The megalithic monuments of Neolithic and Chalcolithic Europe required considerable collective effort not only for their erection but also for the movement of their stones from source to monument. In the vast majority of cases, these stones were moved only short distances of less than 10 km or so (Boaventura et al. 2020). Even so, the weights of some megalithic stones such as the almost 150-tonne capstone of the Menga tomb at Antequera in southern Spain (Rodríguez et al. 2023) and the 330-tonne Grand Menhir Brisé in Brittany, France (Le Roux et al. 2007) mean that movement of these huge stones even over relatively short distances of a kilometre or more from their sources was still an extraordinary achievement.

Stonehenge is unique among megalithic stone circles for several reasons: the raising of lintels on top of uprights, the extensive stone-dressing of its surfaces, and – most significantly – the non-local origins of its stones. Yet it was not the first Neolithic monument with raw materials sourced at a distance. Beginning before the end of the fourth millennium cal bc, there is a trend of composite monumentality in which megalithic tombs were built of materials from different sources.

The great passage tombs of Newgrange and Knowth in Brú na Bóinne were constructed with stones brought from at least six source areas as far away as 40 km to the north and south along Ireland’s east coast (Cooney 2000, 136, Figure 5.2; Stout 2002, 30–1). The largest of these weigh around half a tonne – the greywacke blocks brought from up to 5 km away (see Figure 1).

The great passage tomb of La Hougue Bie, Jersey, incorporates different rocks from across the island (Bukach 2003). Apart from passage tombs, stone circles and other monuments at this time were also composed of different types of rock, although the distances travelled were substantially less. The two Orcadian stone circles, the Ring of Brodgar and Stones of Stenness incorporate monoliths derived from up to seven sources, covering distances of more than 13 km (Downes et al. 2013; Richards 2013a). Also, different stone circles among the complex on Machrie Moor, Arran, are constituted of different types of rock: red sandstone and white granite, both derived from different places on the island (Richards and Wright 2013, 50–9).

Stonehenge’s stones are grouped into three based on their lithology: the large uprights and lintels of sarsen (a silcrete or hard sandstone), the ‘bluestones’ of a variety of igneous and sedimentary rocks originating in the Preseli Hills (Mynydd Preseli) of west Wales and the sandstone Altar Stone. The sarsens form the bulk of Stonehenge, making up the circle of lintelled uprights, the inner horseshoe of five trilithons (pairs of uprights with a lintel) and a number of outer stones consisting of the Heel Stone, the Slaughter Stone and the Station Stones (see Figure 2). All of these may have been erected during Stonehenge’s Stage 2, beginning in 2740–2505 cal bc and ending in 2470–2300 cal bc.1 More precisely, the sarsen circle is dated to 2580–2475 cal bc (Parker Pearson et al. 2022b, 78). A few sarsens may have been erected in Stonehenge’s first phase, beginning in 3080–2950 cal bc and ending in 2865–2755 cal bc (Parker Pearson et al. 2020, 301).

It is estimated that there may originally have been as many as 83 sarsens at Stonehenge, although there is uncertainty about whether the sarsen circle was ever completed (Bowden et al. 2015, 43–5; Parker Pearson et al. 2022b, 84).

The bluestones are composed of a variety of rocks of dolerite, rhyolite, andesite, dacite and Lower Palaeozoic sandstone, all of which are sourced to the area of the Fishguard Volcanic Group and its environs (Ixer and Bevins 2017). At Stonehenge these 43 stones form the outer Bluestone Circle and inner Bluestone Horseshoe, having been erected in a series of formations that were rearranged during four stages of construction during c. 3000–2000 bc. There may originally have been around 80 of them installed at Stonehenge, with 56 initially occupying the circle of Aubrey Holes close to the monument’s perimeter in Stage 1, together with another 25 or so in a small circle (Bluestonehenge), 2 km away beside the River Avon (Parker Pearson et al. 2020, 224–300).

The third type of stone at Stonehenge is the Altar Stone, a 6-tonne monolith of Old Red Sandstone from the Orcadian Basin (Clarke et al. 2024). This lies flat within the south-western part of Stonehenge’s centre, pinned in place by the fallen sarsen upright of the Great Trilithon and by its fallen sarsen lintel.

The distant sources of Stonehenge’s stones have been inferred for centuries. In the seventeenth century, John Aubrey reckoned that he had found a quarry pit for Stonehenge’s sarsen stones 23 km to the north (Scurr 2015, 105). Aubrey also suggested that the sarsens were sourced from the Marlborough Downs slightly further away, around 32 km to the north. Geochemical analysis has recently revealed that 50 of the 52 surviving sarsen uprights and lintels at Stonehenge can be matched with a source in West Woods, 24 km north of Stonehenge (Nash et al. 2020; 2021). Although a re-analysis of the data obtained by Nash’s team has suggested that West Woods was just one of several possible sources (Hancock et al. 2024), the original identification of West Woods as the likely single source is upheld (Nash and Ciborowski in press).

The source of the two as-yet unprovenanced sarsens (Stones 26 and 160) is not yet established but analysis of stone debris from Stonehenge reveals that this debris, thought to derive from destroyed sarsen blocks, can be matched with sources around Bramdean and Brighton, 55 km and 200 km to the south-east, as well as with the West Woods area (Ciborowski et al. 2024). This raises the possibility that these fragments derive from monoliths erected earlier than those of Stage 2, not just taken down after Stage 1 but broken up.

The bluestones have been geologically identified as coming from the area of the Preseli Hills ever since Herbert Thomas (1923) identified the most numerous of the Stonehenge bluestones as spotted dolerites that could be matched with outcrops in those hills. Although occasional attempts have been made to explain the bluestones’ incorporation in Stonehenge as due to transport by glaciers in a previous Ice Age (John 2024; Judd 1902; Kellaway 1971), there is no evidence that glaciers extended more closely than within 100 km of Salisbury Plain (Clark et al. 2022; also discussed in some detail in Ixer et al. in press). Claims that one or more bluestone fragments from Stonehenge and its environs show evidence of having been transported by glaciers similarly do not stand scrutiny (Bevins et al. 2023a).

Four types of bluestones have been matched geologically with outcrops in the Preseli. The source for most of Stonehenge’s spotted dolerites (classed as Group 1) has been identified as Carn Goedog (Bevins et al. 2014). Two sources for unspotted dolerites (Stones 45 and 62; Group 2) are Cerrigmarchogion and Garn Ddu Fach, to the west and east of Carn Goedog (Bevins et al. 2014; 2021; Pearce et al. 2022). Remaining spotted dolerites (Group 3) are thought to derive from an area to the east of Carn Goedog but are not matched to a specific outcrop (Bevins et al. 2014). Of the three types of rhyolite at Stonehenge, Group C is matched to a specific location within the outcrop of Craig Rhos-y-felin, 3 km to the north of the Preseli ridge (Ixer and Bevins 2011). Finally, Stonehenge’s two Lower Palaeozoic sandstone monoliths are similar lithologically and, in terms of age, to strata exposed to the north and east of the Preseli Hills (Ixer et al. 2017).

Excavations at the bluestone sources of Carn Goedog and Craig Rhos-y-felin have uncovered evidence of megalith quarrying dating to the centuries before and around 3000 bc, consistent with the date of Stonehenge’s first stage. At Craig Rhos-y-felin, that precise part of the outcrop with a match for Rhyolite Group C lies directly adjacent to a niche from which a 2.5-m-long monolith has been removed (Parker Pearson et al. 2015). Quarrying installations include a drystone-revetted, artificial platform at the foot of the outcrop as well as a hollow way or sunken trackway leading from the foot of the platform (Parker Pearson et al. 2019). Quarrying artefacts include three stone wedges still in situ within joints close to the gap left by a removed monolith (Parker Pearson et al. 2022a). Similar evidence of quarrying was found at Carn Goedog, in the form of stone wedges and other stone tools, an artificial platform, niches left by removed pillars and wedge-holes cut into the joints between pillars (Parker Pearson et al. 2019).

The Altar Stone (Stone 80) is 4.9 m (16ft) long by 1 m (3½ ft) wide by 0.5 m (1¾ ft) thick (Atkinson 1956, 45). Just when the Altar Stone arrived at Stonehenge is unknown. It could have stood during Stage 1 within one of the larger now-empty stoneholes (Parker Pearson et al. 2022b, 112). But now that it can be geographically uncoupled from the bluestones, the idea that the Altar Stone might have been present in Stage 1 has less attraction. Cleal et al. (1995, 188, Figure 86) postulated that it might have stood vertically during Stage 2 (according to the revised chronology) in a large hole (WA 3639) in the south-west sector of the bluestone arc known as the Q & R Holes (Figure 2; Darvill et al. 2012).

Unfortunately, the Altar Stone’s only identified stratigraphic relationship is with the collapsed and broken sarsen upright (Stone 55, its two broken halves numbered as 55a and 55b) and lintel (Stone 156) of the Great Trilithon which were erected in Stage 2 and now lie on top of it. Just when they fell is unknown, although this must have occurred before the sixteenth century ad and after the Bluestone Horseshoe was erected in Stage 4 (starting in 2210–2030 bc and ending in 2155–1920 bc).

The top of the Altar Stone now lies level with the grass around it. Numerous excavations, long since filled in, have revealed parts of it at different times. In 1620 the Duke of Buckingham’s men dug a large hole at the centre of Stonehenge that may well have exposed part of it. William Stukeley (1740, 30–1) described it as ‘squar’d’, in other words dressed. William Cunnington dug in front (north-east) of it in 1802, and in 1844 a Mr Brown of Amesbury collected a fragment (Wilts #277; Bevins et al. 2023c) from the underside of the Altar Stone. Finally, Richard Atkinson’s (1979, 211–12) 1958 excavation dug around it, revealing that its north-west end had been ‘squared off at right angles to the stone’ while its south-west end is ‘dressed to an oblique bevelled outline’, comparable to the bases of some of the sarsen uprights (Figure 3).

Atkinson (1956, 45) considered that the Altar Stone’s bevelled end was evidence that it had once stood vertically at Stonehenge. He could not identify a stonehole for it but thought that it might have stood in a hole that William Cunnington found ‘close to the altar’. Underneath the upper half of the fallen upright (Stone 55b) of the Great Trilithon, Atkinson (1979, 212) found what he thought was a large stonehole (WA 3359; see also Cleal et al. 1995, plan 2). Realising that it was not deep enough for the Altar Stone, he tentatively suggested that a stone stood here might have formed a pair with the Altar Stone, potentially forming a midsummer sunrise–midwinter sunset solstice alignment in front of the Great Trilithon (Atkinson 1979, 212).

Putting aside these hypothetical suppositions, the current location of the Altar Stone provides some evidence to work with, although there has been uncertainty about whether it was placed in this location deliberately as a recumbent stone or has fallen from an upright position. It lies neatly in front of the Great Trilithon with its long sides aligned approximately on the midwinter sunrise. As Aubrey Burl (2006, 207–10) noted in his very thorough discussion of the Altar Stone and its position, the probability of this occurring by chance if it were a fallen upright is as low as 1:165 (see also Daw 2015). Thus, the most likely scenario is that the Altar Stone was set here deliberately as a recumbent stone. There is no means of knowing when this might have been done; its location in front of the Great Trilithon would imply that it does not pre-date that structure’s erection in Stage 2 but it could date to any point within Stages 2–5.

On the basis of age dating of detrital zircon, apatite and rutile in the Altar Stone, Clarke et al. (2024) have presented compelling evidence for a derivation of the megalith from the Old Red Sandstone (ORS) Orcadian Basin, in north-east Scotland (see Figure 4). This follows extensive investigations of a possible source in south Wales and the Welsh Borderlands, based on an original west Wales source proposed a century ago by Thomas (1923). This proposal of an origin in Wales was endorsed by K. C. Dunham (former Director of the now British Geological Survey) in a letter to Dr J. F. S. Stone, dated 17 October 1947, in which he stated that rock sample #275, a second type of sandstone found in the Stonehenge landscape, from the Stonehenge Cursus and subsequently described by Stone and Wallis (1951), was from Mill Bay, on the Milford Haven estuary, 0.75 km from Cosheston. This led to the notion of Milford Haven being the exit port for marine transport of the bluestones, as proposed by Stone (1948) and later in popular works by Hawkins (1970) and Atkinson (1979). Ixer and Turner (2006) examined various sandstone samples from Stonehenge and from west Wales and showed the Altar Stone and fragment #275 (now known to be a Lower Palaeozoic Sandstone) to be very different, so concluding that Milford Haven was not the source of the Altar Stone, and that it was probably derived from the Senni Beds elsewhere in south Wales.

More recent investigations (Bevins et al. 2020; 2022; 2023b; 2023c; Ixer et al. 2019; 2020), based on further petrographical studies, combined with analytical investigations using X-ray diffraction (XRD), automated mineralogy (SEM-EDS), portable X-ray fluorescence (pXRF) and Raman spectroscopy, have led to mineralogical and geochemical characterisation of the Altar Stone. Critical in these studies has been the non-invasive in-situ analysis of the Altar Stone itself, comparing results with those obtained from a number of debitage samples. Fragments confirmed to be derived from the Altar Stone have then served as proxies for further, more invasive forms of analysis, leading to a greater understanding of the mineralogy and geochemistry of the Altar Stone itself.

The Altar Stone is distinctive in containing the minerals baryte and tosudite (a rare interstratified dioctahedral chlorite/dioctahedral smectite) with a near-total absence of potassium feldspar, while the geochemical analyses of the Altar Stone show high barium (Ba) concentrations, linked to the baryte cement. These investigations led Bevins et al. (2023b) to conclude that the Altar Stone was not, in fact, sourced from south Wales, or indeed from southern Britain, concluding that it was time to ‘broaden horizons’ and to consider northern England and Scotland as possible source regions. Bevins et al. (2023b) also realised that, given the notable differences between the Altar Stone and the bluestones, especially in terms of size, weight and position in the monument, and also considering the fact that it is not known when the stone arrived at Stonehenge and whether it was ever erected or was always a recumbent stone, it was time to ‘de-classify’ the Altar Stone as a bluestone. Including it in the bluestone assemblage had, for a century, strongly restricted thinking about its source.

Following the recognition that the Altar Stone was not sourced from the Anglo-Welsh Basin, a series of criteria were suggested by Bevins et al. (2023b), which potential Altar Stone source sites should meet. These included a terrestrial, ‘red-bed’ origin for the sediments (encompassing Permian and Silurian/Devonian ‘Old Red Sandstone’ age deposits), a high Ba concentration (related to the presence of baryte as a cement), as well as the presence in the region of evidence for significant Neolithic settlement/activity. Bevins et al. (2023b) suggested several regions where such criteria were met, with the most obvious being Mainland, Orkney, with its ORS geology, areas with high Ba (British Geological Survey 2024; Everett et al. 2019) and abundant Neolithic remains, as well as the numerous ‘cultural’ similarities between Orkney and Stonehenge detailed elsewhere in this paper. Even the sizes of the monoliths making up the Ring of Brodgar and the Stones of Stenness are similar to the Altar Stone.

Portable XRF investigations on Orkney were undertaken by Bevins and Pearce in June 2023, analysing various monoliths comprising the Neolithic stone circles of the Ring of Brodgar and the Stones of Stenness, as well as Old Red Sandstone field samples from across Mainland, Orkney. The pXRF results, along with petrological microscopy, XRD, SEM-EDS and heavy mineral (HM) characterisation by Raman spectroscopy on ORS field samples were presented in Bevins et al. (2024), who concluded that there are no realistic comparisons between the geochemistry and mineralogy of the Altar Stone and the analysed Orkney stones and field samples. In particular, there are clear differences between the compositions of the two sample sets in the binary plot of Ba versus Sr and in the triangular plot of Ba-Rb-Sr (see Figures 5 and 6). In addition, distinctive mineralogical and geochemical features of the Altar Stone, described above, are not seen in the Orkney field samples. Although baryte was detected in two Orkney samples, it was linked to base metal mineralisation, which is not the case for baryte in the Altar Stone. Petrography, XRD and HM studies also show marked differences between the two sample sets.

These data lead to the conclusion that the Altar Stone was not sourced from Mainland, Orkney, and that attention needs to be directed to other areas of the ORS Orcadian Basin in mainland northeast Scotland, or, as an extreme, the Shetland Islands (Bevins et al. 2024). Field investigations in these areas are currently underway.

Despite the considerable distance between the north of Scotland and southern England, certain styles of material culture were shared between these regions and even more widely throughout Britain and parts of Ireland. The most striking of these is Grooved Ware, a style of decorated pottery with earliest start dates in Orkney at 3160–3045 cal bc (Copper et al. 2021, 82–3) and in use in the Stonehenge landscape by the thirtieth century cal bc (Wessex Archaeology 2020, 102).

The ‘circular world’ of Late Neolithic Britain, expressed in its timber and stone architecture (Bradley 1998), embraced common architectural principles shared throughout Britain (Bradley 2019; Madgwick et al. 2021; Pollard 2009; Ray and Thomas 2018, 245–6, Figure 5.10). These need not imply some form of homogenised conformity throughout Britain but the numerous similarities between far-flung regions signal a level of shared traditions of practice (Bradley 2024; Cummings 2017, 232; Greaney et al. 2020; Harding 2003, 56–8).

Orkney and northern Scotland shared a number of cultural styles and practices with other, more distant parts of Britain (Parker Pearson 2020). The classic henge (with the bank outside the ditch instead of inside it), found throughout Britain, has its earliest dated example in Orkney at the Stones of Stenness with a modelled date of 3020–2890 cal bc (Schulting et al. 2010, 35–6). Styles of rock art in Orkney are also closely paralleled at Morwick on the Northumbrian coast (Bradley 2024). Even so, there were distinct cultural differences between the Scottish mainland and Orkney. The northern Scottish mainland may have shared in Late Neolithic Britain’s dominant funerary rite of cremation (Willis 2021) and differed from the collective inhumation rites of Orkney.

Transport of the Altar Stone can only have been achieved in two ways – by land or by water (or a combination of both) – both involving human agency. Clarke et al. (2024) have discussed why theories involving Ice Age glacial transport can be rejected. While ice sheets covered northern Britain in previous glacial periods, glaciers did not move material southwards to Salisbury Plain from northern Scotland; rather, glaciers moved northwards from the Grampian Mountains (Clark et al. 2022; Gibbard and Clark 2011).

A land route in the later third millennium bc would undoubtedly have encountered difficult terrain and dense forests, rather different to the largely deforested conditions on Salisbury Plain for moving Stonehenge’s sarsens (French et al. 2012). Yet it might be premature to write off overland transport of the Altar Stone. From ethnographic evidence from living traditions of megalith-building in Madagascar, Indonesia and elsewhere, people prefer to minimise risks, including choosing to avoid moving monoliths by water (Richards 2013b, 7–13). For Stonehenge’s bluestones, for example, there is a possible land-based route as well as Atkinson’s hypothesised sea route, and even any movement by water along the Bristol Channel and then the Bristol Avon would still have required long stretches of movement overland (Parker Pearson 2012, 279; Pitts 2022, 111–23).

In weighing up the pros and cons of either possibility for the Altar Stone, we should remember that long-distance land routes had been in use from the early fourth millennium bc onwards, taken by people exchanging axe heads and other materials over hundreds of miles (Bradley and Edmonds 1993; Schauer et al. 2020). The relative fragility of the Altar Stone would make it susceptible to cracking on such a journey, necessitating the use of a suitably robust wooden sledge with vegetation-made shock absorbers, sliding on wooden ‘rails’ that could be continuously lifted and re-laid.

From the modern-day events of walking from John O’Groats to Land’s End, we might calculate that such a journey with a large and heavy sledge could take perhaps four times today’s time of under two months for such a distance at walking speeds of about 5 km per hour, needing eight months or so to haul a stone from northern Scotland to Stonehenge.

The sea route, at first glance, might seem more attractive and more easily supportable. Although no boats are known from the Late Neolithic in Britain or Europe, Britain and Ireland’s Neolithic farmers and their ancestors had long experience of open-sea travel. Their genetic ancestry can be traced back to early farmers who migrated from Anatolia through the Mediterranean and the Aegean (Brace et al. 2019). By 4000 bc they were capable of moving cattle and other livestock by boat not only across the English Channel but also across the Celtic Sea to southern Ireland (Kinnes 1988; Sheridan 2010).

More locally to northern Britain, the Orkney vole (Microtus arvalis) was probably introduced to Orkney by seafarers in 3455–3100 cal bc from somewhere in continental Europe, possibly northern France or Belgium (Bayliss et al. 2017, 1178; Martínková et al. 2013; but see Sheridan and Pétrequin 2014 for a critique). Similarities in Late Neolithic rock-art styles along the North Sea coast from Orkney to Morwick also hint at long-distance connections along that coastline (Bradley 2024).

The earliest known boats in Britain have been found further south along the North Sea coast, in the Humber estuary at Ferriby. The earliest of these Early Bronze Age sewn-plank boats (Ferriby 3) dates to 2030–1780 cal bc (Wright et al. 2001), broadly contemporary with Stonehenge Stage 5. Shallow draft was essential in these paddled boats to ensure less drag and less weight (Roberts 2004, 39). They may have had rocker (arching of their bottom profile), making them more easily manoeuvrable on the sea and increasing their seaworthiness (Roberts 2004, 45).

Even so, the size of a sewn-plank boat sturdy enough to take the 6-tonne weight of a monolith would have to be far greater than the estimated 15 m–16 m length of the Ferriby boats. That may not have been possible with sewn-plank boats because their length would have been limited by the weakness of the withy stitching to withstand flexing from hogging and sagging (when the hull bends upwards and downwards) (Roberts 2004, 45). Other alternatives in the form of logboats or hide-covered curraghs would seem even less feasible, as demonstrated in 2001 by a failed attempt to transport a modern-day bluestone – substantially smaller than the Altar Stone – by curragh (Roberts 2004, 41).

What we know of Early Bronze Age boats, limited though our knowledge is, would suggest that they and their Late Neolithic precursors were unlikely to be sturdy enough to carry a 6-tonne stone in coastal waters. As has been proposed for the transportation of Stonehenge’s bluestones, a land route may have been favoured (Parker Pearson 2012, 289–91). Travel by land would have provided much better opportunities for spectacle, pageantry, feasting and celebration that would have drawn people in their thousands to witness and take part in this extraordinary venture (Parker Pearson 2012; Richards 2013b, 7–8). It would also have reaffirmed networks of relationships between distant groups, encouraging collaboration and cooperation on a hitherto unseen scale (Richards 2013b, 7–8; Wunderlich 2019).

During our research into sourcing the Stonehenge bluestones, one thing that became evident as the work progressed was the extraordinary appearance of the quarries (Parker Pearson et al. 2015; 2022a). At both Craig Rhos-y-felin and Carn Goedog, the outcrops are striking and impressive, with their naturally upright pillars making these places anomalous and remarkable when encountered (Parker Pearson 2019). As yet, the actual outcrop from which the Altar Stone was extracted is unknown, but we might speculate that it similarly derived from a dramatic location. Alternatively, it may have been ripped from its original context by glacial action and have been come upon in another part of northern Scotland, standing out as an extraordinarily shaped monolith of a rock type alien to its surroundings. Whether the Altar Stone was then set up as a standing or recumbent stone near its source, later to be dismantled and taken to Stonehenge, can only be guessed at.

History and myth provide insights into motivation for taking megalithic stones over long distances to Stonehenge. In Geoffrey of Monmouth’s pseudo-historical History of the Kings of Britain, the wizard Merlin oversees the taking by force of the Giants’ Dance, a stone circle in Ireland, to rebuild it as Stonehenge to commemorate a massacre of Britons by Saxons (Parker Pearson 2022; 2023, 117–18; Wright 2007). The Giants’ Dance is all the more significant for Merlin because he declares that it has healing properties.

In a historical case, in 1296 Edward I of England captured the Stone of Destiny, the 152-kg Old Red Sandstone block on which the kings of Scotland were crowned, and had it brought from Scone Abbey to Westminster Abbey where it was used to crown English monarchs for the next 500 years (Breeze and Munroe 1997). In both this and the Merlin myth, the motive for transportation is the stone’s perceived spiritual properties of bestowing magical or spiritual powers which legitimate authority – whether as a memorial to fallen nobles or as a symbol of sacral kingship. This might be termed cosmological acquisition, the imbuing of Stonehenge with spiritual power deriving in part or in whole from the exotic and distant sources of its stones. One thing is for sure; just as other monuments across Britain and Ireland become material and social microcosms at a more localised level, Stonehenge stands out in being a material and monumental microcosm of the entirety of the British Isles.

An alternative explanation is that the Altar Stone was brought as a contribution or gift by the Neolithic people of northern Scotland, perhaps to cement an alliance or to take part in the extraordinary long-distance collaboration that building Stonehenge represented and embodied. Gordon Childe (1957, 331) wrote of the Preseli bluestones’ journey to Stonehenge as illustrating a political unification or a sacred peace. Of course, such unification could have been achieved or at least attempted by force or by consent; either as acquisition through violence of a formerly inalienable symbol or as the giving of a gift within a relationship of reciprocal equality.2 The latter seems more likely for the Altar Stone, given that its position at Stonehenge seems designed to celebrate the Scottish recumbent circle tradition from which it may have derived.

Why did certain large stones have such significance in Neolithic society? A long-running hypothesis guiding research into Stonehenge is that megaliths can have ancestral significance, representing and even embodying the ancestors that they commemorate, whether as funerary markers, tombs or standing stones (Parker Pearson 2023, 155–7; Parker Pearson and Ramilisonina 1998). Such symbolic constructions serve to link people with the land – as well as with the heavens – to tie distant places and people into a single cosmological vision of ancestral unity and authority (Parker Pearson 2012, 328–31; 2023, 158–60).

Such a strategy was not invented by the builders of Stonehenge; as discussed at the beginning of this article, it was already mobilised and embodied in the great passage tombs of Brú na Bóinne. On a more local landscape scale, this physical and metaphorical representing in microcosm of the surrounding world by Neolithic monuments has long been recognised (Richards 1996; Tilley 1996, 204–11). The builders of Stonehenge took it to even greater lengths to encompass virtually the entire island of Britain.

The arrival of the Altar Stone at Stonehenge is undated but most likely occurred within Stages 2–4, broadly c. 2500–2020 bc. This was a period of substantial population replacement following the arrival from continental Europe of Beaker-using communities with steppe ancestry (Olalde et al. 2018). The Altar Stone’s incorporation into Stonehenge as an attempt at unity may have been a response to a legitimation crisis brought on by this influx of new people. Yet, by the end of Stonehenge’s five-stage construction sequence (1620–1425 cal bc), Britain’s insular Neolithic population appears to have been largely replaced by people with steppe ancestry (Booth et al. 2021). As an attempt at unification, Stonehenge was perhaps ultimately a failure.

Since Childe’s day we have learnt so much more about the many distant sources of Stonehenge’s stones. Unlike any other stone circle, it is built entirely of stones that are not from its immediate environs. The Altar Stone’s origin in Scotland is confirmation of this observation and its distant source further strengthens the unification hypothesis. More detailed geological investigation of Old Red Sandstone sources in northern Scotland may provide a better understanding of the social and cultural context in which the Altar Stone was extracted and initially displayed.

Sourcing the Altar Stone to the northern Scottish mainland is the latest achievement of a remarkable series of geological projects that have been carried out since the early 2010s. It provides dramatic evidence of Stonehenge’s uniqueness among prehistoric stone circles: it was constructed from materials brought from distances of between 24 km and at least 700 km, well in excess of the average distance of up to 7 km noted for megaliths elsewhere in Britain and Europe. Although the date of the Altar Stone’s arrival and of its placing in its current position are unknown within the Late Neolithic–Early Bronze Age constructional sequence at Stonehenge, its solstice-aligned axis and strategic positioning in front of the Great Trilithon argue for its deliberate placement as a recumbent stone rather than as an orthostat that has accidentally fallen.

The Altar Stone’s northern Scottish origin helps to make better sense of hitherto enigmatic observations. First, its location as a recumbent stone within the south-west arc of Stonehenge resonates with the architecture of recumbent stone circles of north-east Scotland whose distribution extends westwards to the Old Red Sandstone beds of the Orcadian Basin. Second, the unusually strong architectural similarities in ground plans of Late Neolithic houses at Skara Brae and Stonehall Farm in Orkney and at Durrington Walls near Stonehenge can now be considered in relation to this new indication of connectivity between these distant regions, providing a possible time for the Altar Stone’s arrival at Stonehenge in or after Stage 2, c. 2500 bc.

Finally, the distant origin of the Altar Stone confirms Stonehenge’s unique status as the one stone circle built entirely from non-local stone; a material microcosm projecting at an enormous scale. It is consistent with recent interpretations of Stonehenge as a monument whose builders attempted – ultimately unsuccessfully – to establish some form of political unification and shared identity across much or even all of Britain, bringing together these extraordinary and alien rocks which symbolised and embodied far and distant communities within a complex material and monumental expression of unity between people, land, ancestors and the heavens.