The Light of the Stars 45
but this determination, made by methods impracticable in
Newton’s day, detracts in no wise from the credit which
admittedly he deserves for giving the first correct estimate
of the order of magnitude of the remoteness of the
stars.
It is well-nigh a century since Bessel determined the
parallax and distance of 61 Cygni—applying principles
essentially similar to those of the range-finders which are
used with modern artillery. Slowly at first, then faster,
the technique of observation improved, and now, by taking
a dozen or so of photographs with one of our larger tele-
scopes we can measure the distance of a star a million
times as far away as the Sun, and be more likely than not
to get it right within five per cent. There are not many
stars as near as this—only a few score, among the millions
which the telescope reveals—and for the remote ones the
percentage error of our measures increases in proportion
to their distance. But, up to ten million times the Sun’s
distance, the results of direct measures of parallax are still
of value—particularly if we can take the average of three
or four determinations by good observers.
To describe these enormous distances with convenience,
we need a new unit of measurement. Astronomers are
coming very generally to adopt the distance at which the
parallax of a star would be one second of arc. The word
parsec has been coined to denote this distance—which is
206,000 times the Earth’s distance from the Sun, or, in
common units, 19 millions of millions of miles. It may
also be defined as equal to 3.26 light-years—the latter unit
being the distance which light travels in a year. No known
star lies within one parsec of the Sun. Two stellar systems
are nearer than two parsecs, and there are probably three
or four hundred nearer than ten parsecs—the majority