the Theorema Egregium
/ teh-oh-REH-mah eh-GREH-gee-um /
Here is the everyday miracle the theorem explains. Why can a flat map of the Earth never get all the distances right? Why does a slice of pizza stiffen along its length when you fold it at the crust? Why does paper roll smoothly into a tube but crumple if you try to wrap a ball? All of these are the same fact: a surface has an intrinsic curvature that you cannot change by bending without stretching, and you can detect it without ever leaving the surface. That fact is Gauss's Theorema Egregium — Latin for 'Remarkable Theorem.'
The precise statement is startling. The Gaussian curvature K was DEFINED using the principal curvatures, which describe how the surface sits and bends in the surrounding three-dimensional space (extrinsic data). Gauss proved that K can nonetheless be computed entirely from the first fundamental form E, F, G and their derivatives — that is, from measurements an ant confined to the surface could make with a ruler, never seeing the third dimension. In symbols, K depends only on E, F, G (and their first and second derivatives), not on e, f, g. The consequence: K is INTRINSIC. It is preserved by any local isometry — any bending that keeps all surface distances the same. Roll a flat sheet (K = 0) into a cylinder and it is still K = 0; that is forced, not lucky.
The implications ripple everywhere. It is why no flat map of the sphere (K = 1/R^2) can preserve all distances — the metric of a curved surface simply cannot be the flat metric, so every map projection must distort something. It is why the pizza-fold trick works: bending the slice into a curve (one principal curvature nonzero) forces the perpendicular direction to stay straight, because the product K must remain 0, lending rigidity. It launched intrinsic differential geometry and, through Riemann, the entire framework Einstein used for curved spacetime in general relativity. One sharp caution to internalise: this is exactly why a curve's curvature kappa (extrinsic, easily changed by bending a wire) and a surface's Gaussian curvature K (intrinsic, unchangeable by bending) must never be conflated — the 'egregium' is precisely that surface curvature, against all appearances, lives inside the surface.
Take a sheet of paper, K = 0 everywhere. Roll it into a cylinder: although it now visibly curves, the Theorema Egregium guarantees K stays 0 throughout, because rolling does not stretch any distance on the paper. Now try to wrap that same paper smoothly around a sphere, where K = 1/R^2 > 0. It is impossible without tearing or crumpling — the intrinsic curvatures 0 and 1/R^2 simply do not match, and no amount of bending can reconcile them. The everyday crinkling of gift wrap on a ball is the theorem made visible.
Paper (K = 0) rolls onto a cylinder but never smoothly onto a sphere (K > 0).
The theorem says K is intrinsic, NOT that all curvature is intrinsic. The mean curvature H and the second fundamental form remain extrinsic and DO change under bending (a flat sheet has H = 0, a cylinder rolled from it has H = 1/(2R)). Only the special combination K = k_1*k_2 is locked to the metric.