North Uist, Foshigary

NF77 6 FOSHIGARRY

NF7430 7636

The remains of this remarkable cluster of Iron Age structures under a mound in the machair in North Uist include two dug-out aisled wheel-houses, a dug-out wheelhouse and several associated small buildings, one with a souterrain. The site was explored by Erskine Beveridge between 1911 and 1914. He started work again in 1919 but died in the summer of 1920 without being able to complete the work. However a draft report had been written and the publication of this was arranged by J Graham Callendar [2] who included his own comments on the structures and artifacts found.

The site is in a sheltered bay facing north and next to good agricultural land [3]. It is remarkable for the wealth of its material culture, particularly for the many well-preserved bone objects.

1. The excavations

Beveridge distinguished six subterr-anean structures in the group which had been somewhat damaged by the shifting of the seashore. The walls of all were simple stone revetments of the vertical sides of pits in the sand. The total area of the settlement – either a small village or a large single farmstead – was about 50.25 by 30.5m (50 by 100 ft) and there was a small outlying harbour further along the shore. The foundations of a ruined modern cottage stood on the summit of the mound of sand which covered the settlement; its floor was from 2.44 - 3.05m (8-10 ft) above the Iron Age floors.

The wheelhouse ('Chamber A'): this was excavated in 1911-12 and more than half of it had been washed away by the sea. Three radial piers were preserved, 2.14m (7 ft) in length and bonded to the wall. The curvature of the latter suggested an internal diameter of about 9.15m (30 ft). There were aumbries in the wall and traces of a stone kerb joining the inner ends of the piers. Finds in this building included sherds of “patterned pottery”, a complete whale-bone long-handled comb [2, fig. 11.1], various other implements of whalebone including the tip of the long pole handle of a rotary quern [2, fig. 13.1], many hammerstones, a sandstone whorl, several thin discs of mica schist, 3 broken quern stones (presumably rotary), and two bone pins. There were also about 12 whalebone mattocks, a type of tool very common at the site.

The first aisled wheelhouse ('Chamber B') was also only partially preserved and had only three surviving radial piers on the arc on the south and south-west. The north-west arc of the main circular wall had disappeared and a straight wall – perhaps a secondary one reducing the enclosed space – formed a chord across the open side; this also formed the boundary with Chamber C. The remains of a modern cottage sitting on the north half of the building may explain the dilapidation of that area.

The remaining piers were from 1.30 - 1.42m (4 ft 2 in - 4 ft 8 in) in length and survived to a height of 76cm (2.5 ft); the 'aisle' was wider than usual, from 61 - 92cm (2-3 ft). In the central area there was an oblong hearth with convex ends, 1.42m (4ft 8 in) long and 1.07m (3 ft 6 in) wide. It is noteworthy that some of the finds are described as having come from "about a foot below the floor" so there may well have been several periods of occupation at the site. A stone, slab-built box was set into the floor near the hearth and three lintelled drains were found under the floor.

Finds from this building included about 30 notched whalebone mattocks, and a large pot which was found inverted on the floor. This was bucket-shaped, rimless and decorated with one applied, impressed cordon [2, fig. 4]. Also found were a whale vertebra cup, the handle of a long bone comb ornamented with incised lines [2, fig. 11], a fragment of a “small-toothed comb” with dot-and-circle incised decoration (presumably composite) and a possible antler whistle [2, fig. 9.5].

The second aisled wheelhouse ('Chamber C'): this building was next to B and had also been badly wrecked by the construction of a modern cottage on top of it. The house was evidently oval with diameters of about 9.15 - 10.68m (30-35 ft). Five radial piers were preserved and their tops were about 2.14m (7 ft) below the modern surface. The 'aisle' was from 51- 76cm (20-30 in) wide and the piers themselves varied in length from 1.35 - 1.47m (4 ft 5 in - 4 ft 10 in), although one was 3.05m (10 ft). The hearth of the recent cottage was only 25cm (10 in) above that of the roundhouse below, which was rectangular; it measured 1.14m (3 ft 9 in) by 97m (3 ft 2 in) and its stone kerb stood 41cm (16 in) above the floor level. Two floor tanks were nearby, formed of stone slabs and covered with slab lids which were at floor level. One had a lintelled drain running from it. One shelf or aumbry was found in the outer wallface.

Artifacts were fairly numerous in this house and the more interesting included a double-sided bone comb with dot-and-circle ornament [2, fig. 5]; this was found on top of radial pier no. 4 of which only a fragment remained and it evidently therefore dated from a late phase in the history of the roundhouse, after it had fallen into ruins. Also found were a bone die, its long sides marked 3-4-5-6 with dot-and-circle ornament [2, fig. 6], and 2 long-handled bone combs. One of the latter was crudely made and the second was ornamented with fine incised lines [2, fig. 11, nos. 2 and 3].

Other dwellings: there were three other small stone cells near Chamber C and "perhaps with a somewhat lower floor level." [2, 313]. Chamber D seems to have had a stone corbelled roof; many large slabs were found lying on its floor. Chambers E and F were really a single room divided by a cross-wall. A vertical shaft leading to a drain was just outside Chamber E.

The long underground gallery, or souterrain, H led away from Chamber F and was presumably contemporary with it; the plan shows that it ran past the chamber and there was evidently a cross-lintelled doorway leading to F about two fifths of the way along it. This cellar (if that is what it was) was 12.8m (42 ft) long and much of it was still roofed with heavy lintels when found, which were below 90cm - 1.2m (3-4 ft) of sand. Its outer entrance was only 40cm (16 in) wide and 86 - 92cm (34 - 36 in) in height but the passage gradually expanded to to general dimensions of 51 - 66cm (20 - 26 in) in width and 86 - 92cm (34 - 36 in) in height.

There was an oven full of peat ash in the south wall about 1.22m (4 ft) from the east end; it was 60cm (2 ft) in diameter and of the same height. A sloping vent 2.13m (7 ft) long leads upwards to within 30cm of the surface through wheelhouse C so the oven may have been a late construction. A floor-level recess led at right angles from the eastern end for a distance of 1.22m (4 ft). On the souterrain floor were found a group of six hammerstones.

2. Discussion

Although the stratigraphical evidence from this site is not extensive Foshigarry is of importance both because it is probably the largest middle Iron Age settlement known in the Western Isles (even though it may not have been occupied all at once – below) and also because of the exceptionally diverse and well preserved array of bone tools found. However it is clear that the site was occupied for several centuries, though not necessarily continuously; the two intact double-edged composite bone combs and the many ornamental bone pins are of late Iron Age type and suggest that a higher occupation level existed (as at A’ Cheardach Mhor – NF74 2) which was not observed. The discovery of one of the combs on top of a ruined roundhouse pier seems to confirm this. An unknown but possibly large proportion of the bone artifacts may therefore belong to this late occupation.

Because of the lack of data about the find spots it is difficult to surmise much about the way the settlement was used in its primary phase of occupation. The oven inside the souterrain is unusual and suggests that the cellar may have been for grain storage or something similar, where a warm dry atmosphere was needed. The large number of notched whalebone mattocks have not been matched at any other excavated site; they suggest how the machair could be cultivated for barley if iron ards were not available and they also suggest that iron hoes were used but rarely survive. The great number of whalebone examples at Foshigarry suggests that a whale was beached nearby at some point and that the inhabitants made good use of its bones (to shape which sharp iron tools would have been needed).

Scott concluded, probably correctly, that the settlement was in fact a sequence of roundhouses lived in successively and if this was the case the impression of a large settlement would be wrong [3] (Scott was speaking authoritatively as he had already excavated the Clettraval roundhouse (NF77 2)).

"From the evidence of the drains it may be deduced that house C antedated house B and there is some confirmation that its floor was at a lower level …. On plotting these (spot levels described by Beveridge below modern ground surface) it would seem that the floors of house F and the souterrain were 3 or 4 feet lower than the floor of house C, and 4 or 5 feet lower than that of house B. There were thus three successive houses within the complex: F with the souterrain; H; B; and C. The successive lives of these houses can be contained within a period of one or two centuries; for they were slight structures with flimsy piers, and wall facings merely revetting the soft sand."

Most of the illustrated pottery seems to be of the decorated 'native' Vaul vases and also the cordoned Balevullin vase; an exceptionally fine example of the latter is shown here. However there is also some Everted Rim ware including at least one fine, fluted rim of the Clickhimin sub-style [2, fig. 25, 3], so one might guess that the main part of the middle Iron Age occupation was fairly late, after Everted Rim ware had arrived. The evidence from Dun Vulan (NF72 1) suggests that this probably happened during the 3rd century AD in the Uists. The large, almost intact, cordoned pot found in house B could support this idea; it must be among the latest middle Iron Age vessels from the site and, although its rim and base had gone, it looks rather like an Everted Rim jar [2, fig. 4].

In fact it is fairly clear that at least one of the roundhouses was inhabited as late as the 6th or even the 7th centuries. A fine example of a Dun Cuier rim sherd, with a zig-zag cordon in the angle of the neck, was found on the site and is illustrated here (Illus. 9.000). A wall sherd, also shown, probably belongs to it.

The question of whether Foshigarry was a high-status settlement should be considered. The absence of trade goods like small glass beads, and of bronze personal items like finger-rings (there was only one ring-headed pin and none of the pottery seems to have been impressed with the head of one), and of prestige items like fragments of Roman Samian pottery or glass, could mean that this was the settlement of a humbler group. The contrast with A’ Cheardach Mhor (NF74 2) is quite strong in this respect. However small items like yellow ring-beads could easily have been overlooked by the methods of excavation employed (which did however recover many small bone pins).

More pertinent to this problem perhaps is the clear evidence of the extensive occupation of the presumably ruined site in the late Iron Age (6th/7th century onwards). Only a small pro-portion of excavated wheelhouse sites have been found to have such a late occupation and it may be that the newcomers chose deliberately to live where the houses of the ‘gentry’ of the middle Iron Age population lived. There is reasonably clear evidence from Dun Ardtreck (NG33 2) that high-status sites retained their prestige over many centuries, even perhaps when they were in ruins. This is a possibility that future excavators of brochs and wheelhouses in the western islands – presumably settled by Q-Celtic-speaking Scots from about the sixth century onwards, and who seem to have brought an end to the long established middle Iron Age culture – could bear in mind. Perhaps all these monumental structures were occupied by indigenous tribal élites who were dispossessed of their inheritance by the Scotic chiefs, who even so wished to appropriate some of the prestige of the past (not everyone now agrees that Irish Scots settled in the west, e.g. Campbell).

3. The finds

Out of the vast quantity of relics found during the Foshigarry excavations only about half eventually came to the National Museum in Edinburgh and were thus described and published reasonably systematically; these pieces had first been kept in Beveridge's house in Dunferm-line. He apparently left much of the less interesting material in his house on Vallay and many of these were given away to friends. However the totals found were listed and these were included by Callendar in his descriptions of the objects which did find their way to the Museum. Hardly any of the finds have accurate stratigraphical contexts although it is now clear, as noted earlier, that the site must have been inhabited for several centuries. Only the more significant finds are listed here; the figure numbers refer to those in [2].

The bone and antler finds have recently been discussed afresh by Ywonne Hallén [5] and her comments are alluded to in the figure captions.

Iron working: 1 rounded mass of iron, perhaps from a bloomery (a 'bloom' is the product of repeatedly hammering (to drive out impurities) the hot mass of spongy iron which was left in the smelting furnace) ; 19 lumps of slag in Chamber B.

Iron: 2 broken blades, each set between hafts formed of a pair of bone plates, and rivets in the composite bone combs [2, fig. 5].

Bronze: 1 ring-headed pin, all in one plane [2, fig. 8.1], 1 perforated bronze plate [2, fig. 8.2].

Bronze-working: 1 small hemi-spherical crucible, thick-walled and with a pouring spout [2, fig. 8.5].

Bone: 4 long-handled combs of whalebone, 2 decorated with geometric patterns in incised line [2, fig. 11], 2 iron-riveted, double-edged composite combs (1 with dot-and-circle decoration on both faces [2, fig. 5]), 1 parallelopiped die (the long sides marked 3-4-5-6 in dot-and-circle motifs [2, fig 6]), 14 complete pins with 2 pin heads [2, fig. 19] (comprising 7 plain [2, fig. 19, 8-11 and 13-15] and 7 ornamental-headed late Iron Age type of which 3 are nail-headed – 1 with a dot-and-circle motif on the flat top) [2, fig. 19, 5-7], 1 ball-headed [2, fig. 19.4], 1 crutch-headed [2, fig. 19.2], 1 with a flat head nicked on the sides (no. 1), and 1 with a double ball-head (no. 3)), 10 needles [2, fig. 19, 16-24] (most with the pointed butt above the eye, especially no. 20), 1 similar object which is probably a fish gorge [2, fig. 19, 17], 8 supposed socketed lance points [2, fig. 16] (made from sheep or deer tibiae, the points made by a slanting cut across the shaft), 1 point made from a flat strip with a fishtail cut at one end [2, fig. 22.6], 3 awls made of sheep or deer ulnae [2, fig. 14] and 3 of whalebone (not illus.), 5 borers (lengths 4.80 - 9.0cm, or 1.85 - 7.5 in), 23 awls made of thin splinters [2, fig. 18], 1 gouge [2, fig. 22.8], 1 socketed harpoon or arrow point with a barbed head [2, fig. 17] (from Chamber C), 4 spindle whorls or buttons made from what look like human femur heads [2, fig. 7, 1-3 and fig. 12, 4 (the last may not be a femur head), 3 possible whorls or discs of whalebone intervertebral plates [2, fig. 12, 1-3], 2 weaving plates of the same, perforated in several places [2, fig. 12, 5], 40 whalebone blubber mattocks with notches for hafting [2, fig. 1, 1-12 and fig. 2, 1], 4 pointed whalebone mattocks [2, fig. 2, 2-4) and 1 long notched mattock or pick [2, fig. 3, 2], 3 perforated whalebone mallet heads [2, fig. 10], 15 horn socketed handles [2, fig. 20, 1-9], 2 horn handles with dot-and-circle decoration, 4-and-a-half handles pierced through the whole length [2, fig. 20, 12], 3 deer-horn picks [2, fig. 21, 1-3], various other tools like chisels, scrapers and smoothers, 1 rib knife with the point made by a short longitudinal slit (judging from the similar example found at Dun Mor Vaul it may be a potter's knife) [fig. 22, 5], and 6 fragmentary whalebone vertebra cups.

Stone: 1 small axe head in Chamber C (not illus.), 3-and -3/16th in (8.0cm) long), 80 beach pebble hammestones, 7 whetstones, 2 quartzite and 1 other pebble strike-a-lights and 4 probably quartzite ones, 2 fragments of saddle querns (Chamber C), 5 fragments of rotary querns (Chambers A, C and E), 3 whorls, 2 with hourglass perforations [2, fig. 7, 4 and 5], 2 unfinished whorls [2, fig. 7, 6 and 7], 1 broken ring 10.2cm (4 in) in diameter and 5.1cm (2 in) thick with a hole 3.2cm (1.25 in) in diameter, 20 flat discs of mica-schist 10.2 - 22.8cm (4-9 in) in diameter, probably pot lids (Chambers A, B, C and D), 46 pumice fragments (found on a shelf in the back wall of E).

Pottery: many sherds were found of which the better decorated were illustrated by photographs [3, figs. 23 and 24] and by a page of unrelated rim sections without indications of pot diameters . Some are redrawn here.

Beveridge's two pages of photographs [3, figs. 23 and 24] show almost exclusively incised ware (Vaul and Balevullin vases) although some of the cordoned sherds which appear otherwise to be plain (nos. 4-6 for example) could be from Everted Rim jars. Balevullin vases, with prominent geometrical patterns of incised lines above the applied cordon, are particularly prominent (Illus. 9.000) and strongly resemble material from the Broch of Ayre in Orkney (HY40 1: Illus. 5.110).

Some of the pottery was drawn by the author in the early 1960s. The dominance of the Balevullin vases is again clear but there is at least one Dun Cuier rim sherd which fits with the belief that the site continued in use until the beginning of the late Iron Age. There is also one fine fluted Everted Rim sherd of the Clickhimin sub-style, and several neck-band Everted Rim sherds.

Sources: 1. NMRS site no. NF 77 NW 5: 2. RCAHMS 1928, xlii: 3. Beveridge and Callendar 1931: 4. Scott 1948, 74-5: 5. Hallén 1994: 6. Armit 1996, 139, 147, 152, 162, 178 and 180: 7. Crawford 2002, 118, 119.

E W MacKie 2007

ShoreUPDATE

The building at NF 74305 76271 can be identified (see image) and the coastal slope in the immediate vicinity seems to have established vegetation on its face and does not appear to be actively eroding. The coastline to the east of this location is eroding, though it seems to be mainly from wind and particularly stock damage rather than the sea. This is particularly evident near the drainage channel referred to where a former enclosure or sea defence wall right on the dune edge looks as if it is suffering damage.

Visited by Scotland's Coastal Heritage at Risk (SCHARP) 3 December 2015