The Glacial Theory of Prof. Agassiz: 363 
gradual fusion and subsidence of the ice, the later boulders were stop- 
ped in their motion by that mountain and settled on its southern decliv- 
ities. He seems, however, we rather think, to mean, that the glaciers 
of the Aar, the Kaisdei: and the Rhone, were lateral and. auxiliary to 
that of the great valley ; that the dilatation of the ice (and the motion 
of the boulders) following the course of the troughs in which it lay, 
was northwest in the lateral valleys, and northeast or southwest in the 
great valley ; and that the blocks resting on Jura are to be considered 
as stranded on one side of the great glacier, the motion of the eastern 
portion of it being northeast, while that of the western was southwest. 
We see some objections to this conclusion. The transference of blocks 
from B, for instance, (figure 8,) should not have been right across to 
2-2, but diagonally to 3, or 1-1, according as the expansive motion of 
the ice was northeast or southwest. A theory, however, which explains 
so many facts, is not to be rejected on account of minor difficulties, 
which future researches may clear up. 
When the mer de glace was melting, the first openings through it 
would be formed where it was thinnest. The water engulfed in these 
would seek out channels where the fissures or vaults under the ice left 
room for it, and valleys of erosion would thus be excavated, sometimes 
at variance with the natural declivity of the ground, and which would 
afterwards become the channels of rivers. Such valleys do occur, and 
the explanation is simple and probable. But account should have been 
taken of the heat developed along the fissure of upheaval, which would 
produce floods of water at the most elevated points; for when the 
granite ascended from below, though it was in a solid state, it must have 
brought with it the temperature of the region from which it came. The 
heat thus generated must have been increased by the enormous friction 
on the pre-existing primary strata, when they were fractured and bent 
up; and the ice in contact with these strata, which surrounded the high- 
est summits of the Alps, must have been first melted. Here was an 
‘obvious source of formidable debacles, which must have produced great 
changes on the surface of the adjacent countries. 
As portions of the old alluvium, containing bones of the fossil ele- 
phant, have been found turned up on the flanks of the Alps, Agassiz 
infers that deposits of clay and gravel existed before the icy envelope 
was formed ; that these must have been broken up and remodelled by 
the streams arising from the fusion of the ice; and, consequently, that 
part of the existing alluvial cover is derived from the wrecks of one 
more ancient. : 
When the ice retired from the great valley or low country, into the 
lateral valleys of the Rhone, the Rhine, the Aar, and others, the for- 
mation of moraines would begin; and the clay, sand, and gravel thus 
_ Collected at particular localities would be dispersed and remodelled by 
