Eratosthenes Measures the Earth: Ancient Genius with Geometry

Around 240 BCE, the Greek mathematician and librarian Eratosthenes asked an audacious question: how big is the Earth? He never left Egypt to answer it. Using a stick, a shadow, a well 800 kilometers (497 miles) away, and geometry that any modern high schooler learns, he calculated Earth's circumference to within a few percent of the value satellites confirm today. No spacecraft, no telescopes, no trigonometric tables as we know them now — just an idea sharp enough to turn a shadow into a planet-sized measurement.

The observation that started everything

Eratosthenes was living and working in Alexandria, Egypt, where he served as chief librarian of the Library of Alexandria, the ancient world's largest center of scholarship, a post he held from roughly 246 BCE onward (Wikipedia: Eratosthenes). Access to that library mattered: it gave him a written record describing a curious detail from the town of Syene, near modern Aswan.

Travelers and scholars had long reported that at noon on the summer solstice, the sun in Syene shone straight down to the bottom of a deep well without casting any shadow at all. That meant the sun was directly overhead — at a 90-degree angle to the ground — at that exact moment, in that exact place.

Eratosthenes knew something else: in Alexandria, roughly due north of Syene, the sun at local noon on the same day never reached that overhead position. A vertical stick planted in the ground in Alexandria always cast a shadow, even at the sun's highest point of the year.

If Earth were flat, the sun's rays being effectively parallel (because the sun is so far away) would produce the same shadow angle everywhere at the same instant. The fact that they didn't told Eratosthenes something concrete: the ground itself was curving away between the two cities. That single inference — a measurable angle implies a measurable curvature — is the whole idea behind the experiment.

Eratosthenes never traveled between Alexandria and Syene to make his measurement. He combined a secondhand report about a well 800 km away with a single shadow measurement taken from his own rooftop, then closed the loop with geometry. It is one of the earliest examples of "armchair" science producing a planet-scale result.

How the geometry actually works

Picture two vertical sticks, one in Syene and one in Alexandria, both perfectly upright and both pointing, in effect, toward the center of the Earth. Because Earth is a sphere, those two sticks are not parallel to each other — they diverge by a small angle equal to the angular distance between the two cities as measured from Earth's center.

Here's the elegant part: if the sun's rays are parallel (a very safe assumption, since the sun sits about 150 million kilometers, or 93 million miles, away — enormously farther than Earth's 12,742-kilometer, or 7,918-mile, diameter), then the angle between a shadow-casting stick in Alexandria and the noon sun is the same as the angle between the two sticks themselves. Measuring the shadow angle in one city, when you already know the sun is directly overhead in the other, hands you the angular separation between the two locations on the globe.

Think of it like slicing an orange. If you know a wedge is 1/50th of the whole orange, and you know how long the wedge's outer edge is, you can figure out the circumference of the entire orange without ever measuring around it. Eratosthenes' shadow angle told him what fraction of the full 360-degree circle Alexandria and Syene represented; the known distance between them gave him the length of that fraction. Multiply, and you have the whole circle.

How did Eratosthenes measure the sun's angle without modern instruments?

He used a gnomon, a simple vertical rod that is essentially the oldest form of sundial, in use in Egypt since at least 3500 BCE (EAAE / gnomon history overview). At local noon on the solstice, he measured the length of the shadow the gnomon cast, then used the ratio of shadow length to rod height to calculate the shadow's angle through basic triangle geometry. That angle came out to about 7.2 degrees, or 1/50th of a full 360-degree circle.

On the measurement itself: he measured the shadow angle in Alexandria as roughly 7°12′ (7.2°), which he rounded to a clean fraction of the circle — exactly 1/50th (American Physical Society, "Eratosthenes Measures Earth"). Rounding a real-world measurement to a tidy fraction was a common practice in ancient astronomy, and it introduced only a small amount of error here.

Filling in the distance: stadia, camels, and professional pacers

The angle alone was useless without a distance between the two cities. According to later accounts, Eratosthenes did not personally pace out the 800-plus kilometers between Alexandria and Syene. Instead, ancient sources credit the distance to bematists — professional surveyors trained to walk with a highly consistent stride length, whose measurements were a standard tool of Hellenistic-era mapmaking (APS: "Eratosthenes Measures Earth").

Their figure came out to about 5,000 stadia. A stadion (plural: stadia) was an ancient unit of length roughly equal to the length of a stadium track, but its exact modern equivalent is contested because different Greek city-states used slightly different stadion lengths, generally estimated between 157 and 185 meters (515 to 607 feet) depending on which historical convention scholars apply (Wikipedia: Eratosthenes; arXiv: "Eratosthenes and Pliny, Greek geometry and Roman follies").

This ambiguity is exactly why you'll see different "final answers" quoted for Eratosthenes across different books and articles. The arithmetic he performed is not in dispute:

Earth's circumference = distance × (360° ÷ measured angle)

Or, plugging in his numbers directly in stadia: 5,000 stadia × (360° ÷ 7.2°) = 5,000 × 50 = 250,000 stadia.

Convert 250,000 stadia using the commonly cited "Egyptian stadion" of about 157.5 meters, and you land close to 39,400 kilometers (24,477 miles) — within roughly 1.7% of the true equatorial circumference. Use a slightly longer stadion value near 185 meters, and the total balloons past 46,000 kilometers, overshooting by more than 15%. Depending on which stadion length scholars apply, Eratosthenes' result lands anywhere from "eerily precise" to "impressively rough but still the right order of magnitude" — but every version of the analysis agrees his method was sound.

Stadion length used	Implied circumference	Comparison to modern value (40,075 km)
157.5 m (common Egyptian stadion)	~39,375 km (24,466 mi)	About 1.7% too low
~166 m (Attic stadion)	~41,500 km (25,791 mi)	About 3.6% too high
185 m (longer Greek stadion)	~46,250 km (28,745 mi)	About 15.4% too high
Modern equatorial measurement	40,075 km (24,901 mi)	— baseline

Table: Because the exact length of the ancient "stadion" Eratosthenes used is unresolved, his 250,000-stadia result converts to very different modern figures depending on which historical stadion is applied. Sources: arXiv, Eratosthenes and Pliny; Wikipedia, Eratosthenes' circumference of the Earth.

Why the result held up even with rough inputs

None of Eratosthenes' individual measurements were perfect, and modern historians of science are candid about that. Two assumptions in particular introduced error and, by luck, partly canceled each other out.

First, Syene is not exactly on the Tropic of Cancer — the latitude where the sun is truly overhead at the June solstice — and it is not perfectly due south of Alexandria; the two cities differ by a few degrees of longitude (Wikipedia: Eratosthenes). Second, the bematists' pacing distance was an estimate, not a laser-precise survey, and modern GPS measurement puts the true Alexandria–Aswan distance closer to 843 kilometers (524 miles) rather than a perfectly round 800 kilometers.

Individually, either error could have thrown the final number off by several percentage points. Together, in Eratosthenes' case, they happened to trend in offsetting directions, which is part of why his answer landed close to reality. That's an important nuance often missing from popular retellings: the accuracy wasn't purely the result of flawless technique, it was sound method plus a bit of good fortune in how the errors combined. This is a common story in the history of measurement science — even NASA's own early orbital calculations and geodetic surveys carried compensating errors that only later, more precise instruments could fully untangle.

The deeper point holds regardless of exactly which number you land on: Eratosthenes proved you can measure something too large to walk around, using nothing but an angle, a distance, and a sphere's basic geometric property that arc length divided by total circumference equals angle divided by 360 degrees. That relationship is still how modern geodesy (the science of measuring Earth's size, shape, and gravitational field) frames its most basic calculations, even though today's tools are laser rangefinders and satellite constellations instead of a gnomon and a well.

What this meant for exploration, cartography, and one very consequential mistake

Eratosthenes' circumference estimate didn't stay in a scholar's notebook. It rippled through more than 1,700 years of geography, navigation, and — in one famous case — a transatlantic miscalculation.

Ancient Mediterranean sailors used circumference estimates like his for dead reckoning, the practice of estimating a ship's position from known speed, heading, and elapsed time rather than direct observation. Later Hellenistic and Roman geographers used his figure to scale maps and estimate the size of unexplored landmasses, treating "how big is the Earth" as a solved, quotable number rather than an open question (APS: "Eratosthenes Measures Earth").

The most consequential downstream use came nearly 1,700 years later. Christopher Columbus, planning his 1492 voyage, worked from Earth-circumference estimates that traced back through medieval and Islamic scholarship to the ancient Greek tradition Eratosthenes had helped establish. But Columbus applied a shorter definition of the "mile" than the original calculations intended, which had the effect of shrinking his working estimate of Earth's size by a significant margin. That error made a westward sea route to Asia look like a matter of weeks rather than the ocean-spanning, unsurvivable distance it would actually have been on a correctly sized globe. Columbus reached land only because an entire pair of continents he didn't know existed happened to be in the way.

It's a genuinely useful case study in how a good number, handled carelessly two millennia downstream, can produce a very wrong plan that still succeeds for reasons that have nothing to do with the original math being right.

Did other ancient astronomers try to measure Earth too?

Yes. About a century after Eratosthenes, the Stoic philosopher Posidonius of Rhodes (c. 135–51 BCE) attempted his own version using the star Canopus rather than the sun: he measured how high Canopus appeared above the horizon at Rhodes versus Alexandria, and combined that angular difference with the known distance between the cities (MacTutor: Posidonius biography). His conclusions varied wildly depending on which ancient source you trust — one figure translates to roughly 39,000 kilometers (about right), while another, reported by the geographer Strabo, comes out closer to 29,000 kilometers, a serious undercount. The spread illustrates just how sensitive these ancient calculations were to small measurement errors amplified by the same "multiply by 360/angle" method Eratosthenes had pioneered.

What people get wrong about this story

The most persistent myth is that Eratosthenes "proved Earth was round" against a society of flat-Earth believers. That's not accurate. Educated Greeks from at least the time of Pythagoras and certainly by Aristotle, more than a century before Eratosthenes, already accepted that Earth was spherical, based on observations like the curved shadow Earth casts on the Moon during a lunar eclipse and the way ships disappear hull-first over the horizon. Eratosthenes' contribution wasn't discovering that Earth was round — it was figuring out how big the sphere actually was, which is a fundamentally different and much harder problem.

A second common misconception is that his result was essentially perfect, "off by only a couple hundred meters" or similar overly precise claims that occasionally circulate online. In reality, because the ancient stadion's exact length is still debated by historians, the honest answer is that Eratosthenes' figure could be anywhere from very close to noticeably off depending on unit conversion, and no serious historian of science claims certainty down to a fraction of a percent. What's genuinely remarkable, and doesn't need exaggeration to be impressive, is that a method built entirely from a shadow, a well, and a walked distance landed in the right ballpark at all.

A third misconception treats this as pure observation, when it was closer to applied engineering: Eratosthenes had to design the measurement (choosing the solstice, choosing a gnomon, choosing to compare two specific cities on roughly the same meridian), execute it carefully, and then trust a chain of geometric reasoning over gut instinct about how big the world "felt" like it should be. That combination of designed experiment, careful measurement, and mathematical interpretation is the essential shape of the scientific method, even though the term itself wouldn't be formalized for another two thousand years.

Frequently asked questions

Q: Who was Eratosthenes? A: Eratosthenes of Cyrene (c. 276–194 BCE) was a Greek mathematician, geographer, poet, and astronomer who served as chief librarian of the Library of Alexandria starting around 246 BCE. Beyond his Earth-circumference calculation, he is also credited with the Sieve of Eratosthenes, an early algorithm for identifying prime numbers, and wrote foundational works on mathematical geography (Wikipedia: Eratosthenes).

Q: How could Eratosthenes measure the sun's angle so accurately without modern instruments? A: He used a gnomon, a simple vertical stick that functions as a basic sundial. At noon on the summer solstice, he measured the length of the shadow it cast and used the ratio between shadow length and rod height to calculate the shadow's angle through similar-triangle geometry, arriving at roughly 7.2 degrees.

Q: Why didn't earlier Greek thinkers measure Earth's size? A: Philosophers like Aristotle and Plato accepted, more than a century earlier, that Earth was spherical based on indirect evidence such as its round shadow during lunar eclipses. But recognizing that different sun angles at different latitudes could be turned into an actual number for Earth's size required a specific insight and a specific experiment that nobody before Eratosthenes had designed.

Q: Did Eratosthenes' calculation influence Columbus's voyage? A: Indirectly, yes. Circumference estimates tracing back to the Hellenistic geographic tradition Eratosthenes helped found were used by later scholars, but Columbus worked from a chain of unit conversions that effectively shrank his working estimate of Earth's size. That miscalculation made a westward route to Asia look far shorter than it actually was; Columbus reached the Americas instead of Asia, an outcome his flawed math never anticipated.

Q: How accurate was Eratosthenes' measurement, exactly? A: It depends on which ancient "stadion" length is used to convert his 250,000-stadia result into modern units — estimates range from about 1.7% too low to more than 15% too high, depending on the conversion applied. The method itself, however, was mathematically sound and forms the conceptual basis of modern geodesy.

Q: What is a stadion, and why does its length matter so much? A: A stadion (plural: stadia) was an ancient Greek unit of distance, roughly the length of a stadium footrace track. Different Greek regions used stadion lengths ranging from about 157 to 185 meters (515 to 607 feet), and because Eratosthenes' final answer was expressed in stadia, the modern equivalent of his circumference estimate shifts substantially depending on which historical stadion scholars apply.

Q: Did anyone else in the ancient world try to measure Earth's circumference? A: Yes. About a century later, the astronomer Posidonius of Rhodes attempted a similar calculation using the star Canopus instead of the sun, comparing its angle above the horizon at Rhodes and Alexandria. His results varied significantly across ancient sources, from a fairly accurate figure to a substantial undercount, showing how sensitive this style of calculation was to small measurement errors.

Q: What assumptions did Eratosthenes' method rely on? A: Three main ones: that sunlight arrives at Earth in effectively parallel rays (true, given the sun's enormous distance), that Syene and Alexandria sit on the same meridian (they don't, quite), and that the professional pacers' distance estimate between the two cities was accurate (it was reasonably close but not exact). The latter two introduced errors that partially offset each other.

Q: Why is this still taught as a foundational example of the scientific method? A: Because it contains every core element in miniature: an observation that prompts a question, a hypothesis about the underlying cause, a designed measurement to test it, careful execution, and a mathematical interpretation that turns raw data into a testable, verifiable claim. It's one of the earliest surviving examples of that full chain being executed on a problem of planetary scale, roughly 2,000 years before "scientific method" became a named discipline.

The takeaway

Eratosthenes' achievement wasn't a lucky guess dressed up as science; it was a tightly reasoned chain from a single strange observation, a well in Syene with no noon shadow, to a number for the size of the entire planet. The geometry he used is something any student learns today, but recognizing that a shadow's angle could be converted into a fraction of Earth's total circumference required seeing a connection nobody had drawn before. The precision of his final number is genuinely debated among historians because of the ancient stadion's ambiguous length, and that debate is worth keeping rather than smoothing over for a cleaner story. What's not debated is the method: measure a small, accessible thing carefully, then use geometry to reason your way to something you could never measure directly. That is still how modern geodesy, and a great deal of physics beyond it, gets its biggest answers.

---

Read next:

• Archimedes: The Naked Run, the Claw, and the Man Who Almost Invented Calculus
• Pythagoras: The Theorem, the Brotherhood, and the Man Who Feared Beans
• Chaos Theory and the Butterfly Effect Explained