336 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
successive phases of eruption in the history of its formation. The 
first was much more violent than the others; it gave birth to a cone 
about 13 kilometers in diameter. The walls of this crater are still 
standing and form about the present cone a circular wall, a sort of 
“Somma,” about 1,800 meters high. To the second stage of activity 
belongs the rim of a higher crater (3,350 meters). Certain coulées, 
bared of snow, can still be distinguished. Finally, the present small 
crater was built in a symmetric position in the interior of the pre- 
ceding one. It is from this crater that the vapors now issue. Many 
other jets of vapor, not visible from the ship, may have been observed 
by Doctor Wilson. The general appearance of the volcano recalls 
very closely that of Etna; the dome shape is much more perceptible 
than in the better known active volcanoes. 
II. GLACIATION. 
The glaciation of Victoria Land is much less intense and less 
exclusive than is generally supposed. This, however, is far from 
meaning that it is mediocre. This glaciation should be studied by 
and for itself, and in comparison with all that is known in the boreal 
world, with the possible exception of certain far distant portions 
of the Arctic region, as the extreme north of Greenland, constitutes 
a new and original type. The very numerous and beautiful photo- 
graphs, and above all the panoramic views published in the account 
of the expedition, present at a glance the full extent of these 
phenomena. 
The type is characterized by features unknown elsewhere. The 
Piedmont glaciers (which we would call simply “ fringing glaciers,” 
or “de rivage”) and floating barriers, which Mr. Ferrar classes 
wrongly, as we believe, with the Piedmont type. It is also char- 
acterized by the unusual and inexplicable manner in which the snow 
is transformed into ice. One can not see directly this transforma- 
tion. The actual climatic conditions are such that thawing, either 
partial or local, is exceptional, and all the surfaces were formed 
either of white granular snow or of compact ice. -Even at the head 
of Ferrar Glacier the transformation of the snow into ice is abso- 
lutely abrupt, and along the foot of the grand cascades the ice 
presents the banded surfaces so characteristic of snow. In the 
slightly elongated depressions there are local accumulations of snow, 
but the line separating the granular snow from the glacier ice is 
always abrupt. If one can not perceive the manner of the trans- 
formation, nevertheless one is obliged to recognize the fact that it 
is accomplished with extraordinary facility. Very little snow 
suffices to give birth to a glacier, and one is struck with the dispro- 
portionate size of the glacial lobes as compared with the extent of 
