The Use and Reuse of Stone Circles

by Richard Bradley

  The pottery from the pit

  Alison Sheridan

  One hundred and nineteen sherds, together with a few fragments and crumbs, were recovered from the truncated pit; the overall weight amounts to just over half a kilogram. Small parts of eight or nine pots are present, with most represented by just four or fewer sherds. The largest single sherd, from Pot 2, measures 83 × 51 mm; this is the least incomplete pot, although only 10–15% of its body is present (in the form of 39 sherds plus a few fragments). While there had been considerable spalling of some pots (principally Pots 2–4), due to the laminar nature of the fabric, there was generally only minimal or minor abrasion to the sherds, suggesting that they had not lain around for long before being deposited in the pit. The two sherds from Pot 9 are softer and somewhat more abraded than the rest, but these may well have been burnt.

  Where lower body sherds exist, these indicate that the pots had been round-based; and sufficient diagnostic material survives to indicate that at least five or six of the pots had been carinated or (in the case of Pot 2) shouldered. The pots vary in size, in wall thickness, and in fabric and finish, although in no case are we dealing with the very thin-walled, very fine pottery that characterises the earliest Carinated Bowl pottery in Britain.

  Pot 1 (Fig. 3.8, top) appears to have been the smallest, finest and thinnest-walled vessel. It is a small carinated bowl with an estimated rim diameter of c. 145 mm and a projected depth of c. 96 mm; wall thickness varies from 8 mm at the neck to 8.5 mm at the carination, narrowing to 4.6 mm at the upper belly. It has an upright, flattened-rounded rim, an upright straight neck, a gentle carination and a medium-depth belly. The surfaces had probably been wet-smoothed but are slightly uneven, with lithic inclusions protruding through the interior and a fingertip depression from where the rim was shaped on the outside; there are hints of horizontal wipe-marks on the exterior. The pot is a medium to blackish-brown throughout. Inclusions consist of angular and sub-angular fragments, up to 2.5 × 1.0 mm in size, of a creamy-buff mineral, possibly feldspar, and possibly also some quartz; there are also occasional small mica platelets, and the overall density of inclusions is c.5%. Blackish staining on the interior may represent the evaporated (rather than burnt-on) former contents of the pot.

  Pot 2 (Fig. 3.8, bottom) is a slightly larger, shouldered bowl, with an estimated rim diameter of c. 230 mm and a projected depth of c. 143 mm; wall thickness ranges between 6.5 mm and c.10 mm. It has a rounded, everted rim; a fairly short, curving neck that is either upright or minimally everted; a ledge-like shoulder; and a medium-depth belly, with the body curving in unevenly below the shoulder. The surfaces are uneven, with numerous lithic inclusions projecting. The surfaces have a slip-like covering that could have resulted from wet-smoothing; there are traces of wipe marks on both surfaces. The exterior is a dark brown, slightly reddish in places and blackish elsewhere; the core is medium brown to blackish; and the interior is dark brown and black. The fabric is hard and laminar, with many sherds existing as spalls. It is also gritty, with abundant subangular fragments (up to c.4.5 × 4.5 mm) mostly of quartz, some accompanied by a black mineral or by mica; the overall density is c.15%. There is probable black staining on the exterior.

  Pots 3 and 4 (not illustrated) had been large, thick-walled (up to 17.5 mm), coarseware bowls, either carinated or shouldered. A neck sherd from Pot 3 suggests that its diameter at that point had been c. 270 mm, while a neck or upper belly sherd from Pot 4 suggests a diameter there of c.300 mm. Several sherds had clearly broken along coil joints. Surfaces are uneven but had been wet-smoothed, with clear wipe marks visible on both surfaces of the Pot 3 neck sherd and on some of the Pot 4 sherds. Surface colours are generally reddish-brown, with the core colour either echoing this or (in Pot 4) being a dark grey-brown. The same lithic inclusions as those noted in Pots 1 and 2 are present, but they include some larger, angular fragments (up to 4.5 × 5.5 mm in the case of Pot 3, and 7 × 5 mm in Pot 4); the density in Pot 3 is c.5%, whereas in Pot 4 it is as much as 15–25%. No obvious traces of organic residues are visible on these pots, although there may be hints of thin staining or sooting on the outside of Pot 4.

  Figure 3.8. Reconstructions of Early Neolithic Carinated Bowl Pots 1 (above) and 2 (below) from Waulkmill (Marion O’Neil).

  The other pots lie between Pots 1 and 2 on the one hand, and Pots 3 and 4 on the other, in terms of their size, wall thickness and surface finish, with Pot 6 probably resembling Pot 1 (but perhaps with a slightly splaying neck). Detailed descriptions of each exist in archive form. The same inclusions occur, in different sizes and densities, in the various vessels; one sherd from Pot 5 is notable for its abundance of mica and for a large quartz-with-mica inclusion. With the possible exception of one of the Pot 9 sherds, there is no evidence for decoration on any of the pots; the exception has two very shallow parallel lines on its exterior, but it is unclear whether these had been deliberately made, or represent the burnt-out traces of grass blades. No obvious signs of organic residues were noted on Pots 5–9.

  This assemblage clearly belongs to the Early Neolithic Carinated Bowl tradition (Sheridan 2007a), but apparently not to its earliest version, which has a higher proportion of thinner-walled, fine-textured, carefully-finished vessels (as seen, for example, in the assemblage from the ‘hall’ at Crathes Warren Field, Aberdeenshire; Sheridan 2009). However, it is clear from the assemblage found in another ‘hall’ close to Crathes, at Balbridie, that ‘modified’ versions of Carinated Bowl pottery had started to emerge soon after the initial appearance of the tradition as a whole (Sheridan 2009). A date for the Waulkmill assemblage as early as the 38th century BC is therefore possible, although a later date within the second quarter of the fourth millennium is equally possible. The Carinated Bowl tradition in general is well-represented in this part of Scotland, particularly along and between the valleys of the Don and the Dee, which appear to have been important foci for early farming communities. Several assemblages from this area have come to light since the publication of a Carinated Bowl pottery Scottish distribution map in 2007 (Sheridan 2007a, fig. 1). Find contexts in this area range from the aforementioned ‘halls’ to smaller houses (as at Garthdee Road, Aberdeen; Murray and Murray 2014) and individual pits or pit groups (as, for example, at Pitdrichie Quarry, Drumlithie, Aberdeenshire: Murray and Murray 2009). The presence of small parts of several pots, with the sherds not heavily abraded, is characteristic of the pit contexts. The contents of the Waulkmill pit could well constitute refuse from a settlement in the immediate vicinity.

  Worked flint

  Richard Bradley

  A single flint flake was found in a posthole which was not associated with any other artefacts. It was part of a pebble that had been worked down to its full extent, using narrow flake technique. It could date from the same period as the Neolithic pottery, but this is by no means certain.

  The Bronze Age stone circle

  The second phase is represented by the surviving traces of the stone circle levelled nearly two hundred years ago. Four components could be identified, all of them in the area that contained most of the rubble in the ploughsoil.

  Up to four possible stone sockets were found, following an arc that suggested the original monument had been approximately 18 m in diameter. Like the other features, they had been truncated by the plough, and three of them were no more than 25 cm deep. They could be identified by their regular plans and profiles, and by comparison with well-preserved features at Tomnaverie. Another monolith had a double socket which was 40 cm deep, indicating that a stone had been replaced. Large boulders, interpreted as disturbed packing, filled the western half of this feature. The monoliths had obviously been of varied proportions and that affected the extent to which they were bedded in the ground. Of those excavated, only the double socket was large enough to have held the stone that still survives. The stump of another one remained in the ground, but it was not clear whether it had been the base of a monolith or part of the kerb. The structure was built out of
pieces of felsite like the standing stone at the edge of the field (Peter Craig pers. comm.).

  The stone holes followed the same arc as a series of shallower features which were interpreted as the setting of a kerb in which the more substantial boulders were shallowly bedded; it is likely that others were simply laid on the ground surface, as happened at similar sites in the area (Fig. 3.9). These features were irregular in outline and up to 25 cm deep. In contrast to the stone sockets, their edges were difficult to define, and these features were probably disturbed when the site was levelled around 1835. Not all of them need have been associated with the monument, but they appear to have followed its outer limit. If that is correct, then approximately 30% of the area of the original monument was within the excavated area. As expected, it had been built on the end of the glacial ridge where the ground fell away to the north, south and west. To the southwest the spread of rubble in the modern ploughsoil ended just outside the line of the outer kerb, but there was no evidence that the sockets for the monoliths there were any deeper than the others. This point is considered further in Chapter 9.

  The central area of this structure was better preserved. It was represented by a curving gully, cut steeply into the natural subsoil (Fig. 3.10). It was up to 50 cm wide at the surface and 30 cm deep, but had obviously been truncated by the plough. It defined a circular area 4 m in diameter (Fig. 3.11). Comparison with the excavated site at Newton of Petty suggests that it marked the foundation trench for a kerb of upright slabs defining an open ‘court’ in the centre of a ring cairn (Bradley 2000, chapter 6); its dimensions are very similar to those of examples that remain intact. The kerbstones had been carefully removed, but sufficient evidence remained to suggest that the original structure was associated with large pieces of broken quartz whose distribution focused on the western sector of the ring. They were possibly used to hold the kerb in position. Alternatively, they may have been piled up against its inner face. On other sites an open enclosure was filled with rubble when the monument was closed. That could have happened at Waulkmill where the internal space contained patches of larger stones than those in the surrounding area.

  Figure 3.9. The surviving remains of the stone circle. The buckets mark the positions of stone holes and the shallower features in between them are interpreted as the position of the outer kerb. To the right the foundation trench for the inner kerb has been fully excavated, as has the central pit. The rectangular feature top left is Grave 2 (Aaron Watson).

  Figure 3.10. The partly excavated foundation trench for the inner kerb, with a recut pit in the centre capped by a setting of boulders (Aaron Watson).

  Figure 3.11. Plan and section of the circular trench and central pit, with details of the secondary recut and the setting of quartzite cobbles.

  At the centre of the monument was a pit which extended beyond the limit of the excavation. It was 1.8 m across and 1.05 m deep. Its function is unknown. The bottom 35 cm of the filling consisted of banded silts, perhaps resulting from natural erosion, but above them was redeposited soil and gravel containing many boulders. After these layers had accumulated, the uppermost part of this feature had silted naturally. There was no evidence of any burial and no artefacts were found. Its date remains uncertain, but there are other sites, including Newton of Petty (Bradley 2000, chapter 6) and Loanhead of Daviot (Kilbride-Jones 1935), where similar pits were excavated in the centre of a ring cairn. They date from the Late Bronze Age and are associated with cremated bone. In this case the only human remains were in a subsequent recut and date from the Roman Iron Age. This evidence is considered in a later section of the chapter.

  Discussion

  Although so little remained after nearly two hundred years, it is clear that the stone circle was not much smaller than its neighbour at Tomnaverie: it was about 18 m in diameter, compared with 17.5 m on the latter site. The sizes of the excavated stone sockets did not suggest that there had been a setting of taller monoliths towards the south or southwest, but this cannot be conclusive. At the same time it is unusual to find a monument in which the standing stones were set into the outer kerb. Normally, they form two concentric circuits separated by a narrow space, but, as Chapter 9 will explain, there is a local parallel for exactly this arrangement at the Blue Cairn, Ladieswell, only 7 km from Waulkmill (Welfare 2011, 15).

  Figure 3.12. The southern horizon from the Waulkmill monument with Mount Keen in the distance on the right (Aaron Watson).

  The most likely reconstruction of this monument is that it took the form of a low circular cairn, with an open space at its centre. It may have had a boulder kerb on the outside, but the inner court was probably defined by a wall of slabs set upright in the ground and was associated with pieces of flaked and broken quartz. The outer kerb incorporated a circle of standing stones set in their own sockets at intervals of approximately 3.5 m. For that reason it is possible to estimate that the monument was originally defined by a circle of up to 15 monoliths; however, it is not known whether they were equally spaced. In 1905, Coles was informed that originally there had been ‘ten or eleven’ standing stones on the site, plus the one that survives today (Coles 1905, 214). There was no direct dating evidence for any part of the monument, but in plan and dimensions it is very similar to a ring cairn at Sundayswells 14 km to the east. That structure lacked a stone circle, but included a court 3 m in diameter defined by a setting of upright slabs like that postulated at Waulkmill. In the centre of the site was a complete All-Over Corded Beaker (Welfare 2011, 228).

  The circle at Waulkmill would have commanded an all-round view towards the horizon, but, in common with many of the stone circles in this part of Scotland, the main focus was towards the south and southwest. It shares this characteristic with recumbent stone circles. There is no dead ground immediately outside the monument, but the vista is dominated by an intermediate ridge 3 km away and, in the distance, by the mountains of south Deeside. If the structure was directed towards a specific peak, it could have been the pointed profile of Mount Keen at a distance of 18 km (Fig. 3.12).

  Worked stone

  Richard Bradley

  QUARTZ AND QUARTZITE

  The foundation trench for the inner court contained 28 substantial fragments of quartz and quartzite. All were angular fragments, but only five of the smallest pieces and one of the larger fragments were struck flakes. The rest had simply been smashed. They fell into three broad size classes. The smaller ones (14 examples) were between 4 cm and 8 cm in maximum dimensions; for another six the equivalent figure was between 9 cm and 11 cm; and for the eight largest fragments it was between 12 and 20 cm. There is nothing to suggest that artefacts were being made here, and there may have been a greater concern with the appearance of the quartz when the stone was freshly broken. This material compares with the sample from the 2012 excavation at Croftmoraig which is considered in Chapter 4.

  The Roman Iron Age burials

  Inhumation graves

  At an early stage in the excavation two distinctive features were identified close to the edge of the monument and on the outer edge of the spread of rubble identified in the ploughsoil. They were 2.5 m apart and larger than any of the excavated stone sockets, but contained large flakes of rock very like those excavated on the Hill of Tuach, where a ring of monoliths was demolished during the nineteenth century (the report on this project appears in Chapter 5). It seemed likely that the fragments at Waulkmill resulted from a similar process. At first these features were interpreted as the holes left behind when two of the taller monoliths were uprooted. It raised the possibility that the structure had been a recumbent stone circle similar to that at Tomnaverie.

  A problem with this interpretation was that only one of these pits was especially close to the perimeter of the prehistoric monument. Another was that their positions were difficult to reconcile with the layout of its inner kerb, which was the best-preserved part of the structure. It was difficult to see how they could have formed parts of two concentric rings. It was only when
excavation was further advanced that the possibility arose that they were really graves. Soon it was confirmed by the discovery of a gaming set and a penannular brooch, very like those found in 1898. At that point it became apparent that the remains of the stone circle had been reused during the Roman Iron Age.

  The first of these features to be excavated was a rectangular pit (Grave 1), 1.9 m long and a metre deep (Fig. 3.13). The filling was of loose gravel and sand, disturbed by roots and burrows, but in the bottom 20–45 cm of this feature the sides were lined by between two and four vertical courses of rounded or roughly squared stones, up to 35 cm in maximum dimension; slabs performed the same role at its western end (Fig. 3.14). They defined a rectangular cist which was at least 1.4 m long and 60 cm wide; the full length is uncertain as the eastern edge of this structure had collapsed before the grave was filled. That could not have happened if the stones had been packed around a coffin. The bottom of the grave was a layer of smaller, more irregular boulders and clean sand. There was no trace of any bone or wood, but at this level a set of 11 gaming pieces/amulets was found in a compact group in the centre of the cist. They might have been in a bag. Towards the eastern end of the grave a penannular brooch was discovered at the same level. It was strikingly similar to the one from the nearby quarry. The upper part of the cist may have collapsed as there were a number of large boulders in its filling, as well as the flakes mentioned earlier. It must have been reinstated rapidly, as it was dug through a patch of loose sand and its edges were extremely unstable.

  The second grave (Grave 2) occupied a larger oval pit (Fig. 3.15). At the level of the glacial till its maximum dimensions were 3.5 × 2.9 m (Fig. 3.16). It was dug through more than one layer in the natural subsoil, including a level of small boulders, and bands of humus and gravel (Fig. 3.17). At a depth of 2 m was another setting of boulders, the largest of which had maximum dimensions of 20–30 cm. They included pieces of quartz, and substantial flakes like those associated with the other burial. On its south side this setting had a straight edge that recalled a wall foundation, but elsewhere it merely filled the lower part of the pit and lapped against its sides. It seems to have covered the base of the original cutting over an area measuring 2.0 × 1.4 m. The stones enclosed an approximately rectangular space which was 2.5 m long. It was between 50 and 70 cm wide and 20 cm deep and contained traces of a coffin, the limits of which extended a short distance beyond the distribution of boulders (Fig. 3.18). It seems likely that these stones were packed around the edges of the coffin when it was introduced to the grave.