The Kelvin machine gave no bottom at the full length of the line, 370
fathoms. After much labour we made a hole in the ice near the stern-
post large enough for the Lucas machine with a 32-lb. lead; but this
appeared to be too light. The machine stopped at 452 fathoms, leaving
us in doubt as to whether bottom had been reached. Then in heaving up
we lost the lead, the thin wire cutting its way into the ice and
snapping. All hands and the carpenter were busy this day making and
placing kennels on the upper deck, and by nightfall all the dogs were
comfortably housed, ready for any weather. The sun showed through the
clouds above the northern horizon for nearly an hour.

The remaining days of August were comparatively uneventful. The ice
around the ship froze firm again and little movement occurred in our
neighbourhood. The training of the dogs, including the puppies,
proceeded actively, and provided exercise as well as occupation. The
drift to the north-west continued steadily. We had bad luck with
soundings, the weather interfering at times and the gear breaking on
several occasions, but a big increase in the depth showed that we had
passed over the edge of the Weddell Sea plateau. A sounding of about
1700 fathoms on August 10 agreed fairly well with Filchner's 1924
fathoms, 130 miles east of our then position. An observation at noon
of the 8th had given us lat. 71° 23´ S., long. 49° 13´ W. Minus
temperatures prevailed still, but the daylight was increasing. We
captured a few emperor penguins which were making their way to the
south-west. Ten penguins taken on the 19th were all in poor condition,
and their stomachs contained nothing but stones and a few cuttle-fish
beaks. A sounding on the 17th gave 1676 fathoms, 10 miles west of the
charted position of Morell Land. No land could be seen from the mast-
head, and I decided that Morell Land must be added to the long list of
Antarctic islands and continental coasts that on close investigation