Drift Mounds near Olyriipia^Wa><h. — Rogers. 399 
pillars or pedestals of Ice. The stones thus elevated are frequently 
large, and form tables which are nearly always inclined southward. In 
other instances the ice over large areas, especially along the center of 
the medial moraine, was covered with cones of fine, angular fragments 
from a few inches to three or four feet in hight. These were not really 
piles of gravel, as they seemed, but consisted of cones of ice, sheeted over 
with thin layers of small stones. The secret of their formation, long 
since discovered on the glaciers of Switzerland, is that the gravel is first 
concentrated in a hole in the ice and, as the general surface melts away, 
acts like a large stone and protects the ice beneath. It is raised on a 
pedestal, but the gravel at the borders continually rolls down the sides 
and a conical form is the result. 
Allowing conditions like these of the Malaspina, Galiano, and 
Ha^'den glaciers to have existed on the melting ice-sheet of the 
tumuli plain in Washington, we are prepared to look over its area 
and observe thousands of kettle-shaped hollows, pits, and wells in 
the ice, containing water and becoming filled with drift gravel and 
sand. A winter comes and while the temperature is below the 
freezing point the water in the holes is congealed, hence all fur- 
ther action ceases until the spring time brings a higher tempera- 
ture. Then these numberless holes are ready to receive the on- 
coming flood and the fugitive drift matter supplied b}^ the melting 
and receding glacier. It is quite obvious that as the drift debris 
was swept over the ice field, these wells would be more or less 
filled with cla3'ey silt, sand, and fine and coarse gravel, correspond- 
ing to the force of the water current. Perhaps this phenomenon 
was repeated for several or many years. At last the parent glacier 
has become so wasted and remote that no more floods occur, and 
at the same time b}' gradual melting the ice-sheet has quite disap- 
peared, leaving these accumulations of drift matter intact. By 
such processes this large area would be left covered with tumuli 
substantially as the}- exist to-day. It would matter little whether 
the walls of the circular pits or wells in the ice were vertical or in- 
clined at various angles or shaped like an hour-glass, as Russell 
observed on the Malaspina glacier or ice-sheet. The enclosed 
masses of stratified drift, when their supporting ice walls were 
melted away, would naturally assume no other form than the one 
in which they now are found, as broad-based, cone-shaped mounds. 
