This link to Archaeoastronomy on Wikipedia gives background information.  In this section we are primarily considering the Thom/ Ruggles assessments and differences, and mainly those concerning possible lunar alignments. Mensuration (shapes of stone circles ) is not discussed.

In Megalithic Sites in Britain (1967), Thom presented evidence that particular declinations of the sun (calendrical)  and the moon had been observed in the Bronze Age.  In the majority of cases the direction indication was from a single structure (e.g. flat stone) or perhaps a short row.  However some of those apparently indicating a significant lunar declination also involved a distant foresight. Thom expressed surprise when a gaussian plot of the differences of declination from a significant lunar value gave a 'double hump' the peaks being separated by about 30' arc - the lunar diameter.   He took this to indicate that the observers had been observing both the upper and the lower limb of the moon.   This would suggest a precision of about ±0°.1.

In Megalithic Lunar Observatories (MLO) (1971), Thom assessed many sites on the basis of potentially precise alignments (i.e. a backsight and an indicated foresight).  In principle this combination is capable of giving 1' arc precision. In Table 7.1 p 76, Thom gave a list of his 40 best lines to date.  These he plotted against declination.  The resulting graph he believed showed that the small 9' arc 'wobble' had been observed during the Early Bronze Age (circa 1700BC).  In MLO Thom included much detailed but dense information on lunar movement, refraction, lunar parallax, extrapolation and other topics.  These important topics are  covered later in 'Discussion of Results'.

In the following years, Thom (usually now with his son A.S. Thom) published a number of papers on the topic of precise lunar alignments. (Thom &Thom 1978, Thom &Thom 1980a, A.S. Thom in Ruggles & Whittle 1981, Thom & Thom in Heggie 1982, Thom & Thom 1983).

The Oxford series of conferences began in 1981.  These were set up to discuss the general topic of archaeoastronomy which in 1981 meant in particular the work of Alexander Thom.  The proceedings of this conference were published ( Heggie, ed., 1982)

At this time while a few archaeologists supported the ideas of Thom (e.g. E.W. MacKie, R.J.C. Atkinson) and some were strongly opposed (e.g. Glynn Daniel, Jacquetta Hawkes), most were undecided. This is understandable as archaeology tends to lie more towards the humanities side than the scientific, or at least mathematical. Furthermore the topic does involve disciplines with which most archaeologists are not familiar. In addition if Thom was correct, then it would probably imply a different social structure to the one generally accepted and perhaps involving 'wise men' or 'astronomer priests' as proposed by Mackie (1977). This idea was not, in general, favourably received.

The criticism by Daniel and Hawkes was not based on rational analysis of Thoms' results but that by Clive Ruggles was. (Ruggles 1981, 1982, 1983, 1984). In the 1981 paper he introduced the idea of levels of potential precision of observation.  Level 1 uses the structure itself alone, Level 2 similar but with the possibility of stone rows, outliers from circles etc.  (Thom included some with distant horizon foresights but Ruggles (1999, p58) chose, for 'reasons of consistency', to discount the foresight and only consider the orientation of the stone / structure). Level 3 is the 'traditional' arrangement of an indicated distant foresight from a backsight.  Level 4 uses the same measurements as level 3 but then uses refinements to calculate the expected declination. (Ruggles 1981, pp153-154; Ruggles 1999, p49)  For details of the refinements used in Level 4 see Thom & Thom 1980a, Thom & Thom 1983.

The potential precision of declination is as follows:  Level 1, ±1º; Level 2 ±0.1º, Level 3  ±1'or 2' arc, Level 4 <±1' arc.  i.e. Level 1 is a Low Precision Orientation; Levels 3 & 4 are High Precision Alignments which would, in principle, be capable of identifying the lunar 'wobble' of 9' arc.

Ruggles published in two papers a detailed reassessment of Thoms' High Precision Megalithic Lunar Sightlines. (Ruggles 1982, 1983 ).  There is a summary of these findings in Ruggles 1999, pp58-63. Several aspects of Thom's findings were criticised :-

  • Fully half of the 40 lines given by Thom were not properly indicated
  • There were some errors of indentification
  • Several supposed foresights were not in fact visible from the backsight
  • There was evidence that Thom had not selected horizon features fairly.  i.e. without regard to the astronomical possibilities
  • There was little consistency in the types of backsights, indicators and foresights
  • The sightlines were chosen from a large number of sightlines with a wide geographical spread.

In addition there was cited a number of issues that would be expected to make the practical setting up of precise alignments and hence observation of the lunar 'wobble', all but impossible:-

  • Short sightlines would mean that vegetation at the foresight would seriously compromise the accuracy
  • Long sightlines would potentially have serious refraction problems
  • The moon could only be observed when rising or setting (in the dark is assumed). i.e. once per day and in this time the declination could, near maximum declination, change by as much as 10' arc
  • Lunar Parallax would cause (apparently) random changes in declination of up to ±3' arc.
  • Given that the moon can have any of 9 significant declinations within each of the 8 lunar bands, in hilly country it was claimed that chance alignments would be likely to be common.     (Ruggles 1999, pp59-63)
  • It would take about 80 - 100 years (half a parallax cycle) to set up each site   (Thom 1971, p82 ; A.S.Thom 1981, p47; Ruggles & Whittle  1981, p191 ;  Ruggles 1999, p63 )

[See 'Discussion of Results']

On the basis of the above, it was considered that the results as presented by Thom could quite easily be explained by chance alone. i.e. that there was insufficient evidence in the data produced by Thom to make a convincing case for the existence of intended precise lunar alignments. There was no suggestion that Thom had in any way intentionally falsified his data (See note below). Rather, it was a case of 'inadvertant bias'. ( Ruggles and Whittle 1981, p189 ;  Ruggles in Heggie (ed.) 1982, pp 90 & 92 ;  Ruggles 1999, pp63 & 65 ). And more recently in response to claims in a paper by Euan MacKie:- "Earlier publications [1981, 1982, 1983] by Ruggles show beyond any reasonable doubt that all Thom's putative astronomical sightlines at a precision greater than about half a degree can quite easily be adequately accounted for as chance alignments." (Ruggles and Barclay 2000, p 70 )

[Note:- While assessing the Thoms' work re high precision lines, Ruggles independently checked most of the lines.   "Resurveys of 38 of the 44 sightlines produced general agreement with the Thoms' declinations to within about 1' " (Ruggles 1983, p S32)]

In 1976 - 1979 and 1981, Ruggles, together with a variety of co-workers, undertook a substantial project to reassess a large number of declinations as indicated by stones in western Scotland.  This was published in 1984 as :-  Megalithic Astronomy   A new archaeological statistical study of 300 western Scottish sites British Archaeological Reports 123. This included details and data on all sites investigated. However it was not apparently done with a view to investigating the possibilities of precise alignments since :-

  • Although approximate horizon profiles were measured where possible, only the orientation of the stone(s) alone was considered when determining an azimuth with no reference to any specific skyline features, and
  • measurements were given to ±0°.1 ( ±6' ) which is insufficiently precise to determine whether or not precise alignments exist:-

In his conclusions, Ruggles says that:-

  • "...there is marginal evidence of the preference for six particular declination values to within a precision of one or two degrees."  ...   "Three of the declinations (-30°, +18° and +27° ) may indicate a specific interest in the edges of the lunar limiting bands (the lunar "standstills") and would imply that organised observations were undertaken over periods of at least 20 years."(Ruggles 1984, p 303)
  • And later:-   " ....there are no instances in Table 12.3 of low significance levels occurring 0°.5 apart, which would be comparable to Thom's double-peak evidence. (Thom 1967, Fig 10.1 ).   Thus we obtain no evidence that the separate limbs of the sun or moon were observed at any particular event." (Ruggles 1984, p 308 )

Ruggles goes on to say:-

  • "We have not examined our data for very high precision indications using distant horizon features such as notches, on the grounds that there is no motivation from our data at the lower precision to do so."   He goes on to say :-  "However in case anyone should wish to scrutinise our data for such evidence, all the information is available in this volume, in the form of background information, profile diagrams and tabulated data, to enable them to do so." Comment:-  Since the data  is given to ±0°.1 and the horizon profiles are little more than sketches, the information to carry out such a study is not apparently available.


Proved or disproved?

It is important to recognise that Ruggles has written on several occasions that his assessments do not disprove the hypothesis of the possible existence of precise lunar alignments; rather that he does not find that the results presented by Thom are sufficient to prove it.  However in his book 'Astronomy in Prehistoric Britain and Ireland' he goes further, and following a discussion of relevant matters, he writes:- "Taken together, these factors lead us to the unavoidable conclusion that lunar motions were not  in fact observed and recorded to high precision in prehistoric times.  For these reasons the idea of lunar observations and markers precise to a few minutes of arc will concern us no further in this book." (Ruggles 1999: p67).

The discussion above has been concerned with the moon.  The sun should also be discussed:-


The sun reaches an extreme north and south at the solstices.  There is evidence that at least some of the four extreme positions were marked approximately before the Bronze Age by passage grave entrances (New Grange, Knowth, Maes Howe, and the Clava Cairns and others) and perhaps at Stonehenge. Thom provided evidence that the positions had been marked accurately around 1750BC. In MLO chap.4 (Thom 1971) he gives four sites in Argyll where he believed this was done: Kintraw, Ballochroy, East Loch Tarbet (Jura) and Peninver (Kintyre); the first two being the most important. The Kintraw site involves an 'observing platform' above a steep ravine.  This was excavated by MacKie in 1970&'71 giving evidence that it was artificial (MacKie 1974, pp178-185; Mackie 1977, pp84-97;   Wood 1978. pp86-89; Mackie in Ruggles & Whittle 1981, pp115-116; Heggie 1981, pp187-189). Assuming that was correct, then the value for the obliquity of the ecliptic (23° 54′.2 ± 0′.7) found from these sites (Thom 1971, p44) is important as it provides the basis for the lunar declinations at the 'standstills' at that time.  Lunar alignments, if they exist, must fit with these expected values assuming that the moon was observed during the same period.

By 1967 Thom had amassed data on the declinations of many orientations/ alignments. When he plotted these as a gaussian graph various peaks were obtained. (Thom 1967, Fig 8.1). Many of these were for significant solar positions:-

  • Solstices
  • The equinoxes - since the earth is in a slightly elliptical orbit the halves of the year are unequal. The observed declination was close to +0°.5 which would be expected if the mid-point of the year by days was used.
  • A half way point between the above two by days (not angle)  (The 'quarter days')
  • Four further peaks splitting the above - again by days.

Thom interpreted these as evidence for a 16 month calender, each 'month' being of about 23 days. (Thom 1967, pp109-117).   The calculated dates for the 'quarter days' agreeing closely with the still well known Candlemass (2nd Feb.), Whitsunday (15 May), Lammas (1st Aug.), and Martinmass (11th Nov.).  (This is described in detail in the section on the Island of  Mull Introduction:-  1a) Solar Alignments in Mull and Islay: Introduction.)

MacKie has strongly supported the ideas of Thom, particularly with regard to solar orientations and alignments. He has published papers on a variety of related topics including:- Archaeological tests re astronomical sites,  Wise men/ Priesthoods and Calendrical matters.  (MacKie 1974, 1977, 1981, 1982, 1985, 1988, 1997, 2002, 2006, 2009)