ssion, in order to save
the strength of the main party for their final drive. One great
advantage which I had on this expedition was that, owing to the size of
my party, whenever the men in this pioneer division became exhausted
with their arduous labor and lack of sleep, I could withdraw them into
the main party, and send out another division to take their place.

Supporting parties are essential to success because, a single party,
comprising either a small or a large number of men and dogs, could not
possibly drag (in gradually lessening quantities) all the way to the
Pole and back (some nine hundred odd miles) as much food and liquid fuel
as the men and dogs of that party would consume during the journey. It
will be readily understood that when a large party of men and dogs
starts out over the trackless ice to the polar sea, where there is no
possibility of obtaining a single ounce of food on the way, after
several days' marching, the provisions of one or more sledges will have
been consumed by the men and dogs. When this occurs, the drivers and
dogs with those sledges should be sent back to the land at once. _They
are superfluous mouths which cannot be fed from the precious supply of
provisions which are being dragged forward on the sledges._

Still further on, the food on one or two more sledges will have been
consumed. These sledges also, with their dogs and drivers, must be sent
back, in order to ensure the furthest possible advance by the main
party. Later on, still other divisions must be sent back for the same
reason.

But my supporting parties had another duty to perform, only a little
less important than the one already noted; that was to keep the trail
open for the rapid return of the main party.

The magnitude of this duty is clear. The ice of the polar sea is not an
immovable surface. Twenty-four hours--or even twelve hours--of strong
wind, even in the depth of the coldest winter, will set the big floes
grinding and twisting among themselves, crushing up into pressure ridges
in one place, breaking into leads in another place.

Under normal conditions, however, this movement of the ice is not very
great in a period of eight or ten days, so that a party starting back
over an outward trail at the end of several days is able to knit
together all faults and breaks in the trail that have occurred during
that period by reason of the movement of the ice.

The second supporting party, starting back several d