Combining Leavitt's cosmic yardstick with Vesto Slipher's handy red shifts, Edwin Hubble now began to measure selected points in space with a fresh eye. In 1923 he showed that a puff of distant gossamer in the Andromeda constellation known as M31 wasn't a gas cloud at all but a blaze of stars, a galaxy in its own right, a hundred thousand light-years across and at least nine hundred thousand light-years away. The universe was vaster—vastly vaster—than anyone had ever supposed. In 1924 he produced a landmark paper, “Cepheids in Spiral Nebulae” (nebulae,from the Latin for “clouds,” was his word for galaxies), showing that the universe consisted not just of the Milky Way but of lots of independent galaxies—“island universes”—many of them bigger than the Milky Way and much more distant.

This finding alone would have ensured Hubble’s reputation, but he now turned to the question of working out just how much vaster the universe was, and made an even more striking discovery. Hubble began to measure the spectra of distant galaxies—the business that Slipher had begun in Arizona. Using Mount Wilson’s new hundred-inch Hooker telescope and some clever inferences, he worked out that all the galaxies in the sky (except for our own local cluster) are moving away from us. Moreover, their speed and distance were neatly proportional: the further away the galaxy, the faster it was moving.