I have no idea if there is any evidence or even informed speculation about this, so I thought I'd ask the community most likely to know.

Megalithic and other prehistoric structures are widely recognized as designed to align with say, sunrise on a solstice. But how did the builders know for sure that the day in question actually was the shortest or longest day, or for the equinoxes, that day and night were equal in length?

Is there any evidence, or even just reasonable hypotheses, for any kind of early technology that could actually measure the duration of sunrise to sunset from many thousands of years ago?

  • 1
    This is a fascinating question! Equinoxes are pretty hard even with clocks (the sun's light is refracted by the atmosphere, so sunset-sunrise is less than sunrise to sunset on the equinox, and at the solstices the length of day and sun position is changing most slowly, making them hard, too.
    – Mark Olson
    Mar 28 at 19:26
  • 1
    This is totally on topic here. However, I bet they'd love this question over on History of Science and Math as well.
    – T.E.D.
    Mar 28 at 20:07
  • Thanks, I'll cross-post it there.
    – Andrew
    Mar 29 at 16:18

2 Answers 2


In a sunny climate, it's easy to determine the solstices with a gnomon; a monolith or obelisk can be used for this purpose. At the winter solstice the gnomon's shadow is at its longest, at the summer solstice the shadow is shortest.

The equinoxes can also be found using gnomon since its shadow traces a straight line, from west to east, all day - and only on that day. Path of the Sun's shadow at an equinox Image from Earth and Sky

Earth and Sky have a very interesting article about the paths that shadows trace during the year.

If you plot the position of the Sun's shadow at the same time every day, say you put a stick in the ground when the gnomon's shadow is longest, you will get an analemma which is also useful for determining solstices and equinoxes.

Analemma from the shadow of a gnomon

Image from a very interesting article at searchresearch1.blogspot.com

Of course it isn't sunny everywhere all the time, but knowing the time of year would have been vital for the survival of neolithic people. It's likely this sort of knowledge would have been passed from generation to generation for thousands of years.


Mark Olson raises some good points about the difficulties in getting accurate measurements from these methods.

I had a look at the Sun's elevation above the horizon on the winter solstice of 2024 using NOAA solar elevation calculator for the location of the Long Meg standing stone in Cumbria. Long Meg is 3.7 metres tall, and using basic trigonometry I found the length of the shadow, at midday, in millimetres:

Length of standing stone shadows at winter solstice

In mid-winter the Sun's penumbra is at its greatest, but given that the penumbra of the Samrat Yantra sundial at Jantar Mantar in Jaiput is 30mm for a gnomon height of 27 metres (over seven times the height of Long Meg), I think it's safe to assume the penumbra in Cumbria is of the order of a few millimetres.

So if the sky had been cloudless last December, someone marking the length of the shadow of Long Meg on a cleared, flat piece of ground, would have been pretty sure that the solstice occurred between the 19th and the 26th with the mid-point being a good guess.

It's unlikely that the sky was cloudless. The nearest historical data I could find was for Barrow-in-Furness on Weather Underground. This shows high humidity for the whole of that period, with some precipitation every day. Barrow is not a great proxy for Penrith (the nearest town to the standing stones) since a range of mountains separate them, but December in England is often (but not always) wet and cloudy. While the climate in the neolithic was warmer than now, it was also wetter.

However, Stonehenge was in use from around 3000 BCE to 1600 BCE, so I think it's fair to say that, over the generations, a reliable and accurate calendar could easily have been constructed by the people responsible for the neolithic monuments of north eastern Europe.

  • 1
    It's a lot less easy than you think! The rate of change of shadow length is close to zero at the solstices, making determination of the shortest hard. The gnomon has to be absolutely stable over the course of a year (easy enough today; not so much in neolithic times.) The ground on which the shadow falls also needs to be very, very stable from year to year. The gnomon needs to be pointy compared with the amount of noon-to-noon shadow movement you hope to observe (hard to do with rocks, the only real stable thing available). And in winter, especially, it often cloudy (not to mention snowy).
    – Mark Olson
    Mar 29 at 21:17
  • 1
    And you don't need the calendar to be more accurate than about +/- two weeks for anything they actually needed it for -- most things like sowing seed would be done by looking at the world around them and judging how wild plants are sprouting and by how warm the soil is.
    – Mark Olson
    Mar 29 at 21:20
  • OP here: It also begs the question "How would you know that it's noon?" Unless you're on the equator, in which case the sun should be bang overhead (I think!) But, for my purposes (which are to provide a degree of authenticity in a novel set around 8,500 years ago in Doggerland, so long before Stonehenge or any other megaliths that apparently were constructed with architecture based on the heavens) all these discussions are proving immensely helpful. I'm very grateful to all the commenters.
    – Andrew
    Mar 30 at 2:17
  • Oh and in additional comment, given that my concerns are set several thousand years before Stonehenge et al, I've a feeling that allowed more than enough time for people wanting to know where to be at the same time as their sources of food to have refined their astronomical skills to the point that they could reliably put up something as consistently accurate as Stonehenge.
    – Andrew
    Mar 30 at 2:22

By the location of sunrise/sunset

Looking at Stonehenge, we see:

The site, specifically the great trilithon, the encompassing horseshoe arrangement of the five central trilithons, the heel stone, and the embanked avenue, are aligned to the sunset of the winter solstice and the opposing sunrise of the summer solstice.

Also consider this picture (source): Picture of the sun for equinox and solstices We can see that the location of sunrise/sunset changes over the year, and even in prehistoric times they could note that as sunrise/sunset moved further south, the days (northern hemisphere) got shorter. And then as sunrise/sunset moved northward, then the days got longer. So it would be fairly easy to conclude that the day the sunset was the most southerly is most likely the shortest day of the year (winter solstice), and same for the summer solstice. This would not determine the exact time of day, but knowing the date (even if in error by a day or so) would be sufficient for festivals and most other events.

For the equinox, if you have a good idea of east/west, then when the sun rises due east, then you are at an equinox.

Also since these are equally spaced, if you have a good way to determine the solstices, you can calculate the equinoxes, or vice versa. The calculations would involve some error, but again they probably weren't worried about the exact time of the solstice or equinox.

Water clocks existed, but would not have been accurate enough to determine the solstice or equinox to the level of accuracy we have now. In short, they didn't use clocks like we do now, but could use other means to determine when the solstices and equinoxes occurred.

  • 3
    "since these are equally spaced, if you have a good way to determine the solstices" - wrong. Spring has currently 92.8 days, summer 93.6, autumn 89.8 and winter 89. The orbit of the earth around the sun is an ellipse, not a circle, and with the sun positioned in one focus, not the center. Since the earth runs faster the nearer it is to the sun, 3.5% difference in distance make for four days in the length of a quarter year. (Only currently. Earth precession means that in 1246, all four seasons had the same length.)
    – ccprog
    Mar 28 at 23:28
  • Does it mean that Stonehedge's structures are "simply" aligned with the southernmost sunrise and the northermost sunset, something human could define even if they couldn't calculate which day is the solstice ? And about equinoxes, I have no clue how prehistoric human could "have a good idea of east/west".
    – Evargalo
    Mar 29 at 13:24
  • All interesting, but still, it doesn't give any indication HOW! That's the interesting bit!
    – Andrew
    Mar 30 at 2:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.