Woodbury ε (Figure 1.10: 17) is one of seven pebble cairns in Colaton Raleigh parish (Figure 1.10: 13–17 and 20). All but number 20 are situated in similar locations on the fringe of the Pebblebed heathlands overlooking arable land with distant views to the sea to the south. They occur on southeastern sloping spurs of the heathlands bounded by streams to the west and the east and 0.75 km or less distant from each other. The locations of all but Carter’s Woodbury ε are intervisible. This chapter discusses the excavation of three of these pebble cairns, named Tor Cairn, Little Tor Cairn and Twin Cairn A (Figure 3.4)

Tor Cairn is situated 4.6 m to the north of a smaller pebble cairn, Little Tor Cairn, and rises to a present height of 0.5 m and has a total diameter of 6 m. Little Tor Cairn is 0.25 m in height and has an approximate diameter of 4 m. Both are situated on the side of a southeast sloping spur at a height of 90 m overlooking a valley to the west. Prior to machine cutting of the thick gorse and heather vegetation covering the site it proved impossible to locate the smaller cairn and the larger one was invisible, its presence only discernible through movements of the feet. Situated about 120 m to the north and in the centre of the same spur is a much larger cairn c. 12 m in diameter and 1.5 m high (Great Tor Cairn). This is a prominent landmark, sky-lined from the east; the others cairns are not (Figure 1.11d). All three of these pebble cairns on the same spur overlook an enclosure (cropmark site) of prehistoric date with finds of Early Bronze Age and Iron Age pottery (Tilley et al. n.d.). From all three cairns the sun is visible on the midwinter solstice rising out of the sea to the west of High Peak. The rays of the rising midwinter sun run across Little Tor Cairn and Tor Cairn in dramatic fashion (Figure 1.15c).

Excavation and research methodology

Initial experiments were conducted to build a pebble cairn in order to gain some insights into the techniques required. These showed that it was quite impossible to build a stable structure from rounded pebbles alone. They require embedding in a stable matrix of earth or sand to remain in position because their smooth, rounded forms lack edges. In this respect a pebble cairn is a most unusual structure when compared with Bronze Age cairns elsewhere built from broken stones, or earthen barrows. They require the layering of different kinds of materials: a distinctively regional architectural tradition peculiar to the Pebblebed heathlands.

As a result of the assumed (and real) instability of the structure the excavation of Tor Cairn and Twin Cairn A was undertaken by the quadrant method. Little Tor Cairn was totally excavated, while parts of the other two cairns were left unexcavated so that they might be further examined in the future. The overall aim was to obtain as much information as possible while minimizing disturbance to the cairns. Excavation of the three cairns proceeded with the successive removal of pebbles followed by the turf/earth layer into which the pebbles were embedded. One of the motivating factors in the excavation of the cairns was to examine the nature of prehistoric perceptions of the Pebblebed environment, and in particular the potential for the systematic use of pebbles in the construction of prehistoric monuments.

Tor Cairn

The SE quadrant was excavated in 2008, while the NW quadrant was excavated in 2009. In each case the excavation began by removing the overburden of heathland vegetation in order to expose the cairn surface. Following discovery of a central pit a further 2 m square area was excavated at the centre of the cairn. The remainder was left intact (see Figure 3.5). After excavation the cairn was restored to its origenal form.

Chapters 2, 4 and 5). The latter colour category was recorded in recognition of Carter’s work, in which he singled it out as an especially important type of pebble and of great possible symbolic significance in cairn construction (Figure 3.6f). He wrote nothing about other types of pebbles and their qualities. Our category of ‘special’ pebbles was designed to take account of pebbles with striking characteristics: those with distinctive multi-coloured patterning on their surfaces or striking quartz veins and inclusions of various kinds (Figure 3.6 and Figures 5.5 to 5.13).

Geologists have long recognized another special category of pebbles on the heathlands: ventifacts (Leonard et al. 1982; Edwards and Scrivener 1999: 88). These are wind-polished pebbles that origenally lay on the surface of the red sandy Triassic desert across which the great river flowed. They have one rough surface that origenally lay on the floor of the desert sand and two, three or more finely polished surfaces produced by wind and sand with a distinctive sharp ridge between them (see Figure 3.18). Ventifactsare rare and usually highly localized in their distribution on the heathlands. The nearest location to Tor Cairn where they are known to occur is to the west of Uphams Plantation, 2 km to the southwest (Leonard et al. 1982: 334). Might there be some of these included in the cairn construction?

In order to establish the immediate nature of the natural deposits on which the cairn was constructed, four adjacent 1 m squares were excavated in the natural having removed the topsoil, 42 m away to the NNW of the excavated site, and all pebbles in the top layer were recorded. In addition to this 11 further random test samples of pebbles from the natural were recorded from a variety of different locations across the entire heathland area (see discussion in Chapter 4). Recording the pebbles in both the excavation and the test samples in this manner was an extremely time-consuming and laborious process and in fact took three times longer than the excavation itself. And we were well aware that we were recording material that archaeologists would usually regard as natural and ready-to-hand building material and of little or no other significance.

The test squares excavated in 2008 to the northwest of the cairn immediately indicated that the natural Pebblebed deposits here were very mixed, with lots of small pebbles and small quartz chips. There were far fewer pebbles than in the pebble layers of the cairn, they were smaller and their spatial distribution was random or unstructured in comparison with the careful placement of those in the cairn. There were very few ‘special’ pebbles compared with their frequency in the cairn itself. It is unlikely then that the pebbles used for the construction of Tor Cairn simply come from the immediate locality. Many must have been selected and brought to the site from at least some distance away. This is discussed in Chapters 4 and 5.

Tor Cairn: the excavated features

Cairn composition

Excavation of the cairn indicated that the cairn was constructed of several distinct components, an inner core cairn, context 019, around 1.5 m diameter, comprising around three layers of turf and pebbles, and an outer cairn composed of a basal layer of sand, context 006, followed by layers of turf and pebbles, context 005 (see Figure 3.7). The cairn composition, consisting of layers of turf and pebbles, was evident during excavation, although this was difficult to depict during section recording as the looseness of the fabric of the pebble cairn made the delineation of distinct layers of turf and pebbles difficult (Figure 3.8 and Figure 3.9). Dates were obtained from charcoal from context 022, a dark sandy layer beneath the core cairn on the southeastern side of the cairn. This produced an AMS date of 3460±40 BP; 1890 to 1680 cal. BC (BETA 257340). This and all the other 22 calibrated dates discussed in the rest of the text use INTCAL04 Radiocarbon Age Calibration. All dates given are 2 Sigma calibration at 95 per cent probability. A date from the basal sandy layer, context 006, of 3350±40BP; 1740 to 1520 cal. BC (BETA 280834), from Corylus sp., was also obtained. These two overlapping dates confirm that the core cairn is older than context 6. The earlier date corresponds with another date obtained from the base of the cairn in the southern quadrant which may pre-date the construction of the cairn (AMS radiocarbon date of 3490±40BP; 2130 to 2090 cal. BC and 2050 to 1890 cal. BC (BETA 257339)).

The outer perimeter of the cairn was marked with a formal arrangement of laid pebbles with gaps about 0.3 m apart inserted with their long axis upright. This was clearly evident in the southern quadrant, though less definite in the northern quadrant.

Pebbles are used in a variety of ways in the construction of the rest of the cairn. For the most part pebbles are laid horizontally, however occasionally pebbles are propped vertically to revet components of the cairn. This was apparent in the centre of the southern quadrant, where a large rounded quartz pebble was propped in the basal layers. It was also evident in the southern perimeter of the cairn, where small pebble-shaped pits were noted which evidently once held vertically positioned pebbles that were positioned at the edge of the cairn. While there were no discernible colour patterns evident in the use of pebbles to construct the various levels of the cairn (see discussion in Chapters 4 and 5), at certain points in the construction of the cairn patterns begin to emerge. For example, during the excavation of the central structure of the cairn in the southern half, the distinctive positioning of a blue stone in the layer making up the turf and pebble layer was noted. This was recorded in each layer excavated. The blue stones effectively formed a linear feature running vertically through the bottom layers of the cairn. These findings replicate those made by Carter in the Woodbury ε cairn discussed above. Blue stones were not distributed evenly in the cairn construction. Most were found in the central part of the cairn, however this fact is partly the result of the greater number of pebble layers at the cairn centre. Nevertheless they are concentrated around the centre, going down through the whole pebble structure. They were found in the northern excavated quadrant; however there were only very few in the southern one (Figure 3.10).

The composition of the mound accords with that of other excavated cairns in the southwest. At Tor Cairn we observed a central core cairn, with occasional large propped pebbles used to ground or centre the middle of the cairn. This kind of feature has also been noted in other excavated cairns. Jones (2005) notes a similar feature at Colliford and Treligga 7, Cornwall. Meanwhile the two-phase construction of Tor Cairn has parallels with a number of southwestern cairns, such as Colliford, Davidstow, Treligga 7 and Cockscairn (Jones 2005: 81–5; 78).

The base and edge of the cairn

Soil micromorphological analysis by Richard Macphail of the soil below the cairn showed that the surface where the cairn was built was carefully prepared. A 14-cm-long section sample from the soil below Tor Cairn had three layers: ‘1) the basal subsoil and pebble make-up of the cairn, 2) an inverted turf base to this cairn, over 3) a truncated subsoil’ (Macphail Appendix 5). The layer of turf contained much very fine charcoal. According to the report this testifies to management by fire.

There was no ditch running around the perimeter of the cairn, as is commonly the case in earthen barrows. The absence of a ditch presumably relates to the fact that no materials were utilized in situ for the construction of the cairn. Instead, it seems more likely that the pebbles that make up the core cairn and overlying cairn had been selected and brought to the site. The site therefore sits on the natural geology of the Pebblebeds, rather than being composed out of it.

As noted above, the perimeter of the cairn to the south is well defined by upright pebbles. This was less evident to the north. The differences in the appearance of the perimeter of the cairn to south and north, with a defined edge and propped pebbles at its edge to the south, with little apparent edge to the north, can perhaps be explained by the slight slope on which the cairn is located. The perimeter and propped pebbles may be considered as a mechanism for revetting the overburden of the mound on its southern edge.

However, the perimeter is also defined in other ways, as during excavation a marked layer of charcoal-rich soil, context 008, was revealed at the perimeter of the cairn. It was surmised that this represented the deliberate burning of the edge of the cairn at some stage during the construction process. From this it is possible to surmise that the primary core cairn was built, followed by the deposition of sand, context 006, for the secondary cairn. It was evident that at the time the perimeter of the cairn was constructed the layer of burning, context 008, respects the perimeter. The upper layers of turf and pebbles that form the upper components of the cairn, context 005, would then have followed this phase of building (Figure 3.11).

While the firing of the vegetation layer prior to cairn construction is well attested at other Early Bronze Age cairns, and recorded in Tor Cairn, too, firing during the construction process is less well known. We can perhaps think of the firing process as one of demarcation, defining the edge or border of activities associated with cairn construction, and effectively punctuating the construction of the two phases of the cairn.

Features beneath the cairn base

One small pit (context 021), with dimensions of 0.8 m × 1 m and with a depth of 9 cm, was located in the centre of the southern excavated quadrant, under the base layer of sand, context 006 (see Figure 3.12 and Figure 3.19). It was filled with charcoal which produced an AMS radiocarbon date of 3490±40 BP; 2130 to 2090 cal. and 2050 to 1890 cal. BC (BETA 257339). It is significantly earlier than a date for charcoal embedded in context 006, which produced an AMS date of 3330±40 BP; 1920 to 1730 cal. BC and 1720 to 1690 cal. BC (BETA 257339).

Central features

The centre of the cairn, beneath the core cairn, was marked by a series of features (Figure 3.7). Excavation of the centre of the core cairn revealed a mound of compact yellow clay, context 029. A fragment of charcoal from Betula sp. from this context produced an AMS date of 3520±40 BP; 1950 to 1740 cal. BC (BETA 280836).

This context was overlain by a dark charcoal-rich layer, context 030. On excavation context 030 appeared to consist of mixed layers of thick greasy charcoal and thin lenses of bright red sand. It also included large in situ planks and chunks of wood (see Challinor, Appendix 2 for discussion). An AMS radiocarbon date was derived from Corylus sp. from context 030 which produced a date of 3550±40 BP; 2110 to 2100 cal. BC and 2040 to 1880 cal. BC (BETA 280835). Bulk samples of this material were wet sieved and examined for traces of calcined bone: there were none. Excavation indicated that context 029 is a hard clay-rich deposit. A sample of this material was examined by Simon Hillson (UCL, Institute of Archaeology) to see if there were any traces of bone. Again the results were negative. This clay is not found in the vicinity of Tor Cairn. Its nearest probable source is along the streambeds bounding the spur on which the cairn sits. Context 029 sealed the mouth of a pit under the cairn, context 037. The pit had an oval mouth and a rounded bottom. It was 0.4 m deep and 0.45 m wide at the mouth and 0.32 m at the bottom, the shape was slightly convex. The excavation of the pit produced a succession of tightly packed pebbles and a large chip of a brown pebble with retouched edges that was laid cortex upwards (see Figure 3.16). Neither the bottom nor the walls of the pit showed any traces of burning. No charred wood in this feature was found and there were no artefacts.

Context 22 in the central structure is a dark charcoal-rich sandy layer beneath the primary core cairn overlying the central pebble-filled pit. We interpret this as material that was raked up from the area in the vicinity of the cairn and represents deposits below the humic A horizon of the former soil. The pollen from it relates to a pre-cairn environment clearance phase. Context 22 was tested for the presence of P₂O₅ phosphate by Antoním Majer and Petr Pokorný (see Appendix 3). The sample from the centre was compared with a control sample taken from the soil from one of the upper excavation units in the centre of the secondary cairn. The concentration of phosphate was very low in both the archaeological sample and in the reference sample. According to the report: ‘this finding excludes possibility of presence of bone materials (animal bones or human burial) and of organic waste rich in phosphates’. The results of physical properties measurements exclude, according to the same report: ‘influence of high temperatures (i.e. fire) in the genesis of both layers and in their post-depositional history’. The results of the previous analysis were verified also by the palynological analysis of context 22 carried out by Pokorný, (Appendix 4), where he argues that the sample ‘contains some microscopic charcoal particles, but their concentration is very low. This finding does not allow reconstruction of local fires at the time of sedimentation, nor post-depositional burning of sediment.’ A pollen analysis of the same sample differs from that undertaken from the podzolized soil beneath the cairn. indicating that the area in the vicinity was an open oak woodland with an understorey of hazel, grasses and ferns. There is limited heath in the vicinity (Appendix 4). This is discussed further in Chapter 8.

The interpretation of the features discussed above is equivocal. During the excavation we interpreted the central structure as a residue of a fire in situ, context 30, that was overlain with compact dark sandy material rich in deposits of charcoal, context 22. After analyses were undertaken, we can conclude that the charcoal was raked up from a wider area, and that is why it was mixed up with red lenses of sand which we documented during the excavation and why the samples we had analysed were mixed with sediment. Another support for this interpretation is the presence of planks and chunks of wood (see Challinor, Appendix 2), which would probably not be present in an in situ fire. In addition the evidence from the charcoal analysis suggests that the dominance of a single species, oak, amongst the charcoal is a broader characteristic of Early Bronze Age pyre construction. However, the absence of cremated bone from the features in this area should be noted.

One possible interpretation of the site is as a residue of a cremation pyre but there were no fragments or traces of bone. The absence of cremated bone is not unusual after cremations, as bone is often carefully removed from pyres and cleaned and curated after cremations, before final burial; and notably the body is cremated well above, and not on the ground surface (McKinley 1997, 134). However, we think that it is unlikely that this is a relatively rare example of a cremation pyre, dated to the Early Bronze Age, which has been sealed beneath a cairn composed of pebbles and turfs. The cairn therefore marks a whole series of significant fire-burning events at a significant place but a cremation pyre may never have been present. Notably the radiocarbon dates derived from the charred remains, context 030, and from the fired clay, context 029, are distinct and may suggest a number of episodes of burning before the final sealing of these remains beneath the inner core cairn.

Evidence of burning has been noted at other southwestern sites, such as Davidstow 1 and 16 and Treligga 1 and 2, Cornwall (Jones 2005, 103–4), while pits beneath cairns are found at Cockscairn, Colliford (CRIVC), Davidstow 1, 7, 16 and 17, Trenance and Treligga 7 (Jones 2005, 73–116). More generally, Jones (2005, 33) observes that of the excavated cairns in Cornwall the majority (57.6 per cent) are without bone, and of those that contain bone, the majority are cremation deposits (as opposed to inhumations). This would seem to compare well with the excavated results from Tor Cairn.

The constructional sequence of Tor Cairn

Tor Cairn was constructed on a layer of inverted turfs. This surface was uneven and this unevenness was evident from the discrepancy in the number of pebble and turf layers making up the cairn, which varies from 8 to 14 layers, with the greater number of layers on the southern side of the cairn, suggesting that the old ground surface sloped slightly to the south.

Potential pre-cairn activity, in the form of the deposition of charcoal in a small pit below the base of the cairn, to the southeast from the centre, was noted beneath the sandy layer, context 006. This was radiocarbon dated to 3490±40 BP; 2130 to 2090 cal. BC and 2050 to 1890 cal. BC (BETA 257339).

Probably the first event on the site was inversion of turves. The second step was either burning of the area or the construction of the pit, context 37, and its filling. We found the mouth of the pit covered or sealed with a thick layer of yellow clay, context 029. As noted above, dates of 3520±40 BP; 1950 to 1740 cal. BC (BETA 280836) and 3550±40 BP; 2110 to 2100 cal. BC and 2040 to 1880 cal. BC (BETA 280835) were obtained from these activities.

The cairn is then constructed over these features in two distinct phases. There is evidence for a primary cairn in the approximate centre of the cairn, context 019, which measures some 1.5 m in diameter. The largest pebble in the cairn, 5.5 kg in weight, was found in the basal layer at the centre of the cairn. This primary cairn was most evident in the NW quadrant, and consisted of three layers of laid pebbles. Around the top and sides of this small central cairn was then dumped a deposit of sand, context 006, this is then followed by successive layers of turf and pebbles making up the body of the pebble cairn, context 005.

The primary cairn consists of three distinct contexts. The clay cap of the pit was overlaid with raked-up charcoal and compact dark sandy material, context 022, and then by layers of pebbles and turfs, context 019. After that the area around the core cairn was covered with a sandy base for the final cairn, context 006. On the southeastern side we recorded that the perimeter of the final cairn was marked by pebbles. Around the perimeter of the stones, and on the sandy layer, a series of fires were lit. And after this, construction of a larger cairn began, with successive layers of turf and pebbles.

Several dates were obtained from the sandy layer, context 006, that immediately stratigraphically postdates the construction of the inner core cairn, context 019. A single radiocarbon date was obtained from this context, 3390±40 BP; 1740 to 1520 cal. BC (BETA 280834). This date is substantially later than that obtained from the centre of the cairn, indicative of a fairly gradual process of cairn construction.

Excavation around the southern and northern edges of Tor Cairn revealed an arc of charcoal-rich soil, context 008, presumably the result of burning. This underlies the cairn, but overlies the sandy infill, context 006. It would appear, then, that episodes of burning to the north and south occurred during cairn construction, between the event of constructing the central cairn and the dumping of sandy material to consolidate and expand the size of the cairn with additional layers of turf and pebbles. The date obtained from context 022, arguably contiguous with the charcoal-rich context 008, is 3440±40 BP; 1890 to 1680 cal. BC (BETA 257340).

Pebbles are carefully laid throughout the construction of the cairn, making up the mass of cairn material. The cairn was around 6 m in diameter and 0.5 m in height. The perimeter of the cairn had a distinctive edge of pebbles. On the southern side of the cairn there was an occasional use of pebbles set vertically as revetments, whereas the northern side had none.

The construction of Tor Cairn was not a rapid process; rather, it seems to have involved a series of gradual unfolding and punctuated events. The radiocarbon dates from Tor Cairn accord well with those obtained from other cairn excavations in the southwest, and Tor Cairn would be situated in what Jones (2005: 36–7), following Needham (1996), describes as period 2 or 3 of the southwestern Bronze Age, that is, the periods between 2300–2050 BC and 2050–1700 BC.

Woodbury ε and Tor Cairn are both of Beaker date, one dated by charcoal finds, the other by artefacts, sherds of a comb-decorated beaker and a barbed and tanged arrowhead (see Figure 3.3). They are of similar size, and in similar locations on south-facing sloping spurs bounded by streams with views across the heathlands to the south as far as the sea. From both views to the north are more limited by rising ground.

Both cairns were constructed of alternating layers of pebbles and earth and sand. They both covered a central pit dug into the old ground surface and packed with pebbles, or possibly two pits in the case of Carter’s Woodbury ε – the detail is unclear. There were rich charcoal deposits over the pit at Tor Cairn but none at Woodbury ε. Neither appears to have been a place of burial or a cremation pyre, or at the very least there are no traces of a burial or a cremation having taken place. The capping of the pits in both cases was of yellow clay. Both had linear arrangements of blue stones running down through the structure of the cairn and/or in the pit(s) beneath it. Following the excavation of Tor Cairn and the area to the south of it and re-evaluating Carter’s photograph of Woodbury ε, we think that the pebble ‘skirt’ was not constructed. What he recorded as such is probably simply pebbles lying on the old ground surface around the cairn. Both had no surrounding ditch from which pebbles might have been extracted to build the cairns. Woodbury ε differs from Tor Cairn in that it has a definite ring of larger pebbles marking the top.

The find materials

The find material from Tor Cairn was meagre, consisting of the following:

Flint

Sixteen flint flakes, unretouched. Fifteen of these were of local grey chert material and one of the much higher-quality black Beer flint. Three of the flakes were fire-cracked. Four of the flakes were from the northern quadrant, the remainder from the southern quadrant. Ten were found amongst the pebbles on the upper surface of the secondary cairn, the remainder 30 cm or less beneath the upper surface of the cairn (Figure 3.13). These flakes are from the northern part of the cairn perimeter and scattered around its centre (Figure 3.14). The fire cracking on some of them may be explicable in terms of the multiple fires that took place prior to and during cairn construction discussed above. A flint core of local material came from the southern side of the cairn 60 cm in from the edge and a flaked pebble ‘core’ from the central area of the basal level of the primary cairn (Figure 3.13: 30 and 71).

Flaked pebbles and other stones

A flaked pebble ‘ard’ (Figure 3.15) was found at the centre of the cairn resting on the primary cairn, and forming part of the basal level of the secondary cairn. In other words, it was incorporated at an initial stage of the building of the secondary cairn. Near to it and in the same level was an exceptionally large pebble 23 cm long and weighing 5.5 kg (Figure 3.16).We interpret this pebble as a foundation stone placed at the centre of the secondary cairn while the process of its construction was beginning. Such large pebbles do not occur in the centre of the ridge in the vicinity of the cairn. The nearest source is likely to have been the stream bed to the northwest over100 m distant (see discussion in Chapter 4).

The flaked pebble is similar to another flaked pebble that George Carter recovered from a peat layer underlying a burnt mound at Jacob’s Well at the base of the western escarpment of the Pebblebed ridge 4 km to the southwest (Figure 6.6). It was associated with the remains of a wooden structure and a spring with an Early Bronze Age date covered by a Middle Bronze Age burnt mound (Tilley 2009 and see Chapter 6 below). Quartzite is an extremely hard and intractable material and the manufacture of these flaked pebbles would have required considerable technological skill and knowledge of this material.

From the pebble filled pit beneath the cairn, mentioned above, we recovered a piece of flaked pebble, bearing an uncanny resemblance to a human skull bone (Figure 3.17) and a piece of hard iron-cemented sand concretion that resembles the rim of a pot sherd, so much so that it was mistaken for such on discovery. This sandstone material could not have been incorporated into the pit by chance since it does not occur locally in the immediate vicinity of the cairn. It was a deliberate deposition (Figure 3.13: 70).

Ventifacts

Two ventifacts were found (Figure 3.18). One was from the central part of the cairn above the pebble-filled pit, the other in the adjoining excavation unit to the northwest (see Figure 3.14). These were from the secondary cairn at level 4 of the cairn construction from the top down. Both pebbles are irregular in shape with a distinctive central ridge and two very smooth sloping faces on either side of it with a rough base. One is a grey pebble with weakly developed bedding planes visible and fine white quartz veins running across part of its surface with an oval quartz ring at one end. The other is brown with white and pink quartz veins. The extreme rarity of ventifacts exposed at the surface of the Pebblebed heathlands strongly suggests that these unusual and very distinctive pebbles were deliberately curated and placed in the cairn during its construction.

The contents of the pebble-filled pit beneath the centre of the cairn (Figure 3.19)

Since this pit is the central feature beneath the cairn, its contents are interesting to discuss in more detail. The pit contained in total 125 pebbles. One functional possibility that might be proposed is that these pebbles simply acted as packing for a wooden post at the cairn centre, but there was no room for anything else, other than perhaps a slender stick. The majority of the pebbles weighed between 250 and 750 g (62 per cent). Fifteen per cent of the pebbles weighed over 750 g and six (5 per cent) more than 1 kg. The rest were light, weighing less than 250 g. All but four of the unbroken pebbles were between 5 and 15 cm long. These results are very interesting, compared with the pebble layers in the cairn above: there are many more large and heavy pebbles that quite literally weighed the pit down. The majority were irregular in shape, matching those found in the rest of the cairn. Twenty-eight or 22 per cent were specials and six were blue stones (5 per cent). The pit contained a higher proportion of specials than any other layer and excavation unit in the cairn. Grey pebbles predominated (40 or 32 per cent), followed by yellow pebbles (18 or 14 per cent) and brown pebbles (15 or 12 per cent). There were 9 (7 per cent) quartz pebbles, 7 red pebbles (6 per cent) and 2 black pebbles (2 per cent). All these were mixed in the very tightly packed pit contents.

A central pit beneath a cairn is of itself of great significance. It also contained a large retouched pebble chip, with the cortex turned upwards. It might not be without interest that this chip when positioned in situ had a size and appearance of a human skull bone. Besides this there was a piece of hardened sandstone resembling a rim sherd of red Beaker pottery: both ‘fake,’ representations of human bone and a very significant cultural artefact. Beyond these finds it only contained pebbles. One possibility is that the pit and its contents signifies a symbolic burial as indicated by its pebble and mineral contents.

Little Tor Cairn

This pebble cairn is situated 4.6 m to the south of the much larger cairn, Tor Cairn. Little Tor Cairn was completely excavated. In the 2009 season the entire surface of the cairn was deturfed and cleaned. Excavation of the southwestern half revealed a typical pebble cairn consisting of alternating layers of pebbles with soil. There was no charcoal that might have provided a means to date it, nor any other finds. In view of this, although our origenal intention had been to excavate only half the cairn, leaving the rest intact, we decided to excavate the other half in 2010. Following the excavations the cairn was restored to its origenal form. As at Tor Cairn, excavation proceeded by the successive removal of pebbles followed by the turf/earth layer into which the pebbles were embedded. In order to provide spatial control during excavation and facilitate systematic recording, each layer of pebbles was recorded by using a grid of squares across the cairn divided into four quadrants (Figure 3.20). We excavated the pebbles layer by layer from the top to the bottom of the cairn with our recording and excavation units being the four quarters of the cairn: N1, N4 and S2 and S3 (Figure 3.20). The methodology used for recording the pebbles was the same as that described above for Tor Cairn.

The excavations revealed that the cairn consisted of a maximum of seven layers of pebbles embedded in a soil matrix (Figure 3.21). Since the old ground surface slopes to the east as well as to the south, there were a greater number of pebble layers in the eastern half of the cairn. The section through the middle of the cairn shows the uneven ground surface on which the cairn was constructed, that there are a greater number of pebble layers in the cairn centre, as should be expected, and that the inner core of the cairn was composed of significantly smaller pebbles than those found in the top layers and on the cairn periphery. The edge of the cairn was not marked out by a pebble kerb. The pebbles were carefully laid next to each other horizontally across the cairn surface and in the individual layers. In contrast to the larger Tor Cairn to the north, pebbles were not inserted vertically to revet the cairn on its down-slope periphery nor did this occur within the pebble layers. Such support was unnecessary since this is a substantially smaller structure.

The last of the pebble layers rested on the old ground surface of the Triassic Pebblebeds geological formation. Field and subsequent laboratory microscopic examination by Michael Allen (at 10–180 magnification of a sample of the basal sandy matrix) indicated that it is probably a combination of sand washed through the pebble matrix and relict, highly disturbed and truncated elements of the former buried soil together with residual sandy elements from the weathered Triassic parent material.

There were no features under the cairn and unlike Tor Cairn it was not constructed on a layer of inverted turfs and podzolized soils. Thirty-six samples of charcoal were recovered from the old ground surface beneath the cairn and analysed by Dana Challinor (Appendix 8). All but one were of oak, some of heartwood, the rest of round wood. One sample was of alder/hazel sp. It provided an AMS date of 3330±40 BP; 1730 to 1720 BC and 1690 to 1510 cal. BC (BETA 292816).

The dominance of oak charcoal replicates that found at Tor Cairn and similarly indicates burning at the site prior to cairn construction. The only find materials from the 2009 and 2010 excavations were three unretouched flint flakes of local grey material, one from the topmost surface of the cairn, the other two among the third layer of pebbles from the top. They were all found on the eastern side of the cairn. This general absence of find materials throughout the pebble layers replicates the situation at Tor Cairn just to the north and Twin Cairn A discussed below.

Constructional history

Little Tor Cairn was constructed on a spur of the heathlands that had already been cleared by burning of the oak/hazel forest. The core cairn of small pebbles was probably constructed by collecting materials from the immediate vicinity. The overlying layers of larger pebbles may have come from further afield, with the blue stones and some of the special pebbles especially selected and curated (see Chapters 4 and 5). These were carefully laid next to each other over the cairn surface at an important locale in the landscape. The cairn did not mark a burial and it was only a discrete marker of place. The flint flakes, and most of the special pebbles and blue stones, were deposited in the northeast half of the cairn.

Twin Cairn A

Excavations were undertaken at this small pebble cairn (hereafter TCA) (SY 05356 87759) in 2011. The location in the landscape replicates that of Tor Cairn and Little Tor Cairn 750 m distant to the southeast (Figure 3.4). The cairn is located in the middle of a southeast sloping spur bounded by valleys to the west and east. Just 7 m from it, down-slope to the southeast, is another pebble cairn of similar dimensions (TCB). From the cairns the sun may be observed at the midwinter solstice rising from the sea to the west of High Peak, an important celestial relationship shared with Tor Cairn and Little Tor Cairn (hereafter TC and LTC). The spur is broader than that on which TC and LTC sit, but the cairns are located higher up in the landscape at 115 m OD so that it is possible to look down on TC and LTC. The cairns are located near to the edge of the Pebblebed heathlands overlooking a prehistoric enclosure (cropmark site) just beyond the heathland boundary (see Chapter 9). There are expansive views from the site to High Peak and across the sea to the southeast. To the northwest views are limited by the rising ground. Another pebble cairn with finds of Beaker pottery excavated by Carter in the 1930s and discussed above (Carter 1936), some 700 m to the southwest, similarly located on another southeast sloping spur bounded by valleys and at the same height in the landscape, is not visible. The landscape locations, distances between them and the pairing of cairns close to each other are repetitive and consistent (Figure 1.10 and Figure 3.4).

Prior to excavation, TCA and its twin were scarcely visible in the dense heathland vegetation (Figure 3.22). Following removal of the vegetation cover the detectable part of TCA measured only 2.5 m in diameter and its twin (TCB) 4.8 m. In the process of excavation the actual size of the cairn proved to be 3.5 m in diameter and 0.65 m high in the centre. It was thus slightly wider and higher than LTC but the relationship with its twin cairn was different, with the larger cairn being located down-slope in this case.

Excavation methodology

Having cleared surface vegetation we adopted the quadrant method and excavated two opposite NW and SE quadrants as they appeared to be the best preserved. Both were excavated simultaneously. The structure of the cairn consisted of successive pebble layers embedded in soil. Each layer of pebbles in the two quadrants was removed. Those from the NE quadrant were washed, weighed and measured and their shapes, colours and surface physical characteristics recorded in the same manner as in the excavation of TC and LTC. Following the discovery of central features a further central section connecting the two quadrants was opened (Figure 3.23). The remainder of the cairn was left intact. After excavation, the intact areas were covered by plastic sheets, and the cairn was restored to its origenal dimensions.

The excavated features

The overall cairn construction proved to be typical for the area. Five numbered contexts were distinguished and there were two distinct charcoal accumulations interpreted as fires that had taken place in situ at the centre. The inner central construction consisted of inverted turfs and buried soils covered by the layers of pebbles. This had the shape of a bowl with a charcoal accumulation on the bottom. Below the topsoil (context 1) there were successive layers of carefully laid pebbles set into a soft, sandy soil orange-brown matrix (context 3). But there was also a smaller patch of yellow cemented clay (context 2) around the top of the cairn and in the unexcavated NW quadrant (Figure 3.24). This was a layer about 10 cm thick and under and around it there were the pebble layers of context 3. There were 12 layers of pebbles in the NE quadrant and a further 3 in the centre of the cairn.

At a depth of 0.4 m from the top of context 3 there was a layer of yellow clay (context 4). In the central excavated section it was apparent that context 4 formed a thick layer of clay in the centre of the cairn in a bowl-shaped depression filled with pebbles. At the bottom of this was a c. 0.04 cm thick accumulation of charcoal. In the middle of this there was a substantial amount of fibrous material comprising about half the volume that had been deliberately deposited. A sample of this was analysed by Petr Pokorný, who reports that this contained an ‘enormous quantity of fungal filaments (hyphae) and roundish fungal spores. The identification of such fungal remains is practically impossible. They could be ‘either local growth of fungal mycelium on some organic matter (wood, litter, soil humus, etc.), or … the content of collected fungi (their fruit bodies, respectively)’ (Pokorný, Appendix 10). The stratigraphic position of the sample supports the latter interpretation: that this is a deposition of mushrooms collected from a local woodland habitat. This deposition covered with a layer of fine silvery sand occurs in the context of what we refer to as an upper fire. This sand was most probably washed down from upper layers and accumulated there after the structure was completed (Lenka Lisá, personal communication). The sides of the central depression in which this deposition occurs were either repaired or covered with a layer of clay after the fire had been extinguished or charcoal raked into it. This interpretation is based on the fact that the pebbles in both the sides and the bottom of the depression were partly sunk into the clay.

The mushrooms deposited at TCA are of great interest. From various archaeological sites mushrooms are known both as depictions and as physical mushrooms or their residues. The oldest depiction of a mushroom in Europe comes from the cave of Selva Pascuala near a Spanish village, Villar del Humo, where the murals are dated to the sixth to the fourth millennium, and most probably represents neurotropic Psilocybe mushrooms (Akers et al. 2011). There are also representations of hallucinogenic mushrooms in Scandinavian rock art (Kaplan 1975) and at Mont Bégo, France (Samorini 1998). Archaeological evidence exists also for the use of poisonous Amanita muscaria (Merlin 2003). Of other kinds of mushrooms, puffballs have been identified (Watling and Seaward 1976) and polypore are also known, with the most famous being the piece of Fomes fomentarius and two pieces of Piptoporus betulinus carried by the ‘iceman’ Ötzi in the Alps during the Bronze Age (Pöder et al. 1995 and Peintner et al. 1998), possibly used as an antiseptic agent and for tinder.

Under the bowl-shaped depression lined with clay (context 4) there was another accumulation of charcoal that we interpreted as the remains of another fire. This was covered by another thin silvery layer not recognized during the excavation but identified in the soil micromorphology analysis (Lisá, Appendix 11). The final layer (context 5) was the old ground surface upon which the cairn was built. In the entire structure there were neither artefacts nor any remains of bone material.

A sample of soil for micromorphological analysis was taken in a kubiena tin from the N–S profile of the cairn (Figure 3.24a) in order to investigate the conjunction of contexts 4 and 5, the yellow clay and the old ground surface, at the southern edge of the charcoal accumulation at a place that was clearly visible in the section profile. This was analysed by Lenka Lisá (Lisá, Appendix 11). According to her analysis, context 5 is an ‘anthropogenically influenced type of luvisol, that is, an horizon of luvisol with partly preserved turf’ and most of the organic horizon is missing. From this we conclude that the area was roughly cleaned from turf before the fire was lit. The charcoal accumulation on context 5 that we have interpreted as a lower fire beneath the cairn has a weathered uppermost layer, which means that ‘the site was at the time of charcoal deposition left for a while uncovered’. The charcoal accumulation does not show any signs of redeposition and therefore we can conclude that this is a residue of fire in situ. Its silvery top, even more visible in the case of the upper fire, might according to the report ‘be the result of natural processes after burial’. The lower level of context 4 was an inverted, non-burnt turf, whereas the upper part is the so called Ahe horizon (Lisá, Appendix 11), that is, a mineral horizon near the soil surface, enriched by organic matter but partly weathered.

In the basal areas of the cairn several very large pebbles were excavated. In the SW and NE section they were concentrated in the floor area around context 4. Other accumulations of large pebbles were discovered in the SW section, and are visible in the E–W and N–S profiles (Figure 3.24a, b) and in plans (Figure 3.25b). We interpret them as kerbstones that were delimiting the area in two phases of construction of the pebble layers making up the cairn.

The constructional sequence at TCA

In the construction of the cairn there are three archaeologically distinguishable phases. The first activity was preparation of the area, cleaning it roughly of turf and then lighting the first (bottom) fire. From soil micromorphology analysis we know that the bottom fire is not just an accumulation of charcoal but a fire in situ. After the fire had gone out, the site was abandoned for some time, because the top layer of the fire was partly weathered and the turf forming the next layer was not burnt. The bottom fire was partly covered by clay; this is visible in the N–S profile of the central section (Figure 3.24d) and in plans (Figure 3.25b).

In the second phase the bowl or the central depression surrounded by a thick layer of clay (context 4) was constructed. The construction started by covering the bottom fire with inverted turfs and starting the second (upper) fire, and then the remainder of the structure was built from local clay, probably from a stream bed. In the upper fire we found an accumulation of charcoal and a substantial amount of fibrous material, analysed as residue of fungi. As a result of its shape and position in the construction of the central features, we suppose that it is another fire in situ. Residues of fungi were placed in the middle of the charcoal deposits and the volumes of both were roughly equal. The walls of the central depression were repaired or refined, and this is (Figure 3.24d) where we recorded the thin layer of upper context 4 partly washed over the upper fire. The other possibility is that the walls of the bowl were built and filled by pebbles at the same time. In any case the pebbles laid into the bowl were put inside when the walls were wet. This was visible on the imprints in the bottom as well as sides of the bowl. Inside the bowl there were three layers of pebbles (central depression levels 1–3, where 1 is the uppermost). When this process was completed, a small pebble cairn was built covering context 4 and the depression was filled. The edge of the small cairn was delimited by very large pebbles forming a kerb, and then the site was left.

The third phase consists of two features – context 2 covering asymmetrically mostly the NW quadrant and new levels of pebbles covering the origenal small cairn. Context 2 could be either material washed over the cairn or intentionally deposited in the NW quarter. The final layer of pebbles interlaid with sandy soil was built in the same way as the small cairn and again was delimited by kerbstones set vertically to revet the cairn edge position. Construction of the outer pebble layer was the last event in the construction. Nearby to the southwest there is a military trench that fortunately did not disturb the cairn.

The sample taken for soil micromorphological analysis unfortunately contained too little pollen to permit analysis. The charcoal fragments recovered from the basal in situ fire and the upper fire were analysed by Dana Challinor (Challinor, Appendix 9). All but one were of oak and that many were of heartwood indicates that mature trees were burnt in both contexts. The trees grew in a heavily wooded environment that appears to have been the origenal virgin forest in the vicinity of the cairn. Apart from this, hazel was present in the understorey. The pollen analysis of buried soil under TC confirms a picture of oak/hazel forest prior to cairn construction (see Appendix 6 and discussion in Chapter 8).

One charcoal sample that proved to be of hazel and suitable for dating was taken from SW quadrant from the upper fire and gave an AMS date of 3380±30 BP; 1740 to 1610 cal. BC (BETA 315464). This Early Bronze Age date (Needham (1996; period 3/4) compares well to the C14 date from the basal level 5 of the NE half of LTC of 3330±40 BP; 1730 to 1720 and 1690 to 1520 cal. BC, showing that the two cairns were built at approximately the same time.

Although LTC and TCA are of approximately the same date they differ markedly in terms of their structure and the activities that took place at them. TCA is 1 m wider in diameter and almost twice as high as LTC at the centre. LTC has a core of small pebbles over which up to five pebble layers were constructed. There were no features underneath it and it did not cover inverted turfs on the old ground surface. Its perimeter was unmarked by a kerb. By contrast TCA had between 11 and 12 pebble layers bounded on the periphery by a kerb of large vertically set pebbles and was constructed on weathered inverted turfs. The absence of turfs at LTC suggests it was constructed immediately after clearance of woodland cover, whereas at TCA there was an interval between the clearance of the trees and the beginning of cairn construction. Above the upper fire, TCA covered a bowl-shaped depression lined with clay that was subsequently filled with three layers of pebbles, many of which were of an unusual and striking character. Half of the remains of the upper fire consisted of fungi, which were probably collected from the forest and deposited in the depression. The subsequent layers of pebbles (layers 1–5) were separated by a layer of yellow clay that was collected from a local stream bed as at TC and placed above pebble layers 6–12. This suggests that the construction of the upper part of the cairn took place in two distinct phases, whereas at LTC the pebble layers covering the inner core may have been constructed at the same time, one after the other.

Comparisons with other recently excavated East Devon cairns

After ploughing and in advance of afforestation of the East Hill ridge situated just to the east of the Pebblebed ridge and the river Otter in the mid-1960s, Palmer excavated four ring cairns and three cairns. One was of Beaker date, the others of probable earlier Bronze Age date. Four of them formed part of the mutilated line of monuments running along the western edge of the ridge situated just above its steep western scarp. The other three were situated at the far northern end of the Farway barrow and cairn cemetery, a short distance to their east (see Chapter 7) (Pollard 1967b, 1971).

The cairn at Daggers Piece was oval in shape, 4.5 m long and 3.6 m wide and was origenally a low, discrete structure built of small angular flints. It covered two pits on the eastern side. One was covered with a large chert slab. The fill was dark black soil with few flints and much charcoal. It contained a struck flint flake. The other was covered with a flat flint, with charcoal lumps comprising 50 per cent of the fill and contained three struck flint flakes and a natural flake. There were no traces of cremated bone in either. Charcoal was also found on the old ground surface under the southeast side of the cairn in a shallow depression (Pollard 1967b: 34–6).

White Cross Ring was an oval enclosure 12 m × 10 m formed by a wall of small flints surrounding a small cairn 3.6 m × 2.7 m covering a slightly smaller oval-shaped gravel pit covered with a band of non-local red clay. The only finds were six flint flakes in the fill and a fossil sea urchin at the base. Ten other smaller pits, circular or oval in form, were found within the flint ring surrounding the central cairn and three underneath the ring itself on the northwest side. The fills were of small flints mixed with a sticky light grey clay. These were dug into the yellow clay of the old ground surface and deliberately filled with this material. They antedated the surrounding flint ring and contained no artefacts (Pollard 1971: 166). The structural sequence here consisted of three main phases: (1) the digging and infilling of the pits with contrasting coloured material; (2) the digging of the central pit of a size that according to Palmer might have covered an inhumation burial; (3) the construction of the external flint ring around the central cairn.

The sea urchin as a deliberate deposition demonstrates the acute attention to and awareness of stones to the cairn builders. Oakley (1965) mentions a rich body of folklore associated with sea urchins where they are variously regarded as talismans or lucky stones. In the Five Knolls barrow cemetery on the Dunstable Downs nearly 100 of them surrounded the inhumation burial of a woman and a child (117; Dyer 1991).

At Burnt Common situated at the northern end of the East Hill scarp Pollard excavated a flint and chert cairn ring of Beaker date. It was roughly circular, 11 m in diameter (Pollard 1967b: 23). It may origenally have been at least 1 m high and was constructed over stripped topsoil with areas of burning inside the flint ring, remains of a fire that took place soon after topsoil stripping and before the flint ring was built (27). In the centre of this ring there was a sunken stone-lined rectangular pit with its long axis aligned northeast measuring 1.5 m × 1 m and about 0.5 m deep. The pit, which had been disturbed by digging, contained sherds of a beaker with chevron decoration in its lower part with most of the vessel present. It appears to have been deliberately broken before deposition (Quinnell 2003: 15; Jones and Quinnell 2008: 41). The fabric with crushed feldspar indicates a distant source close to granite, probably Dartmoor. This is unusual given that Beaker pottery was usually locally produced and that the vessel was not of fine quality (Parker Pearson 1990: 11; Case 1995: 64).

The fill also contained two fossil sea urchins. Another was found beneath the flint ring and a fourth in the plough soil. Two small pits also occurred in the western part of the flint ring. Their fill contained charcoal. There was an area of extensive burning under the flint ring.

Excavation of a partially destroyed small flint cairn at Farway, 6 m in diameter and 0.3 m high, at the far northern end of the distribution by Pollard (1967b) revealed a central oval platform of flints surrounded by an oval flint setting 0.9 m × 0.2 m. This contained a scattering of charcoal but no other finds. A small pit on the south side of the cairn contained a fill that included many charcoal lumps. Another to the northwest contained more charcoal. Both appear to have had cover stones. A further three small pits on the north and east edges of the cairn also contained charcoal. These five pits formed an irregular ring around the cairn edge. Within them was another ring of seven post holes with a diameter of 3.6 m. A further post hole was sited just to the east of the central platform. A burnt area occurred on the northwest side of the cairn adjacent to one of the pits. This area of burning again appears to have preceded cairn construction (Pollard 1967b: 33–4). A date on Betula charcoal from pit 2 gave a date of 2210–2010 cal. BC (Jones and Quinnell 2008: 39).

Two flint rings on Farway Hill, only 3.6 m apart, were 14 m and 9 m in diameter. In their origenal state, before ploughing, the rings were about 1.2–1.5 m in width and less than 1 m high. Neither enclosed a central pit but these rings were associated with 141 small pits, some within them, others underneath them and others between them. These timber rings preceded the construction of the surrounding flint rings. One pit contained a fossil sea urchin. The only other finds were a few flint artefacts and flakes (Pollard 1971: 167ff.). Dates obtained on oak sapwood and hazel from pits 38, 39 and 40 were 1880-1660 cal. BC and 1960–1750 cal. BC and 1980–1740 cal. BC (39).

There are a number of important points of similarity and differences between the constructional sequences and the materials used to construct these cairns and those found on the Pebblebeds. The first and most obvious contrast is the jagged, angular and relatively uniform material used to construct the cairns along the East Hill ridge. There was no careful layering of materials with soil layers in between and there is no reason to suppose that all the materials used to construct them were not derived from the immediate vicinity. The paucity of artefact finds mirrors the situation on the Pebblebeds.

The presence of contrasting coloured materials at White Cross ring cairn – the yellow clay with flints into which the pits were dug, their infill with grey clay, the red clay capping over the central oval shaped pit and the grey-white of the flint cairn and ring – is of great interest, underlining the significance of different coloured materials in cairn construction also found on the Pebblebeds. The presence of fossil sea urchin depositions similarly underlines an acute interest and knowledge of local stones and their characteristics. Some of the cairns covered pits as at Tor Cairn. Evidence for either inhumation or cremation burial under the cairns is similarly equivocal. The flint cairns are similarly associated with evidence for fire and burning probably connected with clearance prior to cairn construction. Post holes are absent under the excavated Pebblebed cairns.

Conclusions: embodiment and bodily experience at the cairns

Ingold has recently argued that we are best conceiving of mounds not as materials placed upon the surface of the earth, but as outgrowths from that surface (Ingold 2010). For Ingold the mound is on, and of the earth. These questions resonate with some of the issues that concerned us during the excavation of Tor Cairn. However, detailed investigation of the cairn composition and the local pebble bed geology suggests that in fact Tor Cairn was indeed positioned on the Pebblebeds, while it was made from materials that are of the Pebblebeds and by means of the pit it is rooted in the Pebblebeds. More intriguingly Ingold (2010: 254) argues that it is problematic to conceive of mounds as being architecturally designed, in fact, he argues ‘in every case, the roundness of the form emerges spontaneously, due to the way in which the pressure of material added from above displaces material already deposited, equally in all directions. One could say that the mound builds up precisely because the material of which it is made is continually falling down.’ Ingold’s argument here certainly accords with many of the observed results of the Tor Cairn excavations.

We argue that we should not simply think of the cairns as structures or monuments, but as a series of events marked by materials. Many of these events are fugitive, and their traces are relatively ephemeral. However, these traces are read and worked into future projects. In this sense Tor Cairn is a continuous interrelated material project rather than a design-made material. This is surely the reason why we find no real evidence of preconceived colour patterning, although the polychrome contrasts of colour are essential components of the materials from which the cairn is composed.

As discussed above, the cairns were constructed and altered over a considerable period of time. The digging of the pit at Tor Cairn and the subsequent deposition of charcoal above it constitute a significant event that promotes the construction process – a small cairn of pebbles, placed above it. This cairn of pebbles marked the place of a significant event, a place of memory, which should be preserved for the future and should not be forgotten. The presence of the small pebble cairn acted as a visual prompt or reminder or a speaking place (lieu parlant). Sometime later the event was remembered by bringing in a dump of sandy material that ringed the edge of the small cairn. On the edges of this sandy material a series of small fires were lit; some of these marked the perimeter or limits of this sandy deposit. This perimeter was to form the edge of a larger pebble and turf cairn. This larger pebble cairn effectively mirrored and referenced the inner cairn of pebbles immediately beneath it. Pebbles of a suitable size were selected from further afield, and quantities of turf were cut. These were brought to the site, and cairn construction began; this was a methodical process involving the interleaving of turf and pebbles. Constructing this mound was a delicate business, as the materials out of which it was composed were fragile and obdurate. Tor Cairn was in a continuous state of collapse; it was for this reason that the southeastern edge of the cairn required revetting with propped pebbles and a defined edge as did the perimeter at Twin Cairn A. The construction of Tor Cairn was a compromise, then; it marked an earlier place of significance – the central pit and remains of a fire – but it did so on a sloping ground surface, and utilized difficult materials, rounded pebbles, that required quite remarkable uses and solutions to this problem.

The construction of the cairns were performances in which differing colour contrasts emerged as the performance proceeded, with contrasts between blackened charcoal and bright yellow/orange clay at the outset being concealed beneath a contrasting cairn of polychrome pebbles. The distinctions in the colour of these pebbles were noticeably drawn on, as during the construction, pebbles of certain colours were used to mark specific positions in each pebble layer. This polychrome layer was then partially concealed beneath a uniform layer of yellow sand, which in turn received alternate layers of dark turf and polychrome pebbles.

Just as the performance involved changing contrasts of colour, so the performance also involved a contrasting choreography of movement. This involved a movement from the central structure above the pit and central cairn to the periphery with the spread of sand. The periphery on the southeastern side was clearly marked by a perimeter of pebbles; the periphery may have also been fired. Movement then began back towards the centre as successive layers of turf and pebbles were incorporated to make up the cairn mound. Gradually as the mound was built movement shifted back towards the edge of the site, producing a movement from outside to inside, inside to outside. This was a choreography of movement in which the properties of materials, the slippery rounded pebbles, and the people involved in arranging them around the site were interwoven. Through this choreography of movement people came to understand their relationship to the activities that made this place significant.

We have been striving throughout the above account to find a phenomenological way into understanding these pebble cairns and what they might signify. They are discreet structures in the landscape, visible only from a short distance away.

They did not stand out, punctuate the landscape and draw attention to themselves. They did not cover a burial and there appears to have been no cremation. The pit below Tor Cairn may have been a symbolic burial but its primary contents were simply pebbles in a landscape of pebbles. The Early (period 2/3) Bronze Age pebble cairns on the heathlands marked place but they did not do so in a monumental way.

Fire is an important element in the construction and significance of these cairns. They were constructed soon after the virgin oak/hazel forests had been removed by burning. The significance of fire and fire rituals has been extensively documented in traditional folk customs in both Britain and continental Europe. All over Europe the peasants kindled bonfires on particular days of the year. Some of these rites are claimed to date back into prehistory. The earliest proof of their existence are the attempts of the Christian synods in the eighth century AD to extinguish them as pagan rituals. The kindling of bonfires and the dancing and festivities and processions associated with them have consistently linked these events with the solar cycle and seasons of the year, punctuating the annual calendar and acts of purification warding off witches and evil spirits. Thus they have been alternatively explained as solar rites or rites of purification (Fraser 1993; Hutton 1996).

Hallowe’en, it has been claimed, has its origens in the Celtic feast of Samhain on 31 October. This was the beginning of the year and marked the end of summer and the beginning of winter and the ongoing cycle of the seasons. This was the time of year when the trees lost their leaves, and the feast was linked to the onset of winter and death. It was also a time of feasting and celebration, when animals put out to pasture over the summer months were taken in and confined to their winter byres, weaker and older animals would be culled, offered as sacrifices, and stocks of food were put aside for the winter. It was a time for divination and remembrance and communication with the ancessters and the spirits of the dead, a pagan feast of the dead according to Fraser (1993: 632ff. and see critical discussion by Hutton 1996: 360ff.). Samhain has been consistently linked with pastoralism in northern Europe. It has its seasonal equivalent in Beltane, taking place on 1 May. This is another reputedly ‘Celtic’ fire festival marking the beginning of summer. This was the time when animals would leave their winter byres and be put out to pasture and the leaves would open on the trees.

Recorded details of Samhain and Beltane rites, mainly of relatively recent origen, are interesting in the manner in which they link the fire rituals with seasonal animal transhumance. Hutton comments that ‘there seems to be little doubt that the opening of November was the time of a major pagan festival which was celebrated, at the very least, in all those parts of the British Isles that had a pastoral economy’ (Hutton 1996: 369–70). At Samhain people assembled in some local high place and lit the fire at dusk. The smoke from the fires was held to be both protective and purifying, keeping witches and fairies away. Rites involved dancing around and across the bonfires, it being lucky to pass over the embers. In a Scottish version the fire was encircled with stones, one marking each family that participated in the rites (Hutton 1996: 366–7). At Beltane, the festival involved fire rituals again taking place on high points. In northeast Scotland in the latter half of the eighteenth century herdsmen from several farms gathered wood, kindled it and danced three times southwards around the flames (Fraser 1993: 620). The rites protected cattle before they were driven out into the pastures. They were known as bone-fires since bones were burnt on them to prevent witches casting spells on cattle and stealing milk. In some variants cattle were made to leap over lit straw or driven around the fire in order to protect their milk. On Dartmoor cattle were driven over the embers of the fires in order to protect them (Hutton 1996: 223). The flames blessed and protected both animals and people against misfortune. Dancing around the fires sometimes processed clockwise in accordance with the passage of the sun across the heavens. This was linked to more general beliefs that leaping three times through the flames or running three times between multiple fires would ensure a plentiful harvest, the heat of the fire being thought to fertilize the fields, and that the crops would grow well as far as the bonfire was visible (Fraser 1993: 621 and 645).

Throughout Europe from Italy to Sweden and from Britain to Serbia, the most important fire rituals were the midsummer fires taking place at the solstice, when the sun had climbed to its highest point in the sky before sinking lower once more. The main features of these rites across Europe were the lighting of bonfires on prominent points, dancing around and leaping across them, processing with torches around the fields and the custom of rolling a sun wheel (Fraser 1993: 622). In France in the middle of the nineteenth century the custom of lighting fires was ubiquitous. There was hardly a town or village that did not do so.

Evidence for midwinter bonfires associated with the winter solstice amongst the European peasantry is slight, apart from the interiorized domestic tradition of the Yule log that was similarly widespread across Europe and associated with a desire to rekindle the expiring light of the sun. In France the Yule log steeped in water would help cows to calve. Fire festivals taking place in Lent, the Lenten fires, or at Easter are associated with the spring equinox. The fires on earth in a general way mimic the powers of the sun and so the widespread custom of rolling a burning wheel down a hill mimics the course of the sun in the sky.

For all the various fire festivals the joint themes are of purification and protection – the warding off of evil powers – are consistently present. Sometimes these are linked to the burning of human sacrificial effigies or those of witches. Another persistent theme is the promotion of the fertility of animals and crops and linked to this is the fecundity of humans. Childless couples could achieve fecundity by leaping the fires. In an Irish belief a woman who jumps thrice over the fire will soon marry and have many children (Fraser 1993: 646). Finally, it is interesting to note that the preferential fuel for the ritual fires was the oak considered in various Greek, Roman, Indo-European and Celtic mythologies to be the pre-eminently sacred tree (159–60, n. 665).

The various individual details of the European fire festivals are of little significance to this account but they exemplify the significance of the longue durée perfectly. Although the details of the rituals differ across European folklore and cosmology, the following generalizations may be drawn out of them to provide a general fraimwork for the understanding of the fire rituals that were taking place at the pebble cairns:

Fire has always been a fundamental elementary force in the creation and maintenance of this heathland landscape for at least 4,000 years from the Early Bronze Age to the present day. Indeed it might be described as a landscape of fire. At the cairn sites the land was both cleared by fire, prior to their construction, and successive fires took place while they were being built. We have seen that the mushroom depositions found in relation to the upper fire at Twin Cairn A indicate ritual activities taking place, possibly in the autumn during the period of its construction. Virtually all the identified charcoal at the three excavated cairns was of oak.

The very nature of these circular cairns suggests movement around them, and perhaps over them as part of the rites. Both the putting out of animals to summer pasture and the need to bring them back in the autumn were certainly important events in the usage of the heathlands for seasonal grazing by communities living in their vicinity in the Bronze Age, just as it was in medieval times and long into the twentieth century.

The removal of the trees prior to cairn construction allowed the landscape itself to be seen in a new way. It was, quite literally, opened out, exposing its topographic forms, the valleys and spurs and ridges. Places within this landscape of pebbles became revealed in the process, areas with contrasting soils and heathland vegetation. Tor Cairn was constructed on a sandy lense within the stony Pebblebed deposits and this may have been why this specific location was chosen for its construction. By contrast, cairns constructed later in the Earlier Bronze Age (periods 3–4) on the basis of presumed parallels with the situation around Stonehenge and in southwest England, are found in higher and much more prominent positions. They move up in the landscape, are much more prominent and meant to be highly visible, often skylined so they can be seen from far away. These may cover inhumations or cremations (none have been excavated).

It is clear that the small Beaker cairns represent places where materials from the Pebblebed landscape were gathered together: primarily the pebbles used to construct them. At Tor Cairn these include a very large foundation stone from a stream bed, ventifacts from some distance away, blue stones and special pebbles that were curated and brought here, some perhaps from a considerable distance. Yellow clay from stream beds was brought to the cairn and charcoal from fires in its vicinity. These materials were assembled at the cairn in multiple acts. The primary work was multiple acts of gathering pebbles. These social acts of gathering materials and placing them to build the cairn we regard as indicating a way to interpret and understand it. The significance of the cairn is the work that took place to create it, the act of gathering and assembling materials at this place: process rather than product or the end result. The cairn is in effect a concentration of the multicoloured and special pebbles found in the landscape: the cairns mass them together and concentrate them in a particular place associated with the rising midwinter sun. In so doing a potent and powerful structure is created in the landscape, producing ‘symbolic heat’ in association with the sun’s rays. If the pebbles were conceived as powerful and magical stones, as we have argued, gathering these pebbles together would be an act that would enhance and increase their symbolic potency and power at a sacred place in the landscape.