scoveries. Of such a
nature is the building of shelters for iceholes. We knew a good deal
about it in the _Discovery_, but unfortunately did not make notes of
our experiences. I sketched the above figures for Nelson, and found on
going to the hole that the drift accorded with my sketch. The sketches
explain themselves. I think wall 'b' should be higher than wall 'a.'

My night on duty. The silent hours passed rapidly and comfortably. To
bed 7 A.M.

_Thursday, June_ 8.--Did not turn out till 1 P.M., then with a bad
head, an inevitable sequel to a night of vigil. Walked out to and
around the bergs, bright moonlight, but clouds rapidly spreading up
from south.

Tried the snow knife, which is developing. Debenham and Gran went
off to Hut Point this morning; they should return to-morrow.

_Friday, June_ 9.--No wind came with the clouds of yesterday, but
the sky has not been clear since they spread over it except for about
two hours in the middle of the night when the moonlight was so bright
that one might have imagined the day returned.

Otherwise the web of stratus which hangs over us thickens and thins,
rises and falls with very bewildering uncertainty. We want theories
for these mysterious weather conditions; meanwhile it is annoying to
lose the advantages of the moonlight.

This morning had some discussion with Nelson and Wright regarding the
action of sea water in melting barrier and sea ice. The discussion
was useful to me in drawing attention to the equilibrium of layers
of sea water.

In the afternoon I went round the Razor Back Islands on ski, a run
of 5 or 6 miles; the surface was good but in places still irregular
with the pressures formed when the ice was 'young.'

The snow is astonishingly soft on the south side of both islands. It
is clear that in the heaviest blizzard one could escape the wind
altogether by camping to windward of the larger island. One sees
more and more clearly what shelter is afforded on the weather side
of steep-sided objects.

Passed three seals asleep on the ice. Two others were killed near
the bergs.

_Saturday, June_ 10.--The impending blizzard has come; the wind came
with a burst at 9.30 this morning.

Simpson spent the night turning over a theory to account for the
phenomenon, and delivered himself of it this morning. It seems a
good basis for the reference of future observations. He imagines the
atmosphere A C in potential equilibrium with large margin of stability,
i