Using the Moon as a Proxy for the Ancient Sun

When it comes to analysing potential alignments at Stonehenge, there's one problem that always gets in the way. Well, there are several actually but the one I'm going to describe here is pretty fundamental.

The Sun, today, does not rise and set at the solstices at the same positions on the horizon that it did when Stonehenge was built.

This is due to an oddity in the way the Earth's rotational axis behaves. We're all familiar with the idea that Earth's axis is tilted away from the vertical at an angle of about 23.5° - it's in all the diagrams of the planet, it's there in every globe you've ever seen on a desk and it's why we have any seasons at all.

The archetypal desk globe with the planet's characteristic 23.5° tilt

We might imagine that it's always been 23.5°, but it hasn't. Over very long periods of around 41,000 years this angle varies from between 24.2° and 22.5° and - we assume - back again. This is called the "Change in the Obliquity of the Ecliptic" (but don't worry about terminology for now).

4,500 years ago, when Stonehenge was built, the angle of tilt was roughly 24° give or take a couple of minutes of arc. As a result the Sun at summer solstice rose and set fractionally further round to the north than it does today, and at winter solstice it rose and set fractionally round to the south than today.

The different isn't much - about a degree or so, but it really affects observations of solstice rises and sets at Stonehenge.

For example, today the Sun at summer solstice rises out of the tip of the Heelstone, but in 2,500 BCE it rose off to the left of it - as seen from the main axis of the monument between Stones 30 and 1 on the NE side.

Summer Solstice Sunrise position today and 4,500 years ago

What this means is that there is no way to make direct observations of the Sun at the solstices as it would have been seen by the builders of the monument. Instead, we have to calculate its prehistoric position (which we can do to an excellent degree of accuracy) and create mockups like the photo above to visualise how it would have looked.

(Incidentally, if you want more about what would have happened and what you would have seen at summer solstice sunrise back in the day then take a look at the article on this site called "The Shadow of the Heelstone")

Frustrating! Except there's a way around the problem thanks to the movements of the Moon.

The Sun's rising position on the eastern horizon varies throughout the year. In summer it's in the NE then it turns around at solstice and starts heading south. By the time of the winter solstice it's reached the SE before it turns back and begins its journey northwards again.

Back and forth, regular as the pendulum of a clock, and in a human lifetime the endpoints of this swing (and the corresponding ones on the western horizon for sunsets) are effectively fixed.

The Sun's annual swing along the eastern horizon from Summer Solstice (left) to Winter Solstice (right) and back

The Moon also swings back and forth along the horizon between endpoints, but it does it every month. Each month there's a northernmost limit for rising (and setting) and two weeks later a southernmost limit. NE to SE to NE to SE to NE to SE.... tick tick tick.

What's different about the Moon is that the endpoints are not fixed like those of the Sun. The furthest north and south positions move back and forth as well. Every 18.6 years these endpoints get as far apart on the eastern horizon (for risings) as they ever can do, and the Moon can rise and set way further north (and south) than the Sun can ever reach in our era.

In the animation above, the yellow diagonal lines on the far left and right are the endpoints of the Moon's swing when it reaches its maximum extent. See how they're further apart than the Sun's endpoints (the orange diagonal lines)? The one in the middle marks the Equinox.

We're at that point in the Moon's cycle now, in 2024/25. It's called the Major Lunar Standstill. It's a poor label since it's not the same sort of "standstill" than the Sun experiences at the solstices, but we're stuck with the name. Never mind.

As always the Moon still keeps swinging along the horizon from day to day over the course of a month so it can rise anywhere between the endpoints of the swing. It's just that the endpoints are further north and south than the Sun's can ever be today. I know I just said that, but it's a critical point!

So what does this mean?

It means that if you pick your moment, you can find the Moon rising and setting at the points on the horizon where the Sun of 4,500 years ago did, so you can therefore use the Moon as a "proxy" for that ancient Sun and can directly observe how things would have looked when Stonehenge was built.

The only difference is that it's the Moon you're looking at rather than the Sun. Happily, they both appear the same size in the sky and it's less damaging to the eye to stare at the Moon :-)

Back in the late 1990s/early 2000s, Prof. Gordon Freeman of the University of Alberta - who'd spent years doing direct observations on site at Stonehenge of solstices - proposed a remarkable idea.

He suggested there was a secondary solstice axis that ran from Winter Solstice Sunrise to Summer Solstice Sunset and that there was a deliberate sightline through the stones that made use of a notch in the edge of Stone 58 coupled with the edge of Stone 53 on the opposite side of the monument. This combination created a "peephole" of sorts that was directed at Coneybury Hill where the ancient Winter Solstice Sun would have risen.

View through the Notch in Stone 58 whose open side is closed by the edge of Stone 53
to frame a very specific spot on Coneybury Hill

(For more detail on this idea, see the article on this site called "The Secondary Solstice Axis")

On July 21st 2024, the Moon rose very close to the same position that the ancient Winter Solstice Sun once did, so here was an opportunity to attempt a direct observation of Gordon's idea - was his sightline valid? By calculation, it definitely is but a chance to observe the Moon proxying for the Sun along this line was not to be missed.

Unfortunately I wasn't able to gain permission to do the observation from within the monument itself, although if all goes well I might be able to do so during one of the next times things align in the same way, so I had to try and get the data from a position out in the landscape on the projection of the alignment to the NW up towards the Cursus Barrows.

The horizon was perfectly clear and at 21:55:42 BST (20:55:42 UT) the first gleam of the Moon's upper limb appeared behind Stonehenge.

First gleam of the Moon - that tiny orange blob just left of centre on the horizon

By 21:58:52 BST, it had reached half-orb and I was getting excited.

3m 10s after first gleam things are looking promising

A little under four minutes later, the alignment fell into place.

The Moon acting as proxy for the ancient Winter Solstice Sun

It is extremely difficult to get into exactly the right spot for a shot like this - I don't have a differential GPS system to give me sub-centimeter accuracy, so I had to rely on eyeballing the monument in daylight through binoculars to get into the best position I could, so I may be a couple of metres to the left or the right of the ideal spot.

I'm certainly higher up in elevation than I would be at the monument, so the Moon appears above the horizon from this vantage point whereas it'd be resting on it otherwise.

Nevertheless, this is an excellent result and shows that it's definitely worth repeating the observation from within Stonehenge itself.

I did notice other photographers around - one actually on the visitor path next to Stonehenge - but none of them were in the right spot to get the desired (for the purposes of checking the alignment) photo. I'm sure they got fabulous images of the Full Moonrise with these hoary old stones beneath, and I look forward to seeing them.

My photos are somewhat grainy (old camera), but they're worth far more to me - and potentially to future researchers of the ancient astronomical sightlines at Stonehenge - than a pretty shot of Moonrise.

Well done for spotting this one in the first place Gordon, you are an inspiration.

Major Lunar Standstill July 19th 2024

At Moonrise on the 19th July 2024, the Moon was very close (within a quarter of a degree) to its maximum southerly declination in its 18.61 regression of the lunar nodes cycle.

I'd booked to go in to Stonehenge on the late evening "Special Access" session, in order to try and improve on my photo from July 9th 2006 of the Moonrise in line with the northeastern side of the Station Stone rectangle.

July 9th 2006 Major Lunar Standstill - 18 years and 10 days ago

(In the mid-1960s, Peter Newham and Gerald Hawkins had independently suggested that the long sides of the Station Stone rectangle were aligned to the southernmost moonrise and northernmost moonset at the lunar standstill - see this article for the background to Newham's work.)

In the intervening 18+ years, English Heritage have placed a marker at the approximate position of where Station Stone 94 once stood (the Station Stone is no longer present, but its stonehole has been found through archaeology). There is a slight inaccuracy in its position (perhaps half a meter off), but it serves as a useful reference.

Moonrise was timed for 20:41 BST (19:41 UT) but the trees of Luxemborough Plantation on the SE horizon delay the Moon's appearance by around 5 minutes.

I set up my camera so that I was in line with the marker for SS94 and the (still present, though slumped) Station Stone 91 on the opposite side of the monument.

Initial photo on the alignment from SS94 marker to SS91.
Features of other visitors pixellated for privacy.

I knew I was further to the southwest of my 2006 position directly over SS94 (a spot which I'd then chosen by estimating the centre of the area inscribed by the bank of the North Barrow), so was expecting the Moon's position at appearance to be somewhat different to then, relative to the treeline.

With the camera taking a shot automatically every 15 seconds, I spent the next few minutes over the rope a couple of metres to the NE to try and spot the first gleam of the 97% waxing gibbous Moon over the treetops. Fortunately, there was no cloud cover at the horizon at all - which wasn't the case back in 2006.

And then, at 20:46:55, I spotted it peeping out of Luxemborough Plantation's trees. Rats! It was too far south (to the right) for my camera to have caught it - and would be blocked by the large sarsens of the NE side of the outer circle. I grabbed the tripod and quickly moved the camera to catch the Moonrise.

Off-alignment by about 2m to the NE with the Moon's upper limb just appearing.
The shot has been contrast enhanced to improve visibility.

OK, so this is somewhat annoying! First gleam over the true horizon would definitely have been visible from my original spot if there were no trees in the way. We really ought to cut a path through Luxemborough Plantation to restore this sightline.

Back home, I decided that it would a useful exercise to montage these two shots using the treeline as the reference to see what it would have looked like in an ideal world. The fractional parallax introduced by moving 2m NE off-alignment is tiny since the trees are a considerable distance away, so it's worth doing.

Here's the result.

Montage of original shot in the alignment position with the off-alignment photo,
using the treeline as the registration reference.

All very well, but how can I tell where the horizon (sans trees) would be? I can guess, but long experience of carrying out observations at Stonehenge has taught me that subtleties in both the shapes of the stones and the horizon profile are important and need to be considered.

Happily, due to the work of David Hoyle (www.standingstones.org) there is an excellent LIDAR/DTM terrain model for the Stonehenge landscape which can be loaded into Stellarium to give a pretty accurate representation of the actual horizon profile.

This is Stellarium's view of 20:46:55 BST on the 19th July 2024:

Stellarium view of Moonrise at the same instant as the first appearance of the Moon over the trees.
The Archaeolines plug-in is being used to show the arcs of the Major Standstill Moon.

A further montage, using the Moon size and the hill at 135° azimuth as reference points, allows me to see the whole picture.

Final montage with true horizon profile and rising arc of Major Lunar Standstill.
My yellow block for SS91 may be a tad short - it was a quick guess for visualisation.

At first glance, this looks excellent - the Major Lunar Standstill southernmost Moonrise appears as if it will emerge from the intersection point of the tip of Station Stone 91 and the true horizon. I suspect my yellow block is too short, but it was a rough positional indicator I plonked in while doing these montages. It's quite a long stone (3m or so) that's slumped right over and is resting on the earthwork bank.

Remember that the Moon on the 19th July 2024 is not precisely at its southernmost extreme declination, as is evidenced by the fact that Stellarium doesn't have it centred on the green rising arc lines.

However - there's one more factor we need to consider and that's the change in the Obliquity of the Ecliptic.

Earth's rotational axis is presently tilted over at roughly 23.5° to the plane of the Earth's orbit - which is why we have seasons (and indeed solstices that mark the turning points in the year). Back when Stonehenge was built, the tilt was 24°. This additional 0.5° tilt has an impact on the rising azimuths of the Sun and Moon at their extreme north and south limits.

In this specific case, 4,500 years ago when the large sarsens at Stonehenge were erected (or 5,000 years ago for the Station Stones - that's another story!) the Moon would have appeared to rise a further 1° to the right of where we are seeing it in these photos - that's two Moon diameters.

The implication of this is serious.

It means that the northeastern edge (long side) of the Station Stone rectangle is not precisely aligned with the southernmost possible Moonrise position - by something like 1.5° to 2° in azimuth based on these photos.

If we also factor in the potential inaccuracy of the position of the marker for SS94 it gets worse still - we may be looking at an error in alignment of up to 2.5° or 5 lunar diameters!

Newham and Hawkins' suggestion that the Station Stone rectangle's long sides are exactly pointed at the southernmost moonrise (SE) and northernmost moonset (NW) is starting to seem a little off.

We know, as modern astronomers, that the chances of catching the Moon rising or setting exactly at its extremest possible declination north or south of the celestial equator are very slight. The Moon races around its orbit (and hence our sky) really quickly so everything has to come together - orbitally and weatherly - for a precise observation of the extreme to be done. That may only be possible once in a generation.

What all this serves to show is that we need to devote proper resources to researching what is actually seen at Stonehenge during these occasions of rare potential astronomical alignments.

We need to be absolutely sure that if we're going to put modern markers in the ground, that they are accurately positioned which implies confirming earlier work about stonehole positions though new archaeological digs with modern DGPS instead of relying on 1950s surveys and interpretation.

And we need to seriously consider restoring sightlines in key directions by selectively felling some trees that obscure the true horizon. Luxemborough Plantation (SE), Larkhill (NE), Normanton Gorse (SW) and Fargo Wood (NW) will all need attention.

The questions that are raised here deserve further investigation. We have the chance over the next 12 months to refine my observations as we progress though this Major Lunar Standstill season (2025 is the key year).

Or not. In which case, we'll have to wait another 18.61 years for the next opportunity to do so.

Thoughts on Bluestone Trilithons

Within the collection of bluestones at Stonehenge are several interesting examples that show distinct signs of having been part of independent, tooled, structures.

Stone 36 and Stone 150 are clearly lintels having mortise holes worked into one of each of their faces, in much the same way that the lintels of the outer sarsen circle and inner horseshoe of sarsen trilithons do.

Stone 150 is quite rounded, and lies prone in the turf in the NE quadrant of the bluestone circle but Stone 36 is far more elegant and is almost entirely buried in the southern quadrant - in fact 36 is arguably the finest dressed stone on the site.

Stone 150

Stone 36

Stone 36, having been lifted for inspection during the excavations of 1954

Stonehenge plan showing the positions of Stones 36 and 150
© Anthony Johnson, annotations by Simon Banton
https://creativecommons.org/licenses/by/3.0/deed.en

As Julian Richards writes, in "Stonehenge - The Story So Far (2nd ed.)", referring to Atkinson's lifting of Stone 36 in 1954:

Apart from its sheer aesthetic qualities, Stone 36 posed some interesting questions. Along with stone 150 it was the second bluestone lintel to be identified and there were the uprights to go with them, pillars in the bluestone horseshoe that showed signs of having originally had tenons. So Atkinson could suggest a phase of ‘tooled bluestones’ which must, as there were now two lintels, have included at least two miniature trilithons. These, on the evidence of the mortice holes in the lintels would have looked very different from the much larger sarsen examples. The lintels would have extended beyond the edges of the slender pillars on which they perched, the space between the pillars and the height of the lintel sufficient to allow people to pass through.  The remarkable Stone 36 also provided evidence that this was not a short-lived structure. One of its mortice holes was surrounded by a shallow depression, presumably a carefully worked seating for the upright on which it sat, and within this hollow the surface of the stone appeared worn, even polished. This did not appear to be deliberate but more the result of friction, perhaps caused by the expansion and contraction of the touching stones. But such a polish would only develop very slowly, suggesting that these stones must have stood as trilithons for many years.

Could these two bluestone miniature trilithons have been the archetypes that eventually gave rise to much larger echoes of similar design in the enormous sarsen trilithons?

If so, then what could have been the intent behind the original creation of the bluestone versions? And where were they erected? Clearly not close to the current positions of 36 and 150 since they are remote from their supposed companion uprights which are components of the inner bluestone horseshoe arrangement and which carry the battered down remnants of tenons on their upper surfaces.

These are all repurposed stones - the bluestones have been rearranged a number of times in prehistory - and 150 in particular has been used, finally, as a pillar of the bluestone circle which was oriented so that its mortise holes would not be visible from the interior of the monument.

Bluestones have a remarkable capacity for being used as lithophones - "rock gongs" if you like. When found at their outcrops in the Preseli Hills in South West Wales, experimentally striking them with a rubber mallet or a deer antler will soon discover that certain ones that will "ring" like a bell.

They have to be positioned just right - balanced without being buried in the ground so that they can resonate when struck. The following video shows my attempts at getting a note out of a number of examples in 2023.

If there were two bluestone trilithons at some point in Stonehenge (transported there having been dismantled from an earlier monument in Preseli, perhaps), could the design have been in an effort to support two appropriate stones above the ground so that they could be easily "rung"?

What are the implications of such a suggestion? Is there an acoustic aspect to the use of Stonehenge that has been hinted at by a number of other researchers? One which employed the bluestones?

Are the bluestone trilithons in fact engineered sound sources? Maybe the "voices" of the stones were thought to have been in some way special in their own right, and bringing them to sing at the site of the future Stonehenge was important for an unknowable (to us) reason.

Given that the final arrangement of the bluestones at Stonehenge has them all either fallen, half-buried, or deeply embedded in the chalk (in the case of the upright ones) it seems that if they did have an original acoustic purpose then this was subsequently either forgotten or discounted by the people who incorporated them into the much later monument - even if they did echo their form in the horseshoe of enormous sarsen trilithons.

Stone 11 Shadow Effect

Stone 11 is the mysteriously half-height, half-width upright in the southern quadrant of the outer sarsen circle at Stonehenge.

Plan of Stonehenge indicating position of Stone 11

Stone 11 seen from the south

Several suggestions have been made to explain its curious dimensions - that the builders ran out of bigger stones, that it's a reused lintel, that it marks a division of the 30 uprights into three groups of 10 (the other marker being Stone 21) for calendrical reasons...

The 2012 Antiquity paper "Stonehenge Remodelled" (Darvill, T. et al. (2012) ‘Stonehenge remodelled’, Antiquity, 86(334), pp. 1021–1040. doi:10.1017/S0003598X00048225) offers:

...stone 11 (south) is narrower and shorter than the others perhaps to somehow mark the southern entrance (or it may even have been a later replacement).

None of those explanations have ever felt entirely satisfactory to me.

Recently I was watching Dr. Terence Meaden's video presentation to the 2024 Megalithomania Conference, on the topic of Pytheas the Greek's voyage to Britain and Thule.

Terence is always interesting and has had enlightening things to say about megalithic structures for decades. It was he that pointed out, for instance, that the Heelstone casts a shadow that penetrates into the centre of Stonehenge on the Summer Solstice (see "The Shadow of the Heelstone" on this site for more info).

While he was explaining his hypothesis that Stone 11 is of reduced height to represent "half" in the count of sarsens in the outer circle - ie that the circle conceptually held 29.5 stones rather than 30, and so could be used as a lunar calendar count, twelve times round it making 354 days then continuing the count on to Stone 11 would bring the day tally up to 365 and hence one solar year - he put up this diagram from his 2017 paper "Stonehenge and Avebury: Megalithic shadow casting at the solstices at sunrise":

Fig. 7 from "Stonehenge and Avebury: Megalithic shadow casting at the solstices at sunrise"
Journal of Lithic Studies (2017) vol. 4, nr. 4, p. 39-66 doi:10.2218/jls.v4i4.1920

Terence points out that the shadow of Stone 11 falls on Stone 40 at Winter Solstice Sunrise, and suggests that the male Stone 11 deliberately interacts with the female Stone 40 in this way.

Those familiar with his work will recognise this as a common theme in his research - phallic upright stones casting shadows onto recumbent receiving stones at key times in the year.

The next image in his presentation gave me goosebumps.

Stone 11 just after sunrise near Winter Solstice with the Sun at its rising azimuth of 2500 BCE

The shadow of Stone 11 can clearly be seen clipping the fallen Stone 14's right hand end (its upper end when erect), and Stone 40 (a bluestone) is directly behind this fallen sarsen lying prone in the turf - not visible in this shot, exactly under Stone 11's shadow. But that's not what grabbed me.

The shadow of Stone 11 continues across the ground and directly intersects with Station Stone 93! This contrast-enhanced closeup of Terence's image shows it well.

Stone 11's shadow hitting Station Stone 93

For this to happen, Stone 11 has to be a very particular height and width - any wider and its shadow will not match the width of Station Stone 93, any taller and the shadow will extend well beyond 93's peak.

We have to remember to account for the shift in the Earth's axial tilt from 24° to 23.5° since 2500 BCE. 

At winter solstice 2500 BCE, the Sun reached "full orb" over Coneybury Hill at an azimuth of 130° 20':

Full orb winter solstice Sun position, 2500 BCE

In our era, on the 27th December 2014 (when Terence's photo was taken), the Sun is at a higher altitude when it reaches this azimuth:

Sun's position on 27th December 2014 when it reaches azimuth 130° 20'

Its altitude in our era is 1° 6', back in 2500 BCE it was 0° 26' - a difference of 40' of arc. This difference puts the tip of Stone 11's shadow fractionally lower on Stone 93 now than it would have been then.

It seems likely that Station Stone 93 would have been completely and exactly engulfed by the shadow of Stone 11 cast by the full orb risen Sun at winter solstice in 2500 BCE.

Drawing this shadow line on the reference plan of Stonehenge from "Stonehenge in its Landscape" (Cleal et al, 1995), along with the secondary solstice axis identified originally by Gordon Freeman (see "The Secondary Solstice Axis" on this site) highlights their parallel nature.

Stone 11 shadow (black line) and the secondary solstice alignment (red line)

Zooming in on this, and incorporating the plan of the parchmarks for the missing stones 17, 18, 19 and 20 from "Parchmarks at Stonehenge, July 2013" (Banton, S., Bowden, M., Daw, T., Grady, D., & Soutar, S. (2014) Antiquity, 88(341), 733-739 https://doi.org/10.1017/S0003598X00050651) reveals an elegant aspect to this shadow line.

Closeup incorporating the plan of the parchmarks identifying the positions of Stones 17, 18, 19 and 20

... the shadow of Stone 11 (the black line) passes exactly through the gap where missing stone 18 and 19 would have stood (if indeed, they'd ever been erected - the jury's out on that one) - so it would not have been blocked by them.

In the end what does this mean?

It appears possible that Stone 11 was positioned and shaped to create a shadowplay effect at winter solstice when Stonehenge was built, one that targeted Station Stone 93 very precisely both in position and size.

If Stone 11 was fatter and taller, this wouldn't have worked so neatly.

All that remains to be explained is the leaning nature of Stone 11, because it is not perfectly upright.

Stone 11 leaning towards the south

... how does this affect things? Is it intentional? I'm looking at this picture and noticing that the right hand edge of the stone is almost perfectly vertical, compensating for the lean.

The investigation into that is a story for another day.

Many thanks to Terence for giving me permission to use his images.