This variation of temperature must be followed by variation of volume of the earth's crust; that is, it _must produce expansion or contraction_. Experiments have been made, under the direction of the United States government, to determine the expansion of the several kinds of rock used in our public works. It was found that granite expands nearly one two hundred thousandth of its length for every degree of increased temperature, limestone somewhat more than that, and sandstone about twice as much. Taking the expansion of the granite as the basis of calculation, and supposing the crust for a hundred miles in thickness to be undergoing change of temperature, there would be a resulting difference of level exceeding two and a half feet for each degree of change in temperature, or more than two thousand five hundred feet for a change of one thousand degrees. This calculation is made upon the supposition that the law of expansion is the same for all temperatures, and that no new conditions are introduced at high temperatures by the presence of aqueous particles. We know, however, that solids expand more rapidly at high temperatures than at low, and the elasticity of aqueous vapor at high temperatures must increase the rate of expansion of the rock through which it is diffused. Although we are not able to introduce, numerically, the effect of these two circumstances, yet it is obvious that they must be considerable. The mean elevation of land above the level of the sea is about nine hundred feet, the mountain masses above that level not being included; and the estimated mean depth of the ocean, not including its chasms, does not exceed two thousand six hundred feet. The _total elevation of the continental masses_, for which it is necessary to account, does not therefore exceed three thousand five hundred feet. This amount of vertical movement may evidently be produced by the expansion and contraction resulting from changes of temperature.