Everyone knows that ice floats in water, is able to carry some amount of materials attached to it, and if lifted by rise of the water-level is able to lift them with itself. Lyell describes this in the Principles (ch. xvi, xxxi), and probably it is mentioned in every textbook. But ice is also able to lift attached materials without rise of the water-level. It requires no tides, no floods; it can lift without external aid, by its melting alone.

When a crust of ice 1 foot thick has formed on a sheet of fresh water it floats with its upper surface about an inch above water-level, its lower surface about 11 inches below. If by any cause this ice-crust is melted over its upper surface, the products of melting flow off, the upper surface sinks a little, but the lower surface rises through about eleven-twelfths of the thickness melted off. Take an ice-floe uniformly 6 feet thick: its lower surface will be floating, if the water be fresh, at about 5½ feet below water-level. Suppose it reduced by melting over the upper surface till only 6 inches thick, the lower surface will now be about 5½ inches below water-level; this lower surface will accordingly have risen more than 5 feet. If any objects or materials were attached to that lower surface they will have risen with it: they will have been lifted. This is undeniable and obvious. It is undeniable and obvious that ice is able to lift materials by simple melting, independently of any other power or action.