longer than 10,000 years, and that indeed its age, which is also the entire age of the Niagara river and falls, may be a considerably shorter period.

It seems very clear to my mind that the gorge erosion above the whirlpool was much aided by the preglacial St. David's stream for the distance of one mile occupied by the great rapids. Here the major part of the depth and width of the gorge had probably been already eroded before the Ice age, being then filled with drift, which the postglacial river easily removed as soon as its gorge toward Lewiston was sufficiently deepened. No powerful falls have there cut a deep channel, and the river consequently has a constricted and very rapid course. Above the old St. David's ravine, however, a massive waterfall has operated along the latest distance of nearly two miles of the gorge, giving to the river there its great depth.

The action of a high waterfall, with great volume of water, precipitated over a hard rock stratum, of which large blocks give way and fall because they are gradually undermined, as in the Horseshoe falls, is well compared by McGee to the deep wearing of potholes. The fallen blocks are moved under the powerful impact of the high cataract and wear a deep channel, attaining near the foot of the present falls the depth of almost 200 feet under the river level. Such cataract action of deep channel wearing may be supposed also to have produced the great depth of the Niagara river at the mouth of the gorge; but I think that this is better attributed to the usual process of stream cutting at the time of depressed level of this part of Lake Ontario, which is otherwise known by its lower inclined beaches extending here under the lake.

Among the conditions which might cause the Niagara river to vary from its present size, only one would produce a great and long continued diminution of the river, so giving for a large part of its history only very slow erosion of the gorge. This hypothetical factor in our problem, which has been assumed by Gilbert, Spencer, Taylor, and Hitchcock, to considerably prolong the time of the gorge erosion, is the diversion of the outflow from the basins of the three lakes above Lake Erie, then confluent and forming the glacial Lake Algonquin, to forsake its present course and pass eastward from Georgian Bay, at first by the way of Lake Simcoe and the Trent river to Lake Ontario, and later by Lake Nipissing and the Mattawa river to the Ottawa.

But differential elevation of the land from its late glacial or Champlain depression here, as on the area of Lake Agassiz, which is now drained by the Red River of the North to Lake Winnipeg and thence by the Nelson river to Hudson bay, took place as soon as the land was unburdened by the glacial retreat. This northward uplift was in progress while yet the ice barrier remained farther north and northeast, holding in succession the glacial lakes Warren and Algonquin, besides several earlier and smaller glacial lakes which became merged in Lake Warren, on the upper part of the St. Lawrence river basin. In the areas of Lake Agassiz and of the Laurentian lakes alike, the uplift was nearly completed during the existence of the glacial lakes, as is known by the almost undisturbed horizontality of the latest and lowest glacial lake beaches. Finally Lake Algonquin, by the northeastward land elevation, became divided into its successors, Lakes Huron, Michigan and Superior.

longer than 10,000 years, and that indeed its age, which is also the entire age of the Niagara river and falls, may be a considerably shorter period.

It seems very clear to my mind that the gorge erosion above the whirlpool was much aided by the preglacial St. David's stream for the distance of one mile occupied by the great rapids. Here the major part of the depth and width of the gorge had probably been already eroded before the Ice age, being then filled with drift, which the postglacial river easily removed as soon as its gorge toward Lewiston was sufficiently deepened. No powerful falls have there cut a deep channel, and the river consequently has a constricted and very rapid course. Above the old St. David's ravine, however, a massive waterfall has operated along the latest distance of nearly two miles of the gorge, giving to the river there its great depth.

The action of a high waterfall, with great volume of water, precipitated over a hard rock stratum, of which large blocks give way and fall because they are gradually undermined, as in the Horseshoe falls, is well compared by McGee to the deep wearing of potholes. The fallen blocks are moved under the powerful impact of the high cataract and wear a deep channel, attaining near the foot of the present falls the depth of almost 200 feet under the river level. Such cataract action of deep channel wearing may be supposed also to have produced the great depth of the Niagara river at the mouth of the gorge; but I think that this is better attributed to the usual process of stream cutting at the time of depressed level of this part of Lake Ontario, which is otherwise known by its lower inclined beaches extending here under the lake.

Among the conditions which might cause the Niagara river to vary from its present size, only one would produce a great and long continued diminution of the river, so giving for a large part of its history only very slow erosion of the gorge. This hypothetical factor in our problem, which has been assumed by Gilbert, Spencer, Taylor, and Hitchcock, to considerably prolong the time of the gorge erosion, is the diversion of the outflow from the basins of the three lakes above Lake Erie, then confluent and forming the glacial Lake Algonquin, to forsake its present course and pass eastward from Georgian Bay, at first by the way of Lake Simcoe and the Trent river to Lake Ontario, and later by Lake Nipissing and the Mattawa river to the Ottawa.

But differential elevation of the land from its late glacial or Champlain depression here, as on the area of Lake Agassiz, which is now drained by the Red River of the North to Lake Winnipeg and thence by the Nelson river to Hudson bay, took place as soon as the land was unburdened by the glacial retreat. This northward uplift was in progress while yet the ice barrier remained farther north and northeast, holding in succession the glacial lakes Warren and Algonquin, besides several earlier and smaller glacial lakes which became merged in Lake Warren, on the upper part of the St. Lawrence river basin. In the areas of Lake Agassiz and of the Laurentian lakes alike, the uplift was nearly completed during the existence of the glacial lakes, as is known by the almost undisturbed horizontality of the latest and lowest glacial lake beaches. Finally Lake Algonquin, by the northeastward land elevation, became divided into its successors, Lakes Huron, Michigan and Superior.

longer than 10,000 years, and that indeed its age, which is also the entire age of the Niagara river and falls, may be a considerably shorter period.

It seems very clear to my mind that the gorge erosion above the whirlpool was much aided by the preglacial St. David's stream for the distance of one mile occupied by the great rapids. Here the major part of the depth and width of the gorge had probably been already eroded before the Ice age, being then filled with drift, which the postglacial river easily removed as soon as its gorge toward Lewiston was sufficiently deepened. No powerful falls have there cut a deep channel, and the river consequently has a constricted and very rapid course. Above the old St. David's ravine, however, a massive waterfall has operated along the latest distance of nearly two miles of the gorge, giving to the river there its great depth.

The action of a high waterfall, with great volume of water, precipitated over a hard rock stratum, of which large blocks give way and fall because they are gradually undermined, as in the Horseshoe falls, is well compared by McGee to the deep wearing of potholes. The fallen blocks are moved under the powerful impact of the high cataract and wear a deep channel, attaining near the foot of the present falls the depth of almost 200 feet under the river level. Such cataract action of deep channel wearing may be supposed also to have produced the great depth of the Niagara river at the mouth of the gorge; but I think that this is better attributed to the usual process of stream cutting at the time of depressed level of this part of Lake Ontario, which is otherwise known by its lower inclined beaches extending here under the lake.