Hunebedden (Dutch dolmens) in proportion

About orientation patterns with hunebedden (in European context)

Alignments [12.1]

Click here for the astronomical charts concerning the Neolithic Age

The ground plans on this page are drawn based on:
- Heveskesklooster G5 - excavation plan, Lanting [12.2]

On this page the question is treated, if some relation can be appointed between the orientation pattern and astronomical phenomenons. Because of a number of uncertainties, this question cannot be answered easily. Aspecially the association with stars is problematic, since their rising shifts along the horizon considerable over centuries. Archeologists can date the finds from hunebedden and dolmens with a tolerance of 50 years at best [12.3], while the rising of many stars appears to shift with 1° already in such a time span. Before an alignment can be called convincing, supplemental regularities are needed, covering several decennia and latitudes. Furthermore most of the time the field conditions of 5000 years ago are unsure, while the growth or decline of ½° in inclination can cause a deviation of 3° in azimuth. This counts for the rising and setting of stars as well as the sun or moon. Planets and comets should be left out of consideration in any case, because of their fast movement between the stars.

In the astronomy there is the problem of the ’obliquity’ of the earth axis. It changes slowly and with it the declinations of sky objects change a little over the millennia. Various authors give different values for the year 3000 BC. Burl has 24°02’, Langbroek has 23°49’ and González-Garía & Costa-Ferrer have 24°50’. A difference of 1° in declination results in a non-neglectable deviation of 3° of azimuth with a latitude of 53°. On this page the data of the University of Soutern California [12.4] is used, which gives an obliquity of 24°03’ in 3000 BC with this latitude.

Especially the data on the pages Orientation grid and Hunebedden with satelites makes it worthwhile to investigate if people aimed to have alignments at the rising of the sun and moon with the solstitia and moon standstills. Around a latitude of 53° those vary from azimuth 35° till 58° and from azimuth 122° till 145°. Next the mirroring around south stands out for the initial grid or the real orientation grid. Due to their soutern azimuth it cannot be associated with the sun or moon. But they do remind of the double peak in the orientation of the gates from the study of González-García and Costa-Ferrer [12.5]. This peak mirrors around south for sky objects with a declinationn of more or less -35°, having a top near -36°. The autors point out six stars as a candidate for an alignment: the four brightest stars of Crux and α- and β-Centaurus. Although other authors have found an association with these stars, an alignment is problematic since their low soutern right ascention [12.6]. In Drenthe sky objects with a declination of -36° don’t rise more then 2° above the horizon and are faded out by the atmosphere. Nevertheless on this page a working hypothesis is put forward for the alignment of orientation grids on α-Crux. Theoretically the poor visibility could have been the reason for the usage of an orientation grid. The best matching grids give azimuth values of 162° till 171° (see the page Orientation grid) and they confirm with declination of -36° till -37½°. Except for α-Crux the other candidate stars rose nothern of 160° around 3000 BC and must be abandoned. During the building of the hunebedden (from 3350 till 3050 BC) α-Crux was rising at 164° till 171°. Possibly the rising of β-Centaurus fits with the orientation grid at 153° around 3400 BC, but then it’s surprising that people didn’t align at the much brighter neighbour α-Centaurus. This possibility is not studied any further.

Assessment of the sites

  • Hunebedden D21 and D22
  • The design of this pair of hunebedden is discussed on the page Parallel orientation lines. The mentioned orientation lines of D22 haven’t been measured in the field, but depend on the excavation plan of van Giffen (see the page Measurement of the ground plans).

    Hunebedden D21 and D22 seem to have one large grid at 43°/133° in common. In Bronneger the winter and summer solstitia point at resp. 131° and 313° (= 180° + 133°). With D21 the orientation lines at 133° between the trilithons can’t be determined better then having a margin of a few degrees and therefore one cannot choose between an alignment at the winter or summer solstice. The gate deviates qua orientation a little from the grid, but it was absent during the excavation by van Giffen and has been restored later. Since the position of the side stones opposite the entrance doesn’t allow for a trough orientation line, an orientation at the winter solstice is most favourable. With D22 an orientation at the summer solstice is more obvious. D22 has a very precise alignment starting from the corner Sl1/Z1 and running alongside a flat surface (marking) at Z1’a. Between Z1’a and Z1’b there is a very small gap and one would expect one stone only at this position. (Initially also van Giffen suspected one big stone, as one can make up from the field plan in his atlas.) This supports the idea of a deliberate setup of the orientation line. In addition the outer side of Sl2 stands at 133°.

    In D21 the orientation lines at 167° form a crucial part of the structure of the hunebed. They run through the gaps between the side stones from the Z-side to the Z’-side. In both D21 and D22 the orientation lines at 167° can be setup via the common grid. At the moment of the building of D22 (about 3200 BC - Brindley’s horizont 3 [12.7]) an association with α-Crux is possible, but this is not the case with D21 (about 3350 BC - horizont 1[12.8]). Taking the sloping ground into account (-½°), than α-Crux rose at ±163° during the building of D21. Strikingly this corresponds to the orientation of the intial grid for D21 and D22 - see the page Orientation grid. One can ask oneself, if the orientation may have been modified sometime. Van Giffen took pictures of his excavation, which show clearly that after construction people have changed the arrangement of some stones. {Revision}

    On the photographs [12.9] one can see a light-colored stroke a little above and parallel to the primary chamber floor - near to the position of the secondary floor. This stroke runs at the entrance side over the threshold, the side stones, the padding stones and the end stone (yellow arrows on the picture beside) and must have been caused by some wall-to-wall event. The stroke exists on side stone Z1 too, which leads to the conclusion, that it cannot have emerged during the excavation. Van Giffen found this stone laying as a part of the secondary floor [12.11] and erected it again. Maybe the upper layer of the floor has been moisty for some time and lime was soaked into the stones. (Have the small pieces of calcinated bone, that van Giffen found in the chamber, to do with it? [12.10]) Anyhow, the stroke proofs, that the stones of the opposite wall (Z2, Z3 and Sl2) must have been layed aside when the stroke emerged. Afterwards they have been set upright again, so that the stroke on these stones stands vertically nowadays (the red arrow).

    After five thousand years one can speculate about the motive for the event only. Adjusting the initial orientation scheme is one of the possibilities. In order to transform orientation lines at 163° in lines at 167°, people needed to move the side stones for at most 10 cm. This wouldn’t have caused big problems for the construction. A possible renovation can have happened at the same time D22 was build, which explains the common grid of both hunebedden.

  • Dolmen Schuldt 632
  • The orientation grid of this ’Uhrdolmen’ near Frauenmark matches that of D22 (see the page Orientation grid). Both have their north-eastern end stone at 132°-133°. With Spr. 632 this is aligned at the winter solstice precisely. Next, due to the threshold at 167° and a diagonal at 193°, an association with α-Crux is possible around 3500 BC. The building of the dolmen comes under the earliest phase of megalith building in Mecklenburg then (FN-II, see the page Chronology), which agrees with the classification as ’Uhrdolmen’ [12.12]. Also orientation lines at 50° and 140° exist. The combination of 140° and 167° reminds at the structure of D39 and D40 (see the page Hunebedden with satelites). Possibly the orientation at 140° can be taken as an approximation of the soutern major standstill of the moon. Near Frauenmark this standstill happens at 144°. Since a major standstill is a rare phenonmenon, the 4° northern deviation in the alignment should not be surprising.

  • The hunebedden with satelites D23, D24, D25, D38, D39 en D40
  • On the page Hunebedden with satelites it is discussed, how the mutual orientation between the central hunebedden and their satelites is setup.

    In Drenthe the major and minor southern standstills of the moon occur at respectively 122° and 143°. At first glance the mutual orientations of 117° and 140° agree with it quite good. But one should question how the orientation line at 117° can run 5° too far north. Supposing a rough orientation at the standstills, we expect a southern deviation for the 122° because of the oscillation of the path of the moon in relation to that of the sun. The alignment at 117° doesn’t fit in and can be raised as an objection to an alignment on the moon. The alignment at 140° can be accepted as a rough orientation, but here we see something remarkable with the structure of the hunebedden. On the basis of the ground plans, it appears that an original orientation at 143° is rendered into an orientation at 140°. In other words, a correct alignment would have been botched on a later stage. This sounds very implausible.

    In the hunebedden D39 and D40 an orientation grid at 171° is transformed into a grid at 167° at some moment. Compared to the rising of α-Crux around 3200 BC this would mean a retrograde movement, so only the 171° can have functioned as a real alignment. Near Emmen α-Crux rose at azimuth 171° around 3200 BC. This matches with the dating of the ceramics from D40 at Brindley’s horizont 3 [12.13].

    D25 has an orientation grid at 162°, which is an exceptional value for Drenthe. After analysis of the terrain an association with α-Crux still is a possibility. The hunebed lies on a flank of the Hunze valley and at 162° one looks in the direction of a meander of this creek (nowadays the Canal Buinen-Schoonoord). With a sloping hillside of -¼° the star rose at 162° around 3300 BC.

  • Drouwenerveld D26
  • The orientation grid of D26 has an orientation of 167°. In this direction the field is inclining a little (app. 0.1°). Supposing an alignment on α-Crux, this would have occured around 3400 BC. According to the adjusted datings of Lanting (see the page Chronology) this is 50 years too early. Herewith it will have been one of the earliest hunebedden in Drenthe (Brindley’s horizont 1 [12.14]).

  • Dolmens G5 and Sprockhoff 68
  • Dolmen G5 near Heveskesklooster has a general orientation of 143°, which agrees with the rising of the moon at its major soutern standstill. At this latitude the orientation grid at 170° lends to an orientation on all standstills of the moon par excellence. These are: the major soutern standstill at 144°, the minor soutern standstill at 122°, the major nothern standstill at 35° and the minor northern standstill at 56°. In the left figure below the minor soutern standstill is obtained by averaging the 117° and 126° lines. Only the minor northern standstill differs from the grid by 2° and needs an apposition, which truly has been implemented in the hunebed. In addition the northern part of the dolmen doesn’t fit the grid well, as if the side stones were placed towards the end stone too far.

    In spite of the possibilities of the grid at 170°, one must doubt if it really aimed the mentioned alignments. As for the orientation, G5 is identical to Spr. 68 near Langholz (Schleswig). In 1977 the dolmen was rebuilt again - before the restoration only the dark grey stones in the figure above stood upright. From the sizes of the restored dolmen one may assume, that it has had the same orientation grid as G5. Spr. 68 has a more northerly latitude, so that the standstills of the moon no longer match. The major soutern standstill occurs at azimuth 147° for example. Like the dolmen near Frauenmark this would mean a divergence of 4° to the north. This isn’t surprising for a phenomenom which happens rarely, but the ideal formula of the grid should be doubted yet. It sounds more convincing, that a usefull grid was applied to another situation wrongly then that a less usefull grid was copied elsewhere and happend to be usefull by coincidence. This reasoning wants G5 to be errected earlier as Spr. 68, while it’s generally accepted, that the dolmens of Schleswig-Holstein precede the monuments in the Netherlands. For a matter of fact, during the excavation of G5 it appeared, that big pieces of stone were broken from two side stones [12.15]. Maybe the current ground plan is the result of renewed insights. This could explain the shifted side stones in the northern part of the dolmen too - as if the adaptions hadn’t finished yet. The excavators found the dolmen dismantled under a thick package of clay, with its cap stones spreaded out.

    The orientation grid of G5 has an orientation of 170° and a diagonal with the proportion 1:3 at 189°. These values mirror each other around south almost, so one can think of the rising and setting of a celestial body. At the time of the building of G5 (horizont 1, ±3350 BC [12.16]) this can have been α-Crux.

  • Dolmen Sprockhoff 541
  • This is a special dolmen near Serrahn. On the page Orientation grid two orientation types are expound. From them an orientation on the minor and major northern standstill as well as on the minor and major southern standstill of the moon emerge. The longitudinal axis lies at 121° - the minor southern standstill at 122°. The threshold lies at 58° - the minor norhern standstill at 56°. The northerly end stone stands at 32° - the major northern standstill at 33°. From the norh-eastern corner two lines can be drawn towards the threshold. Via the first a grid line runs south of it at 148° and via the other a diagonal with proportion 1:1 runs north of it at 140°. Again (compare Schuldt 632 and Spr. 68) we find a difference of 4° too far north for the major standstill. Also the average between both lines could have been aimed for, but this gives no more then a very rough alignment.

  • Dolmen Sprockhoff 540
  • Dolmen Sprockhoff 540 stands with its longitudinal axis at 167° and with a diagonal at 193°. Those values mirror each other around south and can be associated with the rising and setting of α-Crux around 3550 BC. The building of the dolmen must have taken place in the earliest phase of megalith building in Mecklenburg then (FN-II, see the page Chronology), but this seems to be a little too early for an enlarged dolmen [12.17].

    Except for an alignment on α-Crux the dolmen could have an alignment on the winter solstice. One of the diagonals runs at 135° through the dolmen. This is 2° southerly of the soltice at 133°.

  • The dolmens of Soutern France
  • Here it concerns the five dolmens, which are mentioned on the page Orientation grid. Only with Grammont 1 and Bourbouillet 2 alignments at the major standstills of the moon are found. Grammont 1 has its longitudinal axis and its gate pointing at the setting moon at its major northern standstill (47°) and one of its side walls at the setting of the moon at its major southern standstill (133°). The applied orientation grid at 45° / 135° forms a fair average, from which the true alignments can be derived - adding 2° for the northern standstill and subtracting 2° for the southern standstill. Bourbouillet 2 has a diagonal aligned at the major southern standstill of the moon. Apart from that, no alignments at the solstices, the standstills or at α-Crux our found.

    Assessment of the orientations

  • Sun and moon
  • Allowing a fault of 2° and in compliance with the above, a few alignments on the solstices and the standstills of the moon can be found. But the rising and setting of the sun in its winter solstice and summer solstice cannot be discerned from each other. The major southern standstill appears to be set at 140° persistently, while this is 2° till 4° too little. Heveskesklooster G5 and Langholz 68 are an exception with their orientation at 143°. With G5 this gives the only correct alignment on the major soutern standstill of the moon and with Spr. 68 this appears to be 4° too little again. When hunebedden are orientend towards each other, also the alignment at the minor southern standstill diverges, so that one must be doubtfull about the intention. With two sites (G5 and Schuldt 541) all four standstills are present as an orientation.

    Win. Sol = winter solstice, Sum. Sol = summer solstice,
    N. Mas = major nothern standstill, N. Mis = minor northern standstill,
    S. Mas = major southern standstill, S. Mis = minor southern standstill.

    Values with an underscore concern the setting of the sun or the moon.
    All lines diverging more then 2° are printed italic.

    MonumentOrientation line(s)Sum./Win. SolS. MasS. MisN. MasN. Mis
    Schuldt 540135°133°
    Schuldt 632132°, 140°132°144°
    G5143°, ±120°, 35°, 56°144°122°35°56°
    D25140°, 117°142°121°
    D40140°, 117°143°122°
    Spr. 68143°147°
    Schuldt 541141°, 121°, 32°, 58°144°122°33°56°
    Grammont 1133°, 47°133°47°
    Bourbouillet 2132°132°

  • α-Crux
  • In the introduction of this page the necessity is pointed out to compare the data of different times and latitudes for an alignment on a star. The mentioned sites vary from ±3500 BC in Schleswig-Holstein and Mecklenburg, via ±3200 BC in Drenthe to ±2200 BC in Soutern France. Therefore they meet this demand amply. Possibly supplemental data from the measurements by González-García in the eastern Rodopi in Bulgaria can be added to this survey. Amongst it the orientations of two dolmens near Tsarkvishte stand out. Also their orientations mirror each other around south with 162° and 198° [12.18]. This compares with the situation of the dolmens southerly of the French Central Massif. (A elaborate study is needed here.)

    Near latitude 53° the building agrees quite good with the accompanying rising of α-Crux. In Drenthe the starting date deviates with D22 and D40, because their construction took place half way through the time span in which all of the hunebedden were built. Because of the slant terrain with D21, D22 and D25, the disappearance of α-Crux doesn’t match the final phase of the megalith building.
    Near latitude 54° the alignments with Langholz 68 and Barkvieren 92 come over 100 years too early.
    Near latitude 46° till 42° the time of the alignments becomes unsure. The inclination of the horizon can vary rather much in the mountains. Assuming an inclination of +2° [12.19], at latitude 42° and 46° the megalith building and the occurence of α-Crux do agree again. At latitude 44° the megalith building comes 500 years earlier as excpected on the basis of the orientation on α-Crux.

    MonumentLatitudeRegionOrientation lineTime of α-Crux 1)Building time
    Schuldt 9254½°Mecklenburg193° - end stone3660-3360 BC3550-3200 BC
    Schuldt 54053½°Mecklenburg167° - side wall3540-3250 BC3550-3200 BC
    Schuldt 63253½°Mecklenburg167° - threshold3510-3210 BC3550-3200 BC
    G553½Drenthe170° - diagonal3330-3160 BC3300-3100 BC
    D2153°Drenthe163° - half-diagonal2)3370-2880 BC3300-3100 BC
    D2253°Drenthe167° - trilithon 2)3170-2880 BC3300-3100 BC
    D2553°Drenthe162° - grid 3)3300-2800 BC3300-3100 BC
    D2653°Drenthe167° - trilithon, etc. 4)3400-3100 BC3300-3100 BC
    D4052¾°Drenthe170° - side wall3200-3030 BC3300-3100 BC
    Less clear data:
    Bougon F46½°Southern France162° - short wall2650-2050 BC2600-1900 BC5)
    Pichouno±44°Southern France162° - end stone2100-1500 BC2600-1900 BC5)
    Bourbouillet±44°Southern France162° - end stone2100-1500 BC2600-1900 BC5)
    Tsarkvishte42°Rodopi, Bulgaria162° - side wakk1270-660 BC±1200-600 BC 6)

    1) Here the time is meant from the alignment on α-Crux till the moment that this star disappears under the horizon forever.
    2) With Bronneger D21 and D22 an inclination of -½° is taken into account.
    3) With Bronneger D25 an inclinatie of -1¼° is taken into account.
    4) With Drouwenerveld D26 an inclination of -0,1° is taken into account. Except for the eastern trilithons also the threshold and the end stone S2 stand at 167°.
    5) The building time of the dolmens souterly of the French Central Massif is based on carbon dates concerning the usage instead of the building of the monument [12.20]. A general inclination of +2° is taken into account, but this isn’t based on measurements [12.19].
    6) The dolmens in the eastern Rodopi mountains seem to be built during the archaic time of the Tracians [12.21]


    The orientation on the soltices are quite accurate, but aren’t found so much. Because of the configuration with Z1’a and Z1’b, it’s almost sure, that people planned an alignment in Bronneger D22. Probably hunebed D21 has an alignment too via its gate. Towards the orientation lines of two dolmens in Mecklenburg it can be mentioned only, that alignments by means of a diagonal or end stone occur frequently.

    Orientations on the standstills of the moon are more common. Although the orientation lines pointing at the southern standstill mostly run a few degrees northerly, they seem to be associated with it nevertheless. Especially the construction with alignments on all four standstills with Serrahn 541, while the major southern standstill runs 3° northerly also, pleads for it. The deviation should be taken for granted probably. From the situation with Emmen D39 and D40 it even can be concluded, that people created these deviations deliberately. On account of its whimsical movement along the sky, a less exact alignment on the moon doesn’t need to suprise.

    Alignments on α-Crux are doubtfull. With various monuments from diverse times and latitudes the orientation lines agree with the rising of this star, but there seem to exist rather great deviations, which are clustered in the regions at latitude 44°. Only with the concerning Dutch hunebedden the true building time was tracked down and for them an association could be confirmed. Furthermore it can be noticed, that the time between the first orientation on α-Crux and the moment of its disappearing forever, matches the duration of the megalith building per region quite good. But the time span itself lies shifted now and then. {Revision}

    Lastly a remark of common nature. Possibly the ideal construction was more important then a true alignment. This doesn’t need to surprise, if we think of the fact, that with a north/south movement of 15 km or a time span of 30 years only, the rising of α-Crux moves 1 along the horizon already. For the rising of the sun and moon a shift of 1° per 100 km accounts for latitudes around 53°. The limitedness of a direct alignment on this star would become visible within a human life. Moreover, considering the practical sense, it would be exaggerating to have all monuments near each other equiped with the same alignment. Maybe, for the construction of orientation patterns, one should pay more attention to symbolism then to its alignments. The orientation pattern could be seen as the fossile reproduction of an initial setting, that has been developped in some kind of 'centre' and which spreaded out without changing the construction method. One can think of the movement of groups but also of a 'school' of experts, who applied the same construction techniques in a wide spread area. The preference for one of the two depends on how close or diffuse the culture elements have to be seen (alternation of cultures versus mutual influences). Towards the orientation on the moon a northward influence is most obvious, since it can explain the alignments on the minor southern standstill, which point a few degrees too far northerly. From the chart with the orientation on α-Crux a southward movement should be concluded then.