of stars revolving in ellipses round each other; just as Bradley's
attack on stellar parallax failed, but led to the discovery of
aberration, nutation, and the true velocity of light.

_Parallax_.--The absence of stellar parallax was the great
objection to any theory of the earth's motion prior to Kepler's
time. It is true that Kepler's theory itself could have been
geometrically expressed equally well with the earth or any other point
fixed. But in Kepler's case the obviously implied physical theory of
the planetary motions, even before Newton explained the simplicity of
conception involved, made astronomers quite ready to waive the claim
for a rigid proof of the earth's motion by measurement of an annual
parallax of stars, which they had insisted on in respect of
Copernicus's revival of the idea of the earth's orbital motion.

Still, the desire to measure this parallax was only intensified by the
practical certainty of its existence, and by repeated failures. The
attempts of Bradley failed. The attempts of Piazzi and Brinkley,[1]
early in the nineteenth century, also failed. The first successes,
afterwards confirmed, were by Bessel and Henderson. Both used stars
whose proper motion had been found to be large, as this argued
proximity. Henderson, at the Cape of Good Hope, observed alpha
Centauri, whose annual proper motion he found to amount to 3".6, in
1832-3; and a few years later deduced its parallax 1".16. His
successor at the Cape, Maclear, reduced this to 0".92.

In 1835 Struve assigned a doubtful parallax of 0".261 to Vega (alpha
Lyrae). But Bessel's observations, between 1837 and 1840, of 61 Cygni,
a star with the large proper motion of over 5", established its annual
parallax to be 0".3483; and this was confirmed by Peters, who found