bove the earth's attraction that it is separating itself from the
sphere. If this view be correct, it seems likely that we may look to
great volcanic explosions as a source whence the dustlike particles
which people the celestial spaces may have come. They may, in a word,
be due to volcanic explosions occurring on this and other celestial
spheres.

The question suggested above as to the possibility of volcanic
ejections throwing matter from the earth beyond the control of its
gravitative energy is one of great scientific interest. Computations
(not altogether trustworthy) show that a body leaving the earth's
surface under the conditions of a cannon ball fired vertically upward
would have to possess a velocity at the start of at least seven miles
a second in order to go free into space. It would at first sight seem
that we should be able to reckon whether volcanoes can propel earth
matter upward with this speed. In fact, however, sufficient data are
not obtainable; we only know in a general way that the column of
vapour rises to the height of thirty or forty thousand feet, and this
in eruptions of no great magnitude. In an accident such as that at
Krakatoa, even if an observer were near enough to see clearly what was
going on, the chance of his surviving the disturbance would be small.
Moreover, the ascending vapours, owing to their expansion of the steam
in the column, begin to fly out sideways on its periphery, so that the
upper part of the central section in the discharge is not visible from
the earth.

It is in the central section of the uprushing mass, if anywhere, that
the dust might attain the height necessary to put it beyond the
earth's attraction, bringing it fairly into the realm of the solar
system, or to the position where its own motion and the attraction of
the other spheres would give it an independent orbital movement about
the sun, or perhaps about the earth. We can only say that observations
on the height of volcanic ejections are extremely desirable; they can
probably only be made from a balloon. An ascension thus made beyond
the cloud disk which the eruption produces might bring the observer
where he could discern enough to determine the matter. Although the
movements of the rocky particles could not be observed, the colour
which they would give to the heavens might tell the story which we
wish to know. There is evidence that large masses of stone hurled up
by volcanic eruption have fallen seven mi