outh-east limb of the
sun. It was then about 40,000 miles high, and attracted no special
attention. Half an hour later a marvellous transformation had taken
place. During that brief interval the prominence became very brilliant
and doubled its length. For another hour the mighty flame still soared
upwards, until it attained the unprecedented elevation of 350,000
miles--a distance more than one-third the diameter of the great luminary
itself. At this climax the energy of the mighty outbreak seems to have
at last become exhausted: the flame broke up into fragments, and by
12.30--an interval of only two hours from the time when it was first
noticed--the phenomenon had completely faded away.

No doubt this particular eruption was exceptional in its vehemence, and
in the vastness of the changes of which it was an indication. The
velocity of upheaval must have been at least 200,000 miles an hour, or,
to put it in another form, more than fifty miles a second. This mighty
flame leaped from the sun with a velocity more than 100 times as great
as that of the swiftest bullet ever fired from a rifle.

The prominences may be generally divided into two classes. We have first
those which are comparatively quiescent, and in form somewhat resemble
the clouds which float in our earth's atmosphere. The second class of
prominences are best described as eruptive. They are, in fact, thrown up
from the chromosphere like gigantic jets of incandescent material. These
two classes of objects differ not only in appearance but also in the
gases of which they are composed. The cloud-like prominences consist
mainly of hydrogen, with helium and calcium, while many metals are
present in the eruptive discharges. The latter are never seen in the
neighbourhood of the sun's poles, but generally appear close to a
sun-spot, thus confirming the conclusion that the spots are associated
with violent disturbances on the surface of the sun. When a spot has
reached the limb of the sun it is frequently found to be surrounded by
prominences. It has even been possible in a few instances to detect
powerful gaseous eruptions in the neighbourhood of a spot, the
spectroscope rendering them visible against the background of the solar
surface just as the prominences are observed at the limb against the
background of the sky.

In order to photograph a prominence we have, of course, to substitute a
photographic plate for the observer's eye. Owing, however, to the
difficul