352 THE JOURNAL OF GEOLOGY. 
sidered strongly against the theory of glacial dams, and favors the 
inference that the submergence was an invasion of the sea through a 
strait over Lake Nipissing. 
Previous estimates of the age of Niagara Falls are given in the 
paper by Dr. Spencer, and attention is called to the error introduced 
into the later computations by neglecting to take into account the 
changing episodes of the river. It is noted that previous writers have 
overlooked the presence of an ancient drainage of Lake Erie about 
forty miles west of the Niagara. Moreover, the assumption often 
made that the old course of this river passed through the whirlpool 
ravine is shown to be erroneous. 
There was no preglacial Niagara River, and the present channel has 
been cut almost entirely in limestone. The Horseshoe falls during 
forty-eight years has shown a mean rate of recession of 4.175 feet per 
year. The American falls, however, has retreated but 0.64 feet per 
year during this period. Four different episodes are recognized. The 
first represents the recession of the falls from the escarpment to the 
level of the Iroquois beach, which is computed to have occupied 17,200 
years. The second stage began with the lowering of the water at the 
end of the first and the recession of the falls to the vicinity of the 
whirlpool—a period of about 10,000 years it is thought. The third 
stage was the time passed at the whirlpool rapids—computed to be 
800 years. The fourth episode is characterized by the rising of the 
waters in the Ontario basin. In the first part of this episode the river 
cut its way through a ridge (Johnson’s) of limestones, following which 
comes the modern stage of the falls. The duration of this epoch is 
placed at 3000 years, making a total of 31,000 years, or, allowing rooo 
years for the time before the advent of the falls, the age of the river is 
32,000 years. 
In the deformatory elevation of the district, Johnson’s ridge was 
raised twenty-four feet above the Chicago divide, causing an overflow 
in this direction which threatened to end the falls when the cut through 
the ridge was effected. On this basis the drainage of Lake Michigan 
by way of the Des Plaines ended about 1500 years ago. At the present 
rate of terrestrial deformation, the falls will come to an end in about 
5000 years, by the turning of the waters into the Mississippi. The 
conclusions of the paper are based on along series of observations 
which have been given in detail in a number of papers, a list of which 
accompanies the present article. Cy HG: 
