They are not lateral moraines, for these are borne on the glacier itself, or else stranded on the deep canyon sides.

Neither do I think moraines of this kind would be formed by a glacier emerging from a steep narrow canyon and running out on a level plain; for in such cases, as soon as the confinement of the bounding walls is removed, the ice stream spreads out into an _ice lake_.
It does so as naturally and necessarily as does water under similar circumstances.

The deposit would be nearly transverse to the direction of the motion, and, therefore, more or less crescentic.

There must be something peculiar in the conditions under which these parallel ridges were formed.

I believe the conditions were as described below.
We have already given reason to think that the original margin of the Lake, in glacial times, was three or four miles back from the present margin, along the series of rocky points against which the ridges abut; and that all the flat plain thence to the present margin is made land.