The Use and Reuse of Stone Circles Read online

  None of these monographs was completely self-contained, for the three traditions overlapped. Both Clava Cairns and recumbent stone circles feature rings of uprights which are graded by height towards the south or southwest. Each type encloses a cairn and is associated with the use of coloured stones, deposits of quartz, and pecked decoration (usually cup marks). Nor are they entirely separate from the henges found in the same areas. Some share their characteristic orientation, and the comparatively large example at Broomend of Crichie contained a setting of monoliths. It was close to a recumbent stone circle and the two structures were connected to one another by an avenue. In the same way, there had been a standing stone in the entrance of the Migdale henge. In fact the overlap between these structures goes even further, for earlier examples of all three kinds were reused during the Middle and Late Bronze Ages and were associated with cremation burials. Such observations required more investigation.

  Three specific problems remained after the results of these projects were published. They are considered now. Other important issues emerged during the course of this work.

  Henges and stone settings

  The first issue was the relationship between the earthworks known as henges and the stone settings sometimes found inside them. The problem was addressed at Broomend of Crichie, but that excavation raised almost as many questions as it answered. Although the monument had been described as an enclosed stone circle, there were two difficulties to be addressed. The first is that the earthwork was later in date than the setting of monoliths inside it. The other was that the setting was not a circle in the first place – it was an arc of upright stones flanking the position of a shaft grave, whose position had been marked by an unusually tall pillar. At the same time, excavations at Pullyhour, and to some extent at Lairg, suggested the existence of a distinctive group of circular embanked and ditched enclosures which were smaller and later in date than Broomend of Crichie. This echoed Aubrey Burl’s suggestion, based on a small body of excavated evidence, that in Scotland stone circles of similar proportions were among the last to be built. In contrast to those in lowland Britain, their chronology might have extended into the Middle and Late Bronze Ages (Burl 2000, xiv–xv). Until 2011 each line of argument had been pursued separately. What was needed was new fieldwork on sites where these elements were combined.

  Figure 1.1. Varieties of circular monuments in northern and north-eastern Scotland. Information from Bradley 2000; Bradley 2009; Welfare 2011.

  That requirement was met by two of the structures considered here. The unusual site on the Hill of Tuach (or Tuack) was discussed in Stages and Screens and is located not far from Broomend of Crichie, but it was not possible to excavate there until that volume was complete – in fact the book was published while the new excavation was in progress. It re-examined a monument first investigated in 1855 which seemed to combine the attributes of later henges, like those at Lairg and Pullyhour, with those of the last stone circles previously identified by Burl. The work had the added advantage that antiquarian accounts suggested that these structures had been associated with burials similar to the urned cremations dated to a late phase at Broomend of Crichie.

  The other monument which fulfilled this requirement is much better known. This is the stone circle of Croftmoraig (otherwise Croft Moraig) on Tayside which was first excavated by Stuart Piggott and Derek Simpson in 1965 (Piggott and Simpson 1971). In this case the earlier project had been properly published and there was no doubt that the monument had a lengthy history. For the original excavators it began with an Early Neolithic timber circle, which was replaced in successive phases by an oval setting of monoliths and then by a more substantial stone circle provisionally dated to the Early Bronze Age. The entire monument was enclosed by some kind of rubble bank. As the excavators acknowledged, that sequence was merely their preferred hypothesis, and other versions were possible. In parallel with the excavation at Broomend of Crichie, Alison Sheridan and the writer suggested an alternative interpretation of Croftmoraig and a new chronology for the site. The stone circle appeared to be the oldest feature, and the timber setting came later. In turn it was replaced by a small oval stone setting which recalled the characteristics of the later stone circles of Tayside. It seemed possible that most of the ‘Neolithic’ pottery published in 1971 in fact dated from the Middle or Late Bronze Age. It was only in the final phase that the monument was enclosed by any kind of earthwork (Bradley and Sheridan 2005).

  Figure 1.2. The distributions of Clava Cairns, recumbent stone circles and henge monuments in north-east Scotland and along the inner Moray Firth.

  The new interpretation was controversial and it is right to acknowledge that it was not accepted by some of the most influential researchers in this field (Gibson 2010a, 69; Welfare 2011, 261). The question was important, for it raised almost the same issues as the excavation at Tuach. Were small stone circles still being built in the later Bronze Age, and could they have been enclosed at that time? Was the outer perimeter of Croftmoraig related to the earthworks described as henges – a suggestion first made by Piggott and Simpson – and was the timber circle necessarily an early feature, as they had supposed? That last question gained a new relevance with the discovery of another timber circle outside the south entrance of Broomend of Crichie. In this case it was the latest prehistoric feature at the site and dated from the end of the Early Bronze Age – the same period as the earthworks at Pullyhour and Lairg. The only way to resolve this problem was to carry out small-scale fieldwork at Croftmoraig with the object of recovering samples suitable for AMS dating: a technique which was not available to the previous excavators.

  The limits of classification

  The three monographs published between 2000 and 2011 shared a common feature, for none was confined to new work at a single monument. Stages and Screens published excavations at Balnuaran of Clava, but also included the report on an earlier project at Newton of Petty. The main focus of The Moon and the Bonfire was fieldwork at Tomnaverie, but its results were compared with those of smaller projects at two other recumbent stone circles: Cothiemuir Wood and Aikey Brae. In the same way, the large-scale work at Broomend of Crichie was supplemented by another excavation at Pullyhour, and Stages and Screens also reported earlier projects at Migdale and Lairg. Each of the books compared the results of recent work with those of previously published excavations.

  There was some evidence of changes over time, particularly in the case of henge monuments, but these projects raised another issue which had been addressed by field survey but not by excavation. How much variation existed between monuments of the same type? That was especially important in the case of Clava Cairns and recumbent stone circles which may have been built simultaneously in adjacent areas: the inner Moray Firth in one case, and north-east Scotland, in the other. Both groups of monuments occurred with some frequency over a limited area, but it was by no means clear whether nearby examples had taken quite the same forms. For example, the recently excavated stone circle at Cothiemuir Wood was very different from its neighbour at Old Keig, excavated by Gordon Childe in the 1930s (Childe 1934). Attempts to compare the structural sequences at the two sites have been controversial.

  For these reasons it seemed important to undertake excavation at two neighbouring monuments. That was possible in the Howe of Cromar where the recumbent stone circle at Tomnaverie had been excavated on a large scale in 1999 and 2000. In 2012 a smaller project located its neighbour at Waulkmill where the monument had been levelled in the early nineteenth century. The original objective was to compare these structures with one another. Ironically, a third monument of the same kind was damaged during tree planting at Hillhead in 2012 and partly excavated in the following year. As a result, Chapter 9 can compare three neighbouring circles which appear to belong to the same ‘type’.

  Patterns of later reuse

  Our excavations between 1994 and 2008 had provided evidence for a widespread but unexpected phenomenon. The passage graves at Baln
uaran of Clava contained secondary cremations dating from the Late Bronze Age, and a pit containing cremated bone was excavated through the filling of an older ring cairn at Newton of Petty. At Tomnaverie, the central area of the recumbent stone circle was associated with a large deposit of burnt bone which accumulated a thousand years after the construction of the stone circle. A striking variant of this practice was identified at Aikey Brae. In this case no human remains were discovered, but the perimeter of an older stone circle may have been rebuilt at this time.

  The direct dating of cremated bone has resulted in the identification of other monuments with similar evidence of reuse. With improved understanding of the ceramic sequence in Scotland the same phenomenon can be recognised at the well-known sites of Old Keig (Childe 1934) and Loanhead of Daviot (Kilbride-Jones 1935). Smaller and later ring cairns at Cairnwell (Rees 1997), Balnuaran of Clava (Bradley 2000, 43–44) and on the Sands of Forvie (Ralston and Sabine 2000, chapter 2) share structural features in common with older monuments, and during the course of this project the connection was reinforced by the excavation of a similar monument at Laikenbuie – the results are presented in Chapter 6.

  There were other developments during the later first millennium BC. A large Iron Age roundhouse was built inside the recumbent stone circle at Strichen (Phillips et al. 2006), and another was erected beside the Candle Stane (Cameron 1999). A third example was known from surface evidence at Loudon Wood (Welfare 2011, 161). It was an open question how far the building of all these structures was influenced by a single body of ideas. Were there consistent phases of reuse at older stone monuments in Scotland? That is particularly relevant to the later Bronze Age. How does such evidence compare with what is known about structures that were newly built at the same time? In the event four of the monuments considered here – the Hill of Tuach, Croftmoraig, Hillhead and possibly Waulkmill – could have seen secondary activity. That was not the end of the sequence in Northern Britain, as the small ring cairn at Laikenbuie was built in the Early Iron Age.

  Developments during the project

  Inevitably new questions arose during the course of the fieldwork.

  It was already known that the reuse of Scottish stone circles continued after the end of the Bronze Age. Occasional Roman Iron Age artefacts had been found in association with older monuments, including a ring cairn at Monquitter (Anderson 1902), and a large structure of the same kind at Balnuaran of Clava was associated with a cremation burial dating from the Pictish period. It had been suggested that early medieval sculptures were erected at the sites of much older monuments, including Broomend of Crichie (Clarke 2007).

  What was surprising was that one of the monuments excavated during this project had also been reused during the first millennium AD. This was the stone circle at Waulkmill, which was associated with an unusual Roman Iron Age cemetery. In retrospect its discovery could have been predicted as two or three graves of this date had been discovered 350 m from the monument in 1898. Fieldwork in 2012 added another four, one of which was very similar to a burial found in the nineteenth century. The existence of these deposits was unexpected, but not unique, and it raised a new set of problems. It was tempting to associate them with a series of enormous stone roundhouses which also occur in the Howe of Cromar, but they had not been investigated in modern times.

  An opportunity to remedy this occurred in the following year when a large circular structure at Hillhead was damaged by forestry. It was intervisible with Waulkmill. As a result the structure was excavated and is preserved as a public monument. Unfortunately, the identification of this feature as a large roundhouse was made when its remains were under trees and has turned out to be incorrect. Against all expectation this feature proved to be the site of yet another recumbent stone circle. There was no indication of any Iron Age activity. Parts of the monument were intact and even retained structural elements that had not been observed at stone circles since the nineteenth century. Its chronology is established by a series of radiocarbon dates, and the site is also notable for including an in situ cremation pyre and 12,250 pieces of broken quartz. The latter had its counterpart at Croftmoraig where hardly any of this material survives.

  This unexpected discovery has had an influence over the way in which the new programme of research is being published. When the true character of the site became apparent, reports on Croftmoraig, Tuach, Waulkmill and Laikenbuie had been drafted as journal articles. Now it was possible to embark on a more ambitious analysis, comparing the character of five different monuments. Moreover, the identification of a stone circle at Hillhead made it possible to discuss the similarities and differences between three excavated monuments in the Howe of Cromar. They were located within sight of one another and may have been used concurrently. At the same time, the detailed sequence identified at Hillhead – a site which had never been investigated before – shed new light on the other monuments. It was clear that it would make more sense to bring these reports together in a single volume rather than submit them to journals that would be issued at different times.

  One advantage of the new procedure is that the results of these separate projects can be compared directly with one another, and the separate discussions that had been included in different articles have been brought together to make up Part 2 of this study.

  This account begins with Hillhead. Somewhat ironically, it was the last to be investigated and the only monument whose excavation was not planned at the beginning of this project. Because it was well-preserved and was used over such a long period it provides a vital source of comparison with the other sites.

  CHAPTER TWO

  Excavations at Hillhead, Tarland, Aberdeenshire: a recumbent stone circle and its history

  Richard Bradley and Amanda Clarke

  Cromar is a natural basin 50 km west of Aberdeen, in between the Grampian Mountains and the fertile soils extending to the North Sea. It was described by the Edinburgh antiquary John Stuart (1854, 258) in the earliest account of the archaeology of the region:

  The district of Cromar ... is about eight miles [12.9 km] in length from north to south, by four [6.4 km] in breadth from east to west, surrounded by hills, which isolate it from the adjoining country. It lies in between the rivers Dee and Don, at an equal distance from each ... On the south, the district is divided from Deeside by a range [of] hills. A natural opening in this direction conducts the road from Cromar to the south.

  At the time that Stuart was writing the low ground included a loch, but it has since been drained. The basin contains a variety of prehistoric monuments, including two long cairns, round cairns, roundhouses, enclosures, souterrains and a possible hillfort. In 2000, much of the cultivated land was examined by fieldwalking (Fig. 2.1).

  The prehistoric monument at Hillhead overlooks the lower ground which contained two other stone circles (Fig. 2.2). They include Tomnaverie, which was excavated in 1999–2000 (Bradley 2005, Chapter 2), and Waulkmill which forms the subject of Chapter 3. Another example, the Blue Cairn is on the western edge of the basin, but has not seen any modern investigation (Welfare 2011, 15). Hillhead is located in a saddle beside a long-established drove road communicating between Cromar and another basin to the east, the Howe of Cushnie. The stone circle was built at a height of 345 m and was situated on the watershed of the Dee and the Don (NJ 5072 0714). The site commands an extensive view to the south and southwest, extending as far as Lochnagar and the mountains of South Deeside. The stone circles of Tomnaverie and Waulkmill are also visible from the monument. A remarkable feature is that the site at Hillhead was constructed at the first point where a visitor climbing the slope from the west would glimpse one of the most prominent landmarks in northeast Scotland, the summit of Mither Tap of Bennachie. That happens at the exact location of the ring cairn. The mountain can be identified from inside the monument, but any further down the slope to the southwest it is invisible. The area above the excavated site may have contained other monuments which have been located during field
work by Moyra Simon and Jane Summers. They include the sites of several small roundhouses, one of which dates from the Late Bronze Age. Their results are summarised in a later section of this chapter.

  The circle was first identified by Ken Cooper as a bank which was covered by trees when he saw it. At the time the site was known as Blackhills. He made the entirely reasonable suggestion that it was the remains of a large roundhouse like those at Old and New Kinord on the lower ground 9 km away. He also noted the presence of quartz on the site and the position of a flat stone in the centre of the enclosure. Although he published a note of his observations (Cooper 1998), Hillhead remained little known. No one seems to have been aware of its archaeological significance when it was first forested and it was only when the trees had been removed and trenches had been excavated for a second phase of planting that its potential importance was recognised by the landowner. Following discussion between the McRobert Estate and Aberdeenshire Council it was agreed that the site should be preserved and that excavation should take place there.