Researches relating to the Great Lakes. — Spencer. 121 
before the upper lakes were turned into the Niagara drainage, 
for a long time seemed undeterminable, until the features of 
Foster's fiats were used for measuring the amount of work 
performed in that earl}' episode. This standard has since 
been confirmed by other phenomena not yet published; and 
from a different standpoint the distance of the early recession 
has been agreed to by Prof. Gilbert, who now considers the 
age of the falls far greater than that formerly suggested by 
his paper in [886. From all the available data up to 1894. 
the writer computed the age of Niagara falls at 32,000 years.* 
Of the various episodes, that of the cataract passing the nar- 
rows of the whirlpool rapids still seems the most difficult of 
explanation; but the writer has recently found that the nar- 
rows record a second reduction in the amount of fall in the 
river, before the present descent was established, thus retard- 
ing the recession along this section of the gorge, and increas- 
ing in part the time compensation for the reduced amount of 
work performed. However, further discoveries are neces- 
sary to fully explain the phenomenon of the narrows. It now 
seems probable that the error in determining the time re- 
quired for the recession of the falls through the section of the 
whirlpool rapids would not affect the computation of the 
whole age of the river by more than a few per cent. 
No less important than the determination of the age of the 
river was that of the date when the waters of the Algonquin 
basin (Huron, Michigan and Superior) were first turned into 
the Niagara drainage, owing to the warping of the land, 
with the greatest rise occurring along an axis ti"ending N. 28° 
E.f The date of the diversion of the waters of the upper 
lakes from the Ottawa to the Niagara valley has been com- 
puted by the writer at 7,200 years. This result was obtained 
from the mean of three distinct methods of computation, 
varying from 6,500 to 7,800 years .Ji Mr. F. B. Taylor's more 
recent estimate gives the range of from 5,000 to 10,000 years. 
Niagara as a time piece would be incomplete without indi- 
* Duration of Niagara Falls. Am. Jour. Sci., vol. XLVIII, 1894, 
pp. 455-47-2- 
tThis direction occurs east of Georgian bay, while at the end of lake 
Ontario the direction of rise is N. 25° E. See papers by the writer 
cited before. 
JSee Duration of Niagara Falls, cited before. 
