Stone Circle Surveys

Where to start?

As a physicist and electrical engineer by training and radar engineer by profession, my interests have always tended to the physical sciences, cameras, HiFi, optics etc. My interest in stone circles, alignments etc. stems from seeing a television broadcast in the early 70s when the lunar alignments associated with the Paps of Jura were being described. This grabbed my attention as earlier I had sailed past Jura on a three weeks cruise in the Hebrides, but the interest lay dormant as family and domestic matters then kept me away from Scotland.

Many years later I came across and purchased Thom's 1967 book which set out most of the background relating to megalithic yards, and various astronomical alignments. In 1993 I found his 1971 book with enormous stress on high accuracy, particularly with regard to the sites which he asserted were connected with the 54/55 year lunar cycle. A year later I visited the "Druids' Circle" rings at Penmaenmawr in North Wales and was most impressed by the setting, overlooking Conway Bay with views to the Great Orme; and with the large number of stone circles etc. in such a small area. However the visit raised doubts about both the accuracy of published surveys and the selection of stones whose positions were recorded. I had taken some photographs and was struck when looking at them that there appeared to be significant differences from the published plans.

Burl (p 270) illustrates the point by showing Griffiths 1960 plan side by side with Thom's. Thom's plan of 1966 shows that the large fallen stone points to the east side of the circle centre, whilst Griffiths shows the same fallen stone (see below)

Fallen stone at the "Druid's Circle", Penmaenmawr, Wales

pointing to the west side. I thought that the large fallen stone was far too big to have been moved by vandals between the two surveys. Who was correct? Griffiths' plan clearly shows the bank on which the stone circle is set, whilst Thom ignores this feature completely. Although Thom stated that he had surveyed the stones as found and only occasionally probed to find the edges of buried stones, it wasn't clear why some stones were ignored when others had been selected for the survey. For example, walking round the main stone ring it is apparent that in many places there are more stones than recorded by Thom. To an archaeologist it may be obvious which stones are original and should be measured, but I fail to see how Thom could make this decision whilst working within his published criteria, and his plan only shows the larger stones. In view of Thom's statement that he was recording simply the positions of what was there, I found this puzzling. I began to wonder if there was a degree of selection taking place in what was reported.

I was also puzzled by the ability of early societies to pass on knowledge of rare events happening only once every 54 years, at a time when the average life span was probably less than this and at a period for which written records have yet to be found in this country. Reading archaeological sources it was apparent that there were few accurate independent surveys of stone rings, but that most in the archaeo-astronomy field followed Thom. I decided that perhaps I could add to the fund of knowledge by carrying out surveys, publishing my results and allowing others to select which information they needed for their research. If I was to improve on Thom's results I needed a total station or a theodolite and suitable training.

After a lengthy search I obtained an old Wild T2 (~450) which is classed as a 1" of arc theodolite. At this time (~1996) some used total stations were available but their method of recording the data was primitive and I decided that their considerable extra cost was not worth it. At least I could see what I was writing down in a notebook, although our experience when surveying the circle at Mitchell's Fold showed that even in Spring fingers can get frozen on an exposed site!

My first experiments used the theodolite in a triangulation mode from both ends of an accurately taped baseline, but I quickly realised that this was very time consuming and that I would need to temporarily mark points around the sites so that two sightings could be made from both ends of the measured base line. I was advised by the RCHME (now English Heritage) that even temporary markings are not acceptable for scheduled sites. There was also the problem of measuring to the hidden side of the stone. Other techniques would be necessary.

The earlier surveying textbooks suggested the use of subtense bars or tacheometry using special quality levelling poles and parallel plate micrometers, whilst the more modern extolled the virtues of total stations. I investigated the use of a subtense bar, so that the range to points around stones could be made from a single point, and had excellent results in the garden measuring to about 1" of arc. The bar was mounted at a height of about 1.2m and therefore had the advantage that it was visible over many stones, and ranging to the hidden sides of these stones became possible. However on a windy Welsh hillside, the accuracy was a different matter as the bar, made lightweight for portability, only occasionally remained sensibly in the same place through the series of measurements. (I wrote extra error detecting software to check for this once I realised it was a problem.) Shortly afterwards, I purchased an ex-army subtense bar and dedicated tripod, and whilst an interesting device with its invar rods, provision for illumination at night etc., its weight and wooden trunk case was out of the question for field use.

Realising that my practical knowledge of surveying particularly for archaeology was insufficient, I joined the Oxford University Dept. of Continuing Education's Archaeology Survey Week in June 1997, and discovered the practical delights of total stations! This course was mainly aimed at archaeologists and was designed to give them the skills to carry out basic mapping of a site. (It covered the traditional methods of taping, offsets, plane tables and optical squares, and the importance of researching earlier material. Ground radar, magnetometry etc. wasn't covered.) Whilst on this course I discussed with the lecturers (from English Heritage) whether I was correct in assuming that in many cases accurate surveys of circles, independent of Thom, had still to be done, and they confirmed my suspicions. Following the course I investigated the purchase of both a simple total station, and a Distomat EDM for my T2, but was put off by the then very high price and the need for an expensive prism detail pole target. However a month later, in July 1997 Leica introduced their just affordable Disto Basic utilising a visible red laser beam and different, pulse, technology; and as a result can measure to < natural > targets with normal reflectivity out to 30m to an accuracy of 5mm. Measurements take a few seconds. This removed the need for what was then an expensive prismatic detail pole. (Today's prices of mini-prisms etc. are much more affordable as the result of being mass-produced in China.)

I purchased one of the first batches of Distos (Basic model) into the UK (at ~900), and set about mounting it above the T2's telescope, and accurately parallel to the axis. Mounting the Disto above the T2's telescope obviously prevented me taking the traditional pairs of readings using the opposite faces of the theodolite, and these readings then became subject to any centration errors of the theodolite's vertical circles. A careful set of measurements were taken and these errors were found to be acceptable for my uses. Mounting the Disto in this position also completely obscured the guidance sights on the telescope. These were replaced with a Webley Speedpoint red dot sight designed for the pistol shooting fraternity, and sadly, now discontinued. (It's battery-less and just visible below the T2 telescope.)


T2 with Disto Basic mounted above its telescope. The mounting was devised so as to allow easy mounting and demounting, with the parts stowable in the T2's cylindrical container.

As seen above, the Disto was mounted in a customised cradle above the telescope of the T2. Although the Disto has quite a large lens at the front, the laser beam of the Disto emanates from the side of this lens via its outlet port.


The mounting of the Disto on the theodolite has this laser port as close to the telescope as possible. I wrote software so that errors from the vertical offset, and the distance between the telescope axis and the Disto's measuring point were taken into account. (Total stations normally use the same optical axis for both laser and visual beams.) Results were plotted using Autosketch v2.1. The purchase of the Disto completely revolutionised my methods and accuracy. Careful checks in the garden and using traverse techniques have yielded positional differences to precision targets to within 3mm. As this is within the basic accuracy of the Disto Basic, I considered this satisfactory. In the field, communication with my target pole holder was via walky-talky radio with a 250m range.


Finding an accurate true north remains a problem.


Sun-azimuth readings are difficult at the best of times, and without the benefit of angled vewfinders and preferably a Roeloff prism, I've not had any results which I could remotely trust. In addition, the cradle fitted to the T2 restricts the vertical adjustment of the telescope. (Checks on the T2 before I fitted the cradle indicated that its centration was good enough for this application even if taking sightings on both faces was not possible.) I can obtain accurate time from my GPS receiver. Not recommended for me.

Surveying Gyros

Surveying gyros and associated theodolites appear occasionally on the second-user market but whilst they have the necessary accuracy, they remain expensive and are very heavy, depend on heavy batteries and take quite a time to produce a result (~20-30mins).

Magnetic Compass

The best trough compasses do not claim better than 0.1deg accuracy and experience with my sighting compass which can be read to 0.5deg, indicates that it is rare to find areas free from significant magnetic deviation effects. Even 0.1deg accuracy is hardly sufficient to challenge Thom's results.


Assuming that there are clear unambiguous features visible and that their positions are known to good accuracy, it should be possible to sight onto these features and determine one's position and set out an azimuth reference. The sites I have visited have not had trig. points visible and the mountains have been devoid of church steeples, sharp peaks or cliffs etc. Consequently I have not used this technique. Following the increasing use of GPS by the OS, trig points are being allowed to crumble away or fall into visual obscurity hidden by unrestrained tree cover.


As differential GPS is well outside my price range I have restricted myself to handheld Garmin units. With a long enough base line the use of GPS is possible but as the US DoD reduces the accuracy of the civilian available signal to about 100m, a very long base line is necessary. The Garmin GPS12 incorporates a position averaging feature which in most instances claims an accuracy of about 16m or better. Hence for an accuracy of about 0.1deg or 1/600 part of a radian, a baseline of about 15km is required. In open country and in clear conditions this is feasible, and any longer baseline is a bonus. A clear base line of about 10km range is possible from Mitchell's Fold to the monument clearly visible above Montgomery. Unfortunately when I last visited these sites I was using the less sophisticated Garmin GPS38 which does not have the averaging feature, and the plots will be less accurate as a result. However, in many areas where there are stone circles, there simply aren't landscape features which are easily visible and known to the Ordnance Survey. Here, as before I'm thinking about church towers, OS trig points and pointed mountain peaks. Consequently a need for the marker described in the targets section and designed to be placed some distance away from the survey site, its position recorded by GPS, and then sighted by the survey telescope. The emphasis here was on simplicity, high visibility and low cost - just in case other fell wanderers took a liking to it.

Software - I've written a variety of software programs to record photographic details, calculate bearings between different sites with known grid references, find suitable maps given grid references, produce co-ordinate lists suitable for plotting in Autosketch etc.


Detailed Surveys

Although a number of sites were visited over several years, detailed surveys were made of just two stone circles in Wales at Kerry Hill, Powys (Montgomeryshire) and Cerrig Pryfaid, Gwynedd. The number was limited by me and my partner being able to find the necessary time whilst we were both working, and suitable weather conditions for a decent survey. A carefully planned trip to the Cerrig Duon/Maen Mawr site in the Brecon Beacons at Whit 2000 was a disaster with sustained torrential rain making a mockery of the need for long range sightings for an accurate GPS derived north reference.


Further work - now unlikely

My work in this field declined when I realised that laser cloud-point technology had matured to the point where English Heritage could simply hire an instrument (and operator) plus helicopter, and accurately survey several stone circles in a day's work. The amount of data collected is huge, and software now allows the reults to be turned into 3D virtual models from which any variety of accurate measurmements can be made after leaving the site. Unfortunately this happened at about the time that I retired in 2000, and would then have had the necessary free time! Nine years on and there's now a range of better engineered laser cloud-point data collectors available.

This note is an attempt to collect together my techniques, and few surveys for a wider audience. For a more recent account of Thom's work, I'd recommend Robin Heath's book: Alexander Thom: Cracking the Stone Age Code.


References and Bibliogtraphy

Burl, Aubrey 1976 : The Stone Circles of The British Isles. Yale University Press 0 300 01972 6

Thom, A 1967 : Megalithic Sites in Britain. Oxford 0 19 813148 8

Thom, A 1971 : Megalithic Lunar Observatories. Oxford 0 19 858132 7

Heath, Robin 2007: Alexander Thom: Cracking the Stone Age Code. Blue Stone Press 978 0 9526151 4 9