While I was surveying, the ice, which was sixteen inches thick, undulated under a slight wind like water. It is well known that a level cannot be used on ice. At one rod from the shore its greatest fluctuation, when observed by means of a level on land directed toward a graduated staff on the ice, was three quarters of an inch, though the ice appeared firmly attached to the shore. It was probably greater in the middle. Who knows but if our instruments were delicate enough we might detect an undulation in the crust of the earth? When two legs of my level were on the shore and the third on the ice, and the sights were directed over the latter, a rise or fall of the ice of an almost infinitesimal amount made a difference of several feet on a tree across the pond. When I began to cut holes for sounding there were three or four inches of water on the ice under a deep snow which had sunk it thus far; but the water began immediately to run into these holes, and continued to run for two days in deep streams, which wore away the ice on every side, and contributed essentially, if not mainly, to dry the surface of the pond; for, as the water ran in, it raised and floated the ice. This was somewhat like cutting a hole in the bottom of a ship to let the water out.