ght of 16,400 feet.

It took us three days to climb the "Devil's Glacier." On the first of
December we had left behind us this glacier with its crevasses and
bottomless pits and were now at an elevation of 9,350 feet above sea
level. In front of us lay an inclined block-covered ice plateau which,
in the fog and snow, had the appearance of a frozen lake. Traveling
over this "Devil's Ball Room," as we called the plateau, was not
particularly pleasant. Southeasterly storms and snow flurries occurred
daily, during which we could see absolutely nothing. The floor on which
we were walking was hollow beneath us; it sounded as if we were going
over empty barrels. We crossed this disagreeable and uncanny region as
quickly as was compatible with the great care we had to exercise, for
during the whole time we were thinking of the unwelcome possibility of
sinking through.

On December 6th we reached our highest point--according to hypsometric
measurement 11,024 feet above sea level. From there on the interior
plateau remained entirely level and of the same elevation. In 88 deg. 23'
S. we had reached the place which corresponded to Shackleton's
southernmost advance. We camped in 88 deg. 25' S. and established there our
last--the tenth--depot, in which we left 220 pounds of provisions. Our
way now gradually led downward. The surface was in excellent condition,
entirely level, without a single hill or undulation or other obstacle.
Our sleds forged ahead to perfection; the weather was beautiful; we
daily covered seventeen miles. Nothing prevented us from increasing our
daily distance. But we had time enough and ample provisions; we thought
it wiser, also, to spare our dogs and not to work them harder than
necessary. Without a mishap we reached the eighty-ninth parallel on
December 11th. It seemed as if we had come into a region where good
weather constantly prevails. The surest sign of continued calm weather
was the absolutely level surface. We could push a tent-pole seven feet
deep into the snow without meeting with any resistance. This proved
clearly enough that the snow had fallen in equable weather; calm must
have prevailed or a slight breeze may have blown at the most. Had the
weather been variable--calms alternating with storms--snow strata of
different density would have formed, a condition which we would
immediately have noticed when driving in our tent-poles.

Our dead reckoning had heretofore always given the same results a