Arisaig, General

A non-intrusive archaeological investigation of the Rhu Arisaig peninsula involving 32 separate walkover surveys was carried out between November 2012 and January 2016. The survey was instigated by Elizabeth and Allan MacDonald of Arisaig wanting to find out what evidence was extant on Rhu of people living and working there; historical information pointed to a population of 400 people before the early 19th century and it is now more or less deserted. Ken and Jean Bowker of Rough Bounds

Archaeology were invited to join in on the walkover surveys which were carried out in the late autumn, winter and spring months to maximise the opportunity to record features normally obscured in the summer, in particular by bracken. The MacDonalds and Bowkers were often helped on the surveys by several local residents which meant that there was more chance of recording obscure and ephemeral features. Acknowledgement and gratitude for their input and help is due particularly to Alison Stewart, Judi Budge, Lilian MacDonald, Susan Carstairs, Claire Walters, and everyone else who from time to time was able to join in many hours of walking over rough, boggy and stony ground.

Methodology – The surveys generally involved five to six hours of slow painstaking walking stopping to record and photograph and if possible interpret any features found. Two hand held GPS modules were used along with a compass and a digital camera capable of linking the waypoints

noted and the photographs taken. Hand written notes were made on the spot, recording all the waypoints and features with descriptions and measurements, and these were later transcribed onto computer, with way pointed maps created using the Quo mapping system. To help make the

observations more accessible, an illustrated map to scale was also created using the information on the way-pointed map and the feature records.

Desk top Research – Desk top surveys and research were done to compare our finds with old maps and existing records.

Geology – The bedrock of the area is largely metamorphosed sandstones (psammites and pelites) with exposed intrusive bands of basalt. The sandstone breaks down to poor soils, basalt affords better soil. The basalt intrusions form long ridges averaging 70–100m running N/S with sheltered glens in between the ridges; these are now mainly boggy but evidence on the ground shows that these areas were once better drained. Sea level change over the millennia has left raised beaches between 10m to >25m OD, leaving smoothed stones just under the surface adding to the problems of cultivation. Although the peninsula is low lying, rising only to 103m in the W, the bedrock is so near the surface and the weather so wet, that farming is difficult with the constant run off of nutrients, which have to be replaced every season.

Maps – The Roy Map of 1747 to 1752 shows at least six settlements and some areas of cultivation and woodland. It also shows the large cattle fank/orchard at Druim an Daraich with possibly Clanranald’s house behind. We can see Torbea (Tòrr a Bheithe), Ardgasarig (Àrdgaserie), Sand (Sandaig) and Don Camez (Dubh Chamis – Modern Rhu Farm), also

buildings at Gaoth Dail. The spellings are varying and are usually phonetic, as with most old maps right up to modern times. The positions of the settlements are remarkably accurate, while the delineation of arable ground is often quite inaccurate or does not equate with modern field observations.

John Thomson’s Atlas of Scotland, 1832, is quite small scale but indicates a scattering of buildings/settlements throughout the penisula including Sandaig, Ardgassick (Àrdgaserie now confined to the area by Modern Millburn), Duchamus (Modern Rhu Farm), Dorandrishach (Doire na Drise), Guidail (Gaoth Dail), Drumindarrack (Drium an Daraich), Torramore (Tòrr Mor Gaoth Dail), and Torramore (Tòrr Mor) but not, interestingly, Tòrr a’ Bheithe, although buildings in the general area are noted. Thomson’s map

shows eight named settlements. No arable land is shown.

The 1st Edition OS map of 1873 marks three roofed dwellings still occupied viz Rhu Farm, Porter’s Lodge and Milburn Cottage (the last two now being holiday lets). A fourth roofed building on the current Rhu Farm was then a barn and but is now a converted dwelling and another small

roofed building marked just to the N of the fields at Rhu is now roofless. Nine roofed buildings are marked at Druim an Daraich but now only three exist as houses along with a boat house. Three roofed buildings marked near the road to the W of An Garbh Uillt are now roofless. By Milburn Cottage there were five roofed buildings (now ruins or disappeared) marked just to the N and ‘Kennel’ is noted by them. By 1873 Milburn was occupied by a gamekeeper, but now only one roofed barn exists by the cottage. The large fields at Rhu Farm and the large sheep fank to the N are marked but there is only a scattering of unroofed buildings in Tòrr a Bheithe

and Sandaig and Baile Ùr (this last just E of Millburn but unnamed).

It is likely that of the roofed buildings on the 1873 OS map only seven or eight were dwellings, the rest being byres or barns, illustrating a dramatic decline in population since Thomson’s map of 1832. The 2nd Edition OS map adds the 1885 pier at the road end.

Land holding and population – The peninsula was for some 500 years part of the land holding of the Clanranalds (a sept of the Lord of the Isles) until the sale of Rhu, along with much of the Clanranald lands, in 1827. The steady decline of the population of Rhu, in common with many areas of the Highlands was already set in motion from the middle of the 18th century, caused in part by rents rising rapidly, indicating the huge social change since the 1745 rising. The rents for Dubh Chamus and Tòrr a’ Bheithe

between 1748 and 1798 rose from around 223 merks (Scots merks 1 = 13s 4d) to 360 merks; a rise of over 61%; for Gaoth Dail it was nearly 169% and for Sandaig the rise was even worse, an astonishing 332%.

These rises contrast with rents that had barely changed in the previous fifty years and had probably been stable for centuries. At the same time of rent increases there was a general increase of population in the Highlands which became unsustainable with the collapse of the kelp industry and then the failure of the potato crops. With the introduction of sheep

‘Tenants came to be seen as liabilities rather than assets’.

In 1827 Ranald George, 20th Chief of Clanranald, sold Arisaig including Rhu, to his second wife Lady Ashburton for £48,950. She in turn bequeathed it to her nephew Lord Cranston in 1835. Cranston cleared much of the population from Rubha and by 1842 turned it into a sheep farm. He then sold the land to MacKay of Bighouse, Sutherland, who in turn sold it in 1848 to Francis Astely, a Lancashire industrialist. By 1853, Astely turned most of the peninsula into a deer forest, an area of over 3000 acres which was let out to rich patrons for sport. Rhu Farm and lands to the W of the peninsula were excluded from the deer forest by walls and fences. The selling off of the land in 1827, with the further exchange of ownership and land use, meant that in 25 years the final decline of the population was inevitable by 1853. ‘The replacement of the old indebted families by these new interests had a crucial effect on clearance and it now became possible to put into practice, on a much greater scale than ever before, the policy of compulsory emigration mooted in the 1820s and 1830s. These removals were linked to the provision of assisted passages across the Atlantic for

those who lost their lands’. Not all those evicted immediately emigrated but were instead displaced to fend for themselves.

Statistics gathered from the censuses of Rhu confirms the effect of ownership change. The 1841 census shows the population of Rhu as 134 people. By 1861 there were under 40 residents recorded in only seven households. The accuracy of the censuses in this remote area is not certain but is an indication of the continuing depopulation.

Survey results – A total of 172 sub-rectangular ruined structures were recorded, of which 106 had some walls, the rest either only footings or even just an impression on the ground. The features were of varying sizes, the larger averaging 10 x 4m, many 8 x 4m and smaller structures only 5 x 3m. 117 of the sub-rectangular structures were thought to be former dwellings, while the remaining 55 were probably byres or barns but their use over time cannot be certain. 33 further smaller structures were recorded, some sub-rectangular but most were circular, sub-circular or D shaped, only half of which had some walls. These were thought to be shielings or temporary shelters or stores; most were

43 small circular features often only 1.5m or so in diameter were thought to be cleits for storage but may also have been robbed cairns. Another 35 small features, sub-rectangular or circular, were also initially thought to be storage but were probably used in connection with the kelp industry for

stacking or burning.

133 cairns of all sizes were recorded, the majority thought to be for clearance for cultivation. However, 28 cairns might be prehistoric rather than clearance, but only excavation would help to confirm this theory. Just two cairns, one at Druim an Daraich and the other at Goath Dail have been

positively identified as prehistoric. An incised stone, later confirmed as Neolithic, was found in the wall of a roofless ruin.

Field boundaries and dykes were noted and included on an illustrative map (in the full report) as are the 225 areas of cultivation, some of which merge with each other as at Claigean at c200m long. Others comprise small pockets of rigs of only 20 x 20m. Most of the rigs observed were probably

feannagan, dug using a cas-dhireach or straight spade. As Piers Dixon points out these lazy beds or feannagan ’Unlike plough-rig they are to be found on slopes too steep for a plough team to negotiate.’ Only the lower flatter areas of cultivation would have been created with a horse-drawn

plough. There is also the consideration that some tenants cultivating the small isolated areas of rigs however level, would not have access or money for a horse and plough.

18 areas of peat diggings were recorded also of varying sizes from a few metres long to areas > 40 x 20m. Only 10 piers, ports or slipways were noted but others were probably missed because of the state of the tide at the time of the survey.

The distribution of the features found confirmed that the Rhu peninsula was farmed comprehensively, every bit of land possible being used to the full over time, at least for peat digging or pasture, at best for cultivation. The survey confirmed an earlier report by Jonathan Wordsworth that

certain settlements such as Tòrr a’ Bheithe and Doire na Drise were townships not shielings as described on the modern 1:25000 OS map (1972–8). The survey also noted dozens of smaller and more ephemeral features not recorded on either the modern OS map or on the 1st Edition OS map of 1873.

Although the surveys were carried out when the vegetation was at its lowest, and there over 700 waypoints recording significant features, it is certain that not everything worthy of note was spotted. An area of the peninsula between 2 – Sandaig and 32 – Towards Doire Fhada and especially N and W of 29 – Drium Ban were not surveyed. Deer fencing has led to considerable regeneration of scrub and birch and made it difficult to record features.

We are continuing to discover new features on the Rhu Peninsula each time we venture out—this report is not definitive but goes a long way to appreciating the complex use of the land over time. A full report of the 32 walkover surveys will be lodged with DES/NRHE and Highland HER.

Jean and Ken Bowker – Rough Bounds Archaeology

(Source: DES, Volume 17)