Quite lately Halm, at the Cape of Good Hope, measured
spectroscopically the velocity of the earth to and from a star by
observations taken six months apart. Thence he obtained an accurate
value of the sun's distance.[5]

But the remarkably erratic minor planet, Eros, discovered by Witte in
1898, approaches the earth within 15,000,000 miles at rare intervals,
and, with the aid of photography, will certainly give us the best
result. A large number of observatories combined to observe the
opposition of 1900. Their results are not yet completely reduced, but
the best value deduced so far for the parallax[6] is 8".807 +/-
0".0028.[7]


FOOTNOTES:

[1] In 1480 Martin Behaim, of Nuremberg, produced his _astrolabe_ for
measuring the latitude, by observation of the sun, at sea. It
consisted of a graduated metal circle, suspended by a ring which was
passed over the thumb, and hung vertically. A pointer was fixed to a
pin at the centre. This arm, called the _alhidada_, worked round the
graduated circle, and was pointed to the sun. The altitude of the sun
was thus determined, and, by help of solar tables, the latitude could
be found from observations made at apparent noon.

[2] See illustration on p. 76.

[3] See Dreyer's article on these instruments in _Copernicus_,
Vol. I. They were stolen by the Germans after the relief of the
Embassies, in 1900. The best description of these instruments is