Visual walkthrough — HCF (GCD) — prime factorization method, Euclidean algorithm
Everything below is built with rectangles and tiles. We never assume you know what a "divisor" is — we draw it.
Step 1 — What "divides" even means, as a picture
In the figure: a length of is covered by four sticks of — flush at the end, remainder . Below it, leaves a gap: does not divide .

Step 2 — A common divisor, and the target we're hunting
WHAT the figure shows: two floors, and . A stick of length fits both. A stick of length also fits both — and it's longer. A stick of fits neither. The longest stick fitting both is , so .

Step 3 — The one magic fact: divisors survive subtraction
WHY it's true — look at the figure. If the whole is made of whole sticks, and the part you removed () is also made of whole sticks, then the leftover is whole-sticks-minus-whole-sticks, which is still a whole number of sticks. No stick ever had to be cut.

Step 4 — Subtracting many times at once = the remainder
WHAT the figure does: it lays copies of along , then marks the leftover stub . Since is smaller than , we can't fit another whole — that's why .

Step 5 — Repeat: the problem shrinks toward zero
WHAT the figure shows: a staircase of shrinking rectangles for : Each remainder (pink) becomes the tile for the next row. The pairs march , every arrow preserving the HCF by Step 4.

Step 6 — The floor edge case: why
WHAT the figure shows: a length- floor tiled by one stick of length (perfect), and an empty -floor that any stick "tiles" trivially. The biggest common tile is .

Step 7 — Degenerate & edge inputs (never hit a surprise)

The one-picture summary
Below: the entire derivation compressed. Left, the shrinking staircase of with the pink leftover always feeding the next tile row. Right, the arrows all carry the same HCF (Step 4) down to , where Step 6 reads off . The tile that finally lays flush — — is the answer.

Recall Feynman retelling (plain words)
Imagine two floors, a long one and a short one, and you want the biggest square tile that covers both perfectly. Here's the trick: chop the long floor down by removing as many whole copies of the short floor as fit — the little leftover stub is the remainder. Now here's the magic: any tile that covered both original floors also covers that stub (whole minus whole is still whole tiles). So the biggest common tile hasn't changed — but the numbers got smaller. Now play the same game with the short floor and the stub. Keep going; the stubs keep shrinking and must eventually reach zero. The moment a floor divides evenly (stub ), the tile you were using in that step is the biggest tile that fit everything all along. That's Euclid: divide, take the leftover, repeat, and the last leftover before zero is your answer.
Connections
- HCF (GCD) — prime factorization method, Euclidean algorithm — the parent topic this walkthrough visualises
- Fundamental Theorem of Arithmetic — powers the other method (min-exponent factoring)
- LCM — Least Common Multiple — the mirror image via
- Prime Numbers & Factorization
- Diophantine Equations — Bézout back-substitution rides on this same staircase
- Modular Arithmetic & RSA — coprimality (last-non-zero ) is the key
- Reducing Fractions to Lowest Terms — divide top and bottom by this HCF