Legendre's equation
/ Legendre: luh-ZHAHND /
When you solve a problem with spherical symmetry — the gravitational field of the Earth, the electric potential around a charged sphere, the quantum hydrogen atom — and separate the variables, the polar-angle part keeps producing the same differential equation. Legendre's equation is that equation: the master equation of the angular dependence on a sphere. Solving it once tells you the natural 'shapes' a field can take over a spherical surface.
It is the second-order linear equation (1 - x^2) y'' - 2x y' + n(n+1) y = 0, where x usually stands for cos(theta) so the physical range is -1 <= x <= 1, and n is a parameter. Its singular points sit exactly at x = +1 and x = -1 — the north and south poles — and they are regular singular points. Solving by power series about x = 0 gives a recurrence that, for general n, runs forever and diverges at the poles. The cure is to demand a solution finite at both poles, which forces n to be a non-negative integer; then the series truncates and the finite solution is the Legendre polynomial P_n(x). The second, independent solution Q_n(x) is unbounded at the poles and discarded in problems that include the whole sphere.
This is why the multipole expansion of physics — monopole, dipole, quadrupole — is built from Legendre polynomials, and why spherical harmonics (their two-angle generalisation) describe atomic orbitals, the cosmic microwave background, and the gravitational geoid. The lesson worth keeping is structural: quantization here (the integer n) is not imposed by hand but emerges from the single physical requirement that the solution stay finite where the geometry has a singular point.
Take n = 2: the equation (1 - x^2) y'' - 2x y' + 6 y = 0 has the polynomial solution P_2(x) = (3x^2 - 1)/2, finite everywhere on [-1, 1]. Its second solution Q_2 contains a log term like ln((1+x)/(1-x)) and blows up at the poles, so it is dropped for the whole-sphere problem.
Demanding finiteness at the poles selects the polynomial P_n and quantizes n.
P_n is the equation's polynomial solution; do not equate the equation with the polynomial. The full solution space is two-dimensional (P_n and Q_n); the polynomial is just the branch that stays finite, and only at the special integer values of n.