Before excavation in September 2013 the entire area that had been planted was surveyed by a team from the North East Scotland Archaeological Research Society (NESARS) led by Moyra Simon and Jane Summers who examined the network of trenches extending across the surrounding hillside. Their work is reported on p. 23 and resulted in the identification of an extensive distribution of worked quartz. There was a smaller quantity of worked flint, and 16 m south of the ring cairn sherds of early Beaker pottery were found, together with a barbed and tanged arrowhead. They may be all that remains of a disturbed burial. Unburnt bone would not survive on such an acid subsoil.
Figure 2.1. The locations of the three excavated monuments in the Howe of Cromar in relation to the areas investigated by fieldwalking (contours at 50 m intervals). Information from Bradley 2005.
Figure 2.2. The siting of the four stone circles in the Howe of Cromar, showing the extent of high ground (contours at 100 m intervals) and the area that can be seen from Hillhead. Analysis by Irvine Ross.
Figure 2.3. Outline plan of Hillhead showing the principal features, the extent of excavation and the locations of the published sections. In all the plans of the five excavated monuments the smaller arrow represents true north and the larger arrow the direction of magnetic north at the time the fieldwork took place.
Since the monument was to be preserved, excavation was conducted on a limited scale and most of the stratified deposits were left intact. The aims of the project were:
• To determine the original form of the monument;
• To collect dating evidence; and
• To advise on its preservation and interpretation to the public.
The excavation investigated two quadrants, one to the northeast and the other to the southwest. In view of the size of the bank, parts of it were excluded (Figs 2.3–2.4). An additional area was opened at the centre of the monument where there was the flat stone that Ken Cooper had already noted. It proved to be part of a group of three which might have been the remains of a cist – in the event it became clear that they had been put there comparatively recently. The outer perimeter of the enclosure was exposed in both quadrants, but the bank was sectioned completely at only two points, one on either side of the enclosure. A further small section investigated part of the inner kerb where it was especially well-preserved in the north-eastern sector of the site. Within the excavated areas the full extent of the monument was exposed and recorded by photogrammetry, but only the latest and most vulnerable deposit – a layer of quartz close to the modern ground surface – was excavated in its entirety.
Figure 2.4. Kite photo taken from the east showing forestry trenches and exposed rubble at the start of the 2013 excavation (Nigel Healy).
After the work in 2013 the excavated area has been refilled, as have the forestry trenches cutting across the monument. An exclusion zone has been established around the prehistoric enclosure, and this is to be maintained as grazed grassland, allowing visitors to see what remains of this unusual structure.
The structure of the monument
Despite the damage caused by tree planting, the monument was generally well-preserved. It had suffered from the removal of a number of kerbstones and monoliths, probably in the 1940s, but beneath the surface the cairn was largely undisturbed (Fig. 2.5). The monument seems to have developed over six structural phases.
Figure 2.5. Trench 1 showing the positions of the internal wall, the external kerb and the fallen monolith.
The primary ring cairn
The basic structure at Hillhead was a substantial ring cairn 26 m in diameter which enclosed an open court 15 m across (Figs 2.6–2.7). It was built out of large stones, many of them angular blocks, up to 60 cm in maximum dimensions, all of which could have been collected in the vicinity. They lay directly on the prehistoric land surface. To the southwest the enclosure seems to have been provided with an entrance which was 2 m wide. The area was badly disturbed, but the terminals of the bank remained largely intact.
It was constructed in two stages. Where it was sectioned towards the northeast its earliest component was a stone wall with a rubble core buried beneath the later ring cairn (Figs 2.8–2.9). It was 40 cm high and 2.4 m wide. The upper surface of this structure was poorly preserved. During the excavation it was assumed that the damage was due to twentieth century tree planting, but the radiocarbon dates for charcoal on the old land surface raise the possibility that stones were being removed from the monument during the first millennium AD. On the other hand, several stone circles were reused during the Pictish period and it may be wrong to place too much weight on purely practical considerations. This question is considered in detail in Chapter 8.
It seems possible that the top of this wall was stepped. The idea was suggested by finding parallel rows of boulders at its core, and this interpretation was supported by excavating a small section through the filling of the court to expose the inner kerb where this part of the structure was better preserved (Figs 2.10–2.11). Here it was constructed of vertical slabs which had been pushed outwards by the mass of the cairn. They retained a level of flat stones. There may have been another step, largely destroyed, within the excavated area. It would have been about 10 cm high.
A section across the ring cairn on the south-west side of the enclosure found little sign of a similar structure (Figs 2.12–2.14). Instead it identified a bank of rubble about 2.7 m wide and 45 cm high. It consisted of large angular and rounded stones which had been stacked directly on the old ground surface. Here rather more quartz boulders were incorporated in the core of the monument than on other parts of the site. There was no evidence of any revetment within this part of the enclosure, but a large stone-packed socket was visible in section. It suggests that part of the structure had been removed at some time in the past. In contrast to the surrounding area, there was little worked quartz underneath the bank, but a few pieces of charcoal of Corylus and Salix/Populus were found on the old land surface and submitted for radiocarbon dating. Five samples returned dates in the later third millennium BC and should provide a terminus ante quem for the building of the monument. The dimensions of the rubble bank on the south-west side of the ring cairn were virtually the same as those of the internal wall on the opposite side of the perimeter and it seems likely that they formed part of the same circuit. Perhaps this was the oldest component of the site.
Figure 2.6. (Left) Photogrammetric survey plan of the rubble of the ring cairn in Trench 1 viewed from the interior of the monument (Aaron Watson); (right) the same image, emphasising the unexcavated remains of the inner kerb (Craig Williams).
Figure 2.7. Section of the ring cairn in Trench 1 at Hillhead. The remains of the primary wall are towards the left and the secondary addition and outer kerb are seen towards the right (Aaron Watson).
Figure 2.8. Section 1 through the eastern perimeter of the ring cairn.
Figure 2.9. The primary wall on the east side of the ring cairn (Aaron Watson).
Figure 2.10. Detail of the inner kerb in Trench 4, showing a deposit of flat slabs interpreted as a shallow step (Aaron Watson).
Figure 2.11. The inner kerb in Trench 4 showing how it has been pushed outwards by the mass of the ring cairn (Richard Bradley).
Towards the northeast the main body of the ring cairn was a secondary feature abutting the outer face of the existing wall (Fig. 2.15). It had no equivalent on the opposite side of the circuit. In its final form the bank was 8 m wide and survived to a maximum height of 65 cm. Its rubble core was delimited by a kerb of upright slabs up to 50 cm high, wedged between larger boulders placed on the ground. Much of this kerb had been robbed, leaving a trench up to 40 cm wide which was filled with a layer of topsoil and rubble (Fig. 2.16). Other kerbstones had fallen outwards and lay on the surface of the hill. Beyond this external kerb there was a shallow band of rubble up to 80 cm across and a single undated posthole. The north-eastern section of the ring cairn included a few large fragments of quartz, and others were found against the outer ker
b where they may have eroded from the bank. By contrast, there was very little worked quartz in the buried soil, especially on this side of the monument, suggesting that the scatter of artefacts in the surrounding area must have accumulated after the cairn had been built. That is clear as its distribution extended up the hillside from the position of the stone circle.
Figure 2.12. Section 2 through the western perimeter of the ring cairn.
Figure 2.13. The remains of the ring cairn in Trench 3, showing the rubble in section in the foreground, and the hollowed area in between the terminals which had been partly disturbed by a tree stump (Aaron Watson).
Figure 2.14. Section 2 of the ring cairn on the west side of the circuit. Towards the centre a possible stone hole can be seen in section (Richard Bradley).
Figure 2.15. The outer kerb of the ring cairn viewed from the east showing several large slabs held in place between deposits of boulders, and the trench which remained after others had been removed (Richard Bradley).
It seems that the outer kerb acted as a screen rather than a revetment, for the main mass of the cairn was provided by the remains of the existing wall. Nothing similar was found in the south-western sector of the enclosure, suggesting that only part of the original circuit had been modified. Taken together, these observations suggest that in its final form the ring cairn might have presented a slightly elliptical appearance, with a broader and higher structure on the uphill side and a less impressive boundary towards its lowest point. The exposed kerbstones suggest that it was most substantial towards the north and northeast where the surface of the structure may have included shallow steps. The field evidence is tenuous, but, if it has been interpreted correctly, the effect would have been like that of an open-air theatre. An audience occupying this part of the monument would have overlooked the lower ground in the Howe of Cromar and would also have seen the conspicuous outlines of Morven and Lochnagar on the horizon. The south-western perimeter of the enclosure would not have obstructed the view, as it was comparatively low.
Figure 2.16. Detail of the outer kerb on the eastern side of the monument, showing the remains of two kerbstones and the scarp left where others were removed (Richard Bradley).
Figure 2.17. Trench 1. The broken and fallen monolith is to the right and a displaced kerbstone to the left. They are composed of different kinds of stone (Richard Bradley).
Evidence of a stone circle
Even before the excavation commenced it was clear that the surviving cairn had been accompanied by a ring of standing stones. A local resident, Mabel Lawson, has a clear memory of playing there as a child in the 1940s and even remembers when the circle was dismantled to provide material for a wall. We are most grateful to her for this information. The 2013 excavation confirmed her recollection.
On the north-east side of the ring cairn just beyond the outer kerb there was a fallen monolith, 35 cm long, composed of a distinctive piece of magmatite (Fig. 2.17). It had an unusual texture because the rock had melted to a greater extent than the kerbstones composed of the same material. It could have been brought from nearby (we must thank Peter Craig for this information). It is not certain whether the monolith was complete, but its original position seems to be indicated by a shallow hollow 5 cm deep in the natural subsoil. It was accompanied by the disturbed remains of what was probably a ‘packing cairn’ of the kind associated with stone circles like those at Balnuaran of Clava (Bradley 2000, 73–78).
On the south-west side of the enclosure, directly opposite the fallen monolith, the bank was broken by what seems to have been an entrance 2 m wide (Fig. 2.18). The outer edge of the enclosure was marked by an oval hollow 2 m across but less than 10 cm deep (Fig. 2.19). It seemed possible that a large horizontal stone had occupied this position and had later been taken away. All that remained was the imprint left by part of its base, but the hollow that remained did not show any sign of compression. Equidistant from this feature there were two holes edged by large boulders, 65 and 75 cm in maximum dimensions respectively (Fig. 2.20). Both of them were 45 cm deep. They were entirely empty but their positions had been masked by surface debris. Both are very like the sockets for the flankers at Tomnaverie (Bradley 2005, 24). Assuming the comparison is valid, it would suggest that similar features had been removed in modern times. If so, the south-western limit of the site included a recumbent stone approximately 3 m in length flanked by two standing stones.
There are several arguments in favour of this interpretation. The putative recumbent stone is exactly in line with the fallen monolith outside the kerb to the northeast and also with the centre of the ring cairn. As is common in Aberdeenshire, it is located on the south-west side of the circle. Seen from the highest section of the bank there would have been a view between the flankers towards a cleft on the far horizon 15 km away. It was formed by Lochnagar in the background and by a conspicuous spur running south from Morven. The recumbent stone circle at Tomnaverie was also orientated on Lochnagar.
There may be other evidence of the ‘missing’ stones. Close to each of the sockets attributed to the flankers are two substantial pieces of magmatite containing prominent bands of schist; another was recognised in the inner kerb (Figs 2.21–2.22). The two fragments could have formed a pair, and their proportions raise the possibility that they were the base of one flanker and the upper part of another. In between them was a substantial fragment of a pink rock, identified by Peter Craig as felsite which had fractured along a series of horizontal bedding planes. This kind of rock could have been found locally, but it is the only substantial piece from the excavation. It may have formed part of a recumbent stone, broken and largely removed when the cairn was robbed for building material. It overlay one of the features identified as a stone socket (Fig. 2.23).
Figure 2.18. (Above) Photogrammetric survey plan of the rubble of the ring cairn in Trench 3 viewed from the north, showing the positions of part of the recumbent stone and one flanker towards the left, and traces of a second flanker to the right (Aaron Watson); (below) the same image, emphasising the positions of the bank terminals (1, 2), stone sockets (3, 5) and the hollow (4) left by the recumbent stone (Craig Williams). Scale bar 1 m.
Figure 2.19. The hollow in between the terminals interpreted as the position of the recumbent stone. Here half of it has been excavated (Richard Bradley).
Figure 2.20. The stone-packed hollow interpreted as the socket from which the northern flanker had been removed (Richard Bradley).
In the centre of the monument were two more fragments of magmatite like those used for the putative flankers. Excavation showed that they did not form part of a cist. In fact they overlay the latest filling of the ring cairn and fitted together to form a monolith which would have been more than 2 m high. This may have been another component of a stone circle, but, of course, its original position is unknown.
Figure 2.21. A broken monolith interpreted as part of the southern flanker (Richard Bradley).
There is a little evidence of sequence. There is no way of establishing the relationship between the fallen standing stone and the north-western sector of the ring cairn, but the holes left by the putative flankers did not penetrate the old ground surface. If these features have been interpreted correctly, the monoliths must have been bedded in the edge of the ring cairn and would have been placed there after its construction.
The secondary ring cairn
The next development was the capping of the entire monument by a layer of rounded boulders, most of them between 10 and 20 cm in maximum dimensions. According to Peter Craig, they could have been collected close to the site, but did not come from the same deposit as the larger blocks used to build the original structure. They would have been found further downslope than the other raw material. They were various colours, from white to red, but there was no evidence of patterned stonework of the kind identified at Tomnaverie. At Hillhead, a single deposit covered the full extent of the primary cairn and filled the court to the top of the rubble wall which was left unalter
ed. This material sealed a small deposit of Corylus charcoal in the bottom of the court. It provides a terminus post quem for the covering of the primary monument but actually predated the samples from the land surface beneath the cairn. Together with a date of 2461–2207 BC from the buried soil, they suggest a period of activity between about 2580 and 2300 BC before the monument was built. The question is considered further on p. 139.
Figure 2.22. A broken monolith interpreted as part of the northern flanker (Richard Bradley).
Figure 2.23. Part of the void left marking the socket for the southern flanker, masked by the surviving part of the recumbent stone (Richard Bradley).
Figure 2.24. Secondary features in the centre of the monument.
Secondary settings of boulders and a cremation pyre
At the exact centre of the monument three small features cut through the filling of the court (Fig. 2.24). There may well have been more beyond the area investigated. Two shallow pits, each of them 10 cm deep, were lined by medium-sized boulders and completely filled with charcoal. One provided a radiocarbon date of 1409–1302 BC. Next to them, in the middle of the ring cairn, was a larger area over which the natural ground surface had been scorched. The burnt soil was associated with a deposit of charcoal, but in this case it was mixed with a quantity of cremated human bone which had reached a temperature of over 900°C. The evidence for in situ burning suggests that this was the position of a cremation pyre (Fig. 2.25). More burnt bones were scattered over the land surface around it. The charcoal associated with all these deposits originated from the same two species – alder and oak. According to Fiona Shapland’s report, each of these deposits may have contained the remains of only one person. A sample of burnt bone from the feature identified as a pyre has a radiocarbon date of 1214–1005 BC.
The Use and Reuse of Stone Circles Page 4