ever the moon rotated on its axis with respect to the
earth, and if it were either fluid itself or possessed any liquid ocean,
then the tides caused by the pull of the earth must have been
prodigious, and would tend to stop its rotation. Have they not
succeeded? Is it not probable that this is _why_ the moon always now
turns the same face towards us? It is believed to be almost certainly
the cause. If so, there was a time when the moon behaved
differently--when it rotated more quickly than it revolved, and
exhibited to us its whole surface. And at this era, too, the earth
itself must have rotated a little faster, for it has been losing speed
ever since.

We have thus arrived at this fact, that a thousand years ago the day was
a trifle shorter than it is now. A million years ago it was, perhaps, an
hour shorter. Twenty million years ago it must have been much shorter.
Fifty million years ago it may have been only a few hours long. The
earth may have spun round then quite quickly. But there is a limit. If
it spun too fast it would fly to pieces. Attach shot by means of wax to
the whirling earth model, Fig. 110, and at a certain speed the cohesion
of the wax cannot hold them, so they fly off. The earth is held together
not by cohesion but by gravitation; it is not difficult to reckon how
fast the earth must spin for gravity at its surface to be annulled, and
for portions to fly off. We find it about one revolution in three hours.
This is a critical speed. If ever the day was three hours long,
something must have happened. The day can never have been shorter than
that; for if it were, the earth would have a tendency to fly in pieces,
or, at least, to separate into two pieces. Remember this, as a natural
result of a three-hour day, which corresponds to an unstable state of
things; remember also that in some past epoch a three-hour day is a
probability.

If we think of the state of things going on in the earth's
atmosphere, if it had an atmosphere at that remote date, we shall
recognize the existence of the most fearful tornadoes. The trade
winds, which are now peaceful agents of commerce, would then be
perpetual hurricanes, and all the denudation agents of the
geologist would be in a state of feverish activity. So, too, would
the tides: instead of waiting six hours between low and high tide,
we should have to wait only three-quarters of an hour. Every
hour-and-a-half the water w