, is one
of the windiest places in the world. In July it was completely frozen
over as far as we could see in the darkness from a height of 900 feet.
Within a few days a hurricane had blown it all away, and the sea was
black.

I believe, and we had experiences to prove me right, that there is a
critical period early in the winter, and that if sea-ice has not frozen
thick enough to remain fast by that time, it is probable that the sea
will remain open for the rest of the year. But this does not mean that no
ice will form. So great is the wish of the sea to freeze, and so cold is
the air, that the wind has only to lull for one instant and the surface
is covered with a thin film of ice, as though by magic. But the next
blizzard tears it out by force or a spring tide coaxes it out by stealth,
whether it be a foot thick or only a fraction of an inch. Such an example
we had at our very doors during our last winter, and the untamed winds
which blew as a result were atrocious.

Thus it is that floes from a few inches to twenty feet thick go voyaging
out to join the belt of ice which is known as the pack. Scott seems to
have thought that the whole Ross Sea freezes over.[52] I myself think
this doubtful, and I am, I believe, the only person living who has seen
the Ross Sea open in mid-winter. This was on the Winter Journey
undertaken by Wilson, Bowers and myself in pursuit of Emperor penguin
eggs--but of that later.

It is clear that winds and currents are, broadly speaking, the governing
factors of the density of pack-ice. By experience we know that clear
water may be found in the autumn where great tracts of ice barred the way
in summer. The tendency of the pack is northwards, where the ice melts
into the warmer waters. But the bergs remain when all traces of the pack
have disappeared, and, drifting northwards still, form the menace to
shipping so well known to sailors rounding the Horn. It is not hard to
imagine that one monster ice island of twenty miles in length, such as do
haunt these seas, drifting into navigated waters and calving into
hundreds of great bergs as it goes, will in itself produce what seamen
call a bad year for ice. And the last stages of these, when the bergs
have degenerated into 'growlers,' are even worse, for then the sharpest
eye can hardly distinguish them as they float nearly submerged though
they have lost but little of their powers of evil.

There are two main types of Antarctic berg. The first