74 
II. T. FERRAR. 
If no ice-stream moved faster than does the Ferrar Glacier the number of 
bergs would be almost negligible. Doubtless the numerous ice-streams crossing 
the Prince Albert Mountains must give rise to a certain number of bergs, but even 
then the number is probably small. The piedmonts-on-land, lying on a flat shore 
and receiving only a small supply of snow, can seldom provide enough ice to form 
icebergs. Cliffs that encircle shore-ice hold snowdrifts which are sometimes as 
much as GO feet thick. These always float out to sea during the summer, and 
produce small bergs which may easily be mistaken for the broken-up masses of 
Fig. 40. — An Iceberg, over 150 feet high, tilted through nearly 90°. 
tabular bergs. Piedmonts-aground do appear to produce icebergs which have 
irregular shapes and are produced in the manner of text-book icebergs. 
On the Balleny Isles, -where the snowfall is much greater, the piedmonts creep 
further out to sea and therefore supply larger bergs. It is, however, to floating 
piedmonts that we must ascribe the great majority of Antarctic bergs, and when 
we remember that the edge of the Ross Piedmont, whose ice is advancing north- 
ward faster than any other ice in South Victoria Land, has lost a belt averao-inc 
15 miles in width during the last sixty-five years, we see that it must have given 
rise to innumerable bergs. 
O 
Piedmonts-afloat, from their configuration, can only supply tabular bergs, and 
the sizes of these vary enormously. Although the Ross Piedmont rises over 200 
