The Sizes and Masses of the Stars 77
pheric conditions, is the latter. This makes the problem
seem hopeless—barring the construction of huge telescopes,
far larger than any now in existence, and fabulously costly.
But Michelson—a life-long student of light waves and
their uses—has flanked the obstacle by the apparently mad
course of blocking up most of the opening of the telescope,
and letting the light come in only through two apertures
on opposite sides of the centre. The light from one of
these apertures alone would produce a diffraction disk (or
a figure of some other shape, if the opening were not
circular) of much larger size than that given by the full
aperture of the telescope. But, when the images from the
two apertures are superposed, a new phenomenon appears.
The large fuzzy spurious disk is crossed by fine, parallel,
equidistant, dark bands, or “fringes” of exquisitely clear
definition. These fringes are produced by the “interfer-
ence” of the light waves of the pencils proceeding from
the two apertures, very much as the spurious disk is pro-
duced by the interference of waves from all parts of the
full aperture: but the distance between the fringes is in-
versely proportional to the separation of the two apertures,
and when these are set at opposite sides of the full open-
ing, the fringe width is eighty per cent, of the diameter of
the spurious disk due to the unobstructed aperture.
Suppose now that we have a very close double star—
so close that, even with the full aperture, the spurious
images of the components are confused. If we pass to the
two apertures, we shall have larger spurious disks than
ever, and more completely superposed, but each will be
crossed by its own system of fringes, and the central fringe
of each system will pass through the geometrical image of
the corresponding star. If these geometrical images are
separated by only half the distance of the fringes, the