68 
H. T. FERRAK. 
with, but close to the shore crevasses and pressure-ridges show massive and vesicular 
ice. The vesicular ice contains air * amounting on the average to about 8'5 per cent, 
of its own volume, and the ice-grains are usually less than a quarter of an inch 
across. Near White Island the grains are at least half an inch across and of a 
very uniform size. This occurrence, however, is exceptional, for though no running 
water was seen, the patches of bare ice are quite saturated with water, and it has 
already been proved by Herr Em den f that growth of glacier-grains takes place 
most rapidly at or near the freezing point. All observations made seem to show 
Fig. 38. — The Ross Piedmont from the side of Mount Terror, 
SHOWING THE CLIFF-EDGE AND FLAT Ul’I’ER SURFACE. 
that the Ross piedmont is produced by ice-streams, not able to melt upon land, being 
pushed out to unite in a shallow bay, after which the ice-mass floats off towards 
the deep sea. It is remarkable that along the whole cliff-face from Cape Crozier 
to Cape Colbeck no trace of foreign matter in the ice could be observed. At the 
eastward end the land is completely buried in snow, but along the west side the 
land is comparatively bare. Rock-rie/im was never met with more than a very few 
miles from land. The chasm J that skirts the west side is in itself sufficient to 
* Heim, ‘ Handbueh der Gletscherkunde,’ 1885, p. 113. 
f Emden, Neue Denkschr. schweiz. nat. Ges. 1893, Bd. xxxiii, Abth. 1. 
t Scott, Geog. Journ., April 1905, vol. xxv, p. 366, plate. 
