Archaeoastronomy:

A Survey of Prehistoric Sites in South Wales

by Martin J Powell

- Book Retailer, Cardiff

[Note: This is a preliminary work; a full, revised version of this article will appear here at a later date]

Introduction

The following study is the result of a survey carried out by the author of prehistoric sites in the South Wales area of the United Kingdom. The monuments in question are generally considered to have served a funerary and/or ritual function and were constructed during the Neolithic (New Stone Age) and Bronze Age periods in Britain (ca. 3800 to 1200 BC).

There is considerable evidence from around the world that prehistoric man held the heavens in great regard, and that part of his ritual and belief system involved orientating his monuments towards significant rising or setting positions of the Sun, Moon or stars along the horizon. Such orientations might also have served a secondary function as a primitive calendar.

The study of prehistoric man's apparent fascination with the heavens, and the manner in which he orientated his monuments towards the celestial bodies, whether for ritual or secular purposes, is known as archaeo-astronomy (sometimes referred to as astro-archaeology).

The sites in the survey have previously received scant attention from an astronomical viewpoint. It was with this in mind that the author set out, between 1988 and 1992, to create a database of results from as many sites as was possible, which could be studied either in isolation or in comparison with sites from other regions of Britain.

Survey Area

The survey area covers the former counties of Glamorgan, Brecknockshire and Monmouthshire. In one case a site from the neighbouring county of Herefordshire is included due to its evident relationship with similar sites in Brecknockshire.

The counties are referred to by their pre-1974 names. This allows for easy cross-referencing with the site Inventories published throughout the 20^th century by the Royal Commission on Ancient and Historical Monuments in Wales. The counties have since undergone revision on a couple of occasions for administrative purposes.

All sites under study lie between geographical latitudes 51º.44 and 52º.08 North.

Selection Criteria

Clearly it would be preferable to include every prehistoric site within the database. This is not practicable however since many sites are in a ruinous condition and many are unsuitable for astronomical consideration. The majority of prehistoric monuments in South Wales are the hundreds of round cairns and round barrows which occupy the uplands of the Rhondda Valleys. Most of these sites are ruinous, having been robbed of stone over the centuries for the building of field walls, and others no doubt have been completely destroyed.

For the sites to be considered in an astronomical context it is necessary to define a set of criteria which must be satisfied. Essentially, a site must contain a feature (or features) which indicate a direction in which to face.

The clearest case in question is that of the Neolithic long cairns or long barrows, which by their very nature define a clear orientation. Where a cairn contains chambers set at different directions to the main axis, these are also considered individually. Where the cairn has disappeared, the orientation of the chamber itself is considered.

Round cairns and round barrows contain less well-defined orientations. The features considered astronomically in these cases are varied.

In the majority of cases the central cist (small chamber) will have a clear orientation and this line will be studied (eg. Carn Llechart, Mynydd Drumau). Other sites, particularly ring cairns and enclosure-type sites, have entrances through the ring and it is the direction of these openings, as seen from the centre of the cairn, which will be considered (eg. Great Carn I and II).

There has been some contentious research which suggests that prehistoric man may have deliberately designed some rings and stone circles not in a true circle, but in an ellipse, or an egg shape. The counter-argument is that these are simply the result of poor construction, or that they are the result of accommodating a nearby track-way or a ground slope. Where such sites have been identified, the major axis of the monument has been considered in an astronomical context (eg. Carn Caca).

A couple of round cairns have an outlying stone a short distance away which is thought to be contemporary with the cairn and a deliberate part of it (eg. Rhosili Down NW and SE). In such cases the direction of the outlier from the centre of the cairn is considered.

One site contained two stones within a ring which had been marked with cup-shaped depressions (known as cup-marks) which clearly gave them a greater significance than the other stones in the ring (Crick Barrow). Again, the direction of these stones as seen from the centre of the circle is considered.

The final category of alignment is that of two or more standing stones. These are usually studied in both directions, unless on-site evidence suggests that there was a preferred direction.

Method of Assessment

The majority of sites listed in the table were visited by the author and the orientation of visible features measured using a prismatic magnetic compass.  The accuracy of the compass is ± 0º.5, which is well within the limit of accuracy that can be derived from the monuments themselves.  Unless fully excavated and with an accurately orientated plan available, there is clearly a limit to the accuracy within which a site's orientation can be recovered.  In the case of long cairns and barrows this is usually ± 3º, less for well seperated standing stones.

The upright stones of a site are often too large or irregularly shaped to define a clear sight-line.  In chambers or cists the stones often diverge in such a manner that the intended direction is unclear.  In these cases the average direction of the slabs is taken to represent the intended line.

Checks on compass accuracy were taken at most of the sites by taking bearings to features in the surrounding landscape which could later be identified on an Ordnance Survey map.  On a number of occasions the sites showed no features above ground (eg. they had been re-buried after excavation) and in these cases the excavation plan was referred to.  This was also necessary for a few cases where the site had been destroyed after excavation.  A problem here is the accuracy with which the excavator set the North point on the plan.  In some plans it is also unclear whether the North point refers to Magnetic or True North. Without additional information these details cannot be verified, and wherever there is doubt this is indicated in the tables.

The angular elevation of the horizon was in all cases determined from 1:25,000 scale Ordnance Survey maps.  There is clearly a limit to the accuracy of the value obtained using this method.  Generally, the closer the horizon, the greater the error is likely to be.  In most cases, however, the magnitude of uncertainty in the site's orientation far exceeds the precision with which the horizon elevation can be determined.

The assessment of the lines themselves were determined by a computer program, written by the author in the QBasic language.  The results given in the tables are taken directly from the program's output.

Other Considerations

When examing the astronomical lines shown in the tables it is worth bearing in mind a number of important factors.

The alignments have been selected on the basis of their astronomical potential, based partly on evidence from other sites where features appear to have been intentionally orientated towards astronomical targets (eg. at Stonehenge, Newgrange and similarly renowned monuments).  The fact that many of the alignments in South Wales appear to indicate astronomical targets does not in itself prove that this was the intention of the monument's builders.  Indeed, some astronomical lines may be purely the result of chance.  The monuments may have derived their orientations for other reasons.  The local environment may have dictated the orientation of a tomb, for example, if it was constructed in a narrow valley, was surrounded by dense forest or was close to a stream.  They may have been orientated in the direction from which their ancestors came.  A recent study suggests that prominent landscape features, such as mountains, might have served as focal points for the monuments (Tilley, 1994, pgs. 123-136).

Even if astronomy was built into the monuments, practises are likely to have varied between regions.  Some populations may have taken great efforts to align a monument on the sunrise on a particular day in the year, whilst others may have been quite content to orientate them in the general direction of the sunrise, with no implied precision.  Archaeo-astronomy aims to discover the most likely solution to this problem by studying the monuments in large numbers, thereby providing a statistical argument one way or another.

Another factor is the current surface appearance of the monuments.  Several millennia of soil drift and erosion, stone robbing and plant invasion have no doubt altered many of the monuments considerably.  Breaks in an embankment which today appear to have been deliberate entrances through to a monument may not in fact be original features.  Only a complete excavation of the site can reveal its true nature.  Standing stones may only be the remnants of a much larger structure, the direction indicated by the stones today not necesarily having been the intended direction when the complete monument stood.  The azimuths shown in the tables may therefore be incorrect by up to several degrees in the worst cases.  This would no doubt affect the astronomical results, either more favourably or less so.  Nonetheless, the possible existence of astronomical lines, particularly at poorly understood sites such as enclosure monuments, provides data which might help towards a greater understanding of their original function.

In presenting the tables the author would like to extend his appreciation to the Royal Commission on Ancient and Historical Monuments in Wales, whose excellent inventory of Glamorgan was the primary source of information for the duration of the survey.

Copyright Martin J Powell 2001-2; minor revisions Feb 2006

<< Main Index