the ice. Wherever these
"kettle holes" are found large bodies of ice have become anchored,
while the ice behind it has carried the drift until it is covered over
and piled up at the sides. When these ice mountains melted away
depressions were left which in some cases have resulted in lakes, and in
others simply dry kettle holes. This process has been hinted at in a
former chapter, but we give it here as one of the kinds of lakes formed
during the glacial period. They are found everywhere that glacial action
has prevailed. They are found in great abundance in some parts of New
England on the margin of the terminal moraine. These lakes, however, are
comparatively insignificant as compared with the great inland seas like
Lake Superior and Lake Michigan, that undoubtedly owe their origin
largely to the ice age.

There are other factors, however, that enter into the formation of the
great chain of lakes on the northern boundary of the United States
besides those mentioned, that have brought into existence the smaller
inland lakes.

Glacial lakes may be divided into three classes. Those found in the
"kettle holes" of the terminal or medial moraines, and those that are
formed by the deposition of the glacial drift, as, for instance, Devil's
Lake, and those that are caused by ice forming dams across the valley of
a river that lasted only during the ice age. In some lakes of the
second class erosion undoubtedly entered into their formation as well as
the piling up of glacial drift.

In order, however, that we may understand more fully the formation of
these greater lakes it will be necessary for us to go back and examine
the conditions that seem to have existed before the glacial period.

It is a fact well known that continents have periods of elevation and
depression. There is abundant evidence that the northern portion of the
North American continent was elevated to a much higher level in
preglacial times than it occupies now. This is evidenced in very many
ways by sounding the depths of old river beds now filled with glacial
debris. The old beds show unmistakable evidences of having been worn
down to their present level by the action of running water. They also
prove to be many feet below the present sea-level. This fact seems to be
sufficient to prove the theory of a higher elevation of the North
American continent in preglacial times. It should be said here that
undoubtedly the constant filling up of the ocean with the