French version of this article is available on september 2011 Post. Otherwise you can get a PDF version by sending a mail to me

Tuesday, October 18, 2011

Stonehenge : new interpretation of the architectural remains

1. Preamble1

1. I would like to thank my friends and colleagues who have read this article for their critical support, particularly Françoise Laroche for the discussion which led to the assumptions presented here, Nicolas Bresch, Anne Jacquemin, Marie-Christine Hellmann and Amélie Perrier for their numerous corrections and helpful suggestions. The English translation has been amended by Guy Losson.

Stonehenge (fig. 1), one of the most famous megalithic sites, still appears, though many studies have been devoted to its understanding, as an outstanding monument. Long considered a "solar Temple", which opened the door to all kinds of cosmological theories, the most recent studies emphasize its role as a place for burial, a role which appears in a variety of similar structures in this area. In fact, the originality of Stonehenge is mainly due to the monumental stone construction made of gigantic blocks assembled as "trilithons" in the central part, and as a continuous colonnade forming a ring whose diameter is about 30 meters.

Fig. 1 – General view from the north-east (axis and "access" from the monumental avenue)

2. This word derives from Saracen, which originally means “from pagan origin”, and has no scientific ground.

2. A Short description

2.a. Lithic remains

For both the graphic and written description of this phase of the monument, one should refer to the results of the work of archaeologist Richard JC Atkinson, who conducted the most comprehensive and serious study of the whole site, between 1950 and 19643. For the central part, the most accurate plan, reproduced here (fig. 2), was published by Anthony Johnson in 2008. The numbering of artefacts, especially the stones, was established by Petrie in 1880 and allows to identify the elements of the site. Of course, many additional studies and excavations have been undertaken since the work of Atkinson, and they have completed, or sometimes have even changed, the interpretation of the site4.

3. R..J.C. Atkinson, Stonehenge, Hamish Hamilton, 1956. 225 annotated photographs tracing the research, excavations and restoration work carried out since 1901, including those conducted by Atkinson can be seen on the Website of English Heritage.

4. Among other recent works, research carried out by Professor Mike Parker Pearson, from the Department of Archaeology at the University of Sheffield is of particular significance.

The monument itself is part of a series of prehistoric sites located about ten kilometers north of the city of Salisbury in the County of Wiltshire. The most outstanding part of Stonehenge consists of seventy-five huge stones forming two distinct groups. Five units are found in the centre, each having three stones (trilithons), are placed in an elongated pentagonal figure, facing north-east. On the outskirts a "colonnade" with sixty units: thirty pillars and thirty lintels forming a perfect circle placed about 4.10 m high.

This phase of construction, the best known because of its unique appearance and the problems its interpretation and construction poses, is called "Stonehenge IIIa,” by Atkinson, meaning 2000-1700. However, according to sources or authors, dates vary quite widely, between 2600 and 1500 BC. This is a fairly wide span of time within which it is difficult to have a specific milestone. Basing himself on a radiocarbon dating Mike Parker Pearson has recently located the phase of the sarsen to 2600-2100 BC5.
5. Parker-Pearson and others, "The Age of Stonehenge ', Antiquity, 81 (313). pp. 617-639. The issue of various dating will be enlarged upon below.

Let me only mention the other smaller stones, from more distant origin6, which form two complementary figures: a circle of fifty-seven stones around trilithons, and a figure of nineteen stones shaped like a horseshoe among these trilithons. The outer circle was built with re-used stones7. These two groups, later than sarsen in their present position, are not part of the reasoning developed here. Even if their presence is definitely important in the spacial device, they do not interfere with observations that relate to sarsen. Above all, as they do not show up in their original state, their interpretation is more complex.

6. It would seem that these stones come from Pembrokeshire, a County south-west of Wales.

7. At least two blocks are lintels that belonged in a first phase to a trilithon-type structure in bluestone. Atkinson thinks he has found the location of the stones in their first state (op. cit. P48-49), prior to the erection of the sarsen.

This monumental circular stone setting itself fits into a broader composition marked by ground movements: an irregular ditch lined on the inside by a circular bank that is more polished (320 feet - or 97.5 meters8in diameter), a series of fifty-six holes (called Aubrey holes9) drawing a circle of 88 m in diameter, most of which were occupied by cremation burial places. The original destination of these holes is uncertain. Then, still close to the outside circle of the monument, more than fifty other graves whose position has not always correctly been identified by the first excavators10.
8.  I foot = 30,48 cm

9. Named after a 17th-century antiquarian who first reported the existence of these holes.

10. R.J.C. Atkinson, Stonehenge, p. 13-14.

Fig. 4 a and b – 
Bird's eye views (a. from the north-east; b. from the north-west with numbering blocks).

2.b. The peripheral ring of stones

This is a set of thirty upright stones (which I will call "pillars" here) which are placed on a circle of 29.70 meters (108 feet) in diameter inside, with the same number of "lintels"11. The pillars, spaced from each other by 3.11 metres center to center, and 4.10 m high, were numbered clockwise, from 1 to 30 starting from the axis of the so-called "avenue" that is oriented north-east. The lintels are numbered 101 to 130, number 101 being based on pillars 30 and 1. According to this system, it is easy to understand that lintel 127, for example, rested on pillars 26 and 27.

11.  The lintels ' width is quite regular (1 m or so), so the outside diameter can be estimated to 30.70 m. This dimension does not apply to the pillars below, whose rear face is irregular.

The pillars - this feature is immediately noticeable when looking at the pictures - are very irregular in appearance and size, while the lintels are well cut and regular.

Seventeen pillars still stand in their original location. Only six lintels still rest on the pillars: 130, 101, 102 form a continuous sequence, 105, 107 and 122 are isolated on their supports. If we consider the stones in place or the ones fallen, the circle looks perfectly continuous and does not show, which is strange if one accepts the idea of a monumental entrance in this direction, any variation at the meeting of the axis of the so-called avenue.

R.J.C. Atkinson interpreted the slightly wider passage underneath lintel 101 (It is 1 foot taller than the neighbouring openings) as an indication of an axial entry into the circle of sarsen. The width of this passage (1.23 m) is however smaller than those of other bays in the same circle, for example between pillars 5 and 6 (1.32 m) or between pillars 21 and 22 (1.66 m), both bays that no one thought of as being interpreted as an access. These changes are just due to the great irregularity in the size of pillars.

2.c. The central trilithons

In the center, of the five initial trilithons three trilithons are still standing (they are about 6 meters high), the fourth and the axial trilithon, which is higher (7.3 m), are partly destroyed. Since these trilithons are independent structures, the pillars are completely erected above the lintel and the gap between the pillars is very small.
The state of the monument has changed since the first descriptions as some stones have fallen12 while others were re-erected during restoration work undertaken since 1905, but I will not insist on those variations that do not interfere in the discussion below.

12. In particular the axial trilithon, the largest (55+56+156), which collapsed in 1620. Pillar 56, the largest upright stone in Britain, was re-erected in 1901. The neighboring trilithon (57+58+158) collapsed in 1797 but was restored by Atkinson in 1956.

Despite the missing parts, the overall restitution of the monument in stone seems clear: a central oblong horseshoe opening to the north-east direction which corresponds to a monumental straight "avenue" that is over 500 m long. This figure is enclosed in a continuous circular portico.
3.  Remarks about stonework

This article aims to provide an interpretation of the monument based on material characteristics observed and reported since a long time, but which, to my knowledge, have not been explained by the previous scholars who studied the monument.

Fig. 5 a and b - 
Well dressed interior facings of the outer ring of sarsen.

The "colonnade" in fact shows a curious feature: the lintels are carved on all sides. The inner and outer facings are especially prepared according to the geometry of vertical cylinders corresponding to the rays inside and outside the circle. It is different with the pillars : we can see that they are prepared only on the side towards the inner structure of trilithons13. The other vertical surfaces, that is to say, the two sides and rear, are quite irregular, bearing some traces of cleaning, but most often in their original state after having been extracted from the quarry (fig. 5 and 6).
The pillars show very irregular widths, and the result is a complete lack of homogeneity in the succession of pillars, which contrasts with the regular interior facing where the vertical faces of the lintels and pillars are cut perfectly straight in a concave manner, and perfectly fit the regular shape of a cylinder.
In any context of stone architecture of antiquity, the following conclusion would have resulted from this simple observation: the regular faces were meant to be viewed, and on the contrary the roughly worked faces correspond to faces which remained invisible.

13. R.J.C. Atkinson, Op. Cit., p. 23 : « In every case the better (i.e. the flatter) of the two broad surfaces has been set facing inwards. »

Fig. 6 a et b – 
Contrast between the treatment of external and internal surfaces: the example of pillar 27.
Conversely, to accept this difference in treatment for a monument which is visible from such a distance causes some amazement all the more so because of the chaotic appearance it gives the monument from the outside. At the same time, why is it different for the crown of lintel, which was however the most difficult part to get access to ? The following lines aim at taking advantage of this fact - which in itself is not new but has never been really explained - and see what it implies for the restitution of monumental Stonehenge.
The empty spaces between the pillars are, as mentioned before, really variable, depending on the width of these pillars. They vary around 1 m, more precisely between 80 and 166 cm. These variations are incompatible with the idea of a passage. There is a clear contradiction between the supposed "monumental avenue" several hundred meters long and the absence of any access device along its axis at the very point of its contact with the stone circle. Besides one can only be surprised that a structure whose internal surfaces have been carefully dressed should on the contrary reveal rough surfaces for the visitors who approach along what is considered as a monumentalized “avenue”.

But there is a construction technique which is attested in ancient times, and already in megalithic monuments in the Western world. It consists in structuring the walls by vertical blocks and then fill the voids with stones that are more or less worked. A telling example of this process is still visible in the dolmen called the "Fairies' Cave", located in Brueil-en-Vexin near Paris (fig. 7) and in the covered alley called Crec'h Quillé at Saint Quay Perros (fig. 8). These structures with alternating upright big stones and filling are used, in this funerary architecture of prehistoric megaliths, as retaining wall. The best side is facing an open space designed in a hollow14. On the contrary, the sides and back retain the embankment.

14. In the two examples cited here, the upright stones were linked horizontally by slabs covering the walkway and not by lintels like in Stonehenge.

Fig. 7 – « La Cave-aux-Fées », Brueil-en-Vexin (France)

Fig. 8 - Entrance to the covered walkway at Crech'h Quillé (Côtes d'Armor, France)

My colleague Philippe Fraisse told me that this technique of filling in gaps between vertical slabs of gneiss is still used in Greece in the Cyclades, for instance in the islands of Tinos and Paros. As evidence I produce two pictures here, pictures taken by him in 2011 that show the amazing historical permanence of this ancestral technique.

Fig. 9 - Separation or terrace walls in Tinos, Cyclades, Greece
(photographs : Ph. Fraisse).

The drawing reproduced here is arbitrary considering the size and shape of the stones that provide the filling: the flat stones were inspired by some of the examples discussed here. The wall may have been covered with a rough coating which would ensure the uniform appearance of this stone retaining wall palisade. This coating, uniting the entire wall, would conceal irregularities created by the width of pillars and would settle the issue raised by pillar 11, whose much smaller dimensions have posed many problems for commentators who were puzzled by its exceptional reduced size.

Fig. 10 - The stone structure of Stonehenge in its current state (left) and restored (right) with a filling of rubble (probably covered with a coating).

Fig. 11 - Difference of size between the pillars: the case of pillar 11, particularly narrow.

Notwithstanding the careful work of internal facings, let us signal another feature that clearly distinguishes the elaborately constructed trilithon from former Stonehenge megalithic phases, namely the presence of a system of mortise and tenon cut at the same time as the blocks, ensuring the strength of horizontal and vertical assemblies15. This technique seems to have been inherited from a wood technology, that is to be found in Mycenae or in Lycia (Turkey), again as part of an architecture that implements blocks of considerable size. This tradition continued until the classical period (lower part of the monument of the Nereids in Xanthos16).

15. Completed by another type of assembly inspired by wood construction: horizontal lintels show tongur-and-groove joins that allow the stones to block each other.

16. P. Demargne, Fouilles de Xanthos I, Les piliers funéraires, fig. 10, P. Demargne and P. Coupel, Fouilles de Xanthos I II, Le Monument des Néréides: architecture, Klincksieck, Paris, 1963, p. 34 and n. 6.

4. New restoration proposal of Stonehenge

What would the monumental complex of Stonehenge look like in this case?
We would have a large circular space bounded by a continuous wall retaining a conical mound of earth whose limit would be indicated by the slope perimeter of about 100 meters in diameter which can still be seen on the ground. From the mound, only the ring of stones formed by lintels and carefully prepared because of its impact on the site would emerge and, in the central space forming a circular enclosed courtyard, the tops of the central trilithons. Incidentally the issue of how to obtain the regular surfaces of the ring would no longer arise because instead of being perched more than 4 meters high, it would be accessible at the top of the mound.

Fig. 12 - Restored views of the Stonehenge mound : from far (a) and from close (b).

Stonehenge, in this case, would be part of the well known family and widely developed type of burial mounds or tumuli, here with an open central portion, because I rule out the idea, despite the relatively short distances to cross between the peripheral ring and the trilithons, of a device covering the central circle. Not only is there no evidence to that effect, but such a device would be difficult to design because of the lack of trilithon in "front" of the central pentagon and also the increasing height of the trilithons. The slopes of the "crater" would be the area for burial of the deceased and the crater area was probably reserved for rituals, an area organized by trilithons whose height was slightly higher.
Before further discussion about the consequences of this hypothesis, let me first try to meet objections that could go against this interpretation.

4.a. The disappearance of the fill blocks

This objection, which immediately comes to mind, consists in saying that we should have found stone filling on the site. Yet the answer lies in the many examples of architectural remains of this type including dolmens that present no more than their main stone structure. Let me insist on two more recent cases, where we realize that the major structural elements of buildings stayed in place, while the filling has completely disappeared: a Roman oil factory in Brisgane, Algeria (fig. 13), and two buildings that probably date back to Roman times in the still unexcavated site of Blaundos in Phrygia (fig. 14).

The explanation for this loss applies to Stonehenge all the more as, where everyone knows that the movement of larger blocks required considerable resources: the constructions, transformed into real quarries, offered abundant building materials, which had been already prepared for use. Over centuries, it is no real surprise that in a region lacking building stones such as Stonehenge, all that could be recovered, that is to say primarily transportable stones, should have been recovered.
 Outside the block filling, five pillars of the crown were partially destroyed and five other pillars, on the south-west side, have totally disappeared, probably having been crushed as their transport was not an option. It definitely proves that the site was operated as a quarry during its history. People in the neighbourhood there, after first recovering the stones that were the most easily transportable, have also reused larger monoliths by first reducing them to the state of fragments.

4.b. The disappearance of the earthwork.

My colleague Jacques Bourcier an engineer was kind enough to calculate the volume of soil corresponding to the mound overturned here. The result is 22,300 m3, a value well below that of other nearby mounds. The largest known mound, Silbury, corresponds to a volume of 248,000 m3 of earth! He also calculated that a ramp with a slope of 1% (630 m long) would require 33,400 m3 of earth. Today only a circular bank of earth is visible on the site, it looks perfectly regular but is quite silted, and bordered by a ditch which is itself very irregular both in plan and depth.

The disappearance of earth, which is in sharp contrast with the maintenance of other burial mounds can be explained by the presence of hundreds of graves, whose presence is revealed by the latest research17, maybe thousands if we restore the mound in its elevation. The search for artefacts in the burial was enough motivation for the systematic digging and clearing of the mass of earth, unlike the royal or princely tumuli of the Eastern World which were limited to a single burial chamber. Even in the case of a single grave, the mound may disappear completely as that of the tomb of the Princess of Reinheim that was recently restored on the site located on the French-German border (Archaeological Park Bliesbruck Moselle / Saar Reinheim)18. The mound of Silbury, 30 km from Stonehenge, has passed through the ages because no research done there for centuries has revealed any graves. It is to be noticed that in agricultural areas there is a trend to erase mounds . 

17. « For Parker-Pearson, Stonehenge may have been a cemetery reserved for a local elite, a cemetery which later would have become more and more popular, as indicated by the increasing number of graves over time. His opinion is that from the beginning, the site was dedicated to the realm of the dead [...]  », Archéologia n°460, p.23.

18. « At the location of the pond located east of the villa a small hill of 120 m diameter of 2 m above the level of surrounding fields could be found in 1950. Excavations have shown that it was a burial area with several Celtic mounds, which had been levelled since Roman times » (I underline). Internet Site presentation of Bliesbruck-Reinheim:

Concerning the small bench that is still visible, a close examination of the data from excavations - as far as the few data allow - provides a clear enough explanation. During the excavations Atkinson was surprised to find a bench of chalk below the slope itself and he suggested a partial dissolution of the chalk on the entire site except under the embankment19. This feature is really visible on the photo of Figure 15, where I have also tried to draw the section revealed by the excavation. The explanation by natural erosion put forward by the excavator is inconsistent and refers to a geological period that is incompatible with the archaeological data20.
A far more simple explanation, is to consider that this bench is intentional and was intended not only to mark the limits of the mound in a precise way, but also to maintain the earth in the circle drawn by the crown. Has this bench been obtained by stripping the surrounding areas or by the movement of material during the digging of the parallel ditch (which would explain its great irregularity)? Only one new study of this bench, actually unexpected but probably necessary for the proper maintenance of earth, would provide further elements to respond to this question.

19. « Atkinson, Op. cit. p. 10 : « Indeed its actual height is even less than it appears, for excavations in 1954 revealed the surprising fact that the surface of the natural chalk is nearly a foot higher beneath the centre of the bank than it is elsewhere ».

20. Dean Talboys, « Stonehenge. The Geological Date of the Site is between 8,000 - 27,000 Years ».
Taking the explanation of Atkinson literally, the author deduces a much earlier date for the site due to the natural dissolution speed of chalk!

Fig. 15 – Explanation for the surelevation of the chalky soil.

Fig. 16 – General view showing the insertion of the stone building in the tumulus.

4.c. Positive arguments supporting the hypothesis proposed

In addition to the main argument concerning the processing of surfaces and the regular appearance that would then present all the visible parts of the architecture built on the site, several points must be underlined for which our hypothesis provides an answer to the questioning or solves apparent contradictions.
Beside the surface dressing, we can also notice, when looking at an elevation, that the size of pillars is very irregular and in complete contrast with their precise location, which was the most important because it depended on the circle of lintels, the only visible culmination of the complex. According to Atkinson, only the pace of implementation of the pillars has been the subject of special care.
The orientation of the complex and its prolongation by the "monumental avenue" has been overstated. As we have seen, the relationship between these two entities does not give rise to any particular device that would allow access to the inner round space. It is likely that this "monumental avenue" rather corresponds to a ramp for the transport of huge blocks21, previously delivered by river. If this "avenue", whose interruption between the outer ring and the ring of stone would also be inexplicable in the commonly accepted scenario where all the device remains at ground level, is oriented in a special way , it is actually because it continues the axis of the device of trilithons. The space enclosed by the trilithons was actually oriented to sunrise.

Atkinson himself stressed that it was impossible to exactly determine, from the remains, this axis and, hence, the empty speculations about precise astronomical calculations. For example, I demonstrate , by arbitrary comparison, the similar orientations of Stonehenge and of the temple of Apollo at Delphi. To orient sacred buildings towards the rising sun is a common feature during antiquity22.

21. The process for the erection of the blocks is explained below.

22. Concerning the Greek world, see Marie-Christine Hellmann, Architecture grecque, 2. Architecture religieuse et funéraire, Picard, p. 188-189

When we look at a map (fig. 18) at the route between the River Avon and Stonehenge23, we understand that the first part of the course aims to reach a point on this axis, from which the ramp will start, built with very little inclination that allows to move the biggest blocks to the top of the mound.

23. Portions of the road between the straight part of the “avenue” linked to the Avon River were identified during a campaign of aerial photographs conducted in 1921 (Atkinson, op. Cit. P. 56).

4.d. Setting up blocks

This question, which mainly concerns the larger blocks, has led to many assumptions. The pillars of the outer circle weigh an average of 26 tonnes24, the lintels only 6.75 tonnes, but the difficulty was especially due to raising the pillars to the top, even a little bit higher because they fit into these pillars using mortise and tenon.

24. With large variations depending on the thickness of the blocks.

Fig. 17 – Compared orientations at Stonehenge and Apollo Temple at Delphi.
.           In Delphi, the orientation of the Altar (right, older than the temple) is even closer to the general orientation in Stonehenge (difference 1°).

Fig. 18 - Map of the area of Stonehenge (from Archéologia n°460, p. 24)
A : Stonehenge
B : Cursus (about 3.500 BC)
C : so-called « avenue » creating a link between Stonehenge and Avon River.
G et H : series of tumuli in the Stonehenge area

The blocks of trilithons are still heavier, almost 50 tons for the pillars of central trilithon. The excavations have provided an important indication because, at the foot of the pillars, holes made to receive the buried part of the block, have been found. They include a sloping side (45°) to receive the inclined blocks before they were erected to their final position.

For the ring device, the inclined side, facing out, shows that the blocks were prepared by pulling them toward the center of the monument. For trilithons, the downside instead shows that the blocks were recovered by pulling them from the outside. All the proposals made so far assume that the central trilithons were built first for a problem of accessibility and that the peripheral ring was completed afterwards.

Regarding the peripheral ring, the simplest hypothesis is to consider that access of the blocks was done along the slope of the mound. To switch the pillars, it was enough to raise them to a height that was slightly above their center of gravity, that is to say 2.5 meters25. The techniques used are the same as those of the erecting of Egyptian obelisks26.
The minimum slope of the mound necessary for this task is 4.3°27. To then place lintels, we must raise the mound to its final height (just over 4.10 m). The slope rises this time to 11% but this time it is to transport the "smallest" blocks among the sarsen. 

25. One must indeed take into account the full height of the block, including the buried part.

26. For comparison, the oldest known obelisk, erected during the reign of Sesostris I in Heliopolis (1942 BC) is three times heavier than the largest blocks of Stonehenge. It measures 20.73 meters (high )and weighs 120 tons. The technique of setting up obelisks has been described by J.-Cl. Golvin (La construction pharaonique du Moyen Empire à l'époque gréco-romaine : Contexte et principes technologiques , Picard, 2004 ): it consisted in dragging the block from a platform to its location by removing the sand located in the pit prepared in advance.

27. A parodic text of the scribe Hori that describes the construction of a ramp mentions a slope of 7.5% ("papyrus Anastasi," manuscript of 10 247 British Museum, quoted in J.-Cl. Goyon et al. La construction pharaonique, p. 208-209).

The ring being completed, the axial ramp that I reconstitute in place of the alleged processional route allows to bring and, more importantly, to implement the trilithons in the most economical way. As a matter of fact, this ramp, passing above the lintels, ends by a central platform that enables the implementation of all the blocks of the group of trilithons. Whatever length is considered for the avenue with a minimum of 230 yards = 210 meters (Atkinson p. 67), the slope obtained is very low (maximum 2%)28.
  Not only does the difference in height of the central and the peripheral colonnade make it easy to overcome the obstacle of the lintels of the circular ring, but, again, we can see the perfectly logical placement of the ramp as the trilithons are higher along this direction: 6.10 m (lintel included) for the first pair, 6.55 m for the next pair and finally 7.30 m for the axial trilithon (6.70 m without the lintel29) : see fig. 20.

28. The Avenue goes to the North East on a distance of 590 m, then bends to reach the river, which flows 1,8 km away. To reach the height of the lower face of the lintel of the largest trilithon (6.20 m), a ramp of 1% is sufficient along the whole straight section. If we consider a shorter distance of 342 meters (which corresponds to the visible part of the road), the percentage is 2%. The ramp of Khafra pyramid was over 1.5 km long and in some Egyptian quarries, they can reach 12 km. These calculations show that the axial portion of the avenue of Stonehenge is more than enough to move blocks up with a very low slope.

29. The lintel of the axial trilithon lies on the ground.

Furthermore, by first constructing the ring of stone, well wedged through a vertical and horizontal system of mortise and tenon, it leads to a very strong peripheral structure that can more easily control the implementation of central blocks, since it gives the opportunity to attach tow cables along the diameters on opposite points for each of the trilithons.

Having a stable ground up to the top of the blocks over the entire periphery greatly facilitates operations (fig. 21). There is no need to imagine any lifting system consisting of cranes or pulleys that would be anachronistic at that time and even ineffective.

Fig. 19 – First phase of work, view showing the construction of the stone ring.
                Left : pillars are carried along the mound. A slope of 5% is needed to bring blocks to a level from which they can be toggled to their final position.
Then the void between pillars is filled with rubble so that this structure can retain the elevation of earthwork behind.
Right : to put in place the much lighter lintels, earthwork must be raised to the top of the pillars. Then the slope is about 10%.
No lifting device is required for all these operations because stones are dragged only on logs or rather sled, and descended with ropes.

At the beginning of the article in which he proposed an early dating for the erection of these stones, Mike Parker Pearson underlined how a priori impossible it was to accept all the dating given by radiocarbon, since the erection of trilithons, according to these dating would have been done later than the construction of the stone circle30. For this reason he came up against the interpretation of the pits where the artefacts were extracted from and dated.

But we now see that what seemed impossible, namely the placement of trilithons after the peripheral ring, in the hypothesis of erecting the stones from the ground, would on the contrary fully be explained in the scenario of a unique axial ramp for the implementation of trilithons, from the center, by descent in the cavities provided for this purpose.

30. « The discrepancy between these two sets of dates poses a conundrum. How can the great trilithon, dated to 2440-2100 ca BC, be later than the sarsen circle which encloses it? Was the circle incomplete or even partially dismantled to allow builders to erect this enormous structure within the monument? », Op. Cit. p. 621.

4.e. The central trilithons

The five central trilithon pillars differ from the pillars by the fact that the uprights are placed entirely under the lintels, that is to say that the central space is very small, reduced to an acceptable minimum needed to implement them with the use of ropes. As a result, the central void, that is again totally irregular, does not seem to be interpreted as a "gate" (a human being does not easily pass through it or with real difficulty, depending on the case) or even as a device to intentionally create an opening .

Fig. 20 – Second phase of work: delivery and erection of the trilithons. From a central platform, the trilithons, whose height grows along the ramp, are put in place in their final spot only with the help of ropes. Then the lintels are carried horizontally after the pit has been filled.

However only careful observation on the site could provide evidence that there was a filling that would transform the trilithons into rectangular panels measuring ca 5 m by 6-7 m high. I applied such a filling in Figure 16, noting that such large panels could create qualified surfaces for decorations. There are images of axes31 and some other geometric patterns engraved on the back of some trilithons, but because of their small size and especially their random position, they probably don't belong to any intentional setting.

31. The similarity of these axes with Mycenaean representations had provided one of the arguments used by Atkinson to establish a link between the builders of Stonehenge and the Aegean world from the Bronze Age.

The layout of the five trilithons is reminiscent of the "horseshoe" oval shape which can be found in the oldest religious structures in Greece32. It would be tempting to imagine a covered space, but I have already mentioned that this assumption is not supported by any trace on the remains, which most probably stood in the open.

32. Marie-Christine Hellmann, Architecture grecque, 2. Architecture religieuse et funéraire, Picard, p. 35-42

Fig. 21 – Transport of the trilithons on the ramp placed in the axis of the monument after the construction of the stone circle.

5. Consequences for the functional interpretation of the site.

The hypothesis presented here is consistent with a funeral function of the monument that can be called a tumulus with a central courtyard. On the other hand, the pillars being part of a continuous wall and the trilithons being in an enclosed space, any geometric interpretation related to solar alignments, apart from the general direction depending on the summer (or winter) solstice become meaningless.

We should remember that recent research undertaken on the site33 goes almost all in this direction and point out that Stonehenge was mainly a burial place from the third millennium. This burial place was completed with a monument devoted to a ritual dedicated to the dead probably due to the status of the people buried there. The main change is that this monument is set inside a tumulus which is a very common device in this area .
33. Mike Parker Pearson, as part of the Stonehenge Riverside Project : 

We should notice that this interpretation leads to the absence of a walk-in access. There probably was, to access the central courtyard, a secondary device or a device that was implemented only when needed. Such a situation is well suited to a funeral place open to the winds.
It has already been noticed that the dating of this part of the construction was much later than the other Neolithic monuments of Western Europe. Indeed, the regular geometry presented by the memorial stone with its crown34, and assembly of vertical and horizontal stones, do not belong to the time of the menhirs, which is a much cruder architecture where geometry is not required35. In this case we are oriented towards an environment which is much closer, from a technical point of view, to the constructions of the Eastern Mediterranean in particular Egypte36 and Mycenaean Greece. Depending on the timing of the erection of the Sarsen phase of Stonehenge, the reference to Egypt or to the Mycenaean world is more appropriate.
In the Mycenaean world we can find buildings which, although architecturally much more developed (tholoi like the “Treasury of Atreus” in Mycenae or at Orchomenos in Boeotia) have the same layout plan: a circular central space dedicated to ritual and graves placed laterally under a mound of earth. One could also mention the "circle of graves" at the entrance of the city of Mycenae. The hypothesis of the use of Mycenaean builders had already been advocated by Atkinson and accepted by Jean-Pierre Adam38.

This issue is however related to problems of chronology beyond the scope of this study. If we accept the dates proposed by Parker Pearson (before and after 2400), we will turn more readily to the hypothesis of an influence, even with technological transfer from Egypt, where at the same time the Pyramids of Giza were built including the connected funerary temples. The use of an axial ramp would not be surprising in the case of an Egyptian influence.

34. RJC Atkinson 34 (Op. Cit. p. 26) emphasizes the extraordinary precision of the implementation of the ring of lintels.

35. It is the only one in which the stones are squared, dressed, and provided with lintels or imposts, Frank Stevens, Stonehenge Today and Yesterday, Heywood Sumner, 1916.  Stevens refers here to the many circular funerary structures found in Britain.

36. Note that the construction of the valley temple (or temple of granite) part of the Khafra funerary complex, Giza, has striking similarities, especially the height of the pillars which is the same: 4.10 m!

37. R.J.C. Atkinson, Stonehenge, p. 163-164.

38. L'archéologie devant l'imposture, Robert Laffont, p. 140-141.

Stonehenge anyway seems to be part of a tradition that began in the Neolithic period (including dozens of earlier tumuli nearby Stonehenge) and continues during the 2nd and 1st millennia BC in the East (Lydian tombs or Macedonian tumuli).
Many issues have to be reviewed in the light of this new interpretation. It is known that the first stone circle (bluestones or foreign stones), built before the monumental ensemble mentioned here, was moved and rebuilt, perhaps in the form of the circle seen today in the space between the central device and the perimeter wall. Smaller than the ensemble studied here, it may be the boundary of a sacred area.

5.a. Other upright stones

I have not mentioned several isolated blocks, some of which had their heyday before the most recent studies have shown how inaccurate the cosmological exegesis is: I mean the (Friar's) Heel Stone that stands in the “avenue”, surrounded by its own small tumulus. This megalith existed before the ramp and maybe reappeared after removal of the latter. Atkinson had already settled the fate of the alleged alignment of this block towards sunrise at the summer solstice39.

I also think that the two "stations" on the periphery of the mound are earlier works than the constructions that have been discussed here. All these blocks which, unlike ours, are part of an elaborated design, are instead isolated blocks, that are completely raw and belonging to the megalithic era. As for the alleged stone of sacrifice (Slaughter Stone), this stone is also a standing stone erected in alignment with the Heel Stone, marking the axis of orientation of the complex of Stonehenge before it was "petrified" with the sarsen.

39. R.J.C. Atkinson, Stonehenge, p. 15.

Two stones placed at the ends of a diameter seem to fit with two missing stones from which only remain the foundation holes. These four blocks that are not worked, and whose date is not provided, delimit a rectangle that is perpendicular to the axis of the monument, whose tops are near the limits of the mound40. In all likelihood, they marked a geographic focus or a boundary of space devoted to the cult of the dead.
The stone called "altar stone", that is partly buried under fallen pillar 59 and hence inaccessible, may be a key to understanding the ritual that took place at the center of Stonehenge: exposure of the body, cremation ... but for now, any interpretation remains purely conjectural.

40. These are the four large black spots visible on Figure 2
5.b. Holes Y and Z.

Discovered in 1923-24, this double ring of holes is a problem because of its presence outside of the ring of stone, that is to say in an area covered by the mound in my hypothesis. But it must be recalled that on the one hand these holes have never been used and especially that one of the holes is missing, corresponding to the site of a fallen pillar41. They could be an earlier (abandoned project) or, conversely, from a later period, for instance from a phase of destruction of the entire funeral complex. Atkinson believed that a voluntary destruction could be dated to the Roman occupation or to the Middle Ages.

The fact that these holes are devoid of any filling and that they have never been of use could argue for larger projects begun and then abandoned, with the same number of columns, spaced further apart. These projects would have been abandoned precisely because of the big span, a span difficult to fill. In this case, in fact, the project corresponding to the Z holes would have been 38 m in diameter (axial distance of 3.95 m between the pillars, instead of 3.10 m), the project corresponding to the Y holes with a diameter of 53 m (axial distance of 5.50 m). But the missing hole in front of the fallen pillar 8, mentioned above, makes me rather lean for a development that followed the destruction phase.

41. In addition, the circular figure is very approximate, unrelated to the accuracy of the circle of sarsen.

6. Conclusion

This hypothesis could also affect the interpretation of the remains prior to the lithic phase, that is to say, the post holes that indicate a first state of the site, organized for a circular wooden constructions. Maybe the holes can then be interpreted as receiving posts for ongoing fences, but this is beyond the subject matter here42.

42. Note the discovery in 2010 of a comparable site, 900 m west of Stonehenge, called Woodhenge because of the absence of any remains in stone.

This study aims to reconsider the architectural remains of Stonehenge from a technical point of view, irrespective of any preconceived ideas. Although the appearance of the restored monument in the phase of the Bronze Age can, as a result of this study, appear very differently from how we used to imagine these remains, the conclusions that can be drawn would paradoxically confirm the intuitions of many archaeologists who, like Atkinson, have tried for decades to eradicate Stonehenge from its mysterious and esoteric approach and emphasize its funerary function. We consider this monument, if we accept the proposals presented here, as a typologically more simple building and especially a construction of great logic, due to the adaptation of a traditional form to the constraints of erecting the huge stones that made Stonehenge so reputed.

Appendix: the memorial of Huisnes-sur-mer, a modern Stonehenge?

      My friend and colleague Nicolas Bresch, whom I had explained my ideas about Stonehenge, immediately reported to me the existence, not far from Mont-Saint-Michel, of a memorial which shows a striking similarity with the general layout I propose for Stonehenge (fig. 23).
         Designed in 1961 by the renowned German architect Johannes Krahn, close to the site of the battle of Avranches that sealed the success of the Allied landings of June 6, 1944, the memorial of Mount d'Huisnes gathers the remains of 11,956 German soldiers who died during fighting in western France. We find there too the idea of an artificial mound with a large open courtyard in the center, surrounded by the graves of soldiers placed in crypts housed on two levels in a structure which recall that of the English site.

       The global concept of this monument is every bit the same, but it has not been possible for me to find information about the references used in this project. There is only one significant difference : the "wide avenue" whose reality we deny at Stonehenge as an access device, exists in Huisnes as a ramp for visitors. In this case, there was also no need to transport large blocks of stone. In the central area stands a large cross around which, especially on November 12 (Volkstrauertag), the ceremonies of remembrance initiated by the Volksbund association who initiated the project, take place.

Fig. 22 - German Memorial in Huisnes-sur-Mer, Normandy
- 1rst Floor and 2nd Floor plans
- Bird's eye view (Google)
- the inner space for ceremonies with the burials around


Bibliography :

Jones, I, 1655, The most notable antiquity of Great Britain vulgarly called Stone-Heng on Salisbury plain. Restored by I.J. Edited by J.Webb. London: J. Flesher for D. Pakeman and L. Chapman
Stukeley, W, 1740. Stonehenge: a temple restor'd to the British druids. London: W. Innys  &R. Manby
Cunnington, W, 1884, Guide to the stones of Stonehenge. Devizes: Bull Printer.
Petrie, W M F,  1880,  Stonehenge: plans, description, and theories.  London: Edward Stanford
Atkinson, R.J. C.  1984. 
Barrows excavated by William Stukeley near Stonehenge.  Wiltshire Archaeological and Natural History Magazine, 79, 244-246.
Thurman, J, 1868, On ancient British barrows, especially those of Wiltshire and adjoining counties. Part 1, Long barrows. Archaeologia, 42, 402-21.
Gowland, W, 1902,
Recent excavations at Stonehenge. Archaeologia, 58, 37-82.
Hawley, W, 1921, Stonehenge:interim report on the exploration. Antiquaries Journal, 1, 19-41.
Hawley, W, 1922, Second report o the excavations at Stonehenge.  Antiquaries Journal, 2, 36-52.
Hawley, W, 1923, Third report on the excavations at Stonehenge. Antiquaries Journal, 3, 13-20.
Hawley, W,  1924, Fourth report on the excavations at Stonehenge, 1922. Antiquaries Journal, 4, 30-39.
Hawley, W, 1925, Report on the excavations at Stonehenge during the season of 1923. Antiquaries Journal, 5, 21-50.
Hawley, W, 1926, Report on the excavations at Stonehenge during the season of 1924. Antiquaries Journal, 6, 1-25.
Hawley, W, 1928, Report on the excavations at Stonehenge during 1925 and 1926. Antiquaries Journal, 8, 149-76.
Newall, R. S, 1929, Stonehenge. Antiquity, 3, 75-88.
Newall, R. S, 1929,  Stonehenge, the recent excavations. Wiltshire Archaeological and Natural History Magazine, 44, 348-359.
Newall, R. S, 1952, Stonehenge stone no. 66.  Antiquaries Journal, 32, 65-7.
Atkinson, R J C, Piggott, S, & Stone, J F S,  1952,  The excavations of two additional holes at Stonehenge, and new evidence for the date of the monument.  Antiquaries Journal, 32, 14-20
Atkinson, R. J. C, 1956, Stonehenge. London. Penguin Books in association with Hamish Hamilton.
Vatcher, F de M & Vatcher, H L,  1973,  Excavation of three postholes in Stonehenge car park.  Wiltshire Archaeological and Natural History Magazine, 68, 57-63
Atkinson, R. J. C. & Evans, J. G.  1978.  Recent excavations at Stonehenge.  Antiquity, 52, 235-236.
Evans, J G, 1984,  Stonehenge: the environment in the late Neolithic and early Bronze Age, and a beaker burial. Wiltshire Archaeological and Natural History Magazine, 78, 7-30
Smith, G, 1981, Excavations in Stonehenge car park. Wiltshire Archaeological and Natural History Magazine, (1979-80) 74-75, 181.
Pitts, M. W,  1981, The discovery of a new stone at Stonehenge. Archaeoastronomy, 4, 17-21.
Pitts, M. W,  1982, On the road to Stonehenge: Report on investigations beside the A344 in 1968, 1979 and 1980. Proceedings of the Prehistoric Society, 48, 75-132.
Bond, D, 1983, An excavation at Stonehenge, 1981. Wiltshire Archaeological and Natural History Magazine, 77, 39-43.
Crawford, O. G. S, 1923
, The Stonehenge Avenue.  The Observer, 23rd October 1923.
Clay, R. C. C, 1927, Stonehenge Avenue. Antiquity, 1, 342-4.
Smith, G,  1973,  Excavations of the Stonehenge Avenue at West Amesbury, Wiltshire. Wiltshire Archaeological and Natural History Magazine, 68, 42-56
Stone, J. F. S, 1947,
The Stonehenge Cursus and its affinities. Archaeological Journal, 104, 7-19.
Stone, J. F. S, Piggott, S, & Booth, A, 1954, Durrington Walls, Wiltshire:
recent excavations at a ceremonial site of the early second millennium BC. Antiquaries Journal, 34, 155-177.
Wainwright, G J,  1970, The excavation of prehistoric and Romano-British settlements near Durrington Walls, Wiltshire, 1970. Wiltshire Archaeological and Natural History Magazine, 66, 76-128
Wainwright, G. J. Donaldson, P. Longworth, I. H. & Swan, V.  1971.  The excavation of prehistoric and Romano-British settlements near Durrington Walls, Wiltshire, 1970.  Wiltshire Archaeological and Natural History Magazine, 66, 76-128.
Graham, A & Newman, C, 1993, Recent excavations of Iron Age and Romano-British enclosures in the Avon Valley. Wiltshire Archaeological and Natural History Magazine, 86, 8-57
McKinley, J L, 1999, Further excavations of an Iron Age and Romano-British enclosed settlement at Figheldean, near Netheravon. Wiltshire Archaeological and Natural History Magazine, 92, 7-32
building at Butterfield Down.
Rawlings, M, & Fitzpatrick, A P, 1996, Prehistoric sites and a Romano-British settlement at Butterfield Down, Amesbury. Wiltshire Archaeological and Natural History Magazine, 89, 1-43 Richard Atkinson, Stonehenge, Penguin Books, 1956, et rééditions
Cleal, R. M. J., Walker, K. E. & Montague, R., Stonehenge in its landscape (English Heritage, London, 1995)
Cunliffe, B. & Renfrew, C. Science and Stonehenge (Proceedings of the British Academy - 92, Oxford University Press 1997)
Johnson, A. Solving Stonehenge, Thames & Hudson 2008.
Stonehenge, pèlerinage européen au IIIe millénaire, Archéologia n°460, nov. 2008.

Sites internet :


  1. Nice Idea!

    Sadly, were did they get all the 'in fill' in the first place and where has it all gone now? Your are looking at 250K sq feet of chalk rubble - did they move it to silbury hill??


    1. Thank you for showing interest to my explanations. There is a lot of tumulus which have partially or completely disappeared, because of need for earth. About the stones which filled the void, I think it is not necessary to calculate the amount of buildings which have been dismantled on the planet during past centuries... Only some of the big stones were left because they were too big, even if some of them have gone, as you can see on the site, which made some people think that the circle was not complete ! Yours faithfully.

    2. Sadly you missed my point Didler.

      The challk that built Silbury Hill came from the surrounding base as we can see from the topology - minimum effort. There is not the same at Stonehenge - so to prove your hypothesis you need to show:

      Where the chalk came from? and What happened to it after?

      The fact that the ditch is still showing indicates it was removed rather than natural spreading (as the ditch would be under metres of chalk - which excavations how it is not (see Hawley's excavation notes) Cleal et al 1991. Stonehenge in its landscape.