is an end.

This finality _a parte post_ is instructive. Abstract
considerations, based on geometrical or analytical illustrations,
question the finiteness of some physical developments. Thus our
sun may require eternal time to attain the temperature of the
ether around it, the approach to this condition being assumed to
be asymptotic in

289

character. But consider the legitimate _reductio ad absurdum_ of
an ember raked from a fire 1000 years ago. Is it not yet cooled
down to the constant temperature of its surroundings? And we may
evidently increase the time a million-fold if we please. It
appears as if we must regard eternity as outliving every
progressive change, For there is no convergence or enfeeblement
of time. The ever-flowing present moves no differently for the
occurrence of the mightiest or the most insignificant events. And
even if we say that time is only the attendant upon events, yet
this attendant waits patiently for the end, however long
deferred.

Does the essentially material hypothesis of Kant and Laplace
account for an infinite past as thinkably as it accounts for the
infinite future? As this hypothesis is based upon material
instability the question resolves itself into this:-- Is the
assumption of an infinitely prolonged past instability a probable
or possible account of the past? There are, it appears to me,
great difficulties involved in accepting the hypothesis of
infinitely prolonged material instability. I will refer here to
three principal objections. The first may be called a
metaphysical objection; the second is partly metaphysical and
partly physical, the third may be considered a physical
objection, as it is involved directly in the phenomena presented
by our universe.

The metaphysical objection must have presented itself to every
one who has considered the question. It may

290

be put thus:--If present events are merely one stage in an
infinite progress, why is not the present stage long ago passed
over? We are evidently at liberty to push back any stage of
progress to as remote a period as we like by putting back first
the one before this and next the stage preceding this, and so on,
for, by hypothesis, there is no beginning to the progress.

Thus, the sum of passing events constituting the present universe
should long ago have been accomplished and passed away. If we
consider alternative hypotheses not involving this difficulty, we
are at once struck by the fact th