Paleolithic and Neolithic Cultures



For most of the Paleolithic period [or Old Stone Age, beginning more than ~21/2 million years before present (b.p.)], there are few materials that could be interpreted as relevant to human understanding of astronomy, even in the vaguest terms. Evidence for interest in the heavenly bodies has been suggested only for Australia (see §11) and for Western Europe during the Upper Paleolithic (70,000 to ~10,000 years b.p.). A critical summary of the European Upper Paleolithic is provided by Hadingham (1979) in Secrets of the Ice Age. Despite its provocative title and popular nature, this work reviews the results of modern scholarship about the hunters and gatherers of the last 70,000 years or so, mostly from Italy, France, and Spain. He emphasizes the difference of the environment of that time from any existing today: colder, wetter, but in some ways richer, with vastly different fauna. He discusses both continuities and changes among human populations, their tool kits, and other aspects of the culture. The people were Neanderthals (Homo sapiens neanderthalensis) or Cro-Magnon (Homo sapiens sapiens)—both much like ourselves in physical type and inherent capabilities. They were skilled in making stone tools and had some crude housing, at least in some areas. They depended heavily on game, and some became skilled (and perhaps overspecialized) reindeer hunters. Others depended on wild cattle, and most groups probably killed a wide range of animals. Gathering of vegetable foods was surely of great importance, although usually this must be inferred from sketchy evidence. Fishing was probably of some importance, with more lakes and streams than today. Most sites that were then along the coast, where we might expect some evidence of fishing and indications of whether it was based on use of good watercraft, are now sunk deep beneath coastal waters, which have risen many meters since the melting of so much glacial ice. It has been suggested that in some areas there were substantial attempts to control the animal populations and that some of the reindeer could be considered as having been at least semidomesticated. Similar suggestions have been made for horses.

6.1 Paleolithic Cultures

For most of the Paleolithic period [or Old Stone Age, beginning more than ~21/2 million years before present (b.p.)], there are few materials that could be interpreted as relevant to human understanding of astronomy, even in the vaguest terms. Evidence for interest in the heavenly bodies has been suggested only for Australia (see §11) and for Western Europe during the Upper Paleolithic (70,000 to ~10,000 years b.p.). A critical summary of the European Upper Paleolithic is provided by Hadingham (1979) in Secrets of the Ice Age. Despite its provocative title and popular nature, this work reviews the results of modern scholarship about the hunters and gatherers of the last 70,000 years or so, mostly from Italy, France, and Spain. He emphasizes the difference of the environment of that time from any existing today: colder, wetter, but in some ways richer, with vastly different fauna. He discusses both continuities and changes among human populations, their tool kits, and other aspects of the culture. The people were Neanderthals (Homo sapiens neanderthalensis) or Cro-Magnon (Homo sapiens sapiens)—both much like ourselves in physical type and inherent capabilities. They were skilled in making stone tools and had some crude housing, at least in some areas. They depended heavily on game, and some became skilled (and perhaps overspecialized) reindeer hunters. Others depended on wild cattle, and most groups probably killed a wide range of animals. Gathering of vegetable foods was surely of great importance, although usually this must be inferred from sketchy evidence. Fishing was probably of some importance, with more lakes and streams than today. Most sites that were then along the coast, where we might expect some evidence of fishing and indications of whether it was based on use of good watercraft, are now sunk deep beneath coastal waters, which have risen many meters since the melting of so much glacial ice. It has been suggested that in some areas there were substantial attempts to control the animal populations and that some of the reindeer could be considered as having been at least semidomesticated. Similar suggestions have been made for horses. Possible halters or bridles are shown on carved designs of horses, which appear very convincing in isolation; comparable depictions of bison make the interpretation seem less likely. Despite fanciful popular accounts, we know nothing directly or by firm inference about the social structure of these peoples except for clear evidence of the existence of nuclear families of parents and children, and some evidence for care of partially disabled individuals. The great cave paintings of Altamira and elsewhere are majestic and inspiring art, but the motivations behind them remain obscure. Some degree of religious or magical inspiration seems likely in at least some cases, and few archeologists would regard sheer esthetic pleasure as an adequate explanation. Some sort of symbolic interpretation seems likely. Leroi-Gourhan, cited in Campbell (1988b, Pt. 2, p. viii), notes that only certain species of animal known to exist in the Paleolithic are depicted, and these in discrete pairings, architecturally arranged both geographically and historically in such a way as to convey the sense of a polished system of religious belief.

A series of carved markings have been uncovered on antler and bone that look like tallies. Alexander Marshack has published several articles and a popular book (Marshack 1972b) arguing his case that these apparent tallies were designed to mark lunations and therefore represent mathematical notations, which, for him, form an ultimate basis for all science, a necessary, and before this, a missing, prototype. The marks were regarded in the 19th century as perhaps hunting tallies. Marshack has demonstrated by microscopic analysis that the marks were usually made by different tools, hence, presumably marking a succession of different events, thus confirming the tally hypothesis, but were made over extended time. Marshack’s analysis of three faces of an eagle bone from Le Placard, France, suggested that the “feet” were added after the vertical lines and constituted a reuse of each vertical symbol. Marshack further argued that the doubling of the “foot” and vertical components indicates “waiting” periods perhaps created by bad weather; the true observation of the phase then being made on the following day. The notch indicates that whatever calendrical purpose the bone may have had, it was used as a whistle. Marshack’s claim that they represent a lunar calendar is more controversial, and his view that all the carvings represent a single tradition is in no way demonstrated. Hadingham (1979, p. 253) cites M.A. Littauer, who suggests that the material shows that humans were capable of “cumulative symbolic recording” and that there were “varied and idiosyncratic” ways in which that was done.

Even if no more than this was demonstrated by Marshack, it represents an important advance in our knowledge. This “cumulative symbolic recording” is a necessary base for determining all but the most basic astronomical periodicities. The objects of Marshack’s study include a small engraved bone from the Blanchard site of the Aurignacian period (Marshack 1972b, pp. 40–49) and a ritual “baton” from Le Placard of the Magdalenian III period (Marshack 1972b, pp. 88–89). It is worth emphasizing that the Magdalenian III period is separated from the Aurignacian by a longer interval than separates us from the Magdalenian. The Blanchard bone of the Aurignacian period had 69 marks in 24 groups that had been made at different times, sometimes with different inplements. There were 1 to 8 marks in each single group, and the shape of the marks varies from near ovals through broad crescents to thin crescents. They do indeed appear to be some sort of tally, and the shapes are suggestive of a lunar notation. Marshack thinks that the notations may cover two and a quarter months.

Hadingham (1979, pp. 250–251) concluded that by accepting the notches as marking whole lunations, half lunations, and quarter lunations as divisions that were used, allowing for weather-biased counts, and even for noncontinuous record-keeping, “the potential of proving any string of numbers to be lunar is considerable.”

Marshack (1972b, pp. 141–143) gives two examples of wooden calendar sticks from the Nicobar Islands, Bay of Bengal, which are visually and structurally similar to many of the Upper Paleolithic examples. One of these is shown with a Magdalenian-period baton from Le Placard, France (Marshack 1972b, pp. 88–89). Marshack analyzed the sequence of markings and groupings on the baton and suggested lunar interpretations. Both sets keep a tally by different lengths and different angles of lines and different groupings. Neither tally would necessarily lead an uninformed observer to think that it dealt with lunations. The Nicobarese sequence does not obviously seem to be lunar. The difference between the Le Placard baton and the Nicobarese calendar stick is simply that we know from Nicobarese informants that their calendar sticks were indeed being used for lunar tallies. Despite the obvious differentiation of subgroups on the Magdalenian baton, it is by no means certain that Marshack’s grouping coincides entirely with those of the Magdalenian engraver, particularly for longer intervals than for any short grouping. It is these longer intervals that coincide with lunations (or can be made to do so).
We concur with Hadingham’s (1979) summary:

Marshack has never published analyses of a large number of objects which would allow anyone to repeat the statistical tests which he claims back up his theory. Instead, his most detailed study is of only six objects with notches that do fit the lunar phases to varying degrees, which is not to say that another explanation might not fit them better. No one would deny the attractiveness of the theory of a Palaeolithic lunar calendar, its convenience as a rational explanation of the ‘hunting marks’, and its widespread use by peoples at a comparable level of technology. Common sense seems to tell us that Marshack is right, but as yet the proofs are lacking.

More pointed criticism has been made by d’Errico (1989, 1992). In any case, Marshack’s detailed presentation should be consulted by anyone wishing to form an independent judgment on this important issue. There is another parallel in a North American calendar stick, published by Merrill (1945).

It has been claimed that certain Paleolithic paintings represent asterisms (Huffer Trinklein and Bunge 1967, p. 95). If verified, these would be the earliest such representations known and therefore would be of great importance. More recently, Michael Rappenglück (1997, 1999, 2000) has compared Guide 7 software package sky simulations with star-like representations at several Paleolithic sites. He attempts to match the dates of the asterisms as affected by proper motions (see §3.1.7) to the best estimates of the dates of the paintings, and he considers the contexts of the images within their tableaux. Among the asterisms that he thinks may be represented are Corona Borealis (Cueva del Castillo), the Pleiades and Leo, interpreted as a horse (both at Lascaux). He draws many parallels with later work, as for example, grotto art done for Tiberias, and with Celtic myths and images. Although we cannot comment on the results at this juncture, the broad range of scholarship brought to bear on the work is to be commended.

6.2 Megalithic Cultures

6.2.1 The Megalithic World: Cultural Description

Human endeavors of the post-Paleolithic world of Europe were dominated by two factors: increasing reliance on the sea, from molluscs to large fish, and the development of farming techniques, including the use of domestic animals and a wide range of plants, especially grains. Although we know little about the watercraft involved, people had reached Ireland by about 6500 b.c. and they must have been skilled sailors to get there. The spread of farming techniques occurred later. The invention and spread of the polished stone axe, the hallmark of the Neolithic or “new stone age,” was an important factor in clearing forests to create fields for grain. Pottery was another important innovation that allowed storage of foods and liquids and changed cooking techniques, especially of meat. Stews allow adequate nutrition from a much smaller amount of meat than do other cooking techniques, while reducing the danger of disease. Techniques of building with wood, probably largely developed for housing, may also have been used in building more developed watercraft. The movement of cattle, sheep, goats, and pigs by water involves different craft than do those that are adequate for fishing, even deep-sea fishing.

In the 5th millennium b.c., in coastal areas from the Mediterranean to Ireland and the North Sea, Neolithic people began making monumental structures incorporating massive blocks of stone. These were usually either large collective tombs or temples. The term “megalithic” has been applied to such structures. The distribution of such early structures is shown in Figure 6.1.
Figure 6.1.

Some important megalithic monuments of Europe. Note that many are near coasts. Shading indicates the principal areas where structures are found. For fuller maps, see Mohen (1990). Drawing by Sharon Hanna.

The principal classes of monuments relating megalithic cultures to astronomy are as follows:
  1. (1)

    Mounds, especially those with so-called passage graves, each of which consisted of a central burial chamber and a long entrance passage

  2. (2)

    Menhirs, from the Breton word for large pillars, often isolated or in groups of 2 or 3

  3. (3)

    Rows or fans of pillars in alignments

  4. (4)

    Geometric figures—rings (circles, flattened circles, combinations of arcs of different radii), rectangles, or other shaped figures

Our knowledge of the archeological context of such monuments is increasing rapidly, but many of them are very hard to date. The most sophisticated attempts to date large groups of monuments are those of Aubrey Burl (especially, 1976, 1993). We summarize the chronolog-ical data for a selection of the most important sites in Table 6.1.
Table 6.1.

Chronology of selected megalithic sites.



4600 b.c.

Dissignac: Tomb oriented to midwinter solstice light play


4200 b.c.


3800 b.c.


Le Grand Menhir Brisé


Malta: Temples aligned on asterisms—Numeracy and notation

3400 b.c.

Gavr’inis: Tropical year


Castle Rigg


Long Meg and her Daughters


Maes Howe: Midwinter sunset light play

3000 b.c.


Los Millares: Tombs usually oriented to SE quad Stonehenge Ia: Solar and lunar standstills

Brugh-na-Boinne: Complex tropical year alignments; sundial?


2600 b.c.

Crucuno: Pythagorean triangles, megalithic “yard,” tropical year alignments

Stonehenge Ib suggested eclipse prediction


2200 b.c.

Mzorah: Cardinal directions


Callanish: True N-S alignment, lunar movements


(Hill o’ Stanes, Mid-Clyth)


1800 b.c.

Mull and Argyll: Extremes of sun and moon


Hunnenbette: Solar (and lunar?) alignments


Temple Wood: Extreme N. decl. of Moon

Ballochroy: Midwinter and midsummer alignments at right angles


1400 b.c.


6.2.2 Engineering and Astronomy

Most of the important problems in the astronomical interpretation of megalithic architecture have been raised by Alexander Thom, who also suggested answers to many of them in a voluminous body of writings. Archibald S. Thom (1988) lists 129 articles and books that his father wrote or coauthored. The elder Thom was already reading about archaeoastronomy in 1912 and by the 1930s began mapping sites and studying their geometries. His work as an engineering professor, and his avocation, sailing, helped him in the endeavor. In the Thomistic spirit, we again raise questions he raised and explore, with the help of his field work and subsequent work, the evidence for alignments. The main questions are as follows:
  1. (1)

    How were the megalithic sites laid out?

  2. (2)

    Were they observational sites?

  3. (3)

    Do they incorporate astronomical information?

  4. (4)

    If they do incorporate astronomical information, how was this information transmitted within the culture?


First, how were the sites laid out? Thom suggested a range of well-understood geometrical patterns, incorporating Pythagorean triangles and use of a regular unit of measurement, which he called the megalithic yard. It is logically possible to separate the problem of geometric layout from the problem of mensuration, but here the two seem intimately related and the evidence for one is closely tied to the evidence for the other.

Were they used observationally to determine or check astronomical data?

Do the sites incorporate astronomical information? The possibilities include solar, lunar, stellar, and planetary alignments, and Thom has at various times suggested that particular sites included information about the Sun, the Moon and, rarely, the stars. To our knowledge, planetary alignments have not been suggested in the megalithic material. Thom also maintained that many sites included alignments on distant natural features, such as notches on the horizon, or on very large man-made features, such as Le Grand Menhir Brisé in Brittany, which was visible for many miles around. If such features were used, they would have increased the precision with which astronomical phenomena could have been measured.

How was the information transmitted? If data were actually gathered, how could they be stored and subsequently used to determine periodicities or cyclicities of such phenomena as lunar motions and eclipses? Some scholars consider that the lack of clear written records provides an insuperable objection to the argument that astronomical alignments in megalithic structures were intentional. On the other hand, many archaeoastronomers accept the view that the monuments are, as the festschrift to Thom puts it, “records in stone.” Thom believed that the “cup and ring” markings, which we discuss later, were some sort of notational record, not a record of the language of scribes, but nonetheless a minimal written record. With respect to the Irish monuments, Brennan (1983) has attempted to interpret some iconographic motifs in megalithic art as notations of astronomical information. Finally, O’Kelly’s (1982) discussion of pagan Irish mythology in connection with his studies of Newgrange shows that mythology may have been one of the vehicles for the transmission of astronomical knowledge, although this is not how he phrases it.

The answers to these questions are tied to a priori conceptions about the cultures involved and about how such problems should be studied. This is to some extent true for all investigations involving cultures. Here, the conceptions affect and are affected by our understanding of the overall functions of the monuments. A statistical study of various Scottish sites led Norris (1988, p. 273) to write, “These results constitute the first definitive evidence that megalithic man was interested in marking the southern limits of the Sun and Moon.” A more cautious statement is that there is a low statistical probability that the alignments are random, which suggests interest in particular alignments that coincide with lunar and solar alignments (and that we might regard as significant). One approach, taken by Ruggles (1988) and his associates, is to remove the problem as far from the context as possible in order to achieve “fair” conditions, unbiased by expectations, in order to make objective determinations. But this approach too is not without its biases; for example, is it truly scientific to ignore possible compensations provided by human ingenuity in dealing with observing alignments involving large and often irregular stones? Scholars from ethnology, anthropology, or archaeology tend to see astronomers as useful technicians who can supply important data, but who are unskilled amateurs in the study of human behavior and artifacts and who are unable to integrate the data into the contexts that alone can give them meaning. Ruggles (1989, p. 24) has expressed the view that both approaches are necessary:

British archaeoastronomers have often been accused by their American colleagues of lavishing too much attention on statistical rigour at the expense of cultural context; it may soon be the turn of the British to persuade their colleagues that statistical rigour has a crucial rôle for a great many archaeoastronomers, and to teach them how it can be integrated into an approach which quite properly considers the great diversity of cultural evidence which may be available.

Thom has been attacked on statistical grounds by those who cite unconscious bias in selection of evidence favorable to the expected alignments, and Thom’s attention to the geographical context in attempting to determine if particular alignments are meaningful. Attacks on Thom’s work have come also from the archaeologists (e.g., Burl 1988). Anthropologists and archaeologists tend to regard Thom’s selections on the basis of geographic context as admirable, but criticize his inadequate use of cultural context. Which methodology is appropriate requires examination of the contextual features. Ruggles (1988) has argued that “megalithic culture” is, archaeologically, many definably distinct cultures. This is true, but can we, nevertheless, speak of a single megalithic tradition that may extend for millennia and over much of Europe, especially in coastal zones? Put another way, to what extent can conclusions validated for one area be extended to other areas? To answer these questions, we consider the common cultural contexts.

We know from historical and ethnographic records that farmers in most parts of the world believe that all activities of farming, from clearing the fields to harvesting, are affected by the movements of the Moon. Is it reasonable or not to suppose that the Moon was equally important in the beliefs of European Neolithic farmers? We also know from the distribution of monuments (see Figure 6.1) that at least some members of these groups were able sailors. Navigation plays a crucial role in sailing, and coastal navigation has traditionally relied on real or artificial landmarks, with sailing directions tied to successive alignments of such features. Useful landmarks need not be the most prominent features of the landscape, and they may be as diverse as the trees, rocks, spits, sand bars, and other features of variegated shores. Moreover, navigation, in all cases in which we have adequate records, has been tied to stellar observations and stellar alignments for craft out of sight of land. If builders of megalithic monuments were drawing on experience as navigators, we would expect to find natural landmarks as distant foresights or backsights and we would expect to find stellar alignments. Moreover, important alignments are determined from the information they give about sailing geometry, not whether they are bright stars or dim ones. This should be borne in mind when assessing claims for particular sites in which no stellar alignments were found, when only the brightest stars were examined.

As we have noted already, the Moon has two properties that impact ancient cultures: its light and its gravity. The light of the Moon is especially important on a non-summer night and at high latitudes; given the limited range of illumination of torches, nocturnal travels on the sea or on land require the Moon, and the fuller the phase, the better. Moreover, for seafarers, the Moon is important for more than its night light. The tides arise from the different effects of the Moon’s (and, to a less extent, the Sun’s) gravitational attraction on different portions of the Earth (see §4.5.2, for a discussion of the origin of the tides). The Moon’s association with tides is readily discernible: the time difference between the meridian passage of the Moon and high tide is a fixed quantity for a given site (the establishment of the port), even though it varies from site to site because of the effects of local geography. Because the tide-raising effects of the Moon and Sun add together vectorially, the strongest tides will occur around new and full moon (spring tides), and least at first and third quarter (neap tides). Thus, the strengths of the tides are correlated with lunar phase. Moreover, the strongest tides occur when the Sun and Moon are in syzygy, and in such circumstances, eclipses can occur, some of which may be visible locally. The Moon contributes roughly 2/3 of the tide raising forces, and the Sun, ~1/3. According to Harris (1937, p. 611), the range of lunar tides is ~10% greater at a node crossing and ~10% less when the Moon is 90° away from it. Finally, the tide-raising force varies with the inverse cube (not the square) of the distance, so that the lunar tides depend on the location of the Moon in its eccentric orbit (see §2.3.5). Consequently, the range of the tides is ~17% greater when the Moon is near perigee and about the same amount less when it is at apogee. In waters with dangerous rip tides, these phenomena provide an adequate reason for attempts to determine lunar movements with precision. The importance of stellar horizon marking might easily have been extended to the Moon. These facts provide strong motivation for the study of the movements of the Moon and stars, and they may provide the basis for a broad and lengthy tradition, as ethnological studies of indigenous cultures have shown (see §8 and 11).

Given the evident motivation for and likely continuing tradition of megalithic observations, what makes claims of astronomical alignments controversial?

There are several problems. First, there is the problem of relating the current position of a monument to its placement at the time of its erection and use. Historical records indicate that many of the boulders that are allegedly part of alignments in Brittany and Britain have been reerected. The establishment of the existence of a present alignment is relatively trivial compared with doing so for the remote past. Even if historical records do not indicate the reerection of a standing stone, movement of the stones is highly likely from a number of factors: natural forces (trees, frost heaves, earth tremors, etc.) or by people (because of agricultural activity, subsequent religious usage or the discouragement of it, or mere vandalism).

Second, the precision with which observations could have been made with a particular alignment configuration is not known. There are also questions involving the visibility of distant foresites. The vegetation cover between the alleged foresights and backsights at the time of their presumed use is largely unknown for many sites, especially those in agricultural and village areas. In our highly agriculturalized time, open farm land is far more common than it ever was in megalithic times. The atmospheric conditions that prevailed can be plumbed to a certain extent, as we discussed in §3.1.1, but this is scarcely sufficient to provide an ap­propriate atmospheric pressure and temperature on a day-by-day basis. Therefore, the range of the refraction variation (see §3.1.3) in any given period is only loosely constrained.

Third, given the large number of menhirs, boulders, or tumuli in many sites, and a plethora of distant foresites, there is the problem of selection of relevant sightlines. The arbitrariness of some of the directions selected (and none other) for the observation ascribed to particular sites by current day archaeoastronomers is a principal ground of criticism.

Implicit in all the criticisms is the fundamental problem of the intended usage of the site. In no case do we have direct evidence how or even if any megalithic monument was used; we know only what could have been done with it.1 One possible direction to take is to consider an inverse problem.If an alleged foresight-backsight line were blocked byelevation, vegetation, or construction at the time of its erection, could we not say that that particular astronomical association would be disproven? Unfortunately, it appears that we cannot be sure even of this. At Newgrange, we shall see that some of the kerbstones have decoration facing the mound and, hence, would have been invisible after the mound was completed, and until reexposed by excavation. Yet the symbolism of the site is such that they could be considered to have been so placed by intent.

Many of Thom’s astronomical interpretations rest on his ideas that high precision was sought in the observations of the ancient Britons, and that they possessed considerable understanding of geometry. His ideas about how the figures of large stones were laid out bear some scrutiny because they involve Pythagorean triangles.

6.2.3 Megalithic Mensuration

From a study of the dimensions of megalithic monuments throughout the British Isles, Alexander Thom (1967/1971/1973/1978) concluded that standard units of scale were employed. Some practitioners call this the “quantum hypothesis.” Among the units were what Thom called the Megalithic Inch (MI) equal to 0.82 in or 21 mm, and the Megalithic Yard (MY) equal to 2.720 ± 0.003 ft or 0.829 m (the precision cited is Thom’s). The data, methods, and results have all come under careful scrutiny with the result that most investigators subsequent to the Thoms have concluded that marginal evidence exists for the Megalithic Yard, but not as uniform a measure as Thom’s estimates would imply. It is important to reexamine the evidence to see how this conclusion was reached and, regardless of established opinion, to see if the quantum hypothesis may, nevertheless, be true regardless of rigid tests of significance.

The quantum, or the smallest basic unit of any quantity, can be determined from two basic conditions: it can be no larger than the smallest measured quantity and all larger quantities must be integral multiples of it. This can be considered the case if and only if no more than one such similar-scale unit was being used by different groups in the area. One can indeed question whether it is reasonable to expect a culture that has left no accepted evidence of writing (“cup and rings” notwithstanding) to be able to systematize their units of scale so that the same value is derived for each site. Leaving that question aside, what are the bases for the claims that these units were in use among the megalith builders?

They are the scale of markings on the stones in the case of the MI and the scales of the geometric figures created by the placements of the stones in the cases of larger units. Thom identified the Megalithic Inch to be about 1/40 MY on the basis of a histogram of frequencies of occurence of diameters of markings under 12 inches; the histogram shows that there is an apparent clumpiness of diameters near integral units of the Megalithic Inch. The case for the Megalithic Inch was subjected to Broadbent’s lumped variance test2 by Thom and found to be significant. Heggie (1981b, p. 48) subjected the data to more stringent testing, which did not show significant clumping. On the latter basis, there does not seem to be a unique unit that provides a clearly better histogram than any other. Basically, the situation is this: If one assumes a value of the MI (e.g., 1/40 MY), one gets an indication of significance, but this does not prove that that unit alone was significant. Similar arguments have cropped up for the larger units.

With respect to the Megalithic Yard, the studies have been more positive, but the strictest tests suggest marginal significance only. Again, the evidence is for that unit in the form of a histogram of the frequencies of the diameters of stone “circles” (these are often not circular, as we discuss below). The diameters yield significant results for the existence of the MY once a unit is suggested. The same is true for a number of subsets investigated by Thom and others. There is somewhat greater significance for the more flattened variety of circles and for the Scottish circles than for the more nearly circular configurations and for the English and Welsh circles. Supporting evidence for the quantum hypothesis comes from other measurements [(such as the separations of individual stones, cf. Thom (1964/1971/1973/1978)] and from different types of structures, such as stone “fans.” The separation of stones yields a common unit of 0.5 MY to a significance level of 0.97, meaning that the probability of the clumping being produced randomly is only 3%. An example of a stone fan is seen at Mid Clyth, where the convergent rows of stones march up (or down) a hillside (Figure 6.2).
Figure 6.2.

A sketch of the megalithic fan-shaped configuration of stones at Mid Clyth in Caithness, in northern Scotland: from Thom (1964), who finds in such constructs evidence of stone grids used for lunar calculations. For discussion of the functioning of the site, see §6.2.15. Drawing by Sharon Hanna.

The quantum found here, however, was not the Megalithic Yard, but a value 7.743 ft. = 2.360 m (2.843 ≈ 20/7 MY). Work on sites at Carnac and elsewhere has shown marginally significant results. There are notable examples of a failure to confirm the existence of the MY: in the Irish stone circles examined by Barber (1972) and in rings of the “Sanctuary” near Avebury in Wiltshire (Heggie 1981b, p. 42, based on Burl 1979), there does not seem to be evidence for the Megalithic Yard. This has led to suggestions that the integral values of diameters may be suggesting merely “popular” values, or that units were used in some cases and were randomly selected in others.

Supposing the existence of the MY, the measures of accuracy with which it can be obtained have been studied by Heggie (1981b, 56ff). The diameter of the highly circular Ring of Brogar at Orkney (cf. Figure 6.3, which shows also a portion of a comparably large site, Avebury) was determined by Thom and Thom (1973, p. 171) to be 340.02 ± 0.60 ft or 103.64 ± 0.18 m.
Figure 6.3.

(a) The Ring of Brogar, Orkney, northern Scotland. The monument has a diameter of 103.6 m, surrounded by a circular ditch with diameter of 142 m. These dimensions make it one of the largest of the stone rings, comparable to Avebury. Photo courtesy of Sharon Hanna. (b) Avebury in the SW, looking SE. Large entrance stones marking the South entrance are on the right. Photo by David H. Kelley.

This is close to 125 MY. If that were the intended value, the error in the MY would be ≈0.005 ft. The uncertainty is small enough in whatever unit was intended. And it was most probably the Megalithic Yard as suggested by Thom. The variations from site to site suggest that it may have been some aspect of human dimensions—such as foot pacing—which can be done with high precision yet the resulting structure will vary slightly in dimensions from site to site with the builders.3 This was the hypothesis of Kendall (1974, p. 258), although it was challenged by Thom (1974, p. 179).

Thom’s metrical and geometrical hypotheses were supported by subsequent analysis of the Breton site of Crucuno. Although the site is composed of moderately large standing stones, it is a simple rectangle and not impressive compared with many other sites in the region. This Megalithic site is, nevertheless, one of the most important, as a glance at Figure 6.4 reveals.
Figure 6.4.

The construct features at Crucuno. The Xs mark the approximate positions of the stones. Drawing by E.F. Milone, after Thom et al. (1973).

A considerable number of the stones of the site are known to have been reset. This led Hadingham (1976, p. 162) to declare, “Unfortunately, the enclosure was restored in the last century, so that no reliance can be placed on these remarkable facts.” Such a statement ignores the fact that the rectangle is one of the easiest shapes to restore with substantial accuracy. Moreover, the stones are large enough to make a great deal of natural movement unlikely. In 1882, only 9 of the 22 stones were standing but a plan of 1867 offers strong support for the view that the reconstruction was done very carefully (Burl 1985, p. 133). If the alignments of the stones are even roughly authentic, displacements along the lines would not greatly affect the remarkable properties of this rectangle. The short sides of the rectangle measure 30 MY, the long sides, 40 MY, and thus the diagonals, 50 MY. These are the proportions of the classic “Pythagorean triangle,” which, Thom has maintained, is basic to the construction of most megalithic structures. The absolute values conform with the unit of the Megalithic Yard as worked out by Thom from data of more complicated sites in Britain. The simple rectangle (the existence of the triangle is hypothetical but the rectangular is a manifest description) is in a way the best support yet published for the existence of that unit with a value very close to that given by Thom. The rectangle also possesses astronomical alignments, as it defines solar alignments at both solstices and equinoxes. The long sides are aligned east-west, and the short sides, north-south. Thus, the long sides point to both rising and setting Sun at both equinoxes. At this latitude (~47.5°), the diagonal line from the northwest points to the rising of the midwinter Sun, and the diagonal from the northeast points to the setting Sun on the same date. The diagonal from the southwest points to the rising of the midsummer Sun, and the diagonal from the southeast points to the setting of the midsummer Sun. The dimensions of the features of the Crucuno monument are matched to its latitude if its purpose was to provide solar alignments. Hence, the Crucuno monument furnishes simple but strong evidence of a deep megalithic involvement in astronomy. In terms of our understanding, this interpretation is both straightforward and complex. It suggests complete understanding of the regularities in the correspondence of the figure with the apparent movements of the Sun. It seems highly unlikely that a rectangle with these geometrical and astronomical properties would be set by chance on the line of latitude where they could work. The site is about 9 miles from le Grand Menhir Brisé, which should have been visible at Crucuno. However, we know of no suggestion of an astronomical relationship between the two, and they may not have been contemporary. There was once a somewhat similar rectangle, oriented E-W, at Lanvéoc on the the Crozon peninsula in Finistère. Unfortunately, the surviving drawing of the site is not adequate for determining its geometric or astronomical characteristics with precision (Burl 1993, p. 54).

The principal factor in determining the possible astronomical or calendrical features of a site is the presence of specified types of geometric or astronomical alignments supported or at least not contradicted by the geographical and archeological contexts. For example, if several stones point to a distant artificial foresight, and a standstill Sun/Moon set in a knoll as indicated by the foresight in the epoch (independently determined by archaeological data) when the declination permitted this to happen, and if the probability of a random physical alignment of the stones in this direction were vanishingly small, a moderately convincing case for an astronomical alignment will have been made.

6.2.4 Horizon Astronomy

The basic astronomical formulae are (2.1) to (2.4). Solving these equations for the case in which the altitude, h = 0, when an object of declination δ is on the horizon, we have approximate expressions for azimuths and hour angles of rise and set:

Recall that observations are made through Earth’s atmosphere, which both refracts and scatters light. Therefore, for accurate results, corrections to the altitude for elevation, dip, and refraction must be made to determine the true azimuth and hour angle of the object. Consult §3.1.3 for details.

Twice the hour angle of rise or set is the time spent by the object above the horizon; as far as we know, there is no direct evidence that this type of information was recorded in any way in megalithic times. The azimuth of rise or set, on the other hand, can be and probably was marked by sightlines directed to points on the horizon.

It is the reality and the purpose of the layouts of these sightlines that constitute the main subjects of debate in megalithic astronomy. In Megalithic Sites in Britain, Thom (1967) described 145 stone circles and other monuments, and the solar and lunar alignments. With subsequent works (Thom 1971, 1974, A.S. Thom 1984, among others), the claims were sharpened and better substantiated. But criticisms abound (many are summarized by Hicks 1984a). To see why, we first describe how alignments are measured, and then we discuss the types and results of such measurements.

Sightlines are best measured with a theodolite, a survey device capable of high-precision measurements of altitude and azimuth. A measurement of the altitude and azimuth of the Sun in the daytime, or of a star at night, made at a precisely recorded instant of time (necessitating a precision chronometer ), permits both longitude and latitude to be obtained (see §3.2.1). The determination of azimuth for surveying purposes requires that the direction of north (or some other direction) be known precisely, but measurements relative to some given direction can be made and later corrected. North can be established in several ways: by magnetic compass and application of the correction to true north; by measurements of the apparent azimuth and altitude of Polaris (or other appropriate stellar marker for previous epochs); by solar gnomon; or by bisection of the angle relating sunrise to sunset azimuths. Tables giving the differences in A and h between Polaris and the NCP at a given instant on a given date are provided in astronomical almanacs. To save time, the positions of landmarks that are included on accurate survey maps can be measured, and the true azimuth and altitude readings corresponding to all relative measures can be found later. The reduction requires tables of the right ascension and declination of the Sun for the dates of observation, and the star positions, corrected for precession. Corrections must be made for refraction and dip. Recall from §3.1.3 that refraction lifts the image of the rising or setting Sun (or Moon) above the horizon by an amount that is approximately equal to half a degree, but which varies with atmospheric temperature and air pressure, especially within the last few kilometers of the observer. The effect of refraction is to cause objects to rise sooner at smaller azimuths and set later at larger azimuths4 than they would without refraction. Similar corrections must be applied for measurements of the objects on the horizon (distant foresights) or closer objects. Dip causes the horizon to be depressed with effects similar to those of refraction, in the sense that lower altitudes are visible than with a flat horizon. Finally, there may be instrument corrections (scale and zero-point adjustments).

The kinds of alignment-checking measurements that can be made depend on the type and layout of structure being studied. In the relatively straightforward case of a shaft through a building (or tomb), its bearing can be measured by sighting along its length. In the case of two or more standing stones, the measurement of the bearing can be made by aligning the instrument along the stones. On the other hand, a ring of stones presents many more possible alignments and an accurate survey of the site may take considerable time. In some cases, mountains on a distant horizon may align a rising or setting object with a backsight (an object closer to the observer). In principle, such an alignment is purely arbitrary unless features of the local terrain, such as the side of a hill, greatly restrict the field of view, or unless the backsight is a suitable distance away from the observer. Both types of conditions serve to improve the probability that the alignment is intentional. Such an arrangement seems to have been found at Kintraw (Thom 1971/1973/1978, pp. 36–40; MacKie 1974), in the British Isles, for example, but even this site is not without controversy. The measurements, although sometimes inconvenient to make, are straightforward. Surveying is only the beginning, however. The interpretation is more difficult. Figure 6.5 illustrates an astronomical alignment involving distant foresights, backsights, and an observation of a setting Moon at maximum standstill. The more distant the foresights, the more potentially precise the observation, to the limits set by visual acuity and atmospheric conditions. Note that the “artificial foresight” could function as a backsight.
Figure 6.5.

An example of an astronomical alignment involving distant foresights and backsights, and observations of a setting Moon near maximum standstill. Drawing by E.F. Milone.

Once the measurement of azimuth of a distant foresight has been made and the corrections determined, the declination of an astronomical object conforming to the direction of the alignment can be obtained by the equations of §2.2.4. This result can be compared with stellar positions that have been precessed back to the expected epoch, to the Sun at specified times of the year, or to the Moon sometime within the nodal regression cycle.

As we described previously, Ruggles (1981/1982a,c/1983) reviewed the evidence provided so carefully by Thom and associates and strongly criticized Thom’s conclusions concerning the high precision of megalithic astronomical alignments. Nevertheless, Ruggles (1988a) concluded that alignments exist to three levels of precision:
  1. (1)

    At the highest precision, there is some evidence, although marginal, indicating a preference for six specific values of declination, δ, viz., −30°, −25°, −22.5°, +18°, +27+, and +33°.

  2. (2)

    At an intermediate level, there is a strong preference for alignments indicating −31° ≤ δ ≤ −19°.

  3. (3)

    At the lowest precision, alignments indicating the range −15° ≤ δ ≤ +15° are present but rare, perhaps avoided.


We now turn from general propositions to particular cases.

6.2.5 Brittany

In Brittany, we find early passage graves made from gigantic rocks and covered with earth. Most of these show a clear preference for an orientation of the passage to the southeast. A few show a precise alignment to a position that we can regard as astronomically significant. Burl (1985, pp. 23–24) suggests that only the general orientation to the southeast was important and that within that range, the precision of an alignment, for example, to the winter solstice, was purely accidental. This is a solution that has found substantial favor with many archaeologists, not least because it relieves them both of the necessity to understand the astronomical evidence, and of the tedious labor of checking alignments, horizons, and the possible astronomical-geometrical relationships of the site being studied. But there would seem to be at least two other possibilities. One is that the bulk of the population being buried in these great tombs was only interested in a general orientation, but that a small minority was deeply interested in great precision and that these are the people responsible for such alignments as that of Dissignac, Brittany (to be discussed later in this section), to the winter solstice rising Sun. The second possibility is that most or all of the population was interested in a variety of different precise alignments that were important to them but lack significance for us. At the moment, we know of no objective external criteria that would allow us to choose between these alternatives.

The earliest case that involves a precise alignment is found in one of two passage graves at Dissignac in Brittany. This burial monument has a corridor so oriented that the rising winter solstice Sun illuminates the burial chamber. Both graves are in a common mound that was built in several stages, and the graves may be of slightly different dates. Mohen (1990, p. 304) assigns the initial construction to about 4500 b.c. The tomb having the alignment lies to the southwest, with a 7-m long passage leading to a rectangular chamber, which contained many pieces of broken, extensively decorated pottery, beads, and a range of stone artifacts (Burl 1985, pp. 98–99). The entrance capstone has representations of shepherds’ crooks and axes, appropriate for a small farming community. The northeast tomb has a similar passage, but with different orientation and with a bend that prevents light from entering the inner chamber. At this writing, early in the 21st century, this site appears to mark the beginning of the archaeoastronomy record.

One of the more interesting Breton sites investigated thus far is the passage grave at Gavr’inis. One of the lintels of this monument is a broken piece of a giant menhir, which once stood 14 m high (see Figure 6.6). Another piece of the same menhir is built into the Table des Marchands tomb at Loc Mariaquer.
Figure 6.6.

The large menhir, broken pieces of which have been incorporated into the lintel of the passage grave at Gavr’inis and in the tomb of the Table des Marchands at Loc Mariaquer. Note the enormous scale of the complete menhir and its elaborate decoration. Drawing by Sharon Hanna.

The latter site is also the location of another menhir, the tallest one known, Le Grand Menhir Brisé, which would have stood 20.3 m high above its base when erected. It has been argued that this monument was so large that the people who tried to erect it were unable to do so and that it fell and broke. The discovery of another broken menhir of the same class and similar material at the same site suggests that both were broken in the same way, which is much less likely to be due to technical incompetence in both cases. Both Burl and Mohen have separately suggested that both monuments were deliberately torn down. However, the differing positions of the broken parts of Le Grand Menhir Brisé constitute a good argument for the view that the monument was toppled by earthquake (Thom and Thom 1978a, p. 98ff). Interestingly, Burl and Mohen draw opposite conclusions about the incorporation of part of a broken menhir into a passage grave. Burl (1985, p. 109) suggests that this “is evidence of the indifference prehistoric people could have for the handiwork of earlier societies,” whereas Mohen (1990, p. 172) writes, “Using blocks again in other monuments symbolizes the endurance of a cult whose rites would suffer complete destruction of some of its sites.” In any case, the presence of two broken giant menhirs in the same area is strongly suggestive that the same event was responsible for breaking both, whether due to deliberate human action, as Burl and Mohen suppose, or by natural events, as others have thought.

Thom and Thom (1978a, pp. 100–102) had presented the hypothesis that Le Grand Menhir Brisé acted as a foresight for megalithic astronomers gathering information on the movements of the Moon and suggested the identification of a number of sites that could have functioned as observing points for making these observations. From an archaeological viewpoint, the best support for such a view would have been some identification that the sites involved formed a complex. If they were all the same kind of monuments and all derived from about the same period, the hypothesis would have been supported. However, Burl (1985, p. 136) argues that three of the postulated backsights were probably not Neolithic monuments at all. He further argues that an alignment to a mound at Tumiac was inaccurately determined by Thom (if true this would be a rarity), and that the remaining monuments, menhirs, mounds, and passage grave are dissimilar in both architecture and date. Thom’s postulated astronomical date of 1700 b.c. is later than some of the monuments. The most damaging case against the universal foresight hypothesis is that Le Grand Menhir Brisé was probably toppled before some of the other monuments were constructed and long before 1700 b.c. On the basis of 14C dating (see §4.3), Burl (1985, p. 108) suggests that Gavr’inis was built “in the centuries around 3500 b.c.” The hypothesis demonstrates that it is easy to be misled when there is a plethora of monuments. A straight line, after all, requires only two points, and there is a high probability that any line drawn through the monument laden region will pass over some monument. A line from a particular monument to a specifed point on the horizon can almost certainly be extended backward to pass over some monument from which the foresight could have been seen.

The Gavr’inis site, however, has considerable importance beyond its role in the story of the great menhirs. As at Dissignac, the passage grave at Gavr’inis is illuminated by the winter solstice Sun. Burl (1985, pp. 110–111) notes an added interesting difference between this site and Newgrange (described in §6.2.6):

Looking from Stone 19, at the left-hand entrance of the chamber, toward Stone 1, the bearing is 128°, almost perfectly in line with the midwinter sunrise. The main axis of the passage is 134° towards the low-lying Arzon peninsula and the orientation is close to that of the major southern moonrise. It has been calculated that the two alignments, one solar, the other lunar, intersect halfway down the passage level with Stone 7, the white quartz slab whose undecorated surface may have been illuminated by the light of the rising Sun and Moon.

Gavr’inis is among the most lavishly decorated of the Breton tombs. Representations of long-horned cattle abound on the earliest Breton monuments, and the menhir that became the Gavr’inis capstone shows representations of such cattle. These have often been called oxen or bulls, but if the Breton cattle were similar to the contemporaneous Egyptian breed, the exceptional length of the horns would signify cows rather than bulls.

There is also extensive decoration on the companion passage grave at Le Table des Merchands. Here, however, it may be notational. Müller (1970, pp. 107–108) has drawn attention to the 56 shepherd’s crooks, divided into 29 on the left and 27 on the right, and 19 accompanying curved elements, suggestive of the nearest whole number of days of synodic and sidereal months and of the Metonic cycle. At present, however, there is otherwise little evidence to connect the symbols with the cycles. Moreover, scholars disagree on what is being depicted. Burl (1985, p. 136) sees 53 crooks and an “anthropomorphic figurine,” and Twohig (1981, p. 97) sketches the decoration differently. Whatever the outcome of this dispute, the motifs and style of the Breton passage grave decoration are closely paralleled at Newgrange and Knowth in the Boyne Valley of Ireland, and it is generally accepted that there is a close historical connection. The orientation to a winter solstice sunrise is common to Gavr’inis and Newgrange and suggests that both alignments are indeed intentional and arise from convergent or parallel traditions5 that stem from an earlier passage grave tradition in both areas.

6.2.6 Brugh na Boinne

The remarkable complex at Brugh na Boinne (Figure 6.7) provides the best evidence of a contextual (rather than a statistical) nature for extensive and precise interest in astronomy. The three great mounds of Newgrange, Knowth, and Dowth and associated monuments are for the most part intervisible, and together, they form a kind of massive record, which we may only now be starting to understand. Eogan’s (1986) work at Knowth has revealed two massive passage graves, one aligned to the east (and so to the rising equinox Sun), in which the sunlight directly penetrates the main burial chamber, the other aligned to the west (and the equinox setting Sun), with sunlight penetrating far down the passage but prevented from actually entering the burial chamber by a bend. The bent shaft is a feature reminiscent of the second passage grave at Gavr’inis. The kerbstones around the monument and many interior decorated stones provide the most extensive collection of megalithic art yet known, thanks to Eogan’s work. At Newgrange, Michael O’Kelly (1982/1989) meticulously reset stones and cleaned and reconstructed the mound, while carefully recording information on the building sequence, construction techniques, and chronology of the mound, which stands more than 45 ft high and 300 ft across. Its quartz crystal facing glistens again in the Sun, making Newgrange one of the most visually impressive of all the megalithic monuments (Figure 6.8). He discovered a “roof box,” a 62-ft long window shaft over the top of the entrance passage that still lets the rays of the rising winter solstice Sun illuminate the great corbel-vaulted chamber almost6 as it did 5000 years ago. The discovery has helped to convince archeologists of the deliberate and accurate nature of the astronomical alignment of this passage, which was first demonstrated by Patrick (1974a,b) and refined by Rea (1988), who showed that in neolithic times, the sunrise rays could have penetrated far into the interior, illuminating a three-leafed spiral on the back wall.
Figure 6.7.

The funerary complex in the Boinne Valley, Ireland. It consists of three great mounds—Newgrange, Knowth, and Dowth—and associated monuments. Dra­w­ing by Sharon Hanna.

Figure 6.8.

A photographic mosaic of the front of the quartz crystal facing of the passage grave monument at Newgrange, one of the most visually impressive of all the megalithic monuments. Photos by E.F. Milone.

Although the carbon-fourteen data from Newgrange suggest a date about 3100 b.c. for the construction of the mound (O’Kelly 1989, p. 351), a similar date from Knowth antedates the construction of the main mound there, suggesting that Knowth may be slightly later than Newgrange (O’Kelly 1989, pp. 109–110). Interpretations by Brennan (1983), whose activities and lack of meticulous surveying and incomplete documentation of his work have led to strong personal differences with archaeologists who have worked in this area, nevertheless provide some important insights. Three passage graves show alignments related to the winter solstice: the sunrise of Newgrange, the sunset at Dowth, and a possible noon alignment at Mound K—a badly damaged structure that nonetheless still indicates that the passage must have had a very low roof, which would have allowed sunlight to enter the chamber only at midday, when the Sun crosses the celestial meridian. At such a time, it will be remembered from earlier chapters, see, especially, Figure 3.17, the Sun is due South, and at the highest point of its diurnal arc, at an altitude given approximately by

At midwinter, the solar declination is equal to the negative of the obliquity of the ecliptic, δ = −ε ≈ −24°; the latitude at Dowth is φ = 53°41′, so we find h ≈ ~12°. At the main mound of Knowth, a cruciform chamber is lighted by the rising equinox Sun and an angled passage grave is illuminated by the equinox setting Sun, to the point of the bend. The unique property of these passage graves is the unambiguous direction of the alignments. The shaft subtends a small solid angle on the horizon, and that angle is small due to the lengths of the shafts providing the alignments. Slight changes in the direction of the Sun cause different portions of the passageways and the burial chambers to be illuminated, giving high precision to the astronomical alignment. These passage graves provide strong evidence that precision in alignment direction was of direct concern to the people who built these structures, and so make at least plausible some of the arguments by Thom that distant foresights were used to accomplish similar precise alignments of other structures.

At Newgrange, 31 of the 97 kerbstones are known to be decorated, but only about 1/3 of the stones have been completely exposed (M.J. O’Kelly 1982, p. 15; C. O’Kelly 1973/1978/1982/1984, p. 152). Some are decorated not only on the exterior, but also on the side facing the mound, which would have been invisible at any time since the mound was constructed. The alignment of the passage grave is marked by kerbstone 1 (K1) at the entrance and by kerbstone 52 directly opposite. Both of these are bilaterally divided down the middle. K1 has a group of clockwise spirals on the left of the dividing line and a group of counterclockwise spirals on the right (see Figure 6.9).
Figure 6.9.

Views and close-ups of the spirals on kerbstone 1 and on the opposite side kerbstone of the Newgrange passage grave. Photos by E.F. Milone.

Brennan suggests that the spiral marks the passage of the rising Sun on successive days northward along the horizon. The horizon movement, as we have described extensively (§2.3.1), continues northward from winter solstice until summer solstice, when the movement comes to a stop (the “solstice”) and then reverses. We will return to this spiral motif again and again; it is a motif shared by many cultures.7 Brennan suggests that the CCW spiral describes the northward movement, and the CW spiral describes the southern. Other stones, some not in alignment with extreme or equinox positions, are also marked with spirals, making the interpretation less convincing.

Surrounding the mound is another interesting feature, a ring of standing stones. Figure 6.10 shows the stones in a counterclockwise pacing around the northeast arc of these stones.
Figure 6.10.

Portions of the ring of standing stones around the passage grave monument at Newgrange: three successive views proceeding CCW around the entrance. Photos by E.F. Milone.

MacKie (1977a, pp. 72–73) regards the ring of stones as half ellipse and half circular arcs, with “the arcs of circles centered on the corners of two opposed right-angled Pythagorean triangles.” However that may be, in ~2015 b.c. [Sweetman’s (1984) date for the stone ring], Stone 1 of the ring cast a shadow onto the three-leafed spiral on the kerbstone in front of the entrance, K1, at the winter solstice. Prendergast (1991a,b) confirmed this circumstance (with the standing stone most directly in front of the entrance, GC1 in his designation,8 casting a shadow, for a 20-minute interval after dawn, to K1’s three-leafed spiral only for −23°56′ ≤ δ ≤ −23°22′, and found solar declination ranges for the shadows of other stones on the spiral as well: that of GC-2 for −01°14′ ≤ δ ≤ +01°26′, and that of GC-1 for −12°53′ ≤ δ ≤ −09°51′, corresponding to equinox and midseason declinations. Other determinations were: CG11 to GC7 aligned to δ = −23°49′; GC5 to GC3 to a midseason δ = +11°33′; and GC1 to GC-2 aligned to +23°15′, a possible summer solstice alignment). Figure 6.11, from Prendergast (1991, Fig. 5), shows the shadow play on K1.
Figure 6.11.

Simulation of shadow effects on kerbstone 1 by standing stones in the ring at Newgrange (Prendergast 1991, Fig. 5, modified by Sharon Hanna): (a) Shadows cast by standing stone GC1 at winter solstice over a 20-minute interval starting at sunrise with a solar declination −23°56′, appropriate for 2015 b.c. (b) Shadows cast by stone GC-1 when the Sun had a mid-declination of −11°53′. (c) Shadows cast by stone GC-2 at equinoxes (declination 00°31′).

Claire O’Kelly (1982/1984, p. 149) argues that K1 and K52 were in position before the neighboring stones were placed—possibly before the bulk of the monument was constructed. This raises the possibility that the ring predates the mound, and that stones of the ring bearing known alignments were used in laying out the passage grave. In fact, the size of the mound would have precluded any use of the ring in terms of alignments. If some of the kerbstones were used in conjunction with shadows of the ring stones in laying out the passage grave, the symbols on the interior faces of the kerbstones may become explicable. K13, possibly used in summer solstice sunrise-winter solstice sunset alignments, is extensively decorated on both faces, so that it is possible that it was intended to be used in connection with the construction of the mound and the correct placement of its reciprocal, K67. Later stratigraphic work by Sweetman (1984) suggested that the standing stones were erected a millennium after the positioning of K1. Even if true, and the alignments with K1 are somehow fortuitous, the internal alignments among the standing stones remain. Prendergast (1991) also estimates the probability that the apparent solar alignments are due to chance placements of the standing stones and K1. He uses the uncertainty in azimuth, σA, to produce a probablility of a random alignment in the sector of interest of p = ∑ 2σA/180°, that is, the sum of all ranges of angle ±σA centered on particular azimuths. He follows Schaefer (1986) to obtain an overall probablility that the apparent astronomical alignments are significant.9 If 1 represents the number of such apparent astronomical alignments, and N is the total number of lines between the stones at the site, then the probability that the alignments are due solely to chance is

He concludes that the probability that the Newgrange alignments among the stones are due to chance is ~0.01. Note, however, that this approach is hardly foolproof in general; as the errors in the alignments, σA, become very large, so does p, and as p approaches 1, P(1) approaches 0. Here, however, σ can be assumed to be at most tens of arc-minutes.

The second of the great funerary centers in the Boinne complex is Knowth. At Knowth, 90 of the surviving 123 kerbstones are decorated externally and 11 are decorated on the interior side (Eogan 1986, p. 150). The structure of the principal mound, the placement of its kerbstone elements, and their ornamentation can be seen in Figure 6.12.
Figure 6.12.

The grave monument at Knowth: (a) Diagram of the structure shows the kerbstone placements. Drawing by Rea Postolowski and Sharon Hanna. (b) Examples of ornamentation on the kerbstones related to their orientation at Knowth. The frequent use of spirals in funerary monuments probably implies a profound and widespread association between the human life cycle and the cyclical movements of the Sun and Moon. Drawing by Sharon Hanna.

Calendrical concerns are suggested in the spirals and perhaps temporal concerns in what appears to be the face of a vertical sundial at the top of the fifth kerbstone to the south from the eastern entrance (Figure 6.13).
Figure 6.13.

Ornamentation on a kerbstone at Knowth, as it appeared on a visit to the site in August 1990. Spiral structures strongly suggest solar and calendrical concerns, and the radiating lines from the central shallow depression invoke the features of a vertical sundial. If it was indeed used as such, it would be the earliest known sundial. Photos by E.F. Milone.

The radiating lines and the two depressions (albeit very shallow depressions) resemble those seen on a flat Egyptian sundial found at Luxor and now in the Ägyptische Museum, Berlin (Clagett 1995, Fig. III.56; see Figure 4.5). The Luxor sundial supported a weighted string that indicated the vertical and presumably a horizontal stylus to provide a shadow. If the Knowth kerbstone is indeed a sundial, it may be the earliest known example. A gnomon for such a device could have been horizontal with solid footing on the ground and two fingers to steady it in the two depressions. Because the length of a gnomon’s shadow at any measured instant can give seasonal information, such a sundial might have been useful in the placements of some of the other kerbstones. The markings on the northern half of the Knowth dial include eight major points and eight minor points. This is exactly the sort of division one would expect if the 16 calendar divisions postulated by Thom (1967/1972, pp. 109ff) have validity, but if this were a sundial, it would antedate by more than a thousand years the alignments that led Thom to postulate such a calendar.

It is noteworthy that the kerbstones at the entrances to the two passage graves are also divided down the middle, like those at Newgrange. A bisecting shadow is cast onto the exterior face of the kerbstone at the west entrance at the equinoxes. Brennan has written at length about the possible interpretations of many of these kerbstones, but other than calling attention to the markers, he has not made any explicit attempt to determine the relationship between the “decorations” (or possibly notations) and associated alignments. The concept of a notational system, intermediate between iconographic symbolism and true writing is not one that most megalithic scholars have had occasion to consider prior to Brennan’s work, and Brennan does not phrase his studies in these terms. The nature and implications of such a system are treated most fully, to our knowledge, in a work by Langley (1986) relating to repeated symbols in similar contexts in the art of Teotihuacan. Many of his general remarks are probably applicable here.

Of all Brennan’s interpretations other than for the previously mentioned spiral, DHK finds that K52 (in Eogan’s numbering; SW22 in Brennan) is the most convincing (Figure 6.14).
Figure 6.14.

Ornamentation on kerbstone 52 shows possible lunar representations. Drawing by Sharon Hanna.

It displays a series of 22 crescents and 7 circles, features that could be stylized representations of the Moon over a synodic month. A spiral (representing the moving Sun?) cuts through three crescents and appears ten times on the right and nine on the left. Brennan also discusses possible meaning for other features on the kerbstones, such as wavy lines. Although seemingly ad hoc, these ideas merit consideration because they are applied in a context in which alignment and shadow effect appear to be genuine. One possible alternative has been suggested. Stooke (1994) maintains that the “crescents” are actually fairly realistic images of the maria of the Moon rather than stylistic images of a crescent Moon. We find this unconvincing.

Dowth is the third great funerary site in the Boinne valley complex. The layout is seen in Figure 6.7. Examples of ornamentation at the exterior and interior of this site can be seen in Figure 6.15. Note the spirals on the entrance kerbstone and circle and starburst symbols in the interior.
Figure 6.15.

(a) Exterior and (b) interior views of a tomb at Dowth, Ireland. Note the spiral on the kerbstone and the circle and starburst symbols within. Photos by E.F. Milone.

Differences of motifs have been magnified into separate “styles” in megalithic art to an extent that seems baffling to an outsider. DHK does not think that the decorative symbols should be classified into separate styles unless there is objective external evidence of differences of distribution that can be linked clearly either with specifiable geographic areas, time periods, or both.

The monuments in the bend of the Boinne have been referred to as a complex. This might be taken to imply that all were built at about the same time, or that a major and continuing plan was in effect. In a more limited sense, it must at least imply that similar or identical governing concepts determined the placement of newer features in ways that were congruent with those already there. We think that even a cursory examination of the layout of the area supports the view that it is, indeed, a complex in at least the latter sense. The east-west line through mound U to the great Newgrange mound, to mound L (off-center) and mound K, may be such an indication because it is parallel (as any east-west alignment must be) to the equinoctial alignment of Knowth. Another may be the north-south line through the Newgrange mound and the earthen ring or henge to the south, paralleling an alignment involving mound K. Although one is disposed to give weight to the argument that astronomical alignments were being replicated, somewhat higher weight could be given to less obvious geometric factors that influence the placement of monuments. For example, the alignments through the Dowth mound to the NE and through the Knowth mound to the NW are both at angles between 31° and 34° north of the east-west line through Newgrange. Each is on top of a separate ridge, and this purely topographic feature seems to be of central importance in determining the relative distances. More details can be found in Brennan (1983, pp. 70–71) and O’Kelly (1982/1984, pp. 83–84), where reference is made also to work by J. Patrick.

Not only is the Brugh na Boinne complex well marked in astronomical and artistic contexts, but it is also the only set of megalithic monuments in Europe that is directly relatable to pre-Christian myths and clearly incorporate astronomical motifs. However, words of warning are in order. We know of no one who would maintain that Celtic mythology is an unmodified representation of the views of megalithic farmers.

Even if existing stories and scraps of information allow us to glean some aspects of belief in the late centuries b.c., it must be remembered that Celtic warriors in their chariots at that time were as far removed temporally from the builders of Newgrange as we are from those Celts. Nonetheless, Brugh na Boinne plays a role in several Irish myths, which may well retain in modified form older beliefs. They should be considered in any attempt to determine the meaning of the site. In Irish, the common name for a mound is brugh, “mansion” and the different mounds are assigned to different gods. The mound of Newgrange is assigned to Oengus Mac nOg, spelled and interpreted in various ways. O’Rahilly (1946/1957, pp. 516–517) argues convincingly that Mac nOg is the Gaelic equivalent of the god known from British inscriptions as Maponos, and there identified as Apollo, although one need not accept all the linguistic details of O’Rahilly’s argument. The name Maponos means “youth,” whereas Mac nOg seems to incorporate the words for “boy, son” and for “young.” Such names are typical for sun-gods, born again each morning, so that the aforementioned equation with Apollo may be unnecessary. The sun god is also normally thought of as lord of the year, in which role he is to be born at the winter solstice [cf. Fraser (ed., T.H. Gaster) 1959, p. 633]. The appropriate nature of the association of Oengus Mac nOg with Newgrange is clear. The father of Oengus was Dagda or “good god” who seduced Oengus’s mother, Boand, away from her husband, who was variously called Elcmar, Nechtan, or Nuadu of the Silver Hand. Boand is the later version of the name of the river goddess Buvinda, “White Cow”—the name of the presumed moon goddess whose name was given to the river Boyne.

O’Rahilly thinks that Necthan is a Gaelic name cognate with the Latin Neptune and is a byname of Nuadu; the latter name appears in Britain as Nodons (O’Rahilly 1946/1957, p. 321) and is surprisingly equated with Mars in one inscription (O’Rahilly 1946/1957, p. 527), although O’Rahilly equates him with the sun god and the “Otherworld-god of the Celts.” The “silver-hand” suggests a linkage with Tsiw, the Germanic war god, whose hand was bitten off by the Fenris Wolf. Tsiw’s name is incorporated in our Tuesday, the day of Mars in the planetary week. Elcmar or Elcmaire is supposed to have possessed the Brugh na Boinne before Oengus. An Irish tale tells how Cuchulainn, the “Hound of Cuala,” “speared a salmon in the Boyne and then mutilated Elcmaire, who had entered the river to oppose him.” (O’Rahilly 1946/1957, p. 320). The salmon that Elcmaire attempted to protect may have been the “Salmon of Wisdom,” supposed to have resided in a pool at the head of the Boyne, but the complexities of identification are not adequately treated by O’Rahilly. Celtic lore also knows sun goddesses, of whom the greatest is Grian, “sun,” and another is probably Etain who lives in a crystal grianan or sun-house (O’Rahilly 1946/1957, pp. 287–293)—a description that once would have fitted Newgrange well.

Eogan (1986, p. 20) reproduces from the Seanchas na Relec (the History of the Cemeteries) a list of burials of the Tuatha de Danann,10 the people of the goddess Danu. One of the burials is the “Caisel” (castle) of Aengus, probably Newgrange. Another is the grave of Boinn, the goddess herself; it would be interesting to know which (if any) of the mounds was particularly associated with her name. Features of the area included the “paps of Morrigan,” the war goddess; the mound of Tresc; the mound of the Bones; the cave of Buailcc Bec; and the items associated with Esclam, Aedh Luirgnech, Cirr, Cuirell, Cellach, Cineadh, and “the pillar stone of Buidi, ‘where his head is interred’ ”. The names are more or less obscure, and there is no direct evidence for tying most of them to particular mounds, but the list may yet become important in suggesting interpretations of astronomy. At the moment, the most that can be said is that the gods were associated with the mounds and that at least some of the associated gods are planetary, but we have no evidence for planetary alignments at these sites, nor perhaps could one readily expect them. Planetary movements are much more complex than are those of the Sun or even the Moon due to the effects of the relative motions of planet and Earth, causing variations between synodic period intervals.

Nowhere else do we seem to have the degree of mythic relevance that appears with Newgrange and the Boyne complex, although a considerable amount of lore is known that associates mounds with Irish gods, kings, and heroes. Certain other sites reveal more direct information about alignments and associations with possible notational art.

The Loughcrew complex (Sliabh na Caillaighe or Sliabh na Caillighe (Ríordan 1979), Slieve na Calliagh (Harbison 1970/1992), “Hill of the Witch”) is located in the same county (Meath) on high ground 40 km west of the Boinne complex with magnificent and commanding views of the surrounding lush green hills. The burials are centered on four hills more than 200 m in height. Loughcrew had as many as 30 mounds in the last century but only a few are relatively well preserved (see Figure 6.16).
Figure 6.16.

Sectional panoramas of the Loughcrew (Sliabh na Caillaighe) burial complex, Ireland, as seen from Cairn L, on the westernmost hill of the four-hill complex: (a) Montage of views to the west through south, with a close-up of the southwest horizon. Note the menhir, a potential foresight for midwinter sunset. (b) Further views from L show cairns in various conditions of preservation. Photos by E.F. Milone.

As at Brugh na Boinne, the mounds contain a considerable amount of art, but it is freer and cruder than is the Boyne complex. It has been suggested that it is either a prototype or a provincial copy. Because we have no external evidence about the chronology, this problem cannot be resolved at present. There are some very revealing items at Loughcrew even now. The most important of the mounds is Cairn L (Figure 6.17). This is aligned 9° south of east, according to Brennan, but the Sun shines in at the equinoxes at a time when it is substantially above the horizon. However, the declination of the rising Sun can be slightly different at the sunrise closest to the equinox from year to year and from spring to fall,11 and therefore, the shape of the light beam that actually reaches the chamber can be subtly different. Brennan’s information is that as the Sun rises higher and higher, it successively illuminates different parts of the stone, framing first one, and then the other. If the evidence is accurately presented there would seem to be a relationship between the light beam and the design elements delineating the equinox.
Figure 6.17.

Cairn L of the Loughcrew complex (Sliabh na Caillaighe): (a) Interior views. (b) Exterior view from the entrance (upper), and view from inside toward the entrance. Photos by E.F. Milone.

As in the Boinne complex, one small mound faces due south and hence marks mid-day, but whereas the low roof of the Boinne monument would have stopped the light beam from entering except near the winter solstice, here a floor cut below ground level apparently allows light to enter even when the Sun is at its highest, near the summer solstice. It is interesting to note that at summer solstice, a shadow bisects the symmetrical design of Stone 8 of Cairn U (Brennan 1983, p. 87), made possible by the destruction of the roof of the cairn. As at Newgrange, the summer solstice marker would have been visible only at an early stage of the construction of the mound and invisible thereafter.

Two further points can be derived from this material. One is that many of the mounds present unexpected or un-explained alignments. There seem to be alignments to the cross-quarter points of the horizon. A set of four mounds have successively changing alignments. Although some may mark solar calendar divisions, some do not seem to fit any scheme that we can recognize. The second point is that one important series of mounds is apparently aligned on the summer solstice rising Sun, although the individual passages have different alignments. The situation is like that of the east-west alignment at the Boyne complex, and it supports the inferences drawn from that complex.

Other Irish sites described by Brennan have passage grave alignments to solstices and both astronomical-calendrical investigations, and examination of design motifs should be considerably extended beyond what has been done already. Because of the paucity of human skeletons at these sites, and the alleged use of chambers to determine when phenomena would take place, Brennan called them “observatories.” In no normal sense does the use of this term appear appropriate. The passage graves are oriented to already known phenomena, appearing once or twice a year or, perhaps, less. They could serve as calendrical markers, and they may well have served to indicate when particular festivals or rituals should occur, but this hardly requires the use of an observatory. Burl has suggested that the importance of ritual was beginning to outweigh that of burial. However, we cannot accept the view that the burials in these structures were of secondary importance to the astronomical alignments. They are, rather, inseparably linked. Although it is clear that intrusive burials continued to be made in some of the graves at least until Christianity became the dominant religion, the archaeological evidence both in terms of stratigraphy and of associated grave deposits is equally clear that many of the burials are primary deposits made at the time of construction of the monument. The use of the technical term “secondary burials” to indicate that the burials were not made immediately at the time of death may give nonarchaeological readers a mistaken impression. These were mass burials, and in Ireland, the bodies were normally cremated before burial. Thus, they may represent an accumulation built over several years. There is in fact every reason to think that the primary purpose of the monuments was to be graves. The astronomical associations of the winter solstice with the death and rebirth of the Sun, found in so many other cultures, may have been held by the builders of passage graves also. Similar relationships of gods and humans could be postulated for other alignments.

There are, of course, monuments other than passage graves for which it would be difficult even to postulate as much about the ritual activity and beliefs as has been suggested from passage graves. The burial mound of Newgrange may have been built inside an already erected stone circle, but there are contrary opinions. If it were erected earlier than the mound, however, it could have been laid out with a peg and a rope. This in turn could help explain why its diameter, 103.6 m or 125 MY, is the same as at Brogar and the inner rings of Avebury (Burl 1976, p. 71). At about the same time that the Newgrange burial mound was being built, a major set of stone circles was being built in Cumbria.

6.2.7 The Cumbrian Circles

Cumbria (formerly the counties of Cumberland and Westmorland) is in the extreme northwest of England, roughly bounded on the east by the Pennine Mountains, and the dominant type of circle and henge found here is known as the Cumbrian style, after the ancient district name. Over 50 stone circles are known in the Cumbrian region, but nine extremely large ones are known from the Lake District, “a part of Cumbria where virtually no other Neolithic site exists.” There are also two in the same style in Scotland and one in Ireland. Table 6.2, adapted from Burl (1988, pp. 188, 203–205) lists the sites. These may be the earliest megalithic monuments known from England (perhaps as early as 3400 b.c., Burl 1988, p. 184), and they seem to be associated with the manufacture and distribution of polished stone axes (Burl 1988, pp. 183–184); Cumbrian axe factories were apparently also flourishing by 3400 b.c. Burl (1988, pp. 183–186) suggests that one of the purposes of these great stone circles was as a meeting place where people could trade for axes, a sort of market on neutral ground to which outsiders were allowed access at specific times. This may have been particularly important because this seems to have been a time of deteriorating climate caused by volcanic eruptions (3250 b.c. ±80; tree ring date ~3190 b.c.: Burl 1993, p. 30). The association of axes with megalithic monuments is even more striking in Brittany, where axes seem to play a ritual role, and it is widespread in Europe. This accompanies the even more widespread idea that stone axes are thunderbolts. There is also an association (possibly resultant) of stone axes with gods of thunder and lightning. It does not seem implausible to suppose that this association may have been already present at this time, as Burl (1976, pp. 81–182) thought. The further alleged association of thunder-axes with the Sun seems forced and unlikely unless solar phenomena were being correlated with planetary movements. If there was an astronomical connection of the axe cult, it is more likely to have been with the planet Jupiter, as known from the Mediterranean (cf. the Roman god Jupiter Tonans, “Thundering Jupiter”) and arguable for Scandinavia.12 Such an association, however, is not demonstrable for the megalithic sites.
Table 6.2.

Cumbrian circles.a

Name or location


Source of dating information

Castlerigg, Keswick


Stone axe

Long Meg, Eden River east






Grey Yauds



Brat’s Hill



Elva Plain



Shap Centre



Shap South



Grey Croft


Stone axe




Ash-House Wood



Broomigg A












Kirk, Kirkby Moor



Blakely Raise



The Beacon



Shap North



Lacra B









White Moss NE



White Moss SW



Gretigate NW (B)



Gretigate NE (C)



Low Longrigg SW



Low Longrigg NE



Moor Divock 4


Food vessel

The Cockpit



Lacra D



Bleaberry Haws



a Adapted from Burl (1976, Table 2).

b Burl’s chronological ordering, based largely on early and late traits.

Cumbrian monuments include a predominance of flattened circles (as opposed to circles, egg-shaped ovals, or ellipses), suggesting a consistent style. The best known monuments of this region are Castlerigg (or Castle Rigg) near Keswick (Figures 6.18 and 6.19), and Long Meg and her Daughters, East of the Eden River, near Penrith (Figure 6.20).
Figure 6.18.

The layout of the Cumbrian circle of Castlerigg, near Keswick, England. Drawing by Sharon Hanna.

Figure 6.19.

Sectional panoramas of the Cumbrian circle of Castlerigg, England: (a) View toward the entrance from the North. (b) Looking ENE across a rectangular enclosure—the largest hill on the right is the Threlkeld Knotts, over which the Sun appears at the equinoxes. (c) Three views of the rightmost of the stones in (b), and the largest in the circle. Photos by E.F. Milone.

Figure 6.20.

Sectional panoramas of the Cumbrian circle of Long Meg and Her Daughters, near Penrith, England: (a) to (c) As viewed from the center, proceeding clockwise from the South. (d) As viewed from outside the eastern arc of the circle. Long Meg lies along a line to winter solstice sunset. Photos by E.F. Milone.

Castlerigg is a flattened circle where the geometry and astronomy seem especially well blended (Thom 1966, 1967). Modest in scale, the diameter of the ring is ~97 ft (29.6 m), and the height of the tallest stone is ~7.4 ft (2 1/4 m). The horizon is very uneven.

Long Meg is located on the track to the Tyne Gap in the Pennine Mountains that provides access to northeastern England. The ring is a flattened circle, roughly 361 × 305 ft (109 × 93 m), the largest of the Cumbrian sites. It has a partial henge associated with it. A double-stone entrance is at the southwest, and 18 m beyond in the same direction stands the outlier stone, Long Meg (Figure 6.21), 3.4 m high, weighing ~28 tons (Burl 1976, p. 89).
Figure 6.21.

In this view, Long Meg presents the image of a leader addressing a gathering, or perhaps a teacher and her class or a matriarch and her family. Photo by Sharon Hanna.

According to Burl (1976, p. 92), the bearing of Long Meg from the circle’s center is 223.4°. At a latitude of 54.7°, the azimuth of sunset with the Sun at δ = −24° is actually 227.8° for a level horizon, but Long Meg is the tallest of the stones, 3.7 m high and is moreover set at the top of a ridge. These circumstances mean that the Sun will disappear over the stone at a smaller azimuth by an amount that depends on the relative heights of the observer and the stone. The association with an observational solar function goes beyond the bearing, however. On the face toward the ring, Long Meg bears three markings: cup and ring, spiral, and a series of concentric circles, of which the outer two are incomplete (see Figure 6.22). The spiral is counterclockwise, a feature that has been associated with the winter solstice. From the center of the circle, the outlier, Long Meg, stands in line with the midwinter sunset (Burl 1988, pp. 196–197).
Figure 6.22.

Rain-darkened markings on Long Meg. There are three markings on this monument. (a) The entire stone; (b) cup and ring, and spiral; and (c) concentric rings. The spiral is poorly visible in (b) at the lower right. Photo by E.F. Milone.

There was once a great circle of this group at Lochmaben, Dumfries, near the coast; now only two stones remain. Charcoal associated with one of the remaining stones gave a date of 3275 b.c. ±80. The site is mentioned by the Ravenna Geographer of late Roman times as Locus Maponi (from which the present name derives), and the Roman name derives from Maponos, the British-Celtic form of the name of the god known in Ireland as Mac nOg and associated with Newgrange (approximately contemporary in construction with Lochmaben) (Burl 1976, p. 205; O’Rahilly 1946/1957, pp. 292–293). One of the mythical figures associated with King Arthur is Mabon, son of Mellt, i.e., *Maponos,13 son of *Meldos. Here we begin to see a relationship of the Sun god with the god of thunder and lightning, for *Meldos meant “lightning.” The Irish Mac nOg is said to have been a son of the Dagda or “Good God.” Mac nOg tricked his father out of his rightful possession, the Brugh na boinne (O’Rahilly 1957, pp. 52, 516–517). This association of Sun god and lightning god is definitely not an identity. Burl (1976, p. 205) thought that the name of *Maponos in this connection confirmed “the long use of some stone circles and the continuity of tradition whereby customs were perpetuated even by later comers to the district.” The continuity may have been even fuller than Burl realized.

At Penrith, 10 km SSE of Long Meg, is a henge known as “King Arthur’s Round Table.” The central area is 50 m in diameter, and it is surrounded by a circular ditch that in turn is encircled by a bank. The full diameter of the structure is ~90 m, comparable to that of the flattened ring of Long Meg’s Daughters, and probably had a similar function, to gather large numbers of people, possibly for axe trading (Burl 1976, p. 25). Large numbers of axes have been found at Windmill Hill and at Avebury. The stone circle construction, he notes, took a comparable amount of time to construct (he estimates 70 stones × 60 people × 10 hours/day = 42,000 man-hours) compared with estimates by Atkinson (1961) and Coles (1973, p. 73) of 55,000 man-hours for the Penrith henge (cited in Burl 1976/1989, p. 64). Burl also discusses the distributions of these forms. The stone circles and henges have a somewhat different distribution, with larger (over 200 ft, or 61 m diameter) henges distributed more evenly but with larger numbers in the central regions, whereas stone circles are much more strongly concentrated in the central and western regions. It is instructive that the stone circles in the central region tend to be accompanied by henges, but far less so in the western region (Burl 1976, p. 28). Was there competition for local populations in the central regions? If these great enclosures were used primarily as trade centers, the great number of them become somewhat more explicable, and so does the willingness on the part of the stone circle builders to adopt the henge in the most heavily henged regions. This does not explain if or why astronomy or burials would be involved, but it is clear that some of the most elaborate ones ( Avebury and Stonehenge, for instance) apparently served more than one purpose (social functions, trade, funerary, calendric, and maybe other kinds of astronomy functions). Does similarity in style of a center accompany similarity of purposes?

Burl (1988, pp. 187–190, 195) maintains that there is little evidence of numeracy at this period, although he points out (following Thom) the remarkable similarity in design and size between Castle Rigg and Brats Hill on Burn Moor. There is also a similarity between Swinside and Ballynoe in (Northern) Ireland, that is so great that Burl thinks Ballynoe was actually built by Cumbrians. At Swinside (60 stones in a circle 90 ft in diameter), there is a portal entrance to the SE. The two northern stones of the portal coincide with alignment with sunrise on Samain (a cross-quarter day; modern Halloween) (Burl 1993, p. 38). At Ballynoe, a circle of 72 stones with a diameter of 108 ft, has a large portal with the northern stones of the portal aligned on the equinox sunset (Burl 1993, p. 38).

Burl (1988, p. 189) thinks that the diameter of the circles may be a guide to the size of the populations that created them and gathered in them. He guesses supporting populations of ~400 people at Castle Rigg and ~4000 at Long Meg and Her Daughters; of these numbers, 1/10 would have been present at any one time. The circles normally had an entrance or portal marked by two larger stones, with a bigger than usual distance between them, and another pair outside the circle aligned on the entrance pair. The entrances were roughly on the cardinal points. Burl (1988, p. 200) writes

there is hardly a precise alignment in the Cumbrian rings and if any one site were taken on its own, one would be justified in thinking that a supposed calendrical or cardinal orientation was accidental, even imaginary. Single sites can be misleading. Instead, it is the repetition, in ring after ring of comparable alignments that reinforces the belief that the lines were intended and needed by prehistoric people. General patterns, rather than individual site-lines, buttress the argument in favour of archaeoastronomy.

Burl’s final summary of his views on the Cumbrian circles is much more broadly applicable (Burl 1988, p. 202):

This may have been what a stone circle was to its people, a place where axes and gifts were exchanged, a place where annual gatherings were held, a place to which the bodies of the dead were brought before burial, but, above all, a place that was the symbol of the cosmos, the living world made everlasting in stone, its circle the circle of the skyline, its North point the token of the unchangingness of life, a microcosm of the world in stone, the most sacred of places to its men and women.

South of the Cumbrian circles in the British Midlands are a number of old, flattened circles, one of the best preserved of which is that at Arbor Low (Figure 6.23), although the stones are nearly all fallen over. At the center of its henge and ring is one of only four “coves”—three-sided structures composed of three large stones—to be located at such sites in the British Isles.14 It faces SSW but is now fallen. Arbor Low shares many similarities with Cairnpapple (in scale and shape, of both henge and circle), where the funerary tradition spans millennia (Burl 1976, pp. 279–282).
Figure 6.23.

Panoramic view of Arbor Low at Derbyshire, in the English midlands, a large henge and ring structure comparable in scale to Long Meg and Her Daughters. Note the combination of mound, ditch, and stone circle, characteristic of midland sites. Photos and Montage, courtesy Dr. T.A. Clark.

6.2.8 Callanish: The White Cow

On the Isle of Lewis in the Outer Hebrides, in the Northwest of Scotland, there are a number of stone circles, all located near Loch Roag. The best known is at Callanish (Figure 6.24).
Figure 6.24.

View from the NNE avenue of the cruciform monument of Callanish on the Isle of Lewis in the Outer Hebrides. Photo by Sharan Hanna.

It consists of a ring of stones from which an avenue, defined by flanking stones, extends 82 m to the NNE. The stones are angular with their wider sides facing the avenue, except for the outer two, which face the visitor as if they were guard stones. The features of the site are shown in Figure 6.25.
Figure 6.25.

Features of the Callanish site. Drawing by Sharon Hanna.

Shorter lines of stones emerge to the east and (in the same line) to the west, as well as to the south. The avenue stones are along bearings 190.6° and 189.2° on the west and east sides, respectively (Thom 1971/1973/1978, p. 68). Burl (1993, pp. 61, 57) points out that the stones on the east side of the avenue are consistently about 3/4 as high as are the stones on the west side. This feature is characteristic of northern Irish avenues and double rows and of those on the Crozon peninsula in far western Brittany. Sometimes, the kinds of stones used were different as well. If there is any historical connection among these locations, it is probably due to their mutual accessibility by water. The cruciform shape of the Callanish complex is striking, and its location on a hill on a promontory makes it prominently visible at sea. Within the circle and in contact with its east wall is a smaller circle of predominantly smaller stones about a mound in which there is a passage grave with opening to the east. The tomb postdates the circle, and its kerbstones are still later additions. The astronomical implications of the site have been examined by Somerville (1912b, p. 83), Hawkins (1966, p. 186), Thom (1967, p. 122; 1971/1973/1978, pp. 68–69), Hadingham (1976, especially pp. 101–106), Gerald and Margaret Ponting (1982, 1984a,b), M. Ponting (1988), and Burl (1976, pp. 148–155; 1993, pp. 14–16, 59–61, 63–65, 178–180). Thom’s analysis of the Ordnance Survey data suggest that the extreme Southern setting position of the major standstill Moon could have been observed at this site. The 173d periodicity in the variation of the lunar inclination could have been noted through the interaction between the setting Moon and the terrain near Mt. Clisham in the Harris region of the island because at the range of declinations (−29°11.3′ to −29°26.8′), the Moon would alternate between skirting the top and the bottom of the undulating horizon. Ponting and Ponting strongly challenge this observation and state that an intervening hillock would have prevented this observation from this site, although not from sites in the vicinity. Thom also noted that the limits of the short-term perturbation in i could have been determined more easily at another nearby site, at moonrise, against a foresight with more variation. Stellar alignments claimed for Callanish are more problematical. The avenue was said to have been aligned on Capella, for example, by Lockyer (1906/1909, p. 377), Somerville (1912), and Thom (1967, p. 98), but for dates between 1800 and 1790 b.c., in the Middle-Bronze Age, most likely well after the monument was constructed and when the tradition of standing stones was into decline. The history is discussed by Burl (1982, p. 144).

Callanish illustrates the difficulties of reconciling archaeological and astronomical evidence. Hawkins (1965b) determined 12 alignments based on Somerville’s map. Unfortunately, five of these depend on stone 35, which was reerected ~1860, almost certainly not at the original location (Ponting and Ponting 1984, p. 47). On the other hand, Thom thought that a number of other stones had been reerected, based on early descriptions of some of them as “fallen.” In fact, they were all in position, buried in 5 ft or more of peat in early photographs, with only the upper part showing, sometimes mistakenly giving the impression of much smaller fallen stones. At that time, the cairn was completely covered. The peat began forming about 1000 b.c., with changing weather conditions, and thus establishes a minimum age for the site (Ponting and Ponting 1984a, p. 7). Archaeological materials that were recovered in excavations in 1980 and 1981 indicate that the circle was built ~2200 b.c. The Pontings think that the tall standing stone now at the circle’s center was put up still earlier. Stone 33A was visible in 1857 but then covered; it was rediscovered by the Pontings in 1977, by probing, then excavated in 1980, and reerected in its original hole in 1982. A Glasgow University survey in 1974 determined that the azimuth of the east row was 76°5′ rather than 77°8′ as recorded by Somerville and used in all studies prior to the survey. According to Burl (1993, p. 180), the east row would have been aligned on the rising of the Pleiades ~1550 b.c.

There is a suggestive similarity between what has been determined about Callanish and a famous description of a Hyperborean temple by Diodorus Siculus (1st century b.c.) derived from Hecataeus of Abdera (~500 b.c.). The description has frequently been applied to Stonehenge, but Burl (1993, pp. 64–65, 179–180) shows that interpretation to be highly unlikely. Diodorus wrote

Of those who have written about the ancient myths, Hecataeus and certain others say that in the regions beyond the land of the Celts there lies in the ocean an island no smaller than Sicily. This island, the account continues, is situated in the north and is inhabited by the Hyperboreans, who are called by that name because their home is beyond the point whence the north wind (Boreas) blows; and the island is both fertile and productive of every crop, and since it has an unusually temperate climate it produces two harvests each year. Moreover, the following legend is told concerning it: Leto [mother of Apollo] was born on this island, and for that reason Apollo is honored among them above all other gods; and the inhabitants are looked upon as priests of Apollo, after a manner, since daily they praise this god continuously in song and honor him exceedingly. And there is also on the island both a magnificant sacred precinct of Apollo and a notable temple which is adorned with many votive offerings and is spherical15 in shape. Furthermore, a city16 is there which is sacred to this god, and the majority of its inhabitants are players on the cithara; and these continually play on this instrument in the temple and sing hymns of praise to the god, glorifying his deeds.

The Hyperboreans also have a language, we are informed, which is peculiar to them, and are most friendly disposed towards the Greeks, and especially towards the Athenians and the Delians, who have inherited this good-will from most ancient times. The myth also relates that certain Greeks visited the Hyperboreans and left behind them there costly votive offerings bearing inscriptions in Greek letters. And in the same way Abaris, a Hyperborean, came to Greece in ancient times and renewed the good-will and kinship of his people to the Delians. They say also that the moon, as viewed from this island, appears to be but a little distance from the earth and to have upon it prominences, like those of the earth, which are visible to the eye. The account is also given that the god visits the island every nineteen years, the period in which the return of the stars to the same place in the heavens is accomplished; and for this reason the nineteen-year period is called by the Greeks the “year of Meton.” At the time of this appearance of the god he both plays on the cithara and dances continuously the night through from the vernal equinox until the rising of the Pleiades, expressing in this manner his delight in his successes. And the kings of this city and the supervisors of the sacred precinct are called Boreadae, since they are descendants of Boreas, and the succession to these positions is always kept in their family (Loeb Library, Book II, §47). [Bracketed material added by present authors.]

Burl, with the Pontings, supposes that the “Moon being nearer the Earth” refers to the appearance of the major standstill Moon above the Callanish horizon. At its southern maximum (δ ≈ −29°), the Moon never rises more than ~3° above the horizon and the “big moon” effect (see §3.1.3) would enhance the illusion that the Moon appears closer to the Earth at that time. This is a result of the high latitude of the Callanish site (φ ≈ 58°10′) and the circumstance that the altitude of an object transiting the celestial meridian is equal to the sum of its declination and the site’s colatitude: δ + (90 − φ). The effect is far less striking at Stonehenge (φ = 51°11′), where the altitude of the Moon would be greater. The seemingly sharper details are more difficult to explain, because the greater air mass of the lower altitude Moon at Callanish would tend to obscure details rather than enhance them. Perhaps, in this case, the local observer’s visual acuity played an important role. It is interesting that the Greeks adopted the cult of “Hyperborean Apollo” in about 470 b.c. and that Abaris, possibly a mythical figure, is alleged to have taught Pythagoras (Burkert 1972, pp. 149–150).

Another different line of evidence for Callanish as the island of the Hyperboreans involves the reference to the birth of Leto, mother of Apollo, on the island. If we transfer this into Celtic terms, Mac nOg, the equivalent of Apollo, was a son of *Bu-vinda, “White Cow,” the goddess who gave her name to the Boyne River. Ponting and Ponting (1984, p. 30) relate a legend about the arrival of a Gaelic-speaking white cow, which emerged from the sea during a famine. The cow told the people to come to the Callanish stones and she would give them each a bucket of milk. However, a witch brought a bottomless bucket and milked her dry. DHK thinks this is a version of the cornucopia myth, the horn taken from Amaltheia by Zeus and given to the nymphs Io and (her sister) Adrasteia. Io was a name of the moon goddess in Argos and the name given to a woman beloved of Zeus, who turned her into a white cow to evade Hera’s jealousy. The various motifs linked by the story from Callanish look more like scraps of ancient mythology than most such stories.

Another legend given by the Pontings (1984a, p. 27) says that the Stones were brought to Callanish in ships and erected by black men under the direction of a priest-king, who was always accompanied by wrens. He and other priests wore feather cloaks (which was true of some Gaels in the pre-Christian period). The reference to wrens suggests the golden-crested wren, king of the birds. As sacred birds, wrens could be killed only once a year by the “wren-boys” on St. Stephen’s day (Dec. 26) just after winter solstice (Frazer 1912, Part V, Vol. 2, pp. 317–320).17 Sometimes the killing was done by piercing the bird with sticks, fastened to make a kind of armillary sphere18 (see Figure 6.26). It has always been assumed that Diodorus’s “spherical temple” was merely some curious error for “circular temple,” but the association of Callanish with wrens and the winter solstice raises the possibility of a different meaning, involving an actual spherical structure such as an armilla; certainly one would expect educated Greeks of the 1st century b.c. to be able to distinguish a sphere from a circle.
Figure 6.26.

The (symbolic) kil­l­ing of the sacred wren on Saint Stephen’s Day, just after the winter solstice: It was sometimes depicted as the piercing of the bird with sticks, fastened to make a kind of armillary sphere. Drawing by Sharon Hanna.

The wren king is known in Devonshire as “the cuddy vran,” “Bran’s sparrow,” and Bran is another important bird, “crow, raven” as well as a figure of Welsh traditional legend, King Bran. In the Romance of Branwen (named for Branwen, “white crow,” sister of Bran), it is said that Bran’s decapitated head was buried at Tower Hill in London to guard the city from invasion. Decapitation is associated with eclipses in other culture areas. The decapitation and other associated motifs suggest a solar identity for Bran. Apollo’s bird was the crow, and the traditional phrase “as the crow flies” implies a recognized analogy between the crow’s flight and the movement of the Sun along the ecliptic.

It is reasonable to suppose that the Greeks had already defined the “four winds” as direction markers at an early date; two sons of Boreas are included among the crew of the Argonauts. We may say, therefore, that the line south of the circle, which has a true north-south alignment,19 was, in Greek terms, a line to Boreas. The movement of the Moon in the sky and in the 19-year Metonic cycle, the winter solstice alignment, the “north pointer,” the equinox alignment, the possible Pleiades alignment, the size of the island group (the Hebrides), and even the reference to the sphere all seem to fit Callanish far better than Stonehenge.

Burl has repeatedly emphasized that groups of sites that show similar alignments provide much better evidence of intent than any single site can, however great the apparent precision at a single site. On the Isle of Lewis in the near vicinity of Callanish, three stone circles are also oriented to the southernmost setting of the major standstill Moon, and other monuments suggest the same orientation. Ruggles (1984, 1985) demonstrated that north-south lines were common in SW Scotland. There is also a rarer alignment at Callanish that is seldom mentioned.

Looking from the main site (Callanish I) to the east, Callanish XIV, a single standing stone, becomes a remarkably good marker for the equinoctial sunrise (Ponting and Ponting 1985b, p. 37).

6.2.9 Brogar, Stenness, and Maes Howe

These three monuments date from about the same period, 3000–2800 b.c. according to Mohen 1990, p. 132; Burl (1976, p. 101) suggests that Stenness and Maes Howe are slightly older than Brogar. The Brogar (sometimes, Brodgar) and Stenness monuments are rings of standing stones; Maes Howe is a tumulus or burial mound. All are located on the main island (“Mainland”) of the Orkneys in the extreme north of Scotland. The latitude of these sites is ~59.0°. Stenness and Brogar are about a mile apart on an isthmus separating the Loch of Stenness from the Loch of Harray.

The Brogar stone circle monument (Figure 6.3) has a diameter of 103.6 m (125 MY), the same diameter as at Avebury (A. and A.S. Thom 1973, p. 122) and of the standing stone circle surrounding the great tumulus at Newgrange in Ireland. These dimensions make it one of the largest of the stone rings. The stone circle at Brogar is surrounded by a circular ditch with a diameter of 142 m. Thom and Thom (1973) argued that the site was used for lunar observations, although the dates assigned to this relatively high-precision activity, ~1560 b.c., suggest a very late refinement of use of an ancient site. The use of natural foresights on the horizon provides much greater accuracy as well as measurement precision, as we have noted (§§3.2.1, 6.2.4).

The “Standing Stones of Stenness” (Figure 6.27) are on a circle 31.1 m in diameter, surrounded by a circular bank 61 m in diameter. It has an entrance avenue to the NNW, thus, to the setting direction of the midsummer Sun. It originally had 12 stones.
Figure 6.27.

A few of the Standing Stones of Stenness on the Mainland island of Orkney: The central rectangle is just visible. Photo by Sharon Hanna.

Burl identifies the Stenness henged ring as older than the other two sites on the basis both of stylistic considerations—he classifies Stenness as a class I henge and Brogar as class II—and 14C dates (timber and uncremated bone gave uncorrected 14C dates of 1730 ± 270 bc and 2238 ± 70 bc, respectively, and 2356 ± 65 bc was found for an animal bone found in the ditch). Many broken axe heads were found at Stenness.

Seventeenth-century tales (recounted by Burl 1976, p. 15) tell of Sun worship being carried out at Brogar and moon worship at Stenness. Less than a mile east of Stenness is Maes Howe.

Maes Howe is on a raised platform surrounded by an incomplete ditch. It contains a passage grave with a passage 24 m long and lined with flat slab walls and corbelled roof. It terminates in a square chamber adjoined by three square cellular chambers. The tomb is carefully constructed with fine stone workmanship, and it has been cited among Europe’s finest megalithic constructs. A door into the chamber was deliberately cut lower so that it did not completely block the entrance, allowing light from the mid-winter sunset to enter the tomb (Burl 1993, p. 63). This has an analog in the midwinter sunrise phenomenon at Newgrange, which is roughly contemporary. The Orkneys, formerly the Orcades, were named for the “Pig People.” See § for a discussion of similar megalithic alignments on the “Big Pig” and “Little Pig” islands.

In addition to these sites is another, lesser known monument, the Ring of Bookan on a platform surrounded by a ditch about one mile NW of Brogar, which contains a passage grave. Also near Brogar is the Stones of Via passage grave, and a long mound and ~20 cairns. There are also four other passage graves within ~10 km of Stenness. A village of stone houses at Skara Brae on the west coast of Mainland, Orkney, was recovered in the mid-19th century. From carbon 14 dating, its true dates of occupation were between 2400 to 1800 b.c. (Mohen 1990, p. 315).

6.2.10 Stonehenge

This is the best known of all the megalithic constructs. Figure 6.28 gives the modern appearance of the monument, and the layout of the site is given in Figure 6.29.
Figure 6.28.

Stonehenge on Salisbury Plain in southern England as seen from (a) the East, (b) the ENE, and (c) the NNE permits the intact portions of the trilithons to be identified. Photos by Marie C. Jack.

Figure 6.29.

The general layout of Stonehenge. Several prominent features are noted, among them Aubrey holes 55 and 56, the Avenue, and the four stations. Drawing by E.F. Milone.

Stonehenge has been speculated about since the 12th century,20 and carries back in legend further than this. The comments of the Sicilian writer Diodorus, discussed in the context of Callanish, are often thought to refer to Stonehenge, but they are not Stonehenge-specific. Moreover, the historicity of Diodorus’s source, the 6th-century b.c. writer Hecataeus, is not regarded with any great confidence. Ptolemy refers to both the 19th parallel (511/2°N, longest day of 161/2 equinoctial hours) through southernmost “Brittania” and the 28th (62°N, 191/2 equinoctial hours) through “Eboudae,” by which he means the Hebrides (Toomer 1984, pp. 87–89, fn. 65). Roman references to the island are common; in particular, many historical references can be found to the Druids and their worship of trees, stones, and sky deities. However, the Druids were not the builders of this or of the other megalithic sites, although they may have used and maintained them. The origins are in the Neolithic. The history of speculation about Stonehenge is entertainingly conveyed by Hawkins (1965). The chronology (based on Atkinson 1956/1979 but revised with other data) is given in Table 6.3.
Table 6.3.

The chronology of Stonehenge.

Dates b.c.

Raw C14 datea (corrected date)

Source of C14 date


Stonehenge Ia



2180 ± 105 (2810)

Antlers at ditch bottom

Bank, Ditch; possible hut in center; Causeway postholes


[2450 ± 60 (3080)]

Mean of two antlers


Stonehenge Ib



1848 ± 275 (2305)

Charcoal in Aubrey Hole

Aubrey Holes; Heel Stone in place.

Stonehenge II



1728 ± 68 (2130)

Antler from avenue ditch

Four Station stones (91–94).


1765 ± 70 (2180)

Skeleton from ditch

Erection of bluestones in two (later abandoned) circles (“Q and ‘R’ Holes); first part of avenue.


1720 ± 150 (2120)

Antler near sarsen trilithon


1620 ± 110 (2000)

Antler in “R” Hole


Stonehenge IIIab



1720 ± 150 (2120)

Antler in erection ramp of Sarsen Stone 56

Transport of Sarsen Stones to site; Dismantling of Q and R bluestone circles; Sarsen circle, trilithons, Slaughter & other Portal Stones in place.


1240 ± 105 (1550)

Antler from bottom of Y hole

Dressed bluestones tooled and erected; Digging and then abandonment of the Y and Z Holes.

Stonehenge IIIc




Dismantling of dressed bluestones; Bluestones reerected in circle and horseshoe, within corresponding Sarsen constructs.

Stonehenge IV



800 ± 100 (975 ± 115)

Bone and antler from ditches

Extension of Avenue to West Amesbury.


1070 ± 180 (1345 ± 190)

Antler from ditch at W. Amesbury


a See §4.4 for a discussion of this dating technique and the corrections that need to be applied. The phase divisions are due to Atkinson (1960/1979). Data are from several sources, especially Burl (1987) for most carbon 14 dates. “Raw” dates are often denoted “bc”; corrected dates, “b.c.”

Archeologists agree that Stonehenge appears to have been built in three stages. The 14C age for Stonehenge I is ~3000–2500 b.c. (see Hoyle 1977a, pp. 32–34 for the details of the process and further comments), whereas the last phase, Stonehenge III, was constructed about 1000 years later. We begin our discussion with Stonehenge I.

The earliest stage involved the construction of a 409-ft. (125-m) diameter ditch that defines the outer part of the monument, a large earth bank, 56 chalk-filled holes (the Aubrey Holes21) arranged in a circle within the banked enclosure, mounds located within the bank, an entrance “causeway” created by a 35-ft (10.7-m) break in the bank and ditch to the northeast, a large standing stone (No. 96) known as the “Heel Stone” (also written “Heelstone” or “Hele stone”), post holes near the causeway and near the Heel Stone, and stone holes in the entrance (Figure 6.30).
Figure 6.30.

One of the two main outliers at Stonehenge: the Heelstone. (a) Distance view, photo by Marie C. Jack. (b) In closeup, revealing the cup markings. Photo courtesy of Sharon Hanna.

The Heel Stone is of a kind of sandstone, called a “sarsen” stone (derivation uncertain). It rises 16 ft (8.5 m) above ground (another 4 ft is estimated to lie below the surface) and has a thickness never less than 8 ft (2.4-m) across its length. The Causeway and postholes preceded the erection of the Heel Stone, which otherwise would have obscured the posts on the right-hand side of the Causeway. Wood (1978, pp. 162–164) argues persuasively that the Causeway, originally oriented a few degrees north of midwinter sunrise, and the posts in the Causeway were intended for observation of the northern risings of the Moon. Brinckerhoff (1976) noted that between the Causeway posts the Moon could be observed to rise at 1/4, 1/3, and 1/2 of the length of time between the midpoint and major standstill of the lunar cycle. The Heel Stone would have marked a lunar midpoint rise more permanently than any wooden post. Another group of features seems to involve alignments from a slightly later stage of construction (late in Stonehenge I or in Stonehenge II), and so we discuss these before the Aubrey Holes themselves. Within the Aubrey Hole ring perimeter are four “Stations,” 91, 92, 93, and 94 (indicated on Figure 6.29), which form a rectangle of important solar alignments. Stations 91 and 93 are stones (Station stone 91 is a fallen menhir, 3.66 m long). Stations 92 and 94 are on opposing mounds (covering Aubrey holes and thus indicating that they are newer), with 92 being the site of a filled hole that may have held a stone or post. At the latitude of Stonehenge, 51°11′, this rectangle describes the approximate alignment of extreme aspects of the horizon solar calendar, i.e., the amplitude of the Sun’s motion and those of the Moon for both major and minor standstills (see §2.3.1 for the tie-in between the Sun’s annual variation in declination and the effect of this variation on the azimuth amplitude of sunrise and sunset; see §2.3.5 for the discussion of the Moon motions and of the phenomena of major and minor standstills in the 18.y6 cycle). Figure 6.31 demonstrates the amplitudes of the Sun and Moon at the latitude of Stonehenge in this plan view of the astronomical horizon (for an “elevation” view, see Figure 2.17a) for a late epoch.
Figure 6.31.

A plan view of the horizon shows the northern and southern extreme rise points for the Sun and Moon for a flat horizon at latitude 51°N and, thus, the amplitudes of the major and minor standstills of the Moon and the solstices for those conditions. The set points are symmetrical. Drawing by E.F. Milone.

The declinations indicated by alignments among pairs of the four stations are given in Table 6.4, the data of which are taken from Hawkins (1965). Dibble (1976) noted that the rectangle made by the four Stations is composed of two triangles very nearly matching 5 × 12 × 13 Pythagorean triangles. Among themselves, the stations could not have provided high-precision alignments. In 1966, however, three postholes were discovered in a car-park to the northwest. Their locations are marked in concrete on the road surface. Newham (1972) suggests that in combination with stations 91, 92, and 94, and the Heelstone, both solar (setting, summer solstice) and lunar (setting, northern major standstill; setting, northern minor standstill; and setting, midway between the standstills) alignments were achievable with posts set into these holes. They may have been erected in order to help in the layout of the Stations (Newall 1953/1959/1981, p. 23).
Table 6.4.

Alignments at Stonehenge.





Calendar mark





Summer Solstice





Winter solstice





Major standstill





Minor standstill





Major standstill





Minor standstill

If, and only if, the full moon is considered, the directions 91–94 and 91–93 refer to winter solstice moonsets (that is, a moonset when the Sun is at the winter solstice) at the times of major and minor lunar standstills, respectively. Similarly, if only the full moon is considered, 93–92 and 93–91 refer to summer solstice moonrises at the time of major and minor lunar standstills, respectively. When Hawkins (1965) refers to “Midwinter Moonrise,” for example, in connection with these alignments, he is referring to this phenomenon. It must be remembered that the Moon goes through its entire amplitude in a mere month, but that the amplitude varies in size from lunation to lunation, continuously changing over the period of 18.6 years of the nodal regression cycle. Thus, from night to night, the Moon will rise, and set, at a sequence of intermediate azimuths within its rising and setting amplitude.

It is this circumstance that suggests the use of the Aubrey Holes as an eclipse predictor. The ring of Aubrey Holes is ~250 ft (86.87 m) in diameter (Wood 1978, p. 163, gives the circumference as 271.6 m and the radius as 43.2 m). The holes are about 3.5 ft (1.07 m) in diameter and ~2.5 ft (0.76 m) deep with flat bottoms and were dug out of the chalk that underlies much of the region. They are filled with rubble, crematorial remains, charcoal, and chalky debris. The organic material was used in the 14C dating of Stonehenge I and may well be later than the actual construction date. The works by Hawkins (1965) and later by Hoyle (1977) emphasize the importance of the number of Aubrey holes: 56. The mean of three intervals of 19, 19, and 18 years is 18.y 67, an approximation to the nodal regression cycle, and the sum of these numbers is 56. There are three other potential tie-ins of this number to astronomical phenomena.

The number of the Aubrey holes, which, it must be remembered, were a prominent component of the earliest stage of Stonehenge, in a way already nearly encapsulate the motion of the Moon. Other possible tie-ins that have been suggested are the nearly 55 (542/3) nights in which the Moon has completed its motion among the stars twice (Psid ≈ 27d.3), and the period of the triple Saros cycle, 54y 33d, over which nearly identical conditions for an eclipse will recur. Thus, an earlier eclipse will be followed by another of similar duration, and seen at a similar range of longitudes on Earth. Finally, this number is serendipitously close to a triple Metonic cycle (19 × 3 = 57), which reconciles the lunar and solar calendars.

Hawkins’s (1965) scheme for the use of Stonehenge as a “computer” (Schlosser et al. 1991/1994 suggest an “abacus” to be a better analogy), at its simplest, involved the use of six stones of alternate kinds (say, black and white stones), placed in Aubrey holes at intervals of 10, 9, 9, 10, 9, and 9 holes. Figure 6.32 demonstrates the arrangement. Each year, each of the six stones would be advanced one hole; the direction of advancement adopted here is counterclockwise (as in Hawkins 1965, p. 142), but this is not critical, as he pointed out, as long as the hole intervals between the stones are maintained.
Figure 6.32.

Use of the Aubrey holes Stonehenge as a stone age “abacus.” It may have involved the use of six stones of alternate kinds (say, black and white, as indicated here by black and gray circles, respectively). Each year, each of the stones would be advanced one hole; the direction of advancement shown here is counterclockwise, but this is not critical if the stones are separated by 10, 9, 9, 10, 9, and 9 holes in the direction of rotation. Drawing by E.F. Milone.

At least three holes have special meaning, because an eclipse danger occurs when a particular type of stone falls in one of these three holes: 51, 56, and 5. The eclipse phenomena would be accompanied by the rise of the full moon over one of three standing stones in the Avenue: the northernmost rise of the major standstill full moon (at δ ≈ +29°) over Stone D, the northernmost rise of the full moon at midcycle (δ ≈ +24°) over the Heel Stone, and the northernmost rise of the full moon at minor standstill (δ ≈ +19°) over Stone F. The example provided by Hawkins (1965, pp. 141ff) starts with a white stone in hole 56, in the Avenue, during a year when the Moon rose over the Heel Stone at its northernmost rising (with δ ≈ +24°) and when solsticial eclipses of Sun and Moon could be seen. A white stone in holes 56, 51, or 5 becomes a marker of an eclipse danger. A black stone in hole 56 will produce the same result. A white stone in holes 5 or 51, however, will mark a danger of equinoctial eclipses. The success of the proposed scheme can be assessed in Table 6.5, which lists the eclipse phenomena for selected dates in the mid-16th century b.c. from data provided by Hawkins (1965, pp. 178–180). Although the dates refer to a time 1000 years after the construction of Stonehenge I, the table may illustrate the continuing predictive power of the site. Alternative schemes are provided by Hoyle (1977) and by Schlosser et al. (1991/1993, p. 15), with various degrees of success. Schemes occasionally break down, and eclipse cycles come to an end (see Robinson 1983). Such failures may account for the temporary abandonment of the site at various times in its history. The terminations of such series (here, 2039 or 2104 b.c.; cf., Robinson 1983, p. 128) may provide an important way of dating archeological monuments.
Table 6.5.

An example of sequential eclipse datesa at Stonehenge.

Eclipse type

Moonrise Heel Stone

Moonrise Stone D

Moonrise Stone F

Lunar eclipses

Solar eclipses




1554 Jan. 4





1549 Mar. 23





1545 Jan. 10






1541 Apr. 9


1541 Oct. 2




1536 Jan. 14





1531 Apr. 3


1531 Sept. 28





1527 July 16

1526 June 21





1522 Apr. 9


1522 Oct. 3




1518 Dec. 31


a Based on Hawkins (1965, Tables 2, 3, and 4). All dates are b.c., Julian Calender; these dates post date the actual time of building probably by more than a millennium. The sequence, however, illustrates the possible predictive value of the site.

Of course, there is no direct evidence that the Aubrey holes were actually used for eclipse prediction, nor is their use strictly necessary, because the rise azimuths of the Moon provide the principal clues. The argument is one of plausibility only. The alignments are evidence of interest in lunar and solar phenomena. By association, the placement of particular stones in particular holes with the moonrise over various foresights in what later became the Avenue could have served as an early-warning device for that terrifying phenomenon of the ancient world, the eclipse. Such a warning device could serve well in periods of bad weather when close observations of the Moon’s behavior were not possible.

The recognition by the builders of the astronomical significance of the number 56 is critical for the hypothesis that the Aubrey Holes were used to determine eclipses. Against the idea of the Aubrey Holes as an abacus or computer is the circumstance that while 10 other sites in Britain have chalk circles associated with henges, none has 56 of them (Burl 1987, pp. 86–88). Their numbers range from 5 for Llandegal in Gwynedd (Wales) to 44 or 45 for Maumbury Rings in Dorset, but all the others number between 7 and 14. Moreover, Burl (1987, pp. 89–90) argues that the number 56 may be an artifact of the layout of the Aubrey Holes (investigated by Thom and Thom 1974): At each of the cardinal points, two stones nearly (within 1/2 degree) flank the cardinal directions. If 12 other stones are placed between these pairs, the number 56 falls out immediately. The 12 is somewhat arbitrary but does give approximately even spacing. Burl (1987, p. 90) also points out that both the radius of the Aubrey Hole ring (which he gives as 141.4 ft = 43.7 m) and the separation between stones (16.3 ft = 5.0 m) are not even multiples of the “Megalithic Rod” (6.8 ft = 2.1 m), at 20.8 and 2.4 MR, respectively); Thom and Thom (1973) had found, however, that the length of the circumference amounted to an even 131 MR. Whatever the purpose, Newham (1972) noticed that the length of the chord joining every other Aubrey Hole was about 1/3 that of the radius of the Aubrey circle. Once the circle had been established by means of a rope anchored at the center, and marked, the rope could have been folded in thirds and this length laid out as chords across the perimeter. In three turns around the circle, 56 holes could have been marked.

In the next major phase of construction, “Stonehenge II,” a double ring of 82 “bluestones” (principally dolerite and rhyolite rocks and modified volcanic ash, but other types are also present) was erected in features called Q and R Holes. These stones all seem to come from a localized region in the Prescelly Mountains in Pembrokeshire in Wales (Atkinson 1979, pp. 49, 51). Because bluestone chips are not found below the (current) middle layer of the ditch, and there only sparsely at a level consistent with mere transport to the site (Atkinson 1979, p. 72), it is clear that they were erected later than were the features ascribed to Stonehenge I. The Q and R Hole circles have diameters of 86 ft (26 m), and 75 ft (23 m), respectively. Finally, during Stonehenge II , the ditch was further filled in on the eastern side and the “causeway” was extended into an “Avenue” beyond the monument, toward the River Avon, with a new net orientation: to the mid-summer sunrise.

Stonehenge III is usually considered to have developed in three substages. The first, Stonehenge IIIA, saw massive reconstruction activity, which involved the transporting of the huge sarsen stones from Marlborough Downs, near Avebury, some 20 miles away. This period involved the dismantling of the double bluestone circle, and the subsequent erection of five massive sarsen trilithons (three-stone combinations with mortice and tenon anchoring the top stone slab) in a horseshoe pattern open to the northeast. The horseshoe is encircled by sarsen stones, also possessing lintels, some of which are still in place. The upright components of the horseshoe trilithons weigh as much as 30 tons and the transport of these 80 sarsens represents a considerable engineering feat, involving as many as a thousand laborers over several years. See Atkinson (1956/1979) or Hawkins (1965) for informed speculation about how this could have been done.

Carvings of thirty axe-heads and a hilted dagger have been found on the sarsen stones. The dagger shape most resembles that of daggers found in Shaft Graves at Mycenae in Greece, with dates prior to ~1500 b.c. (Atkinson 1979, p. 92). Atkinson (1979, p. 93) states that grave goods of Wessex Culture burials “provide clear evidence” for trade with Greece in this period. Moreover, axe cults are known to have been widespread (as in Minoan Crete) in this period, but axes are also carved in tombs in Brittany and elsewhere in England. The Stonehenge axe carvings closely resemble bronze axes manufactured in Ireland and brought to England between 1650 and 1500 b.c. We recall our earlier discussion of axeheads, and note that associations with lightning or with the Sun are possible here also.

In Stonehenge IIIB, the bluestones that formally occupied the Q and R Holes were prepared and perhaps temporarily set aside, whereas a new set of holes (the Y and Z Holes) were created in concentric circles outside of the sarsen stone circle. There were 60 (or 59) of these holes, with the Z holes lying closer to the sarsen circle. However, it does not appear that the bluestones were ever set into these holes.

In Stonehenge IIIC, the remaining 22 bluestones were reset into a ring just inside the sarsen ring, and into a smaller horseshoe of trilithons within the sarsen trilithons.

Other features of the site include the fallen “Altar Stone,” a 16-ft long, relatively narrow (31/2 ft × 11/2 ft) block of green sandstone, speckled with mica, from a site near Milford Haven in SW Wales (not from the slightly more distant Prescelly Mountains) near the central sarsen trilithon. Its exact intended placement in the monument is not known. A four-stone “portal” of very large stones commanded the entrance to the site from Stonehenge IIIA until at least the mid-17th century (Burl 1993, p. 39). Of these, only the so-called “Slaughter Stone” is still in place. This portal constitutes a striking similarity to portals found in the great circles of the Lake District, although Burl (1993) is unwilling to suggest that the Cumbrian circles were prototypes of Stonehenge III. However, the Lake District circles were standing even more impressively then than now, and the influence could have been transmitted by an individual of the Stonehenge community who had seen one and liked the idea.

It seems impossible that precision alignments could have been made with the massive, irregular stones of the monument as either backsights or not-so-distant foresights, although rough approximations to the solar and lunar standstills could certainly have been made. Atkinson (1979, pp. 94–96) in discussing and refuting Lockyer’s (1909) theory that the midsummer sunrise alignment along the Avenue matched precisely in 1680 b.c. argues that the alignment if determined by observing through spaces of several feet between the sarsens could not possibly give a precise date for the obliquity, because a movement of 1 in would change the date by 200 years, and that by using, say, the right eye instead of the left would shift it by 500 years! It seems likely, however, that the postholes of the earliest phase of Stonehenge could have held smaller and therefore more useful foresights in combination with smaller backsights, or the posts of the entrance way could have been used as backsites by observers near the center of monument in connection with distant foresights. Brinckerhoff (1976) has suggested that pits on the lintels of the great stones in the Sarsen Circle facing the Avenue could have held small wands to be used as foresights, in a manner suggested for the posts of the Causeway during Stonehenge I. The hypothesis requires a backsight across the Sarsen Circle from a now missing lintel. Alignments to northern, major standstill moonrise and to the summer solstice sunrise both at ~2000 b.c. could have been established in this way. Systematic departures of ~0.°2 from alignments with the Causeway postholes were attributed to a change in the obliquity of the ecliptic (see §2.3.3) over the approximate thousand-year interval between Stonehenge I and IIIA. More distant foresights have been proposed also. Alexander Thom (1971/1973/1978) argued that at many sites, especially in Scotland, a relatively high order of precision was achievable, and apparently achieved, by the use of backsights far removed from the observing platform and foresights on a distant horizon. Thom designed a method by which the builders or users of the sites could have sufficiently high-precision measurements of the position of the Moon to detect the 9 minutes of arc variation in the Moon’s inclination. The Thoms suggested several places in the vicinity of Stonehenge from which distant foresights could have been set up (Peter’s Mound, summer solstice sunrise; Coneybury Barrow, southern minor standstill moonrise; Figbury Rings, southern major standstill moonrise; Chain Hill, southern major standstill moonset; Hanging Langford Camp, southern minor standstill moonset; Gibbet Knoll), but some of these places would have been obscured by an intervening ridge and would have required artificial foresights to have been placed on them. See Wood (1978, pp. 178–181) for a thorough discussion. We review the method in a later section (§6.2.15); at present, we discuss the evidence from some carefully studied sites.

6.2.11 Mull and Argyll: A Test of Precision

Ruggles (1984a, 1988a,b) and Norris (1988) discuss at length a series of sites selected to test Thom’s ideas about lunar and solar alignments. Ruggles used predefined criteria for site selection and examined 300 western Scottish sites that had not previously been examined by Thom. Ruggles (1988, p. 275) argues that his criteria are necessary to avoid bias in selection of the sites and measurement points, and he roundly criticizes the work of the Thoms as suffering from just these kinds of bias. All Ruggles’s data were collected before any analysis was undertaken to preclude preliminary analyses from biasing later data recording. This was designed to be a test of Thom’s ideas, in the sense that the data were comparably rigorous but constituted an entirely independent sample. In the first stage of the study, Ruggles (1988, p. 234) concluded that
  1. (1)
    There were indications of preferred declinations, but at three levels of precision:
    1. (a)

      At the lowest level, δ = ±15° were “strongly avoided”

    2. (b)

      At the midlevel, there was “marked preference” for δ > 27°, and for −31° < δ < −19°

    3. (c)

      At the highest precision level (1° to 2°), there was marginal precision for six declinations: −30°, −25°, −22.5°, +18°, +27°, and +33°

  2. (2)

    Sites in Mull and Argyll, especially those with 3-, 4-, and 5-stone rows, showed specific declination preferences

  3. (3)

    The declination trends became more marked for stone rows, pairs, and single flat slabs in the majority of Mull and Argyll sites, especially for the range, −31° < δ < −19°


A second derivative study concentrated on 92 linear arrangements in Argyle and Mull that included new horizon data. This showed a primary grouping of declinations within a degree or two of −30° and a secondary grouping centered on −23° with somewhat more variation. Of these, a statistically defined group of 15 emerged, somewhat modified by later study. The data indicated “secondary” alignments (in terms of the way the menhirs are arranged) toward −24° at sites that also contained “primary” alignments near −30°. The archeologically defined “primary” and “secondary” alignments were reversed at Duncracaig and at Ardmacross, where only declinations near −24° were found. “The primary orientations appear to present particularly strong evidence of deliberate orientation upon the southern major standstill of the moon” (Ruggles 1988, p. 245). The secondary orientations were interpreted as marking either another point in the lunar cycle (although with no declinations higher than −21°) or the midwinter Sun (statistically indistinguishable). Ruggles’s more general conclusions are that the high-precision results of the Thoms are spurious, but that the Megalith builders were indeed interested in marking the extreme solar and lunar positions. These conclusions, however, say nothing about the precision that may have been achieved at particular sites, where a combination of natural features on the horizon, as well as care, intelligence, and good luck on the part of the builders may well have led to at least some of the precise alignments deduced by Thom. The alternative, skeptical view of the apparent successes of the Thom hypothesis at these sites is that they are merely fortuitous. We discuss some of these sites in the next few subsections.

6.2.12 Ballochroy and Kintraw: Controversial Sites

The site of Ballochroy in Argyll (western Scotland) now consists of a row of three stones with the flat faces, most unusually, across the row. A small cist is the only surviving remnant of a former large cairn or burial mound. There were once two small cairns and another standing stone in the same alignment (Burl 1993, p. 176). The alignment to the southeast (A ≈ 226°) passes over the islet of Cara and would have marked the sunset at the winter solstice [δ = −23.06°; Ruggles (1984, p. 279) finds declination limits −25.5 to −24.5°]. Looking across the flat sides of the stones, one would have seen the Paps of Jura across the sea in the distance, with Corra Bheinn mountain marking the summer solstice sunset. To the northeast, along an elevated horizon, the line marks the extreme northerly position of the Moon (moonrise at major standstill). Burl (1993, p. 177) wrote

The sightline towards Jura and the midsummer sunset was uninterrupted, and knowing that the wide faces of the slabs were atypically angled across the row it is likely that the builders of the row quite deliberately set them transversely to establish the alignment, unaware of the lucky coincidence that theirs was the lone latitude in Britain where the midwinter and midsummer solstices occurred at right angles to each other. A few degrees to the north or south would have rendered a comparable design unworkable.

The view would seem to be only slightly exaggerated. At an epoch such as 1750 b.c., when δ = 23.90° (Thom 1971/1978, p. 42), one finds that the site where twice the amplitude on a flat horizon is 90° has a latitude of 55.0°. For the given latitude of the Ballochroy site, 55°42′44″ (Thom 1971/1978, p. 37), summer and winter solstice azimuths of sunset, are, respectively, ASS = 315.99, AWS = 224.01, so that their difference is ASSAWS = 91.97°, again for a flat horizon. However, the horizon is not flat. Thom (1971/1978, p. 42) assumed that the alignment of the Ballochroy cist to the rolling slope of Corra Bheinn was the intended alignment to midsummer sunset and that the alignment from Ballochroy to the Isle of Cara was intended to mark midwinter sunset. The differences between these two measured azimuths, 316°05′ − 226°16′, is 89°49′. Thom’s analysis of this (and related sites) considers the effect of the elevation of the horizon and refraction and with a corrected true altitude of setting. From Ballochroy, he derived a value for the obliquity: 23°54′.2 ± 0′.7, and from this a date: 1750 ± 100 b.c. Burl’s “coincidence” statement requires a range of dates to be meaningful, but if Thom’s analysis is correct, and if the site were in fact erected ~1750 b.c., it seems to us extremely unlikely that the builders who followed this atypical procedure would not know that they were at a place where it would work.

To the North, at Kintraw, there is a site that has been the subject of a major controversy. In a field, there is a cairn with a posthole in it and a very tilted standing stone. Most unusually, the cairn is not a burial mound. The stone does not seem to be aligned on anything that would have been visible at the site from ground level. However, Thom suggested that it was aligned to a notch created by the profiles of two mountains, one of which was Corra Bheinn in the Paps of Jura (the same one aligned from Ballochroy, nearly 30 miles to the SW). The midwinter setting Sun might have been visible for a moment or two from the top of the cairn (at A = 224°), but the cairn would need to have been higher than it is at present, or the observer would need to have been tall. In fact, the placement of the cairn could have been made without seeing the Sun from this site, if the site line were established with other markers.

On examining the site, Thom found a ledge on a slope from which the Sun could have been seen, and the location of the cairn established. Euan MacKie (1976) realized that the archaeoastronomical usage of the ledge was a hypothesis that could be tested archaeologically, and he tried to do so. He found that there were two boulders placed so that an observer looking between them would see the notch and a line to the notch would pass directly over the cairn. He then excavated the ledge on which the boulders rested, but found only indirect evidence that the site was artificial.22 Controversy developed concerning the visibility of the notch on Jura and the necessary height of an observer to see the notch from the platform (MacKie 1976, 1981; Patrick 1981; McCreery, Hastie, and Moulds 1982). The conclusion of McCreery et al. (1982) was that an observer would always have had to be taller than 5′ and that the amount of foliage on intervening trees could have obscured the view for an observer under 5′7″. Refraction variations, inherently unpredictable, would have contributed to the visibility problem. Whatever the validity of these criticisms for the epoch when the site was in use, the notch is plainly visible in a photograph published by MacKie (1976).

Another aspect of controversy involved the width of the ledge. MacKie (1976) claimed that the ledge was wide enough to observe the Sun for some days before and after the solstice. According to McCreery et al. (1982, pp. 187–198), the ledge is not wide enough. They attribute the MacKie claim to a confusion between the azimuth change caused by a declination change and that of the declination. We will discuss the use of a platform at Kintraw for precise observations of the Sun near the end of §6.2.15, after we describe Thom’s proposed observing method, and compute the necessary distance that would have been stepped by an observer to see the shifted azimuth caused by the changing declination of the Sun and Moon.

From the cairn, the lunar 9′ perturbation (see §3.4.5) could have been detected as the Moon set at minor standstill among the complex silhouette of the hills of Jura (Thom 1971/1973/1978, p. 39). The possibility that such precision could be obtainable in the megalithic is, well, incredible, and it is not surprising that many investigators have refused to believe it. In the next few sections, however, megalithic constructs are described that make it at least possible that the necessary information was indeed encoded in the monuments.

6.2.13 Merrivale Stone Rows

On Dartmoor in Devonshire, England, there are ~60 single, double, and triple rows of stones. One of these sites is at Merrivale, about 4 km west of Princetown. The site consists of two rows of stones with a cairn midway along the southern of the two rows, five other cairns, two cists, four stone circles, and isolated menhirs and stone slabs. The two rows are not quite parallel to each other, and extending to the SW from a point ~1/3 of the way from the west end of the southern row is a shorter row leading from a cairn just off the southern row to a pair of stones. The northern row has a bearing of 83.7°, and it is ~182 m long; the second has a bearing of 81.7° and a length of ~264 m. Both rows are actually two parallel sets of stones. The third row is single, with a bearing of 23°, and a length of 42 m. Wood (1978, pp. 130–139) suggests that there are several alignments, including some to the major standstill moonset, at this site, and that various features of the site provide triangles to compute means of extrapolation to be described in §6.2.15. As intriguing as this site is, there are others where the cases for alignments have been made, and where stone grids, touted by Thom as possible computational devices, exist.

6.2.14 Temple Wood: Clear Evidence of High-Precision Measurements?

The site of Temple Wood in Argylle has been called a lunar observatory by Thom (1971/1973/1978). The site consists of two major sections: the main site, a small circles of stones in which there is a smaller ring and a kist; and a secondary site consisting of a linear arrangement of groups of menhirs. The extremities of the secondary site are marked by pairs of menhirs. The southernmost pair (S4 and S5) is oriented approximately toward a notch in a distant hill at azimuth 317.9°. The bearing of the northernmost pair (S2 and S3) is toward another stone, S6, 110 m distant, at 316.0°. Between these two extremes are two clusters of stones. The southern cluster (“Group Q”), which in a 1939 survey consisted of four stones, three upright and one fallen, had, at the time of Thom’s study, only three—the fallen stone having disappeared in the interim. The bearing of the notch from these stones is 317.2°. Finally, the northern group of stones consists of the largest menhir in the complex, S1, surrounded by four smaller, upright stone slabs. Stone S1 bears cup markings. About 310 m toward an azimuth of 315° stands the circle of the main site of Temple Wood, and 2 km away at an azimuth of 317.0° is the foresight notch, to which the southern pair of menhirs already appears to point. The relationships between these bearings and the declinations of the setting major standstill Moon are shown in Table 6.6, based on Thom’s (1971/1973/1978) Table 5.1. Thom assumes an obliquity ε = 23°54′.3, a height of 1.68 m (5.5 ft) for the observer, a lunar inclination i = 5°8′.7, an apparent lunar radius (“semidiameter”) s = 15′.5 and a “declination” perturbation amplitude,23 Δ = 9′.0 (see §2.3.5). The mean declination of all the northernmost moonset azimuth alignment indicators is 29°3′.0, a respectable enough approximation to the extreme northern declination of the Moon. However, Thom’s conclusion of this analysis was that the various alignment indicators were intended to obtain highly precise refinements of the declination, as noted in the last two columns of Table 6.6. For instance, the azimuth alignment over Group Q to the notch was to the Moon’s center, with declination δ = ε + i; that of S4 − S5 to the notch indicates an azimuth alignment corresponding to δ = ε + i + s, while the alignment from S1 over the main site to the notch indicates a declination, δ = ε + is + Δ. Thom (1971/1973/1978, p. 48) suggests that the bearing differences of the smaller stones around S1 indicate the variation in Δ, about 1 arc-minute! Unfortunately for Thom’s hypothesis that megalithic exploration of the small variables was attempted, the necessary backsights to obtain high-precision alignments for δ = ε + is and for δ = ε + is − Δ are not present. Whether the higher precisions implied by the analysis were actually attained or not, the case for intentional alignments to mark the extreme setting points of the major standstill Moon would appear to be strong enough. But there is more evidence yet.
Table 6.6.

Observations and analysis of Temple Wood.





Obs. Dec.


Calc. Dec.

Group Q

Notch over Circle




ε + i

= +29°03.′0

S4 − S5

Notch over Circle




ε + i + s

= +29°18.′5


Notch over Circle




ε + i − s + Δ

= +29°56.′5


Notch over Circle






S2 − S3

Notch over Circle






Mean (ε + i)

Notch over Circle





Notch B, Bellanoch Hill




−(ε + i + s)

= −29°18.′5


Notch A2, Bellanoch Hill




−(ε + i − s)

= −28°47.′5

Mean [−(ε + i)]

Bellanoch Hill




Thom also discovered an alignment to the southernmost setting direction of the Moon from stones S1 and S2 to separate notches in the rolling profile of Ballanoch Hill, 6.3 km away at bearings of 203.8° and 207.9°, respectively, which imply a mean declination of −29.1°, but separate declinations δ = −(ε + i + s) ≈ −29.3 and δ = −(ε + is) ≈ −28.8°, respectively. The SW alignment from S1 is currently impossible due to groups of trees, but that certainly need not have been the case in the Neolithic.

Thom (1971/1973/1978, pp. 50–51) also suggested that alignments involving stones S3 and S6 could have been used to provide warning of the approach of the Moon to the maximum. If the perturbation in declination (the amplitude or maximum value of which, 9′, has been designated Δ) were to achieve a maximum at the same time that the major standstill did, the lower limb of the Moon (at δ = 28°29′) would have coincided with the notch as seen from S6 one and a half cycles or 260 days prior to the true moment of maximum standstill. If, on the other hand, the minimum were to occur at a major standstill, the lower limb of the Moon would have coincided with the notch as seen from S3 only 1 cycle or 173 days prior to major standstill. Finally, he suggested that the alignment involving the center of Group Q would have achieved the alignment to the Moon at declination δ = ε + i = 29°2′.5, to an uncertainty of 1′ only at the epoch b.c. 1700 ± 100, assuming the Moon’s inclination to be i = 5°8′43″. This is because of the slow variation of ε with time (see §4.4).

We are left wondering if high precision at particular sites, using particular combinations of stones and sightlines, was not being achieved afterall. Statistical arguments do not really touch this issue if there is clear evidence of distant enough foresights, or reasonably precise backsights and space for standing platforms at which to record the measurements in some way.

Thom, of course, carried out his own statistical analyses. In one of them (Thom 1971/1973/1978, pp. 75–79), he selected 40 sightlines at 23 sites and from them deduced four values: the obliquity, ε = 23°53′26″; the inclination of the Moon’s orbit, i = 5°8′52″; the major perturbation of the Moon’s inclination (seen in the effect of the declination on the azimuth), Δ = 9′23″; and a mean semidiameter of the Moon’s disk, s = 15′55″ (allowing for the upper limb, lower limb, or mid-disk to be the point of alignment on a distant foresight). Of course by preselecting these sightlines as being astronomically relevant, Thom’s argument can be construed as circular, but the results are impressive in any case. Now, we will explore possible procedures for observing and recording such observations.

6.2.15 Caithness Sites and a Potential Observing Method

At Mid-Clyth in Caithness in northeastern Scotland, there is a grid that perhaps once held as many as ~250 stones on a gently sloping hillside: “the Hill o’ Many Stanes.” Thom (1971/1973/1978, pp. 23–25; 86–90) saw the establishment of the extreme azimuths of the rising and setting Moon as a principal goal of megalithic astronomy, perhaps for eclipse purposes, or because of interest in the Moon for its sacredotal, luminary, or tidal effects. In §3.2.1, we mentioned Thom’s hypothesis that three components of the Moon’s motion could be measured from at least some selected megalithic sites. Thom believed he had found evidence that the requisite information was obtained by virtue of the alignments; he also believed that the information was stored in the geometric grids at Mid-Clyth and perhaps elsewhere and, moreover, that the grids provided computational assistance to determine those components.

The method of encoding the observational information allegedly began with a distance stepped off nightly on either side of the date of extreme azimuth of moonrise/set. The careful marking of where the observer stood to observe a given phenomenon would provide data to determine the moment of extreme azimuth (and therefore declination). We now discuss how this could be accomplished for particular sites (see Chapters 2 and 3 and references cited therein for the geometry and spherical trigonometry background).

For any particular alignment, the azimuth of a rising (or setting) object depends on the declination of the object, δ, the zenith distance, z, and on the observer’s latitude, φ. In general, the relationship (2.1) can be expressed as
The basic idea is this: The declination of the moon rises to a maximum sometime around its extreme rise (or set) azimuth during the month. The variation of the declination over a couple of days on either side of the maximum is assumed to be representable by a parabola of the form where Δt is the time interval from the moment of extreme declination, Δδ is the change in declination over that interval, and k is a constant.24 For the solstitial Sun, at δ = 24°, k ≈ 13 arc secs/day2. For the Moon, at major standstill, with δ = 29°, P = 27.d32, k ≈ 46 arc mins/day2, and at minor standstill, with δ = 19°, k ≈ 30 arc mins/day2. The interval between successive transits of the meridian by the Moon is given by the angular rate formula in terms of the mean solar day (Table 2.6):
The inverse of ω could be called a “lunar day” and is equal to ~1.0350 mean solar days (MSD). In half of such a lunar day, the declination can be expected to change by approximately and
For the Sun, the relations are simpler:
The change in azimuth arising from a such a change in declination is obtainable through differencing (6.7), keeping the latitude and zenith distance (the complement of the altitude, h) constant:
Therefore, solving for the change in A due to the change in δ alone, we get so that we may write the change in azimuth as
Any change in azimuth can be matched by stepping a number of paces perpendicular to the direction of the distant foresight. If we call the distance to be stepped off in, say, feet, ΔS, then the relation between ΔS and ΔA is where d is the distance to the foresight, and in the same unit as ΔS, and ΔA is in radian measure.
We may then write for the size of step to observe the Moon’s rising or setting at the same distant feature of the horizon: at major standstill of the Moon. This is an application of the principle of parallax discussed in §3.1.3: The larger the distance to the foresight, d, the larger the “baseline” shift, ΔS, required to follow a given shift in angle, ΔA.
Thom’s (1971/1973/1978, Fig. 1.1, p. 14) (see our Figure 6.33 for a similar idea) illustrates clearly that if the observer stepped step back on successive evenings a fixed number of paces, before moving laterally to determine the view to the Moon’s set point against the distant foresight and to set a stake or stone, then the Moon’s change in declination, scaled by some factor, would be illustrated by the curve on the ground created by the placements of the stakes or stones. Indeed, extending the arc between the two markers closest to the moment of extreme declination would give the precise declination (if the scaling factor were known) as well as the moment when this extreme declination occurred.
Figure 6.33.

The geometry of the “step” analysis shows the parallactic shift in azimuth needed to observe a given shift in declination. The foresight is in the direction of time’s arrow. Drawing by E.F. Milone.

Whether this purely geometric determination was ever carried out is completely unknown; that is, at present, no one has come forth with evidence of stones so placed at any relevant site, but the process is transparently clear, at least to us, millennia later! What is not so clear is whether the geometric interests of megalithic circle builders extended to the determination of the arcane motions of the Moon, and whether any class of them living in precarious times (see Burl 1987 for vivid accounts of life at Stonehenge!) really had the leisure to study a practical form of analytical geometry, contemplate the meaning and consequences of these motions, and engage in the painstaking work of making both the observations and the determinations. For the time being, we skip over these questions to explore with Thom further methods of determinations—at least how scientists of today transported back to ancient Britain in a Connecticut Yankee sort-of-way could have done so.

Thom (1971/1973/1978, pp. 86–90) related the lateral distances required to observe the shifts in azimuth to the geometric grids found at Merrivale, Mid-Clyth, and elsewhere. He defined a quantity G as that value of ΔS corresponding to the change in the extreme declination exactly half of a lunar day before or after the extreme. Because the Moon is not likely to be setting at exactly the same moment that it reached its extreme declination, G is not directly measurable. Yet Thom claimed that the value of G was nevertheless recorded at some megalithic sites. Here is how he thought this could happen: If ΔA is expressed in arc-minutes and d is given in kms (1000 m), then because there are 3438 arc-minutes/radian, the expression describes the change in position, in meters, in half a day at major standstill. The corresponding shift at minor standstill is

The difference ratio, ΔA/Δδ, is evaluated from (6.11). Recall that G represents the theoretical baseline shift of the observer in half a lunar day before or after the peak in order to preserve the aspect of the Moon over the same distant foresight. If the stakes are at the same spot, the distance to the peak position would be close to a distance G from the stakes.25 From (6.12), over a whole lunar day, the change in declination will vary by a factor of 4 over the 1/2 day change (12.3′ or 8.1′ at major or minor standstill, respectively) given above. The corresponding motion of the observer would be 4G, which may be a large distance.26 Thom’s suggested method of using this quantity was for the megalithic observer to place a stake each night at the location from which the alignment was seen against the selected and presumably distinctive foresight on the distant horizon. If two stakes were separated by a distance 4G, the position at the peak must lie close to one of the two stakes (stake positions on subsequent days will easily indicate which—if there is enough pacing distance at the observing site). A method of graphical extrapolation in position (corresponding to an interpolation in time) could have been used to obtain the position of maximum declination. Thom (1971/1973/1978, pp. 86–90; Fig. 8.2 illustrates the process). He describes two methods and Wood (1978, pp. 114ff) elaborates on them. The methods involve the use of triangles and sectors, such as those found at Mid Clyth and Merrivale.

Thom (1971/1973/1978, p. 87) shows that if observations of the lunar maximum were made exactly one lunar day apart, and if the alignment-giving backsight locations were marked by stakes or some other means, then the distance between the midpoint of the stakes and the point from which the maximum-declination Moon would have set, would be given by where 2p is the separation of the stakes. By dividing both sides of (6.17) by p, we get η/p = p/4G, from which η can be found by setting up a right triangle of adjacent side 4G and opposite side p, and laying off a distance p along the adjacent side (see Figure 6.34).
Figure 6.34.

The triangle method of determining the time correction to the extreme declination of the Moon, showing relationships among the quantities G, p, and η: See text for details. Drawing by E.F. Milone.

The length of a perpendicular from this point to the hypotenuse then defines a similar right triangle, and the opposite side of this small triangle is η. Because the ratios of like sides of similar triangles are equal, the ratio η/p = p/4G follows. If the leftmost stake represents the Moon’s position prior to the maximum, and the rightmost a lunar day later, after the maximum, then by moving the distance G + η to the right, the position of the Moon at true maximum would be determined. Thom (1971/1973/1978, p. 88) suggests that this is the method that was used in Argyllshire.

In a second analytic method, p is the arc of a sector of radius 4G (see Figure 6.35); if one moves in, along the radius, a distance p, and defines a second arc of length, say, x, at the radius (4Gp), then the difference between p and this arc length, px, is η.
Figure 6.35.

The sector method of determining the time correction to the extreme declination of the Moon, showing relationships among the quantities G, p, and η: See text for details. Drawing by E.F. Milone.

We demonstrate this as follows: If θ is the angle subtended by the arcs, then the relation between θ, p, and 4G is (4G)θ = p, so that θ = p/4G (expressed in radian measure). Then, because x = (4Gp)θ, x = pp2/4G, and η = px follows. Thom suggests that a variety of this method was used at Caithness.

When p > G, that is, when the stake separation (2p) is greater than 2G, so that one of the stakes lies fairly close to the maximum, one can define a quantity m = 2Gp. Then, setting m2/4G = Gp + p2/4G = Gp + η = yL, so that where yL is the distance from the closest stake (the rightmost stake in this example) to the place where the maximum declination would have been seen if it occurred at the moment of setting or rising.

A similar procedure was used, arguably, to find the maximum date from observations at several lunations (Thom 1971/1973/1978, pp. 89–90). The same equations are still applicable, but the time interval is a lunar month instead of a lunar day, and the specific values of the variables, such as G will be different.

In Thom’s view, the relating of the position of stakes to the location of the Moon requires only empirical ideas based on observation. Whether these ideas were in fact carried out is moot; if they were not, however, some alternative explanation for the sectored stone rows is required. This has not yet been provided.

Again, in Thom’s view, the value of the site at Mid-Clyth was that it provided a grid for interpolation, from which the foresight direction at the peak of the Moon’s extreme declination could be determined. The stones are arranged in grids of convergent sectors. The base and height of the main sector are 132 ft or 17L, where L = 20/7 MY, the grid element at Mid Clyth. Although Thom asserts that this is exactly equal to G (Thom 1971/1973/1978, p. 89), the calculated value of G is actually equal to 126 ft [38.36 m from (6.15)], within 5% of the correct value, if one uses the closest of the foresights discussed by Thom. More problematic is the quantity 4G, which should be the radius of the sector, if Thom’s proposed extrapolation scheme was used. The radius of the main sector is 360 ft and that of the SW sector, 413 ft, far from the calculated value, ~503 ft (Thom 1971/1973/1978, p. 104).

The evidence for alignments at the site is also provided by Thom (1971/1973/1978, pp. 93–95). There is a small notch only 1.8 miles (2.9 km) distant in an otherwise featureless western horizon to which a line of stones at the top of the hill seems to point. According to Thom, the directions to the notch from various backsight positions along the “Hill o’ Stanes” ridge mark the direction of rise of the Moon in a full range of variables: perhaps ε + is, ε + is + Δ, ε + i, ε + i + s − Δ, ε + i + s, and ε + i + s + Δ in the direction toward A ≈ 24°, and −(ε + i) with similar variations, to the southeast. However, the distance to the southeastern foresight is ~50 miles (~80 kms) and a value of G calculated for such a foresight greatly exceeds the scale of the known grid.

Three similar sites of stone sectors are found within 20 km of Mid Clyth: at Camster, Dirlot, and Loch of Yarrows (Wood 1978, p. 124). At Dirlot, a similar sectored grid of 70 to 80 stones is found, and as at the Hill o’ Stanes, the narrow end of the sector is uphill. Thom asserts that the radius of the base is 145 MY and that the grid elements are 3 MY apart. Here, there is evidence for alignment to the minor standstill moonrise but not with a distant, natural foresight. There is also evidence for the quantity 4G, here computed to be 398 ft, compared with the radius of the sector, 394 ft. The base of the sector, however, is 147 ft, not a calculated G = 100 ft. The azimuth of a direction of three menhirs of which only one is standing currently is 52.°3, corresponding to a declination of +19°11′, the extreme northern limit of a minor standstill moonrise (δ = ε − i + s + Δ). There is also an (as yet) unexplained precisely laid-out zigzag of stones with bearings 59°28′ ± 2′, 337°59′ ± 1′, and 61°33′ ± 1′.

The Loch of Yarrows is 8 km south of Wick and has a nonorthogonal grid. Several distant menhirs and a cairn could have provided artificial foresights, and Thom (1971/1973/1978, p. 99) suggests that a very small break in the contours of Tannoch Hill, about 1 km away, now filled with peat, could have been greater in the past. Alignments to δ = −(ε + is − Δ) for the artificial foresights and to δ = (ε + i) for the natural one are proposed. The measured length of the radii of the rows is 800 ft, compared with computed values of 756, 765, and 826 ft. Thom (1971/1973/1978, p. 98) indicates the lozenge-shaped grid element size to be 2.5 × 2.75 MY.

At Camster, there are two cairns north of a series of stone rows. Here, only about 33 stones are currently known in place, about half of those at the Loch of Yarrows. An alignment is noted to δ = −(ε + is − Δ), but Thom (1971/1973/1978, p. 100) indicates that excavations are needed to establish the radii of the sparsely filled grid rows with greater precision. From what is extrapolated, it appears that the radius is ~544 ft, against 4G = 606 ft sufficient, Thom felt.

Finally, we discuss the situation at Kintraw. From the platform above the gorge, the midwinter Sun’s upper limb may briefly twinkle when it is seen at a coll between Beinn Shiantaidh (43.7 km distance) and Beinn a’ Chaolais (46.5 km). Thom (1971/1973/1978, p. 38) suggested that a green flash might have been visible at this instant in what he regarded as the clearer skies of the Megalithic. For the Sun, (6.12) is still valid: but the constant k for the Sun is ~13 arc-seconds per mean solar day. With δ = −23.90°, A = 224°, z = 89.4333°, and φ = 56.18833°, ΔA/Δδ = 2.365. Therefore, 13 arc-seconds of declination, the change in the declination of the Sun 24 hours from the summer solstice, causes a shift of only ~30 arc-seconds of azimuth in this interval. For the Sun, so that here, G = 0.063 × 43.7 × 2.365 = 6.5 m (~21 ft), compared with Thom’s (1971/1973/1978, p. 38) value of 19 ft. From the description of the platform and contour chart given by MacKie (1974, especially, pp. 178–181), which shows at least 50 m (±4G) of a path length normal to the direction of the col, there should be ample room to establish the correct vantage point for the Sun and, therefore, to establish the location of the cairn and menhir in the intervening valley. However, McCreery et al. (1982) suggest that the col was not visible over this range. Thus, the controversy continues, but the feasibility of high-precision determination at this site seems clear. We conclude our comments on Kintraw by noting that it is interesting that a boulder is found on the platform in line with the col and the menhir, and that a petrofabric analysis by J.S. Bibby, appended to MacKie’s paper, supports a conclusion that the platform was deliberately created rather than a natural feature of the landscape.

6.2.16 Stone Rows at Carnac: Accidental Alignments?

The largest concentration of megaliths in the world is in Carnac in Brittany. As a site to be studied, however, it has problems. Thom and Thom (1971), who surveyed the site, had great difficulty in finding a stone that had not been moved, with any degree of certainty. Nevertheless, DHK, who has visited the site, notes that the stones are not readily moved, and unless there have been organized efforts by pranksters, farmers, or builders, large-scale systematic movement is not likely to have taken place, although cromlechs (mounds enclosed by stones) have been destroyed.

The dates of the stone rows at Carnac and elsewhere in Brittany are unclear. The cromlechs seem to have been the primary focus of interest to which the rows led. At least some of these cromlechs were in existence by about 3000 b.c. However, many of the rows appear to have been built by a process of accumulation over an interval of many centuries. Burl (1993, p. 146) writes, “It is interesting that the short axis of the Ménec West cromlech, an inverted egg, is in line with midwinter sunrise and that the long axis of its eastern counterpart points in the direction of the midsummer sunset as though the two rings were complementary, each for celebrations at the year’s ends.” He goes on to suggest festivals at the cross-quarter days of Beltane (early May) and Samain (early November). Burl points out that there is clear evidence of ceremonial fires associated with some of the cromlechs and that such fires were frequently built in association with equinoxes and solstices. A summer solstice bonfire was built on a burial mound overlooking the rows of Carnac as late as the 19th century.

We have already mentioned the very early erection of the menhirs and one of the most spectacular features of the Carnac area, Le Grand Menhir Brisé, or Er Grah, the largest known menhir. The broken fragments of this monument now lie on the ground at Lacmariaquer. If erect, they would have had a combined height of ~22.5 m (67 ft) and weighed ~340 tons. It was proposed by Thom and Thom (1971) that this was a universal foresight for several backsights in the Carnac region. From these various proposed backsights (the italicized names are currently existing backsights), the orientation to Er Grah reveals the following alignments:
  1. (1)

    Quiberon:     +(ε + i) rise

  2. (2)

    St. Pierre:     +(ε − i) rise

  3. (3)

    Le Moustoir:     −(ε − i) rise

  4. (4)

    Kervilor:     −(ε − i) rise

  5. (5)

    Kerran:     −(ε + i) rise

  6. (6)

    Trevas:     −(ε + i) set, and toward Quiberon

  7. (7)

    Tumiac:     −(ε − i) set, and toward Kervilor and Le Moustoir

  8. (8)

    Petit Mont:     +(ε + i) set, and toward Kerran


There is what Thom and Thom describe as an “extrapolating sector” near St. Pierre, and there are tumuli at Le Moustoir and at sites very close to Petit Mont and Tumiac. Thom and Thom find values of G of 114 ft for Le Moustoir and 94 ft for Kervilor. These sites could have provided extrapolation sites for the southernmost major standstill moonrise. From the proposed site of the northernmost major standstill moonrise alignment, ~1 km south of St. Pierre, there is no record of stones or other indications of a backsight, but the radius of the sector at St. Pierre is ~700 ft, which compares with the computed value of 4G for the site, 720 ft. Although Thom and Thom do not seem to have been aware of it, reports from the last century make it certain that the rows were once much longer and have been partially submerged by the sea (Burl 1993).

Thom and Thom deduced a value for the obliquity that indicated a date near 1580 b.c., but the uncertainties in measurement permitted an even later date. At the date of their work, the hypothesis that the Grand Menhir had been a major lunar foresight at 1580 b.c. or later seemed reasonable. DHK, however, agrees with Burl (1993) that the archa­e­ological evidence now makes the proposal utterly unlikely and that the Brittany study provides a very good example of the dangers of accidental alignments when the monuments being examined are not part of a single complex.

In addition to these interesting features, there are several local sites of stone rows in the area, which can be examined separately:
  1. (1)

    Le Menec stone rows

  2. (2)

    Kermario stone rows

  3. (3)

    Petit menec stone sectored grid

  4. (4)

    St. Pierre sector

  5. (5)

    Champ de menhirs


The Le Menec row and Kermario rows are in the north of the Carnac region proper. The Le Menec rows show a distinct direction shift about midway of its overall length, ~3000 ft or ~1 km. The Kermario rows, of similar overall length, to the northeast, show three such areas, but the westernmost is short, ~100 m. The width of both of these sets of rows is also ~100 m. Further to the northeast, near Kerlescan, there is another, much shorter sector oriented roughly eastward, toward Petit Menec. The sector at Petit Menec is shorter still; Thom and Thom (1971, Fig. 5) give its overall length ~93.6 m or 112 MY, and compute its radius as 225 MY, and its broad base consisting of 14 squares with sides of length 4 MY is 56 MY on the arc. The previously mentioned St. Pierre sector (Thom and Thom 1971, Fig. 6) seems to have squares of 10 MY sides at the base, 40 MY on the arc.

6.2.17 Megalithic Sites in Central Europe

In Oldenburg, near the city Wildeshausen, are seven pairs of long parallel stone rows known locally as Hünenbetten (“Beds of the Giants”). These are discussed by Müller (1970, pp. 75–81). Müller points out that in 1934, at a time when scholarly work on archaeoastronomy was virtually nonexistent, D. Wattenberg noted that the major alignment of the Hünenbett at Visbeker Braut (“bride”) pointed at the midsummer Moon’s major standstill position in the South (−29°). The row is closed at the end by four large stone pillars that act as a foresight in similar fashion to natural foresights suggested by Thom at numerous sites. It should be noted that the two outside pillars have pointed tops, whereas the two interior pillars have flattened tops, paralleling the situation described by Burl (1993) for British sites. At Visbeker Bräutigam (“bridegroom”), the Hünenbett is aligned nearly on the equinox (90.8° W of N). The other alignments are not as obviously meaningful. The archeological context seems to be in the Neolithic, ~2200–1700 b.c. The Hohe Steine (“tall stones”) site has four large boulders set in the middle of an oval ring of much smaller stones.

At Boitin, in Mecklenburg, are four stone circles referred to as Steintanze (“stone dances”) (Müller 1970, pp. 54–50, 81–84). Lines between the centers of three of these circles form an isosceles triangle, one side of which is only 1/2° off a line to true north, and the base of which aligns to the extreme southern moonrise (−29°) in 1800 b.c.

At Klopzow, another site in Mecklenburg, there is a double ellipse or circle, with an entrance and small stones to the southeast and large ones to the northwest. A line drawn from the entrance across the largest stones gives an azimuth of 315°. The horizon is flat at this site, which has a latitiude of 53.5°. This implies a declination of ~24°, summer solstice sunset (Müller 1970, pp. 84–85).

According to Müller (1970, pp. 85–88), a group of stone circles in Odry, in West Prussia, seem to be strongly inter-related, with alignments marking the equinoxes and the rise and set points of the Sun at both solstices and the major standstill southern moonrise and the northern moonset. There is also an alignment to a setting at δ = 33° (for the equinox −1760), said to be for Capella. All alignments are said to work best at about l800 b.c. It is of importance that the lunar and solar interests recognized for the builders of the British megalithic monuments have also been deduced for these similar continental monuments of about the same date.

6.2.18 Mediterranean and North African Megalithic Sites

Our knowledge of the megalithic archaeoastronomy of the Mediterranean is due particularly to Michael Hoskin and his associates, Allen, Gralewski, Ventura, Serio, Tusa, Morales, Papathanassiou, Papadopoulou, Hochsiedel, Knösel. Many of the monuments of Spain and the Mediterranean islands, especially the burial monuments are decidedly similar to those of the megalithic tradition in Brittany and the British Isles, but no one has yet suggested the presence of the complex astronomy that has been postulated for some of the British sites. Malta

The megalithic temples of Malta (Serio et al. 1992; Agius and Ventura 1981) began to be built about 3600 b.c., and construction continued for about a thousand years. Although the date and the tremendous stones used in the temples justify the term “megalithic,” these sites are markedly distinct from other megalithic sites. The first major difference is implicit in the word “temple.” To our knowledge, no one has challenged the appropriateness of this term in Malta and no one has demonstrated that it is really appropriate in most of the megalithic world. There are some generic parallels of later date in Sardinia and Menorca. Great circles, standing stones, and elaborate graves, usually with an opening to the southeast, are absent. The orientations of the temples show completely different astronomical interests. Shifting alignments through time suggest that asterisms (including single stars) were of primary importance in this culture, in contrast to anything that can be clearly demonstrated in other megalithic sites. Art carved on Maltese monuments included geometric figures, animals, and watercraft. Finally, and perhaps most significantly, there are indications of numeracy, astronomical interest, and astronomical record-keeping at a different level than that found elsewhere in megalithic cultures. The most striking example may be seen in the Tal Qadi stone (Figure 6.36). The carving shows dividing lines between apparently sectored groups of star symbols and one marked-off segment containing a possible lunar symbol. This suggests the possibility that the sky was formally divided into nearly equal areas, perhaps marked by asterisms. There are reported to be five segments in the stone fragment and may have been as many as 14–18 on the entire stone face.
Figure 6.36.

The Tal Qadi stone. The carving shows dividing lines between apparently sectored groups of star symbols and one marked-off segment containing a possible lunar symbol. Drawing by Sharon Hanna.

The work of Serio et al. (1992) shows that the 14 alignments of the temples of Malta (with the exception of the Mnajdra Temple I) are too far south ever to face a rising or setting Sun and at least 12 are too far south ever to face a rising or setting Moon. If the strongly clustered alignments are astronomically based, it is probable that they are associated with stellar risings, settings, or southern transits. They opt for the latter possibility with emphasis on α and β Cen and the stars of the Southern Cross. They think that the difference of alignments between Ggantija I at δ = −27.3 (corresponding to A = 125.5) and the somewhat later Ggantija II with δ = −33.8 (A = 134.5) may best be explained by the precessional shift.

At Mnajdra Temple I, the earliest of the three temples at this site, there are two pillars into which rows of dots have been drilled (Ventura, Serio, and Hoskin 1993). The features of the pillar are sketched in Figure 6.37.
Figure 6.37.

Pillar markings at Mnajdra Temple I on Malta. Drawing by Sharon Hanna.

The holes apparently represent a tally count because of structural similarity between the counts on the two pillars, summarized in Table 6.7, in which the counts are summed in the same way, from top to bottom. The rows are not similar in position on the two pillars, and all but the last two groupings on the East Pillar are horizontal. The upper half of the structure, at least, seems to reflect a common interest on both pillars, but exactly what that interest is remains obscure. Ventura et al. (1993) think that the tally refers to days and that the total of sequenced holes on the East Pillar may represent a half-year. However, the fact that Mnajdra I is aligned closely on equinox sunrise, which is also the direction of the heliacally rising Pleiades, eventually led them to an interpretation of the tally numbers as marking intervals between the heliacal risings of a series of bright stars or asterisms (last column of Table 6.7). They start the tally with April 6 (Day 96 of the current calendar) as the date of the heliacal rising of the Pleiades. The stars were presumably selected because of the proximity of their heliacal rises to the postulated dates at the ends of the sequences; the calculation of heliacal rise dates depends to some extent on the visual extinction coefficients (see § reasonably assumed to be between 0.20 and 0.25 (said to be based on unpublished calculations by Bradley Schaefer). With these assumptions, the dates of agreement are less than about 4 days apart at most; the authors argue that this degree of congruence is unlikely to be due to chance.
Table 6.7.

Structure in pillar markings at Mnajdra Temple I.

West Pillar

East Pillar


Line count


Line count


Postulated datesa

Associated asterism





Apr. 6






Apr. 25

α Tau





May 8






May 24

α Ori



3 + 3


May 27, 30

γ Ori




June 3

β Ori





June 27

α CMa





July 8

β CMa





Aug. 2

α Boo





Sep. 24

γ Cru





Oct. 2/3

β Cen





a Assuming each dot on East Pillar marks 1 day counted from the heliacal rising of the Pleiades.

b Nearly vertical line of dots and separated from the above sequences.

c Three pairs of dots forming a roughly triangular grouping, separated from the sequences. Iberia

One of the major megalithic sites of Spain is Los Millares (Mohen 1990, pp. 124–126, 153; Sieveking 1963, pp. 300–304, 313–315). The site was once regarded as an intermediary one, transmitting ideas from the Fertile Crescent and the eastern Mediterranean to the megalithic world of northern Europe. It is now recognized as a southern manifestation of a more general and ancient megalithic tradition. The Los Millares culture has been dated to ~3000–2500 b.c. One impressive monument is a large burial mound surrounded by three concentric rings of menhirs.

The “tholos” tombs of Los Millares were so named because of similarity to the architectural style of Cretan and Mycenean tombs, which are elaborate corbelled domes within burial mounds with large stone-faced entrances. The latter, however, are now known to be of much later date. The tholos tombs frequently have “porthole” entrances. Hoskin, Allen, and Gralewski (1995) report that of 48 tombs with measurable orientations at Los Millares, two were to the southwest, four were somewhat to the south of the midwinter sunrise point, and the rest were oriented to points between midwinter sunrise and midsummer sunrise. The 11 tholos tombs at Barranquete, southeast of Los Millares, were oriented entirely to points between the east and south. The orientations of the tholos tombs have a range similar to those of the tombs of the Montefrio area (Hoskin et al. 1995, p. S69). Azimuths of 41 tomb entrances were measured. Of these, 30 lie in a range from due east to the azimuth of midwinter sunrise, with one alignment at each extreme. Seven fell somewhat north of east and four were south of midwinter sunrise.

Elsewhere in Andalucia, there are several different kinds of burial mounds with a general resemblance to those of Brittany and the British Isles (Hoskin et al. 1995). The passage leading to the burial chamber usually has its entrance in the southeast quadrant. Many more southerly orientations are found here than at Los Millares and Montefrio. Of 198 tombs, the azimuth of which could be determined with assurance, 164 were to the southeast quadrant, 3 (of which one is doubtful) to the southwest quadrant, 1 pointed due south, and 20 to the northeast quadrant. Individually, the most remarkable single monument is the gigantic Dolmen de Menga at Antequera (Hoskin et al. 1994a, pp. S79–S80), which faces northeast. The burial chamber is 18.5 m long and 6 m wide, narrowing to a passage 5 m long and 3.5 m wide. The entire structure is roofed by only five slabs, the largest of which has a volume of 120 m3.

Hoskin et al. (1994b) think that there is no evidence in Iberian tombs of alignments to stars or to the Moon. They suggest that many tombs were aligned on the rising points of the Sun at various times of the year and that the others were aligned so that the passage of the Sun across the sky would illuminate the tomb interior. The pattern is suggestive of the alignments of medieval churches, many of which were set so that the Sun illuminated the high altar on the Saint’s Day of the church (see Heilbron 1999). Balearic Islands, Sardinia, and Pantelleria

The Island of Menorca contains four megalithic sepulchres; neighboring Mallorca and Formentera each contains one. They apparently date from early in the 2nd millennium b.c. (Hoskin and Morales 1991). All are oriented in western and southwestern directions, with azimuths from 220° to 278°. There are also somewhat later communal burial tombs of large stone blocks. The tombs are called navetas, from a resemblance to overturned boats, and may derive from the earlier tradition. Hoskin and Morales divide the navetas into three groups: oval navetas restricted to the east, and eastern and western elongated navetas. The orientations of the six western elongated navetas all fall within the limits of the orientations of the Balearic megalithic sepulchres. The orientation of the six oval navetas and the six eastern elongated navetas center on the south. Only one in each of the latter groups is within the range of the western navetas. Thus, all of these alignments are substantially different from those that are common elsewhere, but there are also markedly distinct subgroups.

Hoskin (1985, 1989, 1991) also discussed Menorca towers, called talayots, which were constructed between 1400 b.c. and the Roman conquest, and they may represent the last remnants of a megalithic tradition. The towers tend to dominate villages that often also contain a physically bounded area or “precinct” with a taula (“table”). These are massive monuments composed of two limestone slabs, one upright and the other horizontal forming a large tau-shaped object. Whether the towers and precincts were sacred or secular is disputed, but Hoskin has come to think that the precincts were sacred areas or sanctuaries. The precinct entrance is usually southward, and Hoskin has suggested that they were aligned to face the passage of the constellation of the Centaur, Chiron (an attested name for this constellation). One of the taula-precincts contained a statue of the Egyptian architect, Imhotep, later deified and associated with medicine. He was equated by the Greeks with their god Asklepios (or Aesculapius). Asklepios was said to be a pupil of Chiron, who was renowned for his healing powers. One of the taula precincts had an alignment coinciding with the heliacal rising of Sirius. Hoskin (1991) points out that there was a ritual on Mount Pelion at the time of the heliacal rising of Sirius. Mount Pelion was said to be disputed territory between Centaurs and Lapiths, whose royal lines had a common origin (Graves 1955–1957 I, pp. 360–362). The mother of Asklepios was said to be a Lapith princess. Aristaeus, a grandson of the Lapith king Hypsaeus, learned Mysteries in Chiron’s cave and was taught healing by the Muses. He is credited with ending a plague sent to the island of Cos by the Dog Star, Sirius (Graves 1955–1957, I, pp. 277–278). These details offer some support for Hoskin’s views. Hypsaeus was the son of one Naiad and the husband of another (Graves 1955–1957, I, pp. 276–277). Naiads were daughters of Phorcys (Boar). The Latin equivalent of Phorcys was Orcus (Graves 1955–1957, p. 129, II, p. 107), whose name is probably incorporated in the names Menorca and Mallorca.

The towers of Sardinia (Hoskin et al. 1993) resemble the talayots of Menorca, which are simpler, and the so-called “Tombs of the Giants” resemble the Menorcan navetas. They seem to be approximately contemporary. Many scholars have supposed a close cultural relationship between them. Hoskin et al. (1993, p. S24) thinks that the similarites are more likely to be due to limitations imposed by the building materials and techniques. In any case, alignments in the two areas are decidely distinct. The Sardinian monuments share the generalized southeast orientation that is so widespread in megalithic monuments.

On Pantelleria, there is a group of communal tombs, called sesi, dating from ~1800 to 1600 b.c. The largest is a mound with 11 entrances and 12 passageways leading to separate chambers inside the mound. There is no discernible clustering of orientations in or among the 42 monuments examined (Tusa, Serio, and Hoskin 1992). Northern Africa

Somewhat south of Tangiers in North Africa, at the site of Mzorah (5°56.′73 W, 35°24.′59 N; 118 m above sea level), there is a major megalithic monument, a great ellipse of menhirs surrounding a tumulus (Mavor 1977). Mzorah means “Holy Place.” As can be seen in Figure 6.38, the menhirs are closely spaced.
Figure 6.38.

The tumulus and menhirs of Mzorah, North Africa. Drawing by Sharon Hanna.

There are 168 menhirs or their broken stumps still in place, but only ~1/3 are erect and unbroken. Mavor’s careful mapping led him to conclude that there were originally 175 present. Mavor suggested that the ellipse appeared to have been laid out on the basis of a 37-35-12 Pythagorean triangle, a type of structure that Thom claimed as the second most common pattern found in the British Isles. Mavor gives a full list of the distances and azimuths of all the menhirs, as measured from the center. Three tall stones (Nos. 130, 131, 132) mark the west point. The south point is marked approximately by a single stone (no. 90). Mavor notes a substantial number of other possibly intentional alignments, including all solar solsticial and equinoctial alignments and lunar major and minor standstill rising alignments and at least one setting alignment. He assumed that the ellipse had been used for observing for a substantial length of time before the tumulus was built, arguing that somewhat similar tumuli were being built in Morocco in the first millennium b.c. and that the menhir circle had affinities with British megalithic monuments dated by Thom in the early 2nd millennium b.c. on the basis of stellar alignments, which the Mzorah monument shares. However, Thom’s stellar alignments are not generally accepted because archeological evidence makes it unlikely that the British monuments are as late as Thom thought, and, with 175 possible targets, and a center point defined only by assumed geometry of the ellipse, it is difficult to ascertain which of the alignments was indeed intended. The coincidence of the eastward thrust of the major axis with the direction of rise of the Moon at minor standstill (between stones 35 and 36) is suggestive, however. This direction is also toward the summit of the mountain Jbel Si Habib.

Mavor makes the point that the ellipse must have been there before the tumulus because it would have been impossible to lay out if the tumulus had been in place. It seems likely to be in the same megalithic tradition as the stone circles of northern Europe, and Mavor cites anthropological and archeological evidence to connect the cultures of neolithic and early Bronze age Europe to those of Morocco.

The monument is unusual because it is mentioned in the writings of Pindar [518–438 b.c.], a poet from Akragas (Agrigento in present day Sicily), as the burial place of the giant Antaeus, King of Libya (a former generic name for Northern Africa), said to have been a son of Poseidon and Mother Earth, and killed by Heracles. The name Libya is taken from the name of the daughter or granddaughter of Io (who was transformed into a white cow by Zeus) and of a woodpecker (or Zeus in the form of a woodpecker), apparently regarded in the story as the king of birds (Graves 1955–1957).

6.2.19 Megalithic Summary

At present, megalithic material from areas outside Brittany and the British Isles is still too sporadically recorded to be included effectively in a general summary. Extremely large stones were used in building monuments in many parts of the world (Mohen 1989, pp. 42–67). In a purely descriptive sense, such architecture may properly be called megalithic. Many scholars at the beginning of the 20th century assumed that all such monuments belonged to a single historical tradition, ultimately derived from Europe. We know of no attempt by any archaeologically competent scholar to defend such a view in the last 50 years. Certainly, some combinations of earth mounds as burial places with large stone monuments seem strikingly similar in widely separated areas, but independent origins seem likely for most of them. For example, the “megalithic” of India (ca. 800–100 b.c.) is now known to follow the Early Iron Age and seems to be associated with Dravidian languages (Parpola 1994, pp. 172–173). In some cases, approximate alignments are suggested by plans and descriptions although astronomical associations of most of these monuments are still undemonstrated at the present time.

For Brittany and the British Isles, we can provide some rough conclusions about the state of present scholarship. The earliest evidence for astronomical interests appears in connection with large burial mounds in Brittany and western Ireland. Entrance passages to these tombs usually had an orientation to the southeast, sometimes specifically to the winter solstice. Early megalithic cultures probably spread largely by sea, and astronomical observations would have been important in navigation as well as religion. Artistic motifs may have been used as notations of astronomical phenomena. Gigantic single pillars of stone were sometimes erected during this period, but clear indications that they were used astronomically are lacking. Somewhat before 3000 b.c., large stone circles were built. These were apparently used as meeting places and incorporated community beliefs about cosmic order, which embodied astronomical orientations and therefore the timing of ritual events. There is very good evidence at this period for alignments to sunrise at winter solstice and good evidence for equinoctial horizon alignments. There are also possibly deliberate alignments on the cross-quarter days, intermediate between solstices and equinoxes. Evidence for interest in smaller divisions of the year or for alignments on stellar rising points is very weak, as is evidence for lunar alignments, at this period. This situation began to change at about the time of the appearance of avenues in the form of long rows of paired stones in the mid-3rd millennium b.c. An interest in the major and minor lunar standstills, and perhaps in eclipses, probably antedates 2000 b.c. Clear data about such interests seem to be associated with shorter rows of stones. Fans of stone rows seem usually to align on associated cists and to date well after 2000 b.c. The suggested function of the fans for measuring the lunar wobble (the 9′ of inclination variation) is still strongly disputed. However, by about 1500 b.c., short rows of stones seemed to be used for making and recording sophisticated lunar and solar observations. Near the end of the megalithic tradition, pairs of disparate stones (tall and pointed vs. squat and rounded) appear, but have no clear astronomical significance.

6.3 New World Medicine Wheels

At about the time that megalithic astronomy was at its height and Egyptian astronomy was just beginning, a parallel tradition arose in North America. Similar to the megalithic monuments of Europe in structure and perhaps in function, the medicine wheels of North America span the millennia. The term “medicine” is here used in the American Indian sense of sacred power, including the possibility of healing. This tradition shows marked internal continuity but with a poorly defined relationship to later American archaeoastronomical data. It is, therefore, treated separately.

The medicine wheels are found predominantly in Alberta, but are found also in Saskatchewan, Montana, and Wyoming, with others reported as far south as New Mexico and Arizona. More are being reported every year. Brumley (1986) and Brace (1987) have summarized materials on medicine wheels and classified them into descriptive groupings. They reflect the views of “dirt archeologists,” who have first-hand familiarity with many of the sites but who did not attempt detailed appraisals of the archaeoastronomy. Vogt (1993) has provided a comprehensive and critical study of known medicine wheels. Vogt cites references to 134 medicine wheels, of which 94 had adequate information for purposes of his statistical analyses. These analyses were the first attempt to establish groupings based on statistical clusterings of traits. The “important” characteristics for this purpose are those that permit clustering. This creates a degree of circularity in the reasoning and does not allow for the very real possibility that the important characteristics to the builders were those that distinguished one wheel from another.

The layout of medicine wheels is suggestive of megalithic circles, but the use of piled up rocks and boulders rather than standing stones decreases the precision obtainable relative to the megalithic circles of Europe. It has been suggested that poles may have been stuck into the tops of cairns; that could have substantially increased precision. There is some direct evidence that that occasionally occurred. The use of alignments to horizon markers has also been suggested in some cases. The sizes of medicine wheels vary greatly, ranging from less than 10 to more than 100 m. Ideally, there is a central stone cairn, a surrounding circle or oval, a series of spokes, and some outlying cairns. Some of the medicine wheels have a shape approximating that of a turtle rather than a circle, and they may belong in the general category of boulder effigy figures (also called geoglyphs). Modern Indians claim that many of the medicine wheels are burials or memorials to dead chiefs, but only two are known to have burials in the central cairn.

Vogt (1993) makes a number of generalizations about medicine wheels based on his survey of the literature. They are normally on top of the highest local hill, usually with a good view in all directions. They are often regarded by local groups as sacred places. Many sites have associated caches and offerings, extending, at least at Majorville, over several thousand years. There are patterned repetitions of statistically defined types of medicine wheels. These include clusters of directional alignments. Fifteen sites show markers for true north, to the extent that the basic data are trustworthy. Vogt has demonstrated that previously proposed adaptations to local features of topography are inadequate to explain these clusters. He maintains that astronomical targets are the only adequate explanation for the alignments. Moreover, not all can be explained by solar, lunar, or planetary data. Some targets must have been either individual stars or asterisms. Precision may not have been an important factor in alignments. He maintains that these alignments were “scientific elements of what probably were largely religious knowledge systems.” Vogt (1993) argues that the types of horizon-based observations earlier suggested by Eddy and others would often have been obscured by tents if the tipi rings, which are frequently found at such sites, are a good indicator. This throws doubt on the use but not necessarily the purpose because the layout could have been done in accordance with horizon observations and the tipis may not have been erected until the appropriate observations had been made.

Vogt (1990, pp. 48–49) discusses the way in which Plains Indians symbolize the cosmos in their dwelling places and ceremonies. He emphasizes the importance of the central pole of the Sun Dance lodge as representing both the “World Pole” and the center. The World Pole is the conceptual axis around which the sky revolves, associated in modern times with Polaris. The Sun Dance was particularly associated with the full moon following the summer solstice (Vogt 1993). Each family group had its assigned position in a circle of tents surrounding the Sun Lodge and in more permanent camps as well (Brace 1987, pp. 122–124). Among more southerly Siouan groups, it is attested that each clan was associated with a particular star or asterism. Hence, any alignment would indicate a particular clan (as Brace suggests) and a particular asterism. Given the view that the cosmos is reflected in the encampments, it seems likely that the association of clans and asterisms frequently held, even in groups for which the association is not attested. Such an association might have involved strict alignments from the center point toward the horizon rising point of “their” asterism among some groups during some periods and a much looser and astronomically imprecise alignment for other groups or during other periods.

6.3.1 Majorville, Moose Mountain, and Big Horn

There are three sites that are of particular interest for archaeoastronomy because of striking similarities in their structural layout and because of their wide separation in space and time. These are the Majorville site in Alberta, the Moose Mountain site in Saskatchewan, and the Big Horn site in Wyoming (Figure 6.39). The Majorville site is archeologically the most important, but its archaeoastronomy is the least well known.
Figure 6.39.

Three principal medicine wheels: (a) Majorville, Alberta. Air photo, courtesy J. Calder. (b) Ground photo, courtesy Tom Head. (c) Moose Mountain in Saskatchewan on the morning of June 21, 1978. Although impressive in appearance, the photo indicates that the expected alignment was not seen, according to Rodger, leading to doubts about the purpose of the monument. Photo, courtesy David Rodger. (d) Big Horn, Wyoming, view from Cairn A. Principal cairns and directions are marked. (e) Big Horn view from Cairn C. The foreground fence posts indicate, from the left, respectively, (1) cairns D (near) and E (far); (2) F; (3) A; and (4) B. Big Horn photos courtesy, Sharon Hanna. Majorville

Probably somewhat before 3000 b.c., a circle of rocks and a central stone cairn were built at Majorville, near the community of Cluny in southern Alberta. There are several outlying cairns and a series of spoke lines. The site has been badly damaged, but the main features can still be discerned (see Figure 6.40a). Portions of 26 spokes can be seen, although on the basis of later structural parallels and structural analysis of this site, it is likely there were originally 28 spokes. It is possible that all the spokes were added later, but no evidence precludes their presence in the original structure. Major excavations were carried out at the site over a period of several years under the general supervision of Richard Forbis of the Department of Archeology of the University of Calgary with much of the field work by James Calder (1977). A wide range of artifacts were recovered in good stratigraphic sequence. Apparently, the cairn was being increased in size from ~3200 b.c. until after 1800 a.d., except for a hiatus in building, although not necessarily in use between ~1000 b.c. and 200 a.d. Table 6.8 lists the archeological phases, components of which were found at Majorville.
Figure 6.40.

The Minton Turtle effigy in south-central Saskatchewan: The principal alignment from the tail to the head would have coincided with the helical rising of Sirius at summer solstice at ~2300 b.c. Drawing by Sharon Hanna.

Table 6.8.

Archaeological phases at Majorville.






Oxbow Complex










Old Women



Astronomical alignments have been claimed for this site but are not yet fully demonstrated, and the degree of damage to this site is so extensive that convincing demonstration is difficult. A museum model of Majorville at the Department of Archeology of the University of Calgary shows alignments to the summer solstice sunrise and to heliacal rise points of Sirius, Aldebaran, and Rigel. Studies by Gordon and Phyllis Freeman (1992) maintain that there are several types of distant foresights and local equinoctial markers as well. They raise the interesting point that the equinox, defined as the date at which the Sun crosses the equator, does not actually correspond to the date when day and night are equal because of dip and refraction considerations. They maintain that the alignments work best about three days before the vernal equinox, which is the date when observationally day and night were equal. However, they do not explain how people could have observed or calculated the small temporal difference involved. The current value of the site is its demonstrated antiquity and its structural similarity to younger, but less complex sites, for which astronomical alignments can be shown. Moose Mountain

Moose Mountain is located in southern Saskatchewan (see Figure 6.39c). It has five major outliers connected by long spokes to an egg-shaped ring of small stones surrounding a central cairn. There are also smaller, isolated piles of stones that appear to be shaped. The alignments have been studied by Kehoe and Kehoe (1977, 1979) and by Alice Kehoe (1981) in conjunction with J. Eddy. Their work indicates that a line from the southwestern outlier along a continuous spoke that ends at the central cairn and through the center of the cairn points directly to summer solstice sunrise. The three lines of sight from the northwestern outlier suggest stellar alignments: The most obvious is that along the spoke, which again crosses the ring to terminate at the central cairn, and through the cairn’s center. The alignment is to the rise point of the star Sirius, which Eddy indicates is correct over a large range of dates. A line from the NW outlier through an outlier to the NNE aligns with the rise of Aldebaran, α Tau at ~1 a.d. or somewhat earlier, and a line through the eastern cairn points to Rigel for a later date. These alignments are found more convincingly at the Big Horn site. Ovenden and Rodger (1978) argue that the alignments are not convincing and that the monument was constructed for other purposes. Big Horn

Of much more recent vintage, the Big Horn Medicine Wheel is located in Wyoming’s Big Horn mountains, at close to the 3000 m level. At present, the site is inaccessible in winter months, and its remoteness has saved it until recently from despoiling vandalism. A branch of a tree found in the central cairn has been dated by the tree ring method to ~1760 a.d. The site has been discussed at length by Eddy (1974, 1977a,b, 1978a, 1979).

From the layout (Figure 6.39d, e), the number of spokes is 28, suggestive of a sidereal month calendar device and of Indian Sun lodges with their 28 rafters and central pole. There is a flattened ring of small stones distributed at an approximate distance of about 7 m from the central cairn, which is much less prominent than that at Majorville. The central cairn is the oldest part of the site. The circle and spokes were added later. Some archeological evidence suggests use of the site into the 19th century. Five outlying cairns are found just outside the ring, and one is found just inside. The Southwestern cairn is farthest away from the ring and linked to it and to the central cairn by a spoke. John Eddy (1974) recognized the possible astronomical importance of the site and argued that it was designed to emphasize the summer solstice both by solar and stellar alignments. A line across this outlier through the central cairn, along the spoke points to the summer solstice rise part of the horizon. A line drawn through from the southeastern outlier through the central cairn points also to the summer sunset point. The northwestern “outlier,” which alone stands inside the ring, suggests stellar alignments. A line to the NNE outlier points, with a modest precessional correction, to the rise point of Aldebaran; through the eastern outlier, a line points to Rigel, β Orionis; finally, a line through the central cairn points to Sirius. The alignment date to give the appropriate precession correction is ~1700 ± ~200y, in agreement with the archeological evidence for the time of use of the site. The only cairn not used in these alignments was found by Jack Robinson to mark the heliacal rise of Fomalhaut about a month before the summer solstice between a.d. 1050 and 1450, earlier than the other evidence would suggest.

The Big Horn Medicine Wheel is a central feature of an area that has been considered sacred by many American Indian tribes. According to Tribal Elders (Price 1994, p. 260), this area has always been a neutral ceremonial area for all tribes, even during times of warfare. The continuing sacredness of the site has created conflict with the Forest Service, who have used the area for a multitude of economic purposes and with tourists who wish to see this interesting site (see Price 1994 for a full discussion). The sacred status of the area has recently been recognized, and some provision has been made to protect the site from damage that tourists often cause, whether due to carelessness or vandalism.

Vogt emphasizes that the particular stars in alleged use at Moose Mountain and Big Horn are not found as important markers at other sites and that there is no ethnoastronomical evidence for the importance of the alignments. However, our knowledge of ethnoastronomy is not so great that the absence of references should be considered a strong argument. Vogt (1993) also suggests the possibility that alignment targets are more likely to have been asterisms than individual stars, which would substantially reduce precision.

6.3.2 Other Medicine Wheels

Eddy (1977a, pp. 160–162) discusses a medicine wheel at Fort Smith, Montana, which has six spokes. One of these is aligned to the summer solstice sunrise, but the other spokes are not discussed.

Alice and Tom Kehoe (1979) describe effigy figures and medicine wheels that coincide with astronomical alignments. The most striking of these is the Minton Turtle effigy in south-central Saskatchewan (see Figure 6.40).

The principal alignment is from the tail to the head; it would have coincided with the heliacal rising of Sirius at summer solstice at ~2300 b.c. At 1 a.d., there would have been a shift of about 4° off this axis according to Eddy (cited in Kehoe and Kehoe 1979, p. 13). There is also a good summer solstice sunrise alignment and a possible sunset alignment for the same day. The height of the central mound would have prevented observation of the Sirius rising (at the horizon) with any precision unless there were a marker, perhaps a pole or a standing human. It is also possible that the central cairn was built later. But whereas at the 2300 b.c. date, Sirius’s horizon rising could not have been observed along the axis if it was blocked by the central cairn (given the present height of the cairn), at about 1 a.d., Sirius would have appeared to rise over the center of the cairn (again at present height). The height of the central cairn also would have affected the visibility of the Sun at horizon rising, but a precessional shift would not have helped in this case. This monument has been identified as a representation of a badger by Brace (1987, pp. 92–93) on the basis of information from Cree and Assiniboine informants. This reflects the assumption that the uppermost features of the figure are “ears”; however, they also correspond strikingly with the eyes of a turtle viewed ventrally (see, for example, Figure 15.3).

At the Roy Rivers wheel, there seems to be a marker for summer solstice sunset. Although the Kehoes do not make such an argument, it could also be postulated that one of the 15 small cairns at the site marks summer solstice sunrise, which, however, still leaves 14 cairns unexplained. This brings us to the more general idea of the multiple purposes evidenced by the many different forms of medicine wheels.

6.3.3 Astronomical Meaning of the Medicine Wheels

The importance of the medicine wheel alignments lies in their alignments; the solar connection is expected and seen. The stellar alignments are meaningful if the heliacal rise dates of the three bright stars are noted; functioning in this way, the Majorville, Moose Mountain, and Big Horn medicine wheels have a solar calendrical importance. The evidence of these three sites suggests that there was a continual tradition of the use of medicine wheel sites, although it has not been shown thus far if the astronomical usage is older than Moose Mountain or if lunar observations were indeed made from any of these sites.

All wheels that show characteristics that suggest astronomical intent also have cairns or alignments that are still unexplained archaeoastronomically. This weakens the case for accepting the astronomical purpose underlying these monuments. If anyone could offer probable astronomical explanations for any of these unexplained alignments, the case for accepting the interpretations of Eddy and others would be correspondingly strengthened.

This concludes our discussions of the stone monument building traditions in the Old and NewWorlds. We proceed now to the Mediterranean to discuss the roots of western astronomy.


  1. 1.

    The same criticism applies to the use of certain features of a site, for example, the Aubrey holes at Stonehenge, which are to some a stone age computer (or perhaps abacus?), and to others a collection of 56 chalk-filled holes arranged (more or less) around a circle.

  2. 2.

    See Broadbent (1955). The test involves the average of the squares of deviations.

  3. 3.

    The modern inch, span, fathom, and foot all have such anthropological origins.

  4. 4.

    Recall that we have defined azimuth to be measured from the North Point (in the Northern Hemisphere) positive eastward. With a similar convention for treating azimuths at Southern Hemisphere sites, measured from the South Point eastward, the same comment holds for the Southern Hemisphere.

  5. 5.

    Tradition implies a common background. A convergent tradition involves groups becoming more alike, usually because of common external factors. A parallel tradition merely preserves common, older features, but that may not have been originally explicit.

  6. 6.

    We say “almost” because the obliquity of the ecliptic has decreased by about 0.5° over that interval, decreasing the azimuth of rise of the midwinter Sun by about 1°.

  7. 7.

    In her book on Polynesia, Makemson (1941, p. 22) refers to “the spiraling path of the Sun.”

  8. 8.

    The designations of the standing stones are those of M.J. O’Kelly: proceeding clockwise from the tomb entrance, but allowing for gaps, as GC1, GC3, and so on; and the stones proceeding in a counterclockwise direction just to the northeast of GC1 are labeled GC-1, GC-2, and so on.

  9. 9.

    Generally, the chance probability in achieving 1 successes out of N trials in any situation in which the probability of a “success” is p can be written: NC1p1(1 − p)N−1, where NC1 is the number of ways (combinations) to achieve these successes: NC1 + N!/[1! × (N − 1)!] = [N × (N − 1) × (N − 2) × . . . (N − 1 + 1)]/[1 × 2 × . . . × 1].

  10. 10.

    Celtic deities, later euhemerized to become early rulers of Ireland.

  11. 11.

    The rising line of the ecliptic is tilted toward the south in the spring and north in the fall (see Figure 2.20). Thus, the diurnal path of the Sun continuously varies in a small but discernible way and with it the azimuth and altitude that together determine shadowing effects. Here, the effect may be enhanced by altitude variation along the eastern horizon.

  12. 12.

    Thor, the northern axe and thunder god, became the ruler of our day Thursday, the day of Jupiter in the planetary week.

  13. 13.

    Linguists use an asterisk to mark recontructed forms. See §11.4, footnote 10.

  14. 14.

    The other coves are located at Stanton Drew in Somerset (where it lies WSW of the main circle) and faces roughly south, at Cairnpapple in West Lothian and faces east, and at Avebury, where it faces NE.

  15. 15.
    In Greek, σφαιροειδη.
  16. 16.

    “Community” may be a better rendering of the Greek πóλιν in this case.

  17. 17.

    The wren appears as Bασιλισκος or “little king” in classical Greek, as Regulus and rex avium in Latin, and similarly in Italian, Spanish, French, German, Dutch, Danish, Swedish, English, and Welsh.

  18. 18.

    In 18th-century Ireland and on the Isle of Man, the wren was killed on Christmas Day and was hung by a leg from two hoops crossed at right angles.

  19. 19.

    It was this feature that so impressed Thom and was so critical in the development of modern archaeoastronomy.

  20. 20.

    Geoffrey of Monmouth’s History of the Kings of England contains a summary of legends regarding the site.

  21. 21.

    Named after John Aubrey. Aubrey inspected and reported on the site for Charles II, beginning in 1663. His description of the chalk-filled holes is contained in his unpublished manuscript entitled Monumenta Brittanica now in the Bodleian Library, Oxford.

  22. 22.

    A stone layer that extended upstream from the boulders appeared to be artificial in terms of location and in the orientation and dip of the stones themselves.

  23. 23.

    The perturbation is actually on the inclination; at a major/minor standstill, however, the perturbation adds directly to the declination; see Figure 2.17b.

  24. 24.

    The apparent variation in declination of the Moon, whether in the 18.y61, the 27.d32 or 173.d3 cycles, or that of the Sun in the 365.d2422 cycle, is sinusoidal and at the peak of the cycle can be approximated by an expression such as Δδ = δmax − δmax × cos(2π × Δt/P), where Δt is the time interval from the maximum value of the declination, δ. From this, Δδ = δmax × [1 − cos(2π × t/P)] = δmax × [2 sin2 (π × Δt/P)], which for short intervals from the moment of maximum (i.e., for Δt << P), becomes Δδ = 2δmax × (π × Δt/P)2 = kt)2.

  25. 25.

    The quantity G can also be understood geometrically as a sagitta. Given the stepping technique, and the tracing of a parabolic arc on the ground, it is the line between a point on the arc midway between the two stakes and the center of the straight line between the two stakes.

  26. 26.

    One of the questions raised about the site of Kintraw, discussed above, was whether the ledge provided sufficient space to recreate the alignment within a day.


  1. Agius, G., and Ventura, F. 1981. “Investigation into the Possible Astronomical Alignments of the Copper Age Temples in Malta,” Archaeoastronomy 4(1), 10–21.ADSGoogle Scholar
  2. Atkinson, R.J.C. 1956/1979. Stonehenge: Archaeology and Interpretation. (London: Hamish Hamilton, 1956; Penguin, 1979).Google Scholar
  3. Atkinson, R.J.C. 1961. “Neolithic Engineering.” Antiquity 35, 292–299.Google Scholar
  4. Barber, J.W. 1972. The Stone Circles of Cork and Kerry: A Study M.A. thesis. (Cork: National University of Ireland).Google Scholar
  5. Brace, G.I. 1987. Boulder Monuments of Saskatchewan, M.A. dissertation. (Edmonton, Alberta: Department of Anthropology, University of Alberta).Google Scholar
  6. Brennan, M. 1983. The Stars and the Stones: Ancient Art and Astronomy in Ireland. (London: Thames & Hudson). (Reviews by R. Hicks: Archaeoastronomy 6, 141–144, 1983; G. Daniel: Nature 306, 516). Largely reproduced with some additional material as The Stones of Time. Inner Traditions International, 1994.Google Scholar
  7. Brinckerhoff, R.E. 1976. “Astronomically-Oriented Markings on Stonehenge,” Nature 263, 465–469.ADSCrossRefGoogle Scholar
  8. Broadbent, S.R. 1955. “Quantum Hypotheses,” Biometrika 42, 45–57.MATHMathSciNetGoogle Scholar
  9. Brumley, D.J. 1986. Medicine Wheels on the Northern Plains: A Summary Appraisal. (Calgary, AB: Ethos Consultants, Ltd., for the Archaeological Survey of Aberta). Reprinted 1996.Google Scholar
  10. Burkert, W. 1972. Lore and Science in Ancient Pythagoreanism tr. E.L. Minan, Jr. (Cambridge, MA: Harvard University Press).Google Scholar
  11. Burl, A. 1976/1989. The Stone Circles of the British Isles. (New Haven: Yale University Press). 410 pp. (Reviews by R.L. Merritt: American Scientist 65, 376, 1977; S.L. Gibbs: Archaeoastronomy 2(3), 19). (7th printing, 1989).Google Scholar
  12. Burl, A. 1979/1986. Rings of Stone: The Prehistoric Stone Circles of Britain and Ireland. (New Haven: Ticknor and Fields). 2nd ed. (New Haven: Yale Univ. Press). (Reviews of 1st ed. by R.J.C. Atkinson: Nature 282, 175–176, 1979; Nature 284, 700, 1980; G.S. Hawkins: Archaeoastronomy 2(4), 25–27, 1979; G.E. Hutchinson: American Scientist 67, 728, 1979; R. Hicks: Archaeology 33(4), 68, 1980; S. Milisauskas: American Anthropologist 82, 882, 1980).Google Scholar
  13. Burl, A. 1985. “Stone Circles: The Welsh Problem,” Council for British Archaeology Report 35, 72–82.Google Scholar
  14. Burl, A. 1987. The Stonehenge People. (London: J.M. Dent & Sons). 249 pp. (Review by A. Whittle in Journal for the History of Astronomy 19, Archaeoastronomy (12), S85–88. 1988.).Google Scholar
  15. Burl, A. 1993. From Carnac to Callanish: The Prehistoric Stone Rows and Avenues of Britain, Ireland, and Brittany. (New Haven: Yale University Press).Google Scholar
  16. Campbell, J. 1988b. Historical Atlas of World Mythology. I. The Way of the Animal Powers. Part 2. Mythologies of the Great Hunt. (New York: Perennial Library, Harper & Row).Google Scholar
  17. Clagett, M. 1995. Ancient Egyptian Science II: Calendars, Clocks, and Astronomy. (Philadelphia: The American Philosophical Society).MATHGoogle Scholar
  18. Coles, J. 1973. Archaeology by Experiment. (New York: Scribners).Google Scholar
  19. d’Errico, F. 1989. “Palaeolithic Calendars—A Case of Wishful Thinking?” Current Anthropolgy 30(1), 117–118.CrossRefGoogle Scholar
  20. d’Errico, F. 1992. “Reply to A. Marshack,” Rock Art Research 8(3), 122–130.Google Scholar
  21. Dibble, W.E. 1976. “A Possible Pythagorean Triangle at Stonehenge,” Journal for the History of Astronomy 7, 141.MathSciNetADSGoogle Scholar
  22. Eddy, J.A. 1974. “Astronomical Alignment of the Bighorn Medicine Wheel,” Science 184, 1035–1043.ADSCrossRefGoogle Scholar
  23. Eogan, G. 1986. Knowth and the Passage Tombs of Ireland. (London: Thames & Hudson). 247 pp.Google Scholar
  24. Frazer, J.G. 1912. The Golden Bough 12 vol. ed. (1911–1915) (London: MacMillan). Older editions: 2 vols., 1890; 3 vols., 1900.Google Scholar
  25. Graves, R. 1955–1957. The Greek Myths. (New York: Penguin).Google Scholar
  26. Hadingham, E. 1976. Circles and Standing Stones. (Garden City: Doubleday).Google Scholar
  27. Hadingham, E. 1979. Secrets of the Ice Age. (Walker: New York).Google Scholar
  28. Harbison, P. 1970/1992. Guide to National and Historical Monuments of Ireland. (Dublin: Gill & Macmillan). 384 pp.Google Scholar
  29. Hawkins, G.S. 1965b. “Callanish, A Scottish Stonehenge,” Science 147, 127.ADSCrossRefGoogle Scholar
  30. Hawkins, G.S. 1966. Astro-Archaeology. (Cambridge: University Press).Google Scholar
  31. Heggie, D.C. 1981b. Megalithic Science: Ancient Mathematics and Astronomy in Northwest Europe. (London: Thames & Hudson). (Reviews by A.S. Thom: Archaeoastronomy 5(4), 24–27, 1982; G.S. Hawkins: The Sciences 23(4), 56).Google Scholar
  32. Heilbron, J.L. 1999. The Sun in the Church: Cathedrals as Solar Observatories. (Cambridge, MA: Harvard University Press).Google Scholar
  33. Hoskin, M. 1985. “The Talayotic Culture of Menorca: A First Reconnaissance,” Journal for the History of Astronomy 16, Archaeoastronomy Suppl. (9), S133–S151.MathSciNetADSGoogle Scholar
  34. Hoskin M. 1991. “The Taulas of Menorca.” IInd Deya International Conference of Prehistory, Recent Developments in Western Mediterranean Prehistory: Archaeological Techniques, Technology, and Theory. Volume II. Archaeological Technology and Theory. Eds., W.H. Waldren, J.A. Ensenyat, and R.C. Kennard (BAR International Series 573) pp. 217–236. (Oxford: Tempus Reparatum).Google Scholar
  35. Hoskin, M., Allan, E., and Gralewski, R. 1993. “The Tombe di Gigante and Temples of Noraghic Sardinia,” Journal of the History of Astronomy 24, Archaeoastronomy Suppl. (18), S1–S26.ADSGoogle Scholar
  36. Hoskin, M., Allan, E., and Gralewski, R. 1994a. “Studies in Iberian Archaeoastronomy: (1) Orientations of the Megalithic Sepulchres of Almeria, Granada, and Malaga,” Journal of the History of Astronomy 25, Archaeoastronomy Suppl. (19), S55–S82.ADSGoogle Scholar
  37. Hoskin, M., Allan, E., and Gralewski, R. 1994b. “Orientations of Corsican Dolmens,” Journal of the History of Astronomy 25(4), 313–316.ADSGoogle Scholar
  38. Hoskin, M., Allan, E., and Gralewski, R. 1995. “Studies in Iberian Archaeoastronomy: (2) Orientations of the Tholos Tombs of Almeria,” Journal of the History of Astronomy 26, Archaeoastronomy Suppl. (20), S29–S48.ADSGoogle Scholar
  39. Hoyle, F. 1977a. On Stonehenge. (New York: W.H. Freeman). 160 pp. (Review by G. Moir: Antiquity 53, 124, 1979; J.E. Wood: Archaeoastronomy 3(3), 37–38).Google Scholar
  40. Huffer, C., Trinklein, F., and Bunge, M. 1967. An Introduction to Astronomy. (New York: Holt, Rinehart, & Winston).Google Scholar
  41. Kehoe, A.B. 1981. “The Cultural Significance of the Moose Mountain Observatory,” Archaeoastronomy 4(1), 8.Google Scholar
  42. Kehoe, T.F., and Kehoe, A.B. 1977. “Stones, Solstices, and Sun Dance Structures,” Plains Anthropologist 22, 85.Google Scholar
  43. MacKie, E.W. 1974. “Archaeological Tests on Supposed Prehistoric Astronomical Sites in Scotland,” Transactions of the Royal Philosophical Society London. A. 276, 169–194.MathSciNetADSCrossRefGoogle Scholar
  44. MacKie, E.W. 1976. “The Glasgow Conference on Ceremonial and Science in Prehistoric Britain,” Antiquities 50, 136–138.Google Scholar
  45. MacKie, E.W. 1977a. Science and Society in Prehistoric Britain. (New York: St. Martin’s Press). (Reviews by D.C. Heggie: Journal for the History of Astronomy 9, 61, 1978; S. Piggott: Antiquity 52, 62, 1979; S.L. Gibbs: Archaeoastronomy 2(2), 21–22; S. Milisauskas: American Anthropologist 82, 882).Google Scholar
  46. Makemson, M.W. 1941. The Morning Star Rises: An Account of Polynesian Astronomy. (New Haven: Yale University Press). (Review by G. Wendt: Sky and Telescope 1(3), 20).Google Scholar
  47. Marshack, A. 1972b. The Roots of Civilization: The Cognitive Beginnings of Man’s First Art, Symbol and Notation. (New York: McGraw-Hill). 413 pp.Google Scholar
  48. Mavor, J.W., Jr. 1977. “The Riddle of Mazorah,” Almogaren VII, 1976, Yearbook of the Institutum Canarium and the Gesellschaft für interdisziplinäre Saharaforschung Hallein, Austria (Graz: Akademische Druck und Verlaganstalt), 89–121.Google Scholar
  49. Mohen, J.P. 1990. The World of Megaliths. (New York, NY: Facts on File).Google Scholar
  50. Müller, R. 1970. Der Himmel über dem Menschen der Steinzeit, Astronomie und Mathematik in den Bauten der Megalithkultur. (Berlin, Heidelberg, New York: Springer Verlag).Google Scholar
  51. Newall, R.S. 1953/1959/1981. Stonehenge, Wiltshire. Department of the Environment Official Handbook. (London: Her Majesty’s Stationery Office).Google Scholar
  52. Newham, C.A. 1972. The Astronomical Significance of Stonehenge. (Gwent, Wales: Moon Publishers).Google Scholar
  53. O’Kelly, M.J. 1982. Newgrange: Archaeology, Art and Legend. (London: Thames & Hudson). 240 pp.Google Scholar
  54. O’Kelly, M.J. 1989. Early Ireland: An Introduction to Irish Prehistory. (Cambridge: Cambridge University Press.)Google Scholar
  55. Ovenden, M.W., and Rodger, D. 1978. “Megaliths and Medicine Wheels,” in M. Wilson, et al., eds. 1978: Megaliths to Medicine Wheels: Boulder Structures in Archaeology (Proceedings, Eleventh Chacmool Conference, University of Calgary, Calgary, Alberta), 371–386.Google Scholar
  56. Parpola, A. 1994. Deciphering the Indus Script. (Cambridge: University Press).Google Scholar
  57. Patrick, J.D. 1974a. “Investigation into the Astronomical and Geometric Characteristics of the Passage-Grave Cemeteries at the Boyne Valley, Carrowkeel and Loughcrew.” M.Sc. Thesis. (Dublin: University of Dublin).Google Scholar
  58. Ponting, G.H., and Ponting, M.R. 1984a. New Light on the Stones of Callanish. (Stornoway: Essprint, Ltd.).Google Scholar
  59. Prendergast, F.T. 1991a. An Investigation of the Great Standing Stones at Newgrange for Solar Calendar Function. M.Sc. Thesis (Dublin: University of Dublin).Google Scholar
  60. Rappenglück, M. 1997. “The Pleiades in the ‘Salle des Taureaux’, Grotte des Lascaux (France). Does a Rock Picture in the Cave of Lascaux Show the Open Star Cluster of the Pleiades at the Magdalénien Era, ca. 15,300 b.c.?” in Actas del IV Congreso de la SEAC, Proceedings of the IVth SEAC Meeting, ‘Astronomy and Culture’, C. Jaschek and F. Atrio Barandela, eds. (Salamanca: Universidad de Salamanca), 217–225.Google Scholar
  61. Rappenglück, M. 1999. “Sky Luminaries in the Space Orienting Activity of Homo Sapiens in the Middle Palaeolithic,” Astronomical and Astrophysical Transactions 17, 459–473.CrossRefGoogle Scholar
  62. Rappenglück, M. 2000. “Ice Age People Find their Ways by the Stars: A Rock Picture in the Cueva de El Castillo (Spain) May Represent the Circumpolar Constellation of the Northern Crown (CrB),” Migration and Diffusion 1(2), 25–18.Google Scholar
  63. Rea, T. 1988. “The Winter Solstice Phenomenon at Newgrange: Accident or Design?” Nature 337, 343–345.ADSGoogle Scholar
  64. Robinson, J.H. 1983. “The Solstice Eclipses of Stonehenge II,” Archaeoastronomy 6, 124–131.ADSGoogle Scholar
  65. Ruggles, C.L.N. 1981. “Prehistoric Astronomy: How Far Did It Go?” New Scientist 90, 750.ADSGoogle Scholar
  66. Ruggles, C.L.N. 1984a. “Megalithic Astronomy: The Last Five Years,” Vistas in Astronomy 27, 231–289.MathSciNetADSCrossRefGoogle Scholar
  67. Ruggles, C.L.N. 1985. “The Linear Settings of Argyll and Mull,” Journal for the History of Astronomy 16, Archaeoastronomy Suppl. (9), S105–S132.ADSGoogle Scholar
  68. Ruggles, C.L.N., ed. 1988a. Records in Stone: Papers in Memory of Alexander Thom. (Cambridge: University Press). 519 pp. (Review by S. McCluskey: Isis 81, 330–331, 1990.)Google Scholar
  69. Schaefer, B.E. 1986. “Atmospheric Extinction Effects on Stellar Alignments,” Journal for the History of Astronomy 17, Archaeoastronomy Suppl. (10), S32–S42.ADSGoogle Scholar
  70. Serio, G.F., Hoskin, M., and Ventura, F. 1992. “The Orientations of the Temples at Malta,” Journal for the History of Astronomy 23, 107–119.ADSGoogle Scholar
  71. Somerville, B. 1912b. “Astronomical Indications in the Megalithic Monuments at Callanish,” Journal of the British Astronomical Association 23, 23–37.Google Scholar
  72. Stooke, P.J. 1994. “Neolithic Lunar Maps at Knowth and Baltinglass, Ireland,” Journal for the History of Astronomy 25, 39–55.ADSGoogle Scholar
  73. Sweetman, P.D. 1984. “A Late Neolithic/Early Bronze Age Pit Circle at Newgrange, Co. Meath,” Proceedings, R. Ir. Academy 85C, 195–221.Google Scholar
  74. Thom, A. 1964. “The Larger Units of Length of Megalithic Man,” Journal of the Royal Statistical Society A127, 527–533.Google Scholar
  75. Thom, A. 1966. “Megalithic Astronomy: Indications in Standing Stones,” Vistas in Astronomy 7, 1–57.ADSCrossRefGoogle Scholar
  76. Thom, A. 1967. Megalithic Sites in Britain. (Oxford: Clarendon Press). Reprinted 1972, Oxford University Press.Google Scholar
  77. Thom, A. 1971/1978. Megalithic Lunar Observatories. (Oxford: University Press). (3rd printing, 1978). (Review by T.M. Cowan: Journal of the History of Astronomy 2, 202, 1971).Google Scholar
  78. Thom, A. 1984. “Moving and Erecting the Menhirs,” Proceedings, Prehistoric Society 50, 382–384.Google Scholar
  79. Thom, A., and Thom, A.S. 1971. “The Astronomical Significance of Large Carnac Menhirs,” Journal for the History of Astronomy 2, 147–160.MathSciNetADSGoogle Scholar
  80. Thom, A., and Thom, A.S. 1973. “A Megalithic Lunar Observatory in Orkney: The Ring of Brogar and Its Cairns,” Journal for the History of Astronomy 4, 111–123.MathSciNetADSGoogle Scholar
  81. Thom, A., and Thom, A.S. 1974. “The Kermario Alignments,” Journal for the History of Astronomy 5, 30­–47.ADSGoogle Scholar
  82. Thom, A., and Thom, A.S. 1978a. Megalithic Remains in Britain and Brittany. (Oxford: University Press). (Reviews by R.J.C. Atkinson: Journal for the History of Astronomy 10, Archaeoastronomy Suppl. (1), S99–S102; G.S. Hawkins: Archaeoastronomy 2(4), 27–28, 1979.)Google Scholar
  83. Thom, A., Thom, A.S., Merritt, R.L., and Merritt, L.M. 1973. “The Astronomical Significance of the Crucuno Stone Rectangle,” Current Anthropology 14, 450–454.CrossRefGoogle Scholar
  84. Tusa, S., Serio, G.F., and Hoskin, M. 1992. “Orientations of the Sesi of Pantelleria.” Archaeoastronomy-Supplement to the Journal for the History of Astronomy 17, S15.ADSGoogle Scholar
  85. Twohig, E.S. 1981. The Megalithic Art of Western Europe. (Oxford: Clarendon Press). (Review by G. Daniel: Antiquity 55, 234.)Google Scholar
  86. Ventura, F., Serio, G.F., and Hoskin, M. 1993. “Possible Tally Stones at Mnajdra, Malta,” Journal for the History of Astronomy 24, 171–183.ADSGoogle Scholar
  87. Vogt, D. 1993. “Medicine Wheel Astronomy,” in Astronomies in Cultures, eds., Ruggles and N.J. Saunders. (Niwot, Colorado: University of Colorado Press). 163–196.Google Scholar
  88. Wood, J.E. 1978. Sun, Moon, and Standing Stones. (Oxford: University Press). (Reviews by R.J.C. Atkinson: Antiquity 52, 251, 1978; E.W. MacKie: Nature 275, 75, 1978; J.A. Eddy: Journal for the History of Astronomy 11, Archaeoastronomy Suppl., (2), S95; L.V. Morrison: Observatory 100, 173, 1980).Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2011

Authors and Affiliations

  1. 1.Department of ArchaeologyThe University of CalgaryCalgaryCanada
  2. 2.Department of Physics and AstronomyThe University of CalgaryCalgaryCanada

Personalised recommendations