1882.[468] Spots
are manifestly associated with violent eruptive action, giving rise to
the faculae and prominences which usually garnish their borders. It is
accordingly contended that upon the withdrawal of matter from below by
the flinging up of a prominence must ensue a sinking-in of the surface,
into which the partially cooled erupted vapours rush and settle,
producing just the kind of darkening by increased absorption told of by
the spectroscope. Round the edges of the cavity the rupture of the
photospheric shell will form lines of weakness provocative of further
eruptions, which will, in their turn, deepen and enlarge the cavity. The
phenomenon thus tends to perpetuate itself, until equilibrium is at last
restored by internal processes. A sun-spot might then be described as an
inverted terrestrial volcano, in which the outbursts of heated matter
take place on the borders instead of at the centre of the crater, while
the cooled products gather in the centre instead of at the borders.

But on the earth, the solid crust forcibly represses the steam gathering
beneath until it has accumulated strength for an explosion, while there
is no such restraining power that we know of in the sun. Zoellner,
indeed, adapted his theory of the solar constitution to the special
purpose of procuring it; yet with very partial success, since almost
every new fact has proved adverse to his assumptions. Volcanic action is
essentially spasmodic. It implies habitual constraint varied by
temporary outbreaks, inconceivable in a gaseous globe, such as we
believe the sun to be.

If the "volcanic hypothesis" represented the truth, no spot could
possibly appear without a precedent eruption. The real order of the
phenomenon, however, is exceedingly difficult to ascertain; nor is it
perhaps invariable. Although, in most cases, the "opening" shows first,
that may be simply because it is more easily seen. According to Father
Sidgreaves,[469] the disturbance has then already passed the incipient
stage. He considers it indeed "highly probable that the preparatory sign
of a new spot is always a small, bright patch of facula."

This sequence, if established, would be fatal to Lockyer's theory of
sun-spots, communicated to the Royal Society, May 6, 1886,[470] and
further developed some months later in his work on _The Chemistry of the
Sun_. Spots are represented in it as incidental to a vast system of
solar atmospheric circulation, starting with the po