seconds. To map stars we wish to know the exact second when they
cross the meridian, or the north and south line in the celestial
dome above us. The telescope (Fig. 21, p. 61) swings exactly north
and south. In its focus a set of fine threads of spider-lines is
placed (Fig. 23). The telescope is set just high enough, so that
by the rolling over of the earth [Page 65] the star will come into
the field just above the horizontal thread. The observer notes the
exact second and tenth of a second when the star reaches each
vertical thread in the instrument, adds together the times and
divides by five to get the average, and the exact time is reached.

[Illustration: Fig. 23.--Transit of a Star noted.]

But man is not reliable enough to observe and record with sufficient
accuracy. Some, in their excitement, anticipate its positive passage,
and some cannot get their slow mental machinery in motion till
after it has made the transit. Moreover, men fall into a habit of
estimating some numbers of tenths of a second oftener than others.
It will be found that a given observer will say three tenths or
seven tenths oftener than four or eight. He is falling into ruts,
and not trustworthy. General O. M. Mitchel, who had been director
of the Cincinnati Observatory, once told one of his staff-officers
that he was late at an appointment. "Only a few minutes," said the
officer, apologetically. "Sir," said the general, "where I have
been accustomed to work, hundredths of a second are too important
to be neglected." And it is to the rare genius of this astronomer,
and to others, that we owe the mechanical accuracy that we now
attain. The clock is made to mark its seconds on paper wrapped
around a revolving cylinder. Under the observer's fingers is an
electric key. This he can touch at the instant of the transit of