We use essential cookies to run the site (session, security, and your theme/language preferences). With your permission we also load embedded third-party content, such as YouTube videos. Cookie Policy
A 2000-year quest that mathematics ended with a single proof
Author(s):Elier Rodríguez García
Index
The Ancient Challenge
Take a circle. Now, using only a compass and an unmarked straightedge, draw a square with exactly the same area. No ruler markings, no protractor, no calculator — just the two tools the ancient Greeks allowed themselves.
This is squaring the circle, and people chased it for more than two thousand years. It appears in Egyptian papyri, obsessed Greek geometers, and filled the margins of medieval manuscripts. "Squaring the circle" even entered everyday language as a phrase for an impossible task.
Here is the catch hiding in plain sight. A circle of radius 1 has area π. To match it, your square needs a side of length π. So the whole question quietly becomes: can you construct the number π with compass and straightedge alone? That single number is where two thousand years of effort would finally break.
Try It Yourself
Below is a circle of area π next to a square you can resize. Drag the slider, or click a famous historical approximation, and watch how close the square's area gets to the circle's.
<p class="hint">The circle has radius 1, so its area is <b>π ≈ 3.14159</b>. A
matching square needs side <b>√π ≈ 1.77245</b>. Drag the slider or pick a
famous construction and watch the area error — it gets tiny, but never hits zero.</p>
<div class="stage">
<svg id="art" viewBox="0 0 240 200" aria-label="circle and square"></svg>
</div>
<div class="row">
<label for="side">Square side</label>
<input id="side" type="range" min="1.5" max="2.0" step="0.0001" value="1.77245">
<span id="sideval" class="mono">1.77245</span>
</div>
<div class="readout">
<div>Circle area π = <span class="mono">3.141593</span></div>
<div>Square area = <span id="sqarea" class="mono">3.141593</span></div>
<div>Error = <span id="err" class="mono err">0.000000</span></div>
</div>
<div class="btns">
<button type="button" data-side="1.772453851">Exact √π</button>
<button type="button" data-side="1.772877">Kochański 1685</button>
<button type="button" data-side="1.7724">22/7 area</button>
<button type="button" data-side="1.875">Crude 15/8</button>
</div>
const fmt = (s, v) => s.replace(/%(\w+)%/g, (_, k) => String(v[k] ?? ''));
// Squaring the circle: a unit circle has area pi. A square of equal area needs
// side sqrt(pi) ~ 1.7724539. Every length you can draw with compass and
// straightedge is algebraic, but pi is transcendental (Lindemann, 1882), so
// sqrt(pi) is NOT constructible. The approximations below get close and miss.
const PI = Math.PI;
const TARGET = Math.sqrt(PI); // sqrt(pi) ~ 1.7724539
const svg = document.getElementById('art');
const slider = document.getElementById('side');
const sideval = document.getElementById('sideval');
const sqarea = document.getElementById('sqarea');
const errEl = document.getElementById('err');
// SVG layout: unit length = 70px, circle centered at (90,100), radius 1.
const U = 50, CX = 90, CY = 100, R = U; // circle radius 1 unit on screen
function draw(side){
const half = side * U / 2;
svg.innerHTML =
'<circle class="circ" cx="'+CX+'" cy="'+CY+'" r="'+R+'"/>' +
'<rect class="sq" x="'+(CX-half)+'" y="'+(CY-half)+'" width="'+(2*half)+'" height="'+(2*half)+'"/>';
}
function update(side){
side = Math.max(1.5, Math.min(2.0, side));
const area = side * side; // square area for a unit-radius circle
const err = Math.abs(area - PI);
sideval.textContent = side.toFixed(5);
sqarea.textContent = area.toFixed(6);
errEl.textContent = err.toFixed(6);
// The error is "essentially zero" only for the exact (non-constructible) value.
errEl.className = 'mono err' + (err < 1e-9 ? ' ok' : '');
draw(side);
}
slider.addEventListener('input', () => update(parseFloat(slider.value)));
document.querySelectorAll('.btns button').forEach(b => {
b.addEventListener('click', () => {
const s = parseFloat(b.dataset.side);
slider.value = s;
update(s);
const err = Math.abs(s*s - PI);
console.log(fmt('%textContent%: side=%s%, area=%v%, error=%toExponential% %v2%', { textContent: b.textContent.trim(), s: s.toFixed(6), v: (s*s).toFixed(6), toExponential: err.toExponential(3), v2: err < 1e-9 ? '(the unconstructible exact value)' : '(still off — not equal)' }));
});
});
update(TARGET);
<p class="hint">The circle has radius 1, so its area is <b>π ≈ 3.14159</b>. A
matching square needs side <b>√π ≈ 1.77245</b>. Drag the slider or pick a
famous construction and watch the area error — it gets tiny, but never hits zero.</p>
<div class="stage">
<svg id="art" viewBox="0 0 240 200" aria-label="circle and square"></svg>
</div>
<div class="row">
<label for="side">Square side</label>
<input id="side" type="range" min="1.5" max="2.0" step="0.0001" value="1.77245">
<span id="sideval" class="mono">1.77245</span>
</div>
<div class="readout">
<div>Circle area π = <span class="mono">3.141593</span></div>
<div>Square area = <span id="sqarea" class="mono">3.141593</span></div>
<div>Error = <span id="err" class="mono err">0.000000</span></div>
</div>
<div class="btns">
<button type="button" data-side="1.772453851">Exact √π</button>
<button type="button" data-side="1.772877">Kochański 1685</button>
<button type="button" data-side="1.7724">22/7 area</button>
<button type="button" data-side="1.875">Crude 15/8</button>
</div>
const fmt = (s, v) => s.replace(/%(\w+)%/g, (_, k) => String(v[k] ?? ''));
// Squaring the circle: a unit circle has area pi. A square of equal area needs
// side sqrt(pi) ~ 1.7724539. Every length you can draw with compass and
// straightedge is algebraic, but pi is transcendental (Lindemann, 1882), so
// sqrt(pi) is NOT constructible. The approximations below get close and miss.
const PI = Math.PI;
const TARGET = Math.sqrt(PI); // sqrt(pi) ~ 1.7724539
const svg = document.getElementById('art');
const slider = document.getElementById('side');
const sideval = document.getElementById('sideval');
const sqarea = document.getElementById('sqarea');
const errEl = document.getElementById('err');
// SVG layout: unit length = 70px, circle centered at (90,100), radius 1.
const U = 50, CX = 90, CY = 100, R = U; // circle radius 1 unit on screen
function draw(side){
const half = side * U / 2;
svg.innerHTML =
'<circle class="circ" cx="'+CX+'" cy="'+CY+'" r="'+R+'"/>' +
'<rect class="sq" x="'+(CX-half)+'" y="'+(CY-half)+'" width="'+(2*half)+'" height="'+(2*half)+'"/>';
}
function update(side){
side = Math.max(1.5, Math.min(2.0, side));
const area = side * side; // square area for a unit-radius circle
const err = Math.abs(area - PI);
sideval.textContent = side.toFixed(5);
sqarea.textContent = area.toFixed(6);
errEl.textContent = err.toFixed(6);
// The error is "essentially zero" only for the exact (non-constructible) value.
errEl.className = 'mono err' + (err < 1e-9 ? ' ok' : '');
draw(side);
}
slider.addEventListener('input', () => update(parseFloat(slider.value)));
document.querySelectorAll('.btns button').forEach(b => {
b.addEventListener('click', () => {
const s = parseFloat(b.dataset.side);
slider.value = s;
update(s);
const err = Math.abs(s*s - PI);
console.log(fmt('%textContent%: side=%s%, area=%v%, error=%toExponential% %v2%', { textContent: b.textContent.trim(), s: s.toFixed(6), v: (s*s).toFixed(6), toExponential: err.toExponential(3), v2: err < 1e-9 ? '(the unconstructible exact value)' : '(still off — not equal)' }));
});
});
update(TARGET);
Result
Console
Notice what happens. You can get the areas to agree to two, three, even four decimal places — Kochański's 1685 construction is good to about four. But the error never reaches zero. Each approximation is a constructible number; the true target, π, is not. You are forever circling a value you cannot land on.
Why It Is Impossible
This is not an open problem and not merely unsolved — it is proven impossible. The verdict came in two steps.
Constructible means algebraic. Whatever a compass and straightedge can draw, starting from a unit length, is a number you can reach by adding, subtracting, multiplying, dividing, and taking square roots. Every such number is a root of a polynomial with integer coefficients — that is, algebraic — and its degree must be a power of two.
πis transcendental. In 1882, Ferdinand von Lindemann proved that πis not the root of any polynomial with integer coefficients at all. It is transcendental, escaping the algebraic numbers entirely. (See π for the number itself.)
The collision. If π were constructible it would be algebraic, so π=(π)2 would be algebraic too. But Lindemann says π is transcendental. Contradiction. Therefore πis not constructible, and the circle cannot be squared.
The result is final and exact — there is no clever trick, no better diagram, no approximation that ever becomes equality. It joins trisecting the angle and doubling the cube (both settled by Pierre Wantzel in 1837) as a construction that is impossible by the very nature of the tools. Like the halting problem, the answer is not "we don't know yet" but "we can prove it can never be done."
Where It Matters
A puzzle about drawing might sound like a curiosity, but its resolution reshaped mathematics:
Transcendental number theory: Lindemann's proof, building on Hermite's earlier work on e, founded a whole field that classifies which numbers are algebraic and which escape. The same machinery underlies modern results about π, e, and beyond.
The power of impossibility proofs: this was a triumph of showing a task can never be done — the same mindset behind undecidability and computational lower bounds. Proving "no method exists" is often deeper than finding a method.
Knowing when to stop: countless amateurs still send "solutions" to journals. Recognizing a proven-impossible problem saves effort — a practical lesson in any field where people chase perpetual motion of one kind or another.
Constructibility and algebra: the link between geometry and field theory (which lengths are constructible) is a gateway into Galois theory and the algebraic structure behind classical geometry.
Conclusion
Squaring the circle is the rare problem with a definitive obituary. For two millennia it looked like a hard drawing exercise; in 1882 Lindemann showed it was impossible from the start, because πis transcendental and so πcan never be built with compass and straightedge.
The lesson outlives the puzzle. Some limits are not about cleverness or computing power — they are baked into the tools you are using. The approximations in the demo will keep getting closer forever, and forever fall short. Knowing why the gap can never close is more powerful than any drawing that almost does. For another problem that is impossible by proof rather than by difficulty, see the halting problem.
Comments
Loading comments...