seemed rather stuffy in the hut last night--I
found it difficult to sleep, and noticed a good many others in like
case. I found the temperature was only 50 deg., but that the small uptake
on the stove pipe was closed. I think it would be good to have a
renewal of air at bed time, but don't quite know how to manage this.

It was calm all night and when I left the hut at 8.30. At 9 the wind
suddenly rose to 40 m.p.h. and at the same moment the temperature rose
10 deg.. The wind and temperature curves show this sudden simultaneous
change more clearly than usual. The curious circumstance is that
this blow comes out of a clear sky. This will be disturbing to our
theories unless the wind drops again very soon.

The wind fell within an hour almost as suddenly as it had arisen; the
temperature followed, only a little more gradually. One may well wonder
how such a phenomenon is possible. In the middle of a period of placid
calm and out of a clear sky there suddenly rushed upon one this volume
of comparatively warm air; it has come and gone like the whirlwind.

Whence comes it and whither goeth?

Went round the bergs after lunch on ski--splendid surface and quite
a good light.

We are now getting good records with the tide gauge after a great
deal of trouble. Day has given much of his time to the matter,
and after a good deal of discussion has pretty well mastered the
principles. We brought a self-recording instrument from New Zealand,
but this was passed over to Campbell. It has not been an easy matter
to manufacture one for our own use. The wire from the bottom weight
is led through a tube filled with paraffin as in _Discovery_ days,
and kept tight by a counter weight after passage through a block on
a stanchion rising 6 feet above the floe.

In his first instrument Day arranged for this wire to pass around a
pulley, the revolution of which actuated the pen of the recording
drum. This should have been successful but for the difficulty of
making good mechanical connection between the recorder and the
pulley. Backlash caused an unreliable record, and this arrangement
had to be abandoned. The motion of the wire was then made to actuate
the recorder through a hinged lever, and this arrangement holds, but
days and even weeks have been lost in grappling the difficulties of
adjustment between the limits of the tide and those of the recording
drum; then when all seemed well we found that the floe was not rising
uniformly with th