These geometric studies were my re-imagining of some of the prodiguous work of Alexander Thom and Aubrey Burl. My objectives were structure, simplicity and synergy. Some of the following surveys were taken from Alexander Thom's book, Megalithic Sites of Britain (Clarendon Press, 1971) and from Aubrey Burl's book The Stone Circles of Britain, Ireland and Brittany (Yale University Press, 2000).
Circle ▼
Seven Gray Stanes Circle NX 919 739
Seven Gray Stanes would appear to be a fairly typical 12 stone diametrically opposed setting and it has been modelled using the same 8 stones, as both a Circle and an Ellipse (ellipticity 0.956). The major stones A-F and C-I are aligned N-S and E-W, however the axes of the ellipse do not show any particular alignments with the stones. The difference between the models is slight and I would suggest that the circle represents the more likely design. This gives us an approximate rule of thumb for separating circles and ellipses. Without any other supporting evidence of alignment, an Ellipticity of less than 0.91 may suggest an Ellipse, anything greater perhaps a Circle. Survey courtesy of Fred R. Coles, Proceedings of the Society of Antiquaries of Scotland, 1895.
Hoarstones Stone Circle SO 324 999
Hoarstones gives an initial impression of being a tri-flattened circle or an egg (arcs A - D and E - K). That was how it was modelled. Stones H and I were ignored as possible portal stones. The most important feature with all these designs is the Primary Circle and the NRest survey told a different story. The primary circle was modelled on the stones in the arc L - A, however it now also provides a good match on the substantial long arc of F - C. This includes stone I previously discounted. Stone J is now a misleading interior stone. This may have originally been a circle, which was later extended.
Castlerigg NY 293 236
Having surveyed the stones of the Flattened Stone Circle at Castlerigg, NRest delivers the Best Fit Circles for the Primary Circle (Rmse % 2.69) and Short Arc (Rmse % 0.93). While the Primary Circle tallies fairly closely with Alexander Thom's analysis, the centre for the Short Arc is not now on the circumference, but further inside. Getting objective assessments of short arcs is only possible with the modern optimisation algorithm in NRest. There is an amplification effect whereby a small change in curvature leads to a large change in the radius and position of the centre, especially where there are few nodes. Looking at the problem from the other direction, you can choose many centres which will seem to give a reasonable match on the circumference. Note that the blue centre is underneath the red axis in these Flattened and Egg-shaped models.
Class 1
Gamelands NY 640 082
Here is an analysis of Gamelands Stone Circle both as an Ellipse and as a Circle using the same survey data. Geometrically Gamelands makes a poor circle, but a reasonable ellipse. I normally like to include the centre position of all stones in the survey as this avoids introducing selection bias. Here there is a distinct flat section, which may be due to later human interference - these smallish stones have not been included in the Best Fit process. There are plenty of other stones in the survey. I have included two known missing stone positions. Ellipticity can be expressed as the length of the Minor Axis divided by the length of the Major Axis. For a perfect Circle this value is 1. Gamelands Ellipse has an Ellipticity of 0.881. Ellipticity determines the shape of the Ellipse, regardless of scale and allows us to group the Class 1 Ellipses by their similar ellipticities. The flattened section aligned with the minor axis of the ellipse may be deliberate, as a similar feature is observed at Long Meg.
Long Meg & her Daughters NY 571 372
I also examined the elliptical nature of Long Meg (ellipticity 0.893, error 1.93%). The flat Northern section may have compromised the shape of the ellipse, where the circle abuts a known ancient enclosure (but see above). It is interesting to note that the ellipse has a pretty good East / West alignment (compass 89.46°).
Some Stone circles are quite clearly circular as at Swinside. Even for a Circle, a Best Fit Ellipse will give a lower RMSE % error than a Best Fit Circle. Does this mean everything was designed as an Ellipse? No, of course not. If the Ellipticity is close to 1, then the original design would have been a Circle. The Best Fit Circle would be the best model of what was originally intended. It is important never to confuse the model with the original design, it is only our best guess and the numbers have to be correctly interpreted.
Twelve Apostles NX 947 794
The Twelve Apostles is a Cumbrian stone circle accidentally located on the Scottish side of the Solway Firth. When modelled, its ellipticity is 0.883 - very similar to both Long Meg and Gamelands in Cumbria. In addition, the axes of the ellipse are N-S / E-W on the page. The missing large stone at L, noted on the 1789 survey, is on this E-W alignment. Long Meg has a distinct E-W / N-S alignment. The stones A and G match the other direction marked on the plan (Magnetic North?) fairly accurately. 3 pairs of stone are diametrically opposed, but not quite centred on the centre of the ellipse. Stones A and G when considered as a pair, don't align with either. Survey courtesy of Fred R. Coles, Proceedings of the Society of Antiquaries of Scotland, 1894. This would seem to be a 10 stone setting. Stones C and H are supernumerary. It is the stones in position that count (even the missing ones).
Glassonby NY 573 394
The stones of Glassonby Stone Circle are arranged contiguously on top of a tumulus. The axes of the ellipse are not NS / EW and they seem to intersect the circle on particular stones. In addition, the 3 intersected stones are not contiguous with other stones. It is difficult to know if this is significant. We don't know anything about the missing stones. The description suggests the stones were set in cobbles and not in sockets, so it is unlikely that the matter will be resolved. Survey courtesy of W.G. Collingwood, CWAAS Transactions, 1901.
Studfold Gate NY 040 223
Studfold Gate Circle is a bit sparse in the West, but the ellipticity suggests that it is a Class 1 Ellipse. It is not possible to tell from the survey if the ellipse is Flattened on the Short Axes. If the positions of the missing stones could be verified, this might be established. There are too few stones in the survey, to have a high level of confidence in the results. Survey courtesy of J.R. Mason & H. Valentine, CWAAS Transactions, 1925.
Devil's Quoit SP 411 047
Devil's Quoit makes a good Class 1 Ellipse and has 2 opposed entrances - not aligned with the axes. It takes the Henge form and also contains a Stone Circle. Blurring the boundaries, it could also be considered to be a particularly regular Clamshell Henge, I suppose. Time for a real Archaeologist. An interesting variation of the multi-phase sites discussed below.
Waun Gunllwch, south of Builth Wells, would appear to have 2 elliptical sites. The northern ellipse is truncated on the southern side by a gully. There is a distinct hollow on the southern boundary. Both sites are aligned roughly N-S, but diverge on different sides of grid North.
Coed Caeau SO 059 413
Coed Caeau is described as a Medieval Ringwork. At some time, the Eastern side outer and inner banks have been massively built up as a defensive measure (Early Iron Age or later?). This distorts the shape of the outer bank in the East. Hence this NRest survey tracks the more reliable ditch. This results in a Class 1 Ellipse which measures approx. 61.7m by 53m in the ditch, probably dating back to the Late Neolithic. The second ellipse is too ill defined to model accurately.
Adhoc Ellipses
Edin’s Hall NT 772 603
Edin's Hall is a complex multi-generational site. My English interpretation prompted by an analysis using a Best Fit Ellipse (ellipticity 0.562, error 0.63%), starts in the Late Neolithic / Bronze Age with a simple elliptical bank. This is highly accurate and maps very well onto most of the site, even beyond the plotted points in the West. This was followed by a heavy Early Iron Age Multivallate upgrade presumably during a time of strife. A Late Iron Age settlement was established in and beyond the Eastern end on the site, dismantling the protective banks there. The Broch in the Western section is thought to have been built as a defensive measure after the arrival of the Romans. It is too wide to be 'normal' broch, but otherwise clearly has broch-like features.
Habchester Fort NT 944 588
Habchester, while looking like an Early Iron Age hill fort, nevertheless appears elliptical. More then half is ploughed out, but you can still see the shadow of inner bank in the North. I am modelling the ditch here. In the ditch, the Ellipse measures approx. 138m by 103m. The entrance in the East and the outer bank to the South are probably Iron Age additions to the earlier structure. They may form extra protection from the slightly higher ground in that direction. It should be noted that we don't have enough well-defined points around the circumference to be entirely confident in this survey.
Meldon Burn NT 214 429
At the seemingly Iron Age Meldon Burn site, I noticed that the western bank made a poor Ellipse. Looking closer, I spotted that the stone footings outside the bank made a better Ellipse, so I modelled them as such. This proved to be very accurate and was also orientated N / S on the plan. Moreover the Ellipse is carefully centred on the scooped platform with even external spacing on the NW / NE / SE quadrants. The western bank looks similar to the other Iron Age banks and is unsurprisingly ill-fitted to the Ellipse. The Ellipse belongs to a much earlier period, possibly Late Neolithic. The N /S orientation was popular at the time. The similarities with the Edin's Hall Ellipse are strong, both are situated on a scooped platforms stretched along the side of a hill. Note also the similarities with Druid Stones near Patterdale.
Meldon Burn South NT 215 425
On the southern trackway from Meldon Burn, there is another related site, Meldon Burn South. Inside the site, there may be a 20m Iron Age hut circle. It sits inside what might be an earlier Egg shaped enclosure. However it is possible that this is an accidental Iron Age Egg.
Druid Stones NY 398 117
At Druid Stones, I plotted points on the circumference and generated a Best Fit Ellipse (in blue) which worked out well (ellipticity 0.68, error 0.83%, orientation N/S)). The site has undergone significant modification in the Iron Age. Originally I suspected that the Elliptical enclosure was re-enclosed at a later date, by the addition of a new straight wall (in red). Now I think this is a Flattened Ellipse. Note this image was constructed from Lidar images supplied by Joe Ridley using his Gradient software, but not of the usual hillshade type. See Mayburgh Riverine
Park Beck NY 449 025
Park Beck has an avenue leading from the centre towards the North East. This proved to be in alignment with the Summer Solstice sunrise, suggesting a Late Neolithic / Bronze Age origin for the site. When a Best Fit ellipse was calculated on the regular part of the site (ellipticity 0.756, error 1.28%), there was a surprisingly accurate alignment of the short axis with the avenue and the entrance to the NE. Like at Druid Stones, the interior street furniture is Late Iron Age Curvilinear.
Millriggs NY 461 025
Millriggs (ellipse 35.6m x 27.0m) has multiple layers of occupation. The Iron Age layer was still occupied in the Roman Period. This might explain the non-circular building and the 2nd C Armlet found there. The SW side of the site was clearly destroyed by the new field system encroaching on the site. It is interesting that the Ellipticity of Millriggs is identical to that at Park Beck. They are exactly the same sort of Ellipse, although on a different scale.
Robin Green NY 521 027
Robin Green (ellipse 57.0m x 44.8m) has multiple layers of occupation. The Late Iron Age layer spreads out over the Eastern edge of the ellipse. It is interesting that the Ellipticity of Robin Green is similar to Millriggs and Park Beck and the Orientation close to that at Millriggs.
Combe Hill, East Sussex would appear to be an Elliptical Causewayed Enclosure. Causewayed Enclosures date from around 3800 - 3500BC and in Southern England where they are more numerous, they are usually irregular in shape. However some appear to take on more regular Circular and Elliptical shapes. These might be late forms which foreshadow the appearance of Circular and Elliptical Stone Circles and Henges.
Ewesley Enclosure NZ 059 927
Ewesley would appear to be a Causewayed Ellipse (C.E.). These are supposedly rare this far North. The railway line has damaged the site. The primary Enclosure is surrounded by 2 further enclosures (completed by the stream in the South), which also suggests that this might be a C.E.. Ewesley is very symmetrical about the long axis. The Ellipse has been Truncated on the long axis with a circular arc. Cumbrian Ellipses are Flattened on the short axis, so I use the word Truncated here.
Tollis Hill NT 516 580
Another Northern Enclosure at Tollis Hill makes a good Ellipse and has 5 well defined entrances which line up on ditch and bank. The ditch is on the outside of the bank. At Ewesley the ditch is on the inside. It is not clear if this is a significant difference. Henges tend to favour distinct inner ditches, but outer ditches are also found in conjunction. The outer ditches of Henges may have been dug to provide extra material for the bank. Inner ditches of Henges can have Stone or Timber settings on the inner side of the inner ditch.
Allan Water Egg-shaped Circle NT 470 062
Burgh Hill Truncated Ellipse NT 470 062
According to Alexander Thom, Allan Water belongs to the class of Egg-shaped Circles. I have tried to replicate this in NRest. Burgh Hill is the alternative name for Allan Water. However the RCAHM survey differs and is perhaps less convincing as an Egg. It is modelled here as a Truncated Ellipse. So which is correct? I don't think we can be sure without better evidence. Alexander Thom lists 10 Egg-shaped Circles. On a quick examination, others are also open to different interpretations and modelling. Not particularly coherent as a class. More work needed.
Moel Ty Uchaf Tri Flattened Circle SJ 056 372
At first glance from Alexander Thom's survey, Moel Ty Uchaf would appear to have an unusual 5 fold symmetry and perhaps 2 entrances on either side of the main horizontal axis. The stones on this site are almost all in their original positions. The NRest analysis reveals that the site is a variation of the Flattened Circle with 3 symmetrical flattened sections - a Tri Flattened Circle. This accounts for the axes all meeting in the centre. The purple axis is aligned Due North and the axis of symmetry in red is aligned with an Outlier according to the original plan (or vice versa). This relatively pristine site allows us to analyse what the early surveyors were truly capable of.
Avebury Tri Flattened Henge SU 103 700
Nearly half of Avebury Henge is built on a circular plan. A Best Fit Circle in blue (Rmse % 0.91) is calculated from points on the outside of the ditch. My conjecture was that the ditch would be dug on the inside of the surveyed circle and the bank on the outside. My view of Avebury changed after discovering Moel Ty Uchaf Tri Flattened Circle. Avebury also fits this pattern well, although the NW corner has been slightly rounded off. The fourth flattened side at Avebury is unmodelled as it is overlain by the village. However if you look at the inside of the ditch there, it is also clearly flattened. I have estimated the position of the fourth arc (in orange) by using the intercepts and a third estimated point, to give a better idea of main axis of symmetry. Unexpectedly, the Green axis aligns with the southern entrance and a line drawn (in red) between the Eastern and Western entrances passes through the centre of the Best Fit Circle.
Avebury Class 3 Henge
Mock-up using the eastern circular part Avebury to recreate the western part of the Class 3 Henge. The surveyed lines are as before. Note that from the Class 3 Henge, we have nearly half of the perimeter and 3 out of 4 of the entrances in their original positions. The northern entrance has moved fractionally to the west, probably at the time of the transition. The WIKI entry on Avebury includes the following interesting note:
The top of the bank is irregular, something archaeologist Caroline Malone suggested was because of the irregular nature of the work undertaken by excavators working on the adjacent sectors of the ditch. Later archaeologists such as Aaron Watson, Mark Gillings and Joshua Pollard have, however, suggested that this was an original Neolithic feature of the henge's architecture.
Watson, Aaron, A. (2001). "Composing Avebury". World Archaeology. 33 (2): 296-314
Gillings, Mark; Pollard, Joshua (2004). "Avebury". London: Gerald Duckworth & Co
Stotfold Class 3 Henge? TL 217 365
Examining Professor John Blair's plan of Late Anglo Saxon Stotfold, I was struck by the simularities with Avebury. I modelled Stotfold as if it was a Class 3 Henge (diameter 730m). The main axes are located roughly EW and NS as is common. The strong EW axis is defined by 2 roads passing through the Henge and passes very near to the centre of the circle. The NS axis is perpendicular to the EW axis, but is offset slightly to the East of the centre. In the Southern part it aligns with the Anglo Saxon track illustrated in the reconstruction by Albion Archaeology. In the North it joins a path which eventually passes through the centre of Manor Farm. There is a fifth, possibly later, entrance in the NW which again has a junction just outside the Henge. The beck inside the Henge to the West, running down from the Methodist Chapel and later diverted might be the course of an inner ditch. The Geometry must be considered circumstantial unless it can be backed up by hard evidence.
Maiden Castle NY 4510 2435
Maiden Castle in Cumbria also follows the Tri Flattened Circle model. This analysis suggests the possibility that Maiden Castle may have an early origin. The inner edge of the ditch has been surveyed as it offered the best image. The axes on the 2 side arcs, after passing close to the centre of the circle, then carry on to pass close to the near intercepts on the primary circle. The surveying seems to have closely followed the Kerry Hill approach. Modelling accuracy on the primary circle using the 2 far intercepts, remains within acceptable limits at 1.54% RMSE.
Kerry Hill SO 158 861
The CPAT Regional Historic Environment Record for Kerry Hill says:
...the stones lie on the inner edge of (or just inside) a low circular bank. If the circle is prehistoric then it is an Embanked Stone Circle (and locally quite rare), if however it is modern then it could still be a reuse of a prehistoric circular enclosure.
The similarities with Grange Stone Circle below are striking. Lines are drawn through 8 diametrically opposed stones, pass through the centre of the circle. The 9th stone, D, is accurately on the primary circle. The circle is Flattened on 2 sides and each has a central stone at F and J. The arc of the primary circle arc is also tidily divided into 2 by stones at B and C. This is clearly a very organised Tri Flattened Circle design. It may even represent the early stage of a site that was never completed and as such gives us an invaluable insight. Given the similarities with Grange Stone Circle and as no-one would have known about Tri Flattened Circles, this suggests the site is genuine.
Chatton Sandyford Ring Cairn NU 100 267
Chatton Sandyford ring cairn was excavated by George Jobey (Archaeologia Aeliana, Vol. 46). It is interesting that the section on the Primary Circle has been built in several straight line sections, but the other 3 arcs are all nicely curved in the Tri Flattened Circle manner and not further flattened. This suggests some deliberation, rather than that the sections were built ad hoc. The axes of symmetry of the 3 arcs pass through the centre of the primary circle. The construction of the outer facing is similar to that of Breach Farm Barrow, small sections being flattened in places.
Breach Farm Barrow SS 97 74
Breach Farm Barrow was excavated by W.F.Grimes (Cardiff Naturalists' Society, 1936). Note North is not vertical on this plan. The Northern quadrant is not flattened. The Eastern quadrant has some small flattened sections. I had expected the circle to be pushed further out on the Western and Southern sections leading to more extensive flattening of the primary circle on those sides. However the computer knew better. This led to a small prominence on the Western quadrant (albeit leading to flattening of a sort on either side of the prominence). The Southern quadrant ended up tidier than I expected with one distinct flattened section. The diameter of the barrow is stated as 81 feet, but this does not seem to tally with the scales on the main plan or cross-section plans. From the main plan I made it 106.6 feet.
Casterton Stone Circle SD 639 800
The Stone Circle at Casterton has an unusual 20 stone setting, sat upon the inside of a circular bank (as shown on the SE quadrant). The stones are arranged in diametric opposition and there are stones on both the North / South and East /West alignments. There are 7 of these pairs shown on the plan, the remaining 3 pairs being compromised by displaced stones on the southern side. 4 of the pairs align with the exact centre of the circle and 4 of which align with a centre 1m to the West of the true centre (1 pair aligns with both centres).
Grange Embanked Stone Circle
Grange Stone Circle is a 46m diameter Embanked Stone Circle near Limerick (S. P. O'Riordain, 1951. "Lough Gur Excavations: The Great Stone Circle (B) in Grange Townland". Proceedings of the Royal Irish Academy. 54: 37-74). It has 10 diametrically opposed stones and a central posthole. The gaps between the orthostats have been filled in by surface-lain smaller stones. Modelling the Circle proved the accuracy of the Late Neolithic surveying around the central posthole. The excavation of this pristine site revealed the careful organisation which went into its construction. Unusually, the NE / SW quadrants appear to be on a 12 stone setting, while the NW / SE quadrants appear to be on an 8 stone setting.
At Balnuaran of Clava from Excavations in Passage-Graves and Ring-Cairns of the Clava Group, 1952-3, Stuart Piggott, Proceedings of the Society of Antiquaries of Scotland.
Clava SW NH 757 444
Clava SW is a Passage Grave within a Circle. The stones in the East have been compromised or lost. The positioning of the remaining stones is instructive. When the Diametrically Opposed stones are plotted (in red), they don't intersect at the centre of the circle (in blue). The circle must have been surveyed accurately first and then when it came to position the stones, a different centre was used in error, I think. Might this error have occurred due to the Stone Circle being awkwardly constructed after the Cairn? The remaining Diametrically Opposed lines (in green), were plotted using the good positions in the West and the new centre. The Eastern stone is almost in the right position. As with all good Passage Graves, the passage is orientated on the Winter Solstice sunset.
Clava NE NH 758 445
Clava NE is a Passage Grave within a Class 1 Ellipse. However without more extensive evidence, the shape might be coincidental. Again the passage is orientated on the Winter Solstice sunset and we have diametrically opposed stones. While the centre of the ellipse is not aligned with the passage, the centre used for the diametrically opposed stones is aligned with the passage and one of the stones in the ellipse. The axes of the Ellipse are in fairly good alignment with 3 of the stones.
Clava RC NH 757 445
Clava RC is a Ring Cairn within a Circle. Again we have diametrically opposed stones. The centre of the circle is not aligned with the centre used for the diametrically opposed stones. One pair of diametrically opposed stones is aligned with the centre of the circle.
Stones of Stenness HY 307 125
It was certainly intended that the Late Neolithic Class 1 Henge at Stenness would have 12 stones. The missing 9th and 12th sockets are not particularly well-placed on the plan. This is a very carefully surveyed setting with 12 evenly spaced diametrically opposed stones. The spacing has been done with more care than at Grange. The number of stones in these diametrically opposed sites is typically 12 or 8. If you count the entrance and opposing stones as single points, Grange is actually a hybrid with 2 quadrants being based on the 12 stones layout and the other 2 on 8 stones. The Sumerians were also fond of the number 12 as it divides into 2, 3, 4, 6 or 12 parts. That's why we have a 360 degree circle, 24 hour clock and 60 minute hours etc. Neolithic farmers are usually reckoned to have come from the Near East, so did they bring their numeric skills with them when they came?
Broomrigg Henge NY 553 467
Broomrigg Henge looks to be a reasonable Class 1 Henge. It is 167ft wide. The inner ditch is 8ft wide and the outer bank is also 8ft wide. It is part of the extensive Broomrigg Plantation prehistoric landscape. Survey courtesy of K.S. Hodgson, CWAAS Transactions, 1935.
Ring of Brodgar HY 295 134
From this early survey of the the Class 2 Henge at Ring of Brodgar, the purpose of the labour intensive rock cut ditches in Orkney becomes clear. The North East ditch is full of water. On this island surrounded on all sides by water, what is more natural that the inner space is separated from the outer space than by water. Plan of Ring of Brodgar with notes and measurements, surveyed by H.Dryden and G.Petrie in 1851, copied by W.Galloway in 1868.
Broadlea Henge NY 219 747
An aerial photo of Broadlea Henge by J.K.St.Joseph, 'Air reconnaissance of North Britain', J Roman Stud, vol. 41, 1951. The outline of an earlier underlying circular henge (Class 2?) is clearly seen as a fainter shadow in the cropmarks (along with a later Roman fort).
Coupland Henge and Droveway NT 941 331
Material from the western side of the Droveway ditch dates to approx. 3800BC making it much older than the Henges. The projection of the western side of the Droveway across the site in green neatly joins the unmodified North and South sides of the route as if the Henges did not exist. As does the corresponding eastern projection in orange.
A Class 2 Henge at Coupland (in red) followed which was centred on the Droveway and whose entrances were aligned with the Droveway. Note only the western Clamshell side of the southern entrance is not on the circular alignment of this original Henge. The 3 remaining bits are on the circle. The Class 2 Henge is probably a similar age and size to the Class 2 Henges to the South.
The Class 2 Henge was subsequently remodelled as a Clamshell Henge, leaving most of the Henge in its original Class 2 form. The other Henges in the Milfield complex are new build Clamshell Henges, but are tellingly built on a smaller scale and each half not particularly on a circular plan. Neither did they sit on the Droveway, but some were built next to it. I would imagine that Coupland Phase 2 Henge was remodelled to fit in with these surrounding Clamshell Henges. The Droveway was narrowed to fit in with the new Clamshell design.Cairnpapple Henge NS 987 717
Stuart Piggott has carefully documented the developmental layers at Cairnpapple, where there is perhaps over a 1000 years of continuous use. For this, the Lifestyling approach incorporating both horizontal and vertical components gives a more direct visual insight.
Modelling Cairnpapple as a single circle doesn't work well, so I modelled the 2 arcs separately. The timber circle makes a reasonable egg shape. The North and South ends of the western arc were compromised by the construction of a modern octagonal enclosure wall. This has been partially masked by extra grey shading in the West.
A further 2 Borders Clamshell henges at Weston and Normangill should be noted. Also the Clamshell Henge at Milfield North Henge which is shown as having a timber post setting - image can be seen online at the Maelmin website.
Weston Henge NT 030 458
Normangill Henge NS 973 221
A common feature of a number of different classes of monument is the Pronounced Inner Ditch and consequent Platform. Another frequently associated feature is the Timber Post / Stone settings around the inner edge of the Inner Ditch. These features seem to transcend time and space throughout Neolithic and Bronze Age Britain.
King Arthur's Round Table NY 523 284
King Arthur's Round Table as a Clamshell Henge, might naturally be expected to conform internally with either Arbor Low or Cairnpapple. Unfortunately the initial approach adopted by R.G.Collingwood left the results open to criticism by Gerhard Bersu, when Collingwood was unable to complete his work. The jury is still out on this. It is entirely possible that King Arthur's Round Table had a timber circle around the edge of the inner platform like North Mains.
Other possible Clamshell Henges:
Fargo Plantation, Wiltshire SU 112 428
Arbor Low Henge SK 160 636
Harold St.George Gray attempts to model Arbor Low with a circle, but with no better success than I could manage at Cairnpapple. The western arc should certainly be modelled flatter.
Bull Ring SK 079 782
The nearby Bull Ring Henge is also thought to have a stone setting on the inner platform. The only difference would seem to be that the Bull Ring is a Class 2 Henge. At Arbor Low and Avebury, the outer stone circles follow the irregular outer side of the platform.
North Mains Henge NN 929 163
North Mains Henge (Strathallan) may contribute to the interpretation of these multi-period sites. The larger timber circle at North Mains is essentially circular with some minor flattening in the NNW area, whilst the smaller inner timber circle makes a good egg shape. Consequently the surrounding henge is also nicely circular. The 2 entrances are also diametrically opposed - a definite Class 2 Henge.
Stanydale Neolithic "Temple" HU 285 502
Stanydale Neolithic Temple is one of my favourite sites. Stanydale belongs to the Heel-shaped class of monuments, having the distinctive concave facade at the entrance. This it sometimes shares with its close relation, the Chambered / Court Cairn. At first, I thought the right-hand side was more carefully surveyed. However once modelled, it became clear that the bulge on the left lay on the modelled circle and it now became less clear, We seem to have two slightly different designs on each side of the line of symmetry. Image Historic Environment Scotland
Nearby Neolithic House on Shetland
This Neolithic near-Longhouse is remarkably symmetric. Other nearby examples are not so tidy in design, so this is perhaps more a matter of chance than design. The classic Longhouse form resurfaced for much of the Medieval period. Only dying out when the last Blackhouse at Arnol on Lewis was abandoned in the 20th century. Report on the Discovery of Numerous Stone Age House-sites in Shetland. By Charles S.T. CALDER, A.R.I.A.S., F.S.A.Scot., Proceedings of The Society of Antiquaries of Scotland.