and covered by a glacier so rugged that in all probability it would
have been impossible to find foothold on it. Here were also Mounts
Oskar Wisting, Sverre Hassel, and Olav Bjaaland, grandly lighted up
by the rays of the sun. In the distance, and only visible from time
to time through the driving mists, we saw Mount Thorvald Nilsen,
with peaks rising to 15,000 feet. We could only see those parts of
them that lay nearest to us. It took us three days to get over the
Devil's Glacier, as the weather was unusually misty.

On December 1 we left the glacier in high spirits. It was cut up by
innumerable crevasses and holes. We were now at a height of 9,370
feet. In the mist and driving snow it looked as if we had a frozen
lake before us; but it proved to be a sloping plateau of ice, full
of small blocks of ice. Our walk across this frozen lake was not
pleasant. The ground under our feet was evidently hollow, and it
sounded as if we were walking on empty barrels. First a man fell
through, then a couple of dogs; but they got up again all right. We
could not, of course, use our ski on this smooth-polished ice, but we
got on fairly well with the sledges. We called this place the Devil's
Ballroom. This part of our march was the most unpleasant of the whole
trip. On December 2 we reached our greatest elevation. According to
the hypsometer and our aneroid barometer we were at a height of 11,075
feet -- this was in lat. 87deg. 51'. On December 8 the bad weather came
to an end, the sun shone on us once more, and we were able to take our
observations again. It proved that the observations and our reckoning
of the distance covered gave exactly the same result -- namely, 88deg.
16' S. lat. Before us lay an absolutely flat plateau, only broken
by small crevices. In the afternoon we passed 88deg. 23', Shackleton's
farthest south. We pitched our camp in 88deg. 25', and established our
last depot -- No. 10. From 88deg. 25' the plateau began to descend evenly