LENGTH OF GEOLOGICAL TIME. 591 



his study of the falls, that it was "more nearly three feet a century 

 than three feet a year." Taking the rate at one foot a year, the six 

 miles will have required over 31,000 years; if at one inch a year, — 

 which is eight and one third feet a century, — 380,000 years. 



The rate at which coral reefs increase in height affords another 

 mode of measuring the past. From calculations elsewhere stated by 

 the author, it appears that the rate of increase of a coral reef prob- 

 ably is not over a sixteenth of an inch a year. Now, some reefs are 

 at least 2,000 feet thick, which, at one sixteenth of an inch a year, 

 corresponds to 384,000 years, or very nearly a thousand years for five 

 feet of upward increase. If the progressing subsidence essential to 

 the increasing thickness were slower than the most rapid rate at which 

 the upward progress might take place, the time would be proportion- 

 ally longer. The reefs may have been begun in the Tertiary. 



The use of these numbers is simply to prove the proposition that 

 Time is long, — very long, — even when the earth was hastening on 

 toward its last age. And what, then, of the series of ages that lie 

 back of this in time ? Thousands of millions of years have been 

 claimed by some geologists, for time since life began. Sir Wm. 

 Thomson has reduced the estimate, on physical grounds, to one hun- 

 <ired millions of years as a maximum. If the time since the com- 

 mencement of the Silurian were but forty-eight millions, the ratio 

 12 : 3 : 1, above deduced for Paleozoic, Mesozoic, and Cenozoic times, 

 would give for each, severally, thirty-six millions, nine millions, and 

 three millions, of years. 



In calculations of elapsed time, from the thickness of formations, there is always great 

 uncertainty, arising from the dependence of this thickness on a progressing subsidence. 

 In the case of coral limestone, the data employed give the least possible time, as is ob- 

 vious from the above. In estimates made from alluvial deposits, when the data are 

 "based on the thickness of the accumulations in a given number of years, — say the last 

 2,000 years, — this source of doubt affects the whole calculation, from its foundation, 

 and renders it almost, if not quite, worthless. An estimate of the length of the Mio- 

 cene epoch, made from data derived from observations on the deposits then forming in 

 England, would have given no idea of the length of time required for the Miocene 

 Molasse of Switzerland ; and, in the same manner, any such data from observations at 

 the present day must be equally fallacious. When the estimate, as from delta-deposits, 

 is based on the amount of detritus discharged by a stream, it is of more value. But 

 even here there is a source of great doubt, in our ignorance of the oscillations the con- 

 tinent may have undergone in past time, which, especially if an upward movement, 

 would have affected the amount of discharge, and, if attended with glaciers, would have 

 produced immensely larger depositions in a given time. This source of doubt affects 

 also the calculations from the excavation of valleys. 



2. GEOGRAPHICAL PROGRESS. 



The system of oscillations and progress in North America during 

 the ages, to the close of the Tertiary period, and the new system 



