Light at the end of the tunnel > the way megalithic art was viewed and experienced

...he came to a mouth of a passage covered with a square stone similar to that at [nearby] Plasnewydd, anxious to reap the fruits of his discovery he procured a light and crept forward on his hands and knees along the dreary vault, when lo! In a chamber at the further end a figure in white seemed to forbid his approach. The poor man had scarcely power sufficient to crawl backwards out of this den of spirits... (Reverend John Skinner 1802) There is a limited but significant passage-grave art tradition in England and Wales which, although restricted to two passage grave monuments in Anglesey, Bryn Celli Ddu and Barclodiad y Gawres and a destroyed megalithic structure located in a park in Liverpool, the Calderstones, marks the eastern extent of megalithic art in Britain and Ireland. The passage grave tradition is also one of the last megalithic architectural styles of the Neolithic. Outside these two areas are a number of sites that possess simple ABSTRACT – This paper explores how megalithic art may have been viewed during a period when Neolithic monuments were in use as repositories for the dead. The group of monuments discussed are primarily passage graves which were being constructed within many of the core areas of Neolithic Atlantic Europe. Although dates for the construction of this tradition are sometimes early, the majority of monuments with megalithic art fall essentially within the Middle to Late Neolithic. The art, usually in the form of pecked abstract designs appears to be strategically placed within the inner part of the passage and the chamber. Given its position was this art restricted to an elite and was there a conscious decision to hide some art and make it exclusively for the dead? In order to discuss these points further, this chapter will study in depth the location and subjectivity of art that has been carved and pecked on three passage graves in Anglesey and NW England. I suggest that an encoded grammar was in operation when these and other passage grave monuments with megalithic art were in use.


Introduction
...he came to a mouth of a passage covered with a square stone similar to that at [nearby] Plasnewydd, anxious to reap the fruits of his discovery he procured a light and crept forward on his hands and knees along the dreary vault, when lo! In a chamber at the further end a figure in white seemed to forbid his approach. The poor man had scarcely power sufficient to crawl backwards out of this den of spirits... (Reverend John Skinner 1802) There is a limited but significant passage-grave art tradition in England and Wales which, although restricted to two passage grave monuments in Anglesey, Bryn Celli Ddu and Barclodiad y Gawres and a destroyed megalithic structure located in a park in Liverpool, the Calderstones, marks the eastern extent of megalithic art in Britain and Ireland. The passage grave tradition is also one of the last megalithic architectural styles of the Neolithic. Outside these two areas are a number of sites that possess simple ABSTRACT -This paper explores how megalithic art may have been viewed during a period when Neolithic monuments were in use as repositories for the dead. The group of monuments discussed are primarily passage graves which were being constructed within many of the core areas of Neolithic Atlantic Europe. Although dates for the construction of this tradition are sometimes early, the majority of monuments with megalithic art fall essentially within the Middle to Late Neolithic. The art, usually in the form of pecked abstract designs appears to be strategically placed within the inner part of the passage and the chamber. Given its position was this art restricted to an elite and was there a conscious decision to hide some art and make it exclusively for the dead? In order to discuss these points further, this chapter will study in depth the location and subjectivity of art that has been carved and pecked on three passage graves in Anglesey and NW England. I suggest that an encoded grammar was in operation when these and other passage grave monuments with megalithic art were in use.
decoration, usually restricted to single and multiple cupmarks (Darvill & Wainwright 2003;Sharkey 2004;Nash et al. 2006; see Tab. 2). However, these monuments are not classified as passage graves. 1 Outside England and Wales; in Ireland and northwestern France, megalithic art and passage grave architecture is both numerous and complex in form and many of the traits incorporated into these monuments are also replicated in the British examples (Forde-Johnstone 1956;Lynch 1967;1970;Shee-Twohig 1981).
These three passage graves each possess complex carved art that is usually located within the inner section of the passage or forms part of the chamber alignment . Both Anglesey monuments have the passage and chamber architecture incorporated into large covering mounds. Despite the intense and comprehensive archaeological investigations of these monuments and the near complete destruction of the Calderstones monument, one can make an assessment of how these monuments may have functioned during ritual-symbolic events. Recent investigations by the author have shown how light may have played an important role especially in illumination of various parts of the chamber and passage architecture; more significantly, how the rock-art was viewed and experienced.
This chapter will assess the impact of light on the chamber, façade and passage areas, focusing on how and why the rock-art was illuminated. I suggest that megalithic art was deliberately placed in such a way as to visually change the ambience of the space between various sections of the passage and chamber, in particular at the point where natural light in the passage area fades and was no doubt replaced by the illumination of fire. Based on two types of light and their intensity, two cognitive emotions are at play. One based on the familiarity and safety of the façade and the outer passage areas, the other based on the ignorance and anticipation of the individual entering the inner passage and chamber which were initially dark, foreboding and unknown.

The data set
The distribution of the passage grave tradition is within the Atlantic zone of Europe, occupying five major core areas, from Iberia to southern Scandi-

Tab. 1. Radiocarbon date ranges for passage grave construction along the Atlantic Seaboard. 2
1 Frances Lynch however does describe several of these as being 'short passage graves (1969a & 1970)'. However, I am inclined to suggest that these monuments have little architectural or chronological associations with the passage grave tradition per se of the Late Neolithic. 2 There are no radiocarbon dates for the Welsh passage-graves. navia. Its development can be clearly traced and it probably terminates in Wales and north-west England by about the mid-3 rd Millennium BC (Shee-Twohig 1981). The majority of the passage graves, in particular those monuments found in southern Scandinavia, are not decorated with megalithic art and are considered earlier than those that are, dating between 3200 and 3500 BC (Tilley 1991.77) and probably represent an initial wave of immigrant farming or the translocation of ideas (or both). Megalithic art can therefore be considered a secondary tradition that occurs only in certain passage grave areas such as Iberia, Brittany, Ireland, Wales and Scotland. The latter two areas are restricted to no more than 12 sites, suggesting limited and late contact with areas fully immersed in the megalithic art tradition. 3 Discussing the distribution of megalithic rock-art in Britain and Ireland is not new. Significant work has been undertaken in Ireland by a number of eminent archaeologists such as Michael O'Kelly (1982;1993), George Eogan (1986), Muris O'Sullivan (1986;1993), Elizabeth Shee-Twohig (1981) and a comprehensive summary by Gabrial Cooney (2000). The passage graves of Newgrange, Knowth and Dowth, located within the Boyne Valley, and Fourknocks and Loughcrew (Sliabh na Callighe), in County Meath all possess significant megalithic rock-art (Shee-Twohig 1981;Cooney 2000). Some designs arguably originate from the south, along the Atlantic seaboard, within the core areas of Brittany and, according to Eogan, from the Iberian Peninsula (1986.172). Based on a limited but significant radiocarbon dating range, one can trace the development of the passage grave tradition. It is generally considered that the movement is from south to north with its zenith in central and southern coastal Ireland during the latter part of the 4 th millennium BC (Tab. 1).
The demise of the passage grave tradition, in particular those monuments that possess rock-art, appears to occur in North Wales and north-west England, although up to seven sites with megalithic art exist in Scotland and Orkney (Piggot 1954;Shee-Twohig 1981). Simple passage graves located on the Isles of Scilly and referred to as entrance tombs date to the late Neolithic but have no megalithic art. It is probable that the same architectural influences were mo-ving northwards but some communities may have considered megalithic art time consuming and developed new ways of expressing burial rites and artistic endevour.
Recently, in 2005, part of an upright from a nearby destroyed passage grave has been discovered in a Bronze Age barrow at Balblair, Inverness, forming one of the cist walls (Dutton & Clapperton 2005). This carving, present on a former upright, shows a new and altogether unique curvilinear design that is not found on any other megalithic stone in the British Isles. The stone, forming the wall of a Bronze Age cist probably originated from a nearby destroyed passage grave. 4 The stone decoration comprises two large gouges, possibly cupmarks with each cup measuring around 0.20m in diameter and up to three finely carved lines radiating away from one of the cups. One is perforated and may have possessed a functional rather than a decorative role. The second and more intriguing design field is located on the upper section of the slab comprising of a series deeply gouged semi-ovate lines which resemble (but not necessarily is) the trunk and branches of a tree. However, more probable and incorporating the two large gouged cups from the lower section of the slab, the two design fields may represent a stylised penis (with accompanying testes). 5 The upper design, measuring approximately 0.70m x 0.70m is symmetrically placed with the outer gouges running parallel with a possible deliberately cut and shaped recess. Despite the uniqueness of this art, however, elsewhere in Scotland the designs show evidence for a regional style developed from probable contact and exchange with monument builders in Ireland. Spirals and concentric circles at Pickaquoy and Eday Manse, both in Orkney, have similar decorative styles with monuments in central Ireland (Shee-Twohig 1981.136-7).
The people who constructed the monuments in Anglesey and on Merseyside would have been involved in varying degrees of socio-political contact and exchange with passage grave builders in central eastern Ireland, operating in what Herity and others refer to as the Irish Sea Zone (1970.30-3). This zone is identified mainly through the stylistic similarities in monument building and the deposition of material cul-ture within each of the core areas that lie in this zone. It is clear that the architectural traits that form the passage grave tradition within both Ireland and North Wales are similar. 6 Excavations undertaken at Newgrange and Knowth in the Boyne Valley, Ireland, have revealed that megalithic art is not just confined to the passage and chamber areas. Both monuments, two of the largest passage graves in Europe, possess rock-art on in front and back of kerbstones which delineate the extent of each of the mounds (Eogan 1986;O'Kelly 1972). Although specific groups of carvings are used, no two stones are the same.
At Knowth, Eogan recognised two distinct forms of megalithic art, one angular, the other curvilinear (1986). 7 Both forms occur within the passage and chambers of the same monument, mainly within the inner passage and chamber areas. The eastern passage at Knowth measures c. 30m and both the curvilinear and rectilinear art is positioned around the inner passage and chamber areas (ibid. 188-95). 8 These images therefore would be near impossible to view from the façade. Further restricted visual access is hampered by a slight kink in the outer passage area. At Newgrange angular and curvilinear art are positioned in the same way. Also, angular carvings in the form of infilled diamonds, multiple zigzag lines and triangles dominate the chamber and inner passage areas at Fourknocks. This art too, cannot be viewed from the façade area; the art is hidden and can only be seen by people within the central chamber and possibly from the short passage.
A similar design complexity and strategy is found within the British passage grave tradition. At Barclodiad y Gawres (ANG 4) the passage, measuring approximately 6m, leads to a cruciform chamber area which has a series of uprights decorated with chevrons, lozenges, spirals and zigzag designs. These lightly pecked designs are similar to decorated uprights found elsewhere along the Atlantic seaboard. There is also a design association with nearby Bryn Celli Ddu (ANG 1) and monuments found in the Boyne Valley (Lynch 1967.1-22). Carved decoration occurs on six stones, located either within the inner passage or chamber areas. 9 Decorated uprights included Stone 5 (L8), 6 (C1), 7 (C2) 8(C3), 22 (C13) and 19 (C16) (Shee-Twohig 1981.229). 10 On Stone 5 are lozenges and vertical zigzags; Stone 6 exhibits a conjoined circular motif, lozenges and vertical zigzags; Stone 7 has a large chevron and on Stones 8 and 19 are a series of spirals and a series of unrecognisable motifs. Finally, on Stone 22 is a spiral, with supporting motifs, lozenges, a horizontal chevron band and a series of vertical zigzags. This particular stone bears some resemblance to the decorated Pattern Stone carved on a fallen upright found over a central pit at Bryn Celli Ddu. Unlike the monuments found in the Boyne Valley, at Barclodiad y Gawres all six stones have their decoration facing into the chamber and are being hidden from view. A large percentage of the megalithic art within the Boyne Valley appears to be public art, carved onto mound kerbing.

Perceiving scapes: art, doorways, thresholds and passages
Before I discuss the social (and antisocial) divisions of space within and around the monument, I want to look at the perception given to different spaces and how our senses react in different ways within these spaces. The passage grave builders would have been aware of sensory arousal, of how people using the monument should act and react. The passage grave blueprint, used throughout most of the Atlantic zone during the Middle and late Neolithic, consists of the façade area, the outer passage, the inner passage and finally the chamber. It is clear that as one progresses through these quite different spaces, one's sensory perception changes. In order to move from, say, the façade to the outer passage area, the change of space is represented by both a physical doorway (doorstone) as well as a metaphysical one. Static points such as a stone threshold, constricting or protruding uprights or a lowering of a passage capstone usually represent the physicality of a space. The changes in the physical statementing was argu-6 Earlier architectural traditions in Ireland and Western Britain are also present such as the wedge tombs of the Carlingford-Clyde Group and the Court tombs of southern Ireland and Cornwall etc. (Daniel 1950). 7 These terms in many ways do not capture the style but for the sake of this section of the paper I will use them. 8 The chamber measures around 20.21m 9 Powell & Daniel (1956) and later Lynch (1967) record five stones. However, in March 2006 a team from the University of Bristol discovered a sixth stone that formed the northern upright of the eastern passage. This multi-phased decoration comprised a large chevron that was either superimposed or underlying a series of horizontal pecked lines . 10 The numbering system uses Powell & Daniel (1956), later used by Lynch (1967) and Shee-Twohig's numbering (in brackets) (1981).
ably accompanied by changes in ambience; changes in light intensity (from light to dark), changes in smell (from open air to stale musty air, due partly to the putrefaction of decomposing flesh) and changes in audibility; from the familiar noises made by the family group within the façade area to the near silence of the inner passage and chamber.
Christopher Tilley's pioneering work on passage grave monuments in Västergötland clearly shows the strategic importance and intentionality of certain architectural features (i.e. what materials are used and how they are used). This group of monuments, consisting of over 265 passage graves within an area of 38km by 25km is the most northerly Neolithic group in Europe. According to Tilley, tombs become 'socialised' through their construction and use, thus allowing sites to become socially-politically manipulated (1991.68). This process is evident through the changes to the tomb architecture or changes in burial practice. From Tilley's analysis all monuments are standardised in design, comprising an east-west oriented passage leading to a north-south oriented chamber, which are incorporated into a round mound. The mound is delineated by stone kerbing and it is more than likely that the capstones of the chamber were exposed during use. The interplay between the different colours and textures of the stone, the arrangement of the passage and chamber uprights and the way one moves through these different spaces would have been paramount to users of the monument. Not surprisingly many architectural traits are replicated in the two Anglesey monuments and passages along the Atlantic Zone. What is absent though is rock-art.
These single-phased monuments are regularly spaced within the landscape, sometimes in rows of up to twelve and are very visible. Tilley has also identified a series of intriguing architectural traits that are replicated in most of the Västergötland monuments which further suggests a recognised blueprint in design associated with the ritual use of the monument. Nearly all the uprights used to construct the passage and chamber walls are of sedimentary rock, while the capstones (or roofing stones) are of igneous rock (ibid. 70). The entrance to all the passages are narrow, measuring between 0.5m and 0.8m in width. The entrance to the passage is also low, suggesting that during use, people entering the monument would have to crawl in order to gain access. However, as one progresses through the passage, the walls and the roof open out and upwards until one reaches the keystone (or threshold). The keystone, located at the transition point between the chamber and passage is a deliberately placed capstone that is lower than the other capstones and, when entering the chamber area, one has to crouch lower in order to gain access to the chamber. Like other thresholds, the keystone is yet another device to restrict visual access from those looking into the passage from the façade area. It also possibly marks the point where the body (or body parts) finally enter the other world of the ancestors.
Between the entrance to the passage and the keystone and entrance into the chamber, the body has to travel, albeit a short distance through what Tilley and others has termed as 'liminal space ' (ibid. 74-5). This liminal space acts as a rite of passage whereby the body is neither of this world or the next. This simplistic hypothesis can be further dissected to represent a series of journeys and not surprisingly adding further complexity to the rite of passage and the way the dead are deposited.
The doorway at Pentre Ifan in Pembrokeshire, SW Wales stands around 2.2m above the façade floor level and is located between two uprights which support an enormous capstone estimated to weigh 30 tons (Pl. 1). The doorway, itself weighing around 5 tons is set in such a way that human bone and associated offerings can be placed though the gaps between the door stone and the uprights (Children & Nash 1997). I have suggested that the door stone

Pl. 1. The façade and entrance of Pentre Ifan. The doorway to the chamber is located between the two supporting uprights (Photo: GHN).
may have been physically tilted and moved to one side in order to allow access by certain high status individuals (Nash 2006). Once inside the chamber secret rites could be performed where both the dead and living would have interacted in a special way; what goes on behind closed doors! The doorway therefore acts as a thoroughfare to the realm of the supernatural. Special individuals would have the knowledge and power to act as an intermediary between living and the dead. They would deposit the remains of new ancestors, communicate with the old ancestors, replenish the grave goods and offering the new ancestors food and drink for their journey into the next world. However, can the same process of moving the dead across a series of zones be maintained when there are no passages? At Pentre Ifan (Pembrokeshire), the space between the enclosed façade and the chamber is short and the distance between the façade and the chamber, unlike other Neolithic monuments, lacks liminal space. However, one could counter argue and suggest that liminal space is represented by the thickness of the doorstone. This blocking device separates the façade from the large rectangular chamber. Arguably, this journey between the living space and the space for the dead is restricted. Nevertheless, the journey, however short, is still paramount within the act of burial.
Sometimes doorways are set into the fabric of the passage and cannot be moved. Two uprights found within the inner section of the passage at La Hougue Bie, Jersey restricts the visual access between the façade and the chamber (Baal et al. 1925.Fig. 2). La Hougue Bie, one of Europe's largest and most impressive passage graves and dating to the latter half of the fourth millennium BC has, like other passage graves, more than one doorway. One of these is located between the passage and the façade while the other is between the inner passage and the chamber (Nash forthcoming). This careful and contrived arrangement of stone uprights within the passage establishes restricted visual access. The builders have added further components to the passage to ensure what is undertaken in the chamber remains secret, visually hidden from people using the façade area.
At the eastern end of the passage a deliberate but subtle kink within the length of the passage has been made, and like most passage graves, the passage constricts at the façade end, while at the chamber end the passage walls and roof widen and rise. People using the monument would have therefore been forced to crouch at the entrance in order to gain access to the 9m long passage. However, as one progressed along the passage, one could move from a crouching position to an upright position by the time one has reached the inner passage and chamber areas. The living and the dead are also guided through the passage and chamber with a set of strategically placed cupmarks (Fig. 1). Several are carved within the inner passage area onto smooth pink granite, while a further 21 are carved onto the external face of the northern chamber (Pls. 2a & 2b). Arguably, and similar to megalithic art elsewhere these would have been hidden during the Neolithic (assuming that they date from this period). Furthermore, nine subtly-placed cupmarks are located on the roof [capstone] of the northern chamber (Mourant 1974).
The strategic location of these cupmarks, between the inner passage to the chamber possibly demarcates part of the journey for the dead and it is clear that these marks cannot be seen from the façade. Indeed, only fire from a hearth or torch could illuminate these marks along with other cultural goods such as decorated pottery and the colour of the sea pebble floor (also found within the chamber and passage). Similar to art found at Newgrange and else-  Darvill 1982;Grimes 1936;Hemp 1935;Nash 1997). The majority of the monuments within this group conform to a number of architectural rules including locally oriented lateral passages and a false portal located within the façade area.

Pl. 2a (left). Passage and chamber arrangement illuminated by artificial light (Photo: Adam Stanford
The chamber and passage at Arthur's Stone are set within a cairn rubble mound. The chamber has nine upright stones that support an enormous capstone, estimated to weigh 25 tons. Recent investigations by the author suggest the chamber and passage were set within a long mound that extended some 50m north of the present mound, probably trapezoidal in shape and similar to other monuments within the group (Nash 2006). At the north-western end of the chamber is an unorthodox right-angled passage with a pronounced doorway, consisting of a single upright (Pl. 3).
Between the chamber and a door stone the passage is oriented east. However, from the door stone to the façade and entrance, the passage changes direction to the northwest, pointing towards an impressive scarp of the Black Mountains locally known as Hay Bluff (Children & Nash 1994.26;Tilley 1994. 140). Similar with other monuments mentioned in this chapter Arthur's Stone has a narrow and low entrance that enlarges as one progresses through the passage towards the doorway and chamber. There is clearly no visual access between the chamber and the façade beyond. The 90˚kink of the passage alignment where the doorway is sited appears to show a conscious attempt to separate living space and the realm of order and control, from the inside world of death, dismemberment and disorientation. The transition between the two spaces is achieved precisely at that point, equidistant between chamber and entrance, where the passage abruptly changes direction. It is here that the living come into contact with the dead; an act and point in time that is repeated in every chamber monument throughout the Neolithic world. A similar passage arrangement with its restricted visual access is recorded for the majority of passage graves either with or without rock-art.
The three passage graves forming the focus of this discussion, Barclodiad y Gawres, Bryn Celli Ddu and The Calderstones inherently have their problems. The first of these to be considered and problematic to all monuments of this age, is taphonomy. The surviving material culture probably represents a small percentage of what was within each of the chamber and passage areas. Consumables, such as offerings of food, wood and hide would have long disintegrated over time. This is further hampered by acidic soils that cover most of Wales. Finally, there is a problem with antiquarian and non-scientific excavation methods. All three sites have been severely disturbed; the Calderstones completely destroyed. Thankfully, the two Anglesey sites were excavated during the mid-20 th century, albeit after initial antiquarian investigations during the 18 th and 19 th centuries, and attempts to use the sites as stone quarries (Appendix). Following excavation both monuments were restored. However, it is not known if the restoration of both monuments was sympathetic with the original layout of the monument. The excavation of Barclodiad y Gawres in 1952-3 focused on the excavation of the passage and chamber areas although, trenching was also extended through the surrounding cairn (Powell & Daniel 1956). One can assume that, based on the excavation report, this excavation was by far the most meticulous and scientific. Excavations undertaken at Bryn Celli Ddu, first by Lukis in 1865 and later by Hemp between 1925 and1929, reveal the probability of a two-phase monument. It is not clear how meticulous both excavations were. What is known is that, based on a mid-19 th century engraving, the monument comprised a chamber and passage and the mound had been almost completely removed (used as a quarry). It is not known from the Hemp excavation report if the passage uprights were in their original position or whether or not the outer passage uprights supported capstones. The Calderstones, presently standing out of context in Calderstones Park, have a sadder but well documented history. They are known to have formed part of a parish boundary during the 16 th century and are indeed located on a map from this period. 11 Between this period and the 19 th century the stones, probably forming part of a passage grave were slowly but systematically destroyed. In the early 20 th century the remaining stones were placed in storage and later erected within the entrance area of Calderstones Park, before finally being re-erected in a purpose built vestibule in 1964.
Despite problems of taphonomy associated with the early archaeological investigations, one can consider a number of similarities between each of the monuments. Firstly, and most simply, two of the three surviving monuments are architecturally associated with other passage graves within the Irish Sea zone. The architectural style used by monument builders in Anglesey and in Ireland suggests complex contact and exchange networks between these and other groups to the north (Cooney 2000.226-8;Eogan 1986.220). Secondly, are the more complex issues as to the origins of artistic endeavour. The artistic styles from all three monuments are all very diffe- The artistic style from the Barclodiad y Gawres monument is essentially geometric in form with the predominant designs being chevrons, lozenges and zigzag lines. These occupy four of the six stones that are located within the inner passage and chamber and are essentially hidden from view and cannot be completely seen with natural light. On Stones C3, C13 and C16 large spirals are present while on Stone C3 spirals dominate (Fig. 2). The rock-art from this monument is probably contemporary and despite its unique design coding, many individual design components are also found within Irish passage graves. The meticulous excavation programme undertaken by Powell and Daniel suggests that the position of each of the stones remained in situ and what one witnesses today is roughly what was present during the Neolithic.
At Bryn Celli Ddu the art comprises a spiral and, more impressively, a serpentine-style carving (Fig. 3). Both designs configure with Eogan and O'Kelly's curvilinear classification. The small spiral, (13cm in diameter) located within the southern section of the chamber and carved on an upright, may be a later addition. I suggest this because compared with other spirals elsewhere in Anglesey and Ireland, the Bryn Celli Ddu example has been clumsily constructed. The serpentine form, which is carved on what is referred to as the 'Pattern Stone', was discovered lying prostrate next to a pit, centrally located and west of the chamber (Hemp 1930). The style of the decoration on the Pattern Stone is similar to Stone 16 at nearby Barclodiad y Gawres and extends over three faces. Excavation by Hemp between 1925 and1929 revealed an earlier monument phase that he interpreted as being a henge and it is possible, though perhaps doubtful, that this stone belongs to this earlier monument. However, I am not entirely convinced that a henge phase exists. Instead the characteristics of the henge may be merely an initial construction or ground preparation phase associated with the passage grave. Whatever the phase, the Pattern Stone and its location, hidden from view, has more to do with the dead rather than the living in that it was found lying prostrate, the art lying face down. Likewise, the spiral, if contemporary with the use of the monument as a passage grave, is difficult to locate and could have only been viewed under certain artificial light conditions.

Moving through physical and metaphysical spaces
Previously, I have briefly considered the way our senses change when experiencing different types of space (Nash 2006). I now want to focus on visuality and how rock-art may have played a vital role in providing a series of focal points both inside and outside the monument. Of course one can never completely comprehend how and why prehistoric burial was undertaken; there lacks any secure tangible evidence and any burial service cannot survive the archaeological record. A phenomenological approach and the way monuments are approached, has been undertaken by Tilley (2005), while a more personal account of how people and the dead move through the monument has been effectively discussed by Parker Pearson & Richards (1994). Both approaches extend beyond the realms of empirical discourse. Cynically though, the evidence used is limited to just >5% of the potential material culture. However, the architecture of the passage grave tradition is similar throughout the Neolithic core areas of Europe and in each area there is a limited number of sites whose condition fares better than others. From these sites one can ascertain a 'sense of occasion'; the way one approaches the monument and the way one enters the façade, passage and chamber. The ambience for each of these is usually different depending on the condition of the monument. Fortunately, both Barclodiad y Gawres and Bryn Celli Ddu are enclosed within a rubble mound, albeit reconstructed, thus the passage and chamber components are hidden from the outside.
In experiential terms Michael Shanks has probably come close in attempting to understand ambience and the rhetoric of space (1992). Shanks' retrospective accounts of visiting the Cotswold-Severn monuments of Maes y Felin and Tinkinswood in South Wales and Dunstanburgh Castle in Northumberland is personal and subjective. He encounters though, an incomplete past, a pastiche of preconceived ideas that affords a casual glimpse of the present. The experience of approaching and entering the monument would have been somewhat different during Neolithic times. One can assume, based on ethnographic and historical evidence, that people using the monument during Neolithic times were high status individuals, and having a similar role to the modern day priest. They would have possessed a knowledge and view of the world that would have made him or her different to others. The experience would have been sublime, fulfilling a plethora of human emotions based on terror with fascination, as each space is encountered and experienced (Chippindale & Nash forthcoming). Probably accompanying the priest would have been other status individuals who would also have special sacred knowledge of the dead. The architectural traits such as the constricting passage, the doorway and threshold, and the kink in the passage, would have restricted the visuality between the façade and the chamber. Furthermore, the natural light availability would have diminished as one progressed along the passage into the chamber, adding further visual restrictions on people standing within the façade area. One can see a similar visual restriction within the medieval church. A strategically placed rude screen constructed between the nave and the congregation and the altar area, plus various parts of the act of service being performed by the priest with his back to the congregation, would have established a clear divide between what can and cannot be seen (and understood).
The way the living and the dead move through a scape or physical space have serious implications on the way a space is perceived. Each space, either re-presenting a different architectural space or a different stage in the journey from landscape to the spirit world, creates a different experience or ambience. Both the architecture and the art would have played a vital role in how the dead changed physically and metaphysically as they moved from one space to another.
Concerning burial deposition, it is now becoming clear that the passage grave tradition represents a corporate mentality towards death. However, both Anglesey passage grave monuments show little evidence of this. It is probable that corporate mentality actually means the burial of a high status extended family and that each member would have to undergo a series of processes before he or she could enter the next world. I would stress here that the final resting-place is not the chamber, which merely forms one stage before the metaphysical journey begins.
The journey for the dead begins outside the monument, between the moment of death and entrance to the façade area. It is probable that the monument was outside the main settlement area and would have been approached in a certain way using a series of markers within the landscape. According to the limited radiocarbon dating evidence the passage grave tradition in Britain and Ireland suggests many monuments were in use between the Late Neolithic and the Early Bronze Age (3000-2000 cal. BC). At the same time other monuments were being used such as standing stones, stone rows and stone circles. In the case of Bryn Celli Ddu two standing stones are located nearby and are possibly contemporary with each other . It is conceivable that the standing stones provided a pathway to Bryn Celli Ddu from the east; I regard this as the first journey. Once at the monument, the body would enter the façade area and thus establish a second journey. Whilst in the façade area, the dead may take on a different guise. There is evidence from other Neolithic burial sites in England and Wales of excarnation, such as at Gwernvale in Breconshire and Wayland's Smithy in Berkshire. The body may have lain in state over many weeks, at the same time providing a necessary period for the mourning of the deceased. During this time the body would have transformed beyond human recognition, becoming a skeleton, physically leaving the world of the living and entering an ancestral world.
When the time was right the remains would have been collected and transferred from the façade to begin the third journey, along the passage or what Tilley (1991.75) refers to as liminal space. Liminal space, defined as a metaphysical as well as a physical entity where the dead are neither human nor ancestral, would be divided into two spaces, the inner and outer passage areas. Between the entrance and the inner passage, the dead and its entourage would experience a number of visual and audible sensa-tions. Firstly, the natural or artificial light from the façade area would obscure parts of the passage architecture as one progressed toward the inner passage area. The amount of illumination from this light source would begin to fade. At the same time the noises from the façade would get fainter as one entered into a world of ancestors and spirits. Deveraux (1996) suggests that certain stones, either in situ architecture or movable, may have been struck that could resonate around the passage and chamber adding further sublime sensations for the occupants of the tomb. These noises, along with chanting both inside and outside the monument, would have created an audible sensation that would have been both synchronous and harmonious. It is at this point within the area of liminal space that the entourage would have to light torches or construct a hearth, probably within the chamber area. When approaching the inner passage area the rock-art would be fully illuminated. The carved abstract images each with their own meaning would move [and dance] as the flames from the fire flicker. A large hearth area was exposed during the Powell & Daniel excavation within the chamber area at Barclodiad y Gawres. Similar evidence has been found at the highly decorative monument of Gavr'inis. Here, charcoal has been found in both the chamber and passage areas (Le Roux 1985).
Light emitted from the Barclodiad y Gawres hearth would have illuminated nearly all the stones, including the newly discovered Stone C2 as well as Stone C3, each forming the northern and eastern walls of the eastern chamber, as well Stones C13, C16 and L8. The last two stones flank the doorway between the inner passage and the chamber and can only be seen from these two areas (Fig. 4). The ritual activity that would have been undertaken within the chamber probably remained exclusive to people using this part of

Fig. 6. The three sections of a menhir, each section now incorporated into later burial monuments (after La Roux 1985).
the monument. As suggested earlier, using Tilley's analysis on monuments from Västergötland, restricted visual access would have been in place. The carved symbols, their location within the chamber, and their position on the panel, along with their relationship with other symbols would have been restricted. In the cases of the two passage graves in Anglesey, and probably the Calderstones monument, the art is positioned in such a way that it cannot be seen from the façade or the outer passage. Many passage grave monuments with art appear to conform to this basic rule. It is clear that in some cases, such as the northern chamber or cell of La Hougue Bie or the eastern chamber at Barclodiad y Gawres the rock-art is very difficult to view. It is possibly that this art was to be viewed only by the ancestors before they or their spirits embarked onto their final journey to the otherworld. The finely pecked lozenge and spirals on Stones C2 and C3 respectively suggest that the light emitted from the hearth would not be enough to fully read, and therefore understand each narrative. It is probable that torches would be required in order to fully illuminate these panels. Once the remains of the dead had been interred, grave goods would be offered in order to accompany the dead on their final journey. Based on the fragmentary evidence from both excavations, offerings would have included pottery vessels, flint tools and, in the case of the burials at Barclodiad y Gawres, a finely carved bone pin (Lynch 1969b.158, fig. 21). However, based on ethnographic evidence it is thought that the spirits of the ancestors would not have entirely left the monument. According to many non-western societies the ancestral spirits could be summoned back from time to time, maybe to assist with later internment. The rock-art, which was hidden, static and probably possessing a restricted meaning, would have guided generation after generation of ancestors using a prescribed way of moving through the various spaces, maybe emerging from the art itself. It is clear that many passage graves that possess megalithic art have panels that are entirely hidden from view. At Gavr'inis, Newgrange and in some respects, at Barclodiad y Gawres and Bryn Celli Ddu, certain panels are hidden and inturned towards the monuments mound. Some of this art is extremely elaborate in form (Eogan 1986;Hemp 1930;O'Kelly 1982). It could be the case that this inturned art originates from earlier monuments. At the Table  des Marchands passage grave, Morbihan, one third of an enormous menhir has been incorporated as a capstone into the chamber architecture. Carved onto the stone is a visible axe (Pl. 4). The other two sections of the stone have been also used to form capstones and are incorporated into the roofs of Gavr'inis and nearby Er Grah (Fig. 6). However, the sections containing once continuous figurative art, including two goats and a large axe, have its art hidden. The destruction of the menhir and its incorporation into a later monument appears to be a deliberate act. Moreover, the positioning of certainly two of the three sections conform with how megalithic art is being used in passage graves. * References refer to discussions specific to rock-art rather than the monument as a whole.

Pl. 4. Central section of an enormous capstone, later incorporated into the
** Also known as Treflys

Concluding remarks
In this chapter I have concentrated on the passage grave tradition, focusing mainly on how the living and the dead may have viewed and experienced megalithic art. I have suggested that rock-art, located mainly within the inner passage and chamber areas of two of the three British passage graves is strategically placed. The art can only be viewed and read if the reader is positioned in a particular place inside the monument. I stress, however, that the ambience experienced by people moving from the façade into the outer and inner passage areas and finally, in the chamber area, changes forming part of a special experience that is restricted to only a few. These changes affect all the senses as people move through the different physical spaces, and this is arguably irrespective of whether or not rock-art is present. Rockart merely forms a series of focal points for the living and, in particular for the dead, as they move through the different spaces whilst on their final journeys to the spirit world.
I would further suggest that similar emotions are experienced outside the monument as well, and that both the living and the dead are required to undertake a series of journeys that include moving through open spaces via a series of landscape markers, themselves sometimes decorated with rock-art. These markers provide rigid points through which people have to move in a particular way. Similar to the way we bury our dead, people are required to follow protocol, and funeral rites are properly observed. Markers, usually in the form of standing stones and more subtly, exposed rock-outcropping, would have provided a series of way points whereby the dead (accompanied by the living) made a series of journeys before embarking on their final spiritual journey from the chamber to the next world. When in the chamber area, an array of grave goods would have accompanied the remains.
I have suggested that changes in perception are controlled and manipulated by changes in architecture and that how our senses react would depend on the ambience of each space encountered. Likewise, the dead also change both physically and metaphysically as they embark on their journeys through these different spaces, from a fleshed and recognisable corpse, through to decomposition and putrefaction to finally a collection of bleached bone; unrecognisable to those who are involved in the performance of burial.
Incorporated into inner spaces of the monument is rock-art, usually in the form of a series of abstract geometric forms. Although the syntax is limited to, say, ten distinct symbols, the way they are positioned on the panel suggests a complex grammar was in operation. The dominant symbols from the three monuments, including chevrons, concentric circles, cupmarks, lozenges, serpentine forms, spirals and zigzags are all located on uprights that are visually restricted and appear to function in areas of the monument where probably only a few individuals would attempt to go. Alternatively, the responsibility of moving the dead from a known to an unknown space would have been restricted to may be one of two high status individuals. As the burial monument is used over many generations the original meaning of each of the symbols may have been forgotten or

REFERENCES
elaborated on. In order to read each of these symbols the strategic position of a hearth or torch would have been paramount.
Finally, like all architecture, the passage grave tradition indicates that there is an intentionality in design. The design elements repeated throughout the Atlantic zone including mound shape, passage and chamber plans, and the construction methodology would have had a significance to the people using the monument. The idiosyncratic and subtle position of, say, the doorway jambs, the constricting passage, or the location of the rock-art, indicate a status-driven society where by only elite individuals get the privilege of accompanying the dead on part of their journey and experiencing the light at the end of the tunnel! ∴ ∴ excavation. According to Powell & Daniel (1956) no primary pottery was found, but there was one artefact, which may be contemporary with initial use of the monument. This was a bone or antler pin which was found with the cremation burials in the western side chamber, a pin similar to skewer pins found at Loughcrew and Fourknocks in central Ireland. The pin fragments were all burnt and would appear to be associated with the cremation. The location of a cinerary urn, above the collapsed roof area, also suggests that the deposition of the cremation was subsequent to the initial use of the tomb. The urn had a decorated bevelled rim made up of a series of lines of plaited cord impressions. Within the central chamber was discovered a hearth approximately 1m in diameter which contained a mixture of charcoal and stone chips. Also recovered was an assemblage of shells, fish bones, amphibia, reptiles and small mammals.

Bryn Celli Ddu, Llanddaniel-Fab
Bryn Celli Ddu is located on a low ridge of a glacial moraine at around 33m AOD and close to the Menai Straits and extensive views of the Snowdonia peaks.
To the north and west is a slightly undulating landscape. Other Neolithic monuments are sited within the locality including the dolmens of Plas Newydd, Bodowyr and Perthi Duon.
Based on the passage grave sequence in central Ireland Bryn Celli Ddu probably dates from the Late Neolithic and possibly has an association with nearby Early Bronze Age monuments such as the standing stone that is located in a field some 200m west of the monument (SH 50632 70103). Also worth noting is the recent discovery of 26 cup marks on rock outcropping that lies roughly 250m north-west of Bryn Celli Ddu (SH 50623 70240).
The mound, 26m in diameter, may have been possibly larger, but during part-restoration by the Ministry of Works, the monument may have been severely altered. The entrance, with its two uprights (without capstone), is located on the eastern side of the mound. It leads and into a slab-roofed passage approximately 7.5m in length. Intriguingly, the southern wall of the passage is straight, whereas the northern wall is not (Thomas 1988. 45). To the west of the monument and almost in line with the alignment of the walls of the passage is a standing stone, suggesting fore planning of the monument. The pas-sage leads into a polygonal chamber roughly 2.5m across. 13 Between the entrance and the passage are two sets of kerbing, which suggests two phases of passage grave building.
The Reverend John Skinner visited the site in December 1802 and writes an important account of how he entered the passage. The site was excavated by Captain F. du Bois Lukis in 1865 and later by Hemp between 1925 and1929. Hemp revealed a possible complex multi-phased history to the site. Beneath the mound was, according to Hemp a circular henge consisting of 14 upright stones, some of which were broken others, leaning outwards, and within the centre of this was a pit that was a recumbent stone slab. During excavation, socket holes were found which might represent the position of further uprights. Underlying some of these socket holes was evidence of cremation material. Covering the area of this possible henge monument, but underlying the present mound, was a purple coloured clay that Hemp suggests may represent a ritual floor. However, Lynch (1969a.112) argues that it is a palaeo-turf line, the colour of which has been affected by drainage conditions from the overlying mound. Lying next to the pit was the Pattern Stone.
Finds from the 1925-29 excavation are meagre and included a petit tranchet arrowhead that is probably Late Neolithic in date. Also recovered was a rounded scraper (thumb-shaped end scraper), a small lithic assemblage numbering twenty pieces and a mudstone bead which was found within the turf line of the ditch, south of the passage. Previous antiquarian interest in this monument (dating to at least the early 19 th century) has probably seen the removal of much of the artefactual evidence from this site.
Within the entrance area of the passage-grave were the sockets of five post-holes that may represent a possible burial platform for human excarnation and two hearths. Immediately behind this structure was a shallow pit containing the remains of an ox burial. The presence of the post-holes, the burial pit and the two hearths suggests ritual activity is an ongoing process whilst the monument was in use.
Along with Barclodiad y Gawres, this monument has two stones with megalithic art, one within a pit, the other in the chamber. Decoration of one of two stones includes an anti-clockwise spiral approximately 13cm in diameter. The other stone, known usu-ally as the 'Pattern Stone' was found in a possible ritual pit in the centre of the monument. The Pattern Stone was believed by Hemp to belong to the henge phase.
Considering the design and the presence of an ox burial Lynch suggests that this monument has a greater association with passage graves in Brittany than with monuments in the Boyne Valley in southern Ireland (1969a.111). Within the chamber area is a single pillar-stone, which has no structural use and therefore may be considered as possessing some ritual or at least, an aesthetic significance.
The complex decoration of the 'Pattern Stone', located on three of its faces comprises a clockwise spiral which is linked to a meandering curvilinear pattern, referred to by Shee-Twohig as a serpentine form (ibid. 230). This design covers both faces of the stone. Also present is a cup-mark. The simple spiral may have direct similarities with stone C16 within the chamber of Barclodiad y Gawres and Stones A, B, C, D, E & F of the Calderstones, Liverpool.
Outside the upright stones were the remains of a silted, flat-bottomed ditch approximately 6m in diameter and 2.2m in depth. It is this feature that is considered to be the henge. However, it is just as plausible to suggest that the henge uprights may in fact represent the kerbing of the passage gave.

Calderstones, Liverpool
This now destroyed site has, in my opinion the most complex decorated carved megalithic art assemblage in southern Britain. The history of its destruction can be traced as far back as the 18 th century. Surviving today are just six decorated stones that once formed the uprights of a passage and/or chamber. Documentary evidence for this site is good and several early 19 th century engravings exist of the site which by then was much denuded. The original site is not known and today the stones stand within a glasshouse in Calderstones Park. There have been attempts, albeit based on limited evidence to place each of the stones into their original context (Cowell 1981). Based on the location of decorated stones at Barclodiad y Gawres, Bryn Celli Ddu and monuments within the Boyne Valley in Ireland, the Calderstones' uprights would have been erected within the chamber and inner passage areas.
The stones were first recorded properly in 1954 by Forde-Johnston who labelled the stones A to F (Fig.  5). However, Forde-Johnston and later Cowell (1981) were concerned with recording only the megalithic art. The rock-art present on these stones is divided into at least four chronological phases that include: Phase I -megalithic art includes concentric circles, grooves, lines and spirals. Phase II include Late Neolithic/Early Bronze Age arcs, an axe, cupmarks, footprints and a wheel motif. Phase III includes medieval and post-medieval graffiti (a Maltese cross, shoeprints and text) and Phase IV includes 20 th and 21 st century textual graffiti.
Interestingly, the art from Phase II does not superimpose Phase I art suggesting some degree of respect for the earlier tradition. It is probable that both phases are within a generation or two of each other. Many components used within both phases are found elsewhere on Late Neolithic/Early Bronze Age exposed rock-art in Northern Britain (see Beckensall 1999). Common artistic themes found on all three monuments are spirals and concentric circles and these designs are probably the earliest. Footprints, of which there are four on the Calderstones are considered rare in Britain and are restricted to a limited number of sites. The Calderstone footprints probably belong to Phase II, dating to between c. 2500 and 1800 BC. Similar carvings have been found on a capstone belonging to the Pool Farm cist burial in Somerset and this site is clearly Bronze Age in date.