South Uist, Drimore, A' Cheardach Mhor

NF74 2 A' CHEARDACH MHOR ('Drimore Smiddy', 'The Big Smiddy', 'Drimore 2')

NF/7564 4128

This dug-out wheelhouse in South Uist was excavated in 1956 as part of the rescue excavations organised by the then Ministry of Works before the construction of the rocket range (the Gaelic name means 'The big smiddy'). Before the work started it was a broad mound up to 2.0m (6 ft 6 in) high. The house inside it was found to have been excavated into the sand with a single facing of stones set against the pit wall and forming the wall of the house; no trace of subsidiary buildings was found. The site stands 185m (200 yds) from the west coast of the island on the flat, grass-covered sand, or machair. Five phases of occupation were inferred by the excavator, Alison Young, the earliest of which was the stone roundhouse itself.

This site is particularly important in that it produced the first evidence found in the west for the stratigraphical distinctiveness of composite bone combs and ornamental-headed bone pins – artifacts which we now know belong to the late Iron Age. It also demonstrated for the first time a clear sequence of pottery within the middle Iron Age itself; an incised ware phase was found below a phase with Everted Rim ware. The photographs make it clear that this was a well conducted, neat and systematic excavation [2, pls. VI-IX].

1. The excavations

Evidence of considerable disturbance and stone-robbing was found but the original floor of the roundhouse was intact.

Phase 1: the roundhouse was built inside its pit in the sand, the stone lining being usually less than 60cm thick but becoming thicker on the west. There were originally eleven free-standing radial stone piers inside the house of which nine survived, all badly ruined. Some stood up to 1.2m high at their inner ends and all rested against the outer wall without being bonded with it; there was no trace of a filled-in aisle. Thus the building had either never been an aisled house or – much less probably – the once free-standing piers had all been demolished and rebuilt. The inner ends of the piers corbelled inwards for a distance of from 7 - 10cm. A kerb of upright stones set into the sand joined the inner ends of the piers 1-5 and 8 and 9 and the facing stones of the pier ends seem to have been deliberately dressed (see Kilpheder, NF72 3).

A short entrance passage faced just north of east and was protected by a small forecourt in which a number of artifacts were found (the prevailing winds are from the south-west [2, 137]). There were no signs of a door-frame.

Artifacts as well as food refuse were found on the sandy floors of the bays; in Bay 5 for example there were many sherds with other refuse lying on a thick layer of peat ash and Bay 7 had many fragments of almost complete pots. In fact the pottery from Phase 1 seemed to divide into two groups, the earlier of which was stratified in the sand below the floors of the bays [2, fig. 5] and the later of which was lying on the floors "as if abandoned at the onset of some disaster." [2, 143]. The rest of the artifacts are not categorised in this manner.

The central area was sealed in by blown sand and the buried floor was a crust trodden with ash. There was nothing below the floor and only a few trampled sherds on it. Two whale vertebra post-sockets (see Dun Mor Vaul – NM04 4) were set into the floor on which were three hearths. The central one was edged with pebbles and had a clay floor. The rectangular Hearth 2, partly on top of one end of no. 1, was also pebble-edged and paved with flat stones set on clay; the long sides were markedly convex. Hearth 3 was further west and was also paved with stones and with a pebble kerb. Three post-holes about 8cm in diameter were found in the western arc of the interior forming a straight line; a fourth was on the south-east. The two whale vertebra sockets were set one on each side of the pair of central hearths and may have been roof supports.

Phase 1a: a later hearth was found in the sand filling the forecourt area, about 60cm above the roundhouse floor; part of the forecourt wall had collapsed on top of it. Iron slag was associated with it and a few potsherds were found with ash and sand tips beyond the wall.

Phase 2: a crude stone shelter was built inside the roundhouse and out of its rubble, and on top of a layer of blown sand; some of the Iron Age masonry was evidently showing above the sand in places.

Phase 3: a few stretches of crude walling were found about 30cm below the modern turf with midden tips and occupation levels associated with them.

Phases 4 and 5: a semicircular hut wall represented Phase 4 but few finds were associated with it. A bronze, loose -ring-headed pin suggested a date of not earlier than the 7th/8th centuries. Phase 5 incorporated disturbed material above the Phase 4 levels and just under the modern turf.

2. Discussion

The importance of the site. There are several reasons for regarding A' Cheardach Mhor as a major advance in the archaeology of the Iron Age in the west. It was the first of the 1956 'rocket range' excavations to be published and the first in the west to produce a well stratified assemblage of roundhouse pottery together with a comprehensive collection of bone and stone tools. The finding of a sequence of quern types is of particular interest.

Secondly there is a clear sequence of pottery within the middle Iron Age, the possible significance of which is discussed below. Thirdly the site had two important later phases of occupation in which the middle Iron Age decorated pottery of Phase 1 could be seen to be evolving and degenerating, and at a time when the roundhouse itself was falling into ruin.

Finally a third phase produced the same ornamental-headed bone pins and cordoned pottery that were found in Dun Cuier (NF60 5), confirming that the occupation of the Barra site was much later than the middle Iron Age. This was the first time in the Hebrides that the lateness of these pins had been demonstrated stratigraphically, even though R B K Stevenson had already inferred this earlier (1955).

Social stratification and relative dating. The pottery which was in use when this roundhouse was finally abandoned included plenty of Everted Rim pottery, in sharp contrast to its close neighbour A' Cheardach Beag (NF74 1), and this contrast is instructive. The incised wares which were found trampled into the primary floor, and in the sand immediately under it, were of the type of decorated Vaul ware vases found exclusively at A' Cheardach Beag. In addition the present site yielded a greater range of high-status objects including yellow glass ring-beads, the bone zoomorphic pin (which seems more likely to have been used in Phase 1 and to have been churned up into a higher level by one of the later occupations), the ball-headed iron pin and what might be an antler copy of a jet pendant. There is also the Everted Rim ware of the kind associated with the broch-building family at Dun Mor Vaul (NM04 4) and which includes some particularly elegant forms [2, fig. 6, no. 31).

Also to be considered are the ornamental-headed bone pins and the composite bone comb which were used on the site after the roundhouse was in ruins and which suggest that its prestige lasted down into the late Iron Age.

What might be the explanations for these important differences between the material cultures of two very similar roundhouses only about a kilometre apart? Two spring to mind – namely (1) a different in age or (2) a difference in status (or a combination of both). However when considering contrasts in assemblages of artifacts one has to bear in mind that techniques of excavation may be partly responsible for some of the differences. For example the small yellow glass ring-beads (and those of other colours) are easily missed by inexperienc-ed trowellers, or if fine sieving (preferably with water) is not carried out. The following comments are therefore tentative.

Were the two roundhouses of different ages? It is fairly clear that, in the Outer Hebrides and probably elsewhere, Everted Rim pottery was not adopted until perhaps AD 300; before that Vaul ware, with some gritty Dunagoil ware, was the pottery in use in the stone brochs and roundhouses (at Dun Mor Vaul the ware appeared earlier but the situation on Tiree seems to have been exceptional). The construction and first occupation of A' Cheardach Mhor may therefore not have taken place until the late 3rd or early 4th centuries – what might be called middle Iron Age 2. A’ Cheardach Beag in that case probably belongs to middle Iron Age 1 – perhaps the 1st century AD and well before the new pottery arrived.

Were the two roundhouses contemporary but of different social status? Since both excavations were carried out before radiocarbon dating could be easily applied to archaeological sites we cannot be sure that these two roundhouses really were built perhaps two hundred years or more apart, as suggested above. They could have been contemporary but the 'Big Smiddy' could have been occupied by a higher-status family with access to more and better trade goods, as well as to the new high-prestige Everted Rim ware, than the occupants of the 'Little Smiddy'.

One interesting question which arises from this discussion is the following. Can the broch- and wheelhouse-using communities of the Western Isles be divided into two, perhaps contemporary, groups on the basis of their material culture? One gets the impression from the Drimore sites that if a roundhouse had a large proportion of Everted Rim ware in its primary occupation levels it also had other luxury trade goods of the kind mentioned above. Conversely the houses with mainly incised line-decorated Vaul ware appear to have much less of this exotic material.

Economy. The excavator writes "There is ample evidence in Phase 1 for describing that occupation as pastoral" but without going into details [2, 158]. There were of course many animal bones (below) but the finding of rotary querns in several levels makes it clear that cereals (that is, barley) were almost certainly cultivated on the machair. The term 'mixed farming economy' is much more accurate.

The report on the animal bones [2, 169] does not include information on the relative frequency of wild and domesticated species but concludes "Although tenuous, the evidence may be interpreted as indicating the existence on South Uist at this time of cattle similar to the old Highland Black Cattle or Kyloes of which, in general, only the bullocks were slaughtered."

Fish bones were almost non-existent (it is hard to know whether this is a genuine lack or whether the excavation techniques of today would have recovered more) but other edible wild animals identified included shellfish and red deer. Most of the bones were from domestic-ated animals such as pig (small quantities), horse (few), sheep (very numerous) and ox (some, except for the very numerous first and second phalanges). There were also three dog bones.

3. Finds (fig. and pl. nos. refer to source [2]) Phase 1

Pottery [2, figs. 5 and 6]. As noted the sherds could be divided into two stratified groups. The older – which one might call Phase 0 – was incorporated in, or was below, the floor levels of the bays and the younger – Phase 1 – lay on these floors. The on-floor pottery had evidently been abandoned with the house and included large fragments of vessels, some nearly complete.

The Phase 0 sherds are mainly fragments of Vaul ware vases with the usual incised line decoration, though there are a few Everted Rim sherds with neck-band cordons. The characteristic Clettraval sub-style of Everted Rim (with decoration of concentric channelled arches) ware appears to be absent. Presumably much of this material was lying about on the surface when the roundhouse was built and fell into the original circular pit in the sand.

The Phase 1 on-floor pottery consisted mainly of Everted Rim ware with many large pieces of jars in the Clettraval sub-style. These are globular and situlate jars with applied horizontal zig-zag cordons with the concentric channelled arches decoration above these. There is also a good example of the version with the internally fluted rim, a type which seems to relate to Clickhimin ware. Three internally decorated bases were found, the last having a remarkable rosette pattern.

Metal (fig. 7). 1 fragmentary bronze ring (no. 16). Iron slag was found with Hearth 1.

Glass. Half of a small yellow ring-bead was found on the floor of the central area and another complete one was unstratified.

Bone and antler [2, figs. 7 and 8]. There is no indication that these have been divided into Phases 1a and 1b. They included – 3 probable quern handles (2 of antler) (nos. 13, 14 and 23), 1 worn antler pick (no. 21), 2 handled cups of whale vertebrae (nos. 19 and 20), 4 headless bone and antler pins, no. 4 highly polished (nos. 1-4), half of the solid bone ball-shaped head of an iron pin (no. 11), 3 coarse bone needles (nos. 5-7), 2 whalebone pegs (no. 10), 1 trimmed antler awl (no. 8), 1 antler gouge (no. 9), half of a bone toggle or hammerhead (no. 15), 1 knife made from a trimmed scapula (no. 18) and 1 perforated disc made from a whale vertebra epiphysis. There is also half a polished antler ring (no. 12) with a cross-section like a flattened D, which looks very like a copy of the kind of jet pendant found in the Inner Hebrides (sites NG33 2 and NM04 4) and on the mainland.

Stone [2, fig 9]. 1 upper stone of a discoid rotary quern with a socket for a loose handle on the upper surface (no. 1), found in the forecourt, 1 rotary quern stone diagnosed as a lower stone (no. 2), found in Bay 7) but which seems more likely to be an upper stone without a handle socket, 1 broken saddle quern used in a pier (no. 3), 1 pivot stone built into the main wall, 41 pebble hammer-stones, 5 whetstones and 22 pumice fragments.

Phase 1a

Pottery (fig. 10). The sherds found associated with the later hearth include 2 fragments of what appear to be Clettraval jars (nos. 42 and 43).

Bone (fig. 13). 1 fine gouge (no. 47).

Stone. 3 pebble hammerstones.

Phase 2

Pottery A complete pot was found in this horizon, an Everted Rim jar with an atypical applied cordon (no. 44).

Stone About a fifth of the upper stone of a rotary quern with markedly pointed edge (no. 4).

Phase 3

Pottery. The vessels are coarse cooking pots, with sooty exteriors, of harsh grey paste with the 'pulled up' everted rims characteristic of Dun Cuier ware (nos. 45-57) which is dateable to the earlier part of the late Iron Age. At the type site (NF60 5) it occurred with the same ornamental-headed bone pins (below).

Bone and antler. 4 nail-headed pins (nos. 24-27), 2 ball-headed pins (nos. 28 and 29), 1 spatula-headed pin (no. 30), 1 cylinder-headed pin (no. 31), 2 needles with pointed butts (nos. 32 and 33), 3 awls made from splinters (nos. 34-36), 1 awl made from a metatarsal (no. 37), 1 double-pointed 'spindle', perhaps an unfinished fish gorge (no. 38), 1 broken antler fish hook with a perforation at the top of the shaft (no. 39), 2 whorls or buttons made from human femur-heads (nos. 42 and 43), 1 worn quern handle (no. 44), 1 antler tool, trans-versely perforated and 1 broken whalebone 'club', perhaps a beater.

Stone (fig. 9). 2 fragments of upper stones of rotary querns (nos. 5 and 6), 23 pebble hammerstones, 1 whetstone and 4 fragments of pumice.

Phase 4

Pottery (fig. 10). 1 sherd of rilled ware probably of 7th/8th century date (no. 58) (Radford 1956, 59). The sherd was strictly unstratified but thought to be associated with the entrance of the Phase 4 hut.

Bronze (fig. 15). 1 cast ring-headed pin with a loose head (no. 48), a type dated to perhaps the 7th/8th centuries or even later (Stevenson 1956, 292).

Bone (fig. 15). The head of a pin with an ornamental head (no. 49). It is not commented on in the report but appears to be a bone version of the well known bronze zoomorphic pin (with a shank with a square cross-section and twisted along its long axis) probably of 4th/5th century date (Fowler 1963, 103, 121-22). A good example was found at Howe (HY21 6) and a broken one at Dun Ardtreck (NG33 2). If the bone example from this site is of similar age it belongs to the later middle Iron Age and should be considerably older than the rest of the material it was found with.

Phase 5

Pottery. In a thin occupation layer next to one of the late walls were found the sherds of a sooty plain pot (not illustrated). A sherd with an applied boss was also found (fig. 5, no. 18) as was a rim sherd of Dun Cuier ware (fig. 10, no. 49).

Iron (fig. 15). A small tanged knife (no. 50) was found in the same stratum as the potsherds just mentioned.

Bone (fig. 15). The fragments of a double-edged, composite comb (no. 51) were found with the sherds just mentioned. A whalebone shovel was also found.

Stone (fig. 15). In addition there is what looks like part of a shale whorl (no. 52) which is illustrated but not mentioned in the text.

Viking material

Part of a steatite lamp of Viking type was found just below the turf (fig. 15, no. 54).

Dimensions: internal diameters (from the plan), 10.07 - 10.98m (33- 36 ft); diameter of the central area, 6.71m (22 ft) from piers 2 to 8.

Sources: 1. NMRS site no. NF 74 SE 9: 2. Young and Richardson 1960: 3. Armit 1992, 62-3: 4. Crawford 2002, 114, 117, 118, 123, 125-6: 5. Armit 1996, 155: 6. Armit 1997, 251: 7. Brown 1997, 232: 8. Clarke 1971, 53, no.27: 9. Fanning 1983, 325, 334, no.1: 10. Henderson 2000: 11. Hunter 2002, 133: 12. Proc Soc Antiq Scot 108 (1976-77), 384, no. 3 (donation): 13. Sharples 2003, 151-2.

E W MacKie 2007