Galmisdale House and the
Sgurr.
(SC729680)

The survey of the archaeology of Eigg by the Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS) has identified and mapped over 1150 individual structures. Processing this body of data is currently underway and, once it is completed, detailed descriptions of all recorded archaeological sites will be available from the National Monuments Record of Scotland (accessible through Canmore). In the meantime, this interim account summarises and illustrates the more important aspects of the extensive and often spectacular archaeology of the island.

The Small Isles comprise four inhabited islands - Canna, Rum, Eigg and Muck. They lie to the south of Skye, astride the sea routes running up the west coast of the mainland and to and from the Western Isles. From both the sea and the mainland their distinctive outlines make them instantly recognisable landmarks, especially the forbidding peaks of the Rum Cuillin and the dramatic Sgurr of Eigg.

The islands are largely composed of igneous rocks formed during a period of intense volcanic activity beginning about 60 million years ago and lasting around three million years. In the northern half of Rum, however, Torridonian sandstones predominate, and small areas of Jurassic sandstones, mudstones and limestones are visible around parts of the coasts of Rum, Eigg and Muck. The three smaller islands are largely formed from a series of lava flows, chiefly of basalt and mugearite. These have weathered into gently sloping terraces bounded by vertical cliffs, and the rocks readily break down to provide a fertile soil. Towering above the southern end of Eigg, the Sgurr is a block of pitchstone - a cooled lava flow which has filled a river valley cut deeply into the basalt, and has now been left standing proud as the softer valley sides have eroded away.

The igneous rocks on Rum have a different origin; these are mostly intrusive rocks that did not break the surface but cooled deep within an enormous volcano, now eroded down to its roots. Neither these rocks nor the hard-wearing sandstones of the northern half of the island have supplied the nutrients essential for the development of fertile soils; this factor, combined with Rum's high rainfall (about 300cm a year at Kinloch), has produced a landscape dominated by wet heathland, limiting settlement and cultivation to a few more favoured locations.

Apart from the brief notices of 17th- and 18th-century travellers, the authors of the Statistical Accounts and the notes in the Ordnance Survey Name Books, the earliest account of the archaeology of any of the Small Isles is that by Norman MacPherson, proprietor of Eigg. His paper of 1878 on the antiquities of the island included a report of an excavation he initiated of the Viking graves at Kildonnan, as well as details of a number of earlier discoveries, mostly made during land improvement operations. The islands also came to the attention of several other late-Victorian scholars, particularly those attracted by the early medieval crosses on Eigg and Canna. The first attempt at a systematic survey of the islands, however, was the Royal Commission's visit in 1925. Fieldwork for the Inventory of The Outer Hebrides, Skye and The Small Isles (1928) began in 1914, but it was only at the tail end of the project, which had been disrupted by the war and financial stringencies, that the Small Isles were visited, all recording apparently being completed during a single week in early July 1925. Fifteen monuments or groups of monuments were recorded, most of them on Canna. On Rum only one site, Kilmory church and burial-ground, is described, while on Eigg only the fort on the Sgurr and Kildonnan churchyard appear to have been visited, both on the same day.

For the next half century the islands attracted little archaeological attention. The Archaeology Division of the Ordnance Survey spent a month in the summer of 1972, principally to visit previously known sites, though several new monuments, including several hut-circles on Eigg and Canna, were mapped. A study of shielings on Rum by Ronald Miller in the 1960s, followed by John Love's more detailed work on the same topic, drew attention to the archaeological potential of that island, and in 1983 the Royal Commission returned to Rum and published a handlist of archaeological sites under the auspices of the Society of Antiquaries listing programme. The discovery of Mesolithic flints near Kinloch Castle in the course of that survey prompted a programme of high-profile excavations by Caroline Wickham-Jones in the following year. On Eigg, Peter and Susanna Wade-Martins, together with colleagues from Norfolk, completed surveys of a number of sites during the 1980s and 1990s, while in 1986 the Scottish Buildings Study Group at Dundee University surveyed farm buildings on both Eigg and Canna. However, it was not until 1994 that the Royal Commission returned to the Small Isles, embarking on a survey of Canna in partnership with the National Trust for Scotland. A broadsheet on the Canna survey was published in 1999, and a longer account has been incorporated into the fourth edition of the late John Lorne Campbell's Canna: The Story of a Hebridean Island. Further work on the Small Isles includes Muck, surveyed in 2002. Meanwhile, the Threatened Buildings Survey of the Architecture Division of the Royal Commission completed an in-depth photographic record of Kinloch Castle in 1996 and, in 1998-9, the principal buildings on Eigg were photographed and surveyed. Finally, all the early crosses on the islands were drawn for Early Medieval Sculpture in the West Highlands and Islands, published by the Royal Commission in 2001.

Other archaeological organisations have also been involved in the islands in recent years. In 2001 Glasgow University Archaeological Research Division conducted an assessment of the archaeology of Rum for Scottish Natural Heritage. In the same year, stimulated by the Royal Commission survey, the National Museum began excavating a Late Bronze Age metalworking site on Eigg, while an assessment has been made of a wooden boat, possibly 18th century in date, buried in the sand in Galmisdale Bay. Finally, excavations in advance of new pier developments on Eigg and Rum have produced quantities of worked flint.

Rock shelter beneath a
boulder on the east coast
of Eigg. A shell-rich midden
extending downslope from
the entrance is identifable
from a rich growth of flag
iris. (SC759497)

Photograph of a 'Shetland-
type' house at Galmisdale,
possibly belonging to the
Neolithic period.
(SC722842)

The Mesolithic (8000-4000BC)
In the 1980s Rum achieved fame as the home of 'Scotland's First Settlers' following the excavation of a Mesolithic settlement in a field at the head of Loch Scresort, to the north of Kinloch Castle. Radiocarbon dates from the site indicate occupation from about 7500 BC, and until recently these were the earliest dates for settlement anywhere in Scotland. Current work on Applecross and the east coast of Skye suggests widespread Mesolithic activity there, but apart from the chance discovery of stone tools by forestry workers on Eigg no other material of this period has so far been recorded on the Small Isles. Eigg, however, offers some potential here amongst the many rock shelters concealed in the boulder screes beneath the cliffs on the east coast of the island, some of which have substantial shell middens spreading out from their entrances.

The Neolithic (4000-2500BC)
Evidence for the Neolithic period is rather more plentiful. During the Canna survey large quantities of pottery of this period, including Unstan Ware and Grooved Ware, as well as Early Bronze Age Beaker sherds, were discovered in the upcast from rabbit burrows dug into a series of large earthen mounds, probably the remains of settlements. Earthen mounds have also been recorded on Eigg and Muck, but burrow spoil on Eigg (there are no rabbits on Muck) has produced only post-medieval pottery. While it is possible that some of these mounds also have prehistoric origins, it is perhaps safer to see them as the remains of turf shieling-huts of medieval or later date. On Rum, the excavations at Kinloch demonstrated that occupation continued through the Neolithic period, while stray finds on Eigg include two Neolithic stone axes, a flint spearhead and two leaf-shaped arrowheads (the latter discovered during survey in 2001). Thumbnail scrapers, probably of this period, were also picked up on ploughed fields during the Eigg and Muck surveys. What is lacking, however, are any certain examples of burial or ritual monuments of this period, the only two possible examples being a heavily robbed cairn on Canna and the site of a 'long cairn' at Laig on Eigg, removed in the early 1850s to provide packing for field drains. Two stone cists within the latter cairn were said to contain 'articles of stone and bone' including, perhaps, one of the two stone axes mentioned above. One final site that may belong to the Neolithic period is an unusual oval structure near Galmisdale on Eigg, whose closest parallels are to be found amongst prehistoric houses on Shetland.

Hut-circle at Cnoc
Smeordail. (SC759499)

Draft plan of two hut-circles
at Howlin, one
superimposed on the other.

The Bronze Age (2500-700BC)
About three dozen round cairns, presumably of Bronze Age date, are distributed across the four islands, a dozen of them on Eigg. Most of these survive as simple circular stony mounds, though some have visible traces of a kerb of large stones. A cairn at Kildonnan on Eigg is the only one to have been formally excavated, in 1875 (a 'cairn' excavated by T C Lethbridge on Canna in 1924 is now thought more likely to be a shieling-mound). The finds from the Kildonnan excavation belong to a Viking burial, but this had been placed in a ruined cist that is probably prehistoric. Three other cists, each within a cairn, were discovered on Eigg during land improvements in the early 1860s, but there are no records of any finds being recovered from them.

The vast majority of hut-circles in the Small Isles are on Eigg, where thirty have been mapped, and on Canna, where there are twelve. By contrast, only one has been identified on Muck and none of the few suggested examples on Rum are convincing. The visible remains comprise low, circular, stony banks, often faced with large edge-set stones, which would have supported a timber-framed conical roof. Nearly all of the Eigg and Canna examples are situated in heather moorland or other rough pasture, well above the limits of post-medieval cultivation; very few survive within the areas of improved ground. This distribution explains their absence from the other two islands, for there is very little unimproved or uncultivated ground on Muck, while on Rum locations suitable for settlement have presumably always been restricted to the same few niches around the coast, densely settled and heavily cultivated into the early 19th century. Exploratory excavations of a hut-circle at Galmisdale by the National Museums of Scotland in the spring of 2002 have produced Bronze Age pottery, though the tradition of building round-houses probably spans most or all of the last two millennia BC.

Apart from fragmentary traces of field-banks on moorland in north-west Eigg, it is only on Canna that remains of prehistoric field-systems can be identified. At the west end of the island hut-circles are spaced out amongst a network of fields and enclosures defined by edge-set stones. It may be that a mixed agricultural economy can be detected here, with smaller fields defining plots of arable, while longer boundaries, disappearing into the moorland peat, may have divided up areas of pasture.

One of the more remarkable discoveries to emerge during the survey of Eigg also dates from the Bronze Age. In May 2001 Royal Commission staff were shown a collection of metal-working debris discovered by the late Brigg Lancaster while digging a hole to bury his cat. The collection included moulds for Late Bronze Age socketed axes, a knife and a pointed tool, along with fragments of crucible. Finds of Bronze Age metalwork are relatively common in Scotland, but there are no more than a handful of sites with substantial evidence for metalworking. Alerted to the discovery, the National Museums of Scotland immediately began excavating the site, and work has continued into 2002. It is thought likely that this is the temporary workshop of a Late Bronze Age smith, possibly an itinerant craftsman.

Poll Duchaill, Eigg. A
stone-walled fort occupies
the rocky knoll in the
foreground. (SC686860)

Loch Nam Ban Mora and
the Sgurr. The island in the
loch is enclosed by a
drystone wall, possibly the
remains of a prehistoric
dun. (SC729720)

Struidh on the north-east
coast of Eigg. A possible
prehistoric erimitic site
stands on the grassy
platform in the centre of the picture. (SC729717)

The Iron Age (700BC- AD500)
The majority of Iron Age monuments in the Small Isles are small fortifications known as forts and duns, at least sixteen of which have been recorded - four on Canna, three on Rum, one on Muck and eight on Eigg. Nearly all of them are sited on small coastal stacks and promontories, where they have needed no more fortification than a stone wall barring the line of easiest access. The interiors of most of them are either featureless, or contain footings of buildings that are demonstrably later than the defences, though three forts on Eigg - two at Grulin on the south side of the island and one overlooking Laig Bay - have visible remains of circular house-platforms that may be contemporary with the ramparts. One fort, however, on the Sgurr of Eigg, is quite different. A stone wall, up to 1.8m high, cuts off the only approach to the summit and encloses about four hectares, most of it bare rock with hardly any level ground at all. It is inconceivable that this has ever been permanently occupied, and it should perhaps be seen either as a temporary refuge or as a focal point for the local community, perhaps used for gatherings and festivals at particular times of the year - it is not difficult to imagine that the dramatic outline of the Sgurr, instantly recognisable for miles around, may have lent it some symbolic importance.

One final prehistoric site on Eigg remains to be described. It stands amongst a jumble of gigantic boulders beneath the cliffs on the north-east coast, in many respects the most remote corner of the island. Here, a substantial platform has been constructed, on top of which there are the remains of a thick-walled circular enclosure. Opening from the interior there is the entrance to a large boulder cave that runs westward beneath the enclosure wall. The main chamber of the cave measures about 7 metres in length and there are other, smaller, chambers opening off to either side and at the end. The cave entrance and the sides of the chamber have been modified by the insertion of rough walling, and a thick deposit of midden material covers the floor, including animal bones, shells and broken hammerstones, some of which have a concretion of crushed shell on their points.

Thus far, the site is unusual, but not unique - at Usinish on the east coast of South Uist there is another round structure constructed around the entrance to a boulder cave. What makes the Eigg site stand out is its position within the landscape. The cliffs on this side of the island are characterised by horizontal banding of different lavas, but immediately above the site the these bands are broken by a dramatic eruption of vertical basalt columns, soaring to the top of the cliff. Standing in front of the round-house, the eye is immediately drawn upwards, and the view is framed by two enormous boulders, one to each side. The sense that this has been a 'special place' is inescapable. Moreover, it is difficult to argue the case for an ordinary defensive or domestic function for the site. It is evidently a secluded spot, hidden away in the scree, over 400m of difficult terrain separating it from the shore. There are easily accessible shieling-huts and fragments of enclosure walls on grassy terraces close by, but to reach the enclosure involves a tricky scramble and the use of all four limbs. We know next to nothing about Iron Age religious practices in western Scotland, but it is tempting to speculate that this may be some prehistoric 'eremitic' site.

Eigg certainly, and Canna probably, were the sites of important monasteries in the early medieval period. No visible remains of these survive, though both islands have rich collections of early medieval sculpture. There are six cross-slabs on Eigg, dating from the 7th to the 9th centuries. One of them has been known since the 19th century, and one was discovered in 1987 in Kildonnan churchyard, the probable site of the early monastery. The other four, however, are first recorded in the early 1930s, when casts were presented to the National Museum of Antiquities of Scotland. Nothing is known about the circumstances of their discovery, though it is assumed that they also came from the churchyard. Most of the crosses are now on display in the porch of the Lodge, and one has been re-erected at Kildonnan.

Apart from its crosses, Canna also has the remains of an unusual monastic enclosure on its remote south-west coast, known as Sgorr nam Ban-Naomha, which may have been a hermitage subordinate to the monastery at A' Chill, and the presence of the early church on the other two islands is attested by two cross-slabs from each, and also by the placename Papadil ('priests' valley') at the southern tip of Rum. In contrast to these overtly Christian sites and monuments, all products of the Dalriadic church, a cemetery of at least fifteen square cairns was discovered during the survey just above the beach at Laig on Eigg. These may be regarded as pagan, quite possibly Pictish, and serve as a reminder that this part of Scotland was very much a frontier between the two peoples in the mid first millennium AD. Square cairn cemeteries are found throughout the area of the Pictish kingdom, with particular concentrations around Inverness and on the east coast as far south as Fife. They are, however, comparatively rare in the Hebrides, and the Laig cemetery is the largest yet recorded on the west coast.

Norse placenames are common on the Small Isles, as of course they are throughout the Hebrides, but the archaeological evidence for the period is limited to a small, albeit important, group of artefacts, chiefly from Eigg, and a possible house discovered during the Muck survey. The Laig boat stems and the Kildonnan sword are amongst the best known Viking artefacts from the Hebrides. The stems, which were discovered in the 19th century during the draining of a bog, measure about 1.9m in length, and one of them has been stepped and hollowed to accept the timber strakes of the hull. The sword was discovered about 1830, together with a whetstone and fragments of other artefacts, while levelling a 'hillock' to the north-east of Kildonnan farmsteading; only the hilt, of silvered bronze, has survived. Two other Viking burials, this time to the south-west of Kildonnan, each containing a sword and other artefacts, are recorded, at least one of them placed in a prehistoric cairn. Otherwise, the only Norse artefacts from the Small Isles are an ivory gaming piece discovered in a cave on the east coast of Rum and a ring-headed bronze pin found on Canna.

The possible Norse house identified on Muck is situated on the N coast of the island. Bow-sided on plan, it measures 13.4m by up to 5.4m within thick stone footings and has an outshot 6.8m long at its north-west end. In plan, at least, this building is clearly different to those of the 18th- and 19th-century settlements on the island, although, given our patchy understanding of rural vernacular building styles in the Highlands before 1700, a late-medieval date is also possible.

The medieval and post-medieval periods (AD1200-1800)
In the medieval and later periods the islands changed hands several times. In the 13th century Eigg and Rum were held by the MacRuaris of Garmoran, from whom they descended to the MacDonalds of Clanranald. This family retained Eigg until 1827, but in the 15th century Rum passed to the Macleans of Coll, whose ownership lasted until 1845 (though in the 1590s it was recorded as being again in the hands of Clanranald). Canna belonged to the Benedictine Abbey of Iona in 1203, and may have been in Church hands since the foundation of the probable Columban monastery. By the 1590s, it too appears to have been in Clanranald hands and, although the Earl of Argyll was granted the island in 1628, Clanranald may have retained possession, leasing the island in 1672 and acquiring the superiority in 1805, before selling it in 1826. Finally, Muck first appears on record in 1549, when it was the property of the Bishop of the Isles, but it had passed to the Macleans of Coll by 1626. In 1799 Muck too passed to the MacDonalds of Clanranald, though fourteen years later it returned to the Macleans.

There are few visible remains on the islands that can confidently be ascribed to the medieval period. Eigg has the remains of a church, outside which there stands a cross shaft of 14th- or 15th-century date, and there is a possible medieval chapel in the burial-ground on Muck. The medieval parish church on Canna is not visible, though its wall footings were identified by excavation in 1994, and a chapel probably stood in the burial-ground at Kilmory on Rum. The only castle on the islands, Coroghon, on Canna, is probably of 17th-century date. Many of the numerous shieling-huts on the islands are likely to be medieval, but the most interesting monuments that have been tentatively assigned to this period are the deer traps on Rum. The best-preserved of these comprises a high-walled oval enclosure set at the foot of a hillside. The deer are thought to have been driven down through a gap cleared in the scree on the slopes above and funneled into the enclosure.

Stone-walled buildings at
Upper Grulin.
(SC759496)

· Settlement
Until the end of the 18th century most farms were held jointly by a number of tenants. Settlement was concentrated into townships occupied not only by the tenants, but also by sub-tenants and cottars (labourers with little or no land of their own). Many of the townships on the islands are probably medieval in origin, though none of the standing remains they contain need be any earlier than the 18th century. Documentary sources often help in analysing the pattern of medieval settlement, but it is always difficult to reconcile a documented settlement with a particular group of archaeological remains. The best-preserved townships in the Small Isles are Kiel on Muck, Harris on Rum and Five Pennies on Eigg; there are also substantial remains at several other sites on Eigg - Braes (to the north of Kildonnan), Sandaveg (next to the Manse) and Grulin (though most of the surviving buildings here are probably 19th century). The walls of the buildings within them are typically constructed with a turf core and stone faces, in contrast to the the rubble-cored walls of the 19th-century houses. In most cases the surrounding fields have been obliterated by later improvement and cultivation, though substantial field-systems survive at Five Pennies and at Grulin, while Harris on Rum is set within a remarkable landscape of lazy-beds enclosed by a head-dyke.

Plan of Five Pennies
township and field-system.

The township and field-system at Five Pennies is illustrated here. The visible remains comprise the footings of about two dozen buildings, a corn-drying kiln and several small yards and enclosures at the edge of an extensive system of irregular fields. The buildings are all roughly rectangular and usually have rounded external corners. In one instance, close to the centre of the township, two buildings are joined along one side, an arrangement commonly found amongst blackhouses in the Western Isles, where byres and other ancillary buildings would often be added to one side of the main house, rather than at its end. The fields around the township are enclosed by substantial turf and stone dykes, within which fragments of rig cultivation can be traced. Their wide variation in size and their irregular form strongly suggest a gradual, piecemeal process of expansion and enclosure of arable land. Field-systems such as this are common on the west coast of Scotland - examples have been recorded by the Royal Commission at Achiltibuie, Wester Ross, and on Waternish, Skye. The Five Pennies field-system has survived largely because it lay outwith the Cleadale crofting settlement established immediately to the south in 1809, where intensive cultivation has left only fragments of the earlier enclosures.

· Shielings
Until the early 19th century, the use of high or remote pastures during the summer months was an important and integral part of the farming calendar throughout Highland Scotland, and the sites of these 'shielings' are usually betrayed by the presence of clusters of small huts. At least 400 huts have been identified on Rum, and over 500 on Eigg. Most of them are simple rectangular bothies built wholly or largely of turf, often decayed to the point where they appear as little more than low grassy mounds. Another type, common on Rum and Eigg, has two or more linked stone-built cells or chambers. On Eigg, most of the chambered forms have a figure-of-eight plan; the entrance led into the larger compartment, from which a narrow passage, often roofed with stone lintels, gave access to the smaller chamber. The latter was invariably constructed of stone with a turf embankment against its outer face, probably intended to keep the interior cool for the storage of butter and cheese, the traditional products of the shieling months. This type of hut is also well-known in Skye, where several examples have been excavated. Dating, however, remains a problem, though at least some of these huts may be medieval. On Eigg, there is a well-preserved group of these huts on the Allt Bidein an Tighearna, above the Cleadale cliffs, and a plan of one hut is reproduced here.

Aside from the shieling-huts, the remote pastures of Eigg are strewn with about 200 pens, small enclosures, and roughly-built shelters. These are particularly numerous below the cliffs of Beinn Bhuidhe on the north and east coasts of the island. Many of these structures are probably the work of shepherds, and are no more than 100 or 200 years old. Others, though, may be medieval or earlier, and it has been noted above that some of the rock shelters may date from early prehistory.

Draft plan and elevation of
late-19th century house at
Sandavore, incorporating
an earlier round-cornered
house.

A kelp kiln at the western
end of the Bay of Laig.
(SC759498)

The 19th and 20th centuries
· Kelp, crofting and clearances
The late 18th and early 19th centuries saw rapid change throughout the Highlands and Islands, as chiefs and landowners sought to maximise income from their possessions. In the 1780s and 1790s, rising rents and a rising population prompted a wave of emigrations, despite the opposition of landowners. By then the price of kelp, a soda ash produced by burning seaweed, which was important as a bleaching agent and in the manufacture of soap, glass and gunpowder, had begun to soar, and landlords needed a vast workforce for the kelp harvest. To accommodate the growing population, the Clanranald estate, which included Canna, Eigg and (from 1799 to 1813) Muck, reorganised the holdings of its tenants, converting the existing multiple-tenant farms into crofting townships comprising a large number of small single-tenant lots. Thus in 1809 plans were made to divide Muck into 47 holdings, and on Eigg in the same year the neighbouring farms of Cleadale and Knockeiltach were converted into twenty-eight crofts. Cleadale remains a crofting township today, and the straight stone-walled boundaries of its narrow plots stand in marked contrast to the irregular enclosures of the earlier farms. The construction of the tenants' houses on Eigg may also have changed at this time. The walls of most of the ruined houses in the crofting settlements have a core of stone rather than turf and a striking number have an internal length of roughly thirty feet (9.15m). This uniformity of length suggests a degree of control by the estate over the size of crofters' houses. Most of these croft houses were replaced during the late-19th and 20th centuries, though others were adapted and extended, and still form the core of a number of buildings on the island, such as at Sandavore, illustrated here. There are, however, few visible remains of the kelp industry. Kelp was burned in simple kilns, usually comprising no more than a pair of low rubble walls enclosing a narrow trench, in which the kelp was burned. As these rather flimsy constructions would normally have been placed close to the shore, many have no doubt been lost to winter storms, though six have been recorded around the coasts of Eigg at Laig Bay, Kildonnan and Galmisdale, and there are others on Canna.

The kelp boom collapsed when the end of the Napoleonic Wars allowed the importation of cheaper sources of alkali from Spain. Faced with a rising population increasingly unable to pay their rents, proprietors' attitude towards emigration quickly changed. In 1826, about 300 people were cleared from Rum, which was converted into a farm for 8000 sheep, and 50 more followed in 1828, leaving only one indigenous family. In the same year 150 people emigrated from Muck. The population of Canna was greatly reduced after its sale by Clanranald in 1826, and in 1849 it was halved again. On Eigg, 140 people elected to emigrate in 1843 and, ten years later, Laig and Grulin (roughly the western half of the island) were let for sheep and the Grulin crofters were evicted.

Lochaber Bank Barn at
Kildonnan.
(SC759495)

Eigg Lodge.
(SC729682)

· Sheep farms and Victorian sporting estates
Since the middle of the 19th century Eigg, Canna and Muck have each passed through the hands of a series of proprietors. Apart from the crofting areas (now confined to Cleadale and neighbouring Cuagach on Eigg, Sanday on Canna and one croft on Muck), they have been run as sheep and cattle farms, of which there are now three on Eigg and one on each of the other two. The existing farm buildings are mostly of late 18th- or 19th-century date, though parts of Laig farmhouse may be slightly earlier. Howlin was built in the 1770s, and Kildonnan farmhouse is probably early 19th century. At Kildonnan there is a fine example of a Lochaber bank barn - a two-storey barn of a type confined to parts of the West Highlands (there is another example at Coroghon on Canna), though similar structures are also found in the Lake District. The most impressive building of this period, though, is probably the manse, built in 1790 and enlarged in 1889. The minister of the Parish of the Small Isles moved from Rum to Eigg in the 1750s, and thirty years later the Church was given the farm of Sandaveg, which was reputed to be the most extensive glebe in Scotland. It would have been a striking landmark when first built - a prominent symbol of the authority of the established religion over a largely Catholic island.

Rum, with the extravagant Kinloch Castle, perhaps the ultimate shooting lodge, is closely associated with the Victorian and Edwardian passion for sporting estates, but the 19th-century proprietors of Eigg were also concerned more with the sporting opportunities offered by the island, and were content to leave the running of the farms largely to their factors. Norman MacPherson, who inherited the island from his father in 1854, surrounded his house with woodland to encourage pheasants, while Robert Thomson, who bought the island in 1896, considered moving all the tenants to Muck in order to make room for deer. The present Lodge, built in 1926-7, is the fourth proprietor's house built on or near that site, intended, like its predecessors, as a holiday house and base for shooting. Designed by a Newcastle firm, Mauchlen and Weightman, the lodge originally had a flat roof, but this was replaced in 1935 with the existing hipped roof.

Eigg has a rich variety of archaeological monuments, many of which contribute in a highly visible way to the spectacular landscapes of the island. While many of these monuments are of types familiar from other parts of Highland and Hebridean Scotland, others are more unusual, even unique - a reflection, no doubt, of the island's position on the western seaways, which has exposed it to a wide variety of influences throughout prehistoric and historical times. The Royal Commission survey was made possible by a grant from the Scottish Wildlife Trust, with funds provided by the Heritage Lottery Fund, and also by the enthusiastic co-operation of the Isle of Eigg Heritage Trust and the residents of the island.