pear dark, for the light would
be stopped by a tooth, either at starting or at returning, continually.
At higher speeds of rotation some light would reappear, and at lower
speeds it would also reappear; by noticing, therefore, the precise speed
at which there was constant eclipse the velocity of light could be
determined.

[Illustration: FIG. 73.--Diagram of eye looking at a light reflected in
a distant mirror through the teeth of a revolving wheel.]

This experiment has now been made in a highly refined form by Fizeau,
and repeated by M. Cornu with prodigious care and accuracy. But with
these recent matters we have no concern at present. It may be
instructive to say, however, that if the light had to travel two miles
altogether, the wheel would have to possess 450 teeth and to spin 100
times a second (at the risk of flying to pieces) in order that the ray
starting through any one of the gaps might be stopped on returning by
the adjacent tooth.

Well might the velocity of light be called instantaneous by the early
observers. An ordinary experiment seemed (and was) hopeless, and light
was supposed to travel at an infinite speed. But a phenomenon was
noticed in the heavens by a quick-witted and ingenious Danish
astronomer, which was not susceptible of any ordinary explanation, and
which he perceived could at once be explained if light had a certain
rate of travel--great, indeed, but something short of infinite. This
phenomenon was connected with the satellites of Jupiter, and the
astronomer's name was Roemer. I will speak first of the observation and
then of the man.

[Illustration: FIG. 74.--Fizeau's wheel, shewing the appearance of
distant image seen through its teeth. 1st, when stationary, next when
revolving at a moderate speed, last when revolving at the high speed
just sufficient to cause eclipse.]

Jupiter's satellites are visible, precisely as our own moon is, by
reason of the shimmer of sunlight which they reflect. But as they
revolve round their great planet they plunge into his shadow at one part
of their course, and so become eclipsed from sunshine and invisible to
us. The moment of disappearance can be sharply observed.

Take the first satellite as an example. The interval between successive
eclipses ought to be its period of revolution round Jupiter. Observe
this period. It was not uniform. On the average it was 42 hours 47
minutes, but it seemed to depend on the time of year. When Roemer
observed in