90 
ICE-PERIOD IN AMERICA. 
ing country, and a comparison of the na¬ 
ture and level of the drift on the mainland 
with those of the same deposits on the har¬ 
bor-islands, suggest a different explanation 
of these phenomena. The sheet of drift 
was once more continuous and extensive than 
it is now, and the localities in which we 
find these crops of boulders are spots where 
the tide has eaten into the drift, wearing away 
the finer materials, or the paste in which the 
larger fragments were imbedded, and allowing 
them to fall to the bottom, or where the same 
result has been produced by the action of 
rivers cutting their way through the drift, and 
thus finding an outlet to the sea. In short, 
instead of showing the power of currents to 
carry along heavy fragments, these stranded 
boulders prove, on the contrary, the inability 
of water to produce any such effect, since it 
is evident that the tides washing against the 
shore, or the rivers rushing down to the 
sea, were equally incapable of bearing off the 
weightier materials, and allowed them to drop 
to the bottom, while they readily swept away 
the lighter ones. Such localities compare with 
the surrounding drift much as the bottom of a 
