unding was taken and the course was once more set for the
south.

The sky remained overcast, the atmosphere foggy, and a south-south-east
wind was blowing as we came abreast of the "ice island," which, by the
way, was discovered to have drifted several miles to the north, thus
proving itself to be a free-floating berg. The glacier-tongue on the
port side took a sharp turn to the east-south-east, disappearing on the
horizon. As there was no pack in sight and the water was merely littered
with fragments of ice, it appeared most likely that the turn in the
glacier-tongue was part of a great sweeping curve ultimately joining
with the southward land. On our south-south-east course we soon lost
sight of the ice-cliffs in a gathering fog.

On the afternoon of January 6 the wind abated and the fog began to
clear. At 5 P.M. a line of ice confronted us and, an hour later, the
'Aurora' was in calm water under another mighty ice face trending across
our course. This wall was precisely similar to the one seen on the
previous evening, and might well have been a continuation of it. It is
scarcely credible that when the 'Aurora' came south the following year,
the glacier-tongue first discovered had entirely disappeared. It was
apparently nothing more than a huge iceberg measuring forty miles in
length. Specially valuable, as clearing up any doubt that may have
remained, was its re-discovery the following year some fifty miles to
the north-west. Close to the face of the new ice-wall, which proved to
be a true glacier-tongue, a mud bottom was found at a depth of three
hundred and ninety-five fathoms.

While we were steaming in calm water to the south-west, the massive
front, serrated by shallow bays and capes, passed in magnificent review.
Its height attained a maximum of one hundred and fifty feet. In places
the sea had eaten out enormous blue grottoes. At one spot, several of
these had broken into each other to form a huge domed cavern, the roof
of which hung one hundred feet above the sea. The noble portico was
flanked by giant pillars.

The glacier-tongue bore all the characters of shelf-ice, by which is
meant a floating extension of the land-ice.** A table-topped berg in the
act of formation was seen, separated from the parent body of shelf-ice
by a deep fissure several yards in width.

** Subsequently this shelf-ice formation was found to be a floating
glacier-tongue sixty miles in length, the seaward extension of a lar