non-Euclidean geometry
For two thousand years, geometry meant the rules Euclid wrote down for a flat sheet of paper: parallel lines stay the same distance apart forever and never meet. Non-Euclidean geometry is what happens when you let that one rule break. Picture a globe instead of a page. Two lines of longitude both run dead straight, due south, crossing the equator at right angles — yet they crash together at the pole. On a curved surface, lines that look parallel can meet, or fan apart and never come close; on a sphere, in fact, there are no parallel lines at all, because any two of these straight lines eventually cross.
There are two flavors, and the deepest way to tell them apart is to ask how many parallels you can draw through a point beside a given line. On a sphere-like surface that curves outward — elliptic geometry — the answer is none: every straight line eventually meets every other, and the angles of a triangle add up to more than 180°. On a saddle-like surface that curves the other way — hyperbolic geometry — infinitely many parallels fan through that point, and a triangle's angles add up to less. Flat Euclid's exactly-one-parallel, exactly-180° triangle turns out to be just the special case in the middle.
This isn't a mathematician's idle game. Einstein realized that gravity is not a force pulling across empty space — it's the curving of spacetime itself, and planets simply follow the straightest possible paths through that curved geometry. The Sun bends the space around it the way a heavy ball dents a trampoline, and Earth rolls along the dent. Non-Euclidean geometry is the language that made that idea sayable.
On a sphere, a triangle drawn from the North Pole down to two points on the equator can have three right angles — its angles sum to 270°, not 180°.
The flat-paper rule that angles sum to 180° simply doesn't hold on a curved surface.
For centuries mathematicians tried to prove Euclid's parallel postulate from his other rules and always failed. In the 1820s–30s Lobachevsky, Bolyai, and Gauss independently grasped why: you can drop it and build a perfectly consistent geometry instead — the hyperbolic one, where infinitely many parallels pass through a point. Riemann later generalized the idea, handing Einstein the tools he needed.