The work of Herschel in this matter has been checked by comparing
spectroscopic velocities in the line of sight which, so far as the
sun's motion is concerned, would give a maximum rate of approach for
stars near Hercules, a maximum rate of recession for stars in the
opposite part of the heavens, and no effect for stars half-way
between. In this way the spectroscope has confirmed generally
Herschel's view of the direction, and makes the velocity eleven miles
a second, or nearly four astronomical units a year.

The average proper motion of a first magnitude star has been found to
be 0".25 annually, and of a sixth magnitude star 0".04. But that all
bright stars are nearer than all small stars, or that they show
greater proper motion for that reason, is found to be far from the
truth. Many statistical studies have been made in this connection, and
interesting results may be expected from this treatment in the hands
of Kapteyn of Groningen, and others.[9]

On analysis of the directions of proper motions of stars in all parts
of the heavens, Kapteyn has shown[10] that these indicate, besides the
solar motion towards Hercules, two general drifts of stars in nearly
opposite directions, which can be detected in any part of the
heavens. This result has been confirmed from independent data by
Eddington (_R.A.S., M.N._) and Dyson (_R.S.E. Proc._).

Photography promises to assist in the measurement of parallax and
proper motions. Herr Pulfrich, of the firm of Carl Zeiss, has vastly
extended the applications of stereoscopic vision to astronomy--a
subject which De la Rue took up in the early days of photography. He
has made a stereo-comparator of great beauty and convenience for