es in a rushing torrent
of air, pulsating with mighty gust-waves. Lowered from the estate
of upright manhood, they humbly crawl, or make a series of crouching
sprints between the gusts. Over the scattered boulders to the east of
the Hut, across a patch of polished snow they push to the first low
ridge, and there they stop for breath. Up on the side of "Annie Hill,"
in the local phrase, the tide sweeps by with fiendish strength, and
among the jagged rocks the man clutching the puffometer-box has a few
desperate falls. At last both clamber slowly to an eminence where a long
steel pipe has been erected. To the top of this the puffometer is hauled
by means of a pulley and line. At the same time the aluminium sphere is
released, and out it floats in the wind tugging at the spring.

The puffometer was left out for an hour at a time, and separate gusts up
to one hundred and fifty and one hundred and eighty miles per hour were
commonly indicated. I remember the final fate of this invention. While
helping to mount it one day, the wind picked me up clear of the ground
and dashed myself and the instrument on some rocks several yards away.
The latter was badly damaged, but thick clothing saved me from serious
injury.

[TEXT ILLUSTRATION]

The wind velocity and wind direction charts for Midwinter's Day,
when the steady south-by-east gale was broken after noon by a welcome
lull--the wind veering the while all round the compass.

The average velocity for the day 66.9 miles per hour, and the maximum of
the average hourly velocities, ninety-six miles.

The steadiness of the temperature was a subject for debate. The stronger
the wind blew, the less variation did the thermometer show. Over a
period of several days there might be a range of only four or five
degrees. Ordinarily, this might be expected of an insular climate, but
in our case it depended upon the fact that the wind remained steady from
the interior of the vast frigid continent. The air which flowed over the
Hut had all passed through the same temperature-cycle. The atmosphere
of the interior, where the plateau stood at an elevation of, say, eight
thousand feet, might have a temperature -45 degrees F. As the air
flowed northwards over Adelie Land to the sea, it would rise slowly in
temperature owing to the increased barometric pressure consequent on the
descending gradient of the plateau. At sea-level the temperature of the
river of air would be, approximately, -20 degr