Gravel Deposits, Muir Glacier.— Russell. 193 
Trunks and branches of trees are scattered over its surface or im- 
bedded in the superficial layers. Portions of trees probably oc- 
cur throughout the deposit, and beneath it over large areas, there 
must be a buried forest. Conditions are easily conceived whereby 
other forests might become buried above the first. 
The débris carried down by the stream is derived almost en- 
tirely from the moraines on the glacier from which it flows, and 
partakes of their heterogeneous character. Examples of all of 
the rocks bordering the east side of the Atrevida glacier occur in 
the stream deposit. 
Owing to the fact that the stream is overloaded at its source 
and is depositing in that region, it is evident that the apex of the 
cone is rising, and, consequently, that it is increasing in thick- 
ness throughout. The grade of the stream increases towards its 
source and will continue to increase as the ice wall from beneath 
which it flows recedes farther and farther. These conditions, to- 
gether with the constant bifurcation of the stream as it flows down 
the cone, thus decreasing its transporting power, make it evident 
that the cone will continue to grow in all of its dimensions until 
the glacier which feeds it, both with water and débris, ceases to 
recede, or until the ratio between water supply and the amount of 
débris contributed is greatly changed. 
Another boulder-wash which is being formed by Kwik stream, 
at the foot of the Lucia glacier, is several times larger than the 
one described above, but is of the same type. Many others of 
less size were observed at the extremities of the secondary glaciers 
about Disenchantment bay, and about the borders of the Malas- 
pina glacier. These numerous examples seem sufficient to dem- 
onstrate that deposits of boulders, gravel and sand of the nature 
of those noticed above, are dependent on the retreat of the gla- 
ciers about which they are formed. Should a glacier, after re- 
treating and depositing an alluvial cone, become stationary, it is 
evident that the amount of coarse débris contributed to the streams 
draining it would be greatly decreased. The streams would then 
be in a condition to erode the deposits previously laid down, and 
to cut channels through them. Should a glacier then advance, it 
might override or plough into the deposits about its front, and 
thus still further assist in their removal. 
Owing to the rapid disintegration of rocks under the climatic 
conditions prevailing in southern Alaska, alluvial cones, more 
