this can nowhere be found to be the case. The upper course of this ancient channel, therefore, must have coincided with that of the present channel. When, then, the Falls had receded to the side of the present Whirlpool, it reached a point where the greater part of its work had been performed. From here to whatever distance the upper course of the ancient river extended, the only work to do was to remove the loose gravel and boulders with which the glacier had filled its channel. This, of course, was effected much more rapidly than the wearing away of the hard limestone bed. Just what was the depth, and how far this old deserted valley extended, it is almost impossible to estimate. These changes are some of the most potent with which one must reckon in any calculation of the time since the beginning of Niagara's history. However, some work has been done in this line; and a broad field is still open for future investigation. [Illustration: Ice Mountain on Prospect Point.] At a very early date (1790), and when it was supposed by many to be almost sacrilegious to discuss the antiquity of the earth, Andrew Ellicott made an estimate of the age of the Falls by dividing the length of the Gorge by the supposed rate of recession. This gave as a result 55,000 years as the age of Niagara River. The next estimates which commanded attention were those of Bakewell and Sir Charles Lyell. Each of these men made separate estimates, but were compelled to take as the basis of their calculation the recession as given by residents of the district. Bakewell's calculations preceded Lyell's by several years, and resulted in ascribing to the Falls an age of 12,000 years. Lyell found the age to be about 36,000 years. The popularity of the latter caused his estimate to be accepted for a long period; many persons undoubtedly placing more faith in his results than he himself did. This method of dividing the distance by the rate of recession would be correct if there were no variables entering into the problem, and if the rate of recession were known; but these first calculations involved errors in the rate of movement of the Falls besides making no allowance for the variations which have been mentioned above.