Michelson's Stellar Interferometer

Every star except the Sun is a point. Through the largest telescope ever built the disc of Betelgeuse is smaller than the blur the atmosphere puts around it, and no mirror that could be cast would change that. Michelson's answer in 1920, with Francis Pease, was to stop trying to make an image. Put a mask with two widely separated apertures over the telescope and look at the FRINGES. A true point source gives crisp ones. A source with real angular size is a crowd of points, each making fringes shifted a little from its neighbour's, and as you increase the separation between the apertures those shifts eventually span a whole fringe and the pattern washes out. The separation at which that happens is the measurement. They built a six-metre beam across the top of the 2.5-metre Hooker telescope at Mount Wilson and in December 1920 measured Betelgeuse at forty-seven thousandths of an arcsecond. At its distance that is a star which, put where the Sun is, would swallow the orbit of Mars. Nobody had ever measured the size of a star before. The idea is why long-baseline interferometry exists at all. Angular resolution needs SEPARATION, not area — so it can be bought by moving two small mirrors apart, which scales, rather than by casting one enormous one, which does not. What you do not get is light: an interferometer can resolve a star too faint for it to detect, which is why the big telescope and the long baseline are partners rather than rivals.