the self-adjoint form

A symmetric arrangement is one that looks the same from either side — a balanced see-saw, a mirror-image dance. Among differential operators there is an analogous notion of being 'balanced', and an operator written in self-adjoint form is one engineered to have exactly that symmetry. It is the special way of writing a second-order equation that makes the beautiful eigenvalue theory go through.

Concretely, a general second-order operator P(x) y'' + Q(x) y' + R(x) y can almost always be massaged into the tidy shape (p(x) y')' + r(x) y, where the first two terms are bundled into a single derivative of a product. The recipe is to multiply the whole equation by an integrating factor mu(x) = (1/P) exp(integral of Q/P dx), chosen precisely so that after multiplying, the y'' and y' terms combine into (p y')' with p = mu P. This is the same integrating-factor trick used for first-order linear equations, deployed one level up. The result, (p y')' + (q + lambda w) y = 0, is the self-adjoint or Sturm-Liouville form.

Why bother? Because in this form the operator obeys a symmetry, called Lagrange's identity, that lets you integrate by parts twice and have the boundary terms cancel under standard boundary conditions. That symmetry is exactly what forces the eigenvalues to be real and the eigenfunctions to be orthogonal. So self-adjoint form is not cosmetic bookkeeping; it is the structural prerequisite that unlocks every guarantee of Sturm-Liouville theory. An equation that cannot be put in this form lacks those guarantees.