at a target at fixed
intervals of time. If the marksman advances, say twenty paces between
each discharge of his rifle, it is evident that the shots will fall
faster on the target than if he stood still; if, on the contrary, he
retires by the same amount, they will strike at correspondingly longer
intervals. The result will of course be the same whether the target or
the marksman be in movement.

So far Doppler was altogether right. As regards sound, anyone can
convince himself that the effect he predicted is a real one, by
listening to the alternate shrilling and sinking of the steam-whistle
when an express train rushes through a station. But in applying this
principle to the colours of stars he went widely astray; for he omitted
from consideration the double range of invisible vibrations which
partake of, and to the eye exactly compensate, changes of refrangibility
in the visible rays. There is, then, no possibility of finding a
criterion of velocity in the hue of bodies shining, like the sun and
stars, with continuous light. The entire spectrum is slightly shifted up
or down in the scale of refrangibility; certain rays normally visible
become exalted or degraded (as the case may be) into invisibility, and
certain other rays at the opposite end undergo the converse process; but
the sum total of impressions on the retina continues the same.

We are not, however, without the means of measuring this sub-sensible
transportation of the light-gamut. Once more the wonderful Fraunhofer
lines came to the rescue. They were called by the earlier physicists
"fixed lines;" but it is just because they are _not_ fixed that, in this
instance, we find them useful. They share, and in sharing betray, the
general shift of the spectrum. This aspect of Doppler's principle was
adverted to by Fizeau in 1848,[628] and the first tangible results in
the estimation of movements of approach and recession between the earth
and the stars, were communicated by Sir William Huggins to the Royal
Society, April 23, 1868. Eighteen months later, Zoellner devised his
"reversion-spectroscope"[629] for doubling the measurable effects of
line-displacements; aided by which ingenious instrument, and following a
suggestion of its inventor, Professor H. C. Vogel succeeded at Bothkamp,
June 9, 1871,[630] in detecting effects of that nature due to the solar
rotation. This application constitutes at once the test and the triumph
of the method.[631]

The eastern