92 
II. T. FERRAR. 
Frost-action . — Owing to the very low temperatures prevalent in high southern 
latitudes, the denuding action of frost is not strikingly conspicuous. As a rule, the 
wind removes all snow from bare rocks, and a marked line always divides the local 
snow-fields from the areas free from snow. Thawing and freezing only occur 
near the edge of snow-fields, and, therefore, owing to the general absence of water, 
frost-action is once more rendered impotent. Castle Rock was perhaps the best 
example of a frost-riven mass near Winter Quarters. It rises sheer above the 
snow-covered peninsula ; but the side facing the sun (the north side) slopes steeply 
down to shore-ice nearly 1000 feet below. Snow is drifted by the prevailing easterly 
wind on to the north side, and in summer large riven blocks fall down and litter the 
area below. The north side of Cathedral Rocks is similarly shattered ; from its 
pinnacled outline it would appear to be more subject to frost-action than the isolated 
peaks further to the west. 
The dolerite, though forming no prominent talus-slopes, appears more prone to 
split than other less-jointed rocks and a conspicuous ledge is always left at its 
contact-junctions. Where dolerite occurs above sandstone, a terrace of the sandstone 
stands out in front of the steep dolerite-cliff. Fans of dolerite- talus are very 
conspicuous in the smaller of the Dry Valleys, also below the hill D, and along the 
south side of the Kukri Hills. 
No screes encroach upon the upper parts of the Ferrar Glacier, and the dolerite 
usually rises perpendicularly from the ice. Along the ice-streams talus-fans are 
rarely abundant enough to become continuous even at their base. The sandstone 
undergoes little frost-action. Often its surfaces still retain the rounded outlines 
which have been produced by ice-action. All transport of rock-material is now 
accomplished by wind, which carries off the sand-grains as fast as they are loosened. 
An important agent in wearing down the sandstone, and one which can hardly 
be classed with any of the ordinary agents of denudation, may be included here. 
The columns of dolerite, in falling down the cliffs, break away the softer sandstone- 
beds and produce a sort of “ chimney talus-shoot,” which conveys the debris to the 
fan at the bottom. Other fragments follow this line of descent, and thus the 
deepening of the gully is accelerated. On the hill x several such gullies may be 
seen. At a height of about 500 feet above the ice, the edges of the sandstone-beds 
which have been caught up by the dolerite (see p. 46) are serrate, and at one 
spot a groove or gully, 20 feet wide, of U-shaped section, has been produced. This 
groove has perpendicular sides, and the uppermost bed of sandstone has been cut 
back 20 feet from the edge of the cliff. 
The granites of Antarctica, as of other regions, seem prone to form screes. 
In the metamorphic limestone area, the hills are usually so rounded that there is 
seldom an opportunity for a loosened block to change its position, and no transport 
takes place until the rock is so finely disintegrated that it can be carried oft’ by 
the wind. 
