3388 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
networks of tiny crevasses are formed, with loud explosions, as soon 
as the mountains cast their shadow on the ice. The detonations 
sometimes last for an hour and a half, and crevasses of 50 meters 
length, traversing the surface of masses of ice 30 centimeters thick, 
have been seen produced by the blow of an iron rod. The glaciers of 
Victoria Land are thus in a state of equilibrium, or rather or inertia, 
quite without counterpart and wholly unexpected for these latitudes. 
But though this glaciation may appear to be highly developed, yet 
it is nevertheless in a marked state of recession. The Antartic no 
more escapes the law of glaciary decrease, the proofs of which have 
been accumulated in the last years over the entire globe, than do the 
latitudes of Grahams Land, as reported by the expeditions of the 
‘Belgica and by Nordenskjéld. The north fork of the Ferrar Glacier 
no longer reaches MacMurdo Sound; the glacier has retreated 
far toward the interior of the land, leaving in its place a valley 
bed about 15 kilometers long, encumbered with morainic..materials, 
‘and in which occur, not far from the actual extremity, three con- 
stantly frozen glacial lakes. More striking yet seems the retreat 
in the system of glaciers formerly dependent from Snow Valley. 
This valley occupies the depression which separated the Royal 
Society Range from the gneissic foothills; it served as the gathering 
grounds for a series of glaciers, the sole remnant of which is now the 
Blue Glacier. All the other emissaries of Snow Valley are to-day 
cut off by cross melting from their reservoirs of supply. They exist, 
nevertheless, in the state of ice slabs. Thus, one counts a sym- 
metrical succession of seven ice slabs in the ravines which descend 
precipitously from the gneissic foothills toward the MacMurdo 
Sound. 
Without seeking the cause of such retreat, which seems to be gen- 
eral, and which is also observed, as will be noted later, in the Ross 
Barrier, certain reasons present themselves at the outset to one who 
seeks the explanation of the indolent glacial régime which we have 
just described. First of all is the natural dryness of the climate of 
Victoria Land. This climatic trait is clearly brought out by certain 
geological observations; the insignificance of the role of water in 
effecting erosion and the persistence in place of their formation of 
certain salts, as sulphate of soda and carbonate of lime, due to the 
chemical decomposition of the rocks. A white powder sometimes 
covers the surface of the rocks, and there is found on the floating ice 
heaps 2 feet in height and from 4 to 5 feet in diameter, composed en- 
tirely of glauber salt, due to the progressive freezing out of the salt as 
the water congeals. To this dryness of climate, which is attested also 
by the small quantity of annual snowfall, is added the action of the 
wind, which does not content itself with the slow work of polishing, 
drilling, and chiseling the rocks, as in ordinary desert phenomena ; 
