Chapter Four: The World’s First Think Tank
Alexandria, 300–415 CE
In the summer of 240 BCE, a man in the Egyptian city of Alexandria heard a story about a well.
The well was in Syene, a town roughly eight hundred kilometres to the south, near the first great cataract of the Nile. Travellers reported something peculiar about it: on a single day each year — the summer solstice, the longest day — at exactly noon, the sun shone directly down into the well and illuminated its entire bottom. No shadow. The sun was directly overhead, not a degree off plumb. In every other town in Egypt, on the same day at the same hour, upright poles and obelisks cast shadows. But in Syene, on the solstice, they did not.
The man who heard this story was named Eratosthenes. He was a mathematician, a geographer, a poet, a music theorist, and the head librarian of one of the most extraordinary institutions the ancient world had produced — the Library of Alexandria, repository of an estimated half million scrolls, the accumulated written knowledge of the entire Mediterranean civilisation. Eratosthenes was, by the affectionate and slightly cutting nickname his contemporaries had given him, “Beta” — always second-best at everything, because he was competent in so many fields that he never dominated any single one. It is the kind of nickname that contains more admiration than contempt.
What Eratosthenes did with the story of the well was one of the most elegant pieces of applied mathematics in human history. He did not travel to Syene. He did not commission a survey. He sat in Alexandria with a stick, waited for the summer solstice, measured the shadow the stick cast at noon, and used two facts — the angle of that shadow, and the distance from Alexandria to Syene — to calculate the circumference of the entire Earth.
He was right to within about two percent.
The City That Collected Everything
To understand what Alexandria was, you have to understand that it was consciously, deliberately, designed to be the greatest city in the world.
Alexander the Great founded it in 331 BCE, at the mouth of the westernmost branch of the Nile Delta, on a narrow strip of land between the sea and a lake, chosen for its natural harbour and its position on the Mediterranean trade routes. After Alexander’s death his empire fragmented, and Egypt fell to one of his generals, Ptolemy I, who made Alexandria his capital. Ptolemy had watched Alexander collect things — territories, peoples, knowledge — and he understood that collecting knowledge was a form of power. He founded the Library, and his son Ptolemy II expanded it into an institution without precedent.
The Library was connected to a research institution called the Mouseion — from which our word “museum” descends, though the ancient institution was nothing like a modern museum. It was a community of scholars, paid salaries by the Ptolemaic state, fed and housed in a grand complex that included dining halls, covered walkways for philosophical discussion, lecture theatres, and the Library itself. Scholars from across the Greek-speaking world — and beyond — were invited, funded, and given access to a collection of scrolls that had been assembled by purchasing, copying, and sometimes simply confiscating manuscripts from every ship that docked in Alexandria’s harbour.
According to later accounts, books arriving in Alexandria were often copied and sometimes retained in the royal collection. This is, depending on your perspective, either the most impressive act of intellectual acquisition in history or a large-scale programme of cultural theft. Possibly both. The effect was to concentrate the written knowledge of the ancient world — Greek drama, Egyptian religious texts, Babylonian astronomy, Indian mathematics, Phoenician navigation — in a single place, where scholars could read across traditions and synthesise ideas that had previously been entirely separate.
This is the environment in which Eratosthenes worked, and it explains a great deal about what he accomplished. A man who had read the Babylonian astronomical records alongside the Greek geometrical tradition alongside the Egyptian geographical surveys was in a position to connect things that no one had connected before. The circumference calculation is exactly this kind of synthesis: it uses Greek geometry (the properties of parallel lines and angles), Egyptian geography (the measured distance between Alexandria and Syene, kept in the administrative records), and Babylonian astronomical observation (the recognition that on the solstice, at Syene’s latitude, the sun is directly overhead). None of these pieces of knowledge originated with Eratosthenes. What originated with him was the idea of combining them.
The Calculation That Measured a Planet
Here is how it worked.
Eratosthenes knew that the Earth is a sphere — this was established Greek doctrine by his time, argued by Aristotle on the basis of the circular shadow Earth casts on the moon during lunar eclipses, the way ships disappear hull-first over the horizon, and the fact that different stars are visible from different latitudes. He also knew, from the travellers’ reports, that on the summer solstice at noon, the sun is directly overhead at Syene — meaning that a line from Syene to the sun passes straight through the centre of the earth.
At the same moment, in Alexandria, a vertical stick casts a shadow. Eratosthenes measured the angle of that shadow and found it to be about 7.2 degrees — or, in the way he would have expressed it, one-fiftieth of a full circle of 360 degrees.
Now the geometry. If you draw two lines from the centre of the earth — one pointing toward the sun through Syene, one pointing straight up from Alexandria — those two lines diverge by exactly the same angle as the one Eratosthenes measured at Alexandria’s stick. This follows from the basic geometry of parallel lines: the sun’s rays arriving at both cities are parallel (the sun is so far away that the angle between rays at two points eight hundred kilometres apart is negligible), so the angle between them at the earth’s surface equals the angle between them at the earth’s centre.
The arc of the earth’s surface between Alexandria and Syene subtends the same 7.2 degrees — one-fiftieth of the full circle. The distance between the two cities was known from the administrative records to be about 5,000 stadia — measured, according to some accounts, by professional bematists, trained walkers who counted their paces over long distances. If that arc is one-fiftieth of the full circumference, then the full circumference is fifty times 5,000 stadia, or 250,000 stadia.
Translating this into modern units is complicated by uncertainty about the length of the stadion Eratosthenes used — Greek stadions varied by region — but the best scholarly estimates put his result at approximately 40,000 kilometres, strikingly close to the modern measurement of 40,075 kilometres. The two modest errors in his method — Alexandria is not quite due north of Syene, and Syene is not quite on the Tropic of Cancer — cancelled each other out with remarkable good fortune, producing an answer of almost uncanny accuracy.
The genius of the calculation is not the answer. It is the structure of the reasoning. Eratosthenes never left Alexandria. He measured one angle, used one recorded distance, and applied one geometrical principle to deduce the size of an object he was standing on and could not step back to view. This is mathematics as a mode of perception — a way of seeing things that are too large, or too distant, or too abstract, for direct observation. He could not see the whole earth. But he could calculate it, because he understood the geometry that connected the small observable angle to the enormous unobservable circumference.
The Sieve and the Stars
Eratosthenes did not stop at the earth’s circumference. He also calculated the tilt of the Earth’s axis with an accuracy of a fraction of a degree. He constructed a map of the known world, organising it with a grid of parallels and meridians — the direct ancestor of the latitude and longitude system still in use today. He compiled a catalogue of 675 stars, establishing positions for each one. He developed the first scientific chronology, using Egyptian and Persian records to date historical events. And — a small but elegant contribution to pure mathematics — he devised a procedure for finding prime numbers that is still taught in schools under his name: the Sieve of Eratosthenes.
The Sieve works by elimination. Start with a list of all whole numbers from 2 upward. The first number, 2, is prime — circle it, and then cross out every multiple of 2 (4, 6, 8, …) because they cannot be prime. The next uncrossed number is 3, which must be prime — circle it, and cross out every multiple of 3. Continue: the next uncrossed number is 5, then 7, then 11. Every number that is not crossed out when you reach it is prime, because if it had any factor smaller than itself, that factor would already have been circled and the number crossed out in an earlier step.
The procedure is not fast for large numbers, but it is systematic, guaranteed to work, and requires no clever inspiration at each step — just the mechanical application of a rule. It is, in the language of modern computing, an algorithm for generating primes: a finite, deterministic procedure that produces a correct result. Eratosthenes had invented not just a mathematical result but a mathematical process — a way of thinking that prefigures, by two thousand years, the algorithmic thinking of the computer age.
Archimedes and the Edge of the Possible
While Eratosthenes was running the Library, the greatest pure mathematician of antiquity was working in Syracuse, on the island of Sicily, and conducting his mathematical correspondence with the Alexandrian scholars by letter. His name was Archimedes, and he is one of the handful of historical figures who genuinely deserves the word genius without qualification.
Archimedes was born around 287 BCE, the son of an astronomer. He may have studied in Alexandria as a young man — his letters to Eratosthenes and to the Alexandrian mathematician Conon suggest close collegial relationships. He returned to Syracuse, where he spent the rest of his life in the service of the city’s king, Hiero II, solving practical engineering problems while simultaneously pursuing mathematics that was, in its depth and ambition, completely detached from any practical application he could have imagined.
His contributions span an astonishing range. He proved that the area of a circle equals πr², that the volume of a sphere is two-thirds the volume of the cylinder that contains it, and that the surface area of that sphere equals the lateral surface area of the same cylinder. He worked out the areas enclosed by parabolas and spirals. He developed the principle of the lever — “give me a place to stand and a lever long enough and I will move the world” — and the principle of buoyancy (the famous bathtub insight). He designed war machines, including catapults and cranes that could grab Roman ships by the prow and overturn them, that held the Roman siege of Syracuse at bay for two years.
And — most important for our story — he came closer to calculus than anyone in the world would come again for the next fifteen hundred years.
Squeezing π Between Two Polygons
The problem of calculating π — the ratio of a circle’s circumference to its diameter — was ancient by Archimedes’ time. The Babylonians had used 3. The Egyptians had used 256/81, which gives approximately 3.16. But nobody had a systematic method for calculating π to arbitrary precision, and nobody had rigorously proved that any particular value was correct rather than merely close.
Archimedes invented both.
His method was to trap π between two polygons — one inscribed inside the circle, one circumscribed outside it. The inscribed polygon’s perimeter is less than the circle’s circumference; the circumscribed polygon’s perimeter is greater. Dividing each perimeter by the diameter gives a lower bound and an upper bound for π. As you increase the number of sides of the polygon, the bounds close in and π is squeezed more and more tightly between them.
Archimedes started with hexagons — six-sided polygons — and doubled the number of sides repeatedly until he reached polygons with 96 sides each. The calculation at each doubling step requires finding the lengths of the new polygon’s sides from the lengths of the previous ones, which involves square roots and careful arithmetic. Without algebra, without decimal notation, working with fractions that become increasingly unwieldy as the sides multiply — Archimedes carried the computation all the way to 96 sides and showed that:
3 + 10/71 < π < 3 + 1/7In decimals: 3.1408 < π < 3.1429.
The true value of π is approximately 3.14159. Archimedes’ bounds contain it, as claimed. The calculation is entirely rigorous — it is a proof, not an approximation, in the technical sense that no error is unaccounted for. He did not say “π is approximately 3.14.” He said “π is definitely larger than this and definitely smaller than that,” and proved both claims.
The method is called the method of exhaustion, and Archimedes did not invent it — the Athenian mathematician Eudoxus had developed it a century earlier — but Archimedes applied it with a power and range that nobody before him had approached. The underlying idea is to approximate a curved figure with a sequence of simpler figures (polygons, in this case) that fill more and more of the curved shape. As the approximation improves, it exhausts the gap between the polygon and the curve. If you can show that the gap can be made smaller than any given quantity, you have, in effect, found the exact value.
This is the essential idea of integration — the mathematical operation that lies at the heart of calculus — dressed in the language of classical Greek geometry. Archimedes was doing integration, in everything but name, in the third century BCE. He calculated the area under a parabola by filling it with an infinite series of triangles, summing the series to get an exact result. He found the volume of a sphere by a similar process. In a remarkable work called The Method, discovered only in 1906 when a prayer book containing an overwritten version of the text was identified in Constantinople, he described a procedure that is essentially the use of infinitesimals — treating areas as composed of infinitely many infinitely thin slices — to discover geometrical results, which he then proved formally by the method of exhaustion.
He knew, in other words, that his informal infinitesimal method was not rigorous — that “infinitely thin slices” was not a concept that stood up to the Greek standard of proof. So he used it to find the answer, and then used the method of exhaustion to prove it. Discovery and proof, separated into two stages, with different tools for each. This is a sophisticated understanding of mathematical practice — more sophisticated, in some ways, than what many later mathematicians displayed.
The Death of Archimedes, and What It Symbolises
In 212 BCE, the Roman general Marcellus finally captured Syracuse after a two-year siege that Archimedes’ war machines had prolonged far beyond what any Roman strategist had anticipated. Marcellus, who reportedly wept at the sight of the beautiful city he was about to destroy, gave specific orders that Archimedes was to be brought to him unharmed.
A Roman soldier found Archimedes in his house, drawing geometric figures in the sand. The soldier ordered him to come. By one account — recorded by Plutarch three centuries later, so its literal accuracy is uncertain — Archimedes said: “Do not disturb my circles.” The soldier, whether out of impatience or ignorance of who this old man was, killed him on the spot.
Marcellus was furious. He gave Archimedes an honourable burial and reportedly sought out his relatives to provide for them. The mathematician who had kept the Roman army at bay for two years with his machines of war was killed by a foot soldier who did not know who he was.
The story — like the legend of Hippasus, like many of the best stories in mathematical history — may be embellished. But it is historically attested by multiple ancient sources and is broadly believed. And even if the words “do not disturb my circles” are legend, the image they preserve is true: a man so deep in a geometric problem that the collapse of his city registered less urgently than the diagram in front of him.
Archimedes had asked to have a sphere inscribed in a cylinder carved on his tomb, with the ratio of their volumes — 2:3, one of his proudest results — inscribed beneath it. The Roman statesman Cicero, visiting Syracuse more than a century later, found the tomb overgrown and neglected, and had it cleared and restored. He could read the inscription. It was still there.
The Machinery of the Heavens
One of the most startling artefacts of the ancient world was pulled from a shipwreck off the Greek island of Antikythera in 1901. It lay unrecognised in the Athens National Archaeological Museum for decades, a lump of corroded bronze that was clearly mechanical but whose function was obscure. Only in the second half of the twentieth century, with X-ray imaging and later with high-resolution CT scanning, did its structure become clear.
The Antikythera mechanism is a geared computing device, built sometime between 200 and 60 BCE, designed to calculate and display astronomical positions: the phases of the moon, the positions of the five known planets, the timing of solar and lunar eclipses, the dates of the Olympic Games. It has at least thirty bronze gears of varying sizes, meshing with extraordinary precision, driven by a single input — turning a handle on the side — that advances all the celestial displays simultaneously. On the front, a large circular dial shows the position of the sun and moon in the zodiac. On the back, two spiral dials display eclipse cycles.
It is, to put it plainly, a mechanical astronomical computer. It is more mechanically sophisticated than anything else we know of for the next thousand years. And it demonstrates something important about Alexandrian and Greek science in the third to first centuries BCE: the ambition to model the heavens mathematically had reached the point where mathematical models could be embodied in physical mechanisms.
The mathematics behind the mechanism required exact knowledge of astronomical periods — how many days in a lunar month, how many months in the 19-year Metonic cycle that brings the lunar and solar calendars back into alignment, the precise ratios of the planetary orbital periods. All of this came from the astronomical tradition that ran from Babylonian observation through Greek mathematical refinement to the great astronomer Hipparchus of Nicaea, who worked in the second century BCE and whose mathematical models of the sun and moon’s motion were the most accurate in the ancient world.
Hipparchus built on both Babylonian observational records — he had access to centuries of eclipse data — and the Greek geometrical tradition. He constructed the first trigonometric table in history: a table of the chord of an angle (related to what we would call twice the sine of half the angle), calculated for every half-degree from 0 to 180 degrees. He used this table to calculate the distance to the moon with a result, obtained by analysing the geometry of solar eclipses observed from different locations, that was accurate to within a few percent. The moon is roughly sixty Earth-radii away. Hipparchus calculated it to be between 62 and 67 Earth-radii. The modern value is about 60.3.
The Antikythera mechanism encodes Hipparchus’s lunar theory in its gears. The gear ratio that drives the moon pointer is 254:19 — encoding the fact that the moon completes 254 rotations relative to the stars in the same 19-year Metonic cycle during which it completes 235 lunar months. This ratio was known from Babylonian observation, incorporated into Hipparchus’s mathematical model, and then translated into a physical gear ratio by some unknown Alexandrian craftsman. Babylonian astronomy, Greek mathematics, and Greek engineering, fused into a portable calculating machine.
The Woman at the Lectern
We have been speaking mostly of men, because the historical record is mostly of men — women’s contributions to ancient intellectual life were systematically underdocumented, their names absent from the lists of library heads and court mathematicians. But Alexandria produced at least one woman whose mathematical reputation was large enough that even the hostile sources that record her death cannot obscure her stature.
Her name was Hypatia. She was born around 360 CE, the daughter of Theon of Alexandria, himself a distinguished mathematician and the last known member of the Mouseion. She surpassed her father. She wrote commentaries on Diophantus’s Arithmetica — a foundational text in number theory — and on Apollonius’s work on conic sections, and she taught mathematics and philosophy to students who came from across the Mediterranean world. Her students wrote of her with a reverence that suggests not just admiration but something close to awe. One, the bishop Synesius of Cyrene, wrote to her for mathematical advice long after he had left Alexandria and entered the church, asking her to design a hydrometer and an astrolabe — practical instruments whose design required the kind of precise mathematical knowledge that only she, in his estimation, possessed.
She lectured publicly, from a chariot, in the tradition of the ancient philosophers. She had political influence: the Roman prefect Orestes was among her students, and she was drawn into the bitter power struggles between the Roman civil authority and the Christian patriarch Cyril of Alexandria. In 415 CE, during a period of particular tension, a mob — whose precise relationship to Cyril has been debated by historians for sixteen centuries — seized Hypatia from her chariot in the street, dragged her to a church, murdered her, and burned her body.
She was approximately fifty-five years old.
Her mathematical works are lost. We know of them only through references in other texts and through the commentaries that were later attributed to her father but which scholarly analysis suggests were at least partly hers. The murder of Hypatia has been mythologised — made to stand for the death of classical civilisation, the triumph of religious barbarism over rational inquiry — in ways that historians rightly resist, because the reality was more complicated. The Library of Alexandria had already declined significantly before her death; the great fire that supposedly destroyed it in a single catastrophic event is itself a myth compounded from several smaller events spread across centuries. Classical learning did not end with Hypatia.
But her death marks, symbolically and roughly chronologically, the end of Alexandria’s era as the intellectual capital of the world. The political and religious upheavals of the fifth century CE would redirect the centre of mathematical activity — first to the court scholars of the Persian Sasanian empire, then to the great institutions of the Islamic Golden Age, where the Greek inheritance would be absorbed, extended, and transformed. That transformation is the subject of Chapter Six.
What Alexandria Made Possible
The Alexandrian achievement was not any single discovery, though the discoveries were extraordinary. It was an institutional achievement: the demonstration that concentrating scholars, resources, and recorded knowledge in a single place, and giving scholars the freedom and the funding to pursue questions for their own sake, produces an outpouring of intellectual work that no individual, however brilliant, could generate alone.
Eratosthenes measured the earth because he had access to the administrative distance records and the Greek geometrical tradition and the Babylonian astronomical observations — all in one library, all available to one mind. Archimedes worked on problems that Eudoxus had posed and that Conon in Alexandria was corresponding about, refining and extending a tradition rather than starting from scratch. Hipparchus built his lunar model on Babylonian eclipse records that the Library preserved. The Antikythera mechanism’s anonymous maker translated Hipparchus’s mathematics into bronze gears. Hypatia taught the accumulated wisdom of Diophantus and Apollonius to students who would carry it into the next era.
This is what institutions make possible: the accumulation, preservation, and transmission of knowledge across generations. Individual genius is necessary but not sufficient. Archimedes was one of the greatest mathematical minds who ever lived, but he needed the Greek geometrical tradition — Euclid’s axioms, Eudoxus’s method of exhaustion, the Babylonian astronomical data — to do what he did. Remove the tradition, and the genius has nothing to stand on.
Alexandria understood this and built accordingly. The Library was not a monument. It was a machine — a machine for generating knowledge by connecting minds across time and space, allowing each generation to build on the work of the previous one rather than reinventing it. Modern universities are, in their essential structure, descendants of the Mouseion. The idea that scholars should be paid to think, housed together, given access to a great library, and encouraged to pursue difficult questions found one of its clearest and most influential ancient institutional forms in Alexandria under the Ptolemies.
It was one of antiquity’s best institutional ideas.
A Calculation Worth Sitting With
Before we leave Alexandria, let us return to Eratosthenes and his stick.
The earth’s circumference is 40,075 kilometres. Eratosthenes calculated it, alone in a library, with a stick and a piece of recorded information about a well, to within about two percent. He had never seen the earth from outside. He had no satellite, no airplane, no ship that had circumnavigated the globe. He had geometry, and the ability to see that a small local measurement — the length of a shadow at noon on a particular day — was connected, by an exact logical chain, to a global fact about the planet he was standing on.
This is the deepest lesson of the Alexandrian tradition: that the world is legible. That careful measurement, combined with mathematical reasoning, can reveal truths about things that are far beyond direct observation. The sun is millions of kilometres away, but its angle can be measured with a stick. The earth is a sphere with a circumference of forty thousand kilometres, but a shadow in Alexandria and a shadowless well in Syene tell you the number. The moon is four hundred thousand kilometres distant, but the geometry of an eclipse brings it within reach of calculation.
Mathematics, in the hands of the Alexandrian scholars, became a form of sight — a way of seeing what is too large, too distant, too fast, or too abstract for human eyes to perceive directly. Eratosthenes could not see the curve of the earth. But he could see its circumference in the shadow of a stick.
That way of seeing — mathematical, indirect, astonishingly powerful — is the gift Alexandria gave to every subsequent generation of scientists. It is why we can calculate the mass of a black hole without visiting it, determine the composition of a star from its light, predict the path of a comet centuries in advance. The tools have changed completely. The underlying idea — measure what you can reach, and reason your way to what you cannot — has not changed at all.
In the next chapter, we leave the Mediterranean and travel east, to the Indian subcontinent, where a different tradition had been quietly building toward one of the most consequential mathematical discoveries of all time. The Greeks had a concept for every kind of number except one: the number that meant nothing. India was about to fill that gap — and in doing so, make the whole of modern mathematics possible.