52 



becomes more acute and then less acute in its general form. 

 When acuteness is increasing, the rate is a little more 

 rapid than the mean; when it is decreasing it is a little 

 slower than the mean. 



Wherever the rock floor above the falls was unusually 

 flat and wide and without local depressions or small valleys 

 to cause concentration of water in greater depth at some 

 particular place, the water sheet on the crest became 

 relatively thin, the falls widened and the gorge was made 

 wider and proportionally shallower. At such times the 

 rate of recession must have been considerably slower, for 

 the undermining power of the falls was reduced. Such 

 a period affected the falls, while the main cataract was 

 passing in front of the American fall, and Goat island, 

 where the gorge is wider than the average and shallower. 

 The rate here in the widest part was probably reduced 

 one-half. 



The slower rate in the wider, shallower part opposite 

 the American fall and Goat island somewhat reduces the 

 average rate for the whole section making it slightly less 

 than Mr. Gilbert's five feet (i -5 m.) a year and indicating 

 a period of 2,700 to 3,000 years for the making of the 

 Upper Great gorge. 



But the most uncertain element in attempting to 

 determine the whole duration of the falls is the rate in 

 the two Erie sections, the Old Narrow gorge and the gorge 

 of the Whirlpool rapids; and the oldest section near Queen- 

 ston. Spencer makes the rate of recession with Lake Erie 

 waters alone -42 of a foot (-128 m.) a year, the volume 

 then being only 15 per cent of the present volume. But, 

 with a volume so much smaller, the rate may be greatly 

 modified by local conditions, especially by those which 

 affect the depth of the water on the crest of the fall. This 

 is clearly shown by the history of the American fall, which 

 has less than five per cent of the whole volume of the river 

 or less than one-third of the discharge of Lake Erie. 



The American fall is a remarkable instance of lack 

 of concentration of water depth on the crest line. The 

 crest is 1,000 feet (304 m.) long and the water on it 

 is not over three feet (-9 m.) deep, while much of it is 

 one foot or less. If the crest was, say, 100 feet (30-5 m.) 

 long the water would be 18 or 20 feet (5-5 or 6 m.) deep 



