subsoil had been exposed across the site by the burnt mound people as their working surface. F009 overlies glacial till of dense cobbles F061. The red gritty stony subsoil was found to overlie pockets of F009 and bedrock and it may be explained in part as weathered bedrock. F026 was certainly not a compact deposit and it was eroded and interleaved with lenses of slopewashed burnt mound material running down the slope from the F011 and F072 hearth area.

This whole SE corner of the site was difficult to interpret due to the slopewashed deposits of black silt, fire cracked stones and larger stones that had apparently been both dumped down the slope and had found their way there through slope wash. The difficulties in understanding these deposits was compounded by the frequent flooding of the excavation area as rain ran off over this steep slope.

Several feature numbers were given out for deposits that turned out to be the same as others and all of the deposits on this slope had been washed there or in the case of the stones thrown away. The final matrix for the SE area F056 complex was F104 orange slopewash (denuded F026 subsoil) over F099 black slope wash over F056 black charcoal rich deposit over F100 stones over F089 pale orange slope wash over F101 stones over F103 natural (=F009). Three samples were processed from F056, two contained charcoal and one was archaeologically sterile. No further post-excavation work is recommended for these slopewashed deposits.

The burnt mound extends over an area of c.7m N-S x 3.5m E-W and c.0.5m deep. It formed a very rough kidney shape curving from the SW round to the NE with the cooking pit F039 located in about the middle of the curve. While the cooking pits are often found between the arms of the crescent there is no burnt mound material to speak of to the E of the cooking pit between it and the steep river bank. The SE corner of the site and the F056 complex of deposits clearly indicates extensive slumping and slopewash down the slope. The stream bank is almost vertical to the E of the cooking pit and it is almost certain that considerable quantities of burnt mound material were thrown over the bank and have been scoured away by the stream over the centuries.

The shieling hut stone cell F010 was located more or less in the centre of the mound where there may have been a depression in the burnt mound material and at the highest point of the natural mound on which the burnt mound is located. Following removal of all the burnt mound deposits it was apparent that the natural low mound, which had an undulating surface, had been scarped into on its E side. This was perhaps to provide shelter for the cooking pit and the burnt mound waste debris lay up against this slight scarp.

Two post holes F037 and F033 and a number of stakeholes suggest there may have been a structure associated with the burnt mound, however there is no clear plan. It is possible the post hole F037 may have held a timber post to which tenting was fixed to provide shelter over the site.
Baker, F., FIRAT Archaeological Services (2008)