fter having been transmitted through that medium which we call the
ether. To the different colours correspond different wave-lengths--that
is to say, different distances between two successive waves. A beam of
white light is formed by the union of innumerable different waves whose
lengths have almost every possible value lying between certain limits.
The wave-length of red light is such that there are 33,000 waves in an
inch, while that of violet light is but little more than half that of
red light. The position of a line in the spectrum depends solely on the
wave-length of the light to which it is due. Suppose that the source of
light is approaching directly towards the observer; obviously the waves
follow each other more closely than if the source were at rest, and the
number of undulations which his eye receives in a second must be
proportionately increased. Thus the distance between two successive
ether waves will be very slightly diminished. A well-known phenomenon of
a similar character is the change of pitch of the whistle of a
locomotive engine as it rushes past. This is particularly noticeable if
the observer happens to be in a train which is moving rapidly in the
opposite direction. In the case of sound, of course, the vibrations or
waves take place in the air and not in the ether. But the effect of
motion to or from the observer is strictly analogous in the two cases.
As, however, light travels 186,000 miles a second, the source of light
will also have to travel with a very high velocity in order to produce
even the smallest perceptible change in the position of a spectral line.

We have already seen that enormously high velocities are by no means
uncommon in some of these mighty disturbances on the sun; accordingly,
when we examine the spectrum of a sun-spot, we often see that some of
the lines are shifted a little towards one end of the spectrum and
sometimes towards the other, while in other cases the lines are seen to
be distorted or twisted in the most fantastic manner, indicating very
violent local commotions. If the spot happens to be near the centre of
the sun's disc, the gases must be shooting upwards or downwards to
produce these changes in the lines. The velocities indicated in
observations of this class sometimes amount to as much as two or even
three hundred miles per second. We find it difficult to conceive the
enormous internal pressures which are required to impel such mighty
masses of gases al