displacing by its impetus the two
strata of luminous matter beneath, the upper of course to a greater
extent than the lower, and thus wholly or partially denuding the opaque
surface of the sun below. Such processes cannot be unaccompanied by
vorticose motions, which, left to themselves, die away by degrees and
dissipate, with the peculiarity that their lower portions come to rest
more speedily than their upper, by reason of the greater resistance
below, as well as the remoteness from the point of action, which lies in
a higher region, so that their centres (as seen in our waterspouts,
which are nothing but small tornadoes) appear to retreat upwards. Now
this agrees perfectly with what is observed during the obliteration of
the solar spots, which appear as if filled in by the collapse of their
sides, the penumbra closing in upon the spot and disappearing after it."

But when it comes to be asked whether a cause can be found by which a
diversity of solar temperature might be produced corresponding with that
which sets the currents of the terrestrial atmosphere in motion, we are
forced to reply that we know of no such cause. For Sir John Herschel's
hypothesis of an increased retention of heat at the sun's equator, due
to the slightly spheroidal or bulging form of its outer atmospheric
envelope, assuredly gives no sufficient account of such circulatory
movements as he supposed to exist. Nevertheless, the view that the sun's
rotation is intimately connected with the formation of spots is so
obviously correct, that we can only wonder it was not thought of sooner,
while we are even now unable to explain with any certainty _how_ it is
so connected.

Mere scrutiny of the solar surface, however, is not the only means of
solar observation. We have a satellite, and that satellite from time to
time acts most opportunely as a screen, cutting off a part or the whole
of those dazzling rays in which the master-orb of our system veils
himself from over-curious regards. The importance of eclipses to the
study of the solar surroundings is of comparatively recent recognition;
nevertheless, much of what we know concerning them has been snatched, as
it were, by surprise under favour of the moon. In former times, the sole
astronomical use of such incidents was the correction of the received
theories of the solar and lunar movements; the precise time of their
occurrence was the main fact to be noted, and subsidiary phenomena
received but casu