Synopsis

Introduction.

This continuing study has been running for more than 25 years. Its initial objective was the development of a standard methodology capable of determining the siting criteria of the monuments of the Cork-Kerry Stone Circle Complex, whether or not they might be astronomical. A systems approach was used which in essence means the questioning of basic assumptions, a fresh look at possible relationships and development of structured data collection techniques. The process was iterative and its development has resulted in a large body of evidence indicating that all megalithic monuments were primarily for astronomical, calendrical and eclipse prediction purposes. Further, that within a given geographical area, those sites are usually inter-related in a functional manner. The survey was then progressively expanded to look at megalithic tombs and bronze age barrows, with consistent results.

Method | Discussion | Conclusion | Home

Brief Review of the Academic Position.

The Bronze Age monuments of the Cork-Kerry Stone Circle Complex are known as ritual monuments because, like their relatives in other countries, they have provided few finds when excavated and have no obvious purpose. They have also been commonly regarded as grave markers, and though there have from the earliest times been suggestions of solar and lunar alignments, these have generally been disregarded due to lack of conclusive evidence.

Seven of these stone circles have been excavated and four contained human remains but E.M. Fahy, who dug three of those four, regarded the deposits as dedicatory only. In contrast, the name "boulder-burial" was coined after his excavation at Bohonagh when the discovery of "a small shallow pit containing a few fragments of cremated bone" meant that "The dolmen was established as a tomb" (Fahy 1961). Thirty years later, three more were dug and all covered ritual pits but none showed any evidence of burial. "The absence of human remains from all three examples does call into question their role as monumental grave markers. One possibility is that the burial rite involved a token deposit of burnt bone, however a non-sepulchral function may not be entirely ruled out.", their excavator later wrote (O'Brien 1994,215). Four short stone rows and five standing stone pairs have also been excavated, at least partially, without any significant finds. Various radio-carbon dates now place the complex in the period c.1650 - c.800 BCE (Lynch 1999,11). Evidence from pollen analysis indicates that there was widespread forest clearance during that time and climatological indications are that the weather was warmer, calmer and clearer than the present.

To summarise: The archaeological evidence provides no clear indication of the purpose of these monuments but neither does it contradict a possible astronomical explanation. The evidence from pollen analysis and climatology is more supportive of such an explanation than not.

Professor Alexander Thom could perhaps be regarded as fathering the modern discipline of "archaeoastronomy" with his 1967 book claiming that the axial alignments of monuments were used to indicate horizon features that marked astronomical events. He suggested that there had been a solar calendar, dividing the year into sixteen parts and that the extreme positions of the moon had also been observed, in some cases to very high precision, for the purpose of predicting eclipses at those times. He also claimed that some stars were indicated. Gerald Hawkins was another pioneer, famous for his work on Stonehenge and use of computers. In 1973 he published a book that included a set of criteria for the scientific study of possible alignments.

Considerable interest in the field developed for a time, with variable results. Clive Ruggles became prominent in developing methodology, stressing objectivity and rigour. He went to the trouble of re-assessing Thom's investigations as well as doing his own research. By 1988, we see in "Records in Stone (papers in memory of Alexander Thom)" edited by Ruggles, that he and others had by then shown that the high precision lunar alignments did not exist, though it was still held by some at least that there was independent evidence for a sixteen "month" solar calendar. In a discussion paper in the same work, Ray Norris made the important statement:

The work of Ruggles and others has highlighted the real problem of British archaeoastronomy and shown it to be one of methodology. The nature of the problem is such that different methodologies give different results. It is unlikely that we have yet found the optimum methodology which, although free from subjective bias, yet cannot accidentally discriminate against a type of alignment favoured by the megalith builders.

By 1999 Ruggles had concluded that there was little evidence for a solar calendar though possibly some for low precision lunar alignments, especially on sectors of the horizon beyond the sun's travels. A possible but unclear relationship between stone row orientations and prominent hills was also recognised but his general message was that there is no evidence for the accurate indication of specific events. It can now be shown, however, that the whole concept of investigating indicated alignments was based upon a subjective view of how the monuments might have worked that is incorrect. In other words, the accepted methodological criteria were accidentally designed so that they did discriminate against the type of alignments favoured by the megalith builders.

Home | Top | Back

Method.

While the original motivation was to perform a completely independent investigation into whether or not there was any truth in Thom's theories, this study was deliberately structured to look at monument siting criteria and seek alternative explanations first. An initial period was therefore spent investigating sites with just a notebook and compass, collecting details about the monuments and the topography so that standard notational techniques could be developed and potentially meaningful patterns looked for. Monuments were considered to be potential encoding systems. The question "What is special about this particular place?" was always kept in mind, while particular attention was also paid to site grouping and inter-visibility.

At the end of this exercise it was possible to say that conventional explanations of monument siting seemed to have no basis in fact and astronomical considerations appeared to be a distinct possibility. A preliminary investigation of some sites involving map based calculation of alignments was then undertaken and these were checked by actual observation. That process gave valuable insights into the more practical aspects of horizon astronomy and made it clear that accurate survey was required. Thus a theodolite was obtained but then, due to other pressures, active investigation had to be suspended for a time.

The next stage started with a look at the axes of two stone circles. The first was Drombeg, where a visit to observe the famed winter solstice sunset in the horizon notch above the axial stone made it clear that the notch did not and never had marked the solstice. Theodolite survey revealed the actual declination [1] and when this was checked against ephemeris [2] data for the appropriate period it turned out that the notch had indicated dates about two weeks away from the solstice. In other words, it had accurately marked a synodic month [3] centred on the solstice. After that, a five-stone circle at Trawlebane was seen to have a similar prominent skyline notch above the axial stone, but this time the declination turned out to be somewhat beyond the solstice position so that the sun could never have reached it. Thus, despite the intriguing Drombeg result, both were initially regarded as evidence of symbolic but inaccurate solstitial alignments. However it so happened that some time later, consideration of the practical effects of horizontal parallax [4] on apparent lunar positions lead to the realisation that the Trawlebane notch was in fact marking the position of the moon when it was half-way between the major and minor limits of its most southerly sets. This was an important step, as it indicated that the axial alignments of these circles really might mark lunar or solar events accurately and it potentially gave two previously unconsidered classes of significant event. The question then was whether or not this would be confirmed by investigation of further sites.

It so happened that the next batch of sites chosen for examination did confirm the hypothesis and several more examples of the lunar mid­point were found as well as some of the "expected" lunar extremes, solstices etc, but then other sites occurred where the axial alignments did not appear to fit the pattern. The question then was "why not?". It is clearly unsatisfactory to suggest that some monuments are accurately astronomical and others are something different. Such a result clearly says as much about the methodology and its underlying assumptions as about the nature of the objects being studied. Two strategies presented themselves: 1) Check with ephemeris data to see what the unexpected declinations might mean, and 2) Look to see where on the horizon the "expected" events actually occurred. Both these approaches provided interesting answers but new anomalies occasionally presented themselves as more and more monuments were surveyed and these, in turn, called for further methodological modifications. From the beginning, as a basic working rule, an axis would be taken to indicate the nearest notable horizon feature, be it hill, dip or whatever. Should there be any conceivable ambiguity, then a range of points were surveyed and the axis assessed independently. This rapidly became standard practice in all cases. When it came to seeking the location of specific events, even larger sectors of horizon had to be measured. As the body of data grew it became clear that various particular declination values kept recurring, both axially indicated and at other prominent points on the horizon. Every time that problematical sites were found, the solution was always revealed by surveying more of the horizon and the discoveries made by doing this cast further light on sites already examined.

It was finally discovered that, in all cases, the whole horizon should be surveyed and the relationship of the monument's structural components to the horizon noted. In such a way, any correlation between the monument, the horizon and astronomical events becomes obvious. The process is objective because choice, other than sampling rate, is excluded from data collection and, on analysis, there is clarity about what might be regarded as significant and why that should be so. The methodology cannot discriminate against any type of alignment favoured by the megalith builders because the survey is complete. Therefore it may be applied to any group of monuments provided there are enough of them for any apparent patterns to be statistically meaningful. In this particular case, a study area was defined such that it contained approximately 20% of all the known major monument sites in the complex. The area was bounded by either the coastline or contiguous map sheet boundaries and all monuments within it were examined. Since then the work has expanded beyond those limits until 25% of the known major monuments of the Cork-Kerry Stone Circle Complex have been surveyed, plus a number of standing stones and radial stone cairns. Additionally, the survey has been extended to cover a small number of every type of megalithic tomb represented in Ireland and a few barrows as well. Results are 100% in agreement!

Home | Top | Back

Discussion.

Thom's work concentrated on "indicated alignments". This was an obvious place to start but has, until now, effectively been the only aspect of megalithic monuments open to astronomical interpretation. Consideration of the existence of markers for astronomical events on parts of the horizon not specifically pointed at by the monument was ruled out, on the basis that we cannot know whether they are intended or coincidental if they are not indicated by a built structure. This is an understandable viewpoint, but one that does not hold up under careful scrutiny. We can know that relationships are intended if enough sites are analysed to show the existence of logically consistent patterns that are unlikely to have occurred by chance. Such patterns have indeed been revealed by this study. The frequent recurrence of particular declination values has lead to the formulation of theoretical models that have been confirmed both by further surveys and reassessment of data already collected. In fact, it may be stated that every surveyed site has contributed to the weight of evidence. Full data for all surveyed monuments is included in the catalogue.

Even Thom's results for his indicated alignments were regarded as suspect, because it was held that he could have subjectively chosen the horizon target that gave him the desired result but there was no evidence that this was what the monument builders had intended. It then became the rule that the declination spread of monument axes should be accurately assessed but the actual physical features of that horizon segment should be ignored. As a variation on this theme, Ruggles did also record the declination of any prominent hilltop close to the axial direction. Such strategies have lead to results that are claimed to demonstrate a lack of accurate astronomical alignments. In this study, rather than being restrictive, the question "Just what were the choices available?" was asked. This was done by measuring high points, low points and other significant break points on the appropriate horizon sectors. It was then found that, when many monuments are surveyed, there are consistently repeated patterns in the way that particular events are accurately marked and the shape of the horizon is matched with the solar and lunar nodal cycles.

A further impediment to understanding what the monument builders were doing has also been that investigators frequently had some prior notion of what should constitute a significant orientation, discounting others as non-significant. This approach gave results of the type: x% are oriented on significant solar or lunar events and y% are not, though more sophisticated statistical techniques were often applied. A notable exception to this was Thom's original discovery of the solar sixteenths but his own ideas did lead him to misinterpret the overall picture.

In this study, Thom's sixteen-fold division of the solar cycle has been confirmed but it has become clear that they were not, as he thought, trying to split the year into equal periods. Additional clusters of declination values were obtained that, when compared with ephemeris data for the appropriate period, turned out to mark days about one and two weeks from the major solar events. When this data was rationalised, it could be seen that each of the sixteen "divisions" is better interpreted as being the centre of a synodic month and that the solar year had been cleverly split into sixteen overlapping lunar periods. In other words, though they were interested in the solstices and equinoxes, what they were really interested in was the full moon that occurred closest to that solar event. Likewise for the cross-quarter days and also the sixteenths [5].

For the lunar nodal cycle it was rapidly confirmed that they had not just marked the major and minor standstills of the moon but also the half-way points between them. After some time, theoretical declination values for further possible lunar sub-divisions were calculated because there remained clusters of observed declination values that could not possibly have been solar or made no sense when given a solar interpretation. The match between these new values and the unexplained survey results was good, showing that the lunar nodal cycle had also been divided into sixteen parts. The question then was "Why divide both the solar and lunar nodal cycles in this particular way?". The answer, obtained by studying several hundred years worth of prehistoric eclipse data, would appear to be that it was to relate the cycles together in such a way as to enable eclipse prediction and to do it in a much simpler and more comprehensive way than Thom's alleged high precision alignments.

Not just the Cork-Kerry monuments then, but all megalithic monuments and indeed barrows as well, can be shown to mark sites for the practice of a sophisticated form of accurate luni-solar horizon astronomy and specifically for the purpose of eclipse prediction. Leaving aside the technicalities for now, just consider what one would actually see if time was devoted to regularly observing solar and lunar rises and sets in temperate latitudes. Early societies were probably more interested in the moon but we will look at the sun first:

The sun's rises and sets move from south to north and back again along the horizon during the course of a year. To mark the pattern one might find, say, a location from where two hills, one each side of a valley, were in the right places to mark the extreme solar positions (the winter and summer solstices). The two hilltops would be (at these latitudes) somewhere around 83 degrees of azimuth apart. That's nearly a right-angle. It wouldn't matter if they were both marking rises or sets, the trick would be to find a place from where each solar extreme occurred on the appropriate hilltop and the day halfway between them was in the bottom of the dip of the valley. One would then, surely, want to mark the position so that it could be found again - it would be a fairly restricted area. The advantage of such a location would be that not only is the annual cycle permanently marked but it is marked in such a way that it is possible to estimate intermediate intervals from the known accurate points because the sun would rise/set in (more or less) the same position on the same day each year.

Now, having established such a useful solar observatory, let's consider the moon, which would also be seen to move from south to north and back again along the horizon but in only 27/28 days. The extreme positions would be in the general vicinity of the two solstitial hills. However, the moon wouldn't rise/set in the same place on the same day each month and it wouldn't have exactly the same extreme positions each month either. If we kept watching and counting, it would become apparent that trying to mark the individual days of the month was a waste of time but that marking the extreme positions would be useful, because the overall distance traveled along the horizon expands and contracts in a predictable manner, every 18/19 years. The locations of the most southerly and northerly moon rises/sets would, over that 18/19 year period, oscillate around the solstitial hilltops, each covering a spread of about 20 degrees of azimuth between their inner and outer limits. The solstitial markers would be a bit to the north of the central points of these oscillations. We could even improve the situation for lunar observations by seeking another place, from where the the most extreme monthly rises/sets could be seen moving from the inner basal step of each hill to the outer basal step and back again - with the hilltops now marking the mid­point of the lunar nodal cycle.

Having spent so much time observing the sun and moon, we would be well aware of eclipses, especially lunar ones. With the annual solar and 18/19 year lunar nodal cycle marked, we would be able to see something very useful. Every 9 years or thereabouts, when the cycle of extreme lunar positions was at either end of its range, any lunar eclipses that happened would be on the full moon nearest the time when the sun was at the centre of its own range (an equinox). Likewise, in the year when the lunar extremes were at the centre of their range, eclipses would happen on the full moon closest to the times when the sun was at the limits of its own range (the solstices). When the lunar extreme positions were half way between the centre and an end of the range then eclipses would occur when the sun was also half way between the centre and an end of its range.

Such is a hypothetical scenario for starting to develop the type of horizon astronomy used by the megalithic monument builders. It is a reasonable description of a basic methodology for using the shape of the horizon to mark solar and lunar nodal cycles but what we actually find in the case of megalithic monuments is much more sophisticated. They used more complex horizon shapes than those described to sub-divide the cycles more comprehensively and give greater accuracy but that's not all. For us to appreciate the full extent of what they were doing we still need to go into a bit more detail concerning lunar observation.

As stated, it takes 27/28 days for the moon to move from south to north and back again but, because the sun has been moving as well, it is 29/30 days from one full moon to the next. This means that every time the moon reaches its most southerly or northerly position it has a different phase to the last time it did so. One cannot usually see a waxing moon rise nor a waning moon set because these events occur during daylight so, to get full coverage of the cycle, it is necessary to observe the lunar extremes rising at some times of the year and setting at others. This means that while it doesn't matter for the sun whether one observes rises or sets, for the moon one must observe both. Again, because of the phases, one must observe both south and north extreme positions. So, for the moon, all four quadrants of the sky must be used in a simple seasonal pattern and the survey data makes it clear that this was consistently done by the monument builders.

What we find then is that orientation was only a secondary characteristic of the monuments. Their most important physical function was that of being place markers. Monuments were used to mark sites from which preferably all, but certainly a substantial proportion of, the horizon was useful for the measurement of the solar and lunar nodal cycles. Because the horizon is never perfect, not every division of the cycles will coincide with a notable landscape feature but enough will do so that the pattern is obvious and the shape will be such that some estimation of missing ones is possible. Where one monument does not command horizons capable of measuring both the solar and lunar nodal cycles completely then somewhere not too far away will be another, offering complementary facilities. This seems to be true of all the major monument types - barrows, tombs, stone circles, boulder-burials, short stone rows and standing stone pairs.

Having established that large sections of the horizon were used to graphically illustrate and measure solar and lunar nodal cycles, it becomes obvious that monument orientations were "context sensitive". This explains why study of "indicated alignments" has never revealed a clear picture. The builders and users of the monuments knew very well why they were in their specific places, what they were for, and what to do there. Orientations and other architectural features were used to give supplementary visual cues. Sometimes just leading the eye towards a key segment of the horizon. More often, perhaps, indicating a significant event at the end or centre of a sequence. Occasionally an axis was used to mark some key event lacking a natural marker, by accurately pointing in the appropriate direction towards a void in the sequence. Multiple stone circles also used other techniques as indicators. Some have a "secondary axis", where a larger or smaller stone than one might expect on one side of the circle faces a similar pair on the other to form, as it were, a secondary axial and portal pair which indicates a key event that complements one pointed at by the main axis. Other variations in stone and gap sizing were used to emphasise various horizon sectors.

The quest to define a methodology capable of determining, in all cases, the purpose and siting criteria of these monuments has resulted in the probable answers to those very questions. A large body of evidence has been produced which, in the catalogue, is given for every site surveyed. Verbal description combined with annotated photographs has been used as the vehicle, as this is essentially to do with visual perception. Survey data tables are also supplied. Any random sample of this information will show a similar overall picture but the details will obviously vary from site to site.

For many years, patterns could be recognised but no formal statistical analysis was possible because data was only partial and traditional methodologies were unsuitable. Logical analysis was used to compensate and this was the position:

The original study area was defined so as to contain an unbiased sample of about 20% of the major monuments of the Cork-Kerry Stone Circle Complex. Since then, the study has expanded beyond that area and indeed, that group of monuments. To date, 169 sites have been surveyed and in every case conform to the same general patterns of relationship with the surrounding horizons. That's more than 25% of the known major monuments of the Cork-Kerry Stone Circle Complex and 100% of all surveyed monuments.

For there to be no exceptions in such a sizable and largely unbiased sample, there are two possible explanations:

• Either the analysis is correct and the monuments really were sited in places deliberately chosen because the horizons were especially suited to the mapping of solar and lunar nodal cycles, or they were not.
• If they were not, then it follows that it must be true that the same coincidences found at monument sites between the skyline and the astronomical patterns of the period that the monuments were constructed would be found at any random location within the study area.
• The latter case is clearly absurd and easily shown to be untrue.

The logical argument given above still stands but, with more data and improved techniques, accurate full horizon profiles are now possible. Numerical comparisons can be made and statistical analysis of this data quite clearly shows a non-random relationship between monument sites and the horizon.

Home | Top | Back

Conclusion.

A large body of survey data has been accumulated, which appears to demonstrate that the primary purpose of megalithic monuments was the accurate mapping of solar and lunar nodal cycles, for purposes of eclipse prediction. Rather than axial orientations indicating single astronomical events as previously postulated, the monuments actually mark places in the landscape from which a large proportion of the horizon graphically models the solar/lunar nodal cycles and contains a good number of accurate markers. Analysis of the data indicates that by the time even the earliest of these monuments was being built that both the solar and lunar cycles had been sub-divided into sixteen parts. Also that the sixteenth parts of the solar cycle were each further sub-divided into three. When this has been done there is a direct correlation such that one may observe the current place of the moon in its cycle and accurately predict the time of the year that the next eclipse may occur.

While the evidence presented here for sophisticated prehistoric astronomy is substantial, it is obviously up to others to decide whether or not the interpretation is correct. That is why full data is supplied. Whatever the verdict there, it is still claimed that enough work has certainly been done for it now to be possible to define a methodology capable of yielding a definitive answer to some long standing questions. Surveying the whole horizon from a monument site and recording the relationship of the monument components to that horizon provides an objective data set from which the degree of correlation with any given set of astronomical criteria may be obtained. It also allows different monuments to be compared directly. Provided that enough monuments exist to provide a sample of sufficient size, the result will be statistically meaningful. Thus "Whole Horizon Analytical Techniques" are capable of application to any group of monuments where some form of horizon astronomy or special topographical relationship is suspected.

Home | Top | Back

Notes:

1. Declination: A measure of the position of an object in the sky, expressed in degrees North (+) or South (-) of the Celestial Equator (which is the earth's equator projected into space). It is exactly equivalent to the latitude at which the object would pass vertically overhead. In this particular case, -22.78°.
2. Ephemeris: Tabulation of calculated astronomical data. In this case, the Jet Propulsion Laboratory's Horizons Ephemeris.
3. Synodic month: The 29/30 day period (mean 29.53) between one new moon and the next.
4. Horizontal Parallax: Makes the rising or setting moon appear to be further south than it really is. This is due to the fact that NW Europe is quite far from the equator and the moon is much, much closer than the sun, so that when it is near the horizon we are in effect looking down on it rather than straight at it. Lunar parallax (mean) in SW Cork is about 0.85 degree of declination at the horizon but its value depends on latitude and horizon altitude. In technical terms, Lunar Apparent Declination = Apparent Declination - Horizontal Parallax.
5. The tropical year is 365.24 days. Thus, to the nearest whole number, the mean interval between: a) the two solstices is 183 days, b) a solstice and an equinox is 91 days and c) either a solstice or an equinox and an adjacent (astronomical) cross-quarter day is 46 days. A sixteenth division of the year splits this interval into two 23 day periods. If we take the outer limit of a synodic month centred on a solstice, equinox or cross-quarter to be 15 days from it, then it is also 8 days from the sixteenth. If we take the outer limit of a synodic month centred on a sixteenth to be 15 days from it, then it is also 8 days from the adjacent solstice, equinox or cross-quarter. The result is a year split into 48 "weeks" of 7 or 8 days. The slight variations in the true lengths of some of these periods are not significant within the given context (which was the reconciliation of lunar and solar cycles, not the creation of equal length periods).

Home | Top | Back