ObserveTheSky › Explanations › Rung 1 · The size of the Earth
Rung 01 · ~240 BCE · Alexandria, Egypt

The size of the Earth

Eratosthenes · a stick, a shadow, and one summer noon

In Syene, far up the Nile (today’s Aswan), something odd happened once a year. At noon on the longest day, an upright stick cast no shadow at all. The Sun stood straight overhead. A later writer tells of a well there into which the noon Sun shone right down to the water, however deep.

In Alexandria, on the coast, a stick at noon on the same day did cast a shadow. Not a long one, but a shadow.

Eratosthenes, in Alexandria, knew both facts. He also had good reason to take the Earth for a ball: when the Moon passes through the Earth’s shadow, the edge of that shadow on the Moon is always round. One Sun. 2 places. 2 different shadows. History of thinking: Aristotle’s evidence for a ball.

So how big is the Earth?

Source

Cleomedes, The Heavens 1.7; the well, Strabo, Geography 17.1.48.

Move the Sun: one picture, 3 things to notice

Drag the Sun, or use the slider. (1) When the stick at Syene has no shadow, what does the stick at Alexandria do? (2) Flatten the Earth: what happens to the 2 shadows now? (3) And what if the Sun were close?
  • the Sun
  • its rays, all parallel
  • a stick, at Alexandria and at Syene
  • a stick’s shadow
  • the Earth
  • the break line: most of the distance cut out

The slider for where the Sun is at noon runs through the year. On the longest day the noon Sun stands straight over Syene. On every other day it stands lower, further to the south, and on the shortest day it is about 47° further south. It never goes further, so neither does the slider.

The question: answer it, and you have the method

The Sun is so far away that its rays reach both cities as parallel lines. (How far away, rung 4 measures. What a near Sun would do, you tried above, with A flat Earth instead and the slider for how far away the Sun is; What works and why comes back to it.)

Take a pencil and paper. Draw the round Earth, the 2 cities, a stick at each and the Sun’s parallel rays. Then think about the question before you open a hint.

The angle at the top of the Alexandria stick, between the stick and the Sun’s ray past its top: where else in the picture of the whole Earth does that same angle appear?

Hint 1

Draw a line from each city down to the centre of the Earth.

  • its rays, all parallel
  • the ray past the top of the stick
  • the 2 sticks
  • a line from each stick down to the centre
  • the Earth
  • the centre of the Earth
Hint 2

A straight line crossing 2 parallel lines makes equal angles with both. Turn the crossing line by its end, or move it up and down by its middle: the 2 marked angles change together.

  • 2 parallel rays
  • a line crossing both: turn it by its end, move it by its middle
  • the 2 marked angles

Your idea

The orange wedge is a copy of the angle at the top of the Alexandria stick, between the stick and the Sun’s ray. Pick it up by its corner and move it; turn it by the end of its arm (or select a handle and use the arrow keys). Where else on the Earth does it fit? Here the cities are drawn further apart than they are, so the angle is big enough to handle.

  • its rays, all parallel
  • the ray past the top of the stick
  • the 2 sticks
  • a line from each city to the centre
  • the angle at the top of the Alexandria stick, and the copy you move
  • the Earth

And then: if that angle fits into a full circle some number of times, what does the walk between the 2 cities, the distance along the ground, fit into?

Hint 3

Every time the angle fits into the full circle, it cuts off the same piece of the Earth’s surface: one walk from Alexandria to Syene. Lay the angle round the centre, again and again, and count until the circle closes.

Drawn with the cities 15° apart, as above. You measure the real angle next; come back then and count.
  • the angle, laid round the centre again and again
  • the piece of the surface each one cuts off: one walk between the cities
  • the Earth

Measure it: the Alexandria stick, drawn to scale

Now measure the angle the question is about. Here is the stick in Alexandria at noon on the longest day: 2 metres tall, drawn to scale. Use the ruler and the protractor. Drag their round handles, or select one and use the arrow keys.

  • Lay the ruler (blue) along the shadow, from the foot of the stick to the tip of the shadow. How long is it?
  • Put the protractor’s (purple) corner on the top of the stick, one arm down the stick, the other along the ray to the shadow’s tip. What angle is that?
  • the Sun’s rays, all parallel
  • the ray past the top of the stick, to the tip of its shadow
  • the stick, 2 m
  • its shadow
  • the ground

Why a flat triangle is enough. On a ball, the direction of straight up turns as you walk, by an angle in proportion to how far you go. Between the 2 cities it turns by a few degrees. But the walk between them (about 800 km, given below) is millions of times longer than a shadow of a few tens of centimetres, so under the shadow straight up turns by only a few millionths of a degree, nothing any protractor could see. For the stick and its shadow, the ground is flat and the triangle is a true right-angled triangle. The Earth’s curve shows only between the 2 cities, and that is where it is used.

Do you know Pythagoras? The 3 sides make a right-angled triangle: the stick, the shadow, and the ray from the top of the stick to the tip of the shadow. That long side should be √(stick² + shadow²). Measure it with the ruler and check. If it comes out right, your drawing really is a right-angled triangle. You do not need Pythagoras for the angle; it only checks the drawing.

Work it out :fill in what you can; an empty step is fine

Given: Syene and Alexandria are 5,000 stadia apart. That is the distance the measurement used. Cleomedes, whose lectures tell the story, gives the number but not how it was found.

Definition: the stadion. A Greek length of 600 Greek feet. The foot differed from city to city, and the stadion with it: the track at Olympia is about 192 metres, and the common (Attic) stadion was about 185 metres. Whether the 5,000 were counted in it, or in a shorter one, is argued about to this day.

Because nobody knows for certain which stadion was meant, the steps use 800 km for the walk. It is close to the real distance: on a map today, Alexandria lies 788 km north of Aswan; the round-Earth page counts 790.6 km, on a perfect ball. (What moved it since, below, shows what each choice does to the answer.) The stick is 2 m.

Each stick is read at its own noon.

Definition: local noon. Noon as the round-Earth page defined it: the moment the Sun stands highest in a place’s sky, due south of anyone north of the tropics. A place further east has its noon earlier, a place further west later.

Alexandria lies only about 3° west of Syene, so its noon comes some 12 minutes later (the Sun moves a degree west every 4 minutes). And on the longest day the noon Sun’s height hardly changes from one day to the next, let alone in 12 minutes. So the 2 noon shadows can be set side by side as if they were cast at one moment.

2 words for the steps, the first met already on the round-Earth page:

Defined earlier on the ladder: the tangent.

Definition: π (pi). For every circle, big or small, the way round divided by the distance across is the same number, π = 3.1416… So the way round is π times the distance across, which is 2π times the radius, the distance from the centre to the edge.

Definition: the radian. A second way to count angles. One radian is the angle at the centre of a circle whose piece of the rim is as long as the radius. A full circle is 2π radians, so one radian is 360 ÷ 2π = 57.3°. The calculator counts in degrees unless it is switched to radians; an angle 57 times too big or too small has been counted in the wrong one.

StepWhat to findYours
1 shadowm
2 angle °
3 times
4 round km
5 radius km
Source

The calculator opens from the button at the top of this section, or from the tab in the corner of the window. In it, the names beside the steps stand for your numbers: atan(shadow / 2), 360 / angle, times × 800, round / (2 × pi).

Definition: the Tropic of Cancer. The circle of latitude where the noon Sun stands straight overhead on the longest day of the year. Its latitude equals the tilt of the Earth’s axis (the Sun can stand overhead only as far from the equator as the axis leans), which changes slowly: 23.44° now, about 23.72° in 240 BCE.

Show how it is done
  1. The shadow: the drawing is made with the Sun 7.2° from straight up, and the ruler gives 0.253 m, about a quarter of a metre.
  2. The angle: atan(0.253 ÷ 2) = 7.2°. Now carry it to the centre of the Earth. Continue the Alexandria stick down into the ground: it points straight at the centre. At Syene, where the Sun stands straight overhead, the Sun’s ray itself runs straight down to the centre. So the line from Alexandria to the centre crosses 2 parallel rays, the one past the top of the stick and the one into Syene, and makes equal alternate angles with them (Hint 2): the angle at the top of the stick is the angle at the centre between the 2 cities. Set this beside what you wrote under the question.
  3. How many times: 360 ÷ 7.2 = 50. The walk from Alexandria to Syene is one fiftieth of the way round.
  4. The circumference: 50 × 800 km = 40,000 km.
  5. The radius: 40,000 ÷ 2π = 6,366 km.
the circumference  =  50 × the walk between the 2 cities
Source

The numbers are the Given’s, with its sources.

Set it beside what you wrote under the question.

His numbers

The stick you measured in Measure it is drawn with his own angle: the Sun one fiftieth of a full circle, 7.2°, from straight up. His method was yours, in his own unit; your numbers are in the table above.

StepHis
The angle7.2°, one fiftieth of a circle
How many times50
The walk5,000 stadia
The circumference250,000 stadia
Source

Cleomedes, The Heavens 1.7 (I.10, pp. 96–101, in H. Ziegler’s edition, Leipzig 1891): Syene and Alexandria on one meridian, 5,000 stadia apart; the shadow one fiftieth of its circle; the whole 250,000 stadia.

What moved it since

To turn his 250,000 stadia into kilometres you need his stadion, and nobody has it. So start from the common one, and then put in, one at a time, what we can measure today. Each button below is one of those, and moves the answer. The gold mark is yours, from step 4.

  1. His, in the common stadion. 250,000 stadia of 185 m each: 46,250 km, 16 % more than the real way round. In a shorter stadion his answer comes closer, but no surviving account says which stadion he used.
  2. A side road: 252,000. 2 later writers, Strabo and Pliny the Elder, give his answer as 252,000 stadia, not 250,000. It is exactly 700 stadia for each of the 360 degrees of a circle. In the common stadion it is 46,620 km.
  3. The walk, from today’s map. Leave the stadia out altogether, and take the 788 km from today’s map, on the Earth’s real shape. His ×50 on that: about 39,400 km, 1.5 % short.
  4. The angle, measured today. Today’s maps put the 2 cities 7.11° apart north to south, not 7.2°: one fiftieth of a circle was a rounded figure. The angle then fits into the full circle 360 ÷ 7.11 = 50.6 times, not 50, and 50.6 walks of 788 km make about 39,900 km, 0.3 % short.
  5. The Earth is not quite a ball. It is a little flattened at the poles. Where it is flatter, you walk further for the same change of angle, so a degree north to south is not the same length everywhere. Between these 2 cities it is 110.8 km; the average all the way round through the poles is 111.1 km. With the real shape, that way round is 40,008 km. Nothing is left over: the method, with today’s numbers, gives today’s answer.
The Earth’s circumference, on a line from 38,000 to 47,000 km: the coral mark moves with the buttons; the gold one is your answer.
  • the circumference, with that button’s numbers
  • the other positions
  • yours
Source

The common stadion: W. Smith (ed.), A Dictionary of Greek and Roman Antiquities (London: J. Murray), “Stadium” (600 Greek feet, 606 feet 9 inches English) and “Mensura” (the argument over shorter stadia). 252,000: Strabo, Geography 2.5.7; Pliny, Natural History 2.247. The cities’ latitudes, 24.09° and 31.20° north: the site’s own place list (Natural Earth); the distance and the length of a degree worked on the Earth’s shape with the radii 6,378.137 km at the equator and 6,356.752 km at the poles (the WGS 84 ellipsoid, as NASA’s Earth fact sheet gives it).

Today’s values: the Earth’s measured shape gives 40,008 km round through the poles and 40,075 km round the equator. Its mean radius is 6,371 km. That is a little more than 40,008 ÷ 2π = 6,367 km, because the Earth bulges at the equator: a circle through the poles takes the shorter way round, over the flattened poles (NASA’s Earth fact sheet: the WGS 84 radii, and the volumetric mean radius, 6,371.000 km).

Parallel rays reach 2 sticks on the curved Earth. The angle between the stick and the ray at Alexandria equals the angle at the centre between the 2 cities: one fiftieth of a circle, so the circumference is 50 walks. The cities are drawn 15° apart, not 7.2°, so that the angles can be seen. 360 ÷ 7.2 = 50, and 50 walks of 800 km are 40,000 km.
  • the Sun’s rays, all parallel
  • the ray past the top of the Alexandria stick
  • the 2 sticks, at Alexandria and at Syene
  • the Alexandria stick’s shadow
  • the Earth
  • a line from each city to the centre
  • the centre of the Earth
  • the angle between the stick and the ray, and the same angle at the centre: 7.2°
  • the walk between the cities: 800 km

The angle you found between the 2 cities is how far apart they are in something the round-Earth page named:

Defined earlier on the ladder: latitude.

See it in our model. Open the Planetarium over Aswan at noon on 21 June 2027, the longest day: the Sun stands 89.3° up, all but straight overhead. Then over Alexandria at its noon the same day: 82.2° up. The 2 differ by 7.1°, the cities’ difference in latitude. Aswan misses straight up by about two-thirds of a degree: it lies at 24.09° north, and the Tropic today is at 23.44°. In 240 BCE, with the axis tilted 23.72°, the miss was about 0.37°, less than the Sun’s own width, about half a degree. A 2 m stick then cast a shadow about 1.3 cm long, and the Sun’s width blurs every shadow’s edge by more than that: no shadow anybody could see. It is a model of the sky: never look at the real Sun.

Source

The tilt in 240 BCE: J. Laskar, Astronomy and Astrophysics 157 (1986), 59, Table 8 (p. 68) with the value at 2000, p. 69; today’s, NASA’s Earth fact sheet. The Sun’s width: its radius, 695,700 km, at the Earth’s mean distance, 149.6 million km, is 0.53°.

What works and why: what the method rests on, and where it could fail

The method is 3 ideas: the rays are parallel, the Earth is a ball, and the angle is the same fraction of a circle as the walk is of the way round. Here is where each came from, and where it would break.

⚑ Assumptions that made it work
  • The Sun is so far away its rays arrive perfectly parallel. If the Sun were nearby, its light would fan out. The shadow difference would then show the Sun’s angle in the sky, not Earth’s curve. Parallel rays need the Sun to be a huge distance away. This was assumed, not proven. And it quietly meant that if the Earth did go round the Sun, its orbit had to be vast, and nearby stars should show a tiny wobble as it moved. No one could find that wobble for another 2,000 years, and its absence was the main argument against a moving Earth. Rung 4 measures how far the Sun is; rung 5 finds the wobble. History of thinking: Aristarchus.
  • Earth is a perfect sphere. The geometry only works cleanly for a sphere. Earth is slightly flattened at the poles, but the error is small. Nor was a round Earth his assumption alone: a century earlier Aristotle had pointed to the curved shadow on the Moon during eclipses, and to stars that rise and sink as you travel north or south. History of thinking: Aristotle’s proofs.
  • Syene and Alexandria lie on the same north–south meridian. They do not quite — Alexandria is 2.981° west of Syene. This small error was unknown to him.
  • The walk between the cities is known well. Nobody knows how the 5,000 stadia were measured: Cleomedes gives the number, not how it was found, and this page takes it as given. Eratosthenes was head of the Library in Alexandria, and the Ptolemaic kings had distances along the Nile measured by royal surveyors (Martianus Capella, a Latin writer of the 400s CE, says the king’s surveyors gave Eratosthenes the distance in stadia from Syene south to Meroë), so he may simply have taken the number from such records. Any error in that baseline scales the final answer straight through. The ×50 is only as good as the walk it multiplies.
  • A line crossing 2 parallel lines makes equal angles. The step the question asked for. It was not new: Euclid had set it down in Alexandria some decades earlier. History of thinking: Euclid I.29.

The flat Earth, with a near Sun. In the first picture, Move the Sun, a flat Earth under a far Sun gives 2 equal shadows, so it cannot be right. But a flat Earth under a near Sun can give 2 different shadows too. 2 measurements do not tell the models apart; a third city does, as the round-Earth page showed: try it with 3 cities. A model that fits 2 measurements is not yet proved.

⚖ So… was he right? (careful — there is a trap here)

The tempting next line is: “a stadion was about 160 metres, so 250,000 stadia ≈ 40,000 km — almost exactly right!” That line hides 2 tricks.

  • No one knows exactly how long his stadion was. Ancient sources disagree. And some modern guesses at the stadion are themselves worked out by taking it that he was right. Using those to check him goes in a circle: the answer was baked into the ruler before the test began.
  • “40,000 km” was never a target he hit by luck. When the metre was created, in 1791, it was defined as one ten-millionth of the distance from the North Pole to the equator. That makes the Earth’s circumference 40,000 km by definition of the metre, and not because a measurement came out that way. (Later, better surveys showed the first survey had been slightly off, so the metre came out about a fifth of a millimetre short. By then the metre was in use and nobody wanted to change it, so the metre stayed and the Earth came out at 40,008 km instead of a round 40,000.) History of thinking: the metre.

So the question is not “did his number match ours?” but “is the geometry right?” — and that we can check today with no stadia anywhere. The straight north–south distance from Alexandria to Aswan is 788 km. Apply his own ×50 and you get ≈ 39,400 km. That is 1.5 % under the true 40,008 km, and 1.2 % of it is his rounded angle: the 2 cities really differ in latitude by 7.11°, not 7.2° (What moved it since puts each one in). The method is sound: run it today with precise instruments and the correct answer simply falls out.

His own book is lost. The story reaches us second-hand, mainly through lectures written 2 to 5 centuries later. So the oddly tidy numbers — exactly 7.2°, exactly 5,000 stadia — are probably a teacher’s rounded version of a messier real measurement.

⚑ 3 things that catch people out
  • Only north–south counts. Walking east or west adds nothing to the angle. If your friend is off to one side, take just the north–south part of the gap from a map. Otherwise your answer comes out too big.
  • Match the moment. On the same north–south line, your local noons happen at the same instant. So “at noon” and “at the same time” mean the same thing. If you are not lined up, each measure at your own local noon instead. That gets rid of the east–west problem.
  • Subtract most of the time — but sometimes add. The equator itself changes nothing, and it makes no difference which hemisphere either of you is in. What matters is the spot where the Sun stands straight overhead, which drifts between the tropics with the seasons. If that spot is outside the pair of you, both shadows point the same way and you subtract. If it falls between you, your shadows point opposite ways and you must add. And if one of you casts no shadow at all, that person’s angle is simply 0° — the well at Syene, all over again.

Why further apart is better

Your protractor is good to about half a degree, however steady your hand. That error stays the same size no matter where you stand. But the angle you are measuring grows with distance, at 1° for every 111.1 km north or south (your 40,000 km ÷ 360):

That lesson comes back again and again: you cannot make your instrument perfect, so make the thing you are measuring bigger. It is exactly why the distance to a star had to wait for a baseline as wide as the Earth’s whole orbit.

✋ Try it yourself — measure the Earth with a friend far away

You need one friend far away, due north or due south of you, and you need to know how far apart you are. At the same moment, each of you stands a stick upright and measures the Sun’s angle from straight up, as you did above. The rest is the method you have just worked out, with one catch: do you subtract your 2 angles, or add them?

No friend far away? Be both of you. Measure at noon on the longest day where you live, and a year later, at noon on the longest day again, somewhere far to the north or south of it. Now you are not trusting a friend to measure well. You are trusting something else: that the noon Sun stands the same height on the longest day every year. That is very nearly true, and the longest day is the best day to use, because there the noon Sun’s height hardly changes from one day to the next: a day out costs you almost nothing.

Move the Sun, and watch the centre of the Earth.

  • the Sun
  • its rays, all parallel
  • your stick, and the line from it to the centre
  • your friend’s stick, and the line from it
  • a stick’s shadow
  • the Earth
  • the break line: most of the distance cut out
🤔 Think it through: “The Earth is flat”

One day somebody tells you, quite seriously, that the Earth is flat. Before you answer, take a moment with these questions.

  1. What would you say to them, in a sentence or two?
  2. Can you prove what you say, or are you repeating what you were told? Which part of it could you check yourself?
  3. Have you tried it: 2 sticks, a friend and a sunny day, or the pictures on this page? What came out?
  4. Would that persuade them? And what would persuade you, if you were in their place?
  5. They answer with another argument. What then?

Their arguments, and yours

Here is what people who hold the Earth to be flat say. For each one: what would you answer, and how far could you prove your answer?

  1. “The horizon looks flat, and so does the sea.”
    Could you prove your answer?
    Something you could check

    From a beach the horizon is only about 5 km away, so you see a very small piece of the ball, and a small piece of a big ball looks flat. Watch a ship sail away: its hull disappears first, the top of its mast last. The horizon page lets you feel how far you can see.

  2. “Water always finds its level, so it cannot curve.”
    Could you prove your answer?
    Something you could check

    “Level” means at right angles to the pull of gravity, and gravity pulls towards the centre of the Earth, so a level surface curves with the Earth. Across a lake 10 km wide the water stands about 2 m higher in the middle than a straight line between the shores.

  3. “If the Earth were a spinning ball, we would feel it move.”
    Could you prove your answer?
    Something you could check

    In a smooth train or plane at a steady speed you do not feel the movement either; what you feel is a change of speed. The spin shows itself another way: watch the stars near the pole star (south of the equator, near the south pole of the sky) for an hour, and they turn round it. Why the sky turns.

  4. “A plane flying over a ball would have to keep pointing its nose down.”
    Could you prove your answer?
    Something you could check

    At 900 km/h a plane follows the curve by turning through about 8° an hour, a degree every 7 minutes or so: far less than the small corrections made for the wind. Nobody aims down; gravity keeps pulling the plane round.

  5. “You can see cities across the sea that a ball would hide.”
    Could you prove your answer?
    Something you could check

    Sometimes you can, sometimes you cannot: air over cold water bends light down around the curve, and by how much changes with the weather. Look at the same far coast on several days. On a flat Earth it would look the same every day.

  6. “2 different shadows also fit a flat Earth with a small Sun close above it.”
    Could you prove your answer?
    Something you could check

    For 2 cities, that is true; this page says so above. A third city, at a different distance, does not fit the small near Sun, while the round Earth fits all three. One more measurement is the test: 3 cities, one Sun.

  7. “Photos from space are fakes.”
    Could you prove your answer?
    Something you could check

    Then leave the photos out. In an eclipse of the Moon, the Earth’s own shadow falls on the Moon, and its edge is always curved, whatever the time of night and whichever way it falls. Aristotle saw that, and so can you.

Look back at your answers. The ones you can show yourself are the strongest: nobody has to take your word for them. The ones that rest on trusting someone are the ones the other side can always doubt. And being right is not the same as persuading: ask what would change their mind, and have ready what would change yours.

How to check a claim about the sky →