Megalithic Aids to Navigation 1986 Ed 2011

104
 - 1 - MEGALITHIC AIDS TO  NAVIGATION 1986 (Rev2 2011) By P.B. Davidson © Copyright 1986-2009 Privately Circulated Scanned to OCR 2009

description

Research undertaken to investigate the relationship between neolithic stone circles and the trans-marine dissemination of stone tools.

Transcript of Megalithic Aids to Navigation 1986 Ed 2011

  • - 1 -

    MEGALITHIC AIDS

    TO

    NAVIGATION

    1986 (Rev2 2011)

    By P.B. Davidson

    Copyright 1986-2009

    Privately Circulated

    Scanned to OCR 2009

  • - 2 -

    MEGALITHIC AIDS TO NAVIGATION 1986 P B Davidson

    Chapter 1. Preamble

    What is this paper about?

    It is as well to be clear about that as the proposition is at best a complex one; the subjecthas, however, been beset for long enough by what can loosely be called Controversy.

    There is the body of people who claim too much; Alfred Watkins Ley lines will serveas an example. The evidence is not there, but a romantic fringe has grown around theidea to the point where archaeologists cringe at the suggestion that three stones are inline!

    There is a body of people that need to be taken very much more seriously; Sir NormanLockyer surveyed many stone circles and alignments in the early years of the centuryand attributed to them solar or stellar indications. Some of his observations are usefulbut are not on their own evidence enough; the questions have to be answered, did thesepeople really do this and if so why?

    Perhaps we never can answer that sort of question; perhaps if we attempt the answer weshall always find more questions being posed. That at least would be ordered progress.We do have to be satisfied that we are not looking at a series of coincidences thatoccurred by chance. If we can be satisfied in general about that, and I think we can, thenwe may begin to ask a series of questions about the whole body of evidence and slowlyextract from it detailed propositions. Of course our answers will be in terms of theprobability of the propositions not occurring by chance, but as we see that a systemexists that follows rules of logic, or the natural laws of science, then we begin to expectthat the system is deterministic and not random.

    You will find in a supporting paper a description of the Jigsaw puzzle. It is written as anallegory and I hope you will see the way in which the most information may be takenfrom a subject; and how that process of learning can be thwarted.

    So what is this book about? It is primarily a serious exposition of a mass of data aboutthe relics of the inhabitants of Western Europe four thousand years ago, but I hope thetechnique of analysis will be appreciated in its own right as an exposition of theeveryday use of the scientific method.

    2 (note:- numbers in text indicate original pagination in scanned version. dpd) I trustthat people will be able to read the book at the level that suits their interest. Most peoplewill know about jigsaws and if they pitch their comprehension of the acceptance of aproposition at that level it gives a sound basis. Let me repeat here the levels ofacceptance that I think we should use; they are always subjective but we have to startsomewhere and we do then have a uniform measure to work with:

    We are saying that we can use the number of events of coincidences as a measure ofsupport for the proposition; we use the function Support (S) as 0.7 x number of co-incidences.

  • - 3 -

    for three coincidences S = 2 the proposition is worth considering further, ninecoincidences S = 6 the proposition well supported thirteen coincidences S = 9 theproposition strongly supported but provided always that the proposition never fails.We shall work this out as we go along and consider in detail what a particularproposition should be and whether we can combine propositions; that one proposition isconditional on another, or that some larger proposition includes the smaller.

    You will understand that the subject is complex! But if all the pieces of the jigsaw fallinto place we shall have the satisfaction of being pretty certain that that is what wasintended; you do have to read through to the end, any particular item will not stand onits own.

    However, it is only fair to say what I think the picture on the box of jigsaw is. It is this.

    In the two thousand years or so before 1500 BC the inhabitants of Western Europedeveloped the techniques of moving, and to some extent working, very heavy monolithsand placing them accurately for a variety of purposes. They also developed thetechnique of working hard volcanic stones into polished tools of a variety of types anduses. To facilitate the finding and dissemination of these tools they developed routes fortravelling through a lightly populated land; some routes are overland, but many areacross substantial stretches of open sea. To do so they needed to identify routes betweensandy beaches (for landing open boats) and to avoid areas of tidal turbulence. Theirnavigation technique, that of memorising the sequence of stars that set behind a stonealignment. Over a period of many centuries they developed a technique of predicting thetides by observing the moon; and by so doing improved their navigation by sailing atthe shortest neap tides.

    We can never prove that but I believe you will find this analysis compelling.

  • - 4 -

    3 Chapter 2. Rude Stone Monuments

    The scope of the problem

    Spread out through Western Europe, but particularly in Brittany and the Western partsof Great Britain, lie the pieces of a gigantic real life jigsaw puzzle; and I propose toapply the techniques of solving a jigsaw puzzle to it.

    I have used a 19th century term, Rude stone monuments, to describe the pieces of thepuzzle; it covers all types of structure and makes no comment about date, or the peoplethat erected them. In practice we are speaking of a collection of Graves of variousshape, of dolmens, of menhirs, of stone circles and of alignments, credited to unknownpeople or peoples that inhabited the area at least over the period 3500 BC to 1500 BC.

    The people are known by their grave goods, by their methods of burial, by the survivalof some durable artefacts such as stone axes. They were probably not literate; wecertainly have no written record of them. There are plenty of features of all sorts thatlink the rude stone monuments from Shetland to Spain; and plenty of features that arepeculiar to place or period.

    It is, however, the presence throughout the system of very remarkable stone circles,menhirs and rude stone align-ments that has added an extra dimension of interest to theproblem; and provided some major interdisciplinary difficul-ties. Various archaeologistsand engineers have, over a long period, attributed to these rude stone alignments anastronomical function. To do so is to attribute to these primitive people a skill that isinconsistent with the state of development of other aspects of their lives; on the onehand they manifestly performed prodigious feats of engineer-ing; on the other hand theinterpretation of them is to accept unexpected skills in mensuration and an ability tostore astronomical data, probably over hundreds of years.

    We are particularly interested, of course, in the work of Professor Alexander Thom andof his son A.S. Thom. Their work will be frequently used in our analysis; but the bestup to date interpretation of their work and the problems it sets for archaeologists lies inDr D.C. Heggies book Megalithic Science. Thus far are the pieces of the jigsawdefined. I expect to take the matter much further by applying to the data the tests weapply to the fit of pieces of a jigsaw; to bring in some recent archaeological concepts;and to tie up the whole picture.

    But first we must start with some basic ideas.

    4 Rude stone alignments

    A rude stone alignment consists of two or more un-dressed stones, though sometimescleavage planes are used to provide a flat face in the direction of the alignment. Stonescome in a variety of sizes, from handleable boulders to stones 10 ft or more tall andweighing several tons.

    The impression immediately created is that a direction is being indicated; and if theobserver aligns with it he will be impressed by the precision with which his eye iscarried in a particular line. But we have to quantify what actually is observed and whatweight to put on it.

  • - 5 -

    The stones may be 2 or 3 ft wide and of irregular shape, yet we would identifyimmediately a particular and precise line. It seems that the observers binocular visionforms an impression of the two outer sides of the line and combines these with anestimate of the line of the tops. Certainly a surveyor using theodolite or plane table doesnot feel any ambiguity in stating a line to 0.1.

    If, however, we now look forward to the horizon, we may see features there that weconsider notable and that could be attributed to the foresight of the alignment. We shallstill be able to observe the point on the horizon accurate to 0.1 indicated by thealignment; but there will be a wider angle within which a notable object on the horizonmay be said to be indicated. What is that angle?

    Let us consider three rude stones in alignment, the third stone being covered by theprofile of the first two; the stones of width c are distance d apart.

    The observer wishing to accept that the alignment of two stones might indicate a featureon the horizon might angle his observation until he observed from one side of Co to theopposite side of C,; he could do the same with the other pair of sides and these twodiagonals subtend an angle A,.

    Now A, (radians) = C + Cod

    5 But that is as wide an angle as would make it credible that the alignment indicated theangle. If a third stone is present then the full angle of the first pair would not provide acredible indication; we would have to accept the angle subtended by C2

    and A2 (radians) = C2-C0 and so ond1+d2

    This is very similar to the proposition we made concerning the number of dimensionsthat had to be right when one piece of jigsaw fitted another. With an alignment each

  • - 6 -

    stone we had to extend the above configuration we have one more dimension to add toour Support for the hypothesis that this alignment indicates a particular direction.

    Parc y Merw

    Let us take an example from South Wales. Parc y Merw in Cardigan MR. SM 999.359.It Thoms classification and the data is given p.73 and Thom 1967 p.157.

    6 The alignment consists of four large menhirs so that we have three dimensionsdetermining the alignment giving a Support of S = 2 on its own that is not very strong.

    The stones are large; let us assume a width of 3 ft for each. Then the value of A(maximum angle subtended) is

    A radian = 3+3 = = 1.15 degree300

    So we may say the alignment indicates a direction of 301.4 0.6

    If we look now at the profile of Mt Leinster some 91 miles away, and in our climatenear the limit of vision, we see that this alignment could be thought to point at MtLeinster or to something rather to the north of that. That indication is of one of thelimiting declinations for the setting of the moon; indeed 301.4 - 0.6 describes it quiteprecisely.

    Does that happen by chance? Does the alignment say look there is Mt Leinster! ordoes it say sail in this direction for a good landfall in Ireland; or does it say one dayyou will again see the moon set on this line? It might say all three, but on this dataalone we cannot resolve it. We can say that, with a range of 0.6, the probability ofthis occurring by chance is 1.2/360 = 1/300; and that this line indicates 301.4 has asupport value of Ln300 or S = 5.7.

    If the indication is of Mt Leinster then the range is wider than we had expected; 1.7and the support is reduced to S = Ln 3.4/300 = Ln 88 S 4.5 and both these assume thatwe cannot use the alignment in the reverse direction.

  • - 7 -

    7 Stone Circles

    Rude stone monuments in the shape of circles occur, as do alignments, in stones ofvarious undressed sizes; moreover, many are not true circles. Nonetheless the observerimmediately forms a circular impression of them; his visual imagination allows for theirregularity of the stones, but a proper survey is needed to observe the variations inshape arid accurately to measure them.

    We suppose, not necessarily correctly, that these circles will have been set out withstakes to obtain the desired shape; the stones then being set in holes prepared for them.The setting out may have been by measured rods or by some form of rope.

    Substantial precision of shape and of dimensions have been claimed for the shapes sosurveyed; and such precision calls into question the techniques of survey used.

    Thom produces evidence for a Megalithic Yard based initially on the diameter of circlesbut extended later from other very precise surveys; the MY is 2.72 ft or 0.83 m and heclaims a precision difficult to attain and sustain over a wide area and a long period oftime.

    The testing of these claims has led to considerable innovation by statisticians; theoutcome of the controversy is of some importance to the whole puzzle.

    Thom also proposes a set of shapes based on simple setting out geometry. The circlefrequently occurs, of course; there are a number of ellipses which may be presumed tobe set out by a rope attached to stakes at the two foci and with a length~(a +

    The other types of circle proposed are shown in the attached print from Heggie pp.30-31. Methods suggested for setting out by Cowan, using a rope and four stakes, arecomplemented by the suggestion of Angel using 3 or 4 stakes and a rope loop.

    Three stakes will define completely the circumscribing circle or ellipse, but the flattenedcircle will require more; 3 or 4 stakes and the length of the rope. Or, as in a jigsawpiece, 6 dimensions for conic sections, 9 for flattened circles.

    P8 Fig Shapes of Circles

  • - 8 -

    9 Ilkley Moor

    In seeking an example to demonstrate the problems of surveying circles and attributingshapes to them, I was fortunate in recalling a survey I carried out in 1970 but neverpublished; doubly fortunate in the way the field record produced an apt design.

    The circle in question is on Ilkley moor and is described by Raistrick in YAJ 29 (1929)p.357 & Fig.4 as the Twelve Apostles; the diameter is given as 52 ft and thedisposition of the stones is similar to that which we observed, though the north point isnot.

    I visited the site in October 1970 with an engineering colleague; both of us experiencedin setting out engineering structures. We found nine of the twelve stones thrown downand three standing but the stone sockets clean and unspoilt. The heavy heather growthnoted by Raistrick was no longer a problem.

    We carried out our survey using a light plane table and Alidade and a glass fibre tape.The plane table was set up without measurement at what we estimated was the circlecentre (the exact position was not important but we wanted distances to be of the sameorder of size). Noel Edwards made the measurements on the basis of an intelligentestimate of where the centre of the stone socket might have been. The survey wasoriented by reference points on the 1 O.S. map. We then sketched the shape of stonesand sockets and the way they were set down; the profile of the three upright stones wasalso sketched. Beyond noting that it was a circle roughly elliptical with axes of 16 mand 13 m, I did not do any more work on the survey; a few years later I found thecircle had been re-erected and I do not know what shape it now presents.

    Our rapidly-prepared survey (albeit unbiased and using a sound basis for estimation)now turns out to be of some value in. presenting the problem of surveying circles. Threestones were standing; enough to establish a unique circle or ellipse. Most of the stoneswere thrown down; permitting the survey of the holes as set out. The result of thesurvey follows; and the fit of the ellipse to all twelve points is not only pleasing butremarkable. What inference should we draw from the survey?

  • - 9 -

  • - 10 -

    11 The Shape of Circles

    I would first of all like to comment on an aspect of stone circles that has surprised me;that on entering the circle one takes up with some confidence a central position that laterturns out to be the correct centre from which to observe the alignment of outliers. Inpreparing the drawing of the Ilkley circle you will find some confirmation of thisnotion; the survey centre, chosen without measurement, is within 20 cm of the Ellipsecentre. I suspect the eye is automatically judging the midpoint of pairs of stones at theends of diameters.

    On a first inspection of the survey the shape appears to be a flattened circle, but onemust assume that the original shape included the three upright stones; the measuredlocations of the remainder are more or less suspect. We should try the fit of a circle andof an ellipse; but how are we to measure a good fit?

    We may draw a circle through three points; and that is particularly appropriate in thiscase where we have only three upright stones. Then we can measure the deviation of theother stones from the circle; there will be nine measurements on one side or the other ofthe circle. We can treat these deviations as arising from the inaccuracies ofmeasurement or disturbance of stones; we would measure the range of the deviations.We shall find a diameter of 14.9m and a range of 1m.

    We have also drawn an ellipse through the three upright stones. The dimensions of thesurvey suggested a major axis of ca. l6.5m and a minor axis of ca. 13.5m, and we chosefor a first trial (because later we wish to explore the hypothesis that a particulargeometry was used in the setting out) an ellipse of major axis 20 MY (l6.6m) and minoraxis 16 MY (l3.28m). By using an overlay and moving it about to cover the threeupright stones and a good fit for the rest, we arrived at the arrangement shown on thesketch. For this the range of deviations was + O.25m.

    We can do a bit better in determining the fit of a survey to a curve by adapting astatistical process used with great benefit in Quality Control systems, known asCumulative Sum Techniques (Cusum for short). For each measurement in the surveythe difference from the expected value is calculated and a cumulative sum is developed.

    In a quality control system acceptable range of values are developed based on the risksinvolved and expressed as the number of standard deviations from the mean; that neednot primarily concern us here. What is useful is that, in accumulating the differencesaround a closed curve, we can specify that they should add to zero for the completecurve as a measure of a good fit. The range of value for the differences will then bequite adequate for most con-siderations.

    For example, in the case of the ellipse, a = 16.6m, b = 13.28m for our survey:-

  • - 11 -

    The range is 0.25 but by adjusting the difference by

    Cusum = -0.60 = -0.0512 12

    the amended range remains 0.25 but the amended axes area = 16.7mb = 13.38m

    Which brings us back to the fit of pieces in a jigsaw puzzle. The 12 survey points fit theexpected values of this ellipse to the stated tolerance. We cannot use all the 12dimensions as we have used three to specify the ellipse. However, we can say that theproposition is supported by S = 9 Ln 2 = 6.3; the proposition is well supported.

    Alignments and Circles

    In discussing Parc y Merw and Ilkley Moor, we are presenting one example of each ofthe two main components of Rude stone monuments and we must consider what furtherinformation there may be in the study of them and what criteria we shall adopt foranalysing them.

    Most of the survey work has already been done by Prof. Thom and his son; and hisinterpretation cogent and competent. It is none of our business to repeat their work, norto upset their conclusions. The work of Dr Heggie is to present not only this work but tointerpret it for all those interested in prehistory; and, in particular, to evaluate thestatistical requirements and the constraints which statistical consideration place on thedata.

    Dr Heggie states the position succinctly We will not be content if the hypothesismerely fits the facts. On the contrary we will demand that there be something about themegalithic sites which we would find very surprising were the hypothesis false. Ofcourse, one way or another that 13 has to be so; if we are suggesting anything differentit is that the hypotheses we test will not only be compared with the chance situation butwith the choice between hypotheses and with the conditioning of one hypothesis onanother.

    Any criticism of the statisticians must be Professor Thoms, from Thom & Thom 1978p.75 on the Le Menec alignments. Some statisticians have criticised our treatment ofLe Menec alignments and have claimed that using the align-ments alone it is notpossible to show conclusively the use of the Megalithic Rod in its construction. Perhapsif they had started by visiting and studying the site they would never have attempted ananalysis of this kind which ignores the evidence coming from Le Menec cromlechs,from Kermario (chap. 7) and from the cromlechs at Kerlescan (chap. 8). To combine theevidence and draw a conclusion needs a philosophical approach; even the mostadvanced statistical theories do not help. Twenty years ago statisticians were muchmore ready to listen to a reasoned argument without resorting to elaborate mathematicaltheories. Had we in the first place waited for statistical theories to develop we shouldnever have been able to proceed.

    This insistence of the man of science to be free to choose his own hypotheses and tomake his own assessment of them is also expressed by A.W.F. Edwards, p34 speakingof hypotheses concerning heat and temperature:

  • - 12 -

    ... though by the likelihood axiom the support will inform us fully of the contributionto our judgement that the data may make, we shall also be influenced by the simplicityof the hypotheses, by their relevance to other situations, and by a multitude of subtleconsiderations that defy explicit statement. The scientist must be the judge of his ownhypotheses, not the statistician. The perpetual Sniping which statisticians suffer at thehands of practising scientists is largely due to their collective arrogance in presuming todirect the scientist in his consideration of hypotheses; the best contribution they canmake is to provide some measure of support, and the failure of all but a few to admitthe weaknesses of the conventional approaches has not improved the scientistsOpinion.

    Thus fortified, let us with caution proceed!

    14 References

    for information on lunar alignments and for Parc y Merw

    see Megalithic Sites in Britain by A. Thom, Clarendon Press, 1967,and Megalithic Lunar Observatories by A. Thom, Clarendon Press, 1971.

    for the refinement of their techniques

    see Megalithic Remains in Britain and Brittany by A. Thom and A.S. Thom,Clarendon Press, 1978.

    for information on the problems of Megalithic Science

    see Megalithic Science by Dr D.C. Heggie, Thomas & Hudson, 1981.

    for information on Ilkley moor

    see The Bronze Age in West Yorkshire by Dr A. Raistrick in YorkshireArchaeological Journal Vol.29, 1929.

    for information on Cusum techniques

    see Cumulative Sum Techniques, R.H. Woodward & P.L. Goldsmith. I.C.I.Monograph No.3, Oliver & Boyd 1964,

    and The design and analysis of Industrial Experiments, O.L. Davies, ed. Oliver &Boyd 1954. Chap.3. Sequential Test of Significance.

  • - 13 -

    15 Chapter 3. The Stone Axe Trade

    Introduction

    We do not know very much about the Social Organisation of the late Neolithic and earlybronze age; there is a presumption that agriculture is developing, but the balancebetween crops and herding is not easily defined; there is a developing technology, inpottery, in stone axes, and, at the end of the period, in bronze. What we have to go onare funerary monuments often with bones and grave goods, a collection of rude stonemonuments that are not funerary, and a widespread litter of stone axes many of themwithout provenance. Evidence for ships comes rather later, but we shall find we mustpresume them. The interest in axes creates a substantial presumption for forest clearanceand timber working. We should probably presume a population with ample resources;there is, in particular, no evidence for warfare.

    Site Catchment Areas

    Faced with a limited record of prehistoric people, techniques have been developingwhich enables some inference to be made about them. These studies are essential forPalaeolithic, and other hunter gatherer societies; they can equally well be adopted forthe low density areas of the late Neolithic.

    If we encounter a site, or sites adjacent, then we may say that as and when this site wascreated some group had to live here: at that time they will have had to hunt and forageand so on to live; they will have created a Site Catchment Area in which, in a non-competitive society, the sites and artefacts at that time will have been related.

    Such a site catchment area will be defined by the distance from the site that it isnaturally convenient to go and to still be able to return within the working day. We mayhave to be rather arbitrary in defining this; would one hour radius or two be moreappropriate? We have taken for this work a two hour radius or 10 km. However thereare also geographic features that will constrain this radius; a river, a coastline, a marshor ever rising ground for which pre-historians have been allowing 5 minutes for a 50m.contour in fell country or 10 minutes for a 50m. contour in rough country.

    When we have defined such a Site Catchment area we may consider that, if it provideda living at one moment, that would have been true over a period and that we should be16 able to consider the assemblage of sites and artefacts as representing the record ofthe use of the area for some period.

    Where we find fairly open country as in the Wolds of Yorkshire and a high density bothof funerary monuments (barrows) and artefacts (stone axes) we will have difficulty inidentifying with any confidence a particular site catchment area. We must start with amore isolated example.

    The area we have chosen as a prototype we have called Hutton Moor (MR. SE 35.75); itlies to the north of Boroughbridge on the limestone ridge that runs from the Tees to theDon; it is bounded essentially by Ure and Swale.

    On the ridge at the centre of the area is an assembly of some eleven tumuli and twohenges; from here to the natural boundary of the river is 5 to 6 km. To the north the

  • - 14 -

    ridge runs unimpeded, and with clear views to Cleveland and Pennine Hills; to the souththe ridge is for some ten kilo- metres interrupted by a terminal moraine; sandy, irregularcountry with marshes (or carrs) and which would have been wooded as the easternextremity of the Forest of Knaresborough. The, river crossing to the south is marked bythe Devils Arrows, the three remaining megaliths of four, (Thoms reference L 6/1);these stones do not have an astronomical significance but they do indicate the dry-shodroute to the south, avoiding the large Carr at Marton. That may not have been itspurpose; we are setting out a proposition and if we have a satisfactory prototype weshall find other examples in support.

    The area is a crossroads in the axe trade (or traffic) from north to south and particularlyfrom Cumberland to the settlements of Yorkshire Wolds, Howardian Hills, andCleveland hills. The routes are described by Manby in CW2, 1964, and in the moreparticular vicinity of Hutton Moor by Bradley in YAJ 1974.

    By crossing the river at Asenby a dry shod route can be followed to the HowardianHills. The main routes from Cumberland are by Swaledale (reaching the area at theThornborough rings), and by Craven (reaching the area at the henge at Nunwick). Wedo not know what Henges (or stone rings) were for, but here we make an associationwith the trade routes and part of our proposition is that they are an essential part of theassembly or marketing arrangements; but of course we do not know if they traded inthe modern sense.

    17 Fig

  • - 15 -

    18 Langdale Axes

    Now that we have established the specification for a prototype Site Catchment Area wecan analyse other regions in these terms. Let us start with Cumbria, from which manystone axes come. These axes are Group VI in the implement petrology classification ofCBA. They derive, from outcrops to be found on Scafell and the Langdale Pikes; that isto say from the high centre of the Lake District. Group VI axes are widely distributedthroughout the British Isles but their biggest concentration is centred on Humberside.

    Manby in CW2 (1964) identifies the location not only of Group VI axes in Cumbria butthe location of part-finished axes, or roughouts, and the axe factories in which theywere worked. The roughouts are found along the routes that lead to Yorkshire andHumberside and provide an indication not only of the routes, but the suggestion that theaxes were being polished as the purveyors moved; suggesting perhaps the operation of atranshumance route. There are five axe factories on Scafell and the Langdales forroughing out axes, but the finishing work seems to have been done lower down. Thesesites are at Portinscale near Keswick, at Gosforth, Edenside and Drigg in Eskdale, atMillom, and at Stainton-in-Furness. Concentrations of roughout axes also suggest axefactories in the vicinity of Penrith, of Carlisle and of Silloth on the coast.

    If we look at the Site catchment Areas of these axe factories, we see that they eachcontain a stone circle (or two in some cases) and that they are each able easily to exploita particular part of the outcrop on Scafell and the Langdales. While there are no circlesassociated with Silloth or Carlisle, there are two in the vicinity of Penrith and, althoughsomewhat farther removed from the Langdales than the other areas, we may consider itas one of the group.

    So we have five Site Catchment Areas in a ring around Cumbria, each with at least oneaxe factory and one or more stone circles. These circles are all identified in Thom 1967;some of them have alignments and we should see what information we may obtain fromthem:- Castle Rigg (L1/1) we will for the moment classify as an observatory for sun andmoon and deal with it later. It has, however, one alignment to an outlier that does not fitthe observatory; it does suggest a dry shod route to the Penrith area.

    Long Meg (L1/7) on the River Eden is a large circle with a large outlier (Long Meg)and a small subsidiary circle. Thom suggests a solar alignment for each of these, but weshould note that the alignment to Long Meg is also 19 for Scafell and provides a dry-shod route to the axe factories. Perhaps it is for both, but we shall return to that subject.

    Burnmoor (L1/6) is a collection of five circles, two of which are some distance from theother three. Thom suggests a variety of solar, lunar and stellar alignments foralignments circle to circle. If, however, we confine ourselves to the alignments of thetwo circles adjacent to the principal circle, we have two bearings (azimuth) of 311.9and 292.3. These do seem to point in the general direction of the areas of Galloway,also much occupied by Late Neolithic people. Is it possible that they were used, as wehave suggested for dry shod land routes, for sea crossings? and, if so, how would it havebeen done? and what advantage would ft have provided?

    On its own, of course, this pair of alignments can demonstrate nothing, but let us makethe proposition and see whether the proposition is supported from other locations. Theproposition is that for sea crossings an inland alignment bearing, transferred to theadjacent shore will provide a viable sea route to another viable beach and SiteCatchment Area associated with the Axe trade. We shall come back to technique and

  • - 16 -

    utility but broadly we are suggesting they learned the succession of stars setting on theinland alignment and used that at sea.

    Sunkenkirk (L1/3) has an alignment with a bearing of 308.8 which, if treated in thesame way as those at Burnmoor, suggests a similar landfall in Galloway.

    Giants Graves (L1/11) on the sea shore near Millom is an alignment of three stones witha bearing of 210.8. We shall see that this also fits the seagoing proposition giving alandfall in Red Wharf Bay in Anglesey.

    By describing the region of Cumbria in this way, as a group of Site Catchment Areasexploiting the stone outcrops for axes, we have provided a basis for describing similarareas and extended our proposition to include a definition of the use of alignments forsea crossings.

    20 Fig

  • - 17 -

    21 Fig

  • - 18 -

    22 Galloway

    Three of our suggested sea crossings from Cumbria are towards Galloway; let ustherefore look at Site Catchment Areas there too.

    In Galloway there are no axe factories; there are circles and alignments; and, what wehave not so far encountered, Megalithic graves. We shall need to study the distributionof these in due course; for the moment we shall mark them on the maps.

    For our immediate purpose we can identify three site catchment areas based on thedistribution of circles, graves and alignments; we shall call them Cree, the Machars andBallantrae. We should probably also define one for the peninsula or the Western shoreof Luce bay; we will call it Logan.

    The Cree peninsula contains three stone circles and three megalithic graves but is alsothe part of the coast indicated by the more northerly alignments from Eskdale; thealignment fits completely our proposition. There, if we look at the alignments in theCree peninsula, we find indications of two dry-shod land routes and two for seacrossings.

    The circle at Cambret Moor (G4/12) has two alignments; at 254.3 (for which there isno obvious objective) and at 296.7 which indicates a dry shod route over the moor tothe alignment on the coast at Ballantrae. The circle at Cauldside (G4/14) has threealignments, of which for 78.2 there is no obvious objective; the alignment at 59.5,however, indicates a dry shod route to the East having its termination at the large circleat Dumfries, Twelve Apostles (G6/1). The third alignment at 156.8 I take to be a seacrossing to the coast of Clwydd near the point of Ayr. The third circle in Cree,Kirkmabreck (G4/13) has an alignment of 5.9 and I take that to be a sea crossing to theIsle of Man, as it marks quite precisely the Eastern side of the headland Point of Ayre.

    We shall discuss elsewhere the sea level in Neolithic times but it does seem to havebeen lower than it is today. This would have appreciably altered the shoreline in thisarea; we have shown these sea crossings starting and finishing at the 10 metre (5fathom) line.

    The Site Catchment Area of Ballantrae probably includes a couple of graves and thecircle at Laggangarn (G3/3); we will note this as an observatory site and return to itlater. On the coast, however, is the alignment Ballantrae (G1/4), which preciselyterminates the alignment from Cambret Moor (G4/12). The alignment itself is for 11.8,23 indicating a route along the coast to Ardrossen.

    Perhaps we should pause at this point to consider where this is leading us; theproposition that alignments may show a dry shod route or a sea crossing is leading us todefine quite a coherent route to the North and to extend the network in other directions.This is beginning to seem reasonable; the routes fit the Site Catchment areas and theTopography. In particular the indicated landfall at Ardrossan fits into a pattern ofPrehistoric trade routes into Scotland. Sir Lindsay Scott in PPS Vol.XVII (1951) Thecolonisation of Scotland in the second Millennium BC indicates the Black Cart Pass,running inland from this part of the coast as a natural route leading through variousGlens to the North and East.

    Which brings us back to Luce Bay and the Machars. The other alignment from Eskdaleindicates the area of Logan on Luce Bay, but the alignment on 308.8 indicates the

  • - 19 -

    Eastern shore of Luce Bay; the general vicinity of Monreith Bay. There is onealignment in the Machars and it is close to Monreith Bay; it is Drumtroddan G3/12 witha bearing of 223.

    The bearing of 223 from Monreith Bay takes us through seas clear of the tide races ofthe North Channel to the Mouth of the Boyne; than which there is no more importantarea of Neolithic settlement. The pattern is developing and gaining credibility; we muststate all the evidence before attempting to quantify our confidence in it.

    24 Fig

  • - 20 -

    25 Gwynedd

    Looking South from Cumbria, indeed following the suggested sea passage of 210.8from Millom, we come to the District of Gwynedd (otherwise Anglesey andCaernarven). The bearing of 210.8 takes us into the centre of Red Wharf Bay inAnglesey an area rich in Megalithic graves and standing stones. Anglesey probablyrepresents three Site Catchment Areas; Red Wharf Bay, Holyhead and Newborough. If,as seems probable, sea levels were some 30 ft lower than present day, Menai Strait maywell not have existed and the southern end, Newborough Warren, and the Northern end,Penmaenmawr, both substantial sandy beaches.

    Penmaenmawr is the important Site Catchment Area as it contains the Craig Llwyd axefactory (Group VII). Nearby is a stone circle, Penmaenmawr W2/1. This circle has twooutliers giving bearings of 18.6 and 240.9. From the beach at Penmaenmawr thebearing of 18.6 leads to the mouth of the Esk in Cumbria; the objective of the bearing240.9 is not so obvious, but from the beach at Newborough Warren would lead to theIrish Coast near Wexford. As an entry point to Ireland that is not very obvious; a ratherfeatureless coast not marked by prehistoric remains, but we shall consider it again in thesection on Dyfed.

    On the western edge of the Penmaenmawr area are a pair of henges; unusual in that atthe entrance to one is burned a stone axe and at the entrance to the other a beaker. Thisdedication to each of two important artefacts does suggest their purpose as tradingmarts. However, from Llandegai runs a prehistoric route, its course described byGresham and Irvine in Antiquity V01.XXXVII (1963). The route crosses the riverConway and the Dee, over the Berwyn mountains and eventually to the Severn, wherewe find a site catchment area centred on the Clun Forest in Shropshire and rich in flintaxe working sites.

    Along this prehistoric route and on the Berwyn Mountain, and where another prehistorictrack runs up from the coast at Ardudwy, will be found the stone circle Moel Ty Ucha(W5/1). Associated with it is a circle across the river Dee, Twyfos (W512). Moel TyUcha has an outlier, giving a bearing of 17.6 with a distant foresight of a prominenthill. The bearing indicates a nearby crossing of the Dee and leads to a standing stone onthe hill just West of the Penbedw circle in Clwyd. So this circle looks like an importantsignpost on an important axe route.

    Returning to the West we find an important Site Catchment Area at the southern end ofthe Lleyn peninsula; there are two axe factories and five graves, two of which are atopposite ends of Forth Neigwl. We shall see that an alignment in Dyfed leads across thebay to Red Wharf Bay.

  • - 21 -

    26 Fig

  • - 22 -

    27 Fig

  • - 23 -

    28 Dyfed

    Let us move South to the other Welsh peninsula; here we find a similar prehistoricregion; similar but, once again, with substantial peculiarities dictated in part at least bythe topography.

    South West of Carmarthen and adjacent to the Pendine Sands we find a self-containedSite Catchment Area with a similar distribution of graves and standing stones to thatwhich we have already observed. The same is true of the Gower peninsula; and in thiscase one of two beaches is the landfall for an indicated sea route from Cornwall. We canidentify other areas on the Western end of the peninsula; at the southern end of St.Brides Bay and West of Fishguard, but neither is central to our argument.

    The activity is in this region more concentrated than we have come across before; itconcerns the region North and South of Prescelly Mountain. On the north coast we haveone sheltered bay at Newport, to which we shall transfer the alignments relevant to seatravel. To the South of the mountain we have a difficulty; there are prehistoric remainsbut they have been much despoiled. The description of what was there in 1910 and wasthen known to have been there is described by Bushell Among the Prescelly Circlesin Arch. Cambrensis, 6th Series Vol.XI. We have put this information as best we can onthe grid system.

    In this same vicinity we also have reference to the location of the source of theStonehenge Blue stones. This is given by Thomas in Antiquaries Journal, Vol.3, 1923.His assessment .. the three main varieties of bluestone (spotted dolerite (prescellite),shyolite and volcanic ash) were matched exactly by the outcrops ... brought together byglacial action with a small area on the South East slopes near Cil-maen-Llwyd. We haveshown this as the regional axe factory on the map; the concentration of circles close toit is remarkable. We shall put further consideration of these circles aside until weconsider from all the regions the characteristics of what we have loosely calledobservatories.

    In the area south of Prescelly there remains a small circle with outliers at Gors Fawr,W9/2, and also described by Bushell. Perhaps the small size of the stones has savedthem from destruction. We give a survey of the outliers of this circle which we made in1971 and which we believe provides a key clue to the use we have been suggesting forthese alignments. Of the six bearings that we recorded, only one (242) has no obviousobjective, though its projection does reach the coast by a dry shod route in the vicinityof the Site catchment area at the Southern end of St. Brides Bay. One other (103.5)indicates a dry-shod 29 route to the Af on Cynin (and thence to the Site Catchment areaSW of Carmarthen) and it is also accurately marked at that point by a standing stone.

    The remaining four bearings (291.5, 307.5, 8 and 14) become interesting if taken assea bearings from the beach at Newport. The bearing 8 marks the beach on the Lleynpeninsula and the bearing 15 marks the beach in the vicinity of Pwthelli (it depends onwhere the shoreline was).

    Which leaves us with the bearings 291.5 and 307.5. You will recall that in the firstsection we took as an example the bearing from a site north west of Prescelly, Parc ymerw, W9/7, with a bearing of 301.4 indicating either Mt Leinster or the northernsetting of the Moon; now here we have a bearing on either side and we are suggestingthem as navigation routes. We shall need to look at the landfalls in Ireland again; for themoment we can say that 291.5 leads to the Wicklow shore at essentially the same place

  • - 24 -

    as the bearing 240.9 from Gwynedd. The bearing 307.5 leads to a part of the coast atthe northern end of the Wicklow mountains.

    So the region Dyfed and Newport Bay in particular fits the proposition we have beendeveloping.

    In addition to the regional map similar to the other regions, we have included details ofthe Circle at Gors Fawr and the outlook from each outlier. The sketch of Pentre Ifanwith its view over Newport Bay is suggestive of our association between these stonesand the traffic of the bay.

    30 Fig

  • - 25 -

    31 Fig

  • - 26 -

    32 Fig

  • - 27 -

    33 Fig

  • - 28 -

    34 Cornwall

    To the South again, Cornwall as a region shows much the same pattern of SiteCatchment areas with axe factories and sea routes. We are setting on one side forconsideration under observations the circles on Bodmin Moor, solely for the reasonthat they do stand apart in type and location. We must also note the isles of Scilly, richin graves of various types; too many for the amount of land now visible. Crawford, in apaper on Lyonesse in Antiquity, Vol.1 (1927), discusses the change of sea level andconcludes that it must have been some 30 ft. lower. We have shown a map to includethe 5 fathom contour.

    The peninsula of Penwith is densely populated with standing stones, circles, graves andwhat are there called Quoits (or Dolmen in France). One cannot, with certainty, separatesite catchment areas for this but we might say there is the southern half, running fromSennen to Penzance and including the source (probably submerged) of Group I stoneaxes; and the northern half, mostly on high ground leading in the East to St. Ives andincluding there the source of Group II stone axes. Group III stone axes at Marazionmight have been exploited by either.

    We shall see that one of the sea routes indicated from Brittany leads to Mounts Bay.But the interesting sea route from this area is one indicated by the stone circle NineMaidens, S1/11, on a bearing of 332.7. This bearing, taken from the only possiblebeach, at St. Ives, indicates a landfall in. Ireland at Tramore. This bearing follows all therules we have proposed so far, but it surprises by the length of the sea passage (250 km).

    Which leaves two areas to the East. The area around Camborne includes the axe factoryat Cam Brea, the source of Group XVI stone axes; it is a good representation of a self-contained collection of stones and graves. Then we come to the area inland and south ofPadstow; with a north-facing beach at Trevose. There is a collection of prehistoricgraves between the beach and Trevose Head; there are several coastal graves, but themain group are 10 km to the south; and that includes the important alignment NineMaidens, S1/9. This is a line of nine large stones on a bearing of 26.3, with the Easternend of a prominent hill precisely marked by the alignment. Now this bearing fromTrevose beach leads to the Port Eynon Point (very precisely) at the western end of PortEynon Bay in Cower. So that the headland, on arrival, must be very similar to thatshown by the alignment.

    We shall evaluate this alignment with the others; we have identified fourteen fromWigtown to St. Ives.

  • - 29 -

    35 Fig

  • - 30 -

    36 Fig

  • - 31 -

    37 Fig

  • - 32 -

    38 Brittany

    And so we come to Brittany. Which presents us with two problems; an embarras derichesse of menhirs, dolmen, allees-couvertes alignments and so on; and the lack of agood corpus. As elsewhere we have noted the main centres of population and avoidedanalysis in the terms of this study; so we have noted for the moment the observatoryof Carnac and the inland settlements, of Mur de Bretagne, and of Landes de Lanvin. Wenote also the source of three types of stone axe and show them on the map; Type Afrom Seledin near Plussulien, a site that was worked for a thousand years and whoseproducts are found throughout France; Type B from an unidentified area in theMontagnes Noires found only in Southern Finistre and Morbihan; Type C from anoutcrop near Plenven, now destroyed by quarrying and found both in Southern Finistreand Morbihan and across France; and Fibrolite, occurring as loose stones in fields andessentially in Finistre north of Brest.

    Burl has recently published Megalithic Brittany (Thames & Hudson 1985) which,although not a true corpus, does go to some trouble to identify the exact location andnature of prehistoric monuments, including museums and similar related items. Wehave used Burls reference numbers, located them on the Carte de Fr /100,000 ofInstitut Geographique National; and we use the rectilinear grid of this series.

    While Thom has made important studies of the observatories at Carnac, I know of nosurvey of alignments by him or anyone else in the rest of Brittany. The record of thesealignments is therefore my own; the surveying was done in 1975, using a plane tableand elevating alidade with orientation on local features identifiable from the Carte deFrance 1/100,000, being typically churches and water towers (chteaux deau).

    The alignments surveyed at Brignogan, at Plozevet, and at Guilvinec have not beenchecked; none of them was surveyed with any objective in mind and those at Guilvinecare marked on my survey notes as rough. Their consistency, therefore, with thepattern already found is impressive. The sites at Erdeven and on the seaward side ofPresqule de Quiberon were re-surveyed in 1985 and the original measurementsconfirmed.

    My identification of Site Catchment areas is therefore confined to the coast starting onthe North coast of Brittany at Lannion (Cote du Nord). I have not surveyed; partlybecause the rocky coastline, even allowing for a prehistoric shore on the l0m line, madea suitable beach improbable. Next along the coast is the large passage grave site ofBarnenez, which appears to stand much on its own in 39 a similarly rocky coast.

    At Brignogan, however, we have a characteristic Site Catchment area and a promisingbeach in the Grve de Goulven, particularly at the l0m line. On the headland north ofBrignogan Plage is a large Christianised Menhir, near the lighthouse. We identifiedto the S. East a smaller menhir that provided a good sight of the large menhir; that givesa bearing of 329 and leads to Newlyn in Mounts Bay. I thought there was a secondalignment from this small menhir to a prominent natural outcrop of le Garo with abearing of 39.00. Burl records another menhir south west of Brignogan at Kervizour(F81b) that appears from the map to give the same bearing. That needs to be checked,but if confirmed leads to Poole Bay and the mouth of the R. Avon. There is evidencetoo of a lower level for the shoreline; the passage grave (F70a) is recorded andillustrated by Burl as being submerged at low tide.

  • - 33 -

    In North West Finistre we find a Site Catchment area centred on the naturaloccurrence of Fibrolite stone axes. On the mainland the record is mainly of large andisolated menhir, but on the coast are two islands with a number of substantial passagegraves; at i Guernioc (F52) and at i Carne (F51). At both, these sites there is apresumption of a lower level for the shoreline; Burl comments on i Guernioc that... once a low hill joined to the Mainland.; that implies at least the 5m line.

    In the Baie de Douarnenez we find the peninsula of Camaret on which there are stillsome stone rows, but we have not surveyed them: but on the south side of the Bay wehave the peninsula leading to the Pointe du Raz on which we can identify a SiteCatchment area. There is a collection of Graves on the northern side; but on the shore ofthe southern side, on the Baie dAudierne, there is a large and very flat menhir, south ofPlozevet Pouldrenzic (F86a). This flatness is sufficient to give a good bearing of 173.5.I think some foresight may have gone, which is a pity as one would like to be quite sureof the prehistoric intention of the alignment, since it suggests a sea route right across theBay of Biscay to a beach in the vicinity of Picos de. Europa, a passage of 485 km. Thismenhir is in close proximity to an area rich in graves, generally described by JacquesBriand in describing the grave at Kersandy in LArchitecture Megalithique (1977); henotes also the abundant local working of flint from local rognons de silex.

    In the South Western corner of Finistre , in the general area inland from Pen March wehave a considerable concentration; centrally there were stone rows (F80), now muchdespoiled. On the southern coast East of Guilvinec near Lohan I recorded in 1975 twoalignments which I 40 described as rough but which should be included because theyappear to be part of the network of sea bearings. They are (F76) Lechiagat, Lehan,which Burl describes as this broad granite slab stands in marshy terrain ... and mycomment in my survey notes was ... is just behind the sand dunes and is flat pointingout to sea. At the position where the tip cuts the horizon there is a natural rocky area buttoo (un)impressive to give a bearing better than 5; and (F90a) Le Ruen and Burl saysThis is a remarkable jagged pillar of granite, 5.5m high with its broad S face towardthe sea and my survey notes say ... is also a flat and large stone but points roughly E-W. There is no obvious sighting platform close-to. The prehistoric sea line would havebeen 1 or 2 km farther out and no doubt the position of-these stones would have beenmore sure. However, that at (F90a) leads us to the coast at Erdeven, avoiding manyrocks and other hazards.

    The Carnac district is so dominated by the stone rows and other elements of theobservatory that few eyes are turned either to Erdeven or to the Cote Sauvage on thePresqule de Quiberon, but that is where we find two more areas with this samecombination of alignment and beach. Near Erdeven there is an alignment of eight stones(Kerascouet? dpd) (CA187b), these are large and some are thrown down, the whole is ina clump of gorse. However one can obtain a bearing; it gives us another location in thevicinity of Picos de Europa.

    Finally there are a group of five stones of fair size, (Manemeur? dpd) not recorded byBurl on the southern end of the Cote Sauvage. The northerly stone (Beg-er-Gonhennec?dpd) is seen to clip the horizon when viewed from three of the others. The remainingstone, large though it is, seems to be something other. From these three stones we obtainbearings which we shall see help, with the others we have noted in Southern Brittany, tocreate an extension of the network of sea passages round the Bay of Biscay to theNorthern Coast of Spain. It really is quite remarkable!

  • - 34 -

    41 Fig 1

  • - 35 -

    41 Fig 2

  • - 36 -

    42 Fig

  • - 37 -

    43 Fig

  • - 38 -

    44 Fig

  • - 39 -

    45 Biscay

    The sea passages suggested by the alignments in Brittany surprise by the length ofpassage implied across open seas notorious even today for foul weather. We shallconsider separately whether techniques existed for navigating over this distance; as tothe dangers of the open sea, they may well be a lot less, particularly in a small boat, thantidal and cross currents near rocks and headlands.

    We have extended our study round the shores of the Bay of Biscay, to the extent that wecan suggest a feasible pattern of sea passages. While we have followed the constraintsof our proposition, we do not have as much data for Aquitaine (Landes) or for Catalonia(Northern Spain); we do have, however, pairs of bearings indicating the same landfall.We have had to make some assumptions; in particular that the sailing point fromQuiberon is a stand-off in the lee of Belle-Isle to clear the small archipelago of islands.One of these suggests a landing on the coast at Lacanau, as does the rough bearing fromGuilvinec-Plozevet in Southern Finistre ; the closest prehistoric remains I am aware ofis at Le Gurp some 40 km. north. However, the line from Quiberon is a very preciseone.

    The other two bearings from Quiberon (with the stand off) lead to the coast north ofSable dOlonne. This coast is marked by a dolmen and a menhir, but we cannot tell howmuch a lower sea level would have altered the coastline. Inland from Sable dOlonne,however, is a Site Catchment Area of various substantial dolmen and two alignments;further inland in the Vende are various graves. The area fits the pattern and one of thealignments (Avrill? dpd)gives a good alignment for Bilbao; the other cannot easily bemeasured because one large stone is wired off and the other, behind the Mairie, is nolonger visible from the first. I have measured as best I can from the map, and onlyincluded it because it also suggests a landing at St. Vincente de la Barquera.

    The two landfalls in Spain are where we might expect them to be; that at Bilbao beingthe gateway to the Valley of the Ebro and Catalonia; that at St. Vincente de la Barquerato Leon and Portugal. The suggestion lacks the precision of the rest of the proposition,but contains sufficient merit to be made.

  • - 40 -

    46 Fig

  • - 41 -

    47 Hebrides

    There remains one other region that displays the same pattern of prehistoric sitecatchment areas connected by alignments indicating sea passages; that is the Hebrides.

    The sea distances involved in crossing the Clyde Estuary from Galloway to Arran andKintyre are short and not notably hazardous. Arran and Kintyre are typical of the sitecatchment areas we have been studying. To the north we find another around Kilmartinand we shall look again at that in Observatories. Beyond that we come to the openwaters of the Minch and on the island of Mull we find again our alignment and a beach.Mull should probably be regarded as two areas, North and South, however most of theother islands are of a size where the coast line defines the catchment area.

    The alignment of 334 from Mull leads to the Eastern coast of North Uist, rich instanding stones and graves. There is an associated alignment at 342, suggesting theroute to Canna (adjacent to Rhum); this may be a passage or it may be an ancillary tothe route to N. Uist.

    Most of the alignments in the Outer Hebrides we propose to deal with underObservatories but we should note two particular areas. To the North we find the areaof Callanish on Lewis; manifestly an observatory, but on the Eastern shore an alignmenton the mainland mountain of Suilven. This bearing at 98.3 leads to a point of entry toSutherland, the Eastern side of which is an important catchment area. Far to the south,indeed on the most southerly island of Berneray, is an alignment shown by Thom H6/5at 342 to Hecla is probably a reverse bearing from an adjacent beach to the large beachin Northern Ireland at Coleraine; the point of entry to the large settlements of Tyrone.

    Two other bearings we should note. From Brevig on the Eastern shore of Barra abearing of 135 leading to Coll. From Islay and its large beach of Laggan Bay thereappears to be a bearing for the beach at Coleraine; this depends on the alignment of twostanding stones shown on ordnance survey and needs to be checked, but they do lie in atypical set of stones in a typical catchment area.

  • - 42 -

    48 Fig

  • - 43 -

    49 Ireland

    Let us draw all this together by looking at the map of Ireland and observing therelationship between the suggested sea passages across the Irish sea and the pattern ofprehistoric settlement.

    The map we have used is enlarged from that in a paper by M. Davies in AntiquariesJournal Vol.25 (1945) and the suggested sea passages added. Davies, in her paper, isconcerned to show that there were sea lanes across the Irish sea and that they wereinfluenced by the terrain and vegetation on the one hand and by the strong tidal currentson headlands and in the North Channel; but she is not specific, as we have been, aboutthe sea passages.

    It will be seen from the map that the points of entry to Ireland that this paper proposesare very relevant to the settlement pattern on both sides of the sea. Let us look at themfrom the South.

    Tramore. The bearing from St. Ives leads to Tramore bay on the coast south ofWaterford. The graves concentrated here are known as end graves; they do not occurelsewhere in Ireland; they do not occur elsewhere in Britain except in the Scilly Isles.

    Wexford. Two bearings lead to the coast north of Wexford harbour; from Newport Bayin Dyfed and from Newborough Warren in Anglesey. The coast here is open, butchanging sea levels may have revealed an island offshore;based on the present 5 fathom line, these bearings would have lead north and south ofthe Island. As a point of entry to Ireland it would have lead to the settlements in theSouth and West.

    Arklow. The more northerly bearing from Newport Bay leads to the open shore Southof Arklow, between Ballymoney and Courtown. It will be remembered that these twobearings from Newport lie to North and South of Mt. Leinster; and so too would theroutes inland; this northerly one leading to the settlements West of the WicklowMountains.

    Drogheda. The bearing from Galloway leads to the open shore just south of the mouthof the Boyne; from which point of entry it is a short distance to the large collection ofgraves dominated by New Grange. A point of entry might be expected in this regionbecause of the typographic similarity of the Court cairns in Northern Ireland and SouthWest Scotland.

    We have shown no points of entry on the North Channel. There is one candidate; thealignment from Ballochray in Kintyre (Thom A.4/4) does give a bearing on RathlinIsland 50 off Ballycastle bay on the North Coast of Antrim. Despite the strongprehistoric presence in that area, we have not included it; preferring to treat it underobservatories.

    Coleraine. The only point of entry on the north coast seems to be the bay of Coleraine.For this there are two bearings; from Berneray and from Islay. As a point of entry theroute leads either to the extensive settlements of Tyrone or to the settlements of Antrimand the axe factory of Tievebulliagh. Taken with the routes between regions on theWest cost of England and Wales that we have discussed in some detail, the result is aremarkably coherent network of sea communications.

  • - 44 -

    50 Fig

    From Davies

  • - 45 -

    52 Chapter 4. Interpretation

    Support for Navigation

    Now that we have set down the Corpus of data for prehistoric sea passages on the WestCoast of Britain and France we must consider how the method of support may be usedto evaluate the proposition.

    We have already indicated that alignments may be for more than one purpose so thatfirst of all we have to decide how to separate them into mutually exclusive categories;not forgetting that any one may be or appear to be usable for more than one purpose.The categories that we propose to define are as follows:-

    A. Sea PassagesB. Dry-shod land routesC. Observatories; essentially concerned with the limiting declinations of sun and

    moonD. Fans and Rows; as an essential part of an observatoryE. Double purpose alignmentsF. Alignments with no obvious objective.

    We may wish, as our knowledge increases, to add to or remove from any of thesegroups. For the moment we are considering Categories A and B and allocating toCategories E and F if necessary.

    Category A. Sea Passages. Where we have a Site Catchment Area for late Neolithic andearly Bronze Age activity which is close to a sandy beach and there are alignments:

    If we have alignments of any of the categories:-

    1) the flat face of a slab2) two or more stones aligned3) a circle with an outlier4) One circle observed from another,

    and we transfer that alignment to the beach, and the alignment (in either direction) isincluded in the seaward arc of the beach,

    Then: the sea passage indicated will be free of nautical hazard (rocks, strong currentsand so on) and will lead to a sandy beach marked by stones or graves of anotherSCA, or be the natural point of entry to another SCA. 53

    Category B. Dry-shod routes. Where we have a Site Catchment Area for late Neolithicor early Bronze age activity, with or without a beach, and there are alignments asspecified in Category A;And the alignment is not included in any seaward arc from a beach,Then, from the alignment will be indicated a dry-shod route free of natural hazards(rivers and marshes, broken country) and will lead to a menhir or circle or grave or be anatural point of entry to another SCA.

    Now in setting up either of these propositions we have, by specifying four types ofalignment, used up four degrees of freedom; and so too for the target beach we havespecified two conditions. So from our total observations we have to deduct six for

  • - 46 -

    degrees of freedom. On the other hand, we have stated that having specified thealignment that two conditions will -result; the one dependent on the other and so we canadd the support for a safe passage to and a suitable beach will be found.

    So we have a simple go/no go situation m = 2 the number of observations will be n,and S = 2(n-6) Ln 2. What we have recorded is, of course, open to debate, but is:-

    Region Category A B E FHutton Moor 1Langdale 4 1 1Galloway 4 2 2Gwynedd 2 2Dyfed 4 2 1 2Cornwall 2Brittany 6Biscay 4Hebrides 6

    32 8 2 4

    So at this stage we may say we havefor Category A Support = 2 x 26 x 0.7 = 36.6

    and for Category B Support = 2 x 2 x 0.7 = 2.8

    54 Navigation

    Having identified over 30 alignments indicating a sea passage between beaches andbetween areas of prehistoric interest, we must consider whether that is feasible; orconversely, if we are satisfied from the support for the proposition, to consider what thattells us about navigation techniques at that period.

    We know very little about the ships involved; we may presume dugout canoes; we mustpresume some heavy rafts if we accept the movement of bluestones from Prescelly toStonehenge. In the Bronze age there is evidence for sewn planked boats but we do notknow if they were rowed or sailed or both. They are presumably developing through theperiod.

    The evidence we have produced implies that these ships could stay at sea for quite sometime in open water and out of sight of land. The length of the passages ranges from 50km to 500 km; the shorter possible in daylight by direct observation; the longestimplying up to three nights at sea. The passage lengths may be classified roughly asfollows : -

    1) Up to ~ km (48 miles) there are 8 and we presume direct observation, journeytime say 8- 10 hrs.

    2) 90- 120 km (up to 75 miles) there are 9 and we presume a night passage steeringby the stars, journey time 12- 15 hrs.

    3) 135- 195 km (up to 120 miles there are 7 and we presume a night and a day(direct observation at start and finish), journey time up to 20- 24 hrs.

  • - 47 -

    4) 200- 500 km (up to 300 miles there are 8; we need to presume two or threenights at sea with a full day out of sight of land. This requires a lot of seamanship, butwe cannot dismiss it; there is as much support for the long journeys as the short.

    We observed that most of the alignments were away from the beach; this is onlyfeasible if they are assumed to be a training aid for learning the succession of stars thatindicate the bearing required. Without a compass this method of setting a course bystars of a constant declination will be found in several societies; the Portugese in theMiddle Ages (they used the Cross staff as an aid); the Arabs in the Indian Ocean (theyused the Kemal, the precursor of the Cross staff as an aid); and the island people of theSouth Pacific until recent times.

    We give some details about the Kemal and about the 55 South Pacific sailing techniquesin a supporting paper. The South Pacific sailing makes use of 3 to 9 stars (or groups ofstars) for a particular passage; perhaps we have a hint about why there are so manynine maidens? A particular star or group will be used for an hour or so as it moves; itwill be observed at an altitude of several degrees, to avoid the effects of low cloud onthe horizon. The Arabs used their Kemal to allow for the angle of the star to the course;do we have in this device a possible use for V buttons & rings often referred to as beltfasteners? We have set this idea out in a supporting paper.

    The South Pacific sailors travel long distances and will be at sea for a week or more;during the day they will steer by observation of the Sun, by an innate sense of position,and by the set of the long waves of the open ocean. For the middle period of the passagesome such techniques would have been essential.

    The observation of the stars is at an altitude of several degrees; and we note that thatcondition applies to many of the alignments. The alignments are only at one end of thepassage; so that the journey must also be made on the back bearing. We would expectships with a high prow and stern for this reason alone. We should also deduce theorganisational feature that sea passages were controlled from one end.

    Sailing by the stars, the ship will follow a curved path and we have drawn a straight lineacross the grid. For short journeys this is probably not a large error; for the longestjourney we have calculated the true course and it is shown by a dotted line on the mapof Biscay. What shape that curve will be will depend on how the navigation is carriedout; how much direct observation of the land is made at start and finish; how navigationis carried out in mid passage; the elevation of the stars observed; and the bearing beingfollowed. The curve on the Biscay map is obtained by assuming that the observations atPlozevet are made at an altitude of 2.

    We have (after Thom 1967)

    Sin d = Sin 1 Sin h + Cos 1 Cos h Cos A.

    d = declinationh = horizon altitude1 = latitudeA = Azimuth or bearing

    for the observing site d = constanth = constant

  • - 48 -

    but clearly, if we continue to observe the same stars (which retain the same d) then thebearing A will vary with the latitude (l).

    56 In the particular curve drawn we presumed that although h = 2 that it was 00 at sea(probably incorrect), and that on a bearing of 173.5 we move from latitude (1) 4800N to 43030t N.

    We would need to be more precise about the longer passages for detailed extension ofthis calculation to yield further useful data.

    57 Observatories

    In considering the site catchment areas concerned with the Stone axe trade we noted thatthese could be assembled into regions that are geographically independent. In theseregions we set aside particular sites of circles and alignments as observatories. Let uslook more closely at what we mean.

    The regions are all connected by our network of sea passages identified by a stonealignment, an adjacent sandy beach, and the indication being of a satisfactory seapassage to a target sandy beach also associated as a point of entry to a neolithic area. Allthe regions originate sea passages; there are others that do not; Ireland, for instance;Yorkshire; Northumbria; East Anglia.

    They are from South to North:- Brittany, Cornwall, Dyfed, Gwynedd, Cumbria,Galloway, Argyll, Hebrides.

    Each of these regions, with the exception of Gwynedd contain one observatory site atleast; though all are of various types and complexity. With two exceptions no othercoastal region has an observatory site. Let us look at the exceptions:-

    Gwynedd. There is only one site, W2/i Penmaenmawr in the region. Without too muchanticipation of our argument, this is a late site so that for most of the period the regiondepended on sea passages controlled from other regions. We cannot be too precise aboutearly or late sites, at any rate at this stage, but by early we might mean largeundressed stones, and by late we might mean small stones in specialised geometricarrays and association with beakers.

    Caithness, Orkney and Shetland. This region, dominated by difficult sea passages, (butfrom the archaeological record, sea passages that were made in the period), is onlyindirectly linked to the other eight regions. But we must include it and its twoobservatory sites of N1/1 Mid Clyth (and other stone row sites in the district) and thering of Brodgar in Orkney.

    Stonehenge is clearly in the category of observatory but the regional association withthe other seafaring regions is less direct. The bluestones came from Prescelly; and weidentify a sea passage from Brignogan in Brittany to Poole Bay. The association ofStonehenge with seafaring must rest on its position in the region of Southern England;the distribution map of Stone Axes (Clough & Cummins i979) in Southern England issuggestive of a coastal traffic along the south coast and into Southern Essex.

    58 So we have nine regions with observatories, and at the same time an active interest inseafaring. On top of the support for seafaring between regions, we can now say thatthere is a strong support for observatories being associated with seafaring.

  • - 49 -

    What then are these observatories and show could they have assisted the seafarer?

    There seem to me to be two principle techniques that would develop from thewidespread observation of stars and moon needed for navigation. Very quickly theywould observe the association of springs and neaps with the phases of the moon. Itwould take a lot longer to unravel the complexities of the tides and of associating themwith the 9 perturbation of the lunar orbit and its 18.6 year cycle. Over a protractedperiod they would observe that the stars precessed. That being so, we must expect avariety of sites reflecting this developing set of ideas; we would expect not onlydevelopment at a site but between sites and regions.

    Tides themselves are directly proportional to tidal currents; it is this aspect which makesthem of such concern to primitive navigators.

    The site at Castle Rigg L1/1 in Cumberland is a useful model of the primaryobservatory ot sun and moon; we reproduce the panorama from Thom 1966. Thepanorama of the bowl of hills surrounding the site enables an opinion to be made on theuses to which the site may have been put. However, Thom (1967) lists here limitingdeclinations for Sun at +24.3 and Moon at - 29.8~

    In Galloway the site Laggangarn G3/3 is similar in the type of observation that may bemade; various solar calendar lines are also possible but limiting declinations are for Sun- 23.7 and moon at - 30.4 and - 19.6.

    In Dyfed we have a problem; there were several major sites south of the PrescellyMountains but they have essentially been destroyed. These circles are close by thelocation from which the bluestones were sent to Stonehenge. The identification of thesource of bluestones from a small area on the South East slopes (of Prescelly) at Cil-maen- Llwyd will be found in Antiquaries Journal Vol.3, 1923. A review of the circlesand alignments from this area, dating from about 1910, is by Rev. W. Dene Bushell,FSA, Among the Prescelly Circles in Archaeologia Cambrensis 6th Series, Vol.XI.We have located these sites on an enlarged part of the 1 O.S. Sheet 139; but thealignments and so on given by Bushell are based on stellar alignments and even in hisday they were much depleted. Gors Fawr W9/2 he includes and is intact and we havereferred to 59 earlier. For the rest we must content ourselves with the categoryobservatory.

    There is one precise observation in Dyfed to which we have already referred; Parc yMerw W917. This is a large stone, long alignment indicating the northern slopes of Mt.Leinster and one of the limiting declinations of the moon. Precisely though this line isindicated, it raises the question of how it could have been observed. The moon sets onthis line once every 18.6 years; the period over which the Stone Axe trade ran wasbetween 1000 and 1500 years; we have only between 50 and 80 opportunities to makethe observation; but only very occasionally is the moon setting actually on the limit ofits orbit; the setting on the day preceding and succeeding the limiting day will besubstantially short. We might suppose that by chance the moon is actually observed onone of the occasions when it is truly at its limit; that that is recorded by an alignmentand the repeat performance awaited; but for any useful data to be obtained about thebehaviour of the moon a technique has to be devised for estimating, from the shortfallon the days before and after, by how much the limiting position should be increased.Thom suggests that, if such precise observations are made, there must have beendevised an offset technique (he uses the phrase a stake method) and that from the

  • - 50 -

    parabolic approximation to the lunar movement over those two days there is acharacteristic offset dimension for each site, C, for which we should require someevidence if a precise lunar observation is to be presumed.

    The region of Cornwall has on Bodmin Moor a site, the Hurlers S1/1, which appears toconcentrate on the alignment of circumpolar stars; an interest stimulated perhaps by theobservation from the sea route indicated by S1/9 that the star Deneb set on that line.Beyond that speculation we find (outside Cornwell) the site Merrivale S2/2 onDartmoor that does show evidence for the extrapolation length C. This site wassurveyed by Wood & Penny and reported in Nature, Vol.275, 1975. We include a copyof this note for a description and survey not performed by Thom, and for its compactdescription of the importance and use of extrapolation.

    Turning now to Scotland. There are two observatories, both in the Hebrides, that do nothave evidence of extrapolation but one more comprehensive than the sites we have sofar noted; they are at Callanish and North Uist.

    There are, however, three observatories in which there is evidence for the extrapolationlength and upon which much of the argument for observation of the 9 perturbationrests. The most southerly of these is at Temple Wood A2/8 near Lochgilphead inArgyll. There are in the region a number of sites that individually indicate a particular60 position on sun and moon but at Temple Wood are a series of stones and circlesarranged so that, by moving along a sequence of them, the setting of the moon behind aparticular profile of the hills can be observed very precisely around one or more of itslimiting declinations, much as at Merrivale.

    At Brogan in Orkney we have a similar configuration; it is, there, complex butessentially the sequence of observations is performed by moving along a slightly raisedridge.

    However, on the mainland of Caithness we have not only an observatory of this type buta configuration which is called Stone Rows or fans; a number of rows of fairly smallstones are laid out in the shape of a narrow fan. The radius of the fan is identified as theextrapolation length G and the distance between stones provides a measure of thecorrection that must be applied to the limiting positions observed on successive days, todetermine the true limit. There are, in close proximity in northern Caithness, four ofthese stone rows; suggesting certainly a speciality of the district, but probably itsgenesis and development.

    Finally we come, way down South, to Brittany and to the very complex array of stonesaround Carnac. Thom has identified two lunar observatories; one based on the menhirLe Menec; and a much larger one based on the Grand Menhir brise at Lochmaraquier.It is sometimes protested that the Grand Menhir was never erected; but there wascertainly the other one operative; in any event there may have been some tree ortemporary wooden structure in anticipation of the stone. The scope of the other stones inthe district do tend to encourage the idea that the main observatory was operated.

    Here we find, as in Caithness, stone rows set out fan- shaped and based on theextrapolation length G for the Grand menhir; there are two fans, for the alignments onQuiberon, and for the northern sector. The novel feature at Carnac, however, is theenormous stone rows at Kermario and Kerlescan.

  • - 51 -

    So you will see that we have eleven observatories (for Stonehenge has the equivalent ofstone rows as post holes on the axis); and they develop in their complexity through atleast three technical phases;

    a) fixed alignments on specific lunar, solar, or stellar positions. There are six ofthese:- The Hurlers on Bodmin Moor, Prescelly, Castle Rigg, Laggangarn, N. Uist andCallanish. 61

    b) Offset alignments with indication of C for the site. There are three: Merrivale onDartmoor, Temple Wood and Brodgar in Orkney.

    c) Offset alignments with fans of stone rows. There arethree, Stonehenge, Caithness and Carnac.

    I take it that the jigsaw puzzle is just about complete; with the most advancedobservations they were competent to comprehend the details of lunar movement and torelate it to the tides.

    I wrote a paper ten years ago, suggesting that the Alignments at Le Menec usedfortnightly observations of the lunar declination on Le Grand Menhir to predict the finemovement of the tides.

    That paper, re- read as the closing piece of the Jigsaw, remains convincing to me. It is acomplex argument and I have added it as a supporting paper.

    The site at Le Menec (Carnac), then, is the end product of a long period of observationwhich is represented by the periods : -

    i. Determination of lunar and solar limiting positions by fixed alignments.

    2. Refinement of limiting positions by defining a traversing platform to observe thechanging position for a period either side of the limit.

    3. calculating from these traverse positions what the limiting position would be, byusing the stone fans.

    4. Increasing the number of traverse positions so that the whole range of lunarmovement may be observed; making those observations for the (e + 1) and

    the (e - i) sectors; scaling both to one range, using stone fans; relating the change inthose measurements to a measurement of tides on the same set of stone rows.

    That is as effective a closing of the jigsaw as you will find; the above argument isincredible unless related to the whole activity of navigation; the alignments andobservatories seem pointless unless they lead up to some purpose of this magnitude.

  • - 52 -

    63

    References

    1. J. Radley (1974) The Prehistory of the Vale of York, YorkshireArchaeological Journal. Vol 46.

    2. T.G. Manby (1965) The Distribution of rough-cut Curnbrian and relatedstone axes of Lake District origin in Northern England. Trans. Cumb. &Westmoreland Antiquary & Archaeological Soc. Vol LXV new series, 1965.

    3. lain Davidson (1983) Site variability & prehistoric economy in Levante in C.Bailey. Ed. Hunter- gatherer economy in prehistory. CUP.

    4. Sir Lindsay Scott (1951) The colonisation of Scotland in the secondMillennium BC. PPS. new series Vol XVII.

    5. M. Davis (1945) Types of Megalithic Monument of the Irish Sea and NorthChannel Coastline; a study in distribution. Antiquaries Journal XXV.

    6. M. Davis (1946) Diffusion and Distribution patternsof the Megalithic Monuments of the Irish Sea and North Channel Coastline.Antiquaries Journal XXVI.

    7. T.H.McK. Clough & W.A. Cumins Stone Axe Studies. CBA Research ReportNo.23.

    8. O.G.S. Crawford (1927) Lyoness. Antiquity Vol.1.9. P. Fowler & C. Thomas Lyoness revisited. The early walls of Scilly. Antiquity

    Vol LIII.10. C. Thomas & P. Pool (1964) The Principal Antiquities of the Lands End

    District. Cornwall Archaeological Society. Field Guide No.2.11. J. Michell (1974) The Old Stones of Lands End. Garnstone Press.12. M.H. Ridgway (1946) Prehistoric Flint Workshop site near Abersoch,

    Caernarvonshire. Arch.Camb. Vol XCIX Pt 1.13. J.B. Lewis (1971- 2) An account of the Penbedw Papers in the Flintshire

    Record Office. Flintshire Historical Society Vol 25.14. Sir Cyril Fox & B. Dickens (1950) The Early Cultures of North West Europe.

    H.M. Chadwick Memorial Studies.15. C.A. Gresham & H.C. Irvine (1963) Prehistoric routes across North Wales.

    Antiquity XXXVII.16. D. Lees (1984) The Sanctuary: A Neolithic Calendar? Institute of

    Mathematics & its Applications Vol 20.17. C. Larnbrick (1983) The Rollright Stones. Oxford

    Archaeological Unit.18. Roger Joussaume (1985) Des Dolmens pour les Morts. 64 Hachette.19. (1977) LArchitecture Megalithique. Colloque du 150a anniversaire de la

    Societe Polymathique du Morbihan. Chateau Gaillard Vamey.20. Sir Norman Lockyer (1906) Stonehenge and other British stone monuments

    astronomically considered. McMillan.21. Sir Norman Lockyer (1905) On the Observations of Stars made in some British

    Stone Circles - Preliminary note. Proc. Royal Soc. Vol 76- A.22. Rev. W.D. Bushell (1910) Amongst the Prescelly Circles. Arch. Cambrensis

    6th series Vol XI.23. G. Williams et al (1963) Swansea Bay to Worms Head. Gower Society.24. J.E. Wood & A. Penny (1975) A Megalithic Observatory on Dartmoor. Nature

    Vol 257. 18 Sep 1975.25. - (1972) Ancient Astronomy at the Royal Society.

    Nature Vol 240. 29 Dec 1972.

  • - 53 -

    26. 1.0. Angell (1978) Megalithic Mathematics, Ancient Almanacs or NeolithicNonsense. The Institute of Mathematics and its Application. Vol 14. No 10.

    27. P.R. Freeman (1976) A Bayesian Analysis of the Megalithic Yard. Journalof the Royal Statistical Society A. Vol 139. Pt I.

    28. A. Thom (1955) A Statistical Examination of the Megalithic sites in Britain.Journal of the Royal Statistical Society. Vol 118 Pt III. 1955.

    29. A. Thom (1966) Megalithic Astronomy: Indications in Standing Stones.Vistas in Astronomy. Vol 7.

  • Peter Davidson 1986-2009

    MEGALITHIC AIDS

    TO

    NAVIGATION

    1986

    Support Papers I

    THE JIGSAW PUZZLE

    By P.B. Davidson

    Copyright 1986-2009

    Privately Circulated

    Scanned 2009

  • Peter Davidson 1986-2009

    Part I Jigsaw Puzzle

    1) The Jigsaw Puzzle: an allegory2) Logic and Algebra: some additional considerations

    a. Propositionsb. And . Orc. Sampling a Multinomial Distributiond. PLnPe. Expected Valuesf. Degrees of Freedom

    See Also, in separate files:

    Part II Constant Declination NavigationPart III Alignments at le Menec as a Tide PredictorPart IV Mensuration

    Keywords:

    Multivariable entities; statistical fit; likelihood; Measure of support;Cluster analysis; test of significance; A W F Edwards

  • Peter Davidson 1986-2009

    3)

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986-2009

  • Peter Davidson 1986 - 2009

    MEGALITHIC NAVIGATION 1986 - SUPPORT PAPER II

    Navigation

    1) Constant Declination Navigation

    2) Arab Navigation

    being an extract from Proceedings of the Asiatic Society of Bengal, Dec. 1836Found at the Cambridge Univ. Library UL No Q 620.C5.1.

    3) the Ancient art of land finding in the Pacific

    being an abstract from We, the Navigators by David Lewis (1972)Australian National University Press, Canberra, pp.5461

  • Peter Davidson 1986 - 2009

    S2.2

    CONSTANT DECLINATION NAVIGATION

    We must consider the way in which Megalithic man would. have used his alignmentsif he had used. them for n~vigation (on land. or sea).

    Through the alignments there will set regardless of season a constant succession ofstars. Only a few of these will be of first magnitude and they will as a result havedifferent extinction angles (that is to say the smaller stars will cease to be visible ata height above the horizon depending on the opacity of the atmosphere).

    It is possible to memorize the succession of stars through the alignment and. to sail (atnight) on the constant course so indicated. At any particular season the star firstvisible at sunset will change. This in itself would have given a guide to desirableseasons for making journeys or it could. have been used in conjunction with othercalendar alignments, (or the observation of the one may have stimulated thedevelopment of the calendar).

    This technic~ue is still used. by the Arabs to make open sea voyages from India toZanzibar and. back in d.hows. They are driven and. guided. basically by the monsoonbut apparently refine their course by using a ring and string. This ring and stringconsists of a (triangular?) ring held in the eye to which is attached. a string and a(triangu.lrr?) ball. The ball is held at arms length and. moved in tLe arc of a circle.The attention of the eye is concentrated on a limited target. The ball can be movedthrough two stars of constant declination and will give a constant inter9ection with thehorizon.

    PBD/PAH7th October 1971

  • Peter Davidson 1986 - 2009

    V-BUTTONS AND RINGSIt is frequently reported in Beaker burial context that t there were one or more 7-boredbuttons. These have remarkable constant characteristics. The diameter is between 11/4" and 1 1/2" and forms a flat cone about 1" deep. In the centre of the flat face arebored two conical holes angled s