In fact, all the evidence from the Old World goes to confirm the long duration of the post-glacial period, and the immensely greater antiquity of the glacial period taken as a whole. It is only from the New World that any serious arguments are forthcoming to abridge those periods, or rather the post-glacial period, for that alone is affected by the facts adduced. It is said that recent measurements of the recession of the Falls of Niagara show, that instead of requiring 35,000 years, as estimated by Lyell, to cut back the gorge of seven miles from Kingston to the Falls, 10,000 years at the outside would have been amply sufficient; and that this is confirmed by the gorges of other rivers, such as that of the Mississippi at St. Paul's. The evidence is not conclusive, for it depends on the rate of erosion going on for the last twenty or thirty years, which may obviously give a different result from the true average, and in fact older estimates, based on longer periods, gave the rate adopted by Lyell. But if we admit the accuracy of the modern estimates, it does not affect the total duration of the glacial period, but simply that of a late phase of the post-glacial, when the ice-cap which covered North America to a depth often of 2000 or 3000 feet, had melted away and shrunk back 400 miles from its original southern boundary, so as to admit of the waters of the great lakes finding an outlet to the north-east instead of by the old drainage to the south. Nothing is more likely than that, as the great Laurentian ice-cap of America was deeper and extended further than the Scandinavian ice-cap of Europe, it may have taken longer to melt the larger accumulation of ice, and thus postponed the establishment of post-glacial conditions and river-drainage to a later period than in the warmer and more insular climate of Europe. It is a matter of everyday observation, that the larger a snowball is the longer it takes to melt, and that when the mass is large it requires a long time to make it disappear even after mild weather has set in. The only other argument for a short glacial period is drawn from the rate of advance of the glaciers in Greenland, which is shown to be much more rapid than that of the glaciers of Switzerland, from which former calculations had been made. But obviously the rate at which the fronts of glaciers advance when forced by a mass of continental ice down fiords on a steep descending gradient, into a deep sea, where the front is floated off in icebergs, affords no clue as to that of an ice-cap spread, with a front of 1000 miles, over half a continent, retarded by friction, and surmounting mountain chains 3000 feet high. Nor does the rate of advance afford the slightest clue to the time during which the ice-cap may have remained stationary, alternately advanced and retreated, and finally disappeared.[9] [9] The following is the latest pronouncement on the subject from a well-known American geologist:--