Introduction
A survey of the steep wooded hillside above the north shore of Loch Moidart between An Dun in the west and the Coul burn in the east was carried out during the early months of 2004 and 2005 by S.Evans and T.J.Evans. In the west between An Dun and the mill burn, the area covered stretched from the shore to above the tree line. Between Kinacarra just to the east of the mill burn, a stone dyke runs along the bottom of the hillside. It passes to the north of the house at Kinacarra, the old school house, the farm and the Episcopalian church and to the north of the grounds of Kinlochmoidart house. In this eastern part, the surveyed area extended from the north of the wall to just above the tree line.

Sketch Map showing Survey Area

The terrain was difficult to survey being rocky and precipitous in many places and it is likely that a great deal of archaeology remains unrecorded in spite of multiple visits to some sections. The thick growth of bracken in all open areas makes any survey impossible during the summer period.

Historic Background
After the time of Somerled, Moidart became part of the Garmoran lands that were inherited by the heiress Aimie MacRuari in the 14th century. Aimie became the first wife of John of Isla and their son Ranald inherited her lands in Moidart and Uist, thus this part of Moidart became part of the Clanranald lands. From 1593 the area to the north of Loch Moidart was in the hands of a Cadet branch of the Clanranald MacDonalds that lived at Kinlochmoidart.

The Kinlochmoidart MacDonalds were prominent in their support for the Jacobite cause in 1745 and Prince Charles Edward Stewart stayed at Kinlochmoidart House before travelling via Dalilea and Loch Shiel to Glenfinnan where the standard was raised. Following the defeat at Culloden, Donald MacDonald of Kinlochmoidart was executed and the estates forfeited to the crown. The house at Kinlochmoidart was burned in 1746. The Roy map of 1748 indicates that this original house may have been sited further west than the subsequent houses but local tradition indicates that the old house was just north of the present house and that in 1746, the lady of the house had sheltered under yew trees to watch her house burn.

In 1758 The Kinlochmoidart Estate was surveyed by David Bruce, Surveyor for the Forfeited Estates and the woods listed and valued1 The woodland north of Loch Moidart was listed as consisting of two parts named Craigkenloch (presumably the section of woodland above Kinlochmoidart) and Craigleabegg (probably in the region of Craig Liath Beag) that together were valued at £15. The woods at Kulis (Presumably, Caolas) were valued at £2. and those at Torvicklunton (Site uncertain) at £5. The woods to the south of the river Moidart and the Loch that belonged to the estate were named "Balcraggan" and were valued at £2.

Although the land remained forfeit, Alexander MacDonald, one of Donald's sons, maintained interest in the family lands. He was married to Margaret Campbell, the daughter of Campbell of Airds and although he served as an officer in one of the Highland regiments he visited Moidart frequently2. A tack between Henry Buller, Factor for the Forfeited estates and Captain Alexander MacDonald late of the 42nd Regiment of Foot, " for the Lands of Kinlochmoidart including Glenforslen and Duilet" was agreed on July 9th 17673.The family petitioned for the return of the estate in 1776 and records show that a Peter McTavish, Mason, was engaged in building a new Kinlochmoidart house in 178126 but the MacDonald family did not regain the estate until 1786 when it was restored to Alexander's son John for the payment of £1,111.15s.5d to the government4. The Duke of Argyll appeared to have had some claim to part of the woodland and to the salmon fishing rights. In 1787 the Duke wrote to his "chamberlain", James Maxwell, on Mull asking him to seek out information about both5. The chamberlain's report of 1788 included a reference to a letter from Mr Campbell of Inverscaddle that contained relevant information but unfortunately did not detail the contents. The chamberlain later reported that the whole of the Kinlochmoidart woods had been valued and that they were expected to yield 802 dozen coals (charcoal) 109 tons of oak bark and 840 solid feet of ash timber. To clarify the position James Maxwell visited Moidart with a wood agent and inspected those woods to which the Duke held rights and reported that 'There is not a single stick of wood of any kind on Duillad or Assary and only a little blackwood without any oaks upon Kenlochuachkerach.' He also reported that at 'Badnagrogan Kilismore and Shonabeg' there were few oaks among blackwood that would be expected to yield only 2 tons, 1 cwt. and 2 Qr. of oak bark. The timber was considered to be too scattered to be worth felling for charcoal production. He mentioned that 'on some other farms of the estate, particularly Kanloch, Kilisbeg or Kiliscolta and part of Bronarie there is a close thriving stool of oakwood of considerable value'.

By April 1791 the Lorne Furnace Company was interested in purchasing the Kinlochmoidart woods but wished to be clear about the ownership6. The company agent at Bonawe, J Harrison, wrote to the Duke of Argyll's factor, Donald Campbell on the 4/4/1791 stating that he had been instructed "to enquire after the Kinlochmoydarts woods if they wear to be disposed off." He went on to say that he had made some enquiries and had been told, "…part belongs to His Grace, The Duke of Argyll but nothing conscionable". He then stated that " I am told that you will be able to give some information concerning Kinloch Moydart's woods and if to be sold, who to apply to, as the company could wish to take a look at them if agreeable to the proprietor."

The Bonawe letter books indicate that the company usually negotiated private sales with individual landowners. However, the disposal of the Kinlochmoidart woodland was different. Among the Robertson MacDonald Papers is a copy of the draft advertisement for the sale of the woods "by public roup at Portappin on the 29th day of October 1792 within the house of Donald Campbell, Vintner there, between the hours of twelve and one afternoon."7 Although I have not yet been able to find any contract for the sale of the woods, that they were purchased by the Lorne Furnace company is confirmed by a copy of a letter written by the company agent on the 7th January 1793 to Mr Campbell, who was acting for the proprietors.8 The letter states that their wood agent, Mr. Satterthwaite, "… was last week at Moidart with the work people to begin cutting the woods so that as the company has taken possession of the woods it will be necessary for the company to have your letter as agreed upon at the roup giving liberty…..for six to nine months after manufacturing to clear the grounds and the shores of the produce - and likewise for sufficient horse grass convenient, if the enclosure pointed out be found not sufficient for all the number of horses that may be necessary to transport the produce of the woods to shore." Later in the year on the 11/11/1793 a bank draft for £405 "being one half of the purchase of the woods at Kinloch Moydart-payable by Arbiter of Roup Martinmas 1793" was sent from Bonawe to a Mr MacDonald, Writer to the Signet, Edinburgh, who was probably representing the interests of the Kinlochmoidart family9.

Not only were the trees at Kinlochmoidart cut for charcoal production and for the sale of oak bark, but a letter written by the company agent to a gentleman in Ballycastle Quay, Northern Island over one year later, dated 1/4/1794 indicates that the large trees were felled and sold as building timber.10 In the letter the agent says, "I observe you wish to know if we have any wood for sale-We have for sale at Moydart, where you was before, both Ash and Oak which would sute for house-building casepols cabbening etc. The ash is remarkably good." Local tradition suggests that timber was processed and taken off by boat from Kinacarra. Angus Peter MacLean, who now lives at Kinacarra has found that the ground in the field to the southeast of the house is full of large pieces of timber.

A note about the management of woods in the vicinity of the farm was made in 179911. It was stated that there were "mostly oak with some ash birch sprush and hazel" and there was concern about the poor state of walls and fences so that the sheep were getting in and damaging the wool on the undergrowth.

In 1799 plans were made for improvements to the farm of Kinlochmoidart12 but the woods to the north of the loch were not mentioned. One of the suggestions for improvement was the construction of the mill. "A mill is much wanted on the estate where it formerly stood where there is a good source of water." The suggestion was then made that attempts should be made to obtain the "works" from the derelict "sleat" mill on the opposite side of Lochmoidart on the Clanranald Estate. Since there is no sign of any machinery in the ruined mill on the south side of the loch the plan was probably carried through. The present mill (now a private house) was therefore built on the site of a former mill soon after 1799 and its machinery was probably transported across the loch from Port a' Bhàta. The mill was operating in 1841 when a miller and his apprentice were recorded in the census. The 1875 first Ordnance survey map shows a mill house (presumably a habitation for a miller) at the site of the house at Kinacarra. The house at Kinacarra may therefore have been built at the same time as the mill in the early C19th but was improved by the architect William Leiper in the 1890s. It is quite possible that the limekiln near Kinacarra was also built at around the same time since lime was needed to improve the land.

A report by the surveyor John Blackadder written in the Spring of 1803 outlines facts about canalising the river Moidart and the use of shell sand from the mouth of the north channel but does not mention the area surveyed.

In 1804, John MacDonald died without issue and the estate passed to his paternal aunt Margaret MacDonald who was married to Lieutenant Colonel Robertson. During the years that followed, there was general hardship on the estate. During the period from 1801 to 1841 there was an increase in population of 25% on the western mainland of the highlands.13 Tenants emigrated to Canada and Australia during the first half of the C19th but after the failure of the potato harvests in the mid 1840s large-scale emigration took place. Many tenants from the Kinlochmoidart estate left for Australia in the early 1850s on the vessels Marco Pollo, Araminta and the Alison. The estate became mainly a sheep farm.

The Robertson MacDonalds sold the surveyed area to the Stewart family in 1882 as part of the Kinlochmoidart estate. The architect William Leiper designed the present Kinlochmoidart house for the family in 188414 and other estate houses were built or improved in the same style in the following years. The large Kinlochmoidart pier was built to bring in the dressed sandstone used in the construction of the house but the 1876 Ordnance survey map shows that a pier already existed at the site. In 1899 the numbers of sheep were reduced and the land was managed as a deer park. The estate is still in the possession of the Stewart family who have restored the house and run a self-catering holiday business.

The age of the old track, that linked Kinlochmoidart and Glenuig before the modern A861 was opened in 1966, is unknown. Since the path passes close to the Iron-age fort of An Dun it is easy to believe that it existed from early times but documentary evidence has proved difficult to find. In the C19th and early C20th it was used to drive livestock to and from the market at Salen15. The brothers Archie and Angus MacDougall of Egnaig were employed to maintain it in the first half of the C20th16.

SURVEY FINDINGS

A. The Woods
Throughout the western part of the surveyed area the trees are mainly mature oaks, of roughly similar size, indicating that they may have been planted at about the same time at some point in the C19th. Although an occasional coppiced tree can be seen, most show no evidence of coppicing.

The oaks are interspersed with scattered birch, rowan, holly, and hazel. Hazels grow mainly along the watercourses and along the sides of the old Glenuig to Kinlochmoidart track. Alders appear in wet areas and birch predominates at higher levels near the tree line. It is interesting that there is no trace of any fence or dyke surrounding the woodland to protect young trees from grazing animals. There is a stone dyke topped by wire that stretches from the modern road northwards via Loch Ard a' Phuill and Lochan na Cloiche Sgoilte to meet the boundary between the estates of Kinlochmoidart and Roshven. The reason for this dyke has been lost but it may have delineated a boundary between common grazings on the Kinlochmoidart estate.

Fig 1. Non-coppiced mature oaks

To the east of the Mill burn (Allt a' Mhuillinn) and just to the northeast of the house at Kinacarra there is a roughly triangular area of open land that is covered in dense bracken and only scattered trees. It is separated from the woodland by the steep sided mill burn and a series of dykes extending from the modern roadside near the lime kiln to meet a steep escarpment to the west of the limekiln burn. The bracken and the presence of clearance cairns indicate that this land was used for cultivation. There are several pollarded oaks in this area that indicate that it was also used for grazing. The woodland to the east of the Limekiln burn becomes more mixed, with scattered growth of oak, birch, rowan, holly and occasional ash and beech trees as well as alder and hazel along the watercourses. There are also a few small Scotch Pines as well as some non indiginous pine trees . There are several oak pollards on the hillside above the farm. Rhododendron growth has become a problem to the north of the church. In the part of the survey area to the north of Kinlochmoidart house, the hillside is more open with a mixed growth of oak, birch, alder, rowan, holly and hawthorn. The woodland flora over this open area indicates that there was probably denser tree growth in the past.

B. The Old Track ( From NM68048.72864 to NM 69277.73057)
The western part of the survey area is traversed by the remains of the old track that linked Kinlochmoidart and Glenuig before the Modern road was built in 1966. The part of the track that climbs the steep slope to the west of An Dun has been damaged by water and motorcycles. At various points the track has been carefully built up with a dry stone wall and in other places stones have been cleared and banks cut away to allow the track to pass through a gully.

"Borrow-pits", from which material has been taken to improve or maintain the track, can be seen at intervals alongside it. At one point the track crosses a possible recessed platform.

At a point to the west of the watercourse that descends to the west of Meall na h-Uamh the track descends a steep hillside in a series of hairpin bends.

Fig. 2. view of built up section of track

A branch leaves the track to descend in a south-westerly direction to a small stone jetty (NM69211.73054) Another branch once left the main track to descend by a more direct north/south route towards the jetty. At a bend in the track between these two branches, a track once continued to the east and is now cut by the new road between the cave and the large pier. There is some evidence that there was once a branch from this track that ascended to the northeast. Where the main track is now accessed from the road, it once appeared to continue at a slightly higher level.

Most of the burns are crossed using flat stones laid side by side and some sections have been paved.

Fig. 3 (above). Track crossing Allt Aileen Fig. 4 (right). Paved section of track

C. Ruined Buildings.
1. A remnant of a rectangular building sited between two springs on the west side of Allt Aileen (Allt Allein) Map Ref: NM68694 73382.

There are two enclosures surrounded by the remnants of stone and turf dykes on the opposite bank of the burn.

2. KLM9. Small Ruined dry-stone building on level ground above and to the west of the dyke and the burn that descend the hillside to the west of Meall na h-Uahm. (Map Ref: NM 69051.73412.) Nearby there are outlines of a shelter. On the opposite side of the burn there are outlines of other subrectangular shelters and two shallow pits.

3. Remains of a house and outbuilding to the east of the Mill Burn. (Map Ref: NM 69830.72901). There are three other ruined outbuildings nearby and the ruin of what could have been another house that appears to be sited just to the north of a platform. There are lazy-beds in the area to the northeast of the house and an enclosure that could have been a kail-yard.

4. Probable ruined house and byre.
Map Ref: NM 71509.72634.

D. Shieling huts.
Outlines of former shieling huts were found above the treeline near Allt Allein, Allt a' Mhuillin, Allt na Creige Leóbaiche and the Quarry burn.

E. RECESSED PLATFORMS
THE MOST STRIKING FINDING ON THE STEEP WOODED HILLSIDE WAS THE LARGE NUMBER OF RECESSED PLATFORMS.

Seventy-one platforms were identified and a further sixteen possible platforms were found.

The definitely identified recessed platforms were all very similar. All had been made by excavating an arc into the hillside and piling up the excavated material on the slope below to make a roughly circular or oval level area. The front edge was usually strengthened by a semicircular retaining wall made of roughly piled stones or boulders or a mixture of stone and earth.

Recessed platform Map Ref: 69066. 73289 Fig. 5	Fig. 6. Recessed platform to the north of Kinlochmoidart House
Fig. 7. Platform east of the Limekiln Burn

The retaining walls are sometimes neatly built as in the example in Fig. 6. but others are roughly built of large boulders as in Fig. 7.

Because the hillside is south-facing, nearly all the platforms (sixty) face south. Some that have been cut into the banks of larger burns face the burns, and others that are cut into ridges extending out from the hillside, face southeast or southwest.

Figure 8

The size of all these levelled areas is remarkably similar. The widths (i.e. the diameter across the platform in the direction of the contour of the hill) vary between 5.7m. and 10.3m. but most are between 7m. and 8m. The mean width is 7.62m. and one standard deviation from the mean is only 0.888. The diameters from the back of the platforms where they are recessed into the hill to the front edge vary from 5.2m. to 9m. Most measure between 6.5m. and 7.5m. The average is 7.12m. and one standard deviation is only 0.789.

Figure 9

The depths of the recessions into the banks to the rear of the platforms vary from 0m. to 1.7m. and the height of the anterior retaining walls vary between 0.3m. and 2m. These measurements depend upon the gradient of the incline of the hill but do not necessarily give a true impression of the actual steepness of the terrain because the hillside is extremely rocky and irregular with many natural vertical cliffs. Although some of the platforms may have quite low front retaining walls, there may be natural vertical cliffs descending only metres to the south of the front of the platform. An example of this can be seen in Fig. 6. where the tops of trees can be seen below the platform in the right of the photograph. These dimensions do give some indication of the work involved in the platform construction.

Most but not all are built near a water supply. The surfaces of a few have been damaged by small watercourses and on a very few of these charcoal can be seen.

Figure 10

The distribution of the platforms is illustrated on the map below.

In general they appear to be scattered across the hillside. Although some appear to be built along contours with animal tracks linking them, they do not appear to be in groups. They have been constructed between altitudes 30m. and 218m. above sea level. The majority lie between 50m. and 200m. The mean altitude is 114.5m. Those at the higher altitudes are at or near the present tree line. No recessed platforms were identified above the tree line although many sites of small shieling huts were identified.

Figure 11	Figure 12

The Origin of the recessed platforms
In many parts of the world, since early times, man has created level areas of ground on which to build or work on sloping terrain. Platforms are now made using heavy earth moving machinery whereas in the past manual labour was required, however the form of the recessed platform has changed little.

Many unenclosed platform settlements consisting of small groups of platforms cut into hillsides have been identified and recorded in different parts of Scotland such as Midlothian, Roxburghshire, Clydesdale and Peebleshire. The size of these platforms varies considerably but on average they are significantly larger than those in the Kinlochmoidart woods. Most have diameters between 13m and 17m. Excavations, such as that at Greenknowe near Peebles,18 have established that such platforms accommodated Iron-age round houses.

Large numbers of platforms similar to those in the Kinlochmoidart woods have been identified throughout the western mainland of Scotland, from Kintyre and the Cowal peninsula in the south to Loch Garry in the north. Undoubtedly, many more are yet to be discovered. Most occur in large scattered groups. These have been regarded as being recessed platforms constructed and used for the manufacture of charcoal in the 18th and 19th centuries and many have been recorded by the Royal Commission for Ancient and Historic Monuments Scotland as charcoal burning stances19.

Charcoal burning stances
Until the 17th century, iron production in the Scottish Highlands took place at small bloomeries, and the iron produced would have been used locally20. The charcoal-fired blast furnace was introduced to Scotland about 1600 and over the following 150 years such furnaces operated at Gairloch, Glenkinglass on Loch Etive, Culnakyle and Invergarry. All were unsuccessful and shortlived. In 1752 the English Iron Masters Richard Ford, William Ford, James Blackhouse and Michael Knott (The Newland Company) agreed a contract with Sir Duncan Campbell of Lochnell for the long term lease of land at Bonawe on Loch Etive on which to erect the Lorn furnace and Iron Works21. The company was attracted to the west of Scotland because of the availability of deciduous woodland suitable for charcoal production. As part of the agreement, the Newland Company gained the use of extensive woodland on Lochnell's estate. In the same year, the iron masters also entered into agreements with the Earl of Bradalbane for the purchase of wood on large tracts of his estate. These woodlands proved to be inadequate and two company letter-books from Bonawe, that are kept in the National Library of Scotland and cover the period from 1786 to 1813, show that the company purchased wood from land-owners over a large area of the western Highlands.

These copied letters not only provide information about individual business transactions but also give information about the process of harvesting the woodland. The operation was run by a company agent based at Bonawe. He directed the activities of a wood-agent and his assistants. A few master "colliers" were contracted to cut the timber and produce the charcoal. These men employed their own gangs of labourers. Woods were "valued" by the wood agent and his assistants and an estimate of the quantity of charcoal that they could yield was made before the company negotiated the purchase. Nearly all types of timber except pine were utilised. Oak was coppiced and would be reharvested at 20 to 30 year intervals. There was a good market for oak bark in industries on the Clyde and elsewhere so that oak was usually felled in early summer when the bark could be easily stripped. The felling of other species such as birch (known as black wood) usually started in January. The individual purchase contracts included provision for the construction of huts, saw pits, "roads" for transportation of the charcoal and where necessary for piers or shipping banks. The contracts all contain agreements for the construction of "pitsheads" or "charcoal hearths" but unfortunately their exact nature is not described. It seems likely that the "pitshead" would be a stance for building the wood stacks for firing. On steep or uneven ground these stances would be recessed platforms. Since these pitsheads are likely to have been made by the same teams of men throughout all the woods purchased by the company, they would probably all be similar in pattern. A letter written by the agent at Bonawe, J Harrison dated 24/ 03/1798 indicates that the pitsheads were permanent structures22. The company were about to harvest woods at Nabdale for a second time. J Harrison was concerned that during the later part of the period of regrowth, the estate tenants had been "plowing" land in the woods. He wrote to the estate factor, Mr. Patterson at Inverary warning that " the company will use the pitsheads, roads and shipping bank this year even if a crop of corn is on them". It is interesting that the surface of one of the recessed platforms in the Kinlochmoidart woods has lazy-beds upon it.

The woodstacks for "coaling" probably consisted of logs from 50 to 200mm thick carefully arranged around a central stake and covered with turf23. The whole stack would measure about 4.5m. in diameter at the base.. The stake would be removed and fire introduced in the central hole. This would then be covered and the slowly burning stack tended for between 2 and 10 days, depending on the condition and dampness of the wood. Wind breaks were errected around the stack. The colliers would have needed some shelter during the period of combustion and would probably have erected a small hut of wood and turf. With diameters of between 5.2m. and 10.3m. most of the recessed platforms at Kinlochmoidart could have accommodated at least one charcoal hearth and a hut. A few of the platforms had traces of hut stances adjacent to them.

Fig. 13. Charcoal burners at work in the early C20th	Fig. 14. Sketch of a Charcoal hearth surrounded by a wind-break as seen in a photograph in the Hutton Getty Collection.

The quantity of charcoal was measured in "dozens" consisting of twelve sacks, each measuring 7ft 6ins. by 3ft and filled to a depth of 5ft 4inches24. The sacks were taken to the shore by ponies and loaded onto small coastal vessels to be transported to Loch Etive. Local supply of fodder for the horses was often a problem.

Large mature trees were unsuitable for charcoal production but it is known that they were felled and sold as building timber.

Stances for timber-framed huts
Ten of these west highland recessed platforms at different locations were excavated in the 1990s and at each site a circle or circles of post-holes were found. Elizabeth Rennie, who directed the excavations, concluded that these indicated that the platforms had originally been stances for timber-framed huts25. No artefacts were found at any of the sites and dating evidence consisted of carbon dating of samples from only five sites and from datable features overlying the platforms. Elizabeth Rennie concluded that the available evidence indicated that some of the platforms were in use in the Medieval or Early Historic periods. At least one of the excavated platforms had been also used as a charcoal hearth in the 18th century. It is therefore possible that some, if not all, of the very numerous west highland recessed platforms are early hut stances and that some were reused for charcoal production in the 17th and 18th centuries.

However, it is evident from photographs e.g. Fig. 14, that charcoal hearths could also be surrounded by a circle of post holes supporting wind breaks, therefore the reliability of the dating evidence is extremely important.

Early Historic or Medieval hut stances or charcoal burning stances in the Kinlochmoidart woods?
Position
The site of An Dun at the west end of the surveyed area shows that the north shore of Loch Moidart was occupied in the Early Historic Period.

The recessed platforms in the area surveyed at Kinlochmoidart were scattered over wide areas throughout the east and west parts of the woodland. They were absent on land immediately to the east of Allt a Mhuillinn that showed evidence of past cultivation, and few were sited in the area between Allt a Mhuillinn and the old stone jetty to the west. Although they did not occur in small groups as may be expected if they had been settlements, some were just within sight of others and this may support the idea that there may have been two settlements of widely spaced dwellings. Very few were sited near any land that could have been used to grow crops. Some of the platforms to the east of the Allt a Mhuillinn and those at low levels between the limekiln and Orchard burns would have been well sited to use the available cultivatable land on the flood plain of the river and the better land between Allt a Mhuillinn and the limekiln burn. All the others would have been inconveniently sited for farming.

All the recessed platforms were sited below the tree line. Although some were near the limit of tree growth, none were found above 218m. The hillsides above the tree line were surveyed, but only the remains of small subrectangular shelters, interpreted as shieling huts, were found clustered near the burns. These were usually sited on small patches of naturally flat ground and not on recessed platforms. This may indicate that the recessed platforms were built within the woodland for the manufacture of charcoal but could also indicate that the people of the Early Historic Period preferred to build within the woodland where building material was readily available and the trees afforded some shelter.

Many of the platforms were built in exposed positions on very steep terrain adjacent to precipitous cliffs. Anyone living on these platforms would have had excellent views of anyone approaching from the sea, but their homes would have been easily visible to invaders. It is difficult to understand why such sites, that would have been inconvenient to climb to and dangerous for young children, would be chosen for dwellings. It seems more likely that such platforms would have been constructed for charcoal production in the late 18th century, although it can be argued that these sites would also have been inconvenient for the "colliers" and that exposure to high winds could have made the controlled burning process more difficult.

All but a very few of the platforms were sited near a water supply. Some were built adjacent to water. If used as dwellings, the proximity of fresh water would have been advantageous. It would also have been useful for the production of charcoal where it was used to douse the fire on opening the charcoal clamps.

Platform morphology
All the definitely identified recessed platforms were remarkably similar in form. The majority have a stone retaining wall at the front (downhill) edge of the platform but the arc cut into the bank at the rear of the platform has been retained by a wall in only two. There was little variation in size. The similarity in appearance suggests that they were all constructed at the same time for a similar purpose and would support the argument that they were all produced for the industrial purpose of charcoal burning in the late 18th century. If the platforms were originally made as hut stances, a greater variation in size might be expected to provide for variation in the size of buildings made for different purposes. A "scoop" in the hillside providing a stance for a small shelter was found near only one of the platforms. However, it is possible that recessed platforms of various sizes were altered by the colliers of the 18th century to meet their requirements. The fourteen areas of even ground that were identified as possible platforms may represent original platforms that were unaltered, but they could also be naturally occurring flat areas used for a variety of purposes both human and animal. Locally it is evident that spots where cattle like to stand often become flattened by repeated use.

Charcoal is evident on a few of the platforms that have been eroded by water. The surface of other platforms was not disturbed to search for charcoal, but since there is no doubt that the woods were felled for charcoal, it is likely that most would have been used as charcoal hearths whatever their origin.

Numbers
If all the seventy one definite recessed platforms were originally stances for early timber buildings, this would indicate that the population of the area was quite high. Even if only one third of the buildings were dwellings, thirty households in this inhospitable area would be surprising. However, it is possible that the stances date from different periods and could represent different eras of human occupation.

Such a large number of platforms may well have been required for charcoal hearths.

Overlying features
Lazy-beds cover the surface of one recessed platform. The cultivation probably took place in the early 19th century when the local population had increased and a great deal of marginal land was utilised for crops.
A ruined stone house stands behind the north edge of another platform and probably was built after the platform. Prior to the very end of the 18th century and the beginning of the 19th century, most houses in the area were constructed of timber and turf. The house probably belongs to the early 19th century.

One recessed platform has been cut into by a borrow-pit for the track to Glenuig. Another possible platform appears to be crossed by the track and has an excavated shallow hole in the south part of it. Although it seems possible that the track through the woodland, which passes close to An Dun, could have been in use in the early Historic Period, the original route is uncertain. The track that is visible today was in use and was maintained until the mid 20th century. It would have been improved on many occasions and relatively modern work could account for the "borrow pit" cut into one platform and for the excavation in the other.

CONCLUSION.
There are arguments in favour of at least some of the recessed platforms being originally hut stances from the Early Historic or Medieval Periods but the vast majority are quite small to accommodate substantial roundhouses. These woods were certainly felled for charcoal production in the last decade of the 18th century. Most or all of the platforms would have been used for charcoal hearths. All could have been constructed specifically as charcoal burners' platforms but it is also possible that the colliers reused ready-made platforms that they found in situ. Perhaps there is a mixture of early recessed hut stances reused for charcoal hearths and platforms made specifically for charcoal production in the late 18th century. Excavation would be necessary to find out if any of the platforms date from the Early Historic or Medieval Periods.

References

1. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3981 f4-7
2. MacDonald, Charles. Moidart or Among the Clanranalds. Birlinn 1997. Ch 11. p189.
3. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3983
4. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3981 f31-32
5. Argyll Estate Instructions 1771-1805 Edited by Green, Eric. Scottish History Society 1964
p 145-154.
6. Lorne Furnace Company Letter Books. National Library of Scotland. MS995 f3.
7. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3983 f96
8. Lorne Furnace Company Letter Books. National Library of Scotland. MS995 f16.
9. Lorne Furnace Company Letter Books. National Library of Scotland. MS995 f18.
10. Lorne Furnace Company Letter Books. National Library of Scotland. MS995 f25.
11. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3983 f125
12. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3945
f126-129.
13. Devine, T.M. Clanship to Crofters' War. p45.
14. Jefferson, Stephen. Kinlochmoidart House.
15. Wood, Wendy. Mac's Croft.
16. Wood, Wendy. Moidart and Morar p145.
17 Sanderson, Stewart F. A Packman's Bivvy, Scottish Studies 1957, Vol I p243.
18. Jobey, George. 1978-80 GreenKnowe Unenclosed Platform Settlement and Harehope Cairn
Peebleshire Pro Soc Antiq Scotland 110, 72-113
Feacham, R.W. 1960-61 Unenclosed platform settlementsProc Soc Antiq Scotland 94,
79-85
19. RCAHMS Argyll II No.361. p281
20. Lewis, John H. The charcoal-fired blast furnaces of Scotland: a review. 1984 Proc Soc Antiq
Scot. 114, 433-479.
21. National Library of Scotland MSS993 f1-16
22. Lorne Furnace Company Letter Books. National Library of Scotland. MS993 f56.
23. Bonawe Iron Furnace, Historic Scotland 2003. p 8.
24. Clanranald Papers. N.A.S. GD201/1/281.
25. Rennie, Elizabeth B. 1997 The Recessed Platforms of Argyll, Bute and Inverness. British
Archaeological Reports, British Series 253, Oxford.
26. MacDonald/Robertson MacDonald Papers. National Library of Scotland. MS3983.

Acknowledgements

I wish to thank the landowner, Nino Stewart, for her helpful advice and Angus Peter MacLean of Kinacarra for providing valuable local information.
The map of Moidart is reproduced with the kind permission of Collins Bartholomew 2006.
A Map was also obtained from the O.S. Web Site Get a Map Service and is reproduced by kind permission of Ordnance Survey.