which might be easily broken down, still the new fall would be five feet higher than the old one. But, so far as can now be discovered, there is no geological necessity, so to speak, for making any such allowance. In the new cataract the American Fall would still be the higher, and its line across the channel nearly straight. The Canadian Fall would undoubtedly present a curve, but more gradual and uniform than the present horseshoe. But there might possibly occur one new feature in the chasm-channel of the river as the result of future recession. That would be the presence in that channel of rocky islands, similar to that which has already formed just below the American Fall. The points at which these islands would be likely to form are those where the indurated rock of either the Medina or the Niagara group lies near the surface of the water. This probably was the case at the narrow bend below the Whirlpool, before noticed, and from thence up to the outlet of the pool. After considering what must have occurred in the last case, we may form some opinion concerning the probabilities in reference to the first. We can hardly resist the conclusion that masses of fallen rock must have accumulated below the Whirlpool as we now see them under the American Fall. But if so, where are they? The answer to this question brings us to the consideration of the most remarkable phenomenon connected with this wonderful river. To the beholder it is matter of astonishment what can have become of the great mass of earth, rock, gravel, and bowlders, large and small, which once filled the immense chasm that lies below him. He learns that the water for a mile below the Falls is two hundred feet deep, and flows over a mass of fallen rock and stone of great depth lying below it; he sees a chasm of nearly double these dimensions, more than half of which was once filled with solid rock; he beholds the large quantities which have already fallen, which are still defiant, still breasting the ceaseless hammering of the descending flood. For centuries past this process has been going on, until a chasm seven miles long, a thousand feet wide, and, including the secondary banks, more than four hundred feet deep, has been excavated, and the material which filled it entirely removed. How? By what? Frost was the agent, ice was his delver, water his carrier, and the basin of Lake Ontario his dumping-ground. Although there is little likelihood that islands similar to Goat Island have existed in the channel from Lewiston upward, still it is probable that, when the Fall receded from the rocky cape below the Whirlpool up to the pool, it left masses of rock, large and small, lying on the rocky floor and projecting above the surface of the water. As there were no islands above, there were no broken, tumultuous rapids. As has been before remarked, the water poured over in one broad, deep, resistless flood. When frozen by the intense cold of winter, the great cakes of